COMMISSION STAFF WORKING DOCUMENT IMPACT ASSESSMENT Accompanying the document Proposal for a Regulation of the European Parliament and of the Council on general product safety, amending Regulation (EU) No 1025/2012 of the European Parliament and of the Council, and repealing Council Directive 87/357/EEC and Directive 2001/95/EC of the European Parliament and of the Council

EN EN
EUROPEAN
COMMISSION
Brussels, 30.6.2021
SWD(2021) 168 final
COMMISSION STAFF WORKING DOCUMENT
IMPACT ASSESSMENT
Accompanying the document
Proposal for a Regulation of the European Parliament and of the Council
on general product safety, amending Regulation (EU) No 1025/2012 of the European
Parliament and of the Council, and repealing Council Directive 87/357/EEC and
Directive 2001/95/EC of the European Parliament and of the Council
{COM(2021) 346 final} - {SEC(2021) 280 final} - {SWD(2021) 169 final}
Europaudvalget 2021
KOM (2021) 0346 - SWD-dokument
Offentligt
1
Table of contents
1. INTRODUCTION: POLITICAL AND LEGAL CONTEXT.................................................4
2. PROBLEM DEFINITION.....................................................................................................10
2.1. What are the problems? ..............................................................................................10
2.1.1. Product safety challenges linked to new technologies .......................................... 12
2.1.2. Product safety challenges in the online sales channels.......................................... 15
2.1.3. Ineffective product recalls..................................................................................... 17
2.1.4. Market surveillance rules are complex and not fully effective.............................. 19
2.1.5. Inconsistent application of product safety rules for food-imitating products........ 21
2.1.6. Problems related to the legal form......................................................................... 21
2.1.7. Regulatory burden and costs of the GPSD (REFIT problem)............................... 22
2.2. What are the problem drivers?....................................................................................22
2.3. How will the problem evolve?....................................................................................24
3. WHY SHOULD THE EU ACT?...........................................................................................25
3.1. Legal basis ..................................................................................................................25
3.2. Subsidiarity: Necessity of EU action..........................................................................25
3.3. Subsidiarity: Added value of EU action .....................................................................26
4. OBJECTIVES: WHAT IS TO BE ACHIEVED? .................................................................27
4.1. General objectives ......................................................................................................27
4.2. Specific objectives......................................................................................................27
5. WHAT ARE THE AVAILABLE POLICY OPTIONS?.......................................................29
5.1. What is the baseline from which options are assessed?..............................................29
5.2. Description of the policy options................................................................................32
5.3. Options discarded at an early stage.............................................................................42
6. WHAT ARE THE IMPACTS OF THE POLICY OPTIONS? .............................................43
6.1. Impacts of Option 1 ....................................................................................................44
6.2. Impacts of Option 2 ....................................................................................................48
6.3. Impacts of Option 3 ....................................................................................................55
6.4. Impacts of Option 4 ....................................................................................................61
7. HOW DO THE OPTIONS COMPARE? ..............................................................................64
8. PREFERRED OPTION.........................................................................................................76
8.1. Preferred option – Option 3 ........................................................................................76
8.2. REFIT (simplification and improved efficiency) .......................................................78
9. HOW WILL ACTUAL IMPACTS BE MONITORED AND EVALUATED?....................79
ANNEX 1: PROCEDURAL INFORMATION .............................................................................83
ANNEX 2: STAKEHOLDER CONSULTATION........................................................................87
ANNEX 3: WHO IS AFFECTED AND HOW?............................................................................96
2
ANNEX 4: ANALYTICAL METHODS.....................................................................................101
ANNEX 5: EVALUATION REPORT...........................................................................................30
ANNEX 6: EFFECTS OF THE COVID-19 CRISIS IN THE CONTEXT OF THE
POLICY OPTIONS AND THEIR EXPECTED IMPACTS...............................................102
ANNEX 7: SMES TEST..............................................................................................................104
ANNEX 8: EFFECTIVENESS OF RECALLS OF CONSUMER PRODUCTS........................108
ANNEX 9: COOPERATION ACTIVITIES UNDER GPSD......................................................125
ANNEX 10: STANDARDISATION PROCEDURE ..................................................................128
ANNEX 11: SUMMARY OF REPLIES TO THE OPEN PUBLIC CONSULTATION............129
ANNEX 12: STAKEHOLDERS OPINIONS ON THE BENEFITS OF THE
DIFFERENT OPTIONS......................................................................................................143
ANNEX 13: MINUTES FROM THE EU WORKSHOPS ADDRESSING THE SALE
OF ILLEGAL GOODS ONLINE........................................................................................146
ANNEX 14: MINUTES FROM THE EU WORKSHOP ON STRATEGIES TO
MAXIMISE THE EFFECTIVENESS OF PRODUCT RECALLS....................................148
Glossary
Term or acronym Meaning or definition
AI Artificial Intelligence
Charter Charter of Fundamental Rights of the European Union
CSN Consumer Safety Network
DSA Digital Services Act
EEA European Economic Area
EU European Union
FIPD Council Directive 87/357/EEC concerning the safety of food-
imitating products
GPSD Directive 2001/95/EC on the general safety of products (General
Product Safety Directive)
ICT Information and Communication Technology
IA Impact Assessment
IoT Internet of things
MSA Market surveillance authority
OPC Open Public Consultation
Pledge Product Safety Pledge
REFIT Regulatory fitness and performance check
Safety Gate/RAPEX EU Rapid Alert System on dangerous consumer products
SME Small and Medium Enterprise
Subgroup Sub-group on AI, connected devices and other challenges for new
technologies to the Consumer Safety Network
GPSD Study The study commissioned by the Commission to support the
evaluation and impact assessment of the GPSD revision and
conducted by CIVIC consulting
TFEU Treaty on the Functioning of the European Union
4
1. INTRODUCTION: POLITICAL AND LEGAL CONTEXT
This report aims at assessing the impacts of a revision of the Directive 2001/95/EC on
general product safety1
(GPSD). It analyses also the impacts of a possible integration of
the Directive 87/357/EEC2
concerning the safety of food-imitating products (FIPD) into
the GPSD.
The objective of the GPSD is to ensure EU consumers are protected from dangerous
products and to ensure the proper functioning of the Single Market. The GPSD provides
the general EU legal framework for the safety of non-food consumer products and
requires that all products placed on the market be safe. The non-food consumer products
include all products (including in the context of providing a service), which are not food
stuff and are intended for consumers or are likely to be used by consumers, and are
supplied or made available to them in the course of a commercial activity, be they new,
used or reconditioned3
.
As illustrated in the Table 1, the non-food product safety framework is mainly made up
of two sets of legislative instruments:
 Union Harmonisation (hereinafter harmonised) legislation: Regulation (EU)
2019/1020 on market surveillance and compliance of products together with the
product-specific safety legislation, such as the toys or the machinery directive form
‘harmonised legislation’ (legislation setting common rules across the EU for specific
sectors)
 The GPSD: As lex generalis, it applies to non-food consumer products to the extent
that there are no specific provisions with the same objective in rules of Union law
governing the safety of the products concerned, such as EU harmonised legislation for
specific categories of products. Therefore it fully applies to non-harmonised consumer
products and also partially to the consumer harmonised products for aspects not
covered by the harmonised legislation. As such, it provides a “safety net” for
consumers and aims to ensure that EU consumers are protected against any safety
risks of consumer products, including future ones.
1
Directive 2001/95/EC of the European Parliament and of the Council of 3 December 2001 on general product safety
2
Council Directive 87/357/EEC of 25 June 1987 on the approximation of the laws of the Member States concerning
products which, appearing to be other than they are, endanger the health or safety of consumers
3
The definition of product in the GPSD excludes second-hand products supplied as antiques or as products to be
repaired or reconditioned prior to being used.
5
Table 1: General overview of the Product Safety Framework
The GPSD, as a safety net, is complementary to harmonised legislation in two ways.
First, it applies in its entirety to consumer products falling outside the scope of
harmonised legislation (e.g. furniture, childcare articles, clothes). Secondly, it applies
partially to consumer products covered by harmonised legislation (e.g. toys or cars) as
long as aspects of product safety covered by the GPSD are not covered in the harmonised
legislation (for example, until very recently EU legislation for cars did not include
provisions on product recalls, that were subject to the GPSD).
The concept of safety under the GPSD always covers levels of protection for the safety
and health of persons, i.e. the dangerous product under the GPSD poses a risk to the
health and safety of consumers. These risks can materialise in different ways but to be
covered under the GPSD the risks always have to relate to health and safety of
consumers. Other types of safety such as material damages are not covered unless they
are linked to the safety and health of consumers.
Regarding market surveillance, there are also two different systems in place: one for
harmonised products (Regulation (EU) 2019/1020) and another under the GPSD for non-
harmonised products and risks falling under the scope of the GPSD.
The GPSD does not cover pharmaceuticals, medical devices and food products. The
safety of food products is regulated separately under the General Food Law Regulation
(EC) No 178/2002. The food products have their own regime, including an alert system
(RASFF). However, the Regulation (EC) No 1935/2004 on food contact materials can
interact with the GPSD when it comes to products containing such materials (e.g.
reusable lunch boxes). Unsafe products containing food-contact materials products might
be subject to safety alerts in both alert systems, RASFF for food and Safety Gate/RAPEX
for non-food products.
The GPSD provides for the obligations of economic operators in the product safety field,
in particular the general obligation for producers to place only safe products on the
market. It contains rules on the power and obligations of Member States and on market
Product Safety Framework
Products
Areas
Non-food Food
Non- Harmonised Harmonised
General Food Law
Regulation Regulation
(EC) No 178/2002
and
Regulation (EC)
1935/2004 on food
contact materials
Non-consumer Consumer Non-consumer
Obligations of economic
operators
National Law
under the mutual
recognition
Regulation
GPSD
Sector specific Union
harmonisation
legislation
+ GPSD as safety net
Sector specific Union
harmonisation
legislation
Market surveillance on the
internal market
Regulation (EU)
2019/1020
+Ability for market
surveillance authorities
to take “more specific
measures” provided
for in GPSD
Regulation (EU)
2019/1020
Safety Gate (RAPEX)
Regulation (EU)
2019/1020 and GPSD
Regulation (EU)
2019/1020 and GPSD
Customs control for
products imported to the
EU
Regulation (EU) 2019/1020
Regulation (EU)
2017/625
6
surveillance. The GPSD establishes provisions on the application of the general safety
requirement and on the adoption of European safety standards supporting the legislation,
which provide presumption of safety and therefore facilitate the compliance with the
safety requirement under the GPSD. The GPSD includes obligations for Member States
and the Commission to inform consumers about dangerous products and contains the
legal basis for the EU Rapid Alert System (Safety Gate/RAPEX), which enables quick
exchange of information between EU/EEA countries and the Commission on measures
taken on unsafe non-food products posing a risk to consumers4
. Finally, the GPSD also
establishes cooperation on product safety between the Commission and Member States
authorities competent for product safety in the context of the Consumer Product Safety
Network ‘CSN’.
Directive 87/357/EEC5
(Food-Imitating Products Directive, ‘FIPD’) sets out rules for
the safety of food-imitating products, i.e. products which can be confused with
foodstuffs, while not being food-stuffs. The FIPD has been originally adopted to address
divergence in national provisions on products which, appearing to be other than they are,
endanger the safety or health of consumers6
. The measures taken against unsafe food-
imitating products by Member States are also notified in the Safety Gate/RAPEX.
The scope of this initiative covers therefore all non-harmonised consumer products, food-
imitating products and also partially also harmonised products for aspects not covered by
the harmonised legislation. However for some aspects the analysis also includes
harmonised products, where data could not be dissociated between the harmonised and
non-harmonised products.
The Commission has adopted guidance to clarify some of the aspects of the GPSD. The
Commission Notice on the market surveillance of products sold online (2017/C 250/01)7
provides guidance for the enforcement of EU legislation on the safety and compliance of
non-food products sold online. The Notice also sets out good practices for the market
surveillance of products sold online and for communication with businesses and
consumers. On 9 November 2018, the Commission also revised the guidelines for the
functioning of Safety Gate/RAPEX8
(originally adopted in 2004 and revised for the first
time in 2010).
ECJ Jurisprudence does not bring particular element helping interpreting the provisions
of the GPSD, most of it being linked either to access to documents or to non-applicability
in the presence of harmonised legislation.
At international level, very different approaches to product safety can be distinguished:
Few jurisdictions, such as Canada and more recently Brazil (in 2019), have adopted a
safety regulatory framework which includes a general safety requirement, similar to the
EU “safety net”. This approach is usually considered in international fora as the best way
4
Regulation (EC) 765/2008 and the new Regulation (EU) 2019/1020 on market surveillance and compliance extends
the scope of Safety Gate/RAPEX also to products to be used by professionals
5
Council Directive 87/357/EEC of 25 June 1987 on the approximation of the laws of the Member States concerning
products which, appearing to be other than they are, endanger the health or safety of consumers
6
These pre-existing national provisions were differing in content, scope and field of application and therefore creating
barriers to the free movement of goods and unequal competitive conditions on the market without ensuring effective
protection for consumers, especially children.
7
Commission Notice on the market surveillance of products sold online (2017/C 250/01) of 28 July 2017
8
Decision 9 November 2018 laying down guidelines for the management of the Community Rapid Information System
(RAPEX). According to Annex II, point 8 of GPSD, the Commission regularly update such guidelines. The new
version of the guidelines updates the scope and purpose of safety Gate/RAPEX, integrates certain aspects of Regulation
(EC) 765/2008 on market surveillance of harmonised products (inclusion of professional products and extension of the
risks to risks other than those for the health and safety of consumers (e.g. environmental risks), includes a reference to
new tools developed over the last years for the proper functioning of Safety Gate/RAPEX, clarifies notification criteria
and enhances traceability, which is essential for follow up by countries in the Safety Gate/RAPEX network.
7
to ensure that regulators have an appropriate legal basis to order corrective measures
against all types of dangerous products, notably when facing emerging safety issues that
are not subject to any regulation yet9
.
In some jurisdictions such as the United States (US) or New Zealand, liability rules play a
major role to complement product safety provisions and private enforcement is a key
aspect of such systems. The US system also relies on a sophisticated injury data
collection scheme and deterrent penalties incentivising businesses to inform as soon as
possible the consumer product safety agency about dangerous products they are
responsible for. In China, governmental approvals (be it certification, license, registration
or individual approval) are required for many products and groups of products before
they can be placed on the market.
Fora for multilateral product safety cooperation, such as the OECD and UNCTAD,
provide opportunities to learn about and get inspiration from other jurisdictions’ best
practices and new regulatory developments.
Political context
The Commission announced the revision of the GPSD in its Work Programme 2020 and
confirmed it in the Adjusted Commission Work Programme 2020 published on 27 May
2020, as one of its REFIT (Regulatory Fitness and Performance Programme)10
initiatives
under the headline objective “A New Push for European Democracy”. This revision is
also one of the legislative proposals mentioned in the Communication of the Commission
on New Consumer Agenda11
, published on 13 November 2020.
In the field of new technologies (including artificial intelligence ‘AI’), the Commission
published a Report on safety and liability implications of AI, the Internet of Things and
Robotics12
accompanying the White Paper on AI13
in February 2020. The report
highlights the need to include clear provisions in the EU product safety legislation,
including the GPSD, to explicitly address safety risks linked to products incorporating
new technologies (connected products and AI).
The other EU institutions have also highlighted the importance of product safety policy.
The European Parliament adopted a resolution on addressing product safety in the
Single Market on 25 November 202014
which also emphasised the need to revise the
GPSD.
In its Presidency conclusions on The Charter of Fundamental Rights in the context of
Artificial Intelligence and Digital Change15
, the Council stated that while these
technologies may enhance the market surveillance of product safety on the EU market,
they may also pose new challenges to consumer protection in the product safety area. The
Council Conclusions on Shaping Europe's Digital Future16
of 9 June 2020 mention that
9
See for instance the OECD Recommendation on Consumer Product Safety from 17 July 2020,
https://legalinstruments.oecd.org/en/instruments/OECD-LEGAL-0459#mainText
10
REFIT is the European Commission's regulatory fitness and performance programme established in 2012 to ensure
that EU law is 'fit for purpose'. It is a process under which existing legislation and measures are analysed to make sure
that the benefits of EU law are reached at least cost for stakeholders, citizens and public administrations and that
regulatory costs are reduced, without affecting the policy objectives pursued by the initiative in question.
11
https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:52020DC0696
12
https://ec.europa.eu/info/publications/commission-report-safety-and-liability-implications-ai-internet-things-and-
robotics_en
13
https://ec.europa.eu/info/files/white-paper-artificial-intelligence-european-approach-excellence-and-trust_en
14
https://www.europarl.europa.eu/doceo/document/TA-9-2020-0319_EN.html
15
https://www.consilium.europa.eu/media/46496/st11481-en20.pdf
16
https://www.consilium.europa.eu/media/44389/st08711-en20.pdf?utm_source=dsms-
auto&utm_medium=email&utm_campaign=Shaping+Europe%e2%80%99s+digital+future+-
+Council+adopts+conclusions
8
digital economy is to be characterised by a high degree of trust, security, safety and
choice for consumers17
.
Finally, the review builds on the results of the Evaluation of the GPSD and the FIPD,
conducted back-to-back to this Impact Assessment (‘IA’) in order to assess the
performance of GPSD in line with the Commission’s Better Regulation rules (see Annex
5). The Evaluation showed the necessity to adapt the GPSD to address product challenges
related to e-commerce as well as the rapid development of new technologies, and to
ensure better enforcement and more efficient market surveillance for consumer products,
including by aligning the systems for harmonised and non-harmonised products.
Furthermore, the Evaluation also showed the necessity to modify some of the provisions
of the GPSD to improve its effectiveness. For example, legislative changes are needed to
improve the effectiveness of product recalls, as well as the treatment of food-imitating
products.
The Commission already proposed to revise the GPSD with the 2013 Product Safety
and Market Surveillance Package. This 2013 Package - made up of two proposed
Regulations, one on product safety (COM(2013)078), the other on market surveillance
(COM(2013)075), aimed at enhancing product safety rules and creating a single set of
market surveillance rules for harmonised and non-harmonised products. The 2013
Package was blocked in the inter-institutional process as the Council did not agree on a
common position because of a proposed provision on the mandatory marking of the
origin of industrial products (the “Made in” provision). The proposals were finally
withdrawn by the Commission in September 2020.
Legal context
The legal framework relevant for the safety of consumer products has evolved with the
adoption of several legal acts:
With the adoption of Regulation (EU) 2019/102018
on market surveillance and
compliance of products, the legal framework for market surveillance of harmonised
products has changed and has been adapted in particular to the challenges linked to the
online sales. The market surveillance of non-harmonised products, also initially included
in the 2013 Product Safety and Market Surveillance Package, remained unchanged. The
differences between the two market surveillance frameworks have been therefore
perpetuated with the adoption of Regulation (EU) 2019/1020, except for the provisions
on products entering the EU market, which apply to all products covered by Union
harmonisation law under Regulation (EU) 2019/1020. This Regulation widened the
difference between the regime applicable to harmonised and non-harmonised products, in
particular by envisaging new powers and cooperation instruments for market surveillance
authorities (‘MSAs’) and introducing some new features such as the “responsible
person”, according to which any economic operators that wants to sell its products into
the EU market has to be represented by an economic operator in the EU (applicable not to
all harmonised products but only to some of them).
17
Council stressed also that some AI applications can entail a number of risks, such as biased and opaque decisions
affecting citizens’ fundamental rights, such as the rights to safety and security. Concerning software, Council
underlined the potential of safe, secure, sustainable and trusted hard- and software value chains to enable and establish
trust in European digital technologies. Council also stressed the need to enhance citizens’ safety and to protect their
rights in the digital sphere across the Single Market and the need for effective and proportionate action against illegal
activities and content online, including the distribution of dangerous goods.
18
Regulation (EU) 2019/1020 of the European Parliament and of the Council of 20 June 2019 on market surveillance
and compliance of products and amending Directive 2004/42/EC and Regulations (EC) No 765/2008 and (EU) No
305/2011
9
In the standardisation area, the Commission adopted Regulation (EU) 1025/201219
on
European standardisation, which provides the general framework for the adoption of
European standards for products and services (to help assessing conformity with Union
legislation), identifies Information and Communication Technology (‘ICT’) technical
specifications, and allow financing for the European standardisation process. It also sets
an obligation for European Standardisation Organisations (CEN, CENELEC, ETSI) and
National Standardisation Bodies on transparency and participation.
The adoption of the new Regulation (EU) 2019/51520
facilitates the application of the
principle of mutual recognition according to which, where no harmonised rules exist at
European level, products lawfully marketed in one Member State can be sold in other
Member States regardless of complying or not with the national technical rules of these
Member States.
Links with other legal initiatives
The GPSD initiative has the following links with other recent and ongoing proposals:
Digital Services Act (DSA)21
: The DSA presented by the Commission on 15 December
2020 includes a new set of horizontal rules to regulate the responsibility of online
intermediaries, including online marketplaces22
. The DSA proposal aims to establish new
obligations for online intermediaries inter alia in relation with how they handle all types
of illegal content hosted on their websites (e.g. unsafe products, counterfeit products, hate
speech, etc). The DSA establishes the general horizontal obligations for online
intermediaries and leaves room for legislation in relation with specific types of illegal
content (such as product safety) to be more specific. For example, the DSA provides the
general framework for the notice-and-action system, but without providing the details on
the timeframe or the procedure, which could then be set up in the revised GPSD (GPSD
would provide a specific timeline and detailed procedure for such notices of unsafe
products). The GPSD may also regulate other product safety aspects of online sales
beyond the role of online intermediaries, such as the role of sellers and the powers of
market surveillance authorities.
Artificial Intelligence (AI) horizontal framework: The new legislative proposal for AI
horizontal framework lays down harmonised rules for the placing on the market, the
putting into service and the use of artificial intelligence systems (‘AI systems’) in the
Union, consistent with a high level of protection of the public interests, in particular
health and safety, and fundamental rights and freedoms of persons. It lays down specific
requirements with which high-risk AI systems must comply and imposes obligations on
providers and users of such systems (for example, regulating inter alia safety aspects of
AI applications in products such as machinery or lifts). Consequently, and with respect to
product safety, it will establish specific requirements for certain AI applications, and the
GPSD would apply as a safety net for products and safety aspects not covered by the AI
horizontal legislation, and therefore complement it. The scope of the initiative is such that
it is likely that some AI applications would remain not covered (e.g. some consumer
19
Regulation (EU) No 1025/2012 of the European Parliament and of the Council of 25 October 2012 on European
standardisation, amending Council Directives 89/686/EEC and 93/15/EEC and Directives 94/9/EC, 94/25/EC,
95/16/EC, 97/23/EC, 98/34/EC, 2004/22/EC, 2007/23/EC, 2009/23/EC and 2009/105/EC of the European Parliament
and of the Council and repealing Council Decision 87/95/EEC and Decision No 1673/2006/EC of the European
Parliament and of the Council
20
Regulation (EU) 2019/515 of the European Parliament and of the Council of 19 March 2019 on the mutual
recognition of goods lawfully marketed in another Member State and repealing Regulation (EC) No 764/2008.
21
https://eur-lex.europa.eu/legal-content/en/TXT/?qid=1608117147218&uri=COM%3A2020%3A825%3AFIN
22
The DSA provides for several due diligence obligations relevant for the product safety area, namely obligations on
notice & action, know your business customer, cooperation with authorities as well as clear terms and conditions
including respect for consumer protection rights.
10
products such as vacuum cleaners). The revised GPSD would therefore need to provide a
legal basis for withdrawing such products from the market to ensure an effective
protection of consumers.
Delegated acts under the Radio Equipment Directive (RED): the RED establishes the
possibility for the Commission of adopting delegated acts in relation with several aspects,
including protection of personal data and fraud for specific categories of radio devices.
The Commission is working on several delegated acts that might partially address the
issue of products presenting cybersecurity risks. However, it will not be possible to cover
all possible consumer products via delegated acts, for instance, devices connected by
cable. Such gaps might be covered by a revised GPSD in its role of safety net.
Directive on Security of Network and Information Systems (NIS): The recent
proposal for a NIS 2 Directive, presented by the Commission on 16 December 2020, lays
down obligations for all Member States to adopt a national strategy on the security of
network and information systems. However, it does not include minimum cybersecurity
requirements for consumer products, so it does not provide any legal basis for authorities
to take action against products presenting such risks. The Cybersecurity Act
(Regulation (EU) 2019/881) introduces an EU-wide cybersecurity certification
framework for ICT products, services and processes. However, it does not include
minimum cybersecurity legal requirements for ICT products. The GPSD is therefore
complementary to these initiatives to fill these gaps.
Circular Economy: According to the new Circular Economy Action Plan, products
placed on the EU market should be more sustainable and designed therefore to last
longer, to be easier to repair and upgrade, recycle and reuse. It is essential that repaired,
upgraded, recycled or reused products continue to meet product safety requirements.
According to the Eco-design directive (Directive 2009/125/EC), safety and health have to
be taken into account in the choice of a specific design solution; however safety issues
related to the end products are not specifically addressed. The Sustainable Product Policy
Initiative (which intends to replace the Eco-design directive and extend its scope) will
notably aim at correcting the fact that many products cannot be easily and safely reused,
repaired or recycled. In case some safety aspects related to products in the circular
economy (such as refurbished appliances or clothing made from recycled plastics) are not
specifically addressed by initiatives from the Circular Economy Action Plan and do not
fall under harmonised legislation, the safety net function of the GPSD comes into play.
2. PROBLEM DEFINITION
2.1. What are the problems?
The Evaluation and the stakeholder views show that the GPSD appears overall to have
met its objectives of ensuring a high level of safety of consumers, while ensuring an
effectively operating internal market for goods; however, still too many unsafe products
reach or remain in the hands of consumers.
On the EU Single Market there should not be obstacles and barriers to the free movement
of goods, which enables unsafe goods to circulate within the EU. Concerning the trend in
the safety of products, the evaluation showed that the notifications of dangerous products
by market surveillance authorities (‘MSAs’) in the Safety Gate/RAPEX increased, from
2005 to 2010, from around 540 to 2000 notifications/year and then fluctuated between 1
550 to 2 100 notifications/year (30% of which concerned non-harmonised products).
11
Also the number of follow-up measures23
reported in Safety Gate/RAPEX has steadily
increased since data started to be gathered in this respect by the European Commission in
2011. Some encouraging signs, such as improvements in product safety perceived by
consumers and a plurality of stakeholders24
can be observed; however, available data
show that unsafe products are still available on the EU market. The share of dangerous
products found by MSAs in inspections represents between 2% and 16% of total
consumer products inspected, with a median value of 4%25
. Unsafe products on the EU
market affects consumers as well as economic operators that play by the rules as they
suffer from lack of level playing field with “rogue operators” from inside and outside the
EU not observing EU product safety rules.
Unsafe products represent an important cost for consumers and society. The GPSD
Study supporting the GPSD Evaluation and IA (hereinafter ‘GPSD Study’)26
estimates
the consumer detriment due to unsafe products today in the following way:
- Consumer detriment linked to product-related injuries and premature death:
The total detriment to EU consumers and society from product-related injuries and
premature death to be EUR 76.6 billion per year. This is the sum of detriment caused by
non-fatal product-related injuries and the cost of premature death where a consumer
product is involved (e.g. accident with tools, strangulation, electrocution, or fire)
occurring outside of work-related locations27
. This figure includes health care utilisation
costs, productivity losses, loss of quality of life, cost of premature death linked to injuries
due to consumer products (both harmonised and non-harmonised) in the EU.
The analysis based on previous research and interviews with product safety experts
concluded that 15% is a reasonable and cautious estimate for the proportion of this total
detriment that was caused by unsafe consumer products, or could have been prevented
through better design, instruction or a safety device. These 15% of accidents could have
been prevented if the products were safe. On this basis, the preventable detriment
suffered by EU consumers and society due to product-related accidents can be estimated
at EUR 11.5 billion per year28
.
- Consumer detriment linked to the loss of value of unsafe products
In addition to the above injury related detriment, the GPSD Study estimates that the
consumers also suffered financial costs of a total value of EUR 19.3 billion in 2019
arising from the fact they have purchased unsafe products that they would not have
23
Follow-ups can be defined as the feedback received from Member States participating in the Rapid Alert System on
actions they have taken following up another country’s alerts on their own market.
24
Source: European Commission 2016 and 2018 survey of consumers’ attitudes toward cross-border trade and
consumer protection and the GPSD Implementation report
25
Source: the GPSD Study. Member States inspections can be targeted so these figures cannot represent the proportion
of all unsafe products on the EU market.
26
Study to support the preparation of an evaluation of the General Product Safety Directive as well as of an impact
assessment on its potential revision, prepared by Civic consulting, December 2020
https://ec.europa.eu/info/files/study-support-preparation-evaluation-gpsd-well-impact-assessment-its-revision-part-1-
evaluation_en
https://ec.europa.eu/info/files/study-support-preparation-evaluation-gpsd-well-impact-assessment-its-revision-part-2-
impact-assessment_en
27
These estimates are based on the best possible approximation of product-related injuries and fatalities. The detriment
cannot be estimated separately by categories of products and therefore include all consumer products, harmonised and
non-harmonised products.
28
15% of EUR 76.6 billion per year. This is the part of the total injury detriment linked to the injuries/deaths caused by
the unsafe aspect of the products. These accidents could have been prevented if the products were safe. The remaining
EUR 65.1 billion are injuries/deaths where a product is involved but the accident is not caused by the unsafe aspect of
the product (e.g. falling from a ladder is a product-related accident but it doesn’t mean that the ladder itself is unsafe).
This part cannot be prevented.
12
purchased if they knew these products were unsafe (hereinafter ‘consumer detriment
linked to the value of unsafe products’). The estimation is based on the fact that the value
of unsafe non-harmonised products per year is estimated at EUR 3.9 billion for online
sales channels, and EUR 15.4 billion for brick-and-mortar shops and other offline sales
channels, for a total of EUR 19.3 billion. This detriment is reduced due to recalls (under
current low recall effectiveness scenario) by approximately EUR 0.4 billion per year,
assuming that consumers are compensated fully for all non-harmonised products they
returned to producers in response to a product recall. This detriment relates to non-
harmonised products covered by the GPSD and is based on the assumption that the loss in
consumer welfare is at least the price to which the product was purchased.29
The presence of unsafe products on the EU market affects therefore both EU consumers
(final users of products) who bear the risk of accidents, injury or death caused by
dangerous products and related costs and Member States who bear increased health
expenditure costs resulting from health treatment of injuries caused by dangerous
products.
The circulation of unsafe products on the EU market also creates a problem for the
Single Market: it does not only contravene the principle of free circulation of goods
(only safe goods are allowed to circulate in the Single Market), but it also risks to create
distortions of competition on the EU Single Market. Economic operators compliant with
EU product safety rules face compliance costs in comparison with non-complaint
operators. At the same time, the presence of unsafe products on the EU market puts in
danger the health and safety of EU consumers, which also undermines consumer’s trust
and confidence in the EU Single Market. Trust is an essential engine of consumers’
consumption and therefore the growth of the EU economy30
. The Eurobarometer data
indicate a decrease between 2016 and 2018 in confidence of consumers in the safety of
products sold in the EU31
.
Also, in the open public consultation (‘OPC’) a large majority of respondents (71%)
expressed that current EU safety rules for non-food consumer products covered by the
GPSD could be improved in specific areas to be more adequate to protect consumers32
.
2.1.1. Product safety challenges linked to new technologies
At the time of the adoption of the Directive the number of consumer products
incorporating new technologies was scarce. This is not the case anymore. The scenario is
likely to evolve with the increasing use of AI, impacting the whole EU market. Moreover,
there were 14.2 billion connected devices in 2019 worldwide, a figure that is estimated to
go up to 25 billion by 2025, of which 4.9 billion estimated to be in Europe33
.
The application of the GPSD to new technology products, such as connected devices or
AI-powered products, is not crystal-clear and the safety of these products is not fully
covered by other EU legislation. The GPSD does not explicitly address the fact that new
29
This relates to non-harmonised consumer products covered by the GPSD. This is based on the assumption that
willingness to pay (WTP) for a product depends on the utility of the product for the purchaser. WTP is equal or higher
as the price for which a product is purchased by a consumer, as otherwise the transaction would not take place. It is
very likely that WTP would be close to zero for an unsafe product (nobody wants to buy e.g., a dangerous childcare
product) – so the loss in consumer welfare is at least the price to which the product was purchased.This calculation
assumes that the consumers do not get reimbursed for the unsafe product.
30
Consumer consumption represented 52,6% of the GDP of the EU in 2019. Source:
https://ec.europa.eu/eurostat/statistics-explained/index.php/Household_consumption_by_purpose
31
See European Commission 2016 and 2018 survey of consumers’ attitudes toward cross-border trade and consumer
protection.
32
See Annex 11 on results of the OPC.
33
Netherlands Entreprise Agency - https://www.cbi.eu/market-information/outsourcing-itobpo/intergrated-internet-
things/market-potential
13
technologies, in particular AI and goods with digital elements34
, can impact product
safety. In the OPC almost half the respondents considered the safety of products
involving new technologies not to be adequately regulated (47%)35
. The Evaluation
showed that this can be problematic, as the development of new technologies mean that
some of the provisions of the GPSD are not well adapted to respond to its objective of
ensuring that all products (including those incorporating new technologies) must be safe
for consumers.
As such, new technologies pose challenges to the concepts and definitions used under
the GPSD. New technology-based products also bring new risks to consumers’ health
and safety or change the way the existing risks could materialise. These new risks, such
as cybersecurity threats, might be possibly present in consumer products and this remains
not explicitly covered in EU legislation for the moment. Besides, the applicability of
software updates for product safety is still not regulated under EU legislation either (lack
of clarify of the responsibilities of economic operators when for example an application is
downloaded into a product modifying its safety features).
A typical example is when a product becomes dangerous by not possessing a minimum
level of cybersecurity, leaving it open to hacking by a malicious party; that was the case
of a passenger car notified in the Safety Gate/RAPEX36
. The radio in the vehicle might
have had certain software security gaps allowing unauthorised third party access to the
interconnected control systems in the vehicle. If these software security gaps were
exploited by a third party for malicious purposes, a road accident could have occurred.
The Sub-group on AI, connected devices and other challenges for new technologies to the
Consumer Safety Network (‘the Subgroup’) highlighted that the lack of explicit mention
in the GPSD of cybersecurity risks affecting safety (‘cybersafety’) posed a challenge for
the protection of consumers and legal certainty for businesses.
Another example relates to consumer’s personal security that can be endangered by third
party accessing their information, as illustrated in another notification in Safety Gate
RAPEX of a smartwatch for children37
. The Icelandic authority argued that this product
would not cause a direct harm to the child wearing it, but lacking a minimum level of
security, it can be easily used as a tool to have access to the location of the child. As one
of the product’s intended function is to keep children safe through localisation, a
consumer would expect that it would not pose security threats to children that may affect
their safety by them potentially being tracked and/or contacted by anyone. As measures
regarding this product were notified to Safety Gate/RAPEX, authorities in the Member
States took follow-up actions. However, the Evaluation showed that many of those
authorities were unsure whether the GPSD38
was applicable to such risks due to the lack
of explicit provisions in this respect. The Sub-group also raised the fact that it is unclear
under which legal or policy instrument such personal security risks should be tackled so
that consumers are effectively protected against such threats.
In addition, the Subgroup stated that there is evidence that new technologies can have an
impact on the mental health of consumers; e.g. connected products as a cause of
34
As defined in Directive (EU) 2019/770 on certain aspects concerning contracts for the supply of digital content and
digital services ‘goods with digital elements’ means any tangible movable items that incorporate, or are inter-connected
with, digital content or a digital service in such a way that the absence of that digital content or digital service would
prevent the goods from performing their functions.
35
See Annex 11 on the results of the OPC
36
RAPEX notification from Germany published in the EU Safety Gate (A12/1671/15) of a passenger car.
37
Example RAPEX notification from Iceland published in the EU Safety Gate’s website (A12/0157/19) of a
smartwatch for children.
38
The legal basis of this notification was the GPSD as at the moment there was no delegated acts under the RED which
could cover this case.
14
depression, loss of sleep, altered brain function and myopia or early blindness in
students39
and children40
. It was also noted however that some mental health challenges
linked to products do not originate from new technologies, they were prevalent before
digitalisation. The subgroup also expressed that it was unclear if mental health risks are
covered under the current definition of safety of the Directive, and that mental health
harm intrinsically caused by a product itself should be covered.
Moreover, it is not clearly stated to what extent the definition of “product” includes
software, whether it is sold with the product or associated with the product later on. This
might impact the safety assessment of the given product. New technologies also pose
challenges related to the notion of placing a product on the market41
. For example,
products including new technologies can evolve and their safety features may change via
software updates or machine learning after they have been placed on the market.
Many of the problems linked to new technologies are crosscutting, so the Commission
has adopted or is working on a number of proposals in relation to those issues. In some
cases specific risks linked to new technologies can be tackled by EU harmonised
legislation. While such proposals may partially address the gaps identified, there are
some aspects that remain or will remain not covered and for which action is still needed
in the context of this initiative.
The Commission is currently developing a delegated act under the Radio Equipment
Directive and assessing whether the provisions of that Directive referring to the
combination of radio equipment and software should apply to certain categories of
products covered by that Directive, as well as to standalone software uploaded onto
connected products that communicate via certain radio modules. The Commission is also
reviewing the Machinery Directive to address those types of risk having an impact on
safety, for example protecting the machinery against malicious third parties or lack of
connectivity. However, despite this, there are still some gaps in addressing safety risks of
consumer products containing new technologies not already covered by other EU
legislation. Home appliances connected to the Internet by cable e.g. will not be covered
under the delegated acts of the Radio Equipment Directive, so cybersecurity risks of such
products will not be covered by such delegated acts. In addition, in view of the highly
innovative potential of the new technology sector, it is difficult to foresee the safety
features and risks of these new technology products.
Finally, the Subgroup also mentioned that one of the common characteristics of AI and
Internet of things (‘IoT’) products is the presence of software that can change/evolve over
time. This challenges the traditional meaning of the concept of placing on the market of
the GPSD. Therefore, the Subgroup recommended that a possible GPSD revision should
clarify that products should be safe over their whole expected lifespan, and should
explore the introduction of the concept of ‘substantial modification’ affecting the safety
of the product after a product was once placed on the market.
This problem affects all consumers purchasing new technology products and causes
particular difficulties to vulnerable consumers that are not familiar with new
technologies, in particular small children and the elderly.
The lack of legal certainty regarding the application of consumer product safety rules to
new technologies may create regulatory costs to businesses (especially SMEs) developing
39
K. Demirci, M. Akgönül, A. Akpinar, 2015. Relationship of smartphone use severity with sleep quality, depression,
and anxiety in university students. Journal of Behavioural Addictions, 4(2): 85–92.
40
Dresp-Langley B. Children's Health in the Digital Age. Int J Environ Res Public Health. 2020 May 6;17(9):3240.
doi: 10.3390/ijerph17093240. PMID: 32384728; PMCID: PMC7246471.
41
The GPSD requires producers to place only safe products on the market (cf Article 3).
15
and producing new technology products and undermines their efforts to design innovative
and cyber-safe products.
2.1.2. Product safety challenges in the online sales channels
While the GPSD applies to consumer products regardless if they are sold offline or
online, the increasing use of e-commerce has negatively influenced the relevance and
effectiveness of the GPSD, creating new challenges for the safety of consumers. Online
sales increased steadily since the GPSD’s adoption: in 2002 only 9% of Europeans
purchased online, while over 70% of them shop online today42
. Furthermore, one out of
five companies in the EU nowadays sells online43
.This trend has been amplified by the
COVID 19 crisis and related lockdowns: in the EU-27, retail sales via mail order houses
or the Internet in April 2020 increased by 30% compared to April 2019, while total retail
sales decreased by 17.9%44
(see Evaluation Annex 5). In addition, many of dangerous
COVID-19 related products (e.g. dangerous masks, hand sanitisers) have been found
online (by 22 October 2020 they represented 16% of all COVID-19 notifications in
Safety Gate/RAPEX from the beginning of the COVID-19 crisis). Furthermore, 39% of
respondents in the OPC expressed that safety rules for products covered by the GPSD
were not adapted to online trade and among respondents who experienced a product
safety incident within the last 5 years, 70% bought this product online45
.
First, the GPSD does not provide for sufficiently effective instruments for online
market surveillance by MSAs. They lack e.g. powers to acquire product samples under
covert identity or block websites proposing dangerous products46
. This creates
inefficiencies in the market surveillance of non-harmonised products sold online, and
therefore insufficient action against such products. This affects the consumer trust in
online sales. While such instruments exist for the harmonised products covered by
Regulation (EU) 2019/1020 on market surveillance, the fact that the latter is not
applicable to non-harmonised products will create an uneven level-playing field between
these two categories of consumer products in the Single Market once this Regulation will
apply47
. This means that an authority could be entitled to take more effective actions
online against a toy bed (a toy being a harmonised product) as opposed to a baby’s crib,
which falls under the GPSD.
Second, new online business models and actors, such as online marketplaces hosting
third party sellers, have become prominent and product safety rules for these economic
operators are unclear under the current GPSD. Among European businesses selling
goods online, 40% have been using online marketplaces to reach their customers in
201948
.The GPSD does not establish clear legal obligations for product safety for
business models that do not fall under the existing categories of producer, importer or
distributor. The online marketplace does not fit to these categories. This affects both
consumer protection and safety of products sold online and the related consumer’s trust,
creates inefficiencies in online market surveillance and creates an uneven level-playing
field between the economic operators selling offline and those selling online in the EU.
42
Nestor Duch-Brown, 2015
43
Eurostat (isoc_ec_eseln2), data for 2019.
44
OECD - E-commerce in the time of COVID-19, http://www.oecd.org/coronavirus/policy-responses/e-commerce-in-
the-time-of-covid-19-3a2b78e8/#biblio-d1e705
45
See Annex 11 on results of the OPC.
46
On 1st August 2017 the Commission issued a Notice on the market surveillance of products sold online to help
public authorities with their market surveillance of online sales but this Notice doesn’t create legal tools as such.
47
The market surveillance provisions under Regulation (EU) 2019/102O will enter into force in July 2021.
48
ESTAT https://appsso.eurostat.ec.europa.eu/nui/submitViewTableAction.do See also (Eurobarometer - TNS, 2016)
for more granular data based on a 2016 survey
16
This has been addressed partially through voluntary action: in 2018 several online
marketplaces signed voluntary commitments to improve the safety of products sold
online: the Product Safety Pledge (hereinafter ’the Pledge’). The current eleven
signatories49
committed, among others, to react within two days when a government
informs them about an unsafe product offered on the platform, to cooperate with national
authorities and to fight against repeat offenders. As the Evaluation has highlighted, while
these voluntary commitments reflect some progress related to the cooperation between
the signatories and authorities, it is challenging to analyse the effectiveness of the Pledge
due to a suboptimal reporting system from the signatories. The Pledge has positive
impacts, as it has set the grounds for an increased cooperation framework between online
marketplaces and market surveillance authorities. However, authorities and stakeholders
have signalled in the GPSD Study that as long as the Pledge remains voluntary, the
infringement of those commitments cannot be penalised by authorities. It is also
challenging to analyse how effective the Pledge is in appropriately ensuring the safety of
products sold online, since the Key Performance Indicators (KPIs) are calculated only on
certain commitments. From this aspect, the Pledge did not help to get information on
specific issues such as on emerging risks of new technologies or improved recalls. The
monitoring reports also showed that there has been a divergence in the way online
marketplaces calculated the KPIs, making it difficult to extract conclusions from those
numbers and properly monitor the effectiveness of the commitments of the Pledge.
Finally, there are also many players on the market that have not decided to adhere to the
voluntary commitments, creating an uneven level-playing field between online
marketplaces targeting EU consumers. Therefore, while the Pledge sets out a very useful
mean of cooperation between online marketplaces and national authorities, its
effectiveness is limited by the limited range of signatories and by its voluntary nature,
limiting enforcement.
Finally, via online sales, EU consumers also purchase more frequently products
offered directly by operators established outside the EU: the proportion of purchases
from sellers outside the EU increased from 17% in 2014 to 27% in 201950
. Around 150
million small consignments are imported free of VAT into the EU each year51
. In 2017
there were 150.000 private consignments coming from China to individual EU consumers
per day52
. This is problematic: first, direct imports make it more complicated to control
the safety of the product before it enters the EU market since it is directly delivered in
individual packages to the consumer without possibly being handled by any economic
operator in the EU subject to products safety obligations under the GPSD53
. Second,
national authorities have difficulties to engage with the trader in case of safety concerns,
if the trader is not represented in the EU but is based in a third country. Article 4 of
Regulation (EU) 2019/1020 on market surveillance creates an obligation, in case a
product is sold in the EU, to have an economic operator in the EU for certain tasks linked
to market surveillance of products’ safety and compliance, but its applicability is limited
to certain categories of harmonised products. If the product safety is not evenly enforced
between EU and non-EU operators, this creates an uneven level-playing field between
these operators. A study provided by Eurocommerce indicates that the cost difference
between products produced in accordance with EU rules and standards, and produced
49
AliExpress, Allegro, Amazon, Bol.com, C-discount, Ebay, eMAG, Etsy, Joom, Rakuten France, Wish.com
50
Eurostat https://ec.europa.eu/eurostat/statistics-explained/index.php?title=E_commerce_statistics_for_individuals#E-
shopping_from_other_EU_countries
51
European Commission , Memo 2017 - Modernising VAT for e-commerce
https://ec.europa.eu/commission/presscorner/detail/en/MEMO_16_3746
52
Eurocommerce – Creating a level-playing field for retail in Europe – August 2019
53
Pure postal and delivery services are exempted from product safety obligations.
17
without taking account of the EU rules may be important for some products54
. While
these conclusions cannot be extrapolated to the overall market or to all products, they
give an indication of the possible detriment due to the presence of rogue traders from
third countries.
The Evaluation has also compiled evidence pointing to the fact that the control of the
safety of products sold online is more problematic than the one for unsafe products
found in brick-and-mortar shops. For example, data coming from the Safety
Gate/RAPEX for the period 2018-2019 show that the share of notifications of unsafe
products in which one of the four traceability information items55
was missing was
between 29,2% and 57,3% (depending on the item) for products 'sold online' and
considerably lower (between 12,6% and 35,7%) for products sold offline. None of the
traceability information was found in 12,8% notifications of products sold online, while it
was only 0,5% for the unsafe products sold offline.
These problems affect also economic operators. EU producers face an uneven playing
field between them and non-EU producers if those do not comply with EU safety rules
and therefore do not bear the compliance costs of the EU product safety legislation.
Online marketplaces targeting EU consumers also do not have a level-playing field since
the signatories of the Pledge bear additional costs and administrative burden compared to
non-signatories that do not take the steps outlined in the Pledge.
In the area of online sales, the product safety obligations of online market places are not
spelled out in any EU legislation and current market surveillance provisions relating to
products imported from outside the EU are tackled only for certain harmonised products
under Regulation (EU) 2019/1020.
Besides, there are also traceability problems with products offered online (the traceability
of the online chain is often deficient and there is a gap between the product information
available to the consumer for a product sold online and offline). The DSA aims to partly
tackle these issues by introducing the “Know-Your-Business-Customer” principle
(KYBC) and traceability provisions when it comes to the online sales via certain online
marketplaces, leaving scope to the revised GPSD to tackle traceability issues for all
online sales.
2.1.3. Ineffective product recalls
Article 5 of the GPSD requires that when a product already sold to consumers turns out to
be dangerous, it needs to be recalled (as a measure of last resort) to protect EU
consumers. But the GPSD does not set any specific rules regarding the modalities of
recalling unsafe products and evidence suggests that the proportion of products
successfully recovered from consumers remains generally low, as recognised by a recent
OECD report56
(even though it varies considerably depending on factors such as channel
of sale57
and product type58
). For instance, one Member State indicated that the return
rate rarely exceeds 10%, except when products have been purchased online59.
Another
national authority estimated that around 80% of products that have relatively low value
and short lifespan remain in consumers’ hands60
.
54
Eurocommerce - Creating a level-playing field for retail in Europe – August 2019.
55
Indication of: manufacturer, brand, type/model, batch number/barcode
56
OECD (2018), Enhancing Product Recall Effectiveness Globally, p. 5.
57
Recalls tend to be more effective if the product was bought online because it’s easier to identify and directly contact
the buyers.
58
Recall effectiveness increases with product price and expected lifespan and decreases with product age.
59
Idem, p. 17.
60
European Commission, 2021, Behavioural study on strategies to increase the effectiveness of product recalls.
18
The EU-wide societal cost of recalled products remaining in consumers’ hands have been
estimated at approximately €378 million in 2019 due to healthcare costs, productivity
losses and losses of quality of life61
. The GPSD Study also estimated that the value of
recalled products that remain with consumers is today EUR 1.3 billion.
The consequences of delayed and ineffective recalls are also exemplified by the deaths
and injuries caused by recent examples of recalled products such as faulty airbags
(estimated to have caused 35 deaths and 300 injuries worldwide62
) and baby sleepers
(associated with 59 baby deaths in the US63
).
The recall procedure is not fully harmonised in the EU, which leads to different
practices, depending on national provisions and economic operators involved. As an
example, fewer than half of EU/EEA countries have established codes of good practice or
guidelines on recalls, and only few of these documents set out requirements as to the
content and channels of recall information or remedies for consumers. The evaluation has
identified this as a significant shortcoming, suggesting that existing requirements are in
themselves currently not sufficient to ensure effective recalls, leading to two problem
areas.
First, many EU consumers are not aware of ongoing recalls of products they own. It is
often difficult to reach the owners of the recalled product. Apart from motor vehicles
(whose registration with public authorities is mandatory), registration schemes are only
available for few higher-value product categories like domestic electric appliances and
communication devices, and even there no link is typically made between registration and
safety64
. In addition, economic operators are hesitant about using customers’ information
collected for other purposes (e.g. in the context of online sales or loyalty programmes) in
the event of a recall because of legal uncertainty about the compliance with the General
Data Protection Regulation.65
Also, there are no comprehensive public sources of recall
information for consumers. For instance, in most EU/EEA countries, the recalling
company has no obligation to put the recall notice on their website or social media and
not all Member States’ authorities publish recall information on their websites, in
addition to reporting recalls to the Safety Gate/RAPEX66
.
Second, consumers may not return a recalled product even if they are aware of the
recall. According to recent surveys, more than a third of EU consumers continue using a
recalled product despite seeing a recall notice67
. This may be caused by recall notices
being unclear and/or minimising consumers’ perception of risk. For instance, the analysis
of existing recall announcements showed that over half of them used terms and
expressions, which could downplay risk, such as ‘voluntary/precautionary recall’,
‘potential concern/problem’, ‘in rare cases/in specific conditions’ or highlighting that
61
Idem
62
https://www.consumerreports.org/car-recalls-defects/takata-airbag-recall-everything-you-need-to-know/
63
https://www.washingtonpost.com/gdpr-
consent/?next_url=https%3a%2f%2fwww.washingtonpost.com%2fbusiness%2f2019%2f10%2f17%2fstudy-concludes-
design-rock-n-play-other-infant-sleepers-led-deaths%2f
64
European Commission, 2021, Behavioural study on strategies to improve the effectiveness of product recalls.
65
European Commission, Notes from EU Workshop on strategies to maximise the effectiveness of product recalls, 23rd
October 2019, p. 2.
66
However, not all recalls need to be notified to Safety Gate/RAPEX. As regards products posing a less than serious
risk, notification is encouraged but not mandatory in the case of voluntary measures taken against products covered by
the GPSD and in the case of both voluntary and compulsory measures taken against products subject to EU harmonised
legislation. In addition, Member States are not required to notify corrective measures in cases where the effects of the
product risk cannot go beyond the territory of the Member State.
67
European Commission, 2021, Behavioural study on strategies to improve the effectiveness of product recalls,
European Commission, 2019, Survey on consumer behaviour and product recalls effectiveness. Final Report
https://ec.europa.eu/consumers/consumers_safety/safety_products/rapex/alerts/repository/tips/ Product.Recall.pdf
19
there have been no reported injuries. Also the procedure for consumers to return the
recalled product may be complex and burdensome and the remedies offered may not be
sufficiently attractive and timely. In a recent consumer survey, recall process taking too
much time and effort was the second-top reason for not responding to a recall (after the
product being cheap)68
.
The stakeholders especially affected by insufficient recall effectiveness are socially
disadvantaged, younger and less safety-conscious consumers (who have shown to be less
responsive to product recalls and less likely to register their products69
) as well as
consumers living in remote areas (for whom returning the recalled product can be costly).
Diverging national requirements (e.g. on recall communication and remedies) also result
in an uneven level-playing field for companies.
2.1.4. Market surveillance rules are complex and not fully effective
Following the adoption of Regulation (EU) 2019/1020, the market surveillance rules
differ for harmonised and non-harmonised products. This Regulation is applicable to
the non-harmonised area under GPSD only regarding the provisions for customs controls.
The market surveillance rules under this Regulation apply only to harmonised products
and differ from those for non-harmonised products in several aspects: responsible
operator in the EU for products entering the EU market, online market surveillance tools
(mystery shopping, blocking websites), strengthened market surveillance rules (e.g.
Single Liaison Office, cross-border mutual assistance).
The Evaluation has also identified coherence problems resulting from the fact that there
are two different sets of market surveillance rules, for harmonised and non-harmonised
products. One good example is toys (e.g. doll’s bed) and childcare articles (e.g. baby’s
bed), that might be conceptually very close and targeting the same consumers, but are
however regulated differently: a toy is a harmonised product regulated by Directive
2009/48/EC, a baby’s bed is a childcare article, which is a non-harmonised product,
falling under the scope of application of the GPSD. Therefore, market surveillance
authorities have different powers for these two products: for example they can carry out
online investigations under covered identity for a doll’s bed, but not for a baby’s bed, as
explained above.70
This has also clear implications for the effectiveness and efficiency of the GPSD71
as it
may lead to market surveillance inefficiencies and thus higher presence of unsafe
products on the EU market in the non-harmonised area. Ensuring coherence between
these rules is important both for market surveillance authorities (they have difficulties to
apply different rules according to the products, e.g. since they do not have the same
market surveillance tools for harmonised and non-harmonised products in online sales)
and for economic operator who might deal, at the same time, with both types of products:
different rules complicate and make more expensive the business activity.
The Evaluation identified also several additional problems for market surveillance of
product safety:
68
European Commission, 2021, Behavioural study on strategies to improve the effectiveness of product recalls
69
Idem, European Commission, 2019, Survey on consumer behaviour and product recalls effectiveness. Final Report
https://ec.europa.eu/consumers/consumers_safety/safety_products/rapex/alerts/repository/tips/ Product.Recall.pdf
70
Regulation 2019(1020) creates e.g. obligation to designate a responsible economic operator in the EU, possibility to
use specific tools for online market surveillance (mystery shopping, blocking websites), etc.
71
This leads to asymmetrical obligations for the different actors based on whether they are dealing with harmonised or
non-harmonised products, leading to administrative burden and complexities for EU businesses.
20
- Market surveillance and customs authorities lack appropriate instruments to enforce
product safety rules for non-harmonised products, in particular in online sales, such as
blocking websites, mystery shopping.
- Products are difficult to trace throughout the supply chain. In particular, products such
as laser pointers, lighters, jewellery, or decorative articles, that fall within the scope of
GPSD and are not subject to sector-specific harmonisation rules, are more likely to
lack relevant information items that are essential to trace them in case they are notified
to Safety Gate/RAPEX.
- The process for adopting voluntary safety standards for the GPSD products is complex
and not sufficiently efficient. It requires a three step process involving comitology.
This could be simplified and streamlined.
- There are differences in the GPSD implementation across Member States. (e.g. the
traceability requirements differ between Member States).
- There is a lack of a mechanism at EU level to solve divergent positions of Member
States regarding the risk assessment of a specific product, which causes a difference in
the treatment of some consumer products inside the Single Market. The number of
notifications to the Safety Gate/RAPEX that were subject to disputes between Member
States has been on average 30 per year.
- The deterrent effect of the GPSD might not be effective enough. A plausible
explanation for this suggested by several stakeholders might be that the sanctions and
penalties for product safety infringements, that are not harmonised across Member
States, remain low. This creates a problem in the context where all products cannot be
controlled by the national authorities, in view of their huge volumes and need to
facilitate trade and free movement of goods.
- The market surveillance system under the GPSD appears to be operating under
considerable resource constraints72
. Market surveillance authorities have indicated
limited staff/financial resources for market surveillance and enforcement most
frequently as a key factor influencing negatively the level of achievement of their
tasks.
- The difficulty of taking enforcement actions against economic operators outside the
EU. This is particularly relevant as the growth of online sales 73
have resulted in an
increase of direct imports; around 150 million small consignments are imported free of
VAT into the EU each year74
.
Uneven and, in some cases, insufficient enforcement actions can harm EU consumers,
since they are exposed to more dangerous products, but also risk to distort competition
for EU businesses and create obstacles to free movement of goods. National market
surveillance authorities suffer from higher administrative costs as a consequence of cross-
border inefficiencies and investigation costs if the relevant operator or the product to be
traced are difficult to find. Discrepancies in the GPSD implementation create an uneven
playing field between Member States and additional regulatory burden for businesses
active across the EU.
The fragmentation of the market surveillance rules between harmonised and non-
harmonised products may also create regulatory burden both for national administrations
and EU businesses. The complexity of the market surveillance legislation creates higher
72
See Annex 11 on the results of the OPC. When asked about the main challenges for enforcement half of the
respondents considered as problematic that Member States’ authorities did not have enough resources (49%), followed
by the difficulty of taking enforcement actions against economic operators outside the EU (46%).
73
Idem
74
European Commission , Memo 2017 - Modernising VAT for e-commerce
https://ec.europa.eu/commission/presscorner/detail/en/MEMO_16_3746
21
costs for economic operators (additional costs of complying with different national
market surveillance and product safety rules for businesses operating in more than one
Member State). The current standardisation process for non-harmonised products also
creates unnecessary administrative burden at EU level and undermines the efficiency of
the standardisation process under the GPSD.
2.1.5. Inconsistent application of product safety rules for food-imitating
products
The legal framework providing rules on safety issues linked to food imitating products is
set out in Directive 87/357/EEC (FIPD). This Directive was adopted before the GPSD
was created the horizontal legal framework for safety of all non-harmonised products,
and it aimed at harmonising the divergent pre-existing national rules on food imitating
products75
. Such separation of rules according to a specific aspect of a product creates
regulatory complexity for national administrations and economic operators.
As reflected in the Evaluation, the number of Safety Gate/RAPEX notifications of food
imitating products is a small percentage of the total. Between 2005 and 2015, a total of
258 notifications (around 17 per year on average) relate to food imitating products.
Moreover, it seems that the product category “Food-imitating products” was only used up
to 2015; afterwards, the products have been categorised according to their use (cosmetics,
clothing, etc.): since 2015, 71 notifications mentioning products not complying with the
FIPD have been submitted under other categories such as cosmetics, kitchen/cooking
accessories, stationery, or decorative articles. Despite that, the number of notification for
food-imitating products remain low. The aspects related to the imitating nature of the
product were incorporated in the risk assessment of the product itself, but not in a
systematic manner by all Member States.
Indeed, the safety provisions of the FIPD are applied differently between Member
States, which have diverging positions on substantial issues, in particular whether all
food-imitating products should be banned per se or measures against these products
should be based on a risk assessment under this Directive. Indeed, the FIPD was adopted
before the GPSD, which sets out the principle of the necessity of risk assessment before
taking appropriate measures against dangerous products and some Member States started
to apply to food-imitating products the GPSD logic while others maintained the primary
interpretation of the FIPD as a ban of these products.
Such different application of this Directive leads to an uneven treatment of these products
across the EU and risks to create distortions of competition in the Single Market. The
stakeholders impacted by this problem are businesses producing such products, due to the
lack of clarity of rules, and also European consumers, who are differently protected
against these products. National market surveillance authorities also suffer from higher
administrative costs due to complexity of the rules. The low number of notifications
related to food-imitating products in the Safety Gate/RAPEX also raises a question
whether a separate legal regime for these products remains justified.
2.1.6. Problems related to the legal form
The current legal form, a Directive, creates several problems linked especially to the
implementation and national differences regarding the date and/or manner of
transposition. Several problems have been encountered in the application of the GPSD,
such as the application of provision on the corrective measures to be adopted in case of a
dangerous product found on the market (Article 8 of the GPSD): the measures to be
75
The first version of the GPSD was the Council Directive 92/59/EEC of 29 June 1992 on general product safety
22
adopted and the type can vary consistently among Member States national transposition
legislation entailing different treatments in different Member States, bringing eventually
to fragmentation in the internal market. Even more relevant, in practice, is the different
transposition of the provision on traceability (Article 5 of the GPSD): the requirements
contained in this article are transposed differently, for example as far as the indication of
the batch of a product or the way and location of the identity and details of the producer
are concerned.
2.1.7. Regulatory burden and costs of the GPSD (REFIT problem)
To summarise, several of the aspects of the current GPSD developed above create
unnecessary regulatory burden for MSAs and companies: discrepancies between market
surveillance procedures for harmonised products and non-harmonised products, market
surveillance inefficiencies between Member States due to diverging assessments and
actions taken against products, lack of clarity inter alia about the scope of GPSD, sub-
efficient standardisation procedures, different implementation of safety rules for food-
imitating products, differences in GPSD implementation, and lack of resources for MSAs
to implement the rules.
2.2. What are the problem drivers?
The figure below summarises the main drivers for the general problem and the five
specific problems identified above:
Figure 1: GPSD General problem tree
The underlying drivers (causes) for the identified problems are multifaceted:
23
Table 2: Nature of problem drivers
Problem
area
Drivers
nature
Drivers’ action
General
problem of
unsafe
products
occurring
on the
market
Market
failure
Under classical economic rationale, the producer’s objective is profit maximisation and
therefore to produce the given product at the best price; in the absence of any safety
regulation, it may tend to sacrifice the product quality and safety by choosing cheaper
components and undergoing less safety verifications. The market price of a dangerous
product does not reflect the real cost for the consumer and for the society, in particular the
costs of injuries linked to a dangerous product. Therefore, the market does not take into
account the negative externalities of dangerous products in terms of reduced public health
and increased public health expenses and such market outcome is therefore not optimal for
the society. However it also has to be taken into account that it is important for some
companies to produce products that are safe for consumers, as their reputation is at stake
and they may lose the consumers’ trust in their brand.
Asymmetry
of
information
Information is needed for markets to operate efficiently. Buyers need to know about the
quality and safety of the product to assess its value. The consumer does not know exactly in
which way the product has been produced and its exact components, while the safety of a
product can be verified including by undergoing a proper laboratory test, which is of course
impossible for an individual consumer to do in advance. Consequently, most of the times
consumers cannot verify the product safety when buying a product and take this into
consideration in their choice; therefore they may make the wrong choice and buy unsafe
products, which leads to sub-optimal societal outcomes.
Split markets In the markets where actors have different and not aligned objectives and the information is
imperfect (as explained above), socially desirable actions are not undertaken and regulation
can then redefine the characteristics of products to be placed on the market, as it is the case
for product safety legislation.
New
technologies
Regulatory
failures
The GPSD does not provide enough legal certainty about the coverage of the specific
features of new technology products, such as software updates or the evolving nature of
some products. Some new types of risks linked to new technologies (such as cybersecurity
risks affecting safety) are not explicitly covered, which leads to legal uncertainty.
Consequently, the current GPSD does not efficiently play its role of safety net for new
technology products.
Online sales Regulatory
failures
When the GPSD was developed, online sales were still at an early stage and therefore the
GPSD’s provisions do not properly address the challenges of the current online
environment. The GPSD does not set out specific obligations related to product safety for
the online marketplaces, while these play today an important role in the online sales. Also,
the GPSD does not provide for effective investigation tools for online sales. Finally, E-
commerce allows for an important increase of direct imports from economic operators
located outside the EU. While the GPSD creates product safety obligations for any products
being placed on the EU market, regardless of their place of origin, it is very difficult to
enforce against traders established outside the Union and offering their products to EU
consumers. There are therefore enforcement difficulties allowing the entering of non-
harmonised consumer products on the EU market without having an economic operator
responsible for these products in the EU.
Recalls Regulatory
failures
Recall procedure as such is not defined under the GPSD. In particular, there are no
minimum requirements on the content and channels or recall communication or remedies
that consumers are entitled to. In some countries, requirements are more prescriptive than
in others, leading to varying levels of consumer protection. One major deficiency is the lack
of legal basis for using existing customers’ data for recall purposes.
Market
failures
(companies
fail to act
Companies may fear the negative reputational impact and other costs created by a recall
and thus avoid communicating clearly about possible safety issues and delay recall
measures and/or underplay the risk when the product turns out to be dangerous76
. Almost
half (47%) of industry respondents to a European Commission’s survey indicated that they
76
An analysis of existing product registration schemes indicated that very few companies make a link between
registration and safety, while a similar analysis of recall announcements showed that over half of them used terms and
expressions, which could reduce consumers’ perceptions of risk European Commission, 2021, Behavioural study on
strategies to improve the effectiveness of product recalls.
24
effectively) did not have a written procedure in place in case the product needs to be recalled (even
though safety-conscious companies were likely overrepresented in the survey)77
. Also
marketing literature suggests that most companies engage in a passive recall strategy rather
than a proactive one78
.
Behavioural
biases
(consumer
inertia)
Consumers do not always act in a rational way in response to recalls. Biases such as
information overload and framing effects mean that if recall notices are lengthy and
unclear, consumers may ignore them, especially if they are time poor. Over-optimism may
result in consumers underweighting the risk posed by a recalled product, while inertia and
endowment effect79
relate to the fact that consumers have an inherent preference for status-
quo, which in the case of recalls means keeping the product.
Market
surveillance
Regulatory
failures
At EU level, the market surveillance rules for harmonised and non-harmonised products are
not only set up in two different legal texts, but also the applicable rules differ in several
aspects for these two categories of products, which creates regulatory complexity for
national administrations and businesses. Furthermore, implementation issues stem from the
fact that the GPSD is a Directive and as such is not directly applicable in the EU and, as the
Evaluation shows, is transposed differently across EU Member States. The GPSD also does
not sufficiently harmonise the provisions on the product traceability, which are therefore
defined at national level and prove to be insufficient. The GPSD does not tackle the
disputes on the risk assessment between Member States, and the standardisation procedure
under the GPSD is not efficient enough.
At national level, the main driver for enforcement problems is lack of resources dedicated
to market surveillance by Member States. Also, the current level of penalties and sanctions
does not create a sufficient deterrent effect for economic operators to prevent the placing of
unsafe products on the market.
Food-
imitating
products
Regulatory
failures
The uneven application of product safety rules for food-imitating products stems from
the fact that the rules are formulated in such a way that it allows a very different application
across EU Member States, some categorically banning all food imitating products, some
others performing a risk assessment before deciding on the measure. The fact that these
rules are set out in another piece of legislation than the rest of the product safety rules
creates unnecessary regulatory complexity for national administrations and businesses and
leads to incoherent measures on the Single Market.
2.3. How will the problem evolve?
Some of the identified problems will remain and even likely get worse: in particular those
linked to online sales and new technologies. There is a clear increasing trend in online
sales in the EU. The COVID-19 crisis and the repetitive lockdowns are accelerating e-
commerce, as well as imports of consumer products from outside the EU. There is also an
increase of new technology consumer products being available on the EU market.
Therefore, the magnitude of problems linked to these new digital challenges is likely to
increase.
At the same time, digital developments offer also opportunities for more efficient market
surveillance by using new technology tools, for example to identify already recalled
products online. Online sales may ease the identification of customers, which is
particularly important in recalls. Also, connected products may be easier to recall and fix
or switch off remotely.
77
Idem.
78
Chen, Yubo & Ganesan, Shankar & Liu, Yong. (2009). Does a Firm's Product-Recall Strategy Affect Its Financial
Value? An Examination of Strategic Alternatives During Product-Harm Crises. Journal of Marketing American
Marketing Association ISSN. 73. 214-226. 10.1509/jmkg.73.6.214
Mukherjee, U., Ball, G., Wowak, K., Natarajan, K. and Miller, J (2021), Hiding in the Herd: The Product Recall
Clustering Phenomenon, Manufacturing & Service Operations Management, https://doi.org/10.1287/msom.2020.0937
Kalaignanam, Kartik & Kushwaha, Tarun & Eilert, Meike. (2012). The Impact of Product Recalls on Future Product
Reliability and Future Accidents: Evidence from the Automobile Industry. Journal of Marketing. 77. 10.2307/23487412
79
In behavioural economics the endowment effect is he finding that people are more likely to retain an object they own
than acquire that same object when they do not own it
25
The new Regulation (EU) 2019/1020 will only have a limited effect on the market
surveillance for products and risks covered by GPSD since only its provisions on customs
(Chapter VII of this Regulation) apply to these products.
Some other problems will also continue to exist and are likely to remain the same or of
the same magnitude in the absence of EU action, in particular the fragmentation,
complexity and ineffectiveness identified in the market surveillance rules. These
problems are mainly linked to regulatory failures of the legal framework itself and would
get worse only if there is an increased trend of non-harmonised consumer products
circulating on the EU market. Problems linked to lack of resources mostly relates to the
political priorities and resources of the Member States.
3. WHY SHOULD THE EU ACT?
3.1. Legal basis
The legal basis for this initiative is Article 114, with due regard to Article 16980
, of the
TFEU. The GPSD has for object ensuring product safety and improving the functioning
of the internal market. GPSD aims at ensuring a high level of consumer protection, by
contributing to protect the health, safety of European consumers and promoting their right
to information81
.
The EU has no exclusive competence on product safety, which is a shared competence.
Therefore, the subsidiarity principle does apply.
3.2. Subsidiarity: Necessity of EU action
The GPSD harmonises the general product safety requirement in the EU. Ensuring safety
of products in the Single Market cannot be achieved sufficiently by Member States acting
alone for the following reasons:
- Data show unsafe products are spread across the EU: unsafe products can be found in
all Member States82
.
- Products circulate freely across the Single Market, including the dangerous ones.
When a dangerous product is identified in a certain country it is very likely that the
same product could be found in other Member States too, not least following the
exponential growth of online selling. This is demonstrated by the number of follow-up
actions taken by Member States in their country after the notification of a dangerous
product in the RAPEX/Safety Gate; while in 2011 there were 2100 follow-up
measures, in 2019 more than 4400 of such measures were notified to Safety
Gate/RAPEX.
- Different rules on product safety at national level can create uneven costs for
businesses to comply with product safety legislation and therefore can cause
distortions of the internal market when /if companies want to operate across borders.
- According to Article 169 of TFEU, in order to promote the interests of consumers and
to ensure a high level of consumer protection, the Union shall contribute to protecting
the health and safety of consumers. If there are different rules concerning product
80
Article 169 make reference to Article 114 to achieve its objectives.
81
Also, product safety is part of the high level of consumer protection that Union policies ensure (see Article 38 of the
Charter of Fundamental Rights of the European Union) and therefore one of the pillars of the EU consumer protection
policy.
82
See Safety Gate/RAPEX annual report –
https://ec.europa.eu/consumers/consumers_safety/safety_products/rapex/alerts/repository/content/pages/rapex/reports/d
ocs/RAPEX.2019.Factsheet.EN.pdf
26
safety and its market surveillance, EU consumers will not be protected against
dangerous products in the same way across the Member States.
- The identified problem drivers mostly do not have any national or sub-national
specificities (problems linked to digital challenges, recalls and food-imitating
products). Problem drivers for market surveillance have partly a national dimension
concerning the lack of resources, level of penalties and availability of market
surveillance tools, which can differ from one Member State to another.
The objective of products safety cannot be sufficiently achieved by the Member States
acting alone, given the need for a very high degree of cooperation, interaction and
coherent action of all the competent authorities in all Member States across the Single
Market to ensure the same high level of protection of consumers and enable effective
action on the Single Market where products circulate freely. Member States cannot
ensure cooperation and coordination by acting independently.
The GPSD establishes the cooperation and coordination between Member States: via the
EU Safety Gate/RAPEX, Member States inform each other about measures taken against
dangerous products. They also take follow-up actions in their territory if the product
alerted is present there. Moreover, authorities consider the implementation of EU
coordinated market surveillance activities on product safety extremely useful, as
economies of scale and the funding provided by the Commission have allowed them to
carry out inspections for some priority categories of products.
The measures under this initiative would not affect the Member States' competences in
market surveillance or assessment of risks, neither would they interfere with national
enforcement or judicial systems, nor would they affect the internal division of
competences among authorities at national level. In the product safety field, Member
States can act first independently to notify the corrective measures taken against
dangerous products, but then follow-up actions are required from all other Member
States.
3.3. Subsidiarity: Added value of EU action
EU level action in product safety for non-harmonised products has clear benefits
demonstrated by the GPSD evaluation:
- Common Union rules allow economy of scale in market surveillance, in particular
nowadays with the exponential development of online selling which intensifies sales
across the EU and direct imports from outside the EU. Sharing costs of market
surveillance occurs also by performing joint market surveillance actions among EU
countries and exchange information.
- EU action allows faster and more efficient circulation of information, in particular
via the Safety Gate/Rapex, thus ensuring fast actions against dangerous products
across the EU and also level playing field.
- Common rules for product safety at EU level have benefits in term of costs savings
and lower administrative burden and complexities for businesses by avoiding them
having to comply with heterogeneous sets of national rules. This enables also free
circulation of goods in the EU and allows for closer cooperation between Member
States.
- Common Union rules enable developing EU product safety standards, which by
giving EU-wide presumption of safety facilitate product safety compliance for
businesses (and potentially decrease the related costs).
- At international level the common set of provisions established by the GPSD has
also allowed the EU to be stronger in promoting high level of safety with
international actors, thus tackling the increasingly high circulation of goods from
third countries via online selling.
27
The functioning of the internal market will be improved by EU level action since
common product safety and market surveillance rules across the EU will ensure a more
even treatment of businesses and therefore less likely distort competition on the EU
Single Market. Better market surveillance and enhanced coordination between Member
States will lead to higher detection of unsafe products, and thus to higher consumer
protection and trust.
The food-imitating product directive is currently subject to very different interpretations
between Member States, ranging from a ban of such products to the inclusion of the food-
imitating aspect in the elements taken into account in the risk assessment. This leads to a
fragmentation of the internal market regarding such products, a more unified approach of
food-imitating products is therefore needed at the EU level, requiring Union action.
4. OBJECTIVES: WHAT IS TO BE ACHIEVED?
Figure 2: Schematic overview of problems and objectives
4.1. General objectives
The general objective of the GPSD is to ensure EU consumers are protected from
dangerous products and to ensure the proper functioning of the Single Market. These two
main objectives of the GPSD are interlinked: if the same high level of safety
requirements applies to all economic operators, it ensures the health and protection of EU
consumers and also level-playing field for all businesses operating on the EU market.
4.2. Specific objectives
This initiative has five specific objectives linked to the five problems identified; it also
seeks to simplify legislation and reduce the administrative burden of the current acts
under consideration (REFIT objective):
28
Table 3: Specific objectives
Figure 3 presents the intervention logic for this initiative:
Specific objectives Description
Ensure that the EU legal
framework provides for a safety
net for all consumer products and
risks, including products and risks
linked to new technologies
This initiative aims to make sure that the GPSD ensures the safety of all non-harmonised
products and addresses relevant risks. In view of the development of new technologies,
there is a need to ensure legal certainty regarding the legal coverage of new technology
products such as connected products and AI, and to be able to address new products safety
risks for health and safety of consumers related to these new technologies, when not
already covered by sectoral legislation. The aim is to preserve the role of the GPSD as a
safety net for consumers.
The initiative does not aim to regulate all risks related to new technologies in general but
only when they create risks to health and safety of consumers (e.g. cybersecurity can entail
risks for privacy or data protection, which are not covered by the GPSD; the GPSD would
only cover the risks related to health and safety (e.g. physical incident) created by e.g. lack
of sufficient cybersecurity features)
Address product safety challenges
in the online sales channels
There is also a need to adapt the GPSD to the new challenges of e-commerce. Product
safety must be ensured irrespectively of the modalities of the supply chain: rules for new
online business models need to be clarified and provisions for market surveillance of
imported products improved to enable appropriate consumer protection and level-playing
field for businesses. Also national market surveillance authorities need to have efficient
tools to perform market surveillance of online sales and the product traceability in the
online sales needs to be improved.
Make product recalls more
effective and efficient to keep
unsafe products away from
consumers
Product recalls play an important role to ensure the safety of EU consumers, since they are
the last resort to keep dangerous products away from them. This initiative aims to increase
the effectiveness and efficiency of recalls by improving the channels and content of recall
information, making recall procedure less burdensome for EU consumers and ensuring
their right to an effective remedy. The initiative takes into account the identified
behavioural biases to increase consumer response. The initiative also aims to ensure
effective monitoring of recall actions.
Enhance market surveillance and
ensure better alignment of rules
for harmonised and non-
harmonised consumer products
This initiative aims to ensure better enforcement of product safety rules by contributing to
more efficient market surveillance. The objective is to improve product traceability so that
dangerous product can be effectively eliminated, the deterrent effect of the legislation for
economic operators not complying with the rules, and to tackle possible discrepancies
about risk assessment between Member States. The aim is also to simplify the procedures
leading up to referencing standards in the Official Journal of the EU for non-harmonised
products.
Following the recent adoption of Regulation (EU) 2019/1020, there is a need to align the
market surveillance legislative framework for non-harmonised products with the one for
harmonised products, the definitions of the GPSD with product harmonisation legislation
and ensure equal treatment for all products and businesses.
The objective is also to ensure more efficient and coherent enforcement and
implementation of the product safety rules across the EU and to monitor that sufficient
resources are dedicated to market surveillance at national level.
Address safety issues related to
food-imitating products
This initiative aims to ensure a consistent application of product safety rules for food-
imitating products by simplifying and clarifying those rules.
REFIT Simplification and
improving the efficiency of the
existing legislation
This initiative aims also to simplify and reduce the regulatory burden of the current GPSD.
29
Figure 3: Intervention logic
Source: GPSD Study
5. WHAT ARE THE AVAILABLE POLICY OPTIONS?
5.1. What is the baseline from which options are assessed?
In the baseline scenario, no new legislative or non-legislative actions specifically
targeting the safety of consumer products will be developed at EU or national level. This
scenario includes several EU-level and national policies and measures which are assumed
to continue being in force or will enter into force in the future:
At EU level, the Commission has recently adopted a proposal for a Digital Services Act
which, if adopted by the co-legislators and once entered into force, should set up new
responsibilities for online intermediaries, including online marketplaces. Also, the
Commission has recently announced its intention to propose new legislative initiatives
linked to new technologies and artificial intelligence, namely the proposal for a
horizontal instrument on AI and the proposal for the revision of the Machinery
Directive, which will clarify certain sectorial safety aspects of new technologies. The
new customs provisions applicable for GPSD products under the new Regulation (EU)
2019/1020 will start to apply in 2021. The Commission will continue, in the frame of the
allocated EU budget, to finance coordinated market surveillance activities on product
Needs/Problems Outputs/results Impacts
Objectives
Global objectives
 To ensure the safety of non-
food consumer productson
the EUmarket
 To contribute to the
functioning of the Single
Market and ensurealevel
playing field for businesses
Inputs/activities
• Implementation of changes
to the EUlegal framework
on product safety
• Provision of funding and
staff at EUlevel for product
safety information
exchange,coordination
and training measures
• Provision of funding and
staff at Member State level
for controlsof importsfrom
third countries,market
surveillance and
supervision of corrective
actionsetc
• Activitiesin the fieldsof:
• Safety Gate/RAPEX
• Risk assessment,incl.new
typesof riskslinked to
new technologies
• Training (e.g.E-enforce-
ment Academy)
• Exchangeof information
with third countries
• Data collection,including
on product-safety related
injuries
• Enforcement of product
safety obligationsof
companiesand
intermediaries,including
relevant new actors
• Standardisation activities
• Consumer awareness
raising regarding product
safety
Reduced discrepanciesin
implementation of EUproduct
safety legislation acrossMS
Consumer are better
protected online and offline
from unsafe consumer
products, including from
productsthat pose new types
of risks– including those
linked linked to new techno-
logies–, and risksdueto envi-
ronmental aspectsthat are
relevant for human health
More effective corrective
actions, including recalls
EUlegal framework modernised
to provide a safety net to
consumersregarding the safety
of non-food products
More standardsprepared in a
shorter timeframe for non-
harmonised productsthat are a
cause of consumer detriment
More efficient standardisation
procedure of mandating and
referencing standardsfor non-
harmonised products
Discrepanciesin implementation
of GPSD acrossMSand
divergencesin risk assessment
between national authorities
Product safety challenges,
including related to new
technologiessuch asIoTor AI
lead to new risksfor consumers
Improved enforcement of
product safety obligations
Reduced compliance costsand
administrative burdensfor EU
businesses
Standardisation processunder
GPSD more burdensomethan for
harmonised products
More effective RAPEXsystem,
e.g.by reducing the time from
identification of an unsafe
product to itsnotification
through the system
Obligationsfor business
operatorsregarding safety large-
ly independent from whether
they are dealing with harmoni-
sed or non-harmonised products
Lack of level playing field for EU
businessescompared to
operatorsin third countries
Better defined rolesand
responsibilitiesof different
actorsincl. when productsare
purchased by consumers
directly from third countries, for
product recallsetc
Better traceability of unsafe
consumer products
Level playing field for EU
businesseswith respect to
product safety obligations,
incl. with operatorslocated in
third countriesand with
respect to new supply chain
actors
Reduction of consumer
detriment due to unsafe
products, including regarding
productspurchased online or
directly from third countries
Market surveillance rulesunder
GPSD are complex and not fully
effective
Fragmented EUlegal framework
lead to increased compliance
costsand administrative burdens
for EUbusinesses
Sufficient enforcement powers
provided to national authorities
in all Member States
Specificobjectives
Ensure that the EUlegal
framework providesfor general
safety rulesfor all consumer
productsand risks,including
product riskslinked to new
technologies
Addressproduct safety
challengesin the onlinesales
channels
Make product recallsmore
effectiveand efficient to keep
unsafe productsaway from
consumers
Enhancemarket surveillance
and ensure better alignment of
rulesfor harmonised and non-
harmonised consumer products
Addresssafety issuesrelated to
food imitating products
Consumersare offered products
online that may be unsafe, not
sufficiently traceable and
covered by market surveillance
Increased trust of consumers
that productsare safe and
consequently, in the Single
Market
EUlegal framework more
consistent and lesscomplex/
fragmented,incl. with respect to
food imitating products
Consumersbetter aware of
product safety and recalls
More innovative and
competitive EUeconomy that
leadson the safety and
sustainability of itsproducts
Increase in e-commerce,incl.
direct B2Ctransactionswith non-
EUcountries, facilitated by new
economic operators, such as
online marketplaces
Insufficient resourcesfor market
surveillance in some MS,incl. for
online market surveillance
Lack of consistent
and actionable data on product-
related injuriesof consumers
Inconsistent application of
product safety rulesfor food-
imitating products
Insufficient effectivenessof
recallsof consumer products
Better dataon product-related
injuriesof consumers
Consumerssuffer detriment due
to unsafe products, and may lose
trust in the Single Market
More uniform risk assessment of
responsible authoritiesin the EU
regarding product safety related
risk, including new typesof risks
linked to new technologiesand
environmental risks
30
safety (see Annex 9). The Commission is expected to continue its advocacy policy on
product safety, in the form of information campaigns and other promotion initiatives
such as the Product Safety Award83
. The Commission will also continue its coordination
role in product safety as Chair of the Consumer Safety Network (‘CSN’)84
. In the area
of product safety in the online sales, the baseline scenario takes into account that the
Commission will continue the cooperation with and steering the commitments of the
online market places in the context of the Product Safety Pledge. Finally, the Commission
will continue adopting safety standards giving presumption of safety for non-
harmonised products under the current procedure, and also its international cooperation
activities.
At national level, it is assumed that Member States will also continue their measures
supporting product safety policy, such as information and promotion campaigns, under
the constraints of the national budgets, and their current market surveillance activities.
Several expected socio-economic developments are also relevant for the product safety
area. Important technological developments bring an increasing number of AI-driven
consumer products and connected products on the EU market. Also the increasing
digitalisation of online sales, dramatically accelerated during the current COVID-19
crisis, will increase the number of products sold online and also those imported directly
from outside the EU. Demographic changes can also have an impact on the safety of
consumers, as for example older people have specific consumption-related needs85
.
The time horizon for this baseline scenario, which will be used for the assessment of
impacts of the different options is a 10 years’ horizon. This takes into account the likely
lifetime of any individual option and on the need to allow for impacts to be realised.
The GPSD Study estimated the costs associated with this baseline scenario for businesses
and Member States as following:
Table 4: Estimated annual cost for businesses to comply with the GPSD, by company size class, in
million EUR
Cost by company size Total costs
Company size (employees) From 0 to 49 50 – 249 250 or more All size categories
Total of manufacturing 79 101 163 343
Total of wholesale 118 81 122 321
Total of retail 232 44 163 439
Total 428 226 448 1 102
Source: The survey conducted in the context of the GPSD Study
The estimated costs for businesses to comply with the GPSD in its current form
amount to EUR 1.1 billion per year86
.
Consumer detriment linked to the unsafe products is expected to grow in the mid-term in
the baseline scenario, due to increasing consumption and a continuing shift to e-
commerce: the GPSD Study evaluates that consumers suffer financial costs of EUR
19.3 billion in 2019 arising from the fact that they have purchased unsafe products that
83
Since 2019 the Product Safety Award rewards every two years businesses going the extra mile for product safety,
beyond their legal requirements.
84
CSN is a network of authorities of the Member States competent for product safety. See Annex 9.
85
COM(2020) 696 final - New Consumer Agenda - Strengthening consumer resilience for sustainable recovery
86
Product safety-related costs that companies would incur anyway (i.e. even in absence of product safety legislation,
e.g. because these costs relate to due diligence) are deducted.
31
they would not have purchased if they knew these products are unsafe. This consumer
detriment in the EU due to unsafe non-harmonised products is estimated on the basis that
the product value is EUR 3.9 billion for the online sales channels, and EUR 15.4 billion
for brick-and-mortar shops and other offline sales channels, for a total of EUR 19.3
billion per year. This estimated baseline consumer detriment in the EU related to unsafe
non-harmonised products is currently reduced due to recalls by approximately EUR 0.4
billion per year87
. This consumer detriment due to the loss of value of unsafe products is
expected to reach EUR 20.8 billion by 2025 and almost EUR 22 billion by 2034 in the
baseline scenario88
.
The cost estimation for Member States takes into account the different organisational
approaches of Member States to market surveillance and is based on staff data for
surveillance of non-harmonised consumer products at national level. The GPSD Study
concluded that the total EU27 staff-related costs for market surveillance of non-
harmonised consumer product amount to approximately EUR 122.4 million per year89
.
Also, four in ten MSAs reported incurring costs other than staff costs (e.g. testing of
products), estimated at most at 0.34% of total staff costs.
The GPSD Study identified that (minor to significant) additional costs due to differences
in the safety requirements in Member States, caused by differences in the national
implementation of the GPSD (e.g. regarding traceability requirements) or legislative
fragmentation between harmonised and non-harmonised products, currently affect 42%
of surveyed companies and 16% of MSAs. These costs are estimated for MSAs to
amount to EUR 0.7 million annually (total for the EU27) and to EUR 119 million
annually for businesses.
On the benefits side of the baseline, the interviews carried out in the context of the GPSD
Study identified that authorities and businesses see moderate to significant benefits
resulting from the GPSD across the board, and in particular through better information on
unsafe products and measures taken by authorities provided through the Safety
Gate/RAPEX, a better functioning internal market and increased consumer trust. 90%
respondents that expressed an opinion considered the costs due to the product safety
requirements of the GPSD to be at least “moderately proportionate” to the resulting
benefits. Close to 60% of respondents that had an opinion even found these costs to be
“largely proportionate” or “very proportionate”, including respondents from companies
and business associations.
For SMEs, the estimated annual costs to comply with the GPSD (after subtraction of
business-as-usual costs) are EUR 428 million per year (companies < 50 employees) and
EUR 226 million per year (companies 50 to 249 employees). The median value for
consumer product safety-related costs in proportion of the total annual turnover appears
to decrease with the company’s size/turnover. This is likely due to scale effects. This
general pattern is confirmed by SMEs’ replies to the business stakeholder survey.
Accordingly, SMEs account for 59% of the total of GPSD-related compliance costs in the
EU, in line with their overall share in the market.
The GPSD Study also analysed the impacts of the COVID 19 crisis on the baseline
scenario. It shows that while the confinement measures have serious expected impacts on
GDP, total retail quickly recovered after the first crisis wave, but new measures in the
87
The GPSD Study estimated the total consumer detriment under the baseline scenario with low recall effectiveness to
be about EUR 1.3 billion per year (calculated as a value of recalled products that remain with consumers).
88
The GPSD Study could estimate the impact of options on the consumer detriment taking as assumption that the
detriment incurred by consumers in case of an unsafe product is equivalent to at least its purchase price.
89
Monetised on basis of population size, number of person hours per year and average wage.
32
current second wave are likely to again lead to substantial impacts on retail and therefore
sales of products. The decline in overall retail sales has been accompanied by a rise of e-
commerce sales that are expected to increase by 16.9% in 2020 in Western Europe90
. The
boost in new spending is expected to leave e-commerce permanently ahead of its
previous pace.
In general terms, product safety processes at companies including with respect to related
supply chain management appear to remain largely unchanged in the COVID-19 context,
except with the increasing reliance on electronic communication instead of physical
meetings (this may pose issues to product assessments). Companies also confirmed the
switch to online sales channels to offer products.
Moreover this crisis required increased market surveillance of COVID-19-related
products, in particular face-covering products (other than the medical or personal
protective equipment devices).Until 1 December 2020, 16 Member States notified in the
Safety Gate/RAPEX 174 alerts (mainly safety masks, some disinfecting gels and UV
lamps).
Finally, there is a strong political commitment for a strong product safety policy at EU
level. This has been recognised by the recently adopted Consumer Agenda91
and several
Council conclusions as explained above. The GPSD evaluation confirms the validity of
the GPSD, but at the same time considers the need for its revision. The European
Parliament has also highlighted the need to revise the GPSD in its resolution on
addressing product safety in the Single Market92
. This also has been largely recognised
by the stakeholders in the consultation process.
5.2. Description of the policy options
To address the objectives developed above, the initiative will intervene on the following
areas:
Figure 4: Interventions to improve the GPSD’s effectiveness
90
https://www.emarketer.com/content/western-europe-see-10-83-billion-more-ecommerce-sales-than-expected
91
https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:52020DC0696
92
https://www.europarl.europa.eu/doceo/document/TA-9-2020-0319_EN.html
33
Beyond the baseline scenario (‘Status quo’) not involving any new actions, the identified
policy options to address the different specific objectives are:
• Option 1. Improved implementation and enforcement of the existing legal
framework, without legal revision of the GPSD (only FIPD revised);
• Option 2. Targeted revision of the GPSD (Directive or Regulation);
• Option 3. Full revision of the GPSD and recasting as Regulation;
• Option 4. Integration of more legal instruments
The range of options includes non-legislative (Option 1) and legislative actions (Options
2, 3 and 4) to address the different specific objectives. All the options defined in the
report propose specific actions to address all five problems identified, but differ in the
level of ambition.
The substantive provisions of the Food-imitating Product Directive were considered
to be revised under all options 1 to 4 with two possible sub-options: (a) a full ban of
food-imitating products per se and (b) application of a product safety risk-assessment on
a case by case basis to this category of products.
Table 5: Overview of the policy options and addressees of the measures
Option
0
Baseline
Option 1
Enhanced
enforcement
Option 2
Targeted legal
revision
Option 3
Full legal revision
Option 4
Integration of
rules
Most relevant
stakeholders
New
technologies
No
change
• Guidance for
economic operators
• Use of standards
for new risks
• Coverage of new risks
• No clarifications on
software related rules
• Coverage of new risks
• Clarify software related
rules
Option 3
+
Integration of
the legal
instruments on
market
surveillance
(GPSD market
surveillance
provisions and
Regulation
2019/1020)
Businesses (for
consumer products
incorporating new
technologies) and
MSAs
Online sales
No
change
Reform, promotion
and expansion of the
Product Safety
Pledge
• Making most
provisions inspired by
the Product Safety
Pledge legally binding
Obligations for economic
operators going beyond
the Product Safety
Pledge (e.g. display same
information online as it is
with the product offline,
duty of care as for
distributors)
Online
marketplaces,
online retailers,
and MSAs
Recall
effectiveness
No
change
Guidance on product
recalls
• Mandatory
requirements on
product recalls
• Legal basis to use
customers’ data for
recalls
• Mandatory elements of
recall notice
Option 2
+ Some additional
mandatory
requirements
(e.g. on product
registration, template for
recall notice, right to
remedy and monitoring)
Businesses
(harmonised and
non-harmonised
consumer
products), MSAs
Market
surveillance
No
change
Increased funding of
joint market
surveillance
activities
• Alignment with market
surveillance and
traceability rules of
harmonised products
• Simplifying
standardisation
procedures
Option 2
+ stronger enforcement
powers to Member
States (penalties) ,
arbitration mechanism
and increased
traceability (delegated
acts)
MSAs and
businesses (in
particular
businesses of non-
harmonised
consumer
products)
Food-
Imitating
Products
No
change
Separate revision of
the FIPD to ensure its
even interpretation
2 sub-options for
treatment of food-
imitating products:
(a) full ban per se.
Integration of the FIP
provisions into the GPSD
2 sub-options for
treatment of food-
imitating products:
(a) Maintaining
dedicated provision on
Idem Option 2 Producers of food
imitating products
and MSAs
34
Businesses and MSAs are the most affected stakeholders by the measures as detailed
in the Table 5. The SMEs and micro-SMEs are not exempted from any of the
obligations foreseen under the different options. EU product safety legislation does not
allow for "lighter" regimes for SMEs since a consumer product must be safe whatever the
characteristics of its supply chain to meet the general objective of product safety and
consumer protection. However provisions are foreseen in the EU legislation e.g. to
facilitate access for SMEs to EU safety standards including those adopted under the
GPSD (see Article 6 of Regulation (EU) 1025/2012).
The detailed description of the actions under the different options, as well as their time
horizon, potential for simplification and reduction of regulatory burden and digital
solutions envisaged to increase efficiency, are developed in the Table 6.
(b) risk-assessment
approach
FIP (recast and
integration) with a full
ban per se
(b) Abandoning any
dedicated provision
(repeal) and reliance on
general provisions for
risk-assessment
approach
Instrument Directive Directive Directive or Regulation Regulation Regulation
Table 6: Option packages
Objectives
Option 0
Baseline
Option 1
Enhanced enforcement
Option 2
Targeted legal revision
Option 3
Full legal revision
Option 4
Integration of market
surveillance rules
Ensure the EU legal
framework provides for
general safety rules for
all consumer products
and risks, including
products and risks linked
to new technologies
No change
• Guidance for businesses
that cyber-security threats
and other risks of new
technologies affecting
physical or mental health.
• Exploring use of European
Standards for new risks
• New risks (cyber-security and other risks of new
technologies affecting physical or mental health) explicitly
covered through legal revision of product safety definition
Option 2
+ Clarify software related rules: Explain how software can
impact the safety of product and clarify responsibilities to
ensure consumer safety
Option 3
+
Integration of the legal
instruments on market
surveillance (GPSD market
surveillance provisions merged
with Regulation 2019/1020 into
one Regulation on market
surveillance
Address product safety
challenges in the online
sales channels
No change
• Review, promotion and
expansion of the Product
Safety Pledge
Making most provisions of the Product Safety Pledge
legally binding for all online marketplaces, such as:
• to consult information on recalled/dangerous products
available on Safety Gate/RAPEX and from other sources
and react quickly;
• to take appropriate action in respect to recalled/dangerous
products, when they can be identified
• to provide single contact points for EU MSAs and to
cooperate with them
• to have an internal mechanism for notice and action
procedure with specific provisions for unsafe products (e.g.
timeframes for action) and other requirements
Option 2 + additional requirements for online operators:
• to display of all safety information online that is also
required to be provided offline; online marketplaces required
to make sure that sellers on their platform provide this
information together with the product offer
• a duty of care to help ensure compliance with the safety
requirements for online marketplaces (in the same vein as the
classical distributors have today: stop supplying unsafe
products, participate in market monitoring, keeping
traceability information, cooperation in corrective actions,
cooperate with MSAs, making efforts to identify dangerous
product offers already removed from their websites but that
keep reappearing. That duty of care would be different than
for distributors as they do not have physical contact with the
product, so their role will focus on doing their most to ensure
that their websites do not offer dangerous products, and if they
do, they cooperate with authorities for corrective actions. This
duty of care would be complementary to the obligations of
actual sellers on the online marketplaces)
Make product recalls
more effective and
efficient to keep unsafe
products away from
consumers
No change • Guidance on product
recalls
Mandatory requirements on product recalls:
• Legal basis to use available customer contact details for
recalls
• Operators need to disseminate recall announcements on
their website, social media, newsletters, retail outlets and
other appropriate channels to ensure the widest possible
reach.
• Mandatory key elements for recall notices (product
description + photo, description of hazard, instructions on
what to do, description of remedy, contact details for
queries)
• Prohibition to use terms decreasing the perception of
risk in recall notices
Option 2 + Further measures to enhance recall effectiveness,
for example:
• Obligation for economic operators to notify
consumers directly whenever possible
• Economic operators who already offer product
registration systems or loyalty programmes should
offer consumers the possibility to register their
contact details specifically to receive safety
notifications
• Possibility to set further requirements for
registration of specific categories of products
through delegated act
• Mandatory template for recall notices to be set
through implementing act
36
• Consumers’ right to an effective, cost-free and
timely remedy for the recalled product (repair,
replacement or refund)
• Less burdensome recall procedure for consumers
(returning a product should not incur any financial
costs, non-portable items to be collected by the
operator)Obligation for businesses to register
voluntary recalls in an EU public database and to
monitor recall effectiveness. Power for authorities
to request monitoring data from operators and decide
if the case can be closed.
Enhance market
surveillance and ensure
better alignment of rules
for harmonised and non-
harmonised consumer
products
No change
• Increased funding of EU
joint market surveillance
activities among EU
Member States
Legal revision of the GPSD to align with market
surveillance and traceability rules for harmonised
products:
• The market surveillance rules aligned with Regulation
(EU) 2019/1020
• Additional requirements for businesses in line with
Regulation (EU) 2019/1020 (notably regarding the
requirement of an EU representative) and other
harmonisation legislation, in particular traceability
requirements from Decision No 768/2008/EC
+ Simplifying standardisation procedures (streamlining
the EU process for elaborating safety requirements and the
standardisation request, e.g. by combining them in one
Commission Decision)
Option 2
+
• More stringent rules on penalties to strengthen their deterrent
effect beyond Regulation (EU) 2019/1020
• Arbitration mechanism in case Member States have
diverging product safety risk assessments (either a group of
Member States or the Commission are called to arbitrate)
• Possibility to set further requirements for traceability
systems through delegated acts, for example regarding
chemicals in childcare articles.
Address safety issues
related to food-imitating
products
No change
Separate revision of the
FIPD,
2 sub-options for treatment
of food-imitating products:
(a) full ban per se
(b) risk-assessment
approach
Integration of the FIP the provisions into the GPSD.
2 sub-options for treatment of food-imitating products:
(a) Maintaining dedicated provision on FIP (recast and
integration) with a full ban per se
(b) Abandoning any dedicated provision (repeal) and
reliance on general provisions risk-assessment approach
Idem Option 2
Instrument Directive Directive Directive or Regulation Regulation Regulation
Digital solutions in
respect of
implementation and
reduction of
administrative burdens
None
Development of digital
solutions, such as an IT
system (web-crawler) to
identify dangerous products
sold online and already
notified via Safety
Gate/RAPEX. It would allow
MSAs to carry out online
market surveillance tasks
more efficiently
Beyond the digital IT systems of Option 1, other digital
solutions can reduce the burden linked to the additional
obligations on recalls (use of internet and social media to
increase recall communication). Aligning to the Regulation
(EU) 2019/1020 will also allow to explore digital interlinks
between existing market surveillance systems at EU and
national level and will therefore make the market surveillance
more efficient.
Idem Option 2 + possible digital solutions in the field of
product traceability
Idem Option 3
37
Time horizon Immediate
Rather short for setting
guidance documents. The
revision of FIPD) requires
adoption of a new legislative
act ( minimum 1 year)
The obligations under this option would become effective by
the revision and entry into force of the revised GPSD.
The obligations under this option would need to be completed
by the revision and entry into force of the revised GPSD.
The obligations under this
option would need to be
completed by the revision and
entry into force of the revised
GPSD and the new market
surveillance Regulation.
Option 1. Enhanced enforcement: Improved implementation and enforcement of the
existing legal framework, without revision of the GPSD
This option does not require a legal revision of the GPSD, and would include:
a) Development of guidance on the safety of new technologies and exploring the use
of European standards to address new risks. The general safety requirement of
the GPSD already encompasses protection against all kinds of risks arising from
the product to the safety and health of persons. The guidance would clarify how
this includes not only mechanical, chemical, electrical risks etc. but also
cybersecurity and personal security threats that affect the safety of persons93
, and
other risks related to new technologies that potentially affect physical or mental
health94
. The standardisation procedure could be used to elaborate European
standards addressing safety requirements for consumer products concerning
certain new risks such as cybersecurity risks of new technologies.
b) More support and promotion of the Product Safety Pledge. To tackle the safety
issues related to online sales (including from third countries), the Pledge would be
updated and promoted through awareness campaigns, and other online
marketplaces would be encouraged to sign the Pledge. No legal requirements will
be introduced for online market places and no person responsible for products in
the EU will be available for non-harmonised products sold online.
c) Development of guidance on product recalls. The guidance would address current
deficiencies concerning the effectiveness and efficiency of recall procedures by
economic operators and market surveillance authorities, relying on the current
legislation. The guidance would concern e.g. the provision of more transparent
recall information to consumers, the use of customer data for direct notifications
and cooperation between different actors in the recall process.
d) Increased funding for joint market surveillance activities among Member States,
so that more coordinated actions of authorities across EU Member States could be
conducted, including the joint testing of consumer products. No legal changes in
the market surveillance rules, including on penalties, where a light approach, with
general provisions on penalties, as it is currently the case in the GPSD, would
continue. In this case, the deterrent effect of sanctions depends on the provisions
adopted by Member States.
e) Revision of the Food-imitating Products Directive to clarify its scope. The
provisions on the Food-imitating products would be kept in the FIPD with two
possible sub-options: (a) food-imitating products could be banned throughout the
Union per se per se and (b) application of a product safety risk-assessment on a
case by case basis to this category of products.
Option 2. Targeted revision of the GPSD (Directive or Regulation)
Option 2 would require a legal revision of the GPSD, which would remain a Directive or
become a Regulation. In case the new instrument is also a Directive, changes to the
GPSD would need to be transposed by Member States into national legislation. The
changes to the legal framework would include:
93
E.g. a smart watch for children, which does not causes a direct harm to the child wearing it, but lacks a
minimum level of security, so that it can be easily used as a tool to have access to the child and therefore endanger its
safety.
94
Mental health risks for consumers deriving e.g. from their interaction with humanoid AI systems.
39
a) Making explicit how the scope of the legal framework and its definitions apply to
risks posed by new technologies but without applying it to standalone software.
The definition of safety in the GPSD would be revised to clarify that the covered
risks arising from the product to the safety and physical/mental health of persons
include not only mechanical, chemical, electrical risks etc. but also cybersecurity
and personal security threats that affect the safety of persons, and other risks
related to new technologies that potentially affect health (similar to the guidance
that would be provided under Option 1). The definition of product in the GPSD
would not be changed, so that safety risks stemming from software are only
covered if the software is integrated in a product at the time of its placing on the
market (as is currently the case). There will be not specific provisions on or
references to software updates.
b) Adding requirements for online marketplaces by making most provisions of the
voluntary Product Safety Pledge legally binding. The Pledges’ commitments e.g.
to consult information on recalled/dangerous products available on RAPEX and
also from other sources; to take appropriate action in respect to recalled/dangerous
products, when they can be identified; to provide single contact points for EU
Member State authorities and to cooperate with them; to have an internal
mechanism for notice and take-down procedure for dangerous products and other
requirements would become legally binding for all online marketplaces targeting
EU consumers95
.
c) Adding requirements for enhancing the effectiveness of product recalls. Create
legal basis for economic operators to use any available customer contact details at
their disposal (e.g. obtained through loyalty schemes or online sales) to directly
notify the owners of recalled products (without the need of consumer consent).
Mandatory key elements would be defined that are to be included in every recall
notice (product description with a photograph, description of hazard, instructions
on what to do, description of remedy, contact details for queries). Prohibition to
use terms decreasing the perception of risk in recall notices (e.g.
‘voluntary/precautionary recall’ or "overheating" instead of fire). In case not all
affected consumers can be contacted directly, businesses would need to
disseminate recall announcements on their website, social media, newsletters,
retail outlets and other appropriate channels to ensure the widest possible reach.
d) Ensuring alignment with harmonised market surveillance rules while keeping
different legal instruments and simplifying standardisation procedures. The
market surveillance rules provided in the GPSD would be aligned with the
provisions in Regulation (EU) 2019/1020. Requirements for businesses would
reflect the current obligations under the GPSD, and include complementary
requirements in Regulation (EU) 2019/1020 (notably regarding the requirement of
an EU responsible economic operator to address the specific issue of direct online
95
All commitments under the Pledge could become legally binding under this option, except most probably
commitment 7 (training to sellers on compliance with EU product safety legislation, etc.) and 12 (exploring new
technologies and innovation to improve the detection of unsafe products).
40
imports from third countries by consumers) and other harmonisation legislation96
.
Traceability requirements would include the requirement to keep supply chain
records (to allow for one-up one-down traceability, i.e. the identification of
suppliers and clients, except final consumers). As a result, general requirements
for businesses and responsibilities and powers of market surveillance authorities
would be largely uniform for harmonised and non-harmonised consumer
products, including on penalties. Also, standardisation procedures at the
Commission level under the GPSD would be simplified.97
e) Integrating the provisions of the Food-imitating Products Directive into the
GPSD. The provisions of the FIPD would be integrated in the GPSD with the
same two substantive sub-options as in the Option 1.
Option 3. Full revision of the GPSD (Regulation)
Option 3 would repeal the Directive and ensure even application of its implementation
through the choice of a Regulation (i.e. it will be directly applicable in Member States).
This option would include all elements of Option 2 and, in addition:
a) Software related rules would be clarified. The GPSD would explain how software
can impact the safety of product and clarify responsibilities to ensure consumer
safety in such cases. Under this option the definition of product under the GPSD
could be adapted to cover the software updates. Specific provisions and
conditions could be elaborated for cases of software updates that affect the safety
of the product after a product is placed on the market, e.g. when the software
operates a substantial modification of the product impacting the risk it poses to
health and safety of consumers. It should be noted that under this option, the
revised GPSD would not regulate cybersecurity aspects in general, as that entails
different issues such as privacy or data protection; however, it would cover cases
when a lack of cybersecurity features can lead to a physical incident and hurt the
consumer, therefore not going beyond the area of consumer protection.
b) Making legally binding most provisions of the voluntary Product Safety Pledge
for online marketplaces (as in Option 2) and include new provisions for actors
across the online supply chain. These new provisions for actors across the online
supply chain would require them to provide all safety information online that is
also required to be provided with a product in 'brick-and-mortar' stores. Online
marketplaces would have a duty of care and they will be required to make sure
that third party sellers on their platform provide this information together with the
product offer (without being required to check the accuracy of the safety
information provided). This duty of care obligations would target online market
places and will be complementary to the obligations of sellers on the online
marketplaces (where the obligations of manufacturer, importer or distributor
would apply depending on the specific case), which would be particularly useful
in cases where the sellers are located outside the EU.
96
See also Decision No 768/2008/EC of the European Parliament and of the Council of 9 July 2008 on a
common framework for the marketing of products, which provides reference provisions, definitions and general
obligations for economic operators for harmonised products.
97
Changes could concern the involvement of Member States committees at various stages of the process. The
elaboration of the European Standards by the European Standardisation Organisations would not be affected.
41
c) Establish further mandatory requirements to enhance recall effectiveness. In
addition to all the elements of Option 2, the following would be introduced:
- Economic operators who offer product registration systems and loyalty
programmes for other purposes (e.g. marketing or technical support)
should offer consumers the possibility to register their contact details
specifically to receive possible safety notifications (personal information
collected for the purpose of product safety should be limited to the
necessary minimum and must not be used for marketing purposes);
- Possibility to set out further requirements for registration of specific
categories of products through delegated act;
- Binding template for recall notices to be set out through implementing act;
- Consumers’ right to an effective, cost-free and timely remedy(repair,
replacement or refund);
- Less burdensome recall procedure for consumers (returning a product
should not incur any financial costs, non-portable items to be collected by
the operator); Binding requirements for businesses to register voluntary
recalls in an EU public database and to monitor recall effectiveness; MSAs
would have the possibility to request monitoring data from economic
operators and decide if the case can be closed.
d) Give stronger enforcement powers to Member State authorities (for example on
penalties) and establish arbitration mechanism in case Member States have
diverging product safety risk assessments. Building on Option 2, general
requirements for businesses and responsibilities of market surveillance authorities
would be largely uniform for harmonised and non-harmonised consumer
products. However, under Option 3 stronger enforcement rules would be
incorporated:
- Penalties: The provisions on penalties would be more clearly defined in a
way to ensure a sufficient deterrent effect, while increasing the sanctioning
powers of Member States.
- Arbitration mechanism: In case Member States have diverging
assessments of the risk posed by a notified product, a mechanism could be
triggered where either a group of Member States or the Commission are
called to arbitrate.
- Traceability: This option would also create a possibility to set further
requirements for traceability systems through delegated acts, for example
regarding chemicals in childcare articles.
- Finally, some further improvements to the market surveillance could be
envisaged based on the feedback from the stakeholders and study results.
e) Integrating the provisions of the Food-imitating Products Directive into the
GPSD with the same two substantive sub-options as in the Option 2.
Option 4. New Regulation merging market surveillance provisions of GPSD and
Regulation (EU) 2019/1020
This option would provide for a new legal instrument including all elements described
under Option 3 and also merging the market surveillance provisions of the GPSD and
42
Regulation (EU) 2019/1020 on the market surveillance and compliance of products into
one new Regulation on market surveillance, so that one single set of rules would apply to
harmonised and non-harmonised products.
5.3. Options discarded at an early stage
Even if options 0 and 1 have received little support by stakeholders they have not been
disregarded and the impacts of all options are assessed.
Initial policy options presented in the Roadmap/Inception Impact Assessment took into
account the results of the GPSD implementation study, which – based on a broad
consultation process – had elaborated key shortcomings of the current legal framework
and stakeholder suggestions for improvements. In the course of the GPSD study, the
completeness of the current set of policy options was validated and no further policy
options for consideration were identified. Also the stakeholders did not raise any other
new real alternatives during the consultation process.
A potential further policy option, discarded at an early stage, is the complete repeal of
the current GPSD. The Evaluation of the GPSD (see Annex 5) concludes that, although
there is a need for specific improvements and simplification, the GPSD is generally
relevant, effective, efficient and coherent, and has EU added value.
This report builds also on the conclusions of the Impact assessment report prepared for
the previous proposal to revise the GPSD tabled in 2013. In particular, some of the
disregarded or eliminated options after analysis of impacts in 2013 were disregarded also
in the current report, e.g. centralisation of market surveillance at EU level, direct
applicability of ad-hoc safety requirements, abolition of the general product safety
requirement.
Also, the introduction of the “Made in” clause, mandatory country of origin labelling for
products (as it was proposed in the 2013 Package), has been disregarded at an early stage
following the technical study the Commission conducted in September 2014,98
assessing
the costs and benefits of the proposed mandatory country of origin labelling for a number
of product categories. The study concluded that there is little evidence of possible
positive impacts of this clause on product traceability and safety for any of the product
groups. Further reinforced traceability requirements will however be analysed in Option 2
and 3 beyond the “Made in” clause.
Some alternatives were considered to certain substantive measures presented in the
intervention logic and have been disregarded:
Concerning new technologies, it was considered whether the sectorial legislation could
cover those new challenges, such as for example the product safety risks linked to
cybersecurity. While particular actions at sectorial level might still be needed, it appeared
that gaps would remain unless a full safety net is ensured, as it is provided under GPSD
in its function of lex generalis.
Regarding setting the requirements for online marketplaces in the product safety area,
an alternative option would have been to define the precise obligations in the proposal for
98
VVA Europe 2015 https://ec.europa.eu/info/sites/info/files/indication_origin_study_2015_en.pdf
43
the DSA. The adopted proposal for the DSA has retained nevertheless a more horizontal
approach proposing the general obligations for all types of illegal content, leaving the
definition of specific product safety obligations of online marketplaces to the product
safety legislation.
When it comes to the traceability requirements, an alternative option could have been
to impose higher traceability requirements for all products. This option has been
considered and disregarded because of lack of proportionality.
An important objective of this initiative is also to create a sufficient deterrent effect to
incentivise the economic operators to comply with the product safety requirements,
which is particularly important in a context where all the products cannot be controlled in
view of their huge volumes. The options 3 and 4 propose to increase the deterrent effect
of the GPSD by reinforcing the provisions on penalties by setting some harmonised
criteria and a minimum threshold for the maximum amount of penalties. A stronger
approach on penalties could have been to set up a precise list of infringements and
corresponding minimum and maximum amounts for penalties. This option has been
disregarded since interfering with Member States’ competences.
6. WHAT ARE THE IMPACTS OF THE POLICY OPTIONS?
The relevant stakeholder affected by the initiative are the economic operators
(manufacturers, importers, distributors and online market places), market surveillance
authorities (MSAs) and EU consumers. The main costs entailed by complying with the
Directive can be classified by stakeholder group as follows:
- Substantive compliance costs for economic operators: costs arising for
manufacturers to ensure compliance with the product safety requirement (setting up
product safety processes, testing, recalls, etc.) and other economic operators’
obligations under GPSD (e.g. for distributors), and possible purchasing of standards.
- Administrative costs for economic operators to comply with obligations to provide
safety information to national authorities on request.
- Enforcement costs for MSAs: costs arising from market surveillance activities
(implementation, enforcement and monitoring), withdrawal of unsafe products from
the market and coordination - internally between MSAs within one country and
externally within the Consumer Safety Network and via the Safety Gate/RAPEX.
Enforcement and coordination costs for the Commission.
- Direct regulatory costs for EU consumers via possible increase of prices or lower
choice of non-harmonised products.
- Direct regulatory benefits: improved health and safety of EU consumers and the
improved environment (decrease of products with both safety and environmental
risks, e.g. due to presence of dangerous chemicals) leads to improved well-being.
Market efficiency improvements in the form of better quality of non-harmonised
products and better information about product safety (e.g. about recalled products)
increase trust of consumers in the market and increased purchasing. Alignment of
market surveillance rules for harmonised and non-harmonised products create a more
level playing field and have therefore positive effect on competition.
- Indirect benefits: the decreased costs of injuries has positive impact on national
health and consumers’ budgets; positive effects on fundamental rights by improving
consumer protection, including the protection of vulnerable consumers (children, the
44
elderly) and on innovation via increased legal certainty regarding the application of
consumer product safety rules to new technologies.
- Cost savings linked to the simplification of procedures (e.g. standardisation),
reduction of regulatory costs for businesses and administrative burdens for MSAs by
alignment of market surveillance rules for harmonised and non-harmonised products,
integration of safety rules for food-imitating products with the rules for non-
harmonised products and choice of a directly applicable legal instrument
(Regulation).
6.1. Impacts of Option 1
Effectiveness in achieving the specific objectives
Table 7: Assessment of Option 1 related to the specific objectives
99
Objectives Areas Assessment Impacts
New
technologies
Certainty regarding coverage of new risks Option will to some extent contribute to certainty regarding coverage
of new risks, without being legally binding. Implementation
differences in Member States may remain.
neutral
/ +
Certainty regarding coverage of software Option will not providing clarity of GPSD’s application to software.
Online sales Safety of products sold on online platforms It is unlikely that safety risks for EU consumers due to products sold
on online platforms will be significantly reduced. neutral
Information of consumers on essential
safety aspects
No change to the current situation.
Recall
effectiveness
Reaching out to consumers affected by
recalls Option will to some extent contribute to certainty regarding recall
procedures, without, however further regulating and therefore
addressing the underlying reasons for limited recall effectiveness. neutral
Information provided in recall notices
Monitoring of recall effectiveness
Remedies for consumers affected by recalls No change to the current situation.
Market
surveillance
Alignment of market surveillance
framework for harmonised and non-
harmonised consumer products
Limited increase of EU funding100
may enhance enforcement, but no
change to the current fragmentation of legal framework for market
surveillance.
neutral
Deterrence effect
No change to the current situation.
Diverging risk assessments by MSAs
Simplification of standardisation
procedures
Food-imitating
products
Addressing risks of food-imitating products Clarify the regime for the food-imitating products : (a) full ban or (b)
risk-assessment approach +
99
Note: Magnitude of impact as compared with the baseline scenario:
neutral = no significant difference to baseline situation;
+ = positive impact compared to baseline;
++ = significant positive impact compared to baseline.
An indication of neutral/+ or +/++ indicates an intermediate assessment, depending on implementation details and/or circumstances.
Costs are indicated as either neutral (no additional costs compared to baseline), or with an indication of the expected increase in EUR
terms, again compared to the baseline situation.
100
Under the current proposal for next Multiannual Financial Framework and the Single Market Programme the yearly
amounts foreseen for coordinated market surveillance actions are only slightly higher (EUR 2,8 million) than the
spending on these activities in the previous years (EUR 2,4 millions)
45
The GPSD Study showed that the overall average assessment of the effectiveness of
Option 1 in addressing the five challenges mirroring the five specific policy objectives
across all respondents and stakeholder groups was 2.9 on a scale of 1 to 5.
Administrative simplification
Guidance provided under Option 1 could to some extent reduce regulatory complexity
and uncertainty regarding the coverage by the GPSD of risks posed by new technologies,
as well as regarding applicable procedures for recalls. Also, complementary measures in
the standardisation field to address safety requirements for consumer products concerning
certain new risks posed by new technologies could have a similar effect. However, as
these guidelines and standards would not be legally binding, this reduction can be
expected to be minor. In addition, legal uncertainty regarding the application of the
GPSD to software will remain. Therefore Option 1 will not significantly reduce the
regulatory complexity and burdens for businesses. The simplification potential is
therefore very limited, stemming mainly from the clarifications provided under the non-
legally binding guidance documents and the revised FIPD.
Responsibilities and powers of market surveillance authorities will remain different for
harmonised and non-harmonised consumer products, and related administrative burdens
for some authorities will continue. In the survey conducted in the GPSD Study, 16% of
authorities reported they currently experience additional costs due to these differences.
Also, administrative burdens on Member States in the field of standardisation would not
be reduced.
Option 1 does not include any additional administrative requirements for specific types of
operators. Only very low burdens are expected for businesses from getting familiar with
new guidance documents.
Economic impacts
The GPSD Study showed that all stakeholders estimate the benefits of Option 1 on a low
level (see Annex 12): companies/business associations estimated benefits ‘moderate’ (3
in a scale of 5 ) and MSAs (2.6 in a scale of 5), and other stakeholders highlight even
lower (‘minor’ benefits, 2 in a scale of 5). Businesses assessed the benefits to be ‘minor’
when it comes to increased business revenue. The assessment of other stakeholders is
particularly low with respect to the reduction of legal complexity and improved supply
chain management due to improved traceability of products (values of 1.8 and 1.9
respectively in a scale of 5).
Concerning the costs incurred under Option 1, the GPSD Study indicates that
implementing this option would not increase companies’ recurrent regulatory
compliance costs (staff costs) or other additional recurrent costs, neither for
manufacturers or distributors. Several business respondents indicated that nothing
substantial would change with the implementation of Option 1 compared to the status
quo, even if better guidance documents could potentially improve clarity and legal
certainty and, as a result, create some cost savings.
Option 1 should create minor additional one-off costs for businesses related to getting
familiar with new guidance provided at EU level. However, the quantitative estimates
provided by company respondents in the GPSD Study confirm that no significant
additional one-off cost are expected at the EU aggregate level. In conclusion, the GPSD
46
Study indicates that the implementation of Option 1 should not change one-off and
recurrent costs of EU businesses.
No significant firm-level impacts are to be expected due to the implementation of Option
1 for specific types of operators, be it SMEs or specific operators such as online traders.
An exception are businesses that are manufacturing or distributing food-imitating
products. Currently, as explained in the Evaluation, the FIPD is applied differently
across EU countries, as some MSAs interpret the FIPD as a per se prohibition of food-
imitating products while others do a case-by-case risk assessment of the safety of
product. To analyse the impact of the provisions on the food-imitating products, two
different sub-options have been considered: (a) a full ban of food-imitating products;
or (b) provisions that would include the food-imitating aspect (and possibly child-
appealing in general) as an element for the assessment of the risk of products and
require a case-by-case risk assessment, as for other consumer products. We can expect
that a targeted revision to better clarify the specific requirements of the FIPD would give
manufacturers and distributors more legal certainty in both sub-options. As both
manufacturers and distributors already have to comply with the current FIPD, we do not
expect additional costs from a revision that merely aims at providing greater clarity and
legal certainty respectively. A greater level-playing field regarding the implementation
and enforcement of the FIPD in the EU could lead to minor cost savings on the side of
manufacturers and distributors of food-imitating products. The GPSD Study concluded
that the negative economic impact of a full ban of food-imitating products would likely
be minor in a broader economic perspective, since the number of these products is
limited101
. At the same time the alternative option (b) is more coherent with the
current risk assessment approach in the GPSD which has already been applied to
food-imitating products by a number of MSAs: restriction of the free circulation of a
given food-imitating product would be based on the assessment of the particular
product’s risks and action would be taken according to this assessment. Applying a risk
assessment would enable a proportionate corrective measure to be taken.
No macroeconomic effects with significant impacts of the implementation of Option 1
on the internal market or trade are expected, since measures under Option 1 are voluntary
in nature and are largely cost neutral. Implementation differences in Member States are
likely to remain at least partially since the additional guidance provided under Option 1
would not be legally binding. The results of the consultation conducted for the GPSD
Study show that stakeholders evaluate the benefits on the internal market and trade minor
to moderate102
. Significant impacts on competition and innovation are also unlikely, as
the benefits of guidance in this respect are limited and all measures are quasi cost-neutral
for businesses (except in the area of food-imitating products, where a slight benefit is
possible due to increased legal clarity). Some benefits are expected due to slightly
increased funding of joint market surveillance activities among Member States.
101
The Evaluation shows that the food-imitating products represent only a very small proportion of the notifications in
the Safety Gate/RAPEX
102
Businesses and business associations assess the potential benefits from better functioning of the EU internal market
and more level-playing field among businesses as ‘moderate’. The deterrent effect on rogue traders is considered
‘minor’ to ‘moderate’, while the benefit of a better access to non-EU/EEA markets is assessed to be ‘minor’. On
average, MSAs expect lower benefits than businesses. When it comes to other stakeholders, their assessment of Option
1 is much lower at an average of only 1.7 (i.e. below ‘minor’).
47
In terms of impact on consumers, the GPSD Study shows that the benefits for consumers
are judged by stakeholders generally minor to moderate in terms of reduced occurrence of
unsafe products, reduced number of accidents and injuries caused by unsafe products and
increased consumer trust. Also since Option 1 would, overall, not result in increasing
product safety-related costs for economic operators, it is not expected to create any
impact on the prices for consumer products in the EU. Due to the voluntary character of
measures under Option 1, there should not be any impact on consumer choice.
Concerning the impact on consumers of the sub-options for the food-imitating products,
both sub-options would lead to the same level of consumer protection against unsafe
products. In particular, lifting the ban in those Member States applying it nowadays,
would not lower the consumer protection if the risk-assessment is applied. Indeed, by
analogy with the risk assessment methodology under the GPSD, the evaluation of the risk
of the product would take into account its food-imitating aspect so the risk for the
consumer would be properly assessed and if the product is found to be unsafe, its placing
on the market would be prohibited.
The guidance provided under Option 1 and the additional funding of coordinated market
surveillance activities could slightly improve the enforcement of the GPSD, with related
benefits for consumers. However, since online market surveillance will not be
substantially improved, consumers would continue to incur detriment, even if the
voluntary safety Pledge would be further promoted. With an increasing role of online
platforms in the EU retail sector in the future, amplified by the COVID 19 crisis, costs for
society due to unsafe products entering the market through online channels from third
countries could increase, although this will also depend whether other measures are taken
at EU level, including in the framework of the new DSA. Option 1 is therefore not
expected to increase the level of consumer protection, including vulnerable consumer
groups such as children and the elderly.
Impacts on Member States
Option 1 is not expected to provide significant benefits for MSAs, except a reduction in
uncertainty about GPSD interpretation thanks to the provision of guidance and the
possible additional funding for joint market surveillance activities. However, it needs
to be considered that the increase of this funding will be limited, as the budget foreseen
for these activities in the Multiannual Financial Framework and the Single Market
Programme will only be around EUR 21 million, which amounts to a very small
increase of yearly average budget for these coordinated activities (from EUR 2.4
million in 2009-2020 to EUR 2.8 million per year in 2021-2027).
Based on the MSAs’ survey, the GPSD Study found that recurrent costs for MSAs should
remain the same under Option 1, compared to the baseline situation, and one-off costs
would be very low (costs resulting from the development of new guidance documents,
and, potentially, the set-up of technical capacities for carrying out market surveillance
activities related to new risks).
The proposed measures would not be expected to have other effects on Member States
since no modifications of market surveillance mechanisms are proposed.
In case the risk-assessment sub-option were chosen for the treatment of food-imitating
products, potential effects could be observed on Member States which have been
applying full ban of these products until now. These Member States would need to make
48
a risk-assessment of these products, but this change should not represent specifically
higher costs for these countries since the number of food-imitating products is very low
and they already apply the risk-assessment approach for all other products103
.
Social impacts, impacts on fundamental rights, environmental impacts
Due to their limited scope and voluntary character, the measures implemented under
Option 1 would not have significant social or environmental impacts, or impacts on
fundamental rights. However, the GPSD Study concludes that if retained, a ban on
food-imitating products from the EU market would have a negative impact on the
freedom to conduct a business as defined under Article 52 of the Charter of
Fundamental Rights of the European Union (‘Charter’), while the GPSD Study,
based on the feedback from Member States, could not identify evidence to prove the
intended benefits (better protection of children) to confirm its proportionality.
Indeed, food-imitating products are not intrinsically dangerous. They can be if they are so
similar to foodstuff that they can be confused with food, and if such confusion could pose
a risk (notably the risk to choke or to be chemically poisoned) if consumers would ingest
such products. Banning all food-imitating products would mean banning also those food-
imitating products that are not dangerous (for instance, those that imitate but cannot be
confused with foodstuff, or those that do not present any risk, notably because they are
not a small part or no small part could be detached from it). Article 52 of the Charter sets
out that limitations on the exercise of the rights and freedoms recognised by the
Charter may be made only if they respect the principle of proportionality, are
necessary and genuinely meet objectives of general interest recognised by the
Union or the need to protect the rights and freedoms of others. A general ban of
food-imitating products would result in banning some non-dangerous products, which
would be an unjustified and non-proportional restriction of the freedom to conduct a
business.
6.2. Impacts of Option 2
Effectiveness in achieving the specific objectives
Table 8: Summary assessment of Option 2, compared to baseline situation
Objectives Areas Assessment Impact
New technologies Certainty regarding coverage of new
risks
Legally binding clarifications will avoid uncertainty. Depending on the
choice of instrument, implementation differences in Member States may
remain
+
Certainty regarding coverage of
software
No change to the current situation
Online sales Safety of products sold on online
platforms
Safety risks for EU consumers due to products sold on online platforms
partly reduced, with the effectiveness also depending on other factors neutral
/ +
Information to consumers on essential
safety aspects
No change to the current situation
Recall effectiveness Reaching out to consumers affected Change can be expected to facilitate the use of available customer data, +
103
We also note that the risk-assessment of the food-imitating products appears to be relatively simple in practice: e.g.
the food-imitating aspect renders highly probable that the product would be put in mouth by children so if small parts
of the products can detach easily, the product will likely present a serious risk and prohibited from being placed on the
market.
49
by recalls and avoid data protection concerns
Information provided in recall notices Improvement in the information provided to consumers is expected to be
achieved
Monitoring of recall effectiveness No change to the current situation
Remedies for consumers affected by
recalls
No change to the current situation
Market surveillance Alignment of market surveillance
framework for harmonised/ non-
harmonised consumer products
Largely uniform general requirements for businesses and responsibilities
and powers of market surveillance authorities for harmonised and non-
harmonised consumer products expected to be achieved
++
Deterrence effect Largely unchanged situation in terms of deterrence of placing unsafe
products on the market
Diverging risk assessments by MSAs No change to the current situation
Simplification of standardisation
procedures
Simplification of standardisation procedures is expected to be achieved
Food-imitating
products
Addressing risks of food-imitating
products
Clarify the regime for the food-imitating products: (a) full ban or (b) risk-
assessment approach +
In the GPSD Study’s survey, all stakeholder groups considered that Option 2 addressed
all challenges at least moderately well. Overall, the average assessment across all
respondents and stakeholder groups was 3.4 out of 5.
Administrative simplification
Option 2 is expected to reduce regulatory complexity and uncertainty, and thereby to
reduce administrative burdens for businesses. Key clarifications regarding the coverage
of new risks posed by new technologies will be provided in the new legal instrument. As
these will be legally binding, the regulatory complexity reduction can be expected to
be more significant than under Option 1. Also, general requirements for businesses
and responsibilities and powers of market surveillance authorities would be largely
uniform for harmonised and non-harmonised consumer products, which is likely to
contribute to reduce regulatory complexity and thereby to reduce administrative
burdens for businesses.
However, depending on the choice of instrument, implementation and interpretation
differences between Member States may remain (if a directive were chosen). The
legal form chosen under this option 2 matters for a certain number of issues and in
particular simplification aspects. Contrary to a directive, a regulation is directly
applicable across the Union; there is therefore no need for Member States to transpose
EU legislation into national law, which can lead to some implementation differences as
analysed in the problem definition. With a regulation, national differences regarding the
date and/or manner of transposition would be eliminated, which would facilitate
consistent enforcement and level-playing field in the internal market. A regulation
ensures better that legal requirements are implemented at the same time throughout the
Union; it also better achieves streamlining of terminology, important for defining the
scope of the legislation, thereby reducing administrative burdens and legal ambiguities.
The choice of Regulation also allows to better deliver on the objective to ensure
coherence with the market surveillance legislative framework for harmonised products,
where the applicable legal instrument is also a Regulation (Regulation (EU) 2019/1020).
50
In addition, as lack of clarity regarding the application to software will remain,
uncertainty in this respect will likely not be reduced. At the same time, Option 2 would
provide limited additional requirements for specific operators, such as requirements
for online marketplaces resulting from making mandatory many provisions of the
Pledge, and requirements regarding mandatory key elements that are to be included in
recall notices.
Finally, ensuring alignment with harmonised market surveillance rules will reduce
administrative burdens on MSAs. Similarly, simplified standardisation procedures at the
Commission’s level should lead to savings for MSAs and the Commission.
Consequently, simplifications would be achieved under this option by aligning market
surveillance procedures for harmonised and non-harmonised products, by simplifying the
standardisation procedure and by merging the rules on food imitating-products with the
ones for non-harmonised products into one single legal instrument and repealing the
FIPD. Also the increase of legal certainty regarding the application of consumer product
safety rules to new technologies will likely reduce the regulatory costs for businesses
producing new technology products. If the new act will be a Regulation the regulatory
burden will decrease even more.
Economic impacts
According to the GPSD Study survey, businesses expect that implementing Option 2
would increase companies’ recurrent regulatory compliance costs to some extent, as
well as the additional one-off cost linked to the implementation of measures (see Table
9). Total costs for businesses in the EU27 in the first year of implementation are
estimated at EUR 36.9 million, equivalent to 0.004% of turnover of EU companies for
manufacturing, wholesale and retail of non-harmonised consumer products. They would
fall in subsequent years down to EUR 29.6 million.
Table 9: Changes in EU companies’ costs within and after the first year of implementation of Option
2, in million EUR
Source: The survey conducted in the context of the GPSD Study
From 0 to 49
employees
50 – 249
employees
250 or more
employees
Total
First year of implementation (total of one-off and recurrent costs)
Total manufacturing sectors 4.8 6.1 9.9 20.7
Total wholesale sectors 2.5 1.7 2.6 6.9
Total retail sectors 5.0 0.9 3.5 9.4
Total additional costs 12.2 8.8 15.9 36.9
Subsequent years (recurrent costs only)
Total manufacturing sectors 4.3 5.5 8.9 18.6
Total wholesale sectors 1.7 1.2 1.8 4.7
Total retail sectors 3.4 0.6 2.4 6.4
Total additional costs 9.3 7.3 13.0 29.6
51
Estimated benefits for businesses linked to cost savings, that are currently caused by
differences in the national implementation of the GPSD and would be partly solved if
the new instrument is a Regulation under the Option 2, would amount to EUR 59 million
annually104
, of which EUR 34 million would be saved by EU SMEs and 26 million EUR
saved by EU large businesses respectively, compared to the baseline.
The GPSD Study’s survey showed that MSAs expect considerably more benefits for
businesses from the implementation of Option 2 (average assessment of 3.3 out of 5,
above moderate; see Annex 12) than businesses/business associations themselves (2.6)
and other stakeholders (2.7)105
, both below moderate. All respondent groups in the
survey assigned similar values to the benefits resulting from better information on unsafe
products/measures taken by authorities provided through Safety Gate/RAPEX.
Concerning impact on SMEs, they generally estimate that a revision of the product safety
requirements of the GPSD according to Option 2 would bring a variety of at least
‘minor’ to ‘moderate’ benefits106
. At the same time Option 2 would impose additional
adjustment (e.g. familiarisation cost) as well as compliance costs on SMEs107
, in
particular for manufacturers. Table 9 shows that SMEs would likely face relative higher
compliance costs than large companies from the implementation of the proposed policy
measures.
Even though the relative cost increases are generally higher for SMEs, the net impact on
SMEs overall costs depends on the benefits that can result from a revised GPSD
aligned to the market surveillance rules and traceability requirements in Regulation (EU)
2019/1020. We expect the SMEs would save EUR 34 million of the costs that
currently arise from inconsistencies in the implementation and enforcement of the
GPSD across the EU. Taking into consideration these benefits and the fact that the
changes in SMEs’ costs from Option 2 are very small, we expect that the overall net
effect from Option 2 on SMEs’ costs is rather low and therefore unlikely to affect
SMEs’ operations.
As regards the impact on online marketplaces, the additional costs from Option 2
making most of the Product Safety Pledge obligations binding, would be minor for those
companies that already signed the Pledge. By contrast, non-signatory marketplaces
would likely face additional compliance costs108
. In particular, some stakeholders were
104
The baseline costs linked to the different implementation of the GPSD are estimated to amount to 119 million EUR
annually (see section 7). As Options 3, 4 and possible 2 foresee to recast the GPSD as a Regulation, implementation
differences would be avoided (due to the direct applicability of the new regulation in Member States), even if some
differences in the national interpretation of rules may remain. Accordingly, we assume a 50% reduction of businesses’
additional costs in this respect in case of choice of Regulation as legal instrument.
105
Indeed other stakeholders are mainly consumer organisations, which showed a clear preference for Option 3 and 4.
106
Significant benefits due to improved quality/lifecycle of products and a deterrent effect on rogue traders, relatively
strong benefits are increased consumer trust, better supply chain management due to improved traceability of products
and better access to the market in non-EU/EEA. These areas are seen as benefits that SMEs assess to be ‘moderate’ to
‘significant’. This is also the case for lower operational risks for businesses and easier compliance with product safety
requirements. By contrast, SMEs considered several benefits to be less than ‘moderate’, including a more level playing
field among businesses and greater legal certainty.
107
This is particularly the case for SMEs that (voluntarily) decide to install and operate customer registration systems.
Similarly, mandatory elements for product recalls (product description with a photograph, description of risk,
instructions on what to do, link to a recall website and free phone number or online service for queries) would increase
the cost of SMEs that have put unsafe consumer products to the market.
108
Two online platforms provided quantitative estimates for the expected impact on recurrent costs, stating that their
companies’ overall consumer product safety-related costs would increase by 10%.
52
concerned that these compliance costs might specifically affect small companies and
create a deterrent effect on new market entrants, with potential negative effects on
competition between marketplaces. Such costs should however be low, as many
obligations under the Pledge that would be implemented relate to a ‘notice and action’
procedure specifically tailored for product safety (i.e. a reactive approach, where
marketplaces only monitor the information provided by the MSAs about unsafe products
but do not need to monitor the safety of products they list by themselves).
Concerning the impact on producers of food-imitating products, the impacts are the
same as in Option 1. Concerning the macroeconomic effects, the interviewed
stakeholders find that the implementation of Option 2 would have a minor to moderately
positive impact on the functioning of the EU’s internal market and international trade109
:
measures to clarify the coverage of new risks by new technologies in a revised legal
instrument, as foreseen under Option 2, can address a part of the uncertainties linked to
the implementation of some of the key concepts in the GPSD and new technology
products, whereby uncertainties would remain with respect to the actual effectiveness of
such measures, but also with respect to the coverage of software. It is possible that
Member States could resort to national measures in this respect, which would create an
obstacle to the free movement of goods or services and lead to an uneven level playing
field for businesses in the future. Still, benefits can be expected from the clarification of
safety risks stemming from new technologies, recall procedures and more coordinated
actions by MSAs. Reduced legal complexity and uncertainty could reduce companies’
administrative burdens to some extent, which could have a moderate positive impact on
the functioning of the EU’s internal market and international trade.
Similar to Option 1, the impacts from Option 2 on EU companies’ competitiveness are
expected to be relatively small as companies’ additional costs incurred by Option 2 in the
first year of implementation would represent only 0,004% of their annual turnover, we do
not expect significant impacts on competition for EU businesses, neither for competition
within the Single Market nor with regard to non-EU competitors.
As concerns innovation, due to the limited impact on companies’ compliance costs no
significant impacts on EU companies’ overall innovative capacities are expected, i.e.
higher budgets resulting from savings in compliance costs that translate to expanded
research and business development activities. On the other hand, new regulatory
requirements for online platforms might result in less innovation in some online platform
business models over time, depending on the extent to which new requirements lead to
additional costs, which appear, however, to be limited under Option 2.
In term of impact on consumers, the alignment with the provisions of Regulation (EU)
2019/1020 on market surveillance and clarifications provided in the new legal instrument
could improve the enforcement of the GPSD, with related benefits for consumers. Also
concerning online sales, making relevant provisions of the Pledge binding should lead to
better monitoring of unsafe products by marketplaces, as there would be a regular
exchange with market surveillance authorities. Option 2 would therefore be expected to
increase the level of protection of EU consumers to some extent, by reducing the
109
MSAs are on average the most positive stakeholder group about the benefits that would result from the
implementation of Option 2 with an average of 3.8 (i.e. seeing close to ‘significant’ benefits). By contrast, the averages
for both companies/business associations and other stakeholders are slightly lower (3.2 and 3.4 respectively, i.e.
between ‘moderate’ and ‘significant’ benefits).
53
incidence of unsafe products in online sales channel. The GPSD Study estimated that
consumers would therefore benefit in terms of reduced consumer detriment based on
the value of unsafe products110
. This detriment reduction is expected to amount to
approximately EUR 333 million in the first year of implementation, increasing to
approximately EUR 1.03 billion over the next decade. As regards improving the recall
effectiveness and therefore reducing number of unsafe products remaining in hands of
consumers would also bring benefits: the GPSD Study estimates under a scenario of
somewhat improved recall effectiveness as expected under Option 2, the consumer
detriment in the EU to be reduced by EUR 205 million in Option 2 compared to the
baseline. This impact on consumers could be also relevant for specific vulnerable
consumer groups such as children, and the elderly, as they are often more affected by
unsafe products.
The survey of the GPSD Study shows that the benefits for consumers are judged by
companies/business associations and MSAs as ‘moderate’ (average values of 3.1 for
companies/business associations and 3.4 for MSAs in a scale of 5). Benefits include a
reduced occurrence of unsafe products and a reduced number of injuries caused by them,
as well as a resulting increase in consumer trust. Other stakeholders are less positive
(below ‘moderate’, average value of 2.6 out of 5). As the implementation of Option 2
would only result in minor increases of consumer product safety-related costs for EU
companies, the impacts from Option 2 on prices of consumer products in the EU are
expected to be negligible. None of the measures considered under Option 2 would be
expected to have a significant impact on consumer choice in the EU.
Concerning the impact on consumers of the sub-options for the food-imitating products,
we consider that both sub-options lead to the same level of consumer protection against
unsafe products, as analysed under Option 1.
Impacts on Member States
The GPSD Study reports that MSAs stated that Option 2 is expected to be more suitable
than Option 1 to improve the current legal framework managing the risk of unsafe
products being placed in the EU market, but the exact benefits would depend on its actual
implementation. Generally, a more uniform framework for harmonised and non-
harmonised consumer products, a simplification of standardisation procedures and a
clarification of rules regarding product recalls foreseen under Option 2 would, over time,
lead to a reduction of administrative burdens for MSAs. The GPSD Study estimates that,
if a Regulation is chosen as legal instrument, benefits for MSAs arising from the
alignment of the provisions for market surveillance of harmonised and non-harmonised
products would lead to improvements in efficiency of market surveillance, and related
cost savings, which are estimated at EUR 0.7 million per year across the EU111
.
Under Option 2, MSAs could be impacted by a broadening of market surveillance
responsibilities (e.g. from modified definitions as regards risks posed by new
110
Consumer detriment linked to the value of unsafe products, calculated on the basis of the purchase price of unsafe
products.
111
See baseline description. The proposed measures under Options 2 (if Regulation), 3 and 4 would fully align
provisions for the market surveillance of harmonised and non-harmonised consumer products so that the cost burden
estimated in the baseline as EUR 0.7 million will be reduced accordingly.
54
technologies). New responsibilities are generally reflected by a greater need for internal
and external resources respectively.
The GPSD Study estimates the total additional recurrent costs for MSAs in EU27 of
approx. EUR 6.7 million annually112
. Concerning one-off costs, the few cost estimates
that were provided by MSA respondents indicate that the one-off adaptation and
implementation costs are considered to be moderate.
Option 2 would align the enforcement powers of MSAs regarding non-harmonised
products with their powers for harmonised products. Thereby, specific gaps such as legal
difficulties to conduct mystery shopping for authorities in some Member States would be
addressed. However, the deterrence effect on rogue traders would not be increased, as
enforcement powers would not be further strengthened through penalties and sanctions.
Likewise, no arbitration mechanism would be created for cases of divergences in the
product safety risk assessment between MSAs, and there would be a continued reliance
on informal approaches in case risk assessments of MSAs diverge to harmonise the
treatment of products on the Single Market.
Social impacts
The implementation of Option 2 is expected to potentially have some positive social
impacts with regard to public health and safety and health systems. The clarification
of covered risks, mandatory obligations for online platforms (in line with the Pledge) and
the alignment with the provisions of Regulation (EU) 2019/1020 will, to some extent,
improve market surveillance and enforcement. To the extent that the number of unsafe
products on the market is somewhat reduced by these measures in the mid- to long term,
this could potentially lead to a lower number of injury cases caused by consumer
products in need of medical attention or hospitalisation, hence decreasing public health
expenditure for the treatment of product related injuries. However, due to the limited
amount of measures taken under Option 2 that could effectively reduce consumer injury-
related detriment in the EU, any impact on health systems would be expected to be
considerably more uncertain and smaller in size than under Option 3.The current cost
of health care utilisation for product-related injuries in the EU are estimated by the
GPSD Study to approximately EUR 6.7 billion per year, with hospitalisation accounting
for the larger part of the total health care costs at about EUR 6.1 billion. A revised GPSD
may contribute thereby to lowering these health care costs for the society.
Environmental impacts
The GPSD Study showed that, while authorities see ‘moderate’ benefits regarding
improved lifecycle/quality of products and a higher level of the protection of the
environment due to the reduction of unsafe products that also have environmental
impacts, companies/business associations and other stakeholders only see between
‘minor’ and ‘moderate’ benefits. The implementation of Option 2 is likely to have
positive environmental impacts, to the extent that it clarifies the application of the
general safety requirement to products containing environmentally harmful substances
that also pose a risk to human health and safety. Already today around 25% of the
112
It should be noted that the actual percentage changes would differ for individual MSAs due to different national
institutional market surveillance systems and organisational characteristics, e.g. the degree of centralisation, MSAs’
product coverage and the actual assignment of new competences and enforcement requirements.
55
products notified in Safety Gate/RAPEX presented a chemical substance risk with
adverse health effects to consumers. The relevant chemicals were often also harmful to
the environment (e.g. lead and mercury).
Impacts on fundamental rights
Option 2 is expected to improve consumer safety to some extent and also to reduce
product-related environmental risks (see above). The implementation of a revised GPSD
according to Option 2 shall hence have a positive impact and ensure a somewhat
higher level of consumer protection and a higher level of environmental protection
in line with the Charter. However, there would be no right to effective, cost-free remedies
for consumers that own a recalled product, which would limit the positive impact of this
option.
At the same time Option 2 imposes additional requirements for businesses, but these do
not affect the fundamental freedom to conduct a business as the former are necessary
to pursue the general European Union interest of increasing consumer protection and are
proportional to the aim pursued, given that the resulting compliance costs are estimated to
be very limited compared to the businesses’ turnover. The negative effect of a potential
ban of food-imitating products would be the same as developed under Option 1.
6.3. Impacts of Option 3
Effectiveness in achieving the specific objectives
Table 10: Summary assessment of Option 3, compared to baseline situation
Objectives Areas Assessment Impact
New
technologies
Certainty regarding coverage of new
risks
Legally binding clarifications will avoid uncertainty. The choice of a
Regulation will avoid implementation differences in Member States
++
Certainty regarding coverage of software Coverage of software by GPSD clarified
Online sales Safety of products sold on online
platforms
Safety risks for EU consumers due to products sold on online platforms
would be partly reduced (and more so than under Option 2)
+ / ++
Information of consumers on essential
safety aspects
Achievement of objectives can be expected
Recall
effectiveness
Reaching out to consumers affected by
recalls
The option can be expected to increase the availability and facilitate the
use of customer data for recall purposes
++
Information provided in recall notices Improvement in the information provided to consumers achieved
Monitoring of recall effectiveness Improvement in the monitoring of recalls is expected to be achieved,
also depending on implementation
Remedies for consumers affected by
recalls
Higher consumer participation in recalls expected
Market
surveillance
Alignment of market surveillance
framework for harmonised and non-
harmonised consumer products
Largely uniform general requirements for businesses and
responsibilities and powers of MSAs for harmonised and non-
harmonised consumer products expected to be achieved
++
Deterrence effect Deterrence effect likely to be achieved, depending on the maximum
levels of penalties and sanctions foreseen
Diverging risk assessments by MSAs Risk assessment are likely to become more harmonised, achieving the
desired effect
Simplification of standardisation Simplification of standardisation procedures is expected to be achieved
56
procedures
Food-
imitating
products
Addressing risks of food-imitating
products
Clarify the regime for the food-imitating products: (a) full ban or (b)
risk-assessment approach +
The stakeholder survey in the GPSD Study confirmed that all stakeholder groups
considered that Option 3 addressed all challenges at least moderately well with the
overall average assessment across all respondents and stakeholder groups at 3.8 on the
scale from 1 to 5.
Administrative simplification
Option 3 would provide several legally binding clarifications, reducing regulatory
uncertainty in this respect. General requirements for businesses and responsibilities and
powers of market surveillance authorities would be largely uniform for harmonised and
non-harmonised consumer products, and implementation differences in Member States
would be reduced, which is likely to contribute to reduced regulatory complexity and
thereby to reduced administrative burdens for businesses.
On the other hand, Option 3 would include some additional administrative
requirements for specific types of operators (e.g. the requirement to provide essential
safety information online for online traders). The most comprehensive requirements
would apply in the context of recalls, which will likely lead to increased administrative
burdens. The GPSD Study concluded that as currently the effectiveness of recalls is
considered to be limited, these additional measures and the related administrative burdens
appear to be proportionate.
Finally, the simplification of the standardisation process has the potential to reduce
administrative burdens on Member States and at EU level by streamlining the related
EU process. There is also potential for decreasing the regulatory burden for
companies thanks to the arbitration mechanism on the risk assessment.
Also, the choice of Regulation instead of Directive under this option will further reduce
the regulatory burden through a consistent application of product safety rules across the
EU.
Economic impacts
The survey conducted in the GPSD Study estimates113
that the additional recurrent
costs would increase under Option 3 as well as the additional one-off cost (see Table
11). Total costs for businesses in the EU27 in the first year of implementation are
estimated at EUR 196.6 million, equivalent to 0.02% of turnover of EU companies for
manufacturing, wholesale and retail of non-harmonised consumer products. They would
fall in subsequent years to EUR 177.8 million. The rise in costs for businesses in
Option 3 is due to the increased substantive provisions under this Option, requiring
investments on the side of businesses, in particular regarding the online sales and recalls.
113
The accuracy of the given estimates depends on the implementation details
57
Table 11: Changes in EU companies’ costs within and after the first year of implementation of
Option 3, in million EUR
From 0 to 49
employees
50 – 249
employees
250 or more
employees
Total
First year of implementation (total of one-off and recurrent costs)
Total manufacturing sectors 26.9 34.4 55.7 17.0
Total wholesale sectors 12.3 8.5 12.7 33.6
Total retail sectors 24.3 4.6 17.1 46.0
Total additional costs 63.5 47.6 85.6 196.6
Subsequent years (recurrent costs only)
Total manufacturing sectors 25.7 32.9 53.2 111.7
Total wholesale sectors 10.2 7.1 10.6 27.9
Total retail sectors 20.2 3.8 14.2 38.2
Total additional costs 56.1 43.8 78.0 177.8
Source: The survey conducted in the context of the GPSD Study
Concerning the stakeholders views, the GPSD Study showed that companies expect the
implementation of Option 3 to cause changes in their recurrent costs, e.g. costs related to
additional staff and additional resources for due diligence measures including the
establishment of IT systems and external services, in addition to one-off costs, such as
familiarisation costs, adaptation costs to regulatory changes (e.g. for external advice).
Businesses expect that implementing Option 3 would increase companies’ recurrent
regulatory compliance costs, generally more for manufacturers than wholesalers and
retailers, as they might have to adjust different stages of the value-adding process to new
regulatory requirements.
Estimated benefits for businesses linked to cost savings, that are currently caused by
differences in the national implementation of the GPSD are the same under Option 3
as in Option 2 due to the choice of Regulation, amounting to EUR 59 million annually
(EUR 34 million saved by EU SMEs and 26 million EUR by EU large businesses). These
costs savings can be deducted from the costs, i.e. net costs in the first year would be EUR
138 million in the first year.
The implementation of Option 3 would be expected to address current gaps in the product
safety regime for non-harmonised products and thereby safeguard the continued free
movement of goods in the Single Market. This would likely contribute to positive spill-
over effects on consumer trust, demand, production and employment, compared to the
baseline scenario, which is beneficial for all undertakings.
Companies and business associations saw less benefits (between ‘minor’ and
‘moderate’) than MSAs and other stakeholders, who assessed benefits to be mostly
considerably more than ‘moderate’ and close to ‘significant’ (see Annex 12).
As concerns the benefits for SMEs, the GPSD Study shows that small companies
generally estimate that Option 3 would bring a variety of at least ‘minor’ to ‘moderate’
benefits, especially due to its deterrent effect on rogue traders and better detection of
unsafe products. However, Option 3 is considered by small companies as less beneficial
58
when it comes to reducing legal complexity or making compliance with product safety
requirements easier for SMEs. In the case of medium-sized companies, Option 3 is seen
as a suitable contribution to an increased level-playing field among businesses and to
have a significant benefit linked to reducing the occurrence of unsafe products and for
contributing to a better functioning of the EU internal market. Finally, moderate benefits
are expected regarding the potential to increase business revenue or consumer trust.
Even though the relative cost increases are generally higher for the SMEs, the impact
on SMEs overall costs is still considered moderate when measured against the benefits
that would result from a greater level of regulatory harmonisation and reduced regulatory
complexity through the choice of a Regulation, savings being estimated at EUR 34
million for EU SMEs. The changes in SMEs costs are estimated to be limited and
Option 3 would not be expected to affect operations considerably114
.
Online marketplaces interviewed generally agree that the measures under Option 3
would bring several benefits115
. In the GPSD Study, some businesses also stated that
obligations for online marketplaces need to go beyond the Pledge’s provisions and be
aligned with those obligations that need to be met by offline importers/distributors,
including applying ex-ante and ex-post measures and meeting traceability requirements.
Marketplaces also indicated that Option 3 would increase in particular their recurrent
costs. The additional costs would generally be relatively limited for signatories of the
Pledge. By contrast, non-signatory platforms would likely face additional compliance
costs. These compliance costs might specifically affect small platforms and create a
deterrent effect on new market entrants, with potential negative effects on competition
between platforms, depending on the size of the additional costs. Due diligence
obligations in terms of product safety might require more efforts, but would likely imply
less efforts than those of brick and mortar distributors for fulfilling their obligations under
the current regime, thanks to the easier product and customer traceability on the online
interface of a given platform.
Online marketplaces and other online sellers would also be affected by a requirement to
ensure that all safety information is provided online in the same vein as it is required
“offline”. We expect these costs to be very limited for both online platforms and online
sellers, as this information is already available and does not go beyond what is indicated
on the packaging.
Concerning the impact on producers of food-imitating products, the impacts are the
same under Option 3 as in Option 1 and 2.
Concerning the macroeconomic impacts, the results of the consultation conducted for
this study indicate that all stakeholder groups see important benefits of Option 3 in
114
This consideration is also true for specific information obligations, such as the obligation for actors across the online
supply chain to provide all safety information online that is also required to be provided with a product in 'brick and
mortar' stores, and the related obligation for online platforms to make sure that third-party sellers, such as SMEs,
provide this information. We expect these costs to be relatively minor for companies selling consumer products on
these platforms, including SMEs.
115
According to them option 3 would improve consumer trust, provide better information on unsafe products and
ensure more effective measures taken by MSAs through Safety Gate/RAPEX, and provide greater legal certainty and
less complexity. Online platforms respondents also tended to agree that the measures in Option 3 would have a
deterrent effect on rogue traders and reduce the occurrence of products presenting health and safety risks in the Single
Market.
59
terms of a better functioning EU internal market and a better level-playing field among
businesses, partly through the deterrent effect on rogue traders. All these potential
benefits were assessed as being ‘moderate’ to ‘significant’ in the GPSD Study’s survey.
The alignment of the market surveillance rules for all products and a clearer legal
framework under Option 3 should overall significantly reduce the businesses’ compliance
costs and administrative burdens, which would level the playing field for companies from
different countries within the EU and may help many European businesses to be more
competitive internationally. At the same time, a more harmonised regulatory level-
playing field within the EU will also induce non-EU companies to market their products
in the EU, with positive impacts on intra-EU competition. The additional gains in EU
companies’ competitiveness are expected to be very moderate as companies’ current
compliance costs with consumer product safety legislation are already relatively low116
,
accounting for relatively small shares of total revenues. Moreover, additional regulatory
requirements would level potential cost reductions.
Depending on the actual implementation, Option 3 should also create a higher deterrent
effect on rogue traders and therefore ensure a better level-playing field for companies by
ensuring that all bear the compliance costs with products safety: this was an important
point raised in the different consultation activities.
Due to the relatively low additional costs for businesses, representing 0.02% of their
annual turnover in the first year of implementation, that would result from Option 3,
we expect neither significant distortions in competition and international trade for EU
businesses, nor significant impacts on EU companies’ overall innovative capacities117
.
The GPSD Study nevertheless expects positive impacts on competition-driven innovation
due to a greater degree of harmonisation and greater legal certainty (e.g. development of
new innovative information and traceability systems).
Concerning the impact on consumers, the implementation of Option 3 would result in
greater benefits for consumers due to broader coverage and greater effectiveness of the
GPSD in protecting consumers from unsafe products, particularly with respect to the
mitigation of risks from new technologies and the coverage of products sold via online
channels. Option 3 could therefore be expected to increase consumer safety in the online
environment and have positive effects on consumer trust, which might translate in higher
demand for consumer goods that are sold via online channels. The GPSD Study estimated
that consumers would therefore benefit in terms of reduced consumer detriment based
on the value of unsafe products118
. This is expected to amount to approximately EUR
1.0 billion in the first year of implementation, increasing to approximately EUR 5.5
billion over the next decade. Improving the recall effectiveness and therefore reducing
the number of unsafe products remaining in hands of consumers would also bring
benefits: the GPSD Study estimates under a scenario of significantly improved recall
effectiveness (under the assumption that return rates of recalled products are doubled due
to legislative measures and more deterrent sanctions and penalties as expected under
116
The GPSD Study found that companies already do a lot for safety as usual business.
117
However, new regulatory requirements for online platforms might result in less competitive dynamism and
innovation in online platform business models over time, depending on the extent to which new requirements lead to
additional costs (and their size compared to other cost factors), similarly to Option 2
118
Consumer detriment linked to the value of unsafe products, calculated on the basis of the purchase price of unsafe
products as explained above.
60
Option 3), the consumer detriment due to ineffective recalls to be reduced by EUR
410 million in Option 3.
The survey in the GPSD Study showed that stakeholders consider that Option 3 provides
‘moderate’ to ‘significant’ benefits for consumers. These include a reduced occurrence
of unsafe products and a reduced number on injuries caused by them, as well as a
resulting increase in consumer trust.
Some of the additional costs incurred to businesses by Option 3 would be passed on, both
up- and downstream the product value chain, and thereby impact consumer prices.
However, as most businesses report relatively low additional one-off and recurrent costs,
the short and medium- to long-term impacts on consumer prices in the EU are expected to
be negligible. Also, the GPSD Study does not expect a significant negative impact on
consumer choice in the EU under Option 3119
.
On the other hand, a limited effect pertaining to the affordability of products is also
possible. While the increase in consumer prices is overall considered negligible under
Option 3, purchase prices for some non-harmonised products might be affected (e.g.
products that are most cheaply ordered through online platforms from non-EU/EEA
traders), and low-income consumers with high price-elasticity may reduce their
purchases.
Concerning the impact on consumers of the sub-options for the food-imitating products,
we consider that both sub-options lead to the same level of consumer protection against
unsafe products, as analysed under Option 1.
Impacts on Member States
Based on the results from the Study, efficiency gains by MSAs would mostly arise from
the alignment of the provisions for market surveillance of harmonised and non-
harmonised products. This would lead to improvements in efficiency of market
surveillance, and related cost savings, estimated at EUR 0.7 million per year across the
EU under Option 3, similarly to Option 2. Also, streamlined standardisation procedures
and an arbitration mechanism that provides clarification regarding risk assessments in
case of disputes between MSAs could lead to additional cost reductions for MSAs over
time.
Concerning the costs, the Study estimates that Option 3 would lead to total additional
recurrent costs of MSAs in EU27 of approx. EUR 6.7 million annually and only
relatively moderate one-off adaptation and implementation costs.
Generally, the efficiency of market surveillance processes with implications across the
EU would be increased under Option 3, mainly via an increased deterrent effect through
provisions on penalties and arbitration mechanism.
Social impacts
The implementation of Option 3 is expected to potentially have positive social impacts
with regards to public health and safety and health systems, higher than in Option 2.
119
This will depend more on other measures taken at EU level (e.g. changes to the VAT regime, the provisions of the
new DSA), which may impact market access for products sold directly to consumers by non-EU/EEA traders through
online marketplaces.
61
The introduction of additional requirements for traceability and product recalls are
expected to improve the effectiveness of recalls of unsafe products from consumers. For
instance, the cost savings from directly informing consumers affected by a recall rather
than using indirect communication channels have been estimated at €73 million in 2019,
i.e. a fifth of the overall estimated cost of recall ineffectiveness120
. In addition, increased
enforcement powers for Member States to impose penalties and sanctions are anticipated
to significantly improve market surveillance and enforcement. Consequently, the current
cost of health care utilisation for product-related injuries in the EU of approximately EUR
6.7 billion per year could be further reduced under Option 3.
Environmental impacts
On the benefits related to environment of Option 3, the Study reports that while
authorities see ‘moderate’ to ‘significant’ benefits regarding improved lifecycle/quality of
products and a higher level of environmental protection due to the reduction of unsafe
products that also have environmental impacts, companies/business associations and
other stakeholders only see benefits that are (close to) ‘moderate’. The same analysis
concerning the hazardous chemicals applies here as under Option 2, but the positive
impact of Option 3 is amplified by the expected better product safety enforcement.
Impacts on fundamental rights
Option 3 is expected to improve consumer safety whilst also reducing environmental
risks (see below). It would thus have a positive impact and ensure a higher level of
consumers’ life as well as consumer protection and environmental protection in line
with the Charter. This positive impact should be amplified by better product safety
enforcement.
The additional requirements imposed on economic operators do not affect the
fundamental freedom to conduct a business, as they are necessary to pursue the
general EU interest of increasing consumer protection and are proportional to the aim
pursued, given that the resulting compliance costs are estimated to be comparatively low
compared to the businesses’ turnover. The negative effect of a potential ban of food-
imitating products would be the same as developed under Option 1.
6.4. Impacts of Option 4
The main significant difference between options 3 and 4 concerns the merger of legal
instruments for the market surveillance rules for harmonised and non-harmonised
products into one single Regulation. Therefore the assessment of impacts will
concentrate only on this additional element (the rest of the analysis under Option 3 is also
valid for Option 4, unless stated otherwise in this section).
Effectiveness in achieving the policy objectives
Creating a single set of rules that would apply to harmonised and non-harmonised
consumer products will simplify the EU legal framework greatly. It can be expected that
the objective to create uniform requirements for businesses and responsibilities and
120
European Commission, 2021, Behavioural study on strategies to improve the effectiveness of product
recalls.
62
powers of market surveillance authorities for harmonised and non-harmonised consumer
products will be fully achieved. All stakeholder groups considered that Option 4
addressed all challenges at least moderately well. The average assessment across all
stakeholder groups was 3.8 out of 5 (similar to Option 3).
Administrative simplification
In addition to Option 3, a single set of rules for market surveillance of harmonised and
non-harmonised consumer products in the EU could, overall, result in even less legal
complexity. This could translate into simplifications for businesses and MSAs in
countries where current national law implements the GPSD and harmonised
product legislation through different legal instruments. However, where all product
safety legislation is already transposed into a single national product safety law (which is
the case in some countries), simplifications through a new EU legal instrument are likely
to be very limited.
Economic impacts
The GPSD Study showed that businesses expected costs from Option to be significantly
higher compared to Option 3 (see Table 12). Total costs for businesses in EU27 in the
first year of implementation are estimated at EUR 331.1 million, equivalent to 0.03% of
turnover of EU companies for manufacturing, wholesale and retail of non-harmonised
consumer products. They would fall in subsequent years to EUR 296.3 million.
Table 12: Changes in EU companies’ costs within and after the first year of implementation of
Option 4, in million EUR
From 0 to 49
employees
50 – 249
employees
250 or more
employees
Total
First year of implementation (total of one-off and recurrent costs)
Total manufacturing sectors 45.0 57.7 93.4 196.0
Total wholesale sectors 21.0 14.5 21.6 57.0
Total retail sectors 41.2 7.8 29.0 78.0
Total additional costs 107.2 79.9 144.0 331.1
Subsequent years (recurrent costs only)
Total manufacturing sectors 42.7 54.8 88.7 186.2
Total wholesale sectors 17.1 11.8 17.6 46.5
Total retail sectors 33.6 6.4 23.7 63.6
Total additional costs 93.4 72.9 130.0 296.3
Source: The survey conducted in the context of the GPSD Study
Compared to Option 3, Option 4 comes therefore to substantial additional costs for
businesses even if there is no substantive difference regarding the regulatory obligations
for businesses and integration of legal instruments should overall reduce regulatory
complexity.
A possible explanation for this is that businesses were generally uncertain about the
precise implications of Option 4 and tended to provide cautious estimates with regard to
additional costs from new regulatory obligations that might arise if one single set of rules
would apply. The data on costs for businesses under different options provided by the
support study are based on the cost estimations provided directly by the businesses
during the survey. When analysing the reasons why businesses were reporting higher
costs for Option 4, it appeared that for businesses the legal certainty, clarity of their
63
obligations, even application across the EU and predictable legal environment are very
important. This was also confirmed by the meetings with businesses associations. If the
rules were set in one legal instrument, that would provide simplicity but proportionally
more for Member Sates authorities (in particular in those Member States where market
surveillance authorities handle both harmonised and non-harmonised products together)
than for businesses. On the contrary, businesses perceived some regulatory risk in Option
4 linked to the possible reopening of Regulation (EU) 2019/1020 for which negotiations
have been recently concluded. Businesses had already to invest into the compliance with
this new market surveillance Regulation and perceived that merging the rules could lead
to reopening of market surveillance provisions already agreed and would create
uncertainty about their future legal environment. It should be noted, however, that these
costs reported by businesses might be inflated since they are not based on any specific
calculation grounds but simple assumptions. If the new market surveillance framework
would integrate the provisions relevant for economic operators in Regulation (EU)
2019/1020, such a new legal instrument would not entail additional costs for businesses.
Since Regulation (EU) 2019/1020 is setting up rules for market surveillance for
harmonised products, any integration of market surveillance rules under one unique legal
instrument would therefore mainly impact Member States rather than businesses, in
particular by creating benefits for those Member States where the same authorities handle
both categories of products, harmonised and non-harmonised.
This difference in viewpoints between businesses and national authorities has been also
confirmed by the general assessment of the benefits of Option 4 in the external survey:
Businesses have been more sceptical, and saw only slightly less than ‘moderate’ benefits
on average (2.9 on a scale of 5), while market surveillance authorities assessed the Option
4 as bringing close to ‘significant’ benefits (value of 3.9).
Other economic impacts (both micro- and macroeconomic impacts) as well as impact
on consumer and households are expected to be identical to the impacts under
Option 3.
The survey of the GPSD Study shows that MSAs expect on average considerably more
benefits that would result from an implementation of Option 4 than businesses/business
associations and other stakeholders121
. It is notable that the expected benefits of Option 4
are considered to be slightly higher by all stakeholders than the benefits of Option 3.
Impacts on Member States
MSAs responding to the Study’s survey stated that having the same rules for all
harmonised and non-harmonised products would induce benefits beyond those already
identified under Option 3.
The Study found that Option 4 would bring an increase in recurrent costs of MSAs of
5% of total annual staff-related costs: this would amount to total additional costs of
MSAs in the EU27 of approx. EUR 3.3 million annually122
. The few numbers that were
121
Overall, MSAs assessed a value of 3.9, or very close to ‘significant’ benefits. MSAs especially expect ‘significant’
benefits from greater legal certainty and reduced legal complexity (values of 4.3 and 4.1 respectively). Also other
stakeholders see this option bringing ‘moderate’ to ‘significant’ benefits (average 3.5). Businesses are more sceptical,
and see slightly less than ‘moderate’ benefits on average (2.9).
122
Actual percentage changes would differ for individual MSAs due to different national institutional market
surveillance systems and organisational characteristics, e.g. the degree of centralisation, MSAs’ product coverage and,
after all, the actual assignment of new competences and enforcement requirements.
64
provided by MSA respondents indicate that the one-off adaption and implementation
costs are considered to be relatively minor (e.g. to prepare some national guidance, new
communication strategy and to strengthen cooperation at the national level).
Social impacts, environmental impacts and impacts on fundamental rights
As the measures implemented under Option 4 are identical to the measures implemented
under Option 3, the two options are expected to have identical impacts in a social or
environmental perspective, as well as on fundamental rights.
7. HOW DO THE OPTIONS COMPARE?
In this section, we compare the results of the impact assessment of the four options, based
on the elements developed in the section 6 and the results of the GPSD Study.
Expected achievement of objectives
Table 13: Comparative assessment of impact on objectives
Source: The survey conducted in the context of the GPSD Study
Option 1 would be expected to achieve only one of the five policy objectives, with some
additional benefits due to reduction of uncertainty (provision of EC guidance). Option 2
appears more effective in reaching objectives, with some identified gaps being closed and
uncertainty reduced by legal measures; however some other gaps remain (e.g. regarding
software, product recalls and online sales channels). Options 3 and 4 would most likely
achieve all the defined objectives of the initiative.
Administrative simplification
Table 14: Comparative assessment of impact on administrative simplification
Option 1 is expected to bring slight reduction of regulatory complexity and uncertainty
(via guidance). There are no new administrative requirements, however: administrative
burdens due to current fragmentation of legal regime continue (experienced by 16% of
MSAs and 42% of companies responding to the survey in the GPSD Study as explained
in the baseline section). Option 2 would bring some additional reduction of regulatory
Area Option 1 Option 2 Option 3 Option 4
Reduction of regulatory
complexity and uncertainty
neutral / + neutral / + + + / ++
Area Option 1 Option 2 Option 3 Option 4
Ensure general safety rules, including for product risks
linked to new technologies
neutral / + + ++ ++
Address safety challenges in the online sales channels neutral neutral / + + / ++ + / ++
Make product recalls more effective neutral + ++ ++
Enhance market surveillance and ensure better
alignment of rules
neutral ++ ++ ++
Address safety issues related to food imitating products + +
+ +
Total effectiveness score by stakeholders (scale 1-5) 2,9 3,4 3,8 3,8
65
complexity and uncertainty, especially if a Regulation was chosen and involve only very
limited additional administrative requirements for specific operators.
Under Options 3 and 4, the reduction of regulatory complexity and uncertainty is the
most significant (all regulatory gaps closed), with related reduction in administrative
burdens for businesses. Some additional administrative requirements concern specific
types of operators, the most comprehensive ones concern recalls, which would be limited
to companies that have brought unsafe products on the market. Option 4 will further
reduce regulatory complexity and bring simplicity, as one single set of rules would apply
to harmonised and non-harmonised products.
Economic impacts
Table 15: Comparative assessment of micro- and macroeconomic impacts
Area Option 1 Option 2 Option 3 Option 4
Benefits for businesses
neutral / +
neutral / +
(Benefits of EUR
59 million/year, if
Regulation)
+
Benefits of EUR
59 million/year
+
Benefits of EUR
59 million/year
Cost of businesses
(EU27)
neutral
Increase by < EUR 37
million/year
Increase by < EUR
197 million/year
Increase by < EUR
332 million/year
Macroeconomic impacts
(Internal market, trade,
competition, innovation)
neutral neutral / + + +
Estimated benefits for businesses linked to costs savings, that are currently caused by
differences in the national implementation of the GPSD and would be partly solved if
the new instrument is a Regulation (Options 2, 3 and 4), would amount to EUR 59
million annually123
, of which EUR 34 million would be saved by EU SMEs and 26
million EUR saved by EU large businesses respectively, compared to the baseline.
Other additional economic benefits for businesses are expected to be minor under
Options 1 and 2, mostly related to reduction of uncertainty due to guidance (Option 1) or
the coverage of certain gaps in a recast GPSD (Option 2). Benefits are expected to
increase with Options 3 and 4, as all legislative gaps identified in the problem analysis
are closed and related uncertainty is avoided. The measures taken regarding online sales
contribute to safeguarding a level-playing field for businesses and the deterrence of rogue
traders, which are expected to have concrete benefits at firm level, especially in those
areas where consumer trust and safety are affected by unsafe products entering the EU
through direct online business to consumer transactions.
123
The baseline costs linked to the different implementation of the GPSD are estimated to amount to 119 million EUR
annually (see section 7). As Options 3, 4 and possible 2 foresee to recast the GPSD as a Regulation, implementation
differences would be avoided (due to the direct applicability of the new regulation in Member States), even if some
differences in the national interpretation of rules may remain. Accordingly, we assume a 50% reduction of businesses’
additional costs in this respect in case of choice of Regulation as legal instrument.
66
Table 16: Changes in EU companies’ costs within and after the first year of implementation of
Options 1 to 4, EU27, in million EUR
Option 1 Option 2 Option 3 Option 4
First year of implementation (total of one-off and recurrent costs)
Manufacturing sectors 0 20.7 17.0 196.0
Wholesale sectors 0 6.9 33.6 57.0
Retail sectors 0 9.4 46.0 78.0
Total additional costs (EU27) 0 36.9 196.6 331.1
Subsequent years (recurrent costs only)
Manufacturing sectors 0 18.6 111.7 186.2
Wholesale sectors 0 4.7 27.9 46.5
Retail sectors 0 6.4 38.2 63.6
Total additional costs (EU27) 0 29.6 177.8 296.3
Equivalent to the share of turnover of EU companies for manufacturing, wholesale and retail of non-harmonised
consumer products (first year of implementation):
Share in turnover 0% 0.004% 0.02% 0.03%
There are no changes in compliance costs for EU companies under Option 1, and only
expected to a minor extent under Option 2. Compliance costs of businesses are expected
to increase more significantly under Options 3 and 4, however still representing a small
fraction of companies’ turnover, maximum 0.03% for Option 4.
Under options 2 to 4 (see Tables 9, 11, 12), the effects of additional compliance costs
will have a larger relative cost impact on SMEs than on large companies. Even though
the relative cost increases are higher for SMEs, the impact on SMEs overall costs is still
considered moderate when measured against the benefits that would result from a
greater level of regulatory harmonisation. The changes in SMEs costs are small and
implementation of any of the options would not be expected to significantly affect SMEs.
Minor impacts on online platforms are expected under Option 2 for those that are not yet
signatories of the Pledge. Under Options 3 and 4 impacts on online platforms are higher,
due to due diligence obligations in terms of product safety. However this would likely
imply less efforts than those the classical distributors have today under GPSD and
therefore proportionate to the general objective.
With respect to macroeconomic impacts, the impacts are expected to be mostly limited,
with most (positive) impacts to be expected under Options 3 and 4. Both options would
be expected to lead to a more aligned and clearer EU legislative framework as well as
reduced legal complexity, which could overall significantly reduce the part of companies’
compliance costs.
Impacts on consumers and households
The consumers will benefit from the reduction of the unsafe products on the EU market.
The expected impact of the different options on the reduction of the consumer detriment
linked to the value of unsafe products is presented in Table 17124
.
124
Estimation of the impact on the injury related detriment could not be done due to the lack of data.
67
Table 17: Expected reduction in consumer detriment due to unsafe products– EU27, in EUR million
per year
Year Option 1 Option 2 Option 3 Option 4
2025 (expected 1st implementation year) 0 333 1 038
As Option 3
2026 0 704 2 153
2029 0 821 3 924
2034 0 1 031 5 491
Options 3 and 4 are likely to be more effective than options 1 and 2 to address the
challenges for product safety posed by online sales channels. The measures taken under
Options 3 and 4 also contribute to aligning the level of product safety between the online
and offline sales channels and increasing it, and thereby to reducing the incidence of
unsafe products on the market overall. Measures taken under Options 3 and 4 are also
expected to reduce consumer detriment estimated on the basis of the value of unsafe
products by approximately EUR 1.0 billion in the first year of implementation,
increasing to approximately EUR 5.5 billion over the next decade, much higher than
in Option 2 and 1. This represents the decrease of financial costs for consumers since
they would avoid buying unsafe products. The reason for this increase over time is that
overall consumer detriment is expected to grow in the mid-term in the baseline scenario,
due to increasing consumption and a continuing shift to e-commerce.
Also, enhancing recall effectiveness would reduce the consumer detriment since less
unsafe products would remain in hands of consumers and they might get compensated for
the recalled products. Guidance measures under Option 1 in the area of product recalls
are not expected to lead to a significantly higher recall effectiveness, and therefore are not
expected to reduce related detriment. In contrast, Options 3 and 4 could be expected to
substantially reduce consumer detriment related to the value of unsafe products
which were not effectively recalled by more than EUR 400 million per year (Option 2
by half of this amount).125
Table 18: Expected reduction in consumer detriment due to ineffective recalls – EU27, in EUR
million per year
Year Option 1 Option 2 Option 3 Option 4
Reduction of consumer detriment 0 205 410 410
The other impacts on consumers and households have been estimated as follows:
Table 19: Comparative assessment of other impacts on consumers and households
Area Option 1 Option 2 Option 3 Option 4
Consumer prices neutral neutral neutral neutral
Consumer choice neutral neutral neutral neutral
Consumer safety and vulnerable consumers neutral + ++ ++
No impacts on consumer prices and choice are expected, as estimated increases in
compliance costs are small compared to baseline costs, and companies’ overall product
125
This estimate is based on a number of scenario assumptions, to provide a reasonable and cautious estimate of
consumer benefits due to improved recall effectiveness. A key assumption is that the detriment incurred by consumers
in case of a recall of an unsafe product is equivalent to at least its purchase price (a recalled, unsafe product could also
cause additional detriment linked to damage to persons, other goods or the environment).
68
safety-related costs, including regulatory compliance costs, account for only very limited
shares of the companies’ turnover.
Regarding consumer safety and the protection of vulnerable consumer groups, the four
options differ: Options 3 and 4 are expected to provide a higher level of protection in
terms of consumer safety and the protection of vulnerable consumer groups, as existing
regulatory gaps are closed and the related policy objectives are better achieved.
Impacts on Member States
Table 20: Comparative assessment of impact on Member States
Area Option 1 Option 2 Option 3 Option 4
Benefits for MSAs neutral / + +
Benefits of
EUR 0.7 million/year
++
Benefits of
EUR 0.7 million/year
++
Benefits of
EUR 0.7 million/year
Costs for MSAs
(EU27)
neutral mostly neutral (<EUR
7 million/year)
mostly neutral
(<EUR 7 million/year)
mostly neutral
(<EUR 4 million/year)
Other effects on
Member States
neutral neutral / + + +
Benefits for MSAs would mostly arise from the alignment of the provisions for market
surveillance of harmonised and non-harmonised products. This leads to improvements in
efficiency of market surveillance, and related cost savings, which are estimated at EUR
0.7 million per year across the EU (for Option 2 if Regulation is chosen and Options 3
and 4)126
. Also, streamlined standardisation procedures and an arbitration mechanism that
provides clarification regarding risk assessments in case of disputes between MSAs could
lead to additional cost reductions for MSAs over time.
Cost for MSAs are not expected to increase significantly under any of the options.
With Option 1, no additional costs are to be expected. Under the other options, estimates
of additional costs are between EUR 3.3 million/year (Option 4) and EUR 6.6.
million/year (in Options 2 and 3 respectively), the difference being related to the
expected degree of legislative alignment (the most far-reaching alignment of the
legislative framework is under Option 4, which leads to most efficiency gains).
Social impacts, impacts on fundamental rights and environmental impacts
Table 21: Comparative assessment of other impacts
Area Option 1 Option 2 Option 3 Option 4
Social impacts neutral neutral / + neutral / + neutral / +
Environmental impacts neutral neutral / + + +
Impacts on fundamental rights neutral neutral / + + +
Option 1 is not expected to have significant social impacts and Option 2 only to have
some positive ones, to the extent that number of unsafe products and product-related
126
See baseline description. The proposed measures under Options 2 (if Regulation), 3 and 4 would fully align
provisions for the market surveillance of harmonised and non-harmonised consumer products so that the cost burden
estimated in the baseline as EUR 0.7 million will be reduced accordingly. Legislative fragmentation between
harmonised and non-harmonised products currently creates costs for MSAs, estimated to amount to EUR 0.7 million
annually (total for the EU27). If the new legislation is a Regulation aligning rules for harmonised and non-harmonised
products, it would create benefits in form of costs savings to MSA which are estimated to fully reach the amount of
EUR 0.7 million per year since the legal fragmentation would disappear.
69
environmental risks are reduced. Most positive social impacts are expected under
Options 3 and 4, due to enhanced market surveillance which should reduce the number
of unsafe products on the market in the mid- to long term, and consequently to a lower
number of injury cases, hence lowering public health costs.
Also reduction of product-related environmental risks decreases in particular in Options 3
and 4, to the extent that the application of safety requirement in this respect is clarified
and effectiveness of recalls of products posing environmental risks is improved.
Options 1 and 2 are not likely to have significant impacts on fundamental rights (Option 2
possibly minor), while Options 3 or 4 are expected to have a positive impact and ensure a
higher level of the consumer protection and environmental protection. Even if Options 3
and 4 impose additional requirements for businesses, these do not affect the fundamental
freedom to conduct a business and appear to be proportional to the general objective
pursued. However, a ban of food-imitating products from the EU market would have a
negative impact on the freedom to conduct a business, and its proportionality regarding
Art 52 of the Charter would need to be proven.
Impacts of sub-options for food-imitating products
A full ban of food-imitating products (sub-option (a)) and case by case risk-
assessment of food-imitating products (sub-option (b)) appear to both deliver the same
level of consumer protection since in both cases unsafe products would be subject to
corrective measures. In both options there is a benefit linked to better clarification and
harmonisation of the rules which would lead to higher legal certainty and level-playing
field for economic operators in both sub-options. The broad economic negative impacts
of a full ban of food-imitating products would likely be minor since the number of
these products is limited. But a full ban on food-imitating products from the EU market
without risk-assessment would have a negative impact on the freedom to conduct a
business, while there is no evidence to prove that it protects better consumers, in
particular children, to confirm its proportionality. At the same time, Option (b) is fully
coherent with the current risk-based assessment approach in the GPSD and more
proportionate to the possible economic impact of corrective measures on economic
operators.
Impacts of the choice of the legal instrument
The analysis shows that a regulation is preferable to a directive in terms of choice of
the legal form. A regulation is directly applicable in all Member States; there is therefore
no need for Member States to transpose EU legislation into national law and no need to
provide them with time to do so. Possible national differences regarding the date and/or
manner of transposition would be eliminated with a regulation, which would facilitate
consistent enforcement and level-playing field in the internal market. A regulation
ensures better that legal requirements are implemented at the same time throughout the
Union; it also better achieves streamlining of terminology, important for defining the
scope of the legislation, thereby reducing administrative burdens and legal ambiguities;
this is particularly true in light of the fact that one of the purposes of the revised GPSD is
to make it as coherent as possible with Regulation (EU) 2019/1020, which is indeed a
Regulation.
From a subsidiarity and proportionality perspective, the choice of the legal delivery
instrument in the form of regulation or a directive does not differ in term of impact. In
70
both cases, the subsidiarity principle is respected since the EU action is necessary to
harmonise the general product safety requirement in the EU and ensure therefore safety
of products, consumer protection and level-playing field in the Single Market. Both
instruments are proportional since the requirements introduced are proportionate to
achieve the level of product safety needed to ensure consumer protection and the level
playing-field for businesses.
The choice of a Regulation instead of a Directive under the Options 3 and 4 (and
possible Option 2) will further reduce the regulatory burden thanks to a more
consistent application of product safety rules across the EU.
Table 22 below presents the overview of all the impacts analysed in this IA report.
Concerning the methodology for the comparison of impacts, the report generally operates
with the “+/-“ rating system for impacts that were qualitatively assessed. However, one
composite indicator representing the “expected achievement of objectives” was computed
based on 5 qualitative indicators measuring the degree the 5 specific objectives of the
GPSD would be achieved. The scale “1-5” was chosen for it to allow for a more accurate
overall score (2.9, 3.8 etc.) of the aggregate of the 5 impacts assessed with the “+/-“
rating (which has only 5 scales: --, -, neutral, +, ++).
Table 22: GPSD Overview of the impacts of policy options
Option 1
Enhanced enforcement
Option 2
Targeted legal revision
Option 3
Full legal revision
Option 4
Integration of rules
Expected achievement of objectives (scale 1-5) 2,9 3,4 3,8 3,8
Administrative simplification neutral / + neutral / + + + / ++
Economic impacts
Benefits for businesses
Benefits if Regulation
neutral / +
NA
neutral / +
EUR 59 million/year
+
EUR 59 million/year
+
EUR 59 million/year
Cost for businesses (EU27) Increase by
Share in turnover
neutral
0%
< EUR 37 million/year
0.004%
< EUR 197 million/year
0.02%
< EUR 332 million/year
0.03%
Macroeconomic impacts (Internal market, trade, competition, innovation) neutral neutral / + + +
Impacts on consumers
Expected reduction in consumer detriment due to unsafe products – EU27,
in EUR million per year
1st
year 0 333 1 038 1038
10th
year 0 1 031 5 491 5491
Expected reduction in consumer detriment due to ineffective recalls –
EU27, in EUR million per year 0 205 410 410
Consumer prices neutral neutral neutral neutral
Consumer choice neutral neutral neutral neutral
Consumer safety and vulnerable consumers neutral + ++ ++
Impacts on Member States
Benefits for MSAs
Benefits if Regulation
neutral / +
NA
+
EUR 0.7 million/year
++
EUR 0.7 million/year
++
EUR 0.7 million/year
Costs for MSAs (EU27) neutral
mostly neutral
<EUR 7 million/year
mostly neutral
<EUR 7 million/year
mostly neutral
<EUR 4 million/year
Other effects on Member States neutral neutral / + + +
Social impacts neutral neutral / + neutral / + neutral / +
Environmental impacts neutral neutral / + + +
Impacts on fundamental rights neutral neutral / + + +
72
Coherence with other EU policy objectives
Options 3 and 4 deliver the most on the Digital priorities of the EU. These options have
the highest impact on the online sales and, in line with the EU’s objectives to Shape EU’s
digital future, these options contribute to make sure that online platforms treat their users
fairly and take action to limit the spread of unsafe products online. Also these options
provide higher safety and legal certainty for connected products and cybersecurity risks,
in line with the EU’s actions in the AI and cybersecurity fields. Options 3 and 4 also
deliver the most in term of positive environmental impacts and are therefore in line with
the EU Green deal priority of the Commission.
Stakeholder views on the options
Figure 5: In your view, to what extent
would Option [...] effectively address
the following challenges for product
safety? – Average across all
challenges
Source: The survey conducted in the
context of the GPSD Study127
Authorities and other stakeholders assessed Options 3 and 4 as being most effective, and
considered them to well address the five objectives of this initiative. In contrast, average
assessments by companies/business associations do not show a considerable variation
between the options. They consider all four options to address the challenges slightly
better than ‘moderately well’.
Figure 6: Where do you see the
greatest additional benefits that
would result from the
implementation of Option […]? –
Average across all benefit categories
Source: The survey conducted in the
context of the GPSD Study
127
In total, 153 survey responses were received, of which 27 responses to the survey of consumer organisations and
other general stakeholder; 48 responses to the survey of authorities, 37 responses to the survey of business associations
and 41 responses to the survey of companies.
1 2 3 4 5
Option 1
Option 2
Option 3
Option 4
Not at all Very well
Other stakeholders Authorities Companies/ Businessassociations
1 2 3 4 5
Option 1
Option 2
Option 3
Option 4
No change in Very significant
benefitsat all
Other stakeholders Authorities Companies/ Businessassociations
73
Businesses find the benefits of Options 3 and 4 to be ‘moderate’ on average, but still
clearly more beneficial than Options 1 and 2. All stakeholders provided the following
ranking of benefits for Options 3 and 4, with average above ‘moderate’ in all three
stakeholder groups: (1) Better functioning EU internal market, (2) Reduced occurrence of
products with health and safety risks, (3) Greater legal certainty, (4) More level-playing
field among businesses, (5) Reduced number of accidents/injuries caused by unsafe
products, (6) Better information on unsafe products, (7) Deterrent effect on rogue traders.
In the OPC, the stakeholders showed a clear support to certain of the proposed
provisions under Option 2, 3 and 4, e.g. to increase the role of online marketplaces as
regards the safety of products offered on their website, along the principles stated in the
Pledge128
and to create an obligation to have a responsible economic operator in the
EU (supported by 70% of stakeholders)129
. Also, a large majority of respondents
expressed that products which resemble foodstuff should be incorporated into the
general product safety legal instrument (69%)130
. Stakeholders expressed also their
support for certain additional provisions foreseen under the Option 3 and 4, e.g. on new
technologies: When asked whether the definition of a product in the GPSD should
specifically encompass software incorporated into the product, the majority of
respondents agreed, even in case the software is downloaded after the product has been
sold (56%). A clear majority of respondents favoured safety obligations for
manufacturers of products incorporating AI applications at the design stage and also
during the lifecycle of the product (75%)131
. Also, a large majority of respondents agreed
that the system of product traceability should be reinforced in the GPSD (82%)132
.
Ranking of options
All the options defined in the report propose specific actions to address all five problems
identified. However, the analysis of impacts shows that some options deliver better on
the defined objectives than others.
Option 1 has considered how to best respond to the specific objectives without revising
the GPSD. Several non-legislative measures have been considered, in particular issuing
guidance documents on the applicability of the GPSD to new technologies and on recalls
and exploring extension of the voluntary measures under the Product Safety Pledge for
online sales. However, the different consultations showed that such non-legally binding
measures would not tackle the identified shortcomings. Also additional funding
possibilities for market surveillance have been considered. But the agreement on the EU
budget for next years showed that the EU funding of the joint market surveillance
activities by the EU budget will remain stable; moreover, in view of the budget
128
See Annex 11 on results of the OPC. When asked about the role that online marketplaces should play regarding the
safety of products offered on their websites, the most commonly supported notions were that they should remove
dangerous products listed on their website when notified (77%), that online marketplaces should prevent the
appearance of dangerous products, including their reappearance once they have been removed (66%) and that they
should inform sellers of their obligation to comply with EU rules on products (64%). More than half of the respondents
agreed that online marketplaces should inform consumers when a dangerous product has been removed from the
marketplace (55%).
129
See Annex 11 on results of the OPC. A large majority of respondents considered that products covered by the
GPSD should only be placed on the EU market if there is an economic operator established in the EU responsible for
product safety purposes (70%).
130
See Annex 11 on results of the OPC.
131
Idem
132
Idem
74
constraints in Member States, aggravated by the current Covid-19 crises, we do not
expect any increase of resources dedicated to market surveillance activities by Member
States themselves.
Under Option 2 and 3 several legislative actions have been considered to tackle the
specific objectives: Option 3 being more ambitious, addresses also better the identified
shortcomings as data shows. Option 4 has considered a full integration of market
surveillance instruments, as it was proposed in 2013, to analyse whether this option
would still be valid after the recent adoption of Regulation (EU) 2019/1020.
Regarding the food-imitating products, different options have been looked at, namely:
1) to maintain a separate Directive on food-imitating products; 2) to merge the provisions
of the current FIPD into the new GPSD; 3) to abandon targeted provisions on food-
imitating products and instead use the general provisions to ensure safety of such
products. For the first two options, the possibility of developing guidance has been
considered in order to overcome the different interpretation by Member States; however,
the consultation of Member States showed that the divergences in interpretation of the
Food-Imitating Products Directive were so important that a legal revision of the rules was
necessary to ensure its even application.
Furthermore, the assessment showed that a general ban of food-imitating products would
result in banning some non-dangerous products, which would be an unjustified and non-
proportional restriction of the freedom to conduct a business. It is therefore essential that
food-imitating products follow the same risk-based approach that prevails for the other
consumer products. Also keeping a separate regime for food-imitating products has been
considered as not necessary in view of the low number of related notifications in the
Safety Gate/RAPEX. To assess the safety of these products, their food-imitating aspect
can be taken into account in the risk assessment under the GPSD, which appears then to
be an appropriate legal instrument to cover the safety of these products. Therefore the
third option consisting in abandoning targeted provisions on food-imitating products and
instead using the general provisions and the risk-based approach contained in the GPSD
to ensure safety of such products appears as the most appropriate.
Table 23: Ranking of the options
Assessment Ranking
Option 1 While Option 1 is causing no costs for businesses and MSAs, it is unlikely to be
adequate to address the problems identified. While uncertainty will be reduced due to
Commission guidance, and coverage of online platforms is expected to increase
through the promotion of the Product Safety Pledge, safety risks due to products sold
on online platforms are expected to continue, as will the other gaps identified.
4
Option 2 Option 2 is causing extremely limited costs (<0.004% of turnover for business,
mostly neutral for MSAs), and is likely to be partially adequate to address the
identified problems. Gaps will remain regarding the coverage of software, and
implementation differences in Member States will likely remain. Option 2 would
only partly reduce the consumer detriment, in comparison to Options 3 and 4.
3
Option 3 Option 3 is linked to somewhat higher costs (<0.02% of turnover for business, mostly
neutral for MSAs) and is mostly adequate to address the problems identified. Gaps
will be closed, and implementation differences avoided. However, while safety risks
for EU consumers due to products sold on online platforms could be partly reduced
(and more than under Option 2, as online platforms would have a duty of care), their
mitigation will also depend on continued surveillance of platforms and other factors
(adoption of DSA). Option 3 would considerably reduce the consumer detriment (due
to the loss of value of unsafe products) by 1 billion EUR in the first year of
implementation and by EUR 5.5 billion over the next decade.
1
Option 4 Option 4 leads to higher costs for business (<0.03% of turnover) than Option 3 and is 2
75
mostly neutral for MSAs. It is also considered to be mostly adequate to address
problems, as measures under Options 3 & 4 are almost identical. Reduction of the
consumer detriment would be the same as under Option 3.
As indicated in the Table 23, Option 3 seems to deliver the best results in terms of
meeting the defined objectives while keeping the economic impact limited, and is
therefore the preferred option. Option 4 could deliver broadly the same results in term
of objectives but with higher costs for businesses and administrative burdens (mainly
linked to uncertainties in revising the Regulation (EU) 2019/1020 on market surveillance
which has not entered fully into force yet).
In this IA, the ranking of the options has been done on the basis of the general
comparison of the impacts and not specifically on a Multi-criteria analysis (MCA)133
.
Indeed, the comparison table of the impacts of policy options clearly shows that Options
3 and 4 perform better overall. The evidence was gathered from multiple data sources
and the results were triangulated to ensure the robustness of the methodology. Based on
the analysis, policy Options 3 and 4 would perform equally well except under three
dimensions: administrative simplification, the costs for businesses and the costs for
Member States. Under Option 3, the costs for business would be lower by ~EUR 135
million/year, while under Option 4 the costs for Member States would be lower by ~EUR
3 million/year and could further reduce regulatory complexity and bring simplicity.
Option 3 delivers the best results while keeping the economic impact limited, and is
therefore the preferred option. For this reason, a Multi-criteria analysis (‘MCA’) was
considered to bring more complexity compared to the added value it would have in
determining the preferred option.
Nonetheless, the MCA was tested on the criteria (impacts) in the comparison table and
the results are highly sensitive to the weights attributed: either Option 3 or Option 4 are
obtaining the highest overall score, depending on the factors which are given slightly
more importance. The main reason is that the two options have very similar scores (as
explained above). Moreover, in the MCA, the monetary (absolute) values are
standardised: the 40% difference in costs for business (from EUR 332 million/year to
EUR 197 million/year) is considered the same as the 40% difference in costs for Member
states (from EUR 7 million/year to EUR 4 million/year). This economic impact is
significant and the MCA would not pick it up accordingly.
The actual effect of the different options will depend on the concrete implementation and
enforcement of the initiative at national level and in particular on the level of resources
attributed to the MSAs and the EU budgets allocated to market surveillance. The
level of allocated resources would however not change the ranking of the options, in
particular because some of the actions foreseen can deliver on the objectives without
higher budgets, e.g. deterrent effect of penalties foreseen under Option 3 and 4.
The level of allocated resources would nevertheless impact the overall effectiveness of
the enforcement of the options. Concerns related to the lack of adequate resources in the
competent authorities of some Member States has been expressed by stakeholders during
the consultation process. None of the options envisages to set any obligation on the
133
The multi-criteria analysis is one of the tools presented in the Better Regulation "Toolbox" (Tool #63) to compare
the different policy options. It is a non-monetary approach and its main advantage is that it allows to simultaneously
consider a significant number of objectives, criteria and relations. MCA gives the opportunity to deal with policy issues
characterised by various conflicting assessments, thus allowing for an integrated assessment.
76
amount to be invested on product safety tasks by Member States since it is set at national
level for market surveillance, which remains within remit of national powers. The
legislative options envisage rather mechanisms allowing economies of scale and better
functioning of market surveillance, such as reinforced cooperation among MSAs
including in enforcing measures adopted, more power for MSAs and more effective
measures at their disposal, the possibility to reclaim from the relevant economic operator
the totality of the costs of their activities in case of dangerous products, the introduction
of Union testing facilities which can ease the testing activities for MSAs (some of these
measures are foreseen already in Regulation (EU) 2019/1020 for harmonised products).
An effective enforcement of the different options will be ensured by monitoring of
reported data on enforcement capacities of Member States (e.g. in the context of the
Consumer Scoreboard or via possible reporting obligations) and raising awareness about
market surveillance needs in term of resources and tools. Also under Options 3 and 4 the
arbitration mechanism will allow more harmonised enforcement of product safety.
The efficiency of the different options can be also enhanced by improving the operation
of Safety gate/RAPEX (e.g. by tackling the delays identified in the Evaluation between
the detection of a dangerous product in a Member State and its notification to the Safety
Gate/RAPEX) and facilitating international cooperation, in particular in the context of
the exchanges of information on dangerous products between the Safety Gate/RAPEX
and third countries. The Evaluation found that the procedure for setting up such
arrangements to exchange non-public information from the Safety Gate/RAPEX could be
clarified to cover the different levels of exchanges between the EU system and third
countries (in particular via legal revision under options 2, 3 and 4). Such exchanges can
enhance the efficiency of Member State’s market surveillance actions. Also, enhancing
product safety worldwide will have a positive impact on protection of EU consumers by
limiting the entry of dangerous products to the EU market.
All the options, and in particular the legislative options 2 to 4, including the preferred
Option 3, conform to the principles of subsidiarity, since the new legal provisions relate
to areas where EU action brings added value to ensure level-playing field on the EU
market and higher product safety, while fully respecting the national competences (these
options harmonise the obligations on economic operators and respect powers for MSAs).
All options also conform to the principle of proportionality given that the size of the
identified problem is considerable (high presence of unsafe consumer products on the EU
market and related high consumer detriment) and the costs associated with the different
options are limited. Also the choice of Regulation as Union action is coherent with
satisfactory achievement of the objective to ensure level-playing field and effective and
even enforcement at national level.
8. PREFERRED OPTION
8.1. Preferred option – Option 3
In view of the data and the analysis presented in the previous sections, the preferred
policy option is Option 3. This policy option addresses all identified problems and
objectives in the most effective, efficient and proportionate way, proposing a legal
revision of the GPSD to make it not only fit for purpose now, but also in the future by
improving its ‘safety net’ function.
Concerning the objective related to food-imitating products, the sub-option (b), risk-
assessment approach, is preferred since it is more coherent and proportionate than sub-
option (a), a full ban per se of these products.
77
The operational objectives under Option 3 are as follows: it would add legal clarity
concerning the coverage of risks of new technologies (cyber-security and other risks of
new technologies affecting consumer health) and the role of software for product safety.
It would make most provisions inspired by the Product Safety Pledge legally binding for
online marketplaces and add additional requirements to improve transparency and duty of
care by online marketplaces. It would enhance recall effectiveness by introducing
mandatory requirements on product recalls and customer traceability. It would also better
align the GPSD with market surveillance rules for harmonised products, enhance
traceability and integrate and clarify the rules for the food-imitating products. It would
provide for increased enforcement powers for Member States, an arbitration mechanism
to solve divergent risk assessments and the possibility to adopt delegated acts to improve
traceability systems. It proposes a burden reduction measure by simplifying
standardisation procedures.
In particular, the preferred option will revise the Directive to include inter alia the
following provisions to tackle product safety challenges posed by online sales:
 New obligations for manufacturers and distributors to include in their online
offers the same information like for the physical offers, namely information on
the name and contact details of the manufacturer and of the responsible economic
operator in the EU if applicable, safety information and instructions. The online
market places should also ensure that this information is displayed with the online
listings. This information allows better traceability of products needed for market
surveillance and better safety information in the online sales.
 Establishment of the figure of a responsible economic operator in the EU, in
line with Regulation (EU) 2019/1020 on market surveillance, to tackle the issue
of direct imports from outside the EU. This Regulation applies this obligation
only for certain categories of harmonised products and could be extended to all
consumer products to make sure that consumers and national authorities can
always address an operator based in the EU for any consumer product potential
safety issue.
 New enforcement powers for market surveillance authorities to carry out
online investigations, in line with Regulation 2019/1020 on market surveillance.
For example, the possibility for authorities to carry out inspections using a
covered identify or the power to shut down webpages;
 New product safety obligations for online marketplaces, in line with the
general principles set in the DSA. While manufacturers will remain the main
responsible economic operators for the safety of a product, online marketplaces
could play an important role and exercise a duty of care in relation with their
responsibilities, e.g. making efforts to identify dangerous product offers already
removed from their websites but that keep reappearing. That duty of care would
be different than for distributors as they do not have a physical contact with the
product, so their role will focus on doing their most to ensure that their websites
do not offer dangerous products, and if they do, they cooperate with authorities
for corrective actions.
With regards to software updates, the preferred option aims to shift the responsibility
for the safety of a product from the initial producer to the actor in charge of the update in
case of ‘substantial modification’. If certain criteria are met, in the case of modifications
such as software updates that alter the safety of the product, the responsibility for safety
would shift to the actor in charge of such modification. For example, if an application
78
aimed to improve the efficiency of a battery is downloaded into a device and
consequently the hazards of the device increase, the software developer would become
the responsible actor. That would ensure that actors in charge of substantial modifications
take into account the impact of their changes on a specific product. In any case, this
would not apply for most software updates, such as the download of games that do not
interfere with the safety of a device.
Under Option 3, the revised GPSD would be complementary to the other ongoing
policy initiatives mentioned in part 1. The advantage of clearly integrating aspects of
substantive alternative policy areas into product safety legislation is therefore to ensure a
real safety net for consumers, making possible that that all non-food consumer products
on the EU market are safe. The revised GPSD under the preferred option will address the
convergences between product safety and the other policy areas, but it will not go beyond
those to avoid any overlaps.
Beyond the legal revision of the GPSD, the self-regulatory instruments such as the
Product Safety Pledge could be strengthened to complement the legal framework. The
Pledge could be further used to continue the operational cooperation between online
marketplaces and MSAs and to include the commitments that would not become binding,
as well as potential new commitments. This way the Pledge would continue to play a
complementary role to the legal framework.
Other voluntary cooperation actions with online marketplaces could be explored, such as
the actions taken in the beginning of the COVID-19 crisis where the online marketplaces,
upon a call for cooperation from the Commission, acted against online scams and new
dangerous products related to the pandemic.
8.2. REFIT (simplification and improved efficiency)
GPSD being part of REFIT, the report has analysed how the current legal framework
could be simplified, improve the efficiency and decrease administrative burden. The
following actions under Option 3 should lead to such higher efficiencies:
Table 24: REFIT Cost savings under the preferred option
REFIT Cost Savings – Preferred Option – Option 3
Description Amount Comments
Alignment to market surveillance procedures for
harmonised products
would simplify the market surveillance rules
Cost savings for MSAs: around
EUR 0.7 million per year across
the EU.
Cost saving for businesses are
included in the line below on cost
savings linked to more uniform
implementation of market
surveillance rules.
Benefits are mainly for
MSAs and businesses
active in both
harmonised and non-
harmonised product
areas.
Conversion to a Regulation
would ensure a common application of product safety rules
and avoid inefficiencies and regulatory costs/burdens related
to the inconsistent implementation of the GPSD across the
EU
Lower regulatory burden and
costs
Cost savings for businesses :
around EUR 59 million
annually (around EUR 34
million saved by EU SMEs and
Benefits for all
stakeholders (reduced
burdens and costs for
businesses and MSAs
and better enforcement
and product safety for
79
26 million EUR saved by EU
large businesses)
consumers)
Simplifying the standardisation procedure under the
GPSD would decrease the administrative burden
Simplify and reduce regulatory
costs
Beneficial to all
stakeholders since
standards could be
referenced faster
Clarification of scope and definitions regarding the
application of consumer product safety rules to new
technologies
would lead to higher legal certainty regarding the application
of consumer product safety rules to new technologies, which
will likely reduce the costs relating to businesses’ (especially
SMEs’) efforts to design innovative, safe and cyber-secured
products.
Reduced regulatory costs Benefits for businesses
producing new
technology products
and consumers of
these products because
of legal clarity and
better safety.
Repealing Directive 87/357/EEC and integrating rules on
food-imitating products into the revised GPSD would
simplify the product safety legal framework and increase
coherence in implementation by Member States
Lower regulatory burden and
costs
Benefits for producers
of food-imitating
products, for MSAs
and consumers
Arbitration mechanism on diverging risk assessments
would lower the regulatory burden for MSAs by helping to
resolve disputes on risk assessments
Reduced regulatory burden Benefits for MSAs and
consumers
Potential future introduction of improved digital
solutions for product traceability through delegated acts
The Study identified e-labelling solutions for traceability
information as potential complementary measure to increase
efficiency of product safety market surveillance.
Reduced administrative burdens Beneficial for MSAs,
businesses and
consumers
Digital interlinks between existing market surveillance
systems at EU and national level (including customs)
similarly to Regulation (EU) 2019/1020 will make the
market surveillance more simple and efficient through
connecting Safety Gate/RAPEX with the EU Customs
database
Lower regulatory burdens For MSAs, customs
authorities
Beyond these simplifications and higher efficiencies, this initiative endeavours to keep
regulatory burdens to the minimum necessary both for businesses and Member States to
what is strictly needed to ensure consumer protection against unsafe products.
To avoid any legal uncertainties and related burdens, the revised GPSD would avoid any
overlaps between lex generalis and lex specialis, by defining its scope. Also the ongoing
work related to product safety under other initiatives has been and will be duly taken into
account to avoid overlaps and overregulation.
9. HOW WILL ACTUAL IMPACTS BE MONITORED AND EVALUATED?
The Commission will monitor the implementation of the revised GPSD, if adopted along
the preferred option, with regards to the achievement of policy objectives identified in
this Impact Assessment in order to be able to assess its effectiveness in the future
evaluation. A commitment to evaluate the impacts of the new legislative act, if proposed,
will be included in the draft proposal. The Commission will start monitoring the
implementation of the revised GPSD after the entry into force of the initiative. The
indicators proposed to monitor the achievement of policy objectives identified in this
Impact Assessment are presented below.
80
The monitoring will be done mainly by the Commission, based on regular EU-wide
consumer surveys and data provided by businesses and MSAs. The monitoring and
evaluation will be done on the basis of existing data sources where possible to avoid
additional burdens on the different stakeholders. The new legislative act, if proposed, will
set out reporting obligations for Member States. This reporting will be done on the basis
of enforcement indicators which will be further defined by a study. The Commission has
already identified some gaps concerning enforcement indicators, and will launch a study
to establish a new set of enforcement indicators134
.
Tables 25 and 26 below provide an exhaustive list of monitoring indicators. A
methodological study for the design of enforcement indicators is ongoing to identify
which enforcement indicators are the most suitable to measure the achievement of the
different objectives and on which Member States could report in practice so that the
Commission can receive reliable and comparable data for the next evaluation of this
initiative. Through this study the Commission could complement the list of the most
suitable indicators for the monitoring system.
The Commission has already mapped existing sources of injury information and looked
into the possibility of establishing a EU-level injury database to help the implementation
of the product safety legislation135
. It is currently assessing the costs and benefits or
setting up such a EU wide injury database (via coordinated actions with Member States).
Table 25: Monitoring indicators for the main policy objectives
Policy objectives Monitoring indicator Data source Data collected
already?
Actors
responsible
for data
collection
Product safety Number of unsafe products on the
market
Safety Gate/RAPEX
gives a proxy
Yes Commission
Consumer detriment Future study (data for
past available from the
GPSD Study)
Not in a
recurrent way
Commission
via study
Consumer trust in product safety
and experience of product-related
injury
Consumer Conditions
Scoreboard
Yes Commission
via regular
surveys
Proper functioning of
Single Market
Number of unsafe products on the
market
Safety Gate/RAPEX
gives a proxy
Yes Commission
Safety net function, also
new technologies
Number of unsafe new technology
products
Safety Gate/RAPEX
gives a proxy
Yes Commission
Consumer concerns about safety of
IoT products
Eurostat ICT survey,
Consumer Markets
Scoreboard
Commission
via regular
consumer
surveys
Product safety in the
online sales
Number of unsafe products found
online
Safety Gate/RAPEX
gives a proxy
Yes Commission
Product recalls more
effective and efficient
Number of recalls and recalled
products
Safety Gate/RAPEX Yes Commission
134
For the future reporting obligations under the revised GPSD, the Commission will consider the reflections and work
engaged in the harmonised area on the monitoring and reporting in order to aim for consistency and to avoid any
duplication or unnecessary burden on national administrations in the collection of relevant data and information.
135
https://ec.europa.eu/consumers/consumers_safety/safety_products/rapex/alerts/repository/content/pages/rapex/docs/
Final_JRC_Report_Injury_and_acccident.pdf
81
Market practices regarding product
recalls and product registration
schemes/loyalty programmes
(Coordinated) market
surveillance activities
Not in a
recurrent way
Member States,
Commission
Self-declared data on recall
participation and product
registration and on exposure to
recall information
Consumer Conditions
Scoreboard
Yes Commission
via regular
surveys
Hard data on recall participation Monitoring data to be
collected by economic
operators
No Economic
operators
(Member
States will be
able to request
this
information)
Enhanced market
surveillance and ensure
better alignment with
harmonised products
Enforcement indicators as defined
by the study commissioned by the
EC mentioned above
National sources, to be
defined by the
commissioned Study
No or only
partially
Member States
based on
indicators
defined by the
study
Safety of food imitating
products
Number of unsafe food-imitating
products
Safety Gate/RAPEX
gives a proxy
Yes Commission
Number of disputes on risk
assessment of these products
between Member States
Safety Gate/RAPEX Yes Commission
Table 26: Monitoring indicators for the operational objectives – Option 3
Operational objectives Monitoring indicator Sources of data and/or
data collection methods
Data collected
already?
Actors
responsible
for data
collection
Clarify coverage of risk and
products linked to new
technologies
Number of questions raised on
the applicability of new
technology products
RAPEX contact points
Wiki
Yes Commission
Clarify the application to
software
Number of notifications related
to safety issues raised by
software
Safety Gate/RAPEX Yes Commission
Making most provisions of the
Pledge legally binding for all
online marketplaces
KPIs and qualitative data (idem
Pledge)
Monitoring reports of the
Pledge
Yes, to be
reinforced
Online
platforms -
Pledge
signatories
Providing all safety
information online that is also
required to be provided offline
Number of cases where
diverging level of information
offline/online found
Regular checks in the
context of (coordinated)
market surveillance
activities
No Member
States,
Commission
Introducing a duty of care for
online marketplaces
Number of cases where duty of
care not respected
Regular checks in the
context of (coordinated)
market surveillance
activities
No Member
States,
Commission
Introducing mandatory
requirements for recalls
Number of cases where new
recall provisions not fulfilled
Regular checks in the
context of (coordinated)
market surveillance
activities
No Member
States,
Commission
Introducing mandatory
requirements for customer
traceability
Number of cases where new
recall provisions not fulfilled
Regular checks in the
context of (coordinated)
market surveillance
activities
No Member
States,
Commission
82
Align with market
surveillance and traceability
rules for harmonised products
- Legal analysis - Commission
Requiring an economic
operator in the EU
Number of cases where this
economic operator is missing
Regular checks in the
context of (coordinated)
market surveillance
activities
No, can be
included in the
Pledge
monitoring
Member
States,
Commission
, online
marketplaces
Simplifying the
standardization procedures
Average duration of the
standardisation procedure
Observed durations Yes Commission
Strengthening the enforcement
powers of MSAs
Number of enforcement
measures adopted at national
level
Implementation reports
of Member States
Yes (for past) Commission
based on
Member
States input
Level of penalties foreseen at
national level
Legal analysis Yes (for past) Commission
based on
Member
States input
Introducing the arbitration
mechanism
Number of disputes on risk
assessment solved
Safety Gate/RAPEX Yes Commission
Incorporation of provisions on
the food-imitating products in
the new legal act
Number of disputes on risk
assessment of these products
between Member States
Safety Gate/RAPEX Yes Commission
83
Annex 1: Procedural information
10. LEAD DG, DECIDE PLANNING/CWP REFERENCES
This Staff Working Document was prepared by the Directorate-General for Justice and
Consumers (DG JUST).
The Decide reference of this initiative is PLAN/2019/6283 Review of the general product
safety directive -Proposal for a regulation on general product safety.
This includes the Impact Assessment report as well as the GPSD Evaluation Report, in
annex to this Impact Assessment.
11. ORGANISATION AND TIMING
 An Inter Service Steering Group (ISSG) has been established to support the work
of DG JUST on the evaluation and impact assessment of this initiative set up.
 DGs participating in this ISSG: SG, LS, CNECT, COMP, ENV, GROW, JRC
OLAF, SANTE, TAXUD
 This GPSD ISSG held 5 meetings times (one informal meeting on 14/02/2020
and four formal meetings on 12/06/2020, 08/10/2020, 18/11/2020 and
07/12/2020). DG JUST consulted the ISSG on the different steps of this initiative:
Roadmap/Inception Impact Assessment, Consultation strategy, Open Public
Consultation questionnaire, the study underlying the evaluation and impact
assessment (ISSG provided comments on all study steps and reports) and finally
on the draft Impact Assessment report.
 Publication in EUROPA of the Roadmap on the evaluation/Inception Impact
Assessment, 30 June 2020
 Launch of the Open Public Consultation on the combined Roadmap/Inception
Impact Assessment, 30 June 2020 - 6 October 2020 (14 weeks).
12. CONSULTATION OF THE RSB
The RS has been consulted on the Impact assessment report and issues a ‘positive
opinion with reservations’ on it.
The two tables below present the elements of the RSB opinion and how the report has
been updated to take them into account:
Main issues raised by the RSB in its opinion and related updates
(1) The report does not sufficiently explain how the horizontal and sectoral elements of the product
safety framework interact with each other in a coherent manner. The fall-back function of the
GPSD as safety net is not sufficiently elaborated. The links to recent safety related sectoral
initiatives are not sufficiently clear.
Related updates:
 The report clarifies the overall structure of the EU Product safety framework in more detail by
explaining the interlink between the GPSD and the other sectorial and harmonised legislation at
84
EU level and the role of GPSD as safety complementing the other legislation to ensure the safety
of EU consumers for any product now and in the future. The report includes more graphical
presentation of the general products safety framework.
 The interlink with the recent and ongoing initiatives, in particular those on digital platforms,
cybersecurity, circular economy and artificial intelligence, has been better explained.
(2) The available policy choices are not sufficiently clear. The report presents only a limited set of
options and lacks detail on the content of the measures contained therein. It does not explain
sufficiently why some options are discarded.
Related updates:
 The report better explains the structure of the options and how they address the objectives but in
a different level of depth.
 The discarded options has been further developed in the revised IA report. In particular, the
report explains why some options already analysed and disregarded in the IA in 2013 can still be
disregarded now (e.g. to have different safety requirements for harmonised and non-harmonised
products, extending the scope to services, abolition of the general product safety requirement).
Also the report explains that some options have been considered and disregarded because of lack
of proportionality, e.g. higher traceability requirements for all products.
 The report explains that other stakeholders did not have raised any other new real alternatives in
the consultation process until now.
(3) The report does not explain in a convincing manner why the estimated costs for business under
the integration option (option 4) are much higher than those of the full legal revision option,
although in terms of substance the options seem very similar.
Related updates:
 Under Option 4 the businesses reported higher costs to our contractor. The report admits that
these costs might be inflated and a clear disclaimer has been included at this respect in the
revised report.
Specific improvements requested by the RSB How the RSB comments have been addressed in
the revised IA report
(1) The report should explain upfront how the
horizontal and sectoral elements of the product
safety policy framework fit together and how the
GPSD general safety net fallback functions. It
should better explain the coherence with Regulation
2019/1020 on market surveillance, and the
relevance of the recent changes to that Regulation
for the GPSD. It should better describe the links to
recent initiatives, such as on digital platforms,
cybersecurity, circular economy and artificial
intelligence.
Cf main issue 1
Explained the interaction of the GPSD with other
EU legislation and initiatives relating to product
safety
(2) The report should better present the scope of the
initiative, especially on which consumer products
are covered. In this sense, it may help to include a
diagram presenting the product safety regulatory
framework. The safety concept needs elaboration. It
is not clear what types of risks and damages it
covers, ranging from health to cyber issues. The
report should detail the specific mechanisms it will
use to identify future product risks to function as a
safety net.
Cf main issue 1
Included a diagram presenting the product safety
legal framework
Better explained the scope of the GPSD and the
proposal
Better explained the concept of risk in particular in
the context of new technologies (cybersecurity
risks) and how these risks could be assessed (e.g.
when substantial modification of the product)
(3) The report should reinforce the problem analysis
Gaps and deficiencies better explained in the
problem definition.
85
to better reflect the deficiencies and gaps the
initiative wants to solve. It should clarify to what
extent self-regulatory measures under the Product
Safety Pledge have been effective and what lessons
can be learned. It should explain to what extent the
Pledge helped to get information on emerging risks
of new technologies and improved recalls.
Achievements and limitations of the Product safety
Pledge better spelled out as well as lesson learnt.
(4) The range of options analysed should be better
linked with the specific objectives and the problems
the initiative aims to tackle. The report should
provide more detail on the content and functioning
of the proposed policy measures under the various
options. It should explore whether there are
alternative policy choices to the substantive
measures presented for each problem area under the
preferred option. It should expand on how the self-
regulatory elements could be strengthened. It should
provide more details about discarded options and
the reasons for their exclusion from the analysis.
Cf main issue 2
The option packages presented in more detail,
beyond the summary table.
The alternative options, which have been discarded
have been included in the IA report. But no new
policy options.
The further use of self-regulatory instruments such
as the Product Safety Pledge after the adoption of
the initiative has been explained.
Examples of other voluntary cooperation actions
with platforms, e.g. during COVID crisis, have been
introduced.
(5) The full integration option comes with
substantial additional costs as regards market
surveillance for business although there seem to be
no real substantive differences on new regulatory
obligations, compared to the full legal revision
option. The report should review the robustness and
reliability of the costs estimates provided in the
support study given their importance for the overall
comparison and ranking of options.
Cf main issue 3
The report reviewed the underlying cost data under
Option 4 and provides the necessary disclaimers.
(6) The report should provide greater clarity on how
this initiative will tackle safety issues related to
consumers’ online purchase from third countries as
well as software updates. It should explain how the
sanction regime would work under the different
options and clarify whether alternatives with
different deterrence effects can be assessed. It
should better describe how effective enforcement of
the options will be ensured.
A detailed presentation of the measures under the
different options (cf point (4)) includes now these
clarifications. The report presents specific
explanation of measures to tackle safety issues
related to consumers’ online purchase from third
countries as well as software updates. The sanction
regimes have been explained under the different
options and alternatives analysed.
The report now elaborates on ways how to ensure
better enforcement: e.g. introduction of the
arbitration mechanism, collecting data on
enforcement capacities of Member States in the
context of the Consumer Scoreboard to raise
awareness.
(7) The REFIT aspect should be clarified,
explaining how the initiative would endeavour to
keep regulatory burdens to the minimum necessary.
More information is needed on how overlaps
between lex generalis and lex specialis would be
prevented.
Some more technical comments have been sent
directly to the author DG.
The report contains now a reinforced explanation on
the simplifications, avoiding overlaps and
overregulation and how the initiative would keep
the minimum necessary regulatory burden while
ensuring the objectives.
Other technical comments (e.g. providing summary
table of the costs of different options) have be taken
into account.
86
13. EVIDENCE, SOURCES AND QUALITY
Studies commissioned or supported by the European Commission
 Study to support the preparation of an evaluation of the General Product Safety
Directive as well as of an impact assessment on its potential revision, Civic
consulting, March 2021
 Study for the preparation of an Implementation Report of the General Product
Safety Directive, Civic consulting, July 2020
 Study on the assessment of the opportunities for increasing the availability of EU
data on consumer product-related injuries, European Commission’s Joint
Research Centre’s, May 2020
 Behavioural Study on strategies to improve the effectiveness of product recalls,
LE Europe, June 2021
 Survey on consumer behaviour and product recalls effectiveness, April 2019
 Implementation of the new Regulation on market surveillance: indication of
origin, VVA Europe, May 2015
External Expertise
 Consumer Safety Network (CSN)
 Sub-Group on Artificial Intelligence, connected products and other new
challenges in product safety to the Consumer Safety Network
Selective bibliography
 Bernstein A. (2013), ‘Voluntary Recalls’, University of Chicago Legal Forum, 1:
394 ff., available at: http://chicagounbound.uchicago.edu/uclf/vol2013/iss1/10
and Jacoby J. (1984), ‘Perspectives on Information Overload’, Journal of
Consumer Research
 OECD (2020)- ‘E-commerce in the time of COVID-19’, available at
http://www.oecd.org/coronavirus/policy-responses/e-commerce-in-the-time-of-
covid-19-3a2b78e8/#biblio-d1e705
 OECD (2018), ‘Enhancing Product Recall Effectiveness Globally’, available at
https://www.oecd-ilibrary.org/science-and-technology/enhancing-product-recall-
effectiveness-globally_ef71935c-en
Other Sources
 Eurostat
 European Injury Database (IDB)
 Safety Gate/RAPEX
 WHO CHOICE
87
Annex 2: Stakeholder consultation
1. Consultation strategy
The impact assessment (IA) for the revision of the General Product Safety Directive
2001/95/EC (GPSD) was supported by the following consultation activities:
- public consultation on the Inception IA and roadmap;
- an open public consultation (OPC);
- stakeholder workshops;
- ad-hoc contributions and targeted consultations with Member States (MS) and
other stakeholders.
The objective of these consultations was to collect qualitative and quantitative evidence
on all key elements of the IA, from relevant stakeholder groups and the general public.
The stakeholder groups identified as relevant are:
- Consumers and consumer organisations,
- Businesses and business organisations,
- Member States market surveillance authorities,
- Other product safety experts.
The consultations were publicised via social media posts, emails to existing networks
(including in the Safety Gate- RAPEX weekly update newsletter on dangerous products),
regular meetings of the expert groups and networks, as well as in speeches delivered by
high-level Commission officials.
2. Overview of consultations
a) Consultation on the combined evaluation roadmap and inception impact
assessment
The consultation on the combined evaluation roadmap and Inception IA took place
between 23 June and 1st September 2020. 44 answers were received: 20 from business
associations, 9 from company/business organisation, 5 from consumer organisations, 2
from non-governmental organisations (NGOs), 3 from citizens, 2 from public authorities,
1 from a trade union, 2 other, and additional 3 feedback were not relevant (because the
questionnaire was empty or contained almost no information).
Most of the stakeholders supported the GPSD revision, almost half of them being in
favour of the full revision (options 3+4).
Option 0 (status quo) 1 stakeholder
Option 1 (better implementation and enforcement) 7 stakeholders
Option 2 (targeted revision) 10 stakeholders (some willing only changes
regarding the online dimensions)
Option 3 (full revision) 12 stakeholders
Option 4 (integration of legal instruments) 9 stakeholders (mainly consumer organisations)
88
The feedback on the IIA (summarised in the next section) has included the objectives as
well as the set of options to be analysed in the IA.
b) Open public consultation on a new Consumer agenda
The open public consultation (OPC) ran between 30 June 2020 and 6 October 2020, in
order to gather views of the public on the ‘New Consumer Agenda’ as well as on three
legislative proposals in the area of EU consumer policy, including the review of the
GPSD. The public questionnaire available in the 24 official EU-languages was targeting
a wide range of stakeholders, both the general public and relevant organisations and
institutions.
The section on the GPSD in the public consultation included questions related to both the
evaluation of the GPSD and the IA for its revision. The number of respondents that
answered at least one question in this section is 257. The majority of respondents were
business associations and EU citizens (each 26%), followed by company/business
organisations (15%). Other respondents included public authorities (11%), consumer
organisations (8%), non-governmental organisations (7%), academic/research institutions
(3%), non-EU citizens (1%) and other respondents (3%).
The full report on the results of this OPC has been published on the Have Your Say
page.136
Annex 13 contains the summary of the replies of the OPC GPSD part.
c) EU Workshop on strategies to maximise the effectiveness of product recalls
The EC organised on 23 October 2019 a workshop on the effectiveness of recalls in order
to take stock of existing market practices and regulatory approaches, and identify
possible new avenues to maximise recall effectiveness. 68 participants took part in the
workshop, including regulators from around the world, representatives of international
organisations, consumer organisations, industry and academics. The workshop was
divided into three thematic sessions, focusing on i) strategies to facilitate direct consumer
contact, ii) strategies to increase consumer response to recalls and iii) roles and
responsibilities in the recall process.
d) 2020 International Product Safety Week (IPSW)
The EC organised the International Product Safety Week on 9-12 November 2020137
.
This event is the largest gathering of product safety experts from all around the world and
it takes place every two years. More than 500 participants from 73 countries registered
for the 2020 online edition, including regulators, businesses, or consumer organisations.
Two sessions were organised on topics of interest for the revision of the GPSD, namely
136
https://ec.europa.eu/info/law/better-regulation/have-your-say/initiatives/12466-General-Product-Safety-Directive-
review/public-consultation
137
https://ec.europa.eu/info/events/international-product-safety-week-2020-2020-nov-09_en
89
on traceability and recalls. Input on these two topics was collected both from a wide
diversity of panellists and also from the audience via interactive online surveys.
e) European Consumer Summit 2020
The EC organised the European Consumer Summit on “Consumers in the Green and
Digital Transition: Challenges and solutions for a new consumer policy”, which took
place on 30-31 January 2020138
. This event gathered over 500 stakeholders including
policymakers, national enforcement authorities, academia, consumer and business
organisations, and youth representatives, from all Member States. Sessions were notably
organised on “safety and consumer protection in online trade” and “Artificial Intelligence
–a consumer-centric approach”. Input was also gathered from the audience via interactive
surveys.
f) Workshops on online marketplaces and product safety
The EC held a number of workshop sessions related to online marketplaces and product
safety on 8, 10, 13 and 17 July 2020. The objective of the workshops was to gather up-to-
date information on the state of play concerning the main challenges in addressing the
sale of illegal goods online. It focused in particular on measures and good practices from
marketplaces and the cooperation with authorities and relevant third parties. Input
gathered through these sessions aimed at feeding into the revision of the e-commerce
Directive and of the GPSD. The workshops gathered more than 60 participants each,
covering a very wide range of stakeholders, such as online marketplaces, retailers,
industry associations, consumer organisations and MS authorities. Annex 14 contains the
minutes from these workshops.
g) Ad hoc contributions and consultations
Input from a variety of stakeholders (Member States authorities, consumer organisations,
businesses, business organisations…) has been collected, notably via extensive
consultations, in the framework of the dedicated study for the evaluation and impact
assessment of the Directive as well as the study supporting the preparation of the
implementation report of the Directive.
Input has also been received from the following stakeholders via ad hoc contributions
and/or ad hoc consultations: consumer organisations, businesses, business organisations,
national chamber of commerce, trademark association. Further bilateral discussions were
also held with Member State authorities.
Stakeholders’ input was also collected through the expert group ‘Consumer Safety
Network’ and the Sub-Group on Artificial Intelligence (AI), connected products and
other new challenges in product safety.
A workshop of the Consumer Safety Network expert group was organised on 19
November 2020 to discuss the study results supporting the implementation report of the
138
https://ec.europa.eu/info/events/european-consumer-summit-2020-2020-jan-30_en
90
Directive, the evaluation and the impact assessment, as well as specific topics of interest
(penalties, operator-based market surveillance and cooperation with customs authorities).
The consultations for the evaluation and impact assessment of the revision of the General
Product Safety Directive, which forms part of the wider consultation strategy on the New
Consumers Agenda, have also benefitted from the consultations on other ongoing
initiatives of the European Commission, namely linked to the White Paper on Artificial
Intelligence and the proposal for a Digital Services Act.
3. Summary of the results of the consultations
In the following summary, "consumer representatives" means national and EU-level
consumer organisations, "business representatives" includes national and EU-level
business organisations,
"MS authorities" includes national market surveillance authorities and government
authorities in charge of product safety.
1) Preserving the safety net role of the GPSD
The overall feedback among all stakeholder groups is that the GPSD is a useful
legislation and its safety net principle should be preserved. Consumer representatives
also emphasised the precautionary principle being a key pillar of the product safety
legislation.
However, a large majority of respondents expressed that current EU safety rules for non-
food consumer products covered by the GPSD could be improved in specific areas to be
more adequate to protect consumers (71% in OPC).
2) Tackling the challenges posed by new technologies
Stakeholders acknowledge that new technologies raise many challenges. Different
stakeholders favoured different approaches to tackle these. In the OPC, almost half the
respondents considered the safety of products involving new technologies not to be
adequately regulated (47%). The majority of respondents agree that the definition of a
product in the GPSD should specifically encompass software incorporated into the
product, even in case the software is downloaded after the product has been sold (56%).
About a quarter of respondents considered that only software already installed into the
product when sold should be included.
Almost all respondents support the introduction of a requirement for products that could
be modified via software updates/download or machine learning to remain safe
throughout their lifetime (very important for 72%, rather important for 24%). A clear
majority of respondents also favoured safety obligations for manufacturers of products
incorporating AI applications at the design stage and also during the lifecycle of the
91
product (75%), whereas only 9% of respondents expressed that the obligations should be
limited to the design stage.
In the consultations, consumer representatives and several MS authorities expressed their
support in extending the definition of ‘safety’ to include (cyber)security aspects that have
an impact on safety. In the consultation on the Roadmap/Inception IA, technology-
oriented businesses showed more reluctance regarding the inclusion in the GPSD of new
technologies and new risks related to them as they point out the possible overlap with
other pieces of legislation. Their preference is that the GPSD remains a technology-
neutral tool, and that risks linked to new technologies are covered in other more specific
pieces of legislation.
3) Addressing safety issues associated with products sold online
The issue of products coming directly or via online platforms from outside the EU was a
recurrent issue mentioned in the consultations. Businesses and business representatives
stress the level-playing field angle and they point out that currently, many EU retailers
suffer from unfair competition in relation to operators based in third countries. Consumer
representatives call to close loopholes regarding international e-commerce. Consumer
representatives and other stakeholders also mentioned the issue of dangerous products
reappearing on online marketplaces. MS authorities stress the difficulty to control
products coming from third countries and to take enforcement actions against economic
operators outside the EU.
In the OPC, the majority of respondents expressed that they were aware of problems
associated with online marketplaces having no direct legal obligations for the safety of
products hosted on their platform by sellers (53%). When asked about the role that online
marketplaces should play regarding the safety of products offered on their websites, the
most commonly supported notions were that they should remove dangerous products
listed on their website when notified (77%), that online marketplaces should prevent the
appearance of dangerous products, including their reappearance once they have been
removed (66%) and that they should inform sellers of their obligation to comply with EU
rules on products (64%). A lower number of respondents thought that online
marketplaces should do a cursory check on all products offered on their website to
identify products that likely do not comply with safety rules (42%).
Views diverge between stakeholders when it comes to the obligations of online
marketplaces:
- Consumer representatives are in favour of strengthening their responsibilities
across the supply chain.
- Businesses’ views are heterogeneous, in particular:
o Retailers argue that online marketplaces play a key role in the supply
chain, and therefore they should have a corresponding responsibility.
92
o Online marketplaces responding to the OPC expressed that they would
also accept some of the Product Safety Pledge’s provisions being binding,
but not more.
According to the participants of the session on the safety of product sold online at the
EU Consumer Summit 2020, voluntary commitments are not sufficient (89% of
respondents),enhanced responsibility for online marketplaces are needed, as well as
better enforcement (for instance regarding website blocking by authorities).
4) Improving market surveillance rules and enforcement
Stakeholders from all categories are in favour of aligning market surveillance rules
between harmonised and non-harmonised products. Some stakeholders insisted on the
fact that the GPSD relies too much on ex post market surveillance, and more action on ex
ante prevention should be done at different levels. Member States’ authorities lack of
resources for market surveillance was also repeatedly mentioned in consultations. Other
challenges mentioned in the OPC included the insufficient number of control checks
carried out, including by customs (29%), insufficient cooperation between market
surveillance authorities in the EU (27%), and divergences between authorities in the
assessment of product risks (19%).
Regarding the introduction of a “responsible person” in the revised GPSD, a large
majority of respondents in the OPC considered that products covered by the GPSD
should only be placed on the EU market if there is an economic operator established in
the EU responsible for product safety purposes (70% in favour). Consumer
representatives support the introduction of such “responsible persons” in the EU, in line
with Regulation 1020/2019, but stress that their responsibilities should be strengthened.
Consumer representatives also call for increased international cooperation on market
surveillance, product safety, customs and enforcement.
5) Revision of the standardisation process
Stakeholders are mostly in favour of simplifying the standardisation process to develop
new standards. Consumer organisations suggested the Commission Decision on safety
requirements to become legally binding.
6) Including food-imitating products in the scope of the revised GPSD
Most stakeholders are in favour of incorporating the food-imitating legislation into the
GPSD. In the OPC, a large majority of respondents expressed that products which
resemble foodstuff should be incorporated into the general product safety legal
instrument (69%). In the consultation on the inception IA, including this element in the
product safety risk assessment was the favoured approach. No support was expressed to
the full ban of food-imitating products. Consumer representatives also suggest including
risk assessment criteria regarding the child-appealing aspect of products in the revised
GPSD.
93
7) Improving the framework for product recalls
Stakeholders’ opinions differ regarding the need to tackle recalls in the revised GPSD. In
the consultation on the inception IA, some stakeholders stressed that this issue is mostly
linked to consumers’ behaviours or to rogue traders. In the OPC, approximately a fifth of
respondents regarded as problematic that there were no specific requirements for product
recalls (22%). However, a consumer representative pointed out that this issue might not
appear very important precisely because consumers are not sufficiently aware about
recalls. Consumer representatives and many MS authorities are in favour of addressing
them in the legislation or through guidance.
The crucial importance of using direct communication to consumers for recalls has been
repeatedly stressed in the consultations, whenever it is possible, for instance because the
product was registered, bought online, or bought with the use of loyalty card. Direct
notification was also judged as being by far the most effective channel to spread recall
information (according to 92% of respondents in the survey held during the IPSW
session on product recalls).
There was also a general agreement among respondents that companies should be
obliged to use customer data at their disposal to contact affected consumers directly in
case of recalls (60% ‘strongly agreed’ and 32% ‘tended to agree’, IPSW survey). In this
regard, businesses and authorities call for more clarity on data protection aspects and
compliance with the General Data Protection Regulation. Several stakeholders mentioned
that consumers should be able to choose to receive safety notifications only (when
registering a product or subscribing to a loyalty scheme).
Consumer representatives and authorities also stressed that online marketplaces should
play a facilitating role in recalls, taking advantage of the channels and systems they have
already put in place to communicate with both consumers and sellers. In the OPC, more
than half of the respondents agreed that online marketplaces should inform consumers
when a dangerous product has been removed from the marketplace (55%). The potential
benefits linked to connected products were also stressed: when a connected product itself
is subject to a recall, this technology can be used to warn consumers or, if they fail to act,
switch off the product or reduce its performance.
Participants in the workshop on recall effectiveness and IPSW session on recalls agreed
that a recall notice should be easy to read, straight to the point and clearly describe the
risk and action to take. Several stakeholders stressed that some key elements and ground
rules, applicable to all recall notices, should be standardised and made compulsory.
8) Improving traceability along the supply chain
A large majority of stakeholders agree that the system of product traceability should be
reinforced in the GPSD (82% in favour in the OPC). The elements that most respondents
in the IPSW online survey wanted to see as mandatory is the type, batch or serial number
or other element allowing the identification of the product (85%). This was followed by
the manufacturer’s name (81%), the importer’s name (63%) and the trademark (59%).
94
Most respondents also favoured the possibility to have traceability information available
in an electronic format only, for instance via a QR code (55%). However, the consumer
representative noted that vulnerable consumers who do not necessarily have the capacity
to read a QR code should not be left aside.
Stakeholders also appear to be in favour of the introduction of a “one up one down”
traceability requirement in the revision of the GPSD, whereby economic operators have
to keep information about the upstream and downstream economic operator in the supply
chain (93% in favour in the slido survey conducted at the recall session of the IPSW).
Consumer representative explained that a differentiation between durable and non-
durable goods would be relevant when it comes to the number of years during which
such information should be kept. Moreover, the role of online marketplaces in improving
product traceability was also stressed, notably that they should check that traceability
information is available before listing a product.
Stakeholders would also welcome the possibility to set up additional traceability
requirements for the components of the product (74% in favour, IPSW survey).
Consumer representative stressed that because of the growing importance of the circular
economy, traceability is increasingly important not only for the product itself but also for
its components.
9) Better tackling of chemical risks
Consumer representatives consider that the revised GPSD should play the role of a real
safety net for chemicals in all products, by setting detailed chemical safety criteria for
non-harmonised consumer products through implementing measures.
10) Addressing counterfeit products
Brand owner organisations stressed that the GPSD should be amended to tackle
counterfeit unsafe products.
4. Use of the results of the consultations
The results from the consultation activities have been incorporated throughout the IA
from the problem definition to possible options and their impacts. The consultations
have confirmed the relevance of the five objectives identified in the inception impact
assessment, as well as the elements and the options proposed to answer these objectives.
The results have also been taken into account in the Evaluation (Annex 5) for the
assessment of the GPSD against the five evaluation criteria, to reflect the different views
of the stakeholders.
Moreover, some elements raised in the consultations will be included as accompanying
implementation measures of this initiative, notably regarding international cooperation:
consumer representatives have called for good examples of cooperation between
regulators, such as the EU-Canada arrangement on product safety alerts, to be
95
replicated with other countries. This will be included in the larger implementation
strategy.
Regarding the inclusion of the fight against unsafe counterfeit products in the scope of
the revision of the GPSD, this issue was duly taken into consideration, but was not
included in the scope of the IA. Indeed, counterfeit products are already addressed by EU
legislation, and unsafe products are covered including by the GPSD, regardless of their
authenticity. Even though counterfeit products can pose safety risks, the safety of a given
product has to be analysed based on a risk assessment.
96
Annex 3: Who is affected and how?
1. PRACTICAL IMPLICATIONS OF THE INITIATIVE
In case the preferred option 3 is retained, the initiative will have practical implications
on both the economic operators handling products covered by the GPSD (it can be
producer, distributor, importer, online marketplaces or fulfilment house) and market
surveillance authorities in the Member States.
Under Option 3, businesses will have additional requirements: manufacturers and
importers will have additional traceability requirements and online traders (online
retailers, distributors or marketplaces) will have the requirement to provide the same
information online, which is available offline (traceability and other mandatory safety
information). This would imply for businesses setting up internal mechanisms to ensure
they comply with these traceability and transparency provisions. Comprehensive
additional requirements would apply in the context of recalls for all businesses, but these
additional requirements would have practical implications only to those companies that
have actually brought unsafe products onto the market. Online marketplaces will also
have to make sure they set up internal mechanisms to comply with most of the Product
Safety Pledge’s provisions and the duty of care responsibility, which would also apply to
them. In addition, companies selling in the Single Market from outside the EU will have
to set up arrangements to ensure that the products sold in the EU have a responsible
economic operator. Finally, enhanced penalties would have an impact on the non-
complaint businesses.
A broadening of market surveillance responsibilities, new competences and a greater
need for internal and external resources respectively to perform the market surveillance
(e.g. the new tools for market surveillance online broaden the possibilities of MSAs and
may require additional resources and skills), will impact market surveillance
authorities in Member States. However, the provisions on MSAs’ powers are largely
aligned to the existing market surveillance provisions applicable to harmonised products
under Regulation (EU) 2019/1020. Therefore, these provisions are not new to MSs and in
particular for those MSs where MSAs handle already both categories of products,
harmonised and non-harmonised. The practical implications are therefore rather better
synergies and use of existing structures and resources than new additional needs. The
extended coverage of risks from new technologies (e.g. cyber-security risks that have an
impact on safety) would be expected to increase the need for professional staff and
external expertise on the side of MSAs to check the safety of new technology products.
The practical implications will start operating as soon as the revised GPSD has entered
into force.
97
2. SUMMARY OF COSTS AND BENEFITS
I. Overview of Benefits (total for all provisions) – Preferred Option
Description Amount Comments
Direct benefits
Increased safety of non-
harmonised products and
reduced product safety risks
covered by GPSD (and
related reduction of number
on injuries caused by unsafe
products)
- Preventable detriment suffered by EU
consumers and society due to product-related
accidents estimated at EUR 11.5 billion per year.
- the current cost of health care utilisation for
product-related injuries in the EU is
approximately EUR 6.7 billion per year, with
hospitalisation accounting for the larger part of
the total health care costs at about EUR 6.1
billion.
These costs can be reduced under Option 3
Options 3 also expected to reduce consumer
detriment estimated on the basis of the value of
unsafe products by approximately EUR 1.04
billion in the first year of implementation,
increasing to approximately EUR 5.5 billion
over the next decade, This represents the
decrease of financial costs for consumers since
they would avoid buying unsafe products.
The GPSD Study also showed that stakeholder
consider that Option 3 provides ‘moderate’ to
‘significant’ benefits for consumers.
Main impact on EU consumers via broader
coverage and greater effectiveness of the
GPSD in protecting consumers from unsafe
products, in particular in online sales and for
risks of new technologies.
Impact also on MS (positive impact on
health care budget)
Higher return rates during
recalls of unsafe products
Reduced number of deaths and injuries caused by
products staying in hands of consumers due to
delayed and badly managed recalls. Reduced
amount of consumer detriment.
Reduced consumer detriment related to the value
of unsafe products which were not effectively
recalled by EUR 410 million per year.
Examples from ineffective recalls: faulty Takata
airbags (estimated to have cause 35 deaths and
300 injuries worldwide) and Fisher-Price rock ‘n
play baby sleepers (associated with 59 baby
deaths in the US).
Main impact on EU consumers via lower
exposure to unsafe products and on MS
(positive impact on health care budget).
Level playing field and a
better functioning EU
internal market
These potential benefits were assessed as being
‘moderate’ to ‘significant’ in the Study’s survey
Mainly via alignment of the market
surveillance rules for all products, a clearer
legal framework and deterrent effect on
rogue traders.
Main impact on EU businesses.
Reduced regulatory costs
and burdens for businesses
Cost reductions for all businesses and in
particular for the 42% of businesses who reported
additional costs related to the diverging
implementation of the GPSD.
Cost savings for businesses of around EUR 59
million annually (EUR 34 million saved by EU
SMEs and 26 million EUR saved by EU large
businesses respectively) through more
harmonised implementation.
Study showed that companies and business
associations estimate the benefits between
‘minor’ and ‘moderate’ and MSAs and other
stakeholders to be mostly considerably more than
Main impact on businesses via:
-legally binding clarifications and choice of
Regulation as instrument will reduce
regulatory uncertainty and even
implementation
-aligning the general market surveillance
and safety requirements for harmonise and
non-harmonised products will reduce
implementation differences and improve the
traceability of supply chain
98
‘moderate’ and close to ‘significant’.
Efficiency gains in market
surveillance and
enforcement
Cost reductions for all MSAs and in particular for
the16% of MSAs who reported related additional
costs to the diverging legal frameworks between
harmonised and non-harmonised products.
Cost savings for MSAs estimated at EUR 0.7
million per year across the EU.
Main impact on MSAs due to aligning
market surveillance provisions between
harmonised and non-harmonised products,
more aligned enforcement powers, increased
deterrent effect and arbitration mechanism.
Reduced administrative
burden of the standardisation
process
Not quantifiable Via the simplification of the standardisation
process will streamline the related EU
process. As it would accelerate
standardisation work, it would increase legal
certainty for companies on the standards to
comply with.
Main impact on MSs and EC
Indirect benefits
Positive spill-over effects on
consumer trust, demand,
production and employment
Not quantifiable Via increased safety of products and free
movement of goods in the Single Market.
Beneficial for all undertakings
Improved companies’
competitiveness
Additional competitiveness gains expected to be
very moderate as companies’ current compliance
costs with consumer product safety legislation
are already relatively low and additional
regulatory requirements would level potential
cost reductions.
Via a more harmonised regulatory level-
playing field within the EU
Main impact on EU businesses
Positive impacts on
competition-driven
innovation
Not quantifiable Via a greater degree of harmonisation and
greater legal certainty (e.g. development of
new innovative information and traceability
systems).
II. Overview of costs – Preferred option
Citizens/Consumers Businesses Administrations
One-off Recurrent One-off Recurrent One-off Recurrent
New
general
due
diligence
measures
of
economic
operators
for
product
safety
Direct costs
- - Familiarisation
costs,
adaptation costs
to regulatory
changes
Total costs of
businesses in
the EU27 in the
first year of
implementation
are estimated at
EUR 196.6
million (one-
off + recurrent
costs in the first
year),
equivalent to
0.02% of
turnover of EU
companies for
manufacturing,
wholesale and
Additional
regulatory
compliance
costs, related to
staff and
additional
resources (more
for
manufacturers
to adjust
different stages
of the value-
adding process
to new
regulatory
requirements)
Recurrent costs
amount to EUR
177.8 million
(0,02% of
companies’
turnover)
Only
relatively
moderate
one-off
adaptation
and
implementat
ion costs.
total
additional
recurrent
costs of
MSAs in
EU27 of
approx. EUR
6.7 million
annually
99
retail of non-
harmonised
consumer
products.
Indirect costs - Potential impact
on consumer
prices in the
EU, expected to
be negligible
(potentially for
low-income
consumers). No
significant or
negative impact
on consumer
choice in the
EU expected
- - - -
Duty of
care
obligation
s for
online
marketpl
aces
Direct costs
- - Costs
estimation
included in the
total above
Additional
regulatory
compliance
costs, for all
online
marketplaces
and in
particular for
non-signatory
of the Pledge,
but likely less
efforts than
those of brick
and mortar
distributors for
fulfilling their
obligations
today.
Costs
estimation
included in the
total above
- -
Indirect costs - - - - - -
All safety
informati
on is
provided
online in
the same
vein as it
is
required
“offline”
Direct costs
- - - Costs to be very
limited for both
online
platforms and
online sellers
(information
already
available and
does not go
beyond what is
indicated on the
packaging)
- -
Indirect costs - - - - - -
New
requirem
ents on
recalls
Direct costs Reduced
cost of
recall
(improved
- Higher
administrative
burden for
recalls and
-
100
remedy) registration
systems. Costs
mainly limited
to situations
when recall
occurs (unsafe
product placed
on the market)
and in any case
operators
should already
carry out
effective
recalls.
Indirect costs - - - - - -
Integratio
n of food-
imitating
products
into
GPSD
Direct costs - - - Minimal effect
on producers of
food-imitating
products, and in
any case not
exceeding costs
supported by
other producers
- Potentially
some costs for
MSAs which
were applying
a ban per se
of these
products and
will have to
do a risk
assessment.
Considered as
minor in view
of the limited
amount of
these products
Indirect costs - - - - -
(1) Estimates to be provided with respect to the baseline; (2) costs are provided for each identifiable action/obligation of the
preferred option otherwise for all retained options when no preferred option is specified; (3) If relevant and available, please
present information on costs according to the standard typology of costs (compliance costs, regulatory charges, hassle costs,
administrative costs, enforcement costs, indirect costs; see section 6 of the attached guidance).
101
Annex 4: Analytical methods
This Annex provides an overview of the following analytical methods and techniques as
well as the related data sources used for the impact assessment:
 Estimation of the detriment due to product-related injuries and fatalities in the
EU;
 Estimation of costs of compliance with the GPSD for EU businesses;
 Estimation of costs of compliance with the GPSD for Member states;
 Estimation of the costs of implementing specific policy options for the potential
revision of the GPSD;
 Estimation of benefits of measures concerning online sales channels;
 Estimation of benefits of measures in the field of recalls; and
 Methods for other supporting estimations.
They are elaborated in the following sub-section.
1. Estimation of the detriment due to product-related injuries and fatalities in
the EU
The cost of non-fatal product related injuries in the EU
For the calculation of the cost of non-fatal product related injuries in the EU139
, we use
the European Injury Database (IDB) as a source of data on product-related injuries. The
data are voluntarily contributed by the Member States participating in the IDB, which
were 15 out of 28 Member States in 2016140
. Two levels of datasets exist in the IDB: the
full dataset indicated as IDB-FDS and the minimum dataset referred to as IDB-MDS.
The IDB-FDS provides more detailed information with regards to the circumstances of
the injury and the products involved, in comparison to the IDB-MDS, which includes
limited information pertaining to the injury, but provides data that can be used to
extrapolate data to the EU level. For the analysis, both datasets have been used.
The analysis focused on accidental, non-intentional injuries and excluded transport injury
events and work-related injuries. As IDB data has also been used as an indicator for the
European Commission’s Consumer Market Scoreboard, we have selected the same
product groups used by the Consumer Market Scoreboard to define consumer products as
represented in the IDB141
.
139
The analysis refers to the European Union of 27 Member States. The monetary values in the
analysis are expressed in EUR 2017; in cases where 2017 values have not been available, monetary values
were inflated to 2017 values using Eurostat’s Labor Cost IndexEurostat, Labour cost index by NACE Rev.
2 activity - nominal value, annual data [lc_lci_r2_a]. NACE_R2: Industry, construction and services
(except activities of households as employers and extra-territorial organisations and bodies). Extracted
16/06/2020.
140
Ibid., p. 26.
141
See European Commission (2014), ‘Consumer Markets Scoreboard. Making markets work for
consumers’, 10th
edition, p. 60-61.
102
To estimate the number of injuries related to different product groups we have used the
number of injuries recorded in the IDB-FDS between 2013-2017. On basis of the data
provided in the IDB we estimated the total number of injuries in the EU27 on average per
year between 2013-2017, using Eurostat population data to extrapolate the FDS data. The
method for extrapolation is elaborated in detail in Annex IIc.
Health care utilization
Health care utilization costs include the costs of hospitalization/hospital admission, the
costs of treatment in a hospital emergency department, as well as the costs of being
treated in a non-hospital setting e.g. at a doctor’s office or as an outpatient. To calculate
the cost of health care corresponding to the product-related injuries, it is necessary to
retrieve data regarding the consequences of the injuries in terms of the required medical
attention as well as the unit costs for each type of health care. The data contained in the
IDB-FDS enabled us to identify between three different groups of product-related
injuries in terms of the type of treatment required: Patients with product-related injuries
that are sent home after treatment; Patients with product-related injuries that are either
treated and referred to a general practitioner for further treatment or treated and referred
for further treatment as an outpatient; Patients with product-related injuries that are
treated and admitted to hospital or transferred to another hospital.
To arrive at the costs of health care utilization we used the approach as described in the
following box:
Health care utilisation costs for a given injury type can be estimated by multiplying the average cost of
treatment by the number of cases, as indicated below:
𝑎 ℎ 𝑎 𝐸𝑈 = ∑[ 𝐸𝑈,𝐶𝑎 × 𝐴 𝑎 𝐸𝑈,𝐶𝑎 ]
Where:
HealthCareUtilEU is the total cost of health care utilisation at the EU level;
NrInjuriesEU,Cat is the number of product-related injuries by treatment category;
AvgTreatmentCostEU,Cat is the average cost of treatment for the given injury in a
given MS, by treatment category.
For assessing average treatment costs, we used unit cost values for health service
delivery from the WHO-CHOICE project, which are provided for different world regions
in 2010 international dollars142
. After converting the two types of costs into EUR 2010
using the OECD purchasing power parity (PPP) exchange rate143
, we inflated them to
EUR 2017 using Eurostat’s Labor Cost Index. Based on these conversions we calculated
the average cost per inpatient bed hospital day and the average cost per outpatient visit.
We used the calculated values to estimate respectively the cost of the three groups of
treatment (as indicated above).
142 WHO Economic Analysis and Evaluation Team (2010), ‘WHO-CHOICE estimates of cost for
inpatient and outpatient health service delivery’, pp. 1-60, available at: https://www.who.int/choice/cost-
effectiveness/inputs/country_inpatient_outpatient_2010.pdf.
143 OECD (2020), Purchasing power parities (PPP) (indicator), available at: doi: 10.1787/1290ee5a-
en (accessed on 06 July 2020).
103
Productivity losses
The cost of productivity losses is considered for this assessment to correspond to the
value of missed time from work. The cost of productivity losses was calculated first by
estimating the number of work days lost as a consequence of the injury related to a
product and then multiplying this number by the EU average gross daily earnings.
Product related injuries for which the type of treatment is not indicated or recorded are
not taken into account for the assessment of productivity losses. The detailed approach
for determining productivity losses is provided in the following box:
The cost of productivity losses for a given treatment category are calculated as the cost of missed work. In
order to account for the fact that a disproportionate number of injuries occur among children, we take into
consideration the proportion of victims that are of working age. The calculation can be expressed as:
𝐸𝑈 = ∑[ 𝐸𝑈,𝐶𝑎 × 𝐴 𝐸𝑈 × 𝐸𝑈 × 𝑎 𝐸𝑈 × 𝑎𝑦 𝐶𝑎 ]
Where:
ProdLossEU is the total cost of productivity losses in the EU;
NrInjuriesEU, Cat is the number of product-related injuries in a given
treatment category;
WAPopEU is the proportion of the injured persons that are of working age;
LMPEU is the labour market participation rate in the EU for working age population;
WageEU is the average daily wage in the EU; and
DaysLostCat is the average number of days of work lost for a given treatment
category.
Loss of quality of life
To estimate the impact of the injury in terms of reduced life quality we use the Quality
Adjusted Life Year (QALY), a measure that integrates evaluation of the quality and
quantity of life144
. For calculating the cost due to reduced quality of life, we have used
the following approach145
.
Loss of quality of life will be considered for serious injuries, which are considered to be those for which
hospitalisation was required, according to the following equation.
𝑎 𝑦 𝐸𝑈 = ∑[ 𝐸𝑈, , 𝑗 × 𝐴 𝑌 𝑗 × 𝑎 𝐴 𝑌𝐸𝑈 ]
Where:
LossQualityLifeEU is the monetised total loss of quality of life of patients
hospitalised due to product-related injuries in the EU;
NrInjuriesEU,Hosp, Inj is the number of hospitalised cases for each main type of injury
related to products in the EU;
LossQALYInj is the Quality Adjusted Life Year loss for each main type of injury;
ValueQALYEU is the monetary value assigned to a Quality Adjusted Life Year.
For each of the injuries we have identified on basis of IDB data, we used a corresponding
144
Adler, Matthew D. "QALY's and Policy Evaluation: A New Perspective." Yale Journal of Health
Policy, Law, and Ethics 6, (2006), Hammitt, James K. "QALYs Versus WTP." Risk Analysis 22, no. 5
(2002): 985-1001.
145
See Karapanou, Vaia. Towards a Better Assessment of Pain and Suffering Damages for Personal
Injuries. A proposal based on Quality Adjusted Life Years. Cambridge, Antwerp, Portland: Intersentia,
2014.
104
QALY-weight that expresses the impact of the injury in terms of the quality of life of
individuals, using relevant specific estimates. Another approach that has been used to
estimate the WTP for a QALY involves taking advantage of the existing literature on the
Value of Statistical Life (VSL). This approach, the validity of which was also confirmed
by an expert of the European Chemicals Agency (ECHA), is also consistent with the
VSL approach that is used below to calculate the cost of premature death. We followed
this approach to derive the monetary value for one QALY, using the VSL range of
estimates between €3.5 million (lower estimate) and €5 million (higher estimate)
included in the Commission’s Better Regulation Toolbox146
. After expressing them in
EUR 2017 using the labour cost index we converted them to VSLY estimates by
applying a discount factor of 4%147
and a remaining life expectancy of 35 years, which is
commonly considered as the remaining life expectancy of an adult at the time of
injury148
. Finally, considering that the resulting values based on the VSL are upper bound
estimates that tend to overestimate the value per QALY by a factor of two on average, we
divided the estimated amounts by two149
. The resulting range of willingness to pay
estimates per QALY used in this study are listed in the following table.
Source WTP for a QALY estimate
Civic Consulting based on VSL estimates
provided in EU Commission’s Better
Regulation Toolbox
€101 706 (low estimate)
€123 500 (medium estimate)
€145 294 (high estimate) (in EUR 2017)
The cost of product related premature death in the EU
In order to arrive at the number of fatal injuries in Europe, we have used the WHO
Mortality Database (WHO-MDB) which contains data for all countries participating in
WHO150
. To enable a selection of fatal injury incidents that are relevant for this analysis
we have filtered existing data by selecting injury incidents based on specific ICD-10
codes. Based on the incidence figures extracted from the WHO dataset we calculated the
cost of premature death related to the selected fatalities. Our approach is detailed in the
box:
Cost of premature death is estimated for all non-intentional fatalities caused by mechanisms relevant for
product safety (such as tools, strangulation, electric current, or fire) outside of work-related locations, on
basis of the following equation:
𝐹𝑎 𝑎 𝐸𝑈 = 𝐹𝑎 𝑎 𝐸𝑈 × 𝐸𝑈
Where:
146
Better Regulation Toolbox complementing the better regulation guideline presented in
SWD(2017) 350, p. 245.
147
Better Regulation Toolbox complementing the better regulation guideline presented in
SWD(2017) 350, p. 503.
148
To estimate VSLY we use the formula VSLY= r*VSL/(1-(1+r)^-L) where r is the discount rate and L
is remaining life expectancy, see also Joseph E. Aldy & W. Kip Viscusi, 2008. "Adjusting the Value of
a Statistical Life for Age and Cohort Effects," The Review of Economics and Statistics, MIT Press,
vol. 90(3), pages 573-581.
149
Daniel Herrera-Araujo, James K. Hammitt & Christoph M. Rheinberger (2020), “Theoretical
bounds on the value of improved health”, Journal of Health Economics 72, p. 1-15.
150
WHO Mortality Database, accessible at: https://www.who.int/healthinfo/mortality_data/en/.
105
LossFatalEU is the monetised total loss due to the relevant fatalities in the EU;
NrFatalEU is the number of relevant fatalities in the EU;
VSLEU is the monetary value of a statistical life in the EU.
The monetary value used to quantify the value of a statistical human life is derived from
individuals’ willingness to pay (WTP) to eliminate a small risk of dying151
. Numerous
studies exist in which the VSL has been empirically estimated using the hedonic wage
method, the stated preference method or other methods152
. We have used the estimates
provided by the European Chemicals Agency (ECHA) to calculate the cost of premature
death, which are also referred to as reference values in the Better Regulation Toolbox of
the European Commission153
. More specifically we use the average value of the higher
and lower estimate for the value of a statistical life provided by ECHA (EUR 4.25
million) as a standard assumption for the cost of a premature death, while retaining the
low and high estimates for later sensitivity analysis. Expressed in 2017 values (again
inflated by using the labour cost index), we arrived at a VSL estimate of EUR 4.6
million. We have used this estimate to arrive at the annual cost of premature death due to
fatalities caused by mechanisms relevant for product safety.
2. Estimation of costs of compliance with the GPSD for EU businesses
(baseline business costs)
We first focused on the estimation of the baseline market size, i.e. the total turnover of
EU businesses from manufacturing and/or selling non-harmonised consumer products in
the EU154
, before analysing company level compliance cost data, and extrapolating it to
EU level, based on the estimated baseline market size. The analysis is structured
according to six steps:
Step 1: Estimation of EU companies’ total annual turnover from the production and/or
sales of non-harmonised consumer products in the EU
Based on NACE industry codes and sector descriptions, we identified those
manufacturing sectors (NACE Rev. 2, B-E), wholesale services sectors and retail sectors
(NACE Rev. 2, G) in which consumer products are produced and/or sold, i.e. we
excluded sectors that clearly focus on the production and sales of industrial products.
Sectors related to motor vehicles have been excluded, in line with the focus on non-
harmonised consumer products. While retail sale can be assumed to be largely related to
consumer products (although retailers may also sell to professional users, and may sell
services), the wholesale and manufacturing in the listed areas clearly also contain
industrial/professional products, an issue considered in Step 3 below. To arrive at the
151
It can also be derived by the willingness to accept (WTA) a small probability of death.
152
The stated preference method tries to elicit the value of non-market goods by directly asking people
how much they value these goods while the hedonic wage method uses labor market data that reveal
the trade-offs workers make between job risks and additional pay. The hedonic wage method belongs
to the group of revealed preference methods which infer WTP / WTA values from observed behaviour.
See Alessandra Arcuri, 2012, "Risk Regulation” in: Roger J. Van den Bergh & Alessio M. Pacces
(ed.), Regulation and Economics, chapter 6, Edward Elgar Publishing.
153
Better Regulation Toolbox complementing the better regulation guideline presented in SWD
(2017) 350, p. 245.
154
All estimates in this section refer to the EU27 as of 2020.
106
share of non-harmonised products produced and/or sold in these sectors, we applied the
estimate provided in the 2017 EU impact assessment for the new Market Surveillance
Regulation, which estimated that about 54% of products circulating within the European
Single Market are harmonised products and 46% are non-harmonised products155
.
Step 2: Deduction of extra-EU export
To calculate the net turnover for non-harmonised consumer products that are only sold in
the EU, we deducted the share of extra-EU exports from the total turnover of EU
companies. The calculation is based on an approximation of sector-specific export shares.
The extra-EU trade by enterprise characteristics data provided by Eurostat do not exactly
match the sector classification of turnover data by enterprise size class156
. We therefore
approximated the extra-EU export shares of manufacturing, wholesale and retail sectors
on the basis of those sectors for which we found full concordance in the two datasets157
.
The estimated extra-EU export shares of manufacturing, wholesale and retail sectors
were subtracted from the annual turnover of EU companies with non-harmonised
products in the selected sectors.
Step 3: Deduction of industrial and professional products
We corrected the EU turnover derived in Step 2 by the percentage shares of turnover that
can be attributed to the production and/or sales of consumer products in manufacturing,
wholesale and retail sectors. For this purpose, we drew on a different dataset, namely the
final consumption expenditure of households by consumption purpose158
. We again
correct for the share of harmonised products, and arrived at an estimate for total
household consumption of non-harmonised products. For the following analysis we
assumed that this consumption of non-harmonised consumer products is equivalent to the
total turnover from non-harmonised consumer products sold by EU retailers. The
estimated retail turnover from non-harmonised products indicated before was adjusted
accordingly, and the resulting amount was allocated between the three enterprise size
classes. Due to data limitations, the same methodology could not be applied for
manufacturing and wholesale sectors159
. For manufacturing and wholesale sectors, we
estimated the share of turnover that can be attributed to consumer products on the basis
of the share of “consumer-oriented” wholesale services in total wholesale services. It is
assumed that the same share reflects the portion of consumer products produced and/or
sold by manufacturers. Based on this approach, we could calculate the total annual EU
turnover of EU companies from non-harmonised consumer products.
155
SWD(2017) 466 final PART 2/4 Commission Staff Working Document Impact Assessment
Accompanying the document COM(2017) 795.
156
In the Annex, we provided detailed trade volumes of extra-EU exports by NACE Rev. 2 activity
and enterprise size class.
157
These sectors are: “Manufacture of textiles, Manufacture of wood and of products of wood and
cork, except furniture; manufacture of articles of straw and plaiting materials”, “Manufacture of paper and
paper products”, “Manufacture of computer, electronic and optical products”, “Manufacture of electrical
equipment”, “Manufacture of furniture”, “Wholesale trade, except of motor vehicles and motorcycles”, and
“Retail trade, except of motor vehicles and motorcycles”. In the Annex, we provide shares of extra-EU
exports in key consumer products sectors broken-down by enterprise size class.
158
Eurostat, Final consumption expenditure of households by consumption purpose (COICOP 3
digit) [nama_10_co3_p3].
159
Eurostat data do not allow to extract “pure” consumer products for manufacturing and wholesale
sectors, i.e. final products that are consumed by households.
107
Step 4: Derivation of empirical estimates for companies’ product safety-related costs
on the basis of survey responses
In our company cost survey and the complementary interviews conducted with selected
companies, businesses were asked to indicate staff time used for managing product
safety, testing for product safety, recalls and other consumer product safety related
activities. We asked respondents to consider all costs for ensuring product safety of both
harmonised and non-harmonised consumer products (excluding pharmaceuticals, medical
devices or food), as the identification of costs for non-harmonised products only was not
considered to be feasible. In addition to staff requirements, companies were asked to
provide estimates for other costs to comply with safety requirements for consumer
products (e.g. costs for external legal advice, costs for external safety testing, costs for
certification of safety of products etc.)160
. The cost estimates provided by the respondents
also include business-as-usual costs, which would incur even in absence of product safety
regulation (see Step 6). These estimates were used to estimate companies’ annual
regulatory compliance costs in Euro terms. The calculation of Euro-denominated costs
for staff was based on the EU’s (weighted) average wage for the business economy,
which in 2019 was 27.50 Euro per hour161
. To account for overhead costs, a 25% mark-
up was added to staff-related costs. Subsequently, the costs for each company were
related to the EU turnover for consumer products, i.e. we expressed companies’ annual
cost resulting from activities to comply with safety requirements for (harmonised and
non-harmonised) consumer products as a share of the related turnover.
Step 5: Extrapolation of EU companies’ annual costs related to the GPSD incl.
business-as-usual costs that occur also in absence of regulation
For each enterprise size class, we multiplied the empirical median values for companies’
relative product safety-related costs, which were derived in Step 4, with the annual
turnover of EU companies that can be attributed to the production and/or sales of non-
harmonised consumer products in the EU (Step 3). The results of this calculation still
include business-as-usual costs.
Step 6: Deduction of business-as-usual costs and extrapolation of EU companies’
annual compliance cost related to the GPSD
In our company survey and interviews, we asked businesses to indicate the share of the
total product safety-related costs that they would incur anyway (i.e. even in absence of
product safety legislation, e.g. because these costs relate to due diligence), hereafter
referred to as business-as-usual costs, BAU. These estimates reflected the self-
assessment of the companies that are part of the sample, and are therefore subjective in
nature. However, as concerns differences between manufacturers, on the one hand, and
wholesalers and retailers, on the other, we considered the estimates to be in line with
expectations and a credible basis for the final step of the assessment. We applied the
empirical median values of these shares to the product safety-related cost estimates
derived in Step 5. Excluding business-as-usual costs, we obtained compliance costs of
EU companies that can be attributed to non-harmonised consumer products, i.e. the costs
for businesses to comply with the GPSD.
160
Business stakeholders were asked to estimates average costs per month in EUR.
161
Labour cost for LCI (compensation of employees plus taxes minus subsidies), provided by
Eurostat.
108
3. Estimation of costs of compliance with the GPSD for Member States
(baseline costs for Member States)
The estimation of MSAs’ staff-related costs related to market surveillance activities for
non-harmonised consumer products in the EU was based on the following three steps:
Step 1: Identification of MSAs annual FTEs for market surveillance activities related
to non-harmonised consumer products
For our estimate we used the number of full time equivalent (FTE) staff for market
surveillance of consumer products as provided in the country research. Where the
available country estimates related to the market surveillance of non-harmonised
consumer products, this figure was directly used in the calculation. Where estimates
related to the total staff for market surveillance of both harmonised and non-harmonised
consumer products, we allocated staff according to the 54%/46% ratio for
harmonised/non-harmonised products circulating within the European Single Market to
derive an estimate for related market surveillance activities162
. It should be noted that a
share of 46% in staff time for market surveillance of non-harmonised consumer products
is 12 percentage points higher than the empirical median share indicated by MSAs for
activities devoted to non-harmonised products in the stakeholder survey (34%),
potentially causing an estimate at the higher end of MSAs’ actual costs that can be
attributed to market surveillance activities for non-harmonised consumer products. For
seven countries, no information on staff numbers was available at all.
Step 2: Approximation of annual FTEs for market surveillance activities related to
non-harmonised consumer products for countries for which data was not available
For the seven countries, for which no staff data was available (Croatia, Germany,
Hungary, Italy Slovenia, Slovakia, and Spain) we estimated the number of FTEs on the
basis of the data for the remaining 20 Member States. To account for institutional
differences with regard to the level of centralisation, we considered two clusters of
countries, in line with the characteristics of the respective market surveillance systems as
described above: Cluster 1: responsibility for market surveillance is centralised (no sub-
national administrations involved); Cluster 2: responsibility for market surveillance is
(partly) delegated to or competence of sub-national administrations, in line with the
administrative structure of the country.
162
As mentioned before, the 2017 EU impact assessment for the new Market Surveillance Regulation
estimated that about 54% of products circulating within the European Single Market are harmonised
products and 46% are non-harmonised products. See SWD(2017) 466 final PART 2/4 Commission Staff
Working Document Impact Assessment Accompanying the document COM(2017) 795.
109
To derive estimates for the number of FTEs per million population for Slovenia and
Slovakia (more centralised market surveillance), we applied the sample median of 3.5
FTEs per million population. To derive FTE estimates for the number of FTEs per
million population for Croatia, Germany, Hungary, Italy and Spain (more decentralised
market surveillance), we applied the sample median of 4.6 FTEs per million population.
Step 3: Calculation of annual staff costs for market surveillance activities related to
non-harmonised consumer products
In the final step, we calculated the EUR equivalent of the estimated number of staff
required for market surveillance of non-harmonised consumer products by multiplying
the number of FTEs per million population by:
 The size of population for each country (in million);
 The number of person-hours per year (1 720) 163
; and
 The average wage of 28.00 EUR, which corresponds to the EU27 average wage
of “administrative and support service activities” (18.70 EUR) and “professional,
scientific and technical activities” (37.30 EUR) for 2017 (latest figure available in
Eurostat database).
4. Estimation of the costs of implementing specific policy options for the
potential revision of the GPSD
Companies assessed in their responses to our cost survey the change that the
implementation of each option would cause in their recurrent costs, e.g. costs related to
additional staff and additional resources for due diligence measures such as IT systems
and external services, in addition to one-off costs, such as familiarisation costs and costs
from adapting to regulatory changes (e.g. for external advice). Both types of costs were
analysed.
To estimate the impact of the implementation of each option on EU businesses’ recurrent
costs, we applied the percentage change in recurrent (annual) costs as assessed by
respondents to the estimated annual product safety-related costs of companies producing
and/or selling consumer products in the EU (baseline estimates). Applying the sample
median as best estimate for the extent to which recurrent costs would increase under each
option, we calculated the change in the estimated annual consumer product safety-related
costs of EU businesses in Euro terms for manufacturers, wholesalers and retailers.
Our estimation of EU businesses’ total one-off costs was based on individual
respondents’ estimates for the total additional staff needed and the total additional non-
staff costs that arise from familiarisation and implementation efforts under each option.
Based on the respondents estimates, we calculated staff costs in Euro terms and added
other (non-staff) one-off costs. The calculation of Euro-denominated costs for staff was
based on the EU’s (weighted) average wage for the business economy, which in 2019
163
Following EU Horizon 2020 guidelines, one person year corresponds to 1 720 person-hours per
year. See, e.g. the H2020 Programme: User's Guide for the Personnel Costs Wizard.
110
was 27.50 Euro per hour164
. To account for overhead costs, a 25% mark-up was added to
staff-related costs.
The total one-off costs for each company were divided by the EU turnover for consumer
products, i.e. we expressed companies’ total additional one-off costs resulting from
activities to comply with safety requirements for consumer products under Option3 as a
share of the related turnover. Applying the sample median to the estimated annual
turnover for manufacture, wholesale and retail of consumer products in the EU resulted
in estimates for additional one-off cost for manufacturers, wholesalers and retailers.
The estimate of recurrent and one-off costs of MSAs was conducted using a similar
approach, with estimates on how the implementation of each option would change their
recurrent costs derived from the answers to our survey of authorities. Again, we
multiplied the empirical median with baseline costs, to estimate recurrent costs, and
separately assessed one-off costs.
5. Estimation of benefits of measures concerning online sales channels
No consistent data is available on the incidence of unsafe products on the EU market. In
the analysis, we used stakeholder assessments as best available estimate to first analyse
the potential detriment accruing currently to consumers due to unsafe products on the EU
market, and then consider the impact that increasing e-commerce and the implementation
of different policy options could be expected to have on this baseline situation. A key
challenge in this respect is the size of the detriment to consumers posed by unsafe
products. An unsafe product could lead to injuries and fatalities, which cause substantial
detriment in the EU every year. Due to data limitations, it is not possible to quantify the
occurrence of product-related injuries and fatalities, or damage to other goods caused by
unsafe products according to sales channel. We therefore in this analysis use as proxy for
the detriment caused by an unsafe product its value (as expressed by its purchase price).
This approach seems to rather underestimate than overestimate detriment, in light of the
different situations analysed. In our baseline analysis, we have estimated the total EU27
household consumption of non-harmonised consumer products (excluding food and
medical products) at EUR 428 664 million per year. Combining this data with the
estimate of the incidence of unsafe consumer products, we derive the value of unsafe
products per year (which is in our approach equivalent to the related consumer detriment)
at EUR 3.9 billion for the online sales channels, and EUR 15.4 billion for brick-and-
mortar shops and other offline sales channels, for a total of EUR 19.3 billion. This figure
is by its nature an approximate estimate, as the data on which it is based has considerable
limitations, and the result is affected by the underlying assumptions.
6. Estimation of benefits of measures in the field of recalls
A fundamental obligation that derives from the GPSD is the obligation of producers and
distributors to notify the authorities and take the necessary actions for consumer
protection, once one of the products that they have placed on the market is identified as
dangerous165
. The limited effectiveness of recalls also leads to consumer detriment, the
size of which is estimated in this annex.
164
Labour cost for LCI (compensation of employees plus taxes minus subsidies), provided by
Eurostat.
165
GPSD Art 5 (3).
111
For estimating consumer detriment due to ineffective recalls, we follow the approach
explained above, namely to use the value of an unsafe product as a proxy for the
detriment it causes to consumers that have bought it (a detailed justification of this
approach is provided in the same Annex). When using the value of a recalled product to
analyse consumer detriment, two situations can be differentiated:
1. An unsafe product is recalled and returned to a producer. The resulting consumer
detriment can be approximated as being zero166
;
2. An unsafe product is recalled and not returned to a producer. In this case the
consumer detriment is the value of the product, as discussed.
Under a scenario of improved recall effectiveness, consumer detriment in the EU can be
expected to be reduced by more than EUR 400 million per year. As mentioned above,
this estimate is based on a number of scenario assumptions, which have been chosen with
the aim to provide a conservative estimate of consumer benefits due to improved recall
effectiveness. A key assumption is that the detriment incurred by consumers in case of a
recall of an unsafe product is equivalent to its purchase price. This is a very restrictive
assumption, as it does not consider situations in which a recalled, unsafe product caused
damage to persons, other goods or the environment. Also, the return rates underlying the
improved effectiveness scenario are still relatively low and might be further increased
through appropriate measures by producers and authorities, considering e.g., the
increased availability of customer data in online transactions. If return rates were to be
improved beyond our assumptions, consumer detriment would accordingly be further
reduced, compared to the estimate provided.
7. Methods for other supporting estimations
Other supporting estimations include the analysis of costs of mandatory accident
reporting and the extrapolation of the number of parcels imported to the EU. In both
cases, baseline data was extrapolated using relevant data sources from international
organisations or data from non-EU countries in which comparable measures were taken.
For more details on the methodological approach taken in each case, see the relevant
section of the report.
8. Validation and quality assurance of results of analyses conducted
Great care was taken to explore all possible data sources at EU level and from
international databases to use the best available data, which is a key element of quality
assurance. All analyses were validated internally by different members of the team, to
safeguard internal consistency and accuracy. Finally, in major analyses external expertise
was involved, either through advisory roles (e.g. an expert of EuroSafe supported the
data extraction process related to the IDB), or through providing advice on specific
methodological issues. These included the WHO, which was consulted on possible
approaches to group ICD-10 codes, and ECHA, which provided advice on the most
appropriate method to determine VSLY values.
Sensitivity analysis was used to assess robustness of estimates against different
assumptions, where relevant. With respect to the estimation of detriment, we elaborated
166
In reality, even in this situation consumers incur a detriment due to the time spent for the
transaction, e.g., for returning the product by mail or in person to a shop. However, this additional
detriment is not considered here, provide a conservative, simplified estimate.
112
sensitivity scenarios concerning the cost of premature death and the loss of quality of life.
The first scenario to be tested against the main scenario involved using the lower
estimate of the VSL to recalculate the costs incurred as a result of premature death. The
second scenario involves the opposite recalculation, namely using the high estimate of
the VSL and the corresponding QALY value to recalculate the costs incurred as a result
of premature death and of lost quality of life. The third and fourth scenarios take into
account the fact that the type of the injury as such e.g. injury to muscle, burn etc. does
not convey the severity of the injury which may significantly influence the magnitude of
the loss. Therefore, to account for the possibility of a mild and severe occurrence of the
same type of injury we estimated the loss of quality of life using both low and high
QALY losses per each type of injury. The rest of the assumptions (monetary value of a
VSL, a QALY) remained the same as in the main scenario. The fifth and final sensitivity
scenario involved taking into account for the calculation of the cost of premature death
only the fatalities caused by mechanisms relevant for product safety that occur at home
keeping everything else constant.
30
Annex 5: Evaluation report
Table of contents
1. INTRODUCTION .............................................................................................................................. 33
Purpose and scope .................................................................................................... 33
2. BACKGROUND TO THE INTERVENTION ................................................................................... 34
Description of the intervention and its objectives .................................................... 34
Baseline and points of comparison........................................................................... 37
3. IMPLEMENTATION / STATE OF PLAY ........................................................................................ 38
4. METHOD............................................................................................................................................ 41
Short description of methodology ............................................................................ 41
External studies ........................................................................................................ 42
Open public consultation (OPC) and stakeholders’ workshops............................... 42
Limitations and robustness of findings .................................................................... 43
5. ANALYSIS AND ANSWERS TO THE EVALUATION QUESTIONS .......................................... 44
5.1 Relevance ........................................................................................................... 44
5.1.1. General relevance of the Directive........................................................................ 44
5.1.2. Relevance of of the Directive vis-à-vis E-commerce and direct imports.............. 46
5.1.3. Relevance of the Directive vis-à-vis new technologies......................................... 50
5.1.4. Relevance of definitions of the Directive.............................................................. 52
5.1.5. Relevance of environmental issues, including chemical risks .............................. 53
5.1.6. Conclusion on relevance ....................................................................................... 55
5.2 Effectiveness ...................................................................................................... 55
5.2.1. Extent to which the Directive has been effective in contributing to consumer safety. 55
5.2.2. Extent to which the Directive has been effective in contributing to the functioning of
the Single Market .................................................................................................................. 60
5.2.3 Effectiveness of the system of market surveillance under the Directive...................... 61
5.2.4 Effectiveness of Safety Gate/RAPEX .......................................................................... 65
5.2.5 Effectiveness of the standardisation procedure of the Directive .................................. 66
5.2.6 Effectiveness of provisions on product recalls............................................................. 67
5.2.7 Impact of ecommerce in the effectiveness of the Directive ......................................... 70
5.2.8 Impact of new technologies in the effectiveness of the Directive................................ 73
5.2.9 Effectiveness of Food Imitating Products directive ..................................................... 74
5.2.10 Conclusion of effectiveness........................................................................................ 75
5.3 Efficiency ........................................................................................................... 75
5.3.1. Analysis of costs.......................................................................................................... 75
5.3.1.1 Costs for businesses................................................................................................... 75
31
5.3.1.2 Costs for authorities................................................................................................... 76
5.3.2. Analysis of benefits..................................................................................................... 77
5.3.3. Balance and distribution between costs and benefits .................................................. 78
5.3.4. Conclusion of efficiency ............................................................................................. 79
5.4 Coherence........................................................................................................... 80
5.4.1 Internal coherence of the Directive .............................................................................. 80
5.4.2 Coherence with product harmonisation legislation and market surveillance ............... 80
5.4.3 Coherence with standardisation policies ...................................................................... 83
5.4.4 Coherence with consumer protection and product liability legislation ........................ 83
5.4.5 Coherence with E-commerce rules............................................................................... 85
5.4.6 Coherence with wider EU policy ................................................................................. 86
5.4.7 Conclusion on coherence.............................................................................................. 86
5.5. EU added value ................................................................................................. 87
6. CONCLUSIONS AND LESSONS LEARNT .................................................................................... 88
APPENDIX 1: PROCEDURAL INFORMATION ...................................................................................... 90
APPENDIX 2: METHODS AND ANALYTICAL MODELS..................................................................... 92
APPENDIX 3: EVALUATION QUESTIONS............................................................................................. 97
APPENDIX 4: IMPLEMENTATION OF THE FOOD IMITATING PRODUCTS DIRECTIVE.............. 99
32
Glossary
Term or acronym Meaning or definition
AI Artificial Intelligence
BAU Business as Usual
CETA EU-Canada Comprehensive and Economic Trade Agreement
CSN Consumer Safety Network
DGCCRF Direction générale de la Concurrence, de la Consommation et de la
Répression des fraudes, France
DSA Digital Services Act
EEA European Economic Area
EFTA European Free Trade Association
EU European Union
EQ Evaluation Question
FIPD Council Directive 87/357/EEC concerning the safety of food-
imitating products
GPSD Directive 2001/95/EC on the general safety of products
IA Impact Assessment
IoT Internet of Things
NLF New Legislative Framework
OECD Organisation for Economic Cooperation and Development
OPC Open Public Consultation
Safety Gate/RAPEX The rapid alert system for dangerous non-food products
Subgroup The Subgroup on AI, connected devices and other new challenges
in product safety to the Consumer Safety Network
SME Small and Medium Enterprise
UCPD Unfair Commercial Practices Directive
VAT Value Added Tax
33
1. INTRODUCTION
1.1 PURPOSE AND SCOPE
Over two decades, the Directive 2001/95/EC on the general safety of products (‘the
Directive’) has established a product safety framework to ensure the safety of consumer
products.
In February 2013 the Commission adopted the Product Safety and Market Surveillance
Package167
, whose aim was, among other things, to revise the Directive. The proposed
rules however were not adopted by the Council and the Parliament, due to the lack of
political consensus on the so called "made-in" clause. Consequently, the proposals were
withdrawn in September 2020.
Since the adoption of the Directive in 2001, new developments in products and markets
have occurred, in particular as concerns products incorporating new technologies and e-
commerce. In addition, the adoption of new legal instruments, such as the recently
updated Regulation on market surveillance and compliance of products168
, has had as a
consequence that the Directive’s provisions on market surveillance are not fully in line
with market surveillance rules for harmonised products.
The Directive has never been evaluated since its entry into force169
. In light of the
above-mentioned developments, the Commission has carried out an Evaluation of the
Directive to assess its performance. This evaluation has been prepared back-to-back
with the Impact Assessment for a proposal to revise the Directive. This was justified as
at the time of the launch of this initiative the Commission had already quite some
evidence to support the evaluation and the impact assessment. An impact assessment was
already carried out by the Commission in 2013, as well as for the recently adopted
Regulation (EU) 2019/2010 on market surveillance and compliance of products.
Moreover, both the Evaluation and the Impact assessment were supported by a number of
studies and other data gathered through consultation. A report on the implementation
of the Directive also accompanies the Impact Assessment and the Evaluation, as
established in Article 19(2) of the Directive.
The geographical scope of the evaluation covers 31 countries (EU28, Iceland,
Liechtenstein and Norway). It focuses on the period from 2004 (i.e. subsequent to the
deadline for its transposition and application according to Article 22 of the Directive) to
2020, seeking to understand trends over this period wherever possible. This evaluation
167
COM(2013) 78 final - The Product Safety and Market Surveillance Package
168
Regulation (EU) 2019/1020 Regulation on market surveillance and compliance of products
169
The Product Safety and Market Surveillance Package did not include a proper evaluation of the
Directive 2001/95/EC according to Better Regulation rules.
34
also covers Council Directive 87/357/EEC concerning the safety of food-imitating
products (FIPD)170
.
This evaluation assesses the following criteria: relevance (whether the tools of the
Directive correspond to current needs), effectiveness (whether the original objectives
have been achieved), efficiency (the functioning of the Directive from a simplification
and burden reduction perspective), coherence (how the Directive works together with
other legislation in the field of safety of consumer products), and the EU added value of
the Directive. The evaluation's findings have fed into an Impact Assessment of the
policy options, which addresses the problems identified, including those related to food-
imitating products.
2. BACKGROUND TO THE INTERVENTION
2.1 DESCRIPTION OF THE INTERVENTION AND ITS OBJECTIVES
The Directive has a twofold objective. On the one hand, according to its recital (2), the
Directive pursues the aim of improving the functioning of the internal market. As recital
(3) confirms, it has introduced a common legislative framework in order to avoid
disparities between Member States that could have emerged in the absence of Union law.
At the same time, the Directive intends to achieve a high level of consumer protection by
introducing a general product safety requirement and other measures (recital (4)). Both
aims are interrelated, it is the safety requirement for consumer products envisaged by the
Directive, which prevents disparities that would lead to creating barriers to trade and
distortion of competition within the internal market.
The Directive establishes a general safety requirement for all consumer products, as it
obliges the producers to only place safe products on the market.
The Directive applies to all sales channels, offline and online.
Safety of services falls outside the scope of the Directive, but in order to secure a high
level of consumer protection, its provisions also apply to products that are supplied or
made available to consumers in the context of a service used by them. The safety of
equipment used by the service provider, in particular that on which the consumers ride or
travel, is nevertheless excluded. However, products which are actively operated by the
consumer at the premises of a service provider, such as hairdryers available to guests in
hotels rooms, are subject to the provisions of the Directive.
One of the key characteristics of the Directive is its role as a “safety net”, as it applies
insofar as there are no more specific provisions with the same objective in EU product
harmonisation legislation. Therefore, it complements sectorial legislation, as it covers all
aspects and risks not specifically addressed, thus ensuring a high level of protection of
consumers.
170
Council Directive 87/357/EEC of 25 June 1987 on the approximation of the laws of the Member States
concerning products which appearing, to be other than they are, endanger the health or safety of consumers
35
Standards are very important for the implementation of the Directive. The general safety
requirement can be difficult to apply for businesses and national authorities because of
the lack of a common benchmark on what constitutes a “safe” product. Therefore, the
Commission can make use of European standards to make this general safety
requirement more operational. European standards are voluntary and market-driven171
and their advantage is not only that they replace the corresponding national standards in
all Member States, making the life of businesses, including SMEs, easier, but in
particular that products are presumed safe if they conform to voluntary national standards
transposing European standards, which are referenced in the EU Official Journal.
Standards therefore serve a double purpose: facilitating market access and ensuring
product safety. The first step in the standardisation process under the Directive is a
Commission Decision to set the so-called "safety requirements" to be met by the
standards172
. The second step is the adoption of a Commission Decision173
issuing the
formal standardisation request ("mandate") to the European Standardisation
Organisations to develop standards compliant with the EC-adopted safety requirements.
The last step takes place after the European standardisation organisation has developed
the standard in conformity with the safety requirements. At this stage, in case the
Member States in the Committee established in Article 15 of the Directive vote
favourably and the Commission consider that the standard complies with the
requirements of the GPSD, the Commission adopts a Commission Decision to publish
the reference to this standard in the EU's Official Journal. A more detailed explanation of
the standardisation procedure is detailed in Annex 10.
The Directive provides for additional duties for economic operators, which can be
differentiated between measures to be taken before or after the product is placed on the
market:
 Pre-market obligations: Besides the general safety requirement, the Directive
requires producers to inform consumers of any risks associated with the products
they supply. The aim is to enable consumers to assess the risks inherent in a
product throughout the normal or reasonably foreseeable period of its use, where
such risks are not immediately obvious without adequate warnings, and to take
precautions against those risks. This obligation must be fulfilled when the product
is made available on the market. It not only relates to information on the proper
use of the product (as described in user manuals), but also to risks that come, for
example, with the age or the long-term use of the product. Producers should also
171
They are developed by the European Standardisation Organisations (CEN, Cenelec, ETSI), recognised
by the Regulation on European Standardisation 1025/2012. ANEC, the European association representing
consumers in standardisation, is funded by the EU budget and participates as an observer in the Consumer
Safety Network.
172
It is based on preliminary work undertaken with Member States, industry and consumer associations
(i.e. involving the Consumer Safety Network) to check the feasibility, the relevance and to reach consensus
about the contents of the safety requirements.
173
This is done in compliance with the Standardisation Regulation 1025/2012.
36
make sure that any product present on the market can be traced to swiftly enable
removal if necessary to avoid putting consumers at risk.
 Post-market obligations: Producers and distributors shall cooperate with the
competent authorities on actions taken to avoid the risks posed by products which
they supply or have supplied. Producers and distributors are also required to
immediately notify the respective authorities in EU Member States in case they
know or ought to know, on the basis of the information in their possession and as
professionals, that a product that they have placed on the market poses risks to the
consumer that are incompatible with the general safety requirement. Accordingly,
producers shall withdraw unsafe products from the market, publish warnings of
unsafe products or recall products from consumers on a voluntary basis or at the
request of the competent authorities. Distributors are required to act with due care
and within the limits of their respective activities, they need to participate in
monitoring the safety of products placed on the market.
To complement this and in order to ensure appropriate enforcement of the EU product
safety requirements, the Directive also sets out responsibilities for Member States to
establish systematic approaches to perform effective market surveillance. Member States
establish or nominate national authorities competent to monitor the compliance with the
product safety requirements and give the necessary powers to these authorities to take
appropriate measures under the Directive. National market surveillance authorities have a
responsibility to:
 check whether products available on the market are safe;
 ensure product safety legislation and rules are applied by manufacturers and other
actors in the supply chain;
 take appropriate action in case a unsafe product is detected on the market.
The Directive sets up the Rapid Alert System for non-food Consumer Products
(hereinafter "Safety Gate/RAPEX"). This system establishes the circulation of
information among the Commission and Member States' authorities on measures taken
by Member States' authorities and economic operators in relation to products posing a
serious risk to the health and safety of consumers. Information on non-serious risks can
also be circulated under Safety Gate/RAPEX. The Commission publishes relevant
information concerning all notifications on the EU Safety Gate website174
. Member States
are required to follow up the notifications of products posing a serious risk and inform
the Commission of any measures adopted. The system has been expanded by Regulation
765/2008175
to apply also to measures adopted in relation to professional products and to
other public interests beyond safety, such as the protection of the environment and
security.
174
https://ec.europa.eu/consumers/consumers_safety/safety_products/rapex/alerts/repository/content/pages/r
ape x/index_en.htm
175
Regulation (EC) No 765/2008 of the European Parliament and of the Council of 9 July 2008 setting out
the requirements for accreditation and market surveillance relating to the marketing of products and
repealing Regulation (EEC) No 339/93
37
In some exceptional circumstances, Article 13 of the Directive allows the Commission to
adopt temporary measures (valid for 1 year), via a decision, to ensure the safety of certain
products. It can be used in situations where at the same time Member States significantly
differ on the approach how to deal with the safety risk; where the risk needs to be dealt
with a high degree of urgency or where the risk can be eliminated only by adopting
appropriate measures at EU level.
This evaluation also covers Council Directive 87/357/EEC (FIPD), that applies to
products “which, although not foodstuffs, possess a form, odour, colour, appearance,
packaging, volume or size, so that is likely that consumers, especially children, will
confuse them with foodstuffs and in consequence place them in their mouths, or suck or
ingest them, which might be unsafe and cause, for example, suffocation, poisoning, or
the perforation or obstruction of the digestive tract.”
2.2 BASELINE AND POINTS OF COMPARISON
An intervention logic was developed for the purposes of this evaluation (see Figure 1). It
shows the logical sequence and causal relationships between the Directive’s rationale,
based on identified needs, its objectives, the activities undertaken, the intended results
(outputs), outcomes and impacts. The figure also shows other external factors (beyond
the Directive’s control) that may influence the impacts and outcomes.
Figure 1: Intervention logic
Source: GPSD Study
This evaluation takes also into account the accompanying Implementation Report 2019
as established in Article 19 of the Directive.
38
3. IMPLEMENTATION / STATE OF PLAY
The first general product safety directive was adopted in 1992176
. That directive was
amended (resulting in Directive 2001/95/EC on the general safety of products, subject of
this evaluation) in order to complete, reinforce or clarify some of its provisions in light of
experience as well as relevant developments on consumer product safety, together with
the changes made to the Treaty, especially in articles concerning public health and
consumer protection, and in the light of the precautionary principle.
All EU Member States notified transposition measures of the Directive, and this
evaluation showed that there have not been any problems in the transposition of the
Directive into national legislation. There are no open infringements regarding the
implementation of the Directive.
Member States are responsible for appointing competent authorities responsible for the
implementation of the Directive at national level and for ensuring that the Directive is
effectively enforced within their territories. In addition, Article 10 of the Directive sets
up an informal network of the Member States' authorities aimed at further enhancing
administrative cooperation (the "Consumer Safety Network"). Given that the Directive
forms part of the EEA Agreement, the same rules and mechanisms are also in place in the
EFTA countries applying the EEA Agreement, i.e. Norway, Iceland and Liechtenstein.
Most market surveillance authorities in the Member States work on the basis of annual
inspection programmes which take into account, among others, previous experiences and
findings, products that are frequently found unsafe and consumer complaints. If
necessary, all Member States carry out controls and tests which are not necessarily
foreseen in their programming, for example in emergency situations. To provide
assistance to the European network of Member States' product safety authorities, the
Commission has co-funded more than 50 joint actions on market surveillance among
these authorities since 2007.
Regarding the implementation of the Safety Gate/RAPEX, the number of measures
reported in the system has increased progressively over the years and since 2012 has
stayed just above 2,000 alerts a year. In 2019, a total of 2,243 measures were circulated
in the system177
.
176
Council Directive 92/59/EEC of 29 June 1992
177
https://ec.europa.eu/consumers/consumers_safety/safety_products/rapex/alerts/repository/content/pages/r
apex/reports/docs/RAPEX.2019.report.EN.pdf
39
Figure 2: Number of alerts in Safety Gate/RAPEX for the period 2003-2019
Source: Safety Gate/RAPEX
According to Article 12.4 of the Directive, the Rapid Alert System remains open to
applicant countries, third countries or international organisations, within the
framework of agreements between the EU and those countries or international
organisations, according to arrangements defined in these agreements. Based on this
article, a specific module of the system has been created to allow for swift flagging of
notifications to the Chinese authorities concerning unsafe products from China. The
Chinese authorities investigate these cases in order to trace back the manufacturers,
exporters and businesses concerned and take measures, which most often consist of
making companies aware of product safety rules in Europe. In addition, an exchange of
information on unsafe consumer products started with the Canadian authorities in 2019,
enabled by the EU-Canada Comprehensive and Economic Trade Agreement (CETA) and
an Administrative Arrangement between the European Commission’s Directorate-
General Justice and Consumers and the Department of Health Canada. The exchanges
aim to help EU Member States better target their enforcement efforts and identify
emerging product safety risks.
Furthermore, according to Annex II, point 8 of the Directive, the Commission shall
regularly update guidelines concerning the joint management of the Rapid Alert System
by the Commission and the Member States. A first version of the guidelines was adopted
in 2004178
, followed by another version in 2010179
. The most recent version dates from
2018180
.
178
Commission Decision 2004/418/EC (OJ L 151, 30.4.2004, p. 84)
179
Commission Decision 2010/15 (OJ L 22, 26.1.2010, p1-64)
180
Commission Implementing Decision (EU) 2019/417 of 8 November 2018 laying down guidelines for
the management of the European Union Rapid Information System ‘RAPEX’ established under Article 12
of Directive 2001/95/EC on general product safety and its notification system (notified under document
C(2018) 7334). OJ L 73, 15.3.2019, p.121–187
0
500
1000
1500
2000
2500
3000
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
Number of alerts per year
40
When it comes to standardisation work under the General Product Safety Directive, there
are currently a total of 67 standards that have been referenced under the Directive by the
European Commission. In 2019, all referenced standards were included in a new
Decision181
, to "create a complete list of references". These standards concern the
following product types:
 Furniture
 Gymnastic equipment
 Stationary training equipment
 Child use and care articles
 Bicycles
 Paragliding equipment
 Internal blinds
 Diving accessories
 Lighters
 Roller sports equipment
 Decorative oil lamps
 Children's clothing
 Floating leisure articles
 Cigarettes (ignition propensity)
 Child protective products
 Audio, video and similar (safety requirements)
 Information technology equipment (safety - general requirements)
To date, the Commission applied the procedure provided for in Article 13 of the
Directive on five occasions:
 Firstly, it was used to extend the ban on phthalates in toys182
during the period up
to the adoption of the permanent ban under Directive 2005/84/EC183
 The next measure based on this Article was the Decision of 11 May 2006184
requiring Member States to ensure that cigarette lighters placed on the EU market
be child-resistant and to prohibit placing on the market lighters which resemble
objects that are particularly attractive to children (so-called "novelty lighters"). In
the absence of a suitable revised version of the European standard on child safety
requirements for lighters which could be referenced in the Official Journal of the
EU as providing a presumption of conformity with the safety requirement of the
GPSD, it was necessary to extend the period of validity of this Decision ten times
181
Commission Implementing Decision (EU) 2019/1698 of 9 October 2019
182
Commission Decisions 2004/178/EC (OJ L 55, 24.2.2004, p. 66), 2004/624/EC (OJ L 280, 31.8.2004,
p. 34) and 2004/781/EC (OJ L 344, 20.11.2004, p. 35)
183
Directive 2005/84/EC (OJ L 344, 27.12.2005, p. 40)
184
Commission Decision 2006/502/EC (OJ L 198, 20.7.2006, p. 41)
41
(extension which was valid until 11 May 2017185
). A new version of the above-
mentioned European standard was referenced in the Official Journal of the EU in
2017.
 Another measure was taken by the Decision of 21 April 2008186
requiring
Member States to ensure that magnetic toys placed or made available on the
market display a warning about the health and safety risks they pose. Magnets
used in toys have become more powerful, but also detach more easily, thus
presenting life-threatening risks if ingested (as they can perforate the stomach or
intestines). The Commission adopted this temporary measure which was valid
until 21 April 2009, until a relevant safety standard was adopted and referenced
by the Commission in the Official Journal of the EU.
 Another measure was adopted on 19 March 2009187
, requiring Member States to
ensure that products containing the biocide dimethylfumarate are not placed or
made available on the market.
 Finally, on 9 August 2011 a Decision was adopted regarding the compliance of
standard EN 16156:2010 and the assessment of the ignition propensity of
cigarettes188
.
The enforcement of the FIPD has been unequal across the EU as Member States have
divergent opinions on its interpretation, in particular its relation with other pieces of
sectorial legislation, such as the Toys Safety Directive and the Cosmetic, Detergents and
CLP Regulations. The Commission proposed the withdrawal of the FIPD and the
inclusion of its provisions within the General Product Safety Regulation in the 2013
Package on Product Safety and Market Surveillance. The Commission has since
withdrawn this Package and therefore the FIPD remains in force. More information about
the implementation of the FIPD is provided in Appendix 4.
4. METHOD
2.2. Short description of methodology
The evaluation was carried out according to the Commission’s evaluation techniques and
triangulation methods to ensure robustness of the information obtained. The evaluation
received input from consultation activities, official statistics and studies. The evaluation
followed several steps to collect both qualitative and quantitative data from the relevant
stakeholders and national authorities.
185
In chronological order: Commission Decision 2007/231/EC (3) until 11 May 2008, Commission
Decision 2008/322/EC (4) until 11 May 2009, Commission Decision 2009/298/EC (5) until 11 May 2010,
Commission Decision 2010/157/EU (6) until 11 May 2011, Commission Decision 2011/176/EU (7) until
11 May 2012, Commission Implementing Decision 2012/53/EU (8) until 11 May 2013, Commission
Implementing Decision 2013/113/EU (1) until 11 May 2014, Commission Implementing Decision
2014/61/EU (2) until 11 May 2015, Commission Implementing Decision (EU) 2015/249 (3) until 11 May
2016 and Commission Implementing Decision (EU) 2016/575 until 11 May 2017
186
Commission Decision 2008/329/EC (OJ L 114, 26.4.2008, p. 90)
187
Commission Decision 2009/251/EC (OJ L 74/32)
188
Commission Decision 2011/496/EU (OJ L 205, 10.8.2011, p. 31–32)
42
2.3. External studies
This Evaluation is based on two main building blocks:
1) An external study on the Directive aimed to assess its functioning and performance and
to identify potential shortcomings and whether improvements should be envisaged.
2) Complementary studies and research. This includes the following:
a. A study on product recall effectiveness
b. A study to support the preparation of the Implementation report of the
Directive, which ran from August 2019 to March 2020.
c. The Eu opea Co issio s Joi t Resea h Ce t e s study o the assess e t of
the opportunities for increasing the availability of EU data on consumer
product- related injuries189
.
Figure 3: External studies supporting the GPSD initiative
2.4. Open public consultation (OPC) and stakeholders’ workshops
The public consultation on the EU's New Consumer Agenda, that included a specific
section on the Directive, ran between 30 June 2020 and 6 October 2020. The consultation
was available in 24 official EU-languages and respondents could reply in any of these
languages.
The objective of this consultation was to gather the views of the public on the new
European consumer policy for the next period, the so-called ‘New Consumer Agenda',
including a questionnaire on the Directive. These views provided input into the
evaluation and impact assessment of the General Product Safety Directive. The questions
on the review of the Directive were tailored to two main categories of stakeholders,
respectively the general public and stakeholders who are familiar and have certain
experience with market surveillance and product safety rules. There were 257
respondents that answered at least one question related to the General Product Safety
Directive. The majority of respondents were business associations and EU citizens (each
26%), followed by company/business organisations (15%). Other respondents included
189
https://injuryprevention.bmj.com/content/early/2020/09/28/injuryprev-2020-043677.full
43
public authorities (11%), consumer organisations (8%), non-governmental organisations
(NGOs) (7%), academic/research institutions (3%), non-EU citizens (1%) and other
respondents (3%).
Moreover, the Commission set up a Sub-group of the Consumer Safety Network on
artificial intelligence, connected products and other new challenges on product
safety. The Sub-group’s tasks were to assess whether and to what extent existing product
safety frameworks are adapted to emerging market realities (connected products, AI,
software, etc.) and, in particular, to assist the Commission in developing an EU-wide
assessment on the need for the possible adaptations of the General Product Safety
Directive in this regard. The Sub-group was established in November 2019.
The Commission also organised a series of workshops as part of a broader engagement
with stakeholders and evidence-collection strategy, supporting the evaluation of the E-
Commerce Directive and the impact assessment for the Digital Services Act as well as
the revision of the General Product Safety Directive. The workshops aimed to gather
information on the role online marketplaces are playing in the current online supply chain
and views on their role for the future. The main findings of the workshops are detailed in
Annex 13.
Finally, the Commission also engaged with different stakeholders from around the world
in the frame of the International Product Safety Week in November 2020, for the
discussion of different topics including products recalls and traceability.
2.5. Limitations and robustness of findings
One important limitation of this evaluation is the difficulty to differentiate data
between harmonised and non-harmonised products. The Directive plays a
complementary to the product harmonisation legislation, and consequently it is extremely
complicated to isolate the effects and impacts of the Directive. Moreover, although
market surveillance rules differ for harmonised and non-harmonised products, in practice
Member States organise their investigations and enforcement actions in a comprehensive
manner, without distinguishing its legal basis, rendering almost impossible to evaluate
both frameworks separately. As a mitigating measure, these aspects were addressed in
the consultations activities.
Moreover, it is worthwhile to mention that data coming from the Safety Gate/RAPEX
should be interpreted cautiously. Measures reported in the system might be affected by
multiple factors, in particular inspection priorities of authorities, perceived risks, etc.
Besides, notifications only reflect injury events if these are communicated to the market
surveillance authorities, which is not systematically the case and not based on the actual
frequency of injuries. Notifications also refer to products, but not the exact number of
articles or items affected by the measure. This does not in any way limit the value of the
system, but shows that its data cannot be simply used as proxy for product safety
trends or for analysing the preventive potential of enhanced product design or safety
features. Consequently, the evaluation has considered data from Safety Gate/RAPEX as
one indicator among several, and it is complemented by other datasets and sources.
44
During the consultations activities, considerable efforts were made to reach out SMEs.
Otherwise explicitly mentioned, opinions of SMEs do not significantly differ from the
opinion of other business stakeholders.
Information related to the product safety aspects of emerging digital technologies, i.e.
Internet of Things (IoT) and Artificial Intelligence (AI), is scarce and often not
sufficiently mature, since some of these technologies are still under evolution. The
conclusions of the evaluation in this aspect are based on information collected from
stakeholders, studies, media sources and forecasts.
Notwithstanding the specific limitations mentioned above, which could at least partially
be compensated by the answers obtained during the consultation activities, the overall
availability and reliability of data and the approach taken is generally considered
satisfactory.
5. ANALYSIS AND ANSWERS TO THE EVALUATION QUESTIONS
5.1 Relevance
5.2.1. General relevance of the Directive
The Directive was adopted in order to respond to the following needs:
 Consumer products placed on the EU market for consumers do not cause harm to the
safety and health of people.
 Free movement of goods in the Single Market.
These original needs remain relevant and match the general objectives of the Directive.
However, it is important to consider whether the specific tools and provisions provided
by the Directive continue to be relevant and future-proof.
A large majority of respondents of the OPC expressed that the current EU safety rules for
non-food consumer products covered by the Directive could be improved in specific
areas to be more adequate to protect consumers (see figure 4). Nearly one in four
respondents held that the current rules were fully adequate, whereas only a small
minority considered them not to be adequate at all.
45
Figure 4: Replies to the OPC question: “In your view, to what extent are current EU
safety rules for non-food consumer products covered by the GPSD adequate to protect
consumers?”
Source: OPC. Total no. of respondents: 214, Single-choice question
When asked about problems related to the implementation of safety rules for products
covered by the Directive, respondents of the OPC (see figure 5) most commonly
expressed that rules were not adapted to online trade (39%) and that the rules were not
appropriately enforced (39%). More than a third also considered the rules not to be
adapted to new technologies (36%) and perceived legal definitions as not sufficiently
clear or outdated (35%). Slightly less than a third of respondents (30%) reported that
roles and obligations of different economic operators were not appropriately defined and
that there were difficulties for consumers to report unsafe products. Lastly,
approximately a fifth of respondents regarded as problematic that there were no specific
requirements for product recalls (22%) or listed other issues (23%).
Figure 5: Replies to the OPC question: “Are you aware of any problems related to the
implementation of EU safety rules for consumer products covered by the GPSD?”
Source: OPC. Total no. of respondents: 205, Multiple-choice question
23%
22%
30%
30%
35%
36%
39%
39%
OTHER
THERE ARE NO SPECIFIC REQUIREMENTS FOR
PRODUCT RECALLS
DIFFICULTIES FOR CONSUMERS TO REPORT UNSAFE
PRODUCTS
THE ROLES AND OBLIGATIONS OF DIFFERENT
ECONOMIC OPERATORS ARE NOT APPROPRIATELY…
LEGAL DEFINITIONS SUCH AS PRODUCT , SAFE
PRODUCT OR PLACING ON THE MARKET ARE NOT…
RULES ARE NOT ADAPTED TO NEW TECHNOLOGY
PRODUCTS (E.G. CONNECTED DEVICES)
PRODUCT SAFETY RULES ARE NOT APPROPRIATELY
ENFORCED
RULES ARE NOT ADAPTED TO ONLINE TRADE
Fully
adequate
23%
Could be
improved in
specific
areas
71%
Not
adequate at
all
6%
46
Moreover, some stakeholders indicated in their comments that the COVID-19 crisis has
led to emerging needs in relation to product safety. They expressed that consumer safety
was a cardinal value, as the COVID-19 crisis has accelerated existing trends, such as the
increased use by consumers of online retail for non-food purchases.
In the sub-sections below the relevance of specific aspects and tools provided by the
Directive are analysed.
5.2.2. Relevance of of the Directive vis-à-vis E-commerce and direct
imports
While the Directive does not establish specific provisions related to online sales, it
clearly applies to all sales channels, offline and online. Case law190
has clarified that EU
product legislation also applies to cases where online sellers based outside the EU target
consumers in the EU. The assessment to determine if an online offer targets EU
consumers must be done on a case-by-case basis. The following aspects could be
considered: the international nature of the activity, use of a language and currency (for
example the euro) of the Member States, a domain name registered in one of the Member
States, geographical areas to which dispatch is possible.
At the same time, the number of consumers buying products online has drastically
increased since the adoption of the Directive. In 2007, less than half of Internet users
bought or ordered goods or services for private use in the previous 12 months; in
contrast, in 2019 that figured increase to 71%, with five countries (UK, Denmark,
Netherlands, Sweden and Germany) exceeding 80%191
. Furthermore, the COVID-19
crisis has accelerated this trend: in the EU-27, retail sales via mail order houses or the
Internet in April 2020 increased by 30% compared to April 2019, while total retail sales
diminished by 17.9% (Figure 6)192
.
190
Judgment of the Court of Justice of the European Union of 12 July 2011, Case C-324/09, L'Oréal/eBay,
paragraph 65 and Judgment of the Court of Justice of the European Union of 7 December 2010 in joined
Cases C-585/08 and C-144/09 Peter Pammer v Reederei Karl Schlüter GmbH & Co KG, and Hotel
Alpenhof GesmbH v Oliver Heller.
191
Eurostat
192
OECD - E-commerce in the time of COVID-19, http://www.oecd.org/coronavirus/policy-responses/e-
commerce-in-the-time-of-covid-19-3a2b78e8/#biblio-d1e705
47
Figure 6 - Retail turnover, year-on-year change, EU-27 (July 2019- July 2020)
Source: OECD, E-commerce in the time of COVID-19
Evidence gathered in the context of the evaluation showed that online sales have led to
problems in enforcing the Directive for mainly two reasons: difficulties for the
enforcement of the control of products sold online and unavailability of responsible
economic operators measures could be effectively addressed to in case products are
directly imported from outside the EU.
In the frame of the preparation of the Notice on the market surveillance of products sold
online193
, as well as discussions held within the Coordinated Activities on Product Safety
2019194
, the following challenges were identified for market surveillance authorities in
relation with the safety of products sold online:
 Sampling: there is a lack of clear competences for authorities to engage in
mystery shopping at the level of the Directive. There are also legal restrictions
for authorities in some Member States that prevent them from hiding their
identity when making online inspections, which can make online product safety
checks ineffective.
 Testing: authorities face challenges in conducting risk assessments or safety
tests due to the lack of physical access to products.
193
C/2017/5200 Commission Notice on the market surveillance of products sold online
194
CASP 2019 Coordinated Activities on the Safety of Products – Online Market Surveillance – Final
report
https://ec.europa.eu/consumers/consumers_safety/safety_products/rapex/alerts/assets/documents/casp/CAS
P_HA_online_market_final_report.pdf
48
 Resources: the lack of financial resources, or even the lack of credit cards that
authorities can use for online sampling, combined with the lack of competence to
ask traders for reimbursement of the product price makes online market
surveillance difficult. In contrast, during offline market surveillance activities,
authorities can usually seize products free of charge.
 Identification of economic operators: authorities face difficulties finding and
contacting the trader selling through online marketplaces.
 Product identification: there is often lack of sufficient data on products
available online (no images of labels, bad quality pictures, and no technical
data). The Directive does not establish obligations to economic operators on
what specific information needs to be provided in an online offer.
 Reappearance of unsafe products and repeat infringers: it often happens that
offers containing unsafe products that were already removed are re-uploaded on
the same or a different online marketplace.
 Awareness: there is lack of awareness among consumers and businesses about
buying and selling safe and compliant products online.
In addition, e-commerce allows consumers to purchase directly from operators located
outside the EU. This makes it more difficult for authorities to check the safety of
products entering the single market. These direct imports have increased in the latest
years: around 150 million small consignments are imported free of VAT into the EU
each year195
and it has been reported that in 2017 there were 150.000 private
consignments coming from China to individual consumers per day196
. This is a challenge
for product safety as, in the case of non-harmonised products197
, there is often no
economic operator within the EU available that market surveillance authorities could turn
to for enforcement measures in the EU. As noted above, the Directive imposes
obligations on the producer as well as on distributors. In practice this means:
 The manufacturer’s representative and the importer might be considered as
producers according to Article 2(e) of the Directive.
 Where the producer is domiciled in a non-EU/EEA country, it is in most cases
outside the reach of the market surveillance authorities of the Member States to
impose measures on them. Market surveillance authorities may be able to
cooperate with the authorities of the non-EU/EEA country where the producer is
195
European Commission , Memo 2017 - Modernising VAT for e-commerce
https://ec.europa.eu/commission/presscorner/detail/en/MEMO_16_3746
196
Eurocommerce – Creating a level-playing field for retail in Europe – August 2019
197
For a certain number of harmonised products, Article 4 Regulation 2019/1020 foresees the figure of the
responsible economic operator in the EU.
49
domiciled (which is the exception198
), but in principle they cannot take measures
themselves.
 Moreover, if the producer sends non-harmonised products directly to the
consumer, there is no (other) economic operator with product safety obligations
involved in the EU.
The only supply chain actor that is often involved before the consumer gets the product is
an online platform. However, it seems that online marketplaces do not fall under the
definition of distributor under the Directive and therefore they are not subject to product
safety enforcement measures in a way that is foreseen for producers and distributors.
Moreover, even where Member States have taken measures to recall unsafe products,
evidence from different investigations carried out by stakeholders show that recalled
products often continue to be sold or to reappear on the market in online sales channels.
The European Commission already identified these challenges and took some non-
legislative measures to try to tackle the issue. First, it adopted a Notice on the market
surveillance of products sold online199
to assist Member State authorities in the
enforcement of EU legislation on the safety and compliance of non-food products and to
contribute to a more uniform and coherent application of that legislation in the online
environment. In addition, the European Commission facilitated the signature by nine
online marketplaces of the Product Safety Pledge200
. This initiative, originally signed in
2018 and the first of its kind in the product safety area, sets out specific voluntary actions
in 12 different areas by online marketplaces to improve product safety going beyond
what is already established in EU legislation. As part of the Pledge, signatory online
marketplaces have committed to report to the European Commission every six months on
the actions taken to implement the Product Safety Pledge, with the inclusion of key
performance indicators.
However, stakeholders have pointed out that these measures, although positive, do not
resolve the underlying problems related to online sales. The Notice on the market
surveillance of products sold online has proven to be a useful guidance for authorities,
but does not solve the issue that in many countries they have legal and practical barriers
to carry out online investigations and take appropriate measures. The Product Safety
Pledge has set the grounds for an increased cooperation framework between online
marketplaces and market surveillance authorities. However, as pointed by authorities and
stakeholders, it remains voluntary, so it cannot be enforced in case of infringements, and
there are still many actors on the market that have not adhered to the initiative.
198
Direct cooperation of market surveillance authorities with other relevant authorities in non-
EU/EEA countries is only carried out in a minority of countries. Authorities from only five countries
(Germany, France, Ireland, Lithuania, United Kingdom) reported cooperating once every three months or
more often with non-EU/EEA country authorities, see GPSD implementation study, p 105.
199
C/2017/5200 Commission Notice on the market surveillance of products sold online
200
Product Safety Pledge, available at:
https://ec.europa.eu/info/sites/info/files/voluntary_commitment_document_4signatures3-web.pdf
50
In conclusion, it can be noted that the Directive is not adapted to the specific
challenges posed by online sales, including the increased level of direct B2C imports
towards the EU, facilitated by online marketplaces and fulfilment service providers201
.
Many stakeholders have noted that it would be beneficial to adjust the Directive in
relation to these additional economic operators, to address the newly emerged needs
related to the online environment and to increase the safety of products sold online to EU
consumers.
5.2.3. Relevance of the Directive vis-à-vis new technologies
At the time of the adoption of the Directive the number of consumer products
incorporating new technologies was scarce. This is not the case anymore. There were
14.2 billion connected devices in 2019 worldwide, a figure that is estimated to go up to
25 billion by 2025, of which 4.9 billion will be in Europe202
.
Almost half the respondents of the OPC considered the safety of products involving new
technologies to be not adequately regulated (47%), with only 18 % stating the opposite.
The other 35% did not know (see figure 7).
Figure 7: Replies to the OPC question: “Do you think that the safety of products
involving new technologies is adequately regulated?”
Source: OPC. Total no. of respondents: 227, Single-choice question
201
Fulfilment service providers are defined in Regulation 2019/1020 as follows: ‘fulfilment service
provider’ means any natural or legal person offering, in the course of commercial activity, at least two of
the following services: warehousing, packaging, addressing and dispatching, without having ownership of
the products involved, excluding postal services as defined in point 1 of Article 2 of Directive 97/67/EC of
the European Parliament and of the Council ( 31), parcel delivery services as defined in point 2 of Article 2
of Regulation (EU) 2018/644 of the European Parliament and of the Council ( 32), and any other postal
services or freight transport services
202
Netherlands Entreprise Agency - https://www.cbi.eu/market-information/outsourcing-
itobpo/intergrated-internet-things/market-potential
Yes
18%
No
47%
Don't know
35%
51
The European Commission published in February 2020 a report on the safety and
liability of new technologies203
. The report identified uncertainties linked to the
application of the product safety framework with respect to the specific risks posed by AI
systems and other digital technologies. It concluded that the current product safety
legislation, including the General Product Safety Directive and harmonised product
legislation that follows the horizontal rules of the “New Legislative Framework”204
,
already supports an extended concept of safety protecting against all kind of risks arising
from new technology products. However, there were a number of unclear areas according
to the report:
 EU legislation does not comprehensively include explicit provisions regarding
some categories or risks posed by new technologies, such as cybersecurity risks
that affect safety. Such provisions would provide a better protection of users and
more legal certainty for businesses.
 While the Union product safety legislation takes into account the safety risks
stemming from software integrated in a product at the time of its placing on the
market and, potentially, subsequent updates foreseen by the manufacturer, there is
a lack of specific and/or explicit requirements on standalone software (e.g. an
'app' that would be downloaded). The product safety framework does not provide
for additional obligations for manufacturers to ensure that they provide features to
prevent the upload of software having an impact on safety during the lifetime of
AI products205
.
 Regarding the concept of “placing on the market”, as far as the future
“behaviour” of AI products can be determined in advance by the risk assessment
carried out by the manufacturer before the products are placed on the market, the
Union product safety framework already sets out obligations for producers to take
into account in the risk assessment the “use” of the products throughout their
lifetime. However, there may be also situations in the future where the outcomes
of the AI systems cannot be fully determined in advance. In such a situation, the
risk assessment performed before placing the product on the market may no
longer reflect the use, functioning or behaviour of the product.
Moreover, the Subgroup on AI, connected devices and other new challenges in
product safety to the Consumer Safety Network reached the following conclusions
regarding the relevance of the Directive vis-à-vis new technologies:
 The legal definition of safe product of the Directive can be understood as
addressing many types of risks by which a product can, directly or indirectly,
cause harm to consumers. However, it does not explicitly refer to some risks
203
Report on the safety and liability implications of Artificial Intelligence, the Internet of
Things and robotics COM/2020/64 final
204
Regulation (EC) No. 765/2008 and Decision (EC) No. 768/2008
205
The Radio Equipment Directive envisages the possibility to adopt delegated acts under its Article 3(3)(i)
that would partially address this issue.
52
posed to new technologies, such as cybersecurity risks that affect safety (“cyber-
safety”).
 New technologies can also pose risks not only because they can have a direct
impact on the health and safety of the consumers, but also because through
connectivity, they can be indirectly used as a tool to put at risk their personal
security. Having said that, it is unclear under which legal or policy instrument
such personal security risks should be tackled so that consumers are effectively
protected against such threats.
 Regarding mental health, while they are not a new phemonenon, there is
evidence that new technologies can have a psychological impact on users206
.
However, the Sub-group agreed that risks to mental health that are not intrinsic to
the product, but come from the use of a product in particular ways, should not be
considered part of the concept of “safety” in the Directive.
 Despite that the current version of the Directive is in theory broad enough to
cover safety risks resulting from software interacting with the product, it does
not explicitly mention it, creating legal uncertainty in this regard.
 One of the common characteristics of AI and IoT products is the presence of
software that can change/evolve over time. This challenges the traditional
meaning of the concept of placing on the market of the Directive.
Finally, it was mentioned by one stakeholder of the OPC that 3D printing could also
affect product safety. However, it appears that this technology does not seem to present
new safety challenges. 3D printers are already covered by sectorial legislation
(Machinery Directive). For the products created by the 3D printer, if they are placed on
the market, they need to comply with product safety legislation, as 3D printing is just a
new manufacturing technique. Furthermore, there is a similarity between products
“printed” by consumers for their own use and traditional DIY products created by
consumers.
5.2.4. Relevance of definitions of the Directive
In the context of the GPSD Study, stakeholders were asked whether or not they
considered the key concepts of the Directive to be still relevant or whether they saw a
need for it to be adapted to changed circumstances. Often, comparable numbers of
stakeholders even of the same group – companies/business associations, authorities and
other stakeholders – suggested that a concept should be changed or kept as it is. Figure 8
provides an overview of results, and indicates the number of respondents that considered
that a specific concept needed to be clarified and updated:
206
Dresp-Langley B. Children's Health in the Digital Age. Int J Environ Res Public Health. 2020 May
6;17(9):3240. doi: 10.3390/ijerph17093240. PMID: 32384728; PMCID: PMC7246471.
53
Figure 8: Replies to the question of the GPSD Study “Considering the emergence of new technologies
and new business models/actors: Is there a need to clarify and update [the following] terms and concepts
as currently used in the GPSD? – Number of respondents indicating Yes”
Source: GPSD Study
It appears that most of stakeholders detected that the current definitions for some
concepts do not appropriately reflect the needs of the Directive. The reasons highlighted
related to divergences with other pieces of EU legislation (see coherence section), but
also to the fact that some of the definitions are now outdated due to the development of
new technologies and online sales.
5.2.5. Relevance of environmental issues, including chemical risks
The definition of safety in Art. 2(b) of the Directive covers all product-related risks that
can affect the safety and health of persons. This definition therefore also includes risks
related to environmental pollutants in products that can affect human health (e.g. heavy
metals such as lead and cadmium, phthalates, etc.). A broader scope of risks to be
considered in addition to those related to the health and safety of consumers, such as
security and environmental risks, was only introduced with Regulation (EC) 765/2008
setting out the requirements for accreditation and market surveillance relating to the
marketing of products. Since then, the Safety Gate/RAPEX applies to measures which
prevent, restrict or impose specific conditions on the marketing and use of products
4
11
13
17
22
22
25
24
30
38
11
19
17
20
20
22
25
27
24
26
8
4
4
16
15
17
15
16
16
19
0 10 20 30 40 50 60 70 80 90
Other
“Recall”
“Withdrawal”
“Product”
“Dangerousproduct”
“Safe product”
“Producer”
“Seriousrisk”
“Distributor”
“Placing on the market”
Number of respondentsindicating YES
Companies/ Businessassociations Authorities Other stakeholders
54
posing a serious risk to the health and safety of consumers or to measures which prevent,
restrict or impose specific conditions on the marketing and use of products posing a
serious risk to the health, safety or, for harmonised products, other relevant public
interests, including environmental risks.
Data from the Safety Gate/RAPEX shows how the number of notifications indicating
“Environmental” as the type risk was slowly increasing, and in 2019 the number of
notifications suddenly more than tripled. This rapid rise can be explained by the revision
of the guidelines of the Rapid Alert System adopted in 2018 that clarified the cases where
risk assessments related to chemical risks do not need to be performed in order to
consider a product as unsafe. The revised guidelines have indeed set out that if a
chemical substance in a product is already banned or restricted by Union legislation, the
product can be considered to pose a serious risk, without the need to perform a specific
risk assessment. This has simplified the risk assessment process for chemicals, including
environmental pollutants with health impact.
Another factor to take into consideration in the surge of notifications on environmental
risks concerns the increased focus on motor vehicles’ emissions in the aftermath of the
“Dieselgate”, and the subsequent number of alerts regarding vehicles not satisfying
emission standards.
Figure 9: Number of notifications to Safety Gate/RAPEX mentioning “Environment” among the risk
types
Source: Safety Gate/RAPEX.
A further analysis of the recent dataset of the system shows that there seems to be a
general tendency to identify the risk as “chemical” if the substance in the product poses a
direct health risk to the consumer, e.g. acute poisoning. For the years 2013 to 2019, the
dataset contains 3 606 notifications (approximately 25% of all notifications in the period)
of products presenting a “Chemical” risk, more than ten times the number of notifications
that indicate "Environment" as risk type. However, substances presenting a chemical risk
will often also have an adverse effect on the environment. In some cases, the products
notified in the rapid alert system contain substances that can be dangerous for human
health (e.g. cadmium, lead) but that are contained in parts of the product that the
consumer will not be in direct contact with (e.g. solders). The risk will thus materialise
55
during the “end of life” phase of the product, with a possible accumulation of these
dangerous substances in the environment. While the definition of safety in the Directive
is considered to cover risks related to environmental pollutants in products that can affect
human health, this coverage is not explicitly stated. This leaves room for interpretation
regarding products posing long-term risks stemming from the toxicity of environmental
pollutants. The extent to which the Directive is well adapted to environmental issues with
health impact therefore depends on the interpretation of the definition of safety in the
Directive and could be clarified.
As regards emerging chemical hazards, the Directive has showed in a number of
occasions its particular relevance in being a safety net and the role of the rapid alert
system as a ”watchdog” for new chemical hazards. Examples are Dimethilfumarate and
Bisphenol A, for which restrictions were triggered after measures against products
containing these chemicals were reported in Safety Gate/RAPEX (see above section 3).
It can therefore be concluded that the Directive remains relevant in relation with
environmental and chemical risks. However, legal changes might be needed to clarify
that risks related to the indirect and long-term health effects of environmental pollution
under the scope of the Directive.
5.2.6. Conclusion on relevance
Overall, the Directive, its objectives and the product safety framework that it establishes
remain relevant for the needs of avoiding harm to consumers and building trust on
consumer products as a prerequisite for the free movement of goods. However, the
growth of online sales and the development of new technologies show that some of the
provisions of the Directive are not well adapted to respond to its objectives. In addition,
some of the definitions and provisions are now outdated and could be subject to fine-
tuning.
5.2 Effectiveness
5.2.1. Extent to which the Directive has been effective in contributing to
consumer safety
Several indicators and data sources can be used to assess the extent to which the GPSD
and related market surveillance and notification procedures have been effective in
achieving a high level of consumer protection through the reduction of unsafe products
on the EU market. These include the following indicators/sources:
 Trends in the number of Safety Gate/RAPEX notifications;
 Share of unsafe products found during market surveillance inspections;
 Data on product-related injuries;
 Assessment of consumers and stakeholders concerning the level of product
safety achieved.
None of these indicators is without limitations, and to obtain an overall picture they have
to be considered together.
56
First, the number of alerts to Safety Gate/RAPEX show a progressive increase over the
years; since 2012 the total number of notifications was just above 2,000 alerts a year. In
2019, a total of 2,243 alerts were circulated in the system. The three product categories
with the largest number of notifications (toys, clothing and motor vehicles) account for
between 1 050 and 1 350 annual notifications with an average of 1 230 notifications per
year.
Notifications in the Safety Gate/RAPEX may include information concerning the number
of items that are being affected by the measures taken, e.g. the number of items that were
rejected at the EU border, or the number of items that were recalled from the market.
This information is part of the notification that is only accessible for market surveillance
authorities. Table 1 shows an extract of this data, covering a twelve month period from
May 2019 to April 2020, and including information for a total of 536 notifications in
which more than 1 000 items were affected.
Table 1: Number of notifications and number of items affected by measures taken per product category
(May 2019 to April 2020)
Product category
Number of
notifica-tions
Number of items affected, with data referring to ... Total
National
circulation
EU/EEA
circulation
Worldwide
circulation
Unknown
circulation
Motor vehicles 272 27 240 1 049 811 9 424 961 17 462 909 27 964 921
Construction products 1 4 500 000 4 500 000
Protective equipment 11 4 800 4 290 000 16 545 4 311 345
Electrical appliances and
equipment
30 638 177 63 278 1 146 608 210 719 2 058 782
Toys 126a)
183 800 539 534 483 901 1 207 235
Other 7 10 700 528 594 539 294
Cosmetics 12 56 560 208 063 264 623
Lighting equipment 11 12 969 231 657 244 626
Lighting chains 17 105 520 51 600 157 120
Childcare articles and
children's equipment
9 8 111 131 817 139 928
Chemical products 4b)
2 160 75 073 77 233
Kitchen/cooking accessories 3 5 952 57 249 63 201
Hobby/sports equipment 6 13 197 45 734 58 931
Jewellery 5 1 200 51 394 52 594
Clothing, textiles and
fashion items
7 5 031 22 073 24 985 52 089
Machinery 3 28 556 28 556
Decorative articles 4 11 000 5 052 16 052
Pyrotechnic articles 1 14 400 14 400
Measuring instruments 2 3 648 3 000 6 648
Gas appliances 3 6 140 6 140
Recreational crafts 1 2 953 2 953
Gadgets 1 1 008 1 008
Total 536 988 528 6 084 641 10 571 569 24 122 941 41 767 679
Source: Safety Gate/RAPEX. Bold = Non-harmonised product category.
57
As the table shows, the listed notifications in this twelve-month period affected some
41.8 million items in total or 79 900 items per notification on average. The largest
category is “Motor vehicles” with the highest number of notifications (272) and the
highest number of items affected (approximately 28 million items). Notifications that
concern clearly non-harmonised product categories (marked in bold) account for a total
of 169 548 items in this five-month period.
Regarding evidence from market surveillance activities of Member States authorities,
table 2 presents data on the total number of consumer products inspected by market
surveillance authorities in the EU/EEA member countries, as well as the total number of
unsafe consumer products found207
. The share of unsafe products found by market
surveillance authorities in their inspections is frequently between 2% and 16% of total
consumer products inspected (interquartile range), with the median value being 4%208
. In
some countries this share is much higher: from five countries it was reported that the
share of unsafe products of total consumer products inspected is close to 20% or higher.
However, the GPSD Study notes that the data has been reported from various sources
according to different criteria, so that these figures have to be interpreted with care. As
market surveillance authorities often sample according to risk-based criteria (i.e. focusing
on risky products, conducting visual inspections to choose for testing products that can
potentially be unsafe), this figure is not representative of the incidence of unsafe
consumer products on the market209
.
Table 2: Share of inspected consumer products and share of unsafe products found (last
available year, mostly 2018 or 2019)
Country Total number of
consumer products
inspected
Total number of unsafe
consumer products
found
Share of unsafe products
found (of total products
inspected)
Austria : : :
Belgiuma)
710 283 40%
Bulgariap)
4 624 120 3%
Croatiaq)
4 475 47 1%
Cyprusb)
7 105 301 4%
Czech Republicc)
17 088 156 1%
Denmarkd)
2 500 520 21%
Estoniae)
8 317 46 1%
Finlandr)
85 31 36%
207
The table includes combined figures for harmonised and non-harmonised products, as separate statistics
are rarely available.
208
The interquartile range is the data between the 25th and 75th percentile of a data series. The
median is the middle value, or 50th percentile. In other words, the interquartile range comprises the
quartiles below and above the median.
209
This risk-based approach also affects the type and number of Safety Gate/RAPEX notifications,
which may be influenced by changing priorities concerning which risks are considered by authorities when
conducting inspections.
58
Frances)
3 980 760 19%
Germanyf)
27 541 12 715 46%
Greeceg)
850 100 12%
Hungary : : :
Irelandt)
492 : :
Italy : : :
Latviah)
1 144 64 6%
Lithuaniai)
2 000 59 3%
Luxembourgj)
867 15 2%
Maltak)
1 313 22 2%
Netherlands 6 500 n.a. n.a.
Poland l)
8 671 440 5%
Portugalu)
: : :
Romaniam)
15 245 41 0.3%
Slovenian)
605 9 1%
Slovakia : : :
Spain : : :
Sweden : : :
UK : : :
Source: GPSD Study, from data provided by market surveillance authorities
Furthermore, according to the analysis presented in Annex 4, it is estimated that
consumer detriment linked to injuries and premature deaths from unsafe products is
EUR 76.6 billion per year. This is the sum of detriment caused by non-fatal product-
related injuries, and the cost of premature death due to fatalities caused by mechanisms
relevant for product safety (such as tools, strangulation, electric current, or fire) occurring
outside of work-related locations210
. The analysis based on previous research and
interviews with product safety experts concluded that 15% is a reasonable and cautious
estimate for the proportion of the total detriment that was caused by consumer products,
or could have been prevented through better design, instruction or a safety device. On
this basis, the preventable detriment suffered by EU consumers and society due to
product-related accidents can be estimated at EUR 11.5 billion per year.
In addition to this injury related detriment, the GPSD Study estimates that the consumers
also suffer financial costs of a total value of EUR 19.3 billion for 2019 arising from
the fact they have purchased unsafe products that they would not have purchased if
they knew these products were unsafe211
.
210
These estimates are based on the best possible approximation of product-related injuries and fatalities.
The detriment cannot be estimated separately by categories of products and therefore include all consumer
products, harmonised and non-harmonised products.
211
This relates to non-harmonised consumer products covered by the GPSD. This is based on the
assumption that willingness to pay (WTP) for a product depends on the utility of the product for the
purchaser. WTP is equal or higher as the price for which a product is purchased by a consumer, as
otherwise the transaction would not take place. It is very likely that WTP would be close to zero for an
unsafe product (nobody wants to buy e.g., a dangerous childcare product) – so the loss in consumer welfare
59
Consumer survey data can also provide supporting evidence regarding product safety, at
least to the extent that consumers perceive product safety to be an issue relevant to them,
based on their own experiences, the experiences of friends and media reports. EU data
exists concerning the consumer perception of the level of product safety in the EU. The
data derives from the Commission’s regular surveys on consumer attitudes toward cross-
border trade and consumer protection since 2008 (the last relevant survey was conducted
in 2018).
Figure 10: Percentage of consumers who agree that essentially all non-food products
are safe or that a small number of non-food products are unsafe (EU average), 2008-
2018
Source: Compilation of the GPSD Study based on data from the Commission’s 2016 and
2018 survey of consumers’ attitudes toward cross-border trade and consumer protection
Figure 10 indicates that consumer trust in product safety in the EU has shown a slight
increase over time, with the proportion of consumers agreeing that essentially all non-
food products in their country are safe (or that only a small number are unsafe) increasing
from 65% in 2008 to 78% in 2016, before decreasing again to 70%; a possible
explanation provided by stakeholders for this late drop is the increase of unsafe products
found online. The largest increase (9 percentage points) occurred between the 2014 and
2016 surveys, before returning in 2018 to slightly above the 2014 level.
During the GPSD study, stakeholders were consulted on their views about to what extent
the Directive has been effective in reaching its objective of protecting consumers from
unsafe products. Their overall opinion was positive, while SME provided a slightly more
negative opinion in this regard than other businesses.
is at least the price to which the product was purchased.This calculation assumes that the consumers do not
get reimbursed for the unsafe product.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
2008 2009 2010 2011 2012 2014 2016 2018
60
The available data also confirms that large numbers of unsafe products that could affect
the safety of EU consumers are rejected at the borders, withdrawn from the market or
recalled. This implies that a reduction of unsafe products on the market is achieved in
practice, in line with the objective of the Directive.
However, there is still a significant number of unsafe products in the EU market, which
hint to the fact that the deterrent effect of the Directive might not be effective enough. A
plausible explanation for this, suggested by several stakeholders, might be that the
maximum amounts of sanctions and penalties for product safety infringements, that are
not harmonised across Member States, remain significantly low. Sometimes companies
organise product recalls and other measures in jurisdictions first where the level of
penalties is higher than in some EU Member States. This damages the effectiveness of
the Directive and the EU product safety framework as a whole.
5.2.2. Extent to which the Directive has been effective in contributing to the
functioning of the Single Market
According to its recital (2), the Directive pursues the aim of improving the functioning of
the internal market. As recital (3) confirms, it has introduced a common legislative
framework in order to avoid disparities between Member States that could have emerged
in the absence of Union law.
The Directive plays an essential role in the functioning of the Single Market, in line with
the legal basis of the legislation, so that producers ensure safety and market surveillance
authorities can take actions against products, risks and aspects not covered by sectorial
legislation. There is no indication that Member States have tried to stop the income of
products for from other EU Countries for which no harmonisation legislation exists and
to which the Directive therefore applies fully (non-harmonised products) for other
reasons than their insufficient level of safety.
However, stakeholders emphasise that market surveillance authorities of different
Member States may come to different conclusions in relation to the risks posed and
safety of a particular product, and that this in some cases affect their operations and
increases administrative burdens, thus having a negative impact on the functioning of the
Single Market and the level-playing field for economic operators. Disputes between
Member States on risk assessments are discussed within the Safety Gate/RAPEX
network. Over recent years, the number of such disputes to better align the risk
assessments by different Member States' authorities has been relatively stable, as
indicated in table 3. The number of notifications that were subject to disputes has been on
average less than 30 per year212
.
212
The number of actual disputes was slightly higher, as in some cases more than one Member State
provided a different risk assessment in a follow-up notification (or "reaction" as it was named previously)
compared to the risk assessment by the Member State that submitted the original notification.
61
Table 3: Number of disputes on risk assessments that needed to be discussed within the RAPEX network
Year
Number of notifications that
were subject to disputes
Number of follow up disputes
2013 19 21
2014 39 41
2015 33 39
2016 19 24
2017 24 28
2018 26 27
2019 30 30
Total 190 210
Source: Safety Gate/RAPEX.
In this regard, it is important to note that over the years, the European Commission has
issued a number of guidance documents that support the uniform application of the
Directive in the Member States, including the Commission Implementing Decision (EU)
2019/417 of 8 November 2018 laying down guidelines for the management of the
European Union Rapid Information System ‘RAPEX’ according to Article 12 of
Directive 2001/95/EC on general product safety. Moreover, there is evidence that the
training programmes that the European Commission organised for the national market
surveillance authorities and the EU financing of coordinated market surveillance
activities of EU Member States have contributed to a more uniform application of the
Directive in the Member States.
Nevertheless, the lack of a mechanism at EU level to solve divergent positions of
Member States regarding the risk assessment of a specific product remains a challenge
for the effectiveness of Directive.
During the GPSD stakeholders were consulted on their views about to what extent the
Directive has been effective in reaching its objective of contributing to the functioning of
the Single Market. Their overall opinion was positive, while SME provided a slightly
more negative opinion in this regard than other businesses.
As conclusion, on a general level the Directive has been effective in contributing to the
free movement of goods within the Single Market. The lack of a mechanism to arbitrate
disputes on risk assessment contributes negatively to the effectiveness of the Directive.
5.2.3 Effectiveness of the system of market surveillance under the Directive
The Directive establishes the obligation for Member States to carry out market
surveillance activities to enforce is provisions. This evaluation detected three key factors
that have negatively influenced the effectiveness of the market surveillance provisions:
traceability of unsafe products as a precondition for market surveillance; the lack of
resources of authorities; and the coexistence of two different market surveillance systems
for harmonised and non-harmonised products. The latter point is developed in the
Coherence section; the other two here below.
62
The extent to which actions by market surveillance authorities against unsafe products
are effective depends on how easily, quickly and precisely products can be identified and
traced back to specific producers, importers, distributors, sellers and consumers. An
adequate system of product traceability allows market surveillance authorities to
determine if an unsafe product is on their market, to trace the economic operators who
made the product available, and to enforce the appropriate corrective actions. From the
perspective of the economic operator, traceability is fundamental for effectively and
efficiently managing product risks; increased traceability enables more targeted and less
costly corrective actions, e.g. by limiting the size of withdrawals or recalls. Finally,
traceability is also important for consumers because if an unsafe product is already
purchased, clear product identification is necessary for consumers to respond to a recall.
It appears that at present, the Directive’s provisions on traceability are not sufficiently
explicit to guarantee that complete information on supply chains and distribution of the
product is gathered. The Directive does not contain detailed traceability requirements.
Article 5(1) contains a general obligation for producers to provide the necessary
information for tracing a product, without asking for specific or minimum identification
information. According to article 5(1), this information may for example include “an
indication, by means of the product or its packaging, of the identity and details of the
producer and the product reference or, where applicable, the batch of products to which it
belongs, except where not to give such indication is justified”. Apart from producers,
distributors are also required to keep and provide documentation necessary for tracing the
origin of the products (article 5(2)). Furthermore, the emergence of online sales poses
additional challenges to trace an unsafe product back to where and by whom it was
produced and sold from.
The extent to which these requirements based on the Directive achieve adequate product
traceability can be demonstrated through the data available in the Safety Gate/RAPEX.
From 2013 to 2019 a significant share of the alerts that were submitted for unsafe
consumer products involved products with unknown product information items. In 2019
for instance, 36% of alerts for unsafe products did not include information about the
manufacturer; 20% concerned products of unknown brand or batch number/barcode; and
a share of 12% regarded products with no type or model information. Figure 11 based on
alerts registered in the EU Safety Gate shows that only an improvement over time on the
availability of information on the manufacturer and batch number/barcode of the product
(i.e. a decrease of the number of alerts that did not provide such information). There is no
clear trend of improvement over time for the rest of the traceability information.
63
Figure 11: Share of Safety Gate/RAPEX alerts with unknown product information
items (2013-2019)
Source: GPSD Study, from Safety Gate/RAPEX data
The same data also reveal that missing product information is more typical for specific
types of products such as laser pointers, lighters, jewellery, decorative articles, etc. What
these products have in common is that they all fall within the scope of Directive and are
not subject to sector-specific harmonisation rules. It follows that product categories under
the Directive are more likely to lack relevant information items that are essential to trace
them.
Secondly, the market surveillance system under the Directive (consisting of market
surveillance activities by authorities in the Member States, information exchange through
Safety Gate/RAPEX and coordination and support measures) appears to be operating
under considerable resource constraints. In a 2018 evaluation of the product safety-
related actions funded under the EU Consumer Programmes213
, authorities indicated
limited staff/financial resources for market surveillance and enforcement most frequently
as a factor influencing negatively the level of their achievement. A previous study
concluded that the total budget available to authorities in 18 EU Member States for
which data was available declined annually between 2010 and 2013 in nominal terms,
and the total staff resources available to authorities (in full time equivalent units) also
showed a negative trend214
. In a recent survey, both authorities and other stakeholders
213
See Civic Consulting (2018), Ex-post evaluation of the Consumer Programme 2007-2013 and
mid-term evaluation of the Consumer Programme 2014-2020, Part 1 – Mid-term evaluation of the
Consumer Programme 2014-2020 and European Commission,
214
European Commission, Ex-post evaluation of the application of the market surveillance
provisions of Regulation (EC) No 765/2008, Final Report, May 2017, p 35-39.
0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
2013 2014 2015 2016 2017 2018 2019
Pecent
of
all
alerts
No batch number/barcode No brand
No type/model No manufacturer
None of the four
64
agreed that two of the three top problems affecting the functioning of market surveillance
relate to a lack of resources: limited staff resources of market surveillance authorities in
general, and in addition, a specific lack of financial resources for product testing215
. It is
widely acknowledged by authorities and other stakeholders that the staff and financial
resources of market surveillance authorities are often insufficient, with the fragmentation
of responsibilities at national level leading to inefficiencies due to a lack of economies of
scale in some cases, and contradictory measures and approaches for risk assessment
between authorities in others. Also, the number of inspections can be considered as
generally low, with a median of roughly 400 inspections of consumer products per year
and million population, based on data from those Member States that provided such
information.
Over the last 15 years, the European Commission has co-financed more than 50
coordinated market surveillance activities (the so-called Joint Actions or, since 2018,
Coordinated Activities on the Safety of Products or CASP) carried out by Member State
authorities, with a total budget around 27 million EUR since the start of the activities.
These coordinated activities aim at promoting and coordinating administrative
cooperation for the application of Directive and ultimately at ensuring a consistent
approach towards the effective enforcement of product safety legislation across the
internal market. Most coordinated actions have resulted in the identification of a
significant number of unsafe products, with non-compliance rates around 20%216
, leading
to consequent notifications in the Safety Gate/RAPEX for 14 categories of products. The
implementation of these actions have been considered by authorities extremely useful, as
economies of scale allow and the funding provided by the Commission have helped
them to carry out inspections for some categories of products that would have been less
controlled otherwise.
Finally, it was mentioned by several authorities that some Member States complement
product-based market surveillance with a market surveillance of internal processes set up
by economic operators to ensure product safety. In those cases, authorities reported an
increase of the effectiveness of product safety enforcement.
In conclusion, while the Directive sets an efficient system to ensure the safety of
consumer products and contribute to better and more coordinated market surveillance,
challenges related to traceability, institutional fragmentation (including regarding market
surveillance of harmonised and non-harmonised consumer products) as well as resource
constraints limit the effectiveness of the overall system. At the same time, coordinated
surveillance activities have proven to be very efficient, contributing to a consistent
enforcement of product safety legislation across the internal market.
215
GPSD implementation study, p.90.
216
Joint Action reports repeatedly indicate that these high rates of non-compliance were not necessarily
representative for the market, as non-random samples were taken and often samples were tested where a
visual inspection had suggested possible deficiencies.
65
5.2.4 Effectiveness of the Safety Gate/RAPEX
The Safety Gate/RAPEX ensures that information about unsafe products withdrawn from
the market and/or recalled from consumers anywhere in Europe is circulated between
Member States and the European Commission, so that appropriate action can be taken by
market surveillance authorities in all EU Member States (and the EEA countries Iceland,
Liechtenstein and Norway). During the period 2005 to 2019 a total of 25 560
notifications were submitted (close to 5 notifications on average per day), with all
Member States participating in the system on a regular basis. The indispensable character
of the system is emphasised by the diversity of institutional approaches for market
surveillance and the high degree of fragmentation of market surveillance authorities
according to sectoral and/or administrative considerations in many in Member States,
which requires quick and effective information exchange and distribution, for which the
system is a key channel.
The effectiveness of the Safety Gate/RAPEX is also illustrated by the analysis of the
number of follow-up measures. Follow-ups can be defined as the feedback received from
Member States participating in the Rapid Alert System on actions they have taken
following up another country’s alerts. As Figure 12 shows, the number of follow-up
measures have steadily increased since data started to be gathered in this respect by the
European Commission in 2011.
Figure 12: Number to follow-up measures taken by Member States reported to Safety
Gate/RAPEX for the period 2011-2019
Source: Safety Gate/RAPEX
The GPSD Study also showed that authorities to a large extent appreciated the
functioning of RAPEX, with 65% considering the system to function at least 'moderately
well' (48% considered it to be 'rather' or 'very well' functioning). Other stakeholders were
even more positive, with 70% finding the system at least 'moderately well' functioning
(46% considered it to be 'rather' or 'very well' functioning). Only a small minority
provided a negative assessment ('rather not' or 'not at all' functioning).
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
2011 2012 2013 2014 2015 2016 2017 2018 2019
Follow-up
66
Despite this general satisfaction with the rapid alert system, there are also issues that
prevent its optimal operation. In the surveys of authorities and general stakeholders for
the GPSD Study, respondents were asked whether they had encountered impediments
when using the system. Moreover, the recent GPSD implementation study analysed the
average duration between the detection of an unsafe product and its notification to Safety
Gate/RAPEX in each Member State. The RAPEX system sets out guidelines setting
indicative timeframes for notifications, but these are not compulsory for Member States.
In most cases this duration is two weeks or more; in some cases, the duration might be up
to 6 months.
Several authorities emphasised that the duration between detection of an unsafe product
and its notification to the system depended on the type of product, the risk, the required
testing and the behaviour of the economic operator (objections by the relevant economic
operator are in some cases reported to lead to significant delays).
Regarding international cooperation, there is high interest both from third countries and
from the EU in exchanging information on unsafe products between the Safety Gate/
RAPEX, and third countries. This exchange enables more efficient market surveillance
in the EU and enhancing product safety worldwide which can also in turn result in a
better protection of EU consumers. When it comes to the procedure to set up
arrangements to exchange non-public information from the RAPEX system, the Directive
does not provide absolute clarity. Indeed, article 12(4) only mentions “access to
RAPEX”, which must be based on an international agreement. The Directive does not
distinguish between the different levels of exchanges between the EU system and third
countries. In practice, there is a major difference between:
 a third country becoming a full RAPEX member.
 a third country with which selected non-public RAPEX information is shared.
The current provisions of the Directive clearly cover the first type of cooperation, which
consists in full access to Safety Gate/RAPEX and must be based on an international
agreement. However, there is no clarity on the procedure for the second type of
cooperation.
It can be concluded that the rapid alert system achieves its objective to provide a platform
for exchange of information concerning unsafe products. Since its establishment in 2003
in accordance with Article 12 of the Directive, it has become a cornerstone of the EU
market surveillance and product safety framework. Market surveillance authorities and
other stakeholders therefore consider the Safety Gate/RAPEX mostly to be well
functioning and effective. Still, certain issues currently prevent its optimal operation,
such as delays between the detection of an unsafe product in a Member State and its
notification to Safety Gate/RAPEX, as well as the procedure to set up arrangements for
the exchange of information with third countries.
5.2.5 Effectiveness of the standardisation procedure of the Directive
The standardisation system established under the Directive has been effective in the
sense that standardisation requests have been elaborated under the Directive and these led
in most cases to standards which help producers to comply with the general products
67
safety requirement under the Directive by providing presumption of safety of the product.
The list of referenced standards shows the importance of the Directive in this respect, as
most of the listed products have a high potential for consumer harm. The GPSD Study
showed that referenced standards are widely used by manufacturers, as a conforming
product shall be presumed safe. Even a brief Internet research on a major e-commerce
website indicates that products falling under the Directive (such as gymnastic equipment,
bicycles, carry cots and stands etc. ) are regularly advertised as conforming to the
relevant standards, thereby providing evidence of their application in practice. Market
surveillance authorities also make an extensive use of standards to assess the safety of
products and take measures accordingly.
Replies from stakeholders to the GPSD Study have shown that the effectiveness of the
standardisation process might be hampered by several procedural issues. In particular,
stakeholders have stated that the process is considered to be long and complicated, and
sometimes it reduces the possibilities for participation of authorities, SMEs, consumer
organisations and other organisations, as well as universities. However, it should be
mentioned that the standardisation process must strike a balance between speed and the
quality of the outcome, thus, of the standard.
As detailed in Annex 10, the process under the Directive includes one step more than the
procedure applied in relation to harmonised standards. The reason is that the
harmonisation directives contain essential safety requirements on which standards can be
based. In contrast, the wide coverage of the Directive requires specification of the safety
requirements for a specific product, which then serves as a guideline for the work of the
European standardisation bodies217
. The addition of Step 1 of the process under the
Directive therefore seems to be justified and unavoidable. As the European
standardisation organisations attempt to achieve consensus, the duration for elaborating
the standard by the ESO (Step 3) also seems to be justified by the nature of the process;
while this takes time, the consensus-principle has always been regarded to be an essential
element of standard setting procedures. Step 4, the Commission decision to reference
standards, has been recently adapted following developments in case law. However, a
possible merger of steps 1 and 2 would offer a potential area of improvement (in
particular to enhance its efficiency, see section below).
5.2.6 Effectiveness of provisions on product recalls
Product recalls are one of the most common measures to mitigate the risks posed by
unsafe products in the EU218
. Among the over-2000 measures reported each year through
the EU Rapid alert system219
about half concern recalling the products from consumers.
The Directive does not contain any specific provisions for recall procedures and
217
See the GPSD implementation study, at 7.2.1.
218
The term “recall” refers here to the process aimed at achieving the return of a dangerous product that
has already been supplied to consumers, initiated directly by the producer or distributor of the
dangerous product, or ordered by authorities.
219
https://ec.europa.eu/consumers/consumers_safety/safety_products/rapex/alerts/repository/content/pages/r
apex/index_en.htm
68
timelines, communication or the remedies to be offered to consumers. Producers
undertake voluntary action to organise recalls but authorities can also order a recall on
the basis of notifications of unsafe products from other countries or the results of their
own market surveillance activities or if producers’ actions are deemed insufficient. As
shown by the Implementation Study, each Member State follows its own approach, with
some common elements, but also diverging practices.
While the effectiveness of product recalls varies considerably depending on factors such
as channel of sale220
and product category221
, the proportion of products successfully
recovered from consumers remains generally low, as recognised by a recent OECD
report222
. For instance, one Member State indicated that the return rate (understood as the
number of products returned by consumers following the release of a product recall
notice and other related communications) rarely exceeds 10%, except when products
have been purchased online, which makes it easier for suppliers to contact and alert their
customers about a recall223
. Another national authority estimated that around 80% of
products of low value with a short lifespan remain in consumers’ hands224
.
The consequences of delayed and ineffective recalls are also exemplified by the deaths
and injuries caused by products such as faulty airbags (estimated to have caused 35
deaths and 300 injuries worldwide225
) and baby sleepers (associated with 59 baby deaths
in the US226
); see Annex 8.
In the interviews with authorities held in the frame of the GPSD Study, few of them were
able to estimate recall effectiveness in terms of the percentage of the recalled consumer
products that were actually collected. Several authorities also suggested that even though
they collect related data, it is difficult to determine the effectiveness of product recalls.
At a recent EU workshop, it was concluded that there is no systematic approach by the
authorities to monitoring recall effectiveness227
.
The main obstacle to recall effectiveness is the difficulty of identifying and contacting
the owners of recalled products, which means that many EU consumers are simply not
aware that they own a recalled product. The Directive does not contain any requirements
220
Recalls tend to be more effective if the product was bought online because it’s easier to identify and
directly contact the buyers.
221
Recall effectiveness increases with product price and expected lifespan and decreases with product age.
222
OECD (2018), Enhancing Product Recall Effectiveness Globally, p. 5.223
Idem, p. 17.
223
Idem, p. 17.
224
European Commission, 2021, Behavioural study on strategies to increase the effectiveness of product
recalls.
225
https://www.consumerreports.org/car-recalls-defects/takata-airbag-recall-everything-you-need-to-know/
226
https://www.washingtonpost.com/gdpr-
consent/?next_url=https%3a%2f%2fwww.washingtonpost.com%2fbusiness%2f2019%2f10%2f17%2fstud
y-concludes-design-rock-n-play-other-infant-sleepers-led-deaths%2f
227
EU workshop on strategies to maximize the effectiveness of product recalls, 23 October 2019,
p.11. Annex 14
69
on consumer traceability and authorities interviewed for the Implementation Study
mentioned that reaching the final consumer is a challenging task.
For most recalled products, customer data is not available or even if it is available, it is
not used to inform affected consumers. Apart from motor vehicles (whose registration
with public authorities is mandatory), registration schemes are only available for few
higher-value product categories like domestic electric appliances and communication
devices, and even in these sectors actual registration rates tend to be rather low228
.. In
addition, economic operators are hesitant about using customers’ information collected
for other purposes (e.g. in the loyalty programmes or online sales) in the event of a recall
because of a possible legal uncertainty about the compliance with the General Data
Protection Regulation229
. Limitations on customer data were also notified by online
platforms in a Study on contractual relations between online platforms and their
professional users230
. Moreover, there are no systematic sources of indirect recall
information for consumers. For instance, only 12 EU/EEA countries publish recall
information on their websites231
.
A second obstacle to recall effectiveness is the consumers’ reluctance to return a
recalled product even if they are aware about the recall. An EU-wide survey
commissioned by the European Commission showed that over a third of consumers
(35%) did not react to a recall that was relevant to them: 31% continued using the
product with extra caution, while 4% took no action whatsoever232
. The corresponding
figures in a most recent consumer survey in 10 EU countries were 24% and 13%,
respectively233.
This may be caused by recall notices being unclear and/or minimising consumers’
perception of risk. For instance, the analysis of existing recall announcements showed
that over half of them used terms and expressions, which could undermine consumers’
perception of risk, such as ‘voluntary/precautionary recall’, ‘potential concern/problem’,
in rare cases’/in specific conditions’ or highlighting that there have been no reported
injuries234
. In addition, the procedure for consumers to return the recalled product may be
complex and burdensome.
228
European Commission, 2021, Behavioural study on strategies to increase the effectiveness of product
recalls.
229
European Commission, Notes from EU Workshop on strategies to maximise the effectiveness of
product recalls, 23rd
October 2019, p. 2.
230
Study on contractual relationships between online platforms and their professional users, April 2018.
231
European Commission, 2021, Behavioural study on strategies to improve the effectiveness of
product recalls.
232
European Commission (2019). Survey on consumer behavior and product recalls effectiveness, p.
20, available at:
https://ec.europa.eu/consumers/consumers_safety/safety_products/rapex/alerts/repository/tips/Product.Reca
ll.pdf.
233
European Commission, 2021, Behavioural study on strategies to improve the effectiveness of
product recalls.
234
Idem.
70
In addition to the issues mentioned above, behavioural research235
suggests that cognitive
biases and heuristics may also influence consumers to take suboptimal decisions
regarding how to respond to product recalls and may lead them not to take action. Biases
such as information overload and framing effects mean that if recall notices are lengthy
and unclear, consumers may ignore them, especially if they lack time. Over-optimism
may result in consumers underweighting the risk posed by a recalled product, while
inertia and endowment effect (i.e. the emotional bias that causes individuals to value an
owned object higher, often irrationally, than its market value) relate to the fact that
consumers have an inherent preference for status-quo, which in the case of recalls means
keeping the product236
.
To sum up, the increase in the number of product recalls over time and the fact that
recalls are currently for most part organised on a voluntary basis can be considered as
indications that the Directive has contributed to making recalls more widely used as a
corrective measure. However, EU-wide general requirements regarding the recall
procedure or communication are missing. This has been repeatedly reported as a
significant shortcoming, suggesting that existing requirements are in themselves
currently not sufficient to ensure effective recalls. The resulting limited effectiveness of
recalls may negatively affect consumer safety and the degree to which there is a level-
playing field for businesses in the internal market, affecting therefore the extent to which
the objectives of the Directive are achieved in practice.
5.2.7 Impact of e-commerce in the effectiveness of the Directive
The evaluation showed that e-commerce poses additional challenges for the control of
the safety of products sold online, thus questioning the effectiveness of the Directive in
this regard. In particular, there is evidence pointing to the fact that the control of the
safety of products sold online is more problematic than the one for unsafe products
found in brick-and-mortar shops. Data coming from the Safety Gate/RAPEX for the
period 2018-2019 show that the share of notifications of products 'sold online' in which
one of the four traceability information items was missing (see table 4) was between
29,2% and 57,3% (depending on the item); the share for notifications not indicating ‘sold
online’ in which one of the four traceability information items was missing was
considerably lower (between 12,6% and 35%7).
235
Bernstein A. (2013), ‘Voluntary Recalls’, University of Chicago Legal Forum, 1: 394 ff., available at:
http://chicagounbound.uchicago.edu/uclf/vol2013/iss1/10 and Jacoby J. (1984), ‘Perspectives on
Information Overload’, Journal of Consumer Research, 10(4), pp. 432-435.
236
See OECD (2018), Enhancing Product Recall Effectiveness Globally, p. 5.
71
Table 4: Number and share of Safety Gate/RAPEX notifications concerning unsafe
consumer products with unknown product information items (by sales channel, 2018-
2019)
Total
number of
notifications
Number of
notifications
regarding
products not
indicating
'online'
Share :
Notifications
not online
missing
information /
Total
notifications not
online
Number of
notifications
regarding
products
indicating
'online'
Share :
Notifications
online
missing
information /
Total
notifications
online
Total number of
notifications for
consumer products
2018-19
3 864 3590 - 274 -
Notifications in which information item is missing
- No manufacturer 1437 1 280 35,7% 157 57,3%
- No brand 800 700 19,5% 100 36,5%
- No type/model 531 451 12,6% 80 29,2%
- No batch
number/barcode
805 667 18,6% 138 50,3%
- None of the four 52 17 0,5% 35 12,8%
Source: Prepared with Safety Gate/RAPEX data.
Furthermore, recent investigations from market surveillance authorities, consumer
organisations and industry associations show worrying levels of unsafe products in some
categories of products sold online:
 A campaign carried out in 2018 by DGCCRF237
, the French market surveillance
authority, showed that on average 42.8% of the inspected products online were
unsafe. The level of dangerous products reportedly varied a lot depending on the
product category: high rates of unsafe products were found for example in low
priced jewellery (74%) and some electrical products (66%), while for cosmetics
the rate was 16%. The situation also varied greatly considering the marketplace
on which the samples were taken (ranging for example from 22% to 58% of
unsafe products). The authority concluded that it found a significantly higher
share of unsafe products on online marketplaces compared to products sampled
across all distribution channels.
 In February 2020, six consumer groups from the BEUC network238
tested 250
electrical goods, toys, cosmetics and other products bought from a number of
online marketplaces. They selected the products based on possible risks and
found that 66% of them failed EU safety laws with possible consequences such as
electric shock, fire or suffocation.
237
Direction générale de la Concurrence, de la Consommation et de la Répression des fraudes
238
https://www.beuc.eu/publications/two-thirds-250-products-bought-online-marketplaces-fail-safety-tests-
consumer-groups/html
72
 Also in 2020, Toy Industries of Europe239
undertook a project to assess the safety
and the legal compliance of toys bought from operators who sell on several online
marketplaces present in the EU. The assessment was conducted on toys bought in
seven EU markets - Denmark, France, Germany, Italy, the Netherlands, Spain and
Sweden. The project showed that more than 97% of the 193 toys assessed did not
comply with strict EU toy safety rules and 76% of 134 that were safety-tested had
defects that made them unsafe for children. In addition, 83% of the toys bought as
part of the project did not include necessary traceability information.
These results support the other evidence provided above. When interpreting the research
presented in the previous paragraphs, it is important to recall that all quoted studies are
based on risk-based sampling, i.e. they focused on products with a higher probability for
non-compliances. While this is a standard approach used by market surveillance
authorities, it means that results are not necessarily representative of the market overall,
but provide insights into specific problem areas.
As already stated in the section on Relevance, the Commission has taken some actions to
tackle this problematic. One major action has been the facilitation of the signature of the
Product Safety Pledge by 9 online marketplaces. Most market surveillance authorities
and the online marketplaces part of the Pledge see the signature of the voluntary
commitments as positive, as it sets a cooperation framework and defines possible actions
to be carried out by online platforms as an important player in the supply chain. The 3rd
Monitoring Report of the Pledge240
, published in November 2020, showed some progress:
there are indications that signatories are taking actions to monitor Safety Gate/RAPEX
notifications and remove product listings when notified by authorities. However, further
than that it is challenging to analyse the effectiveness of the Pledge to appropriately
ensure the safety of products sold online. The report also showed that there has been a
divergence in the way online marketplaces calculated the Key Performance Indicators
(KPIs), making it difficult to extract conclusions from those numbers and monitor the
effectiveness of the commitments of the Pledge. Further negotiations with marketplaces
to enhance the monitoring and accountability of the Pledge have proved to be complex.
In addition, authorities, consumer organisations and other stakeholders have signalled
that the Product Safety Pledge remains voluntary, so the infringement of those
commitments cannot be penalised by authorities. There are also many players on the
market that have decided not to adhere to the voluntary commitments, creating also an
uneven level-playing field between online marketplaces targeting EU consumers.
In addition, the presence of unsafe products in online channels can have an impact
on the level-playing field for businesses. A study published by Eurocommerce241
showed estimations on how much lower the purchase price would be if the products did
not comply with EU Product Safety Rules. The cost difference between products
239
https://www.toyindustries.eu/wp-content/uploads/2020/06/Executive-Summary-Online-Marketplaces-6-
1.pdf
240
https://ec.europa.eu/info/sites/info/files/3rd_progress_report_product_safety_pledge.pdf
241
Eurocommerce – Creating a level-playing field for retail in Europe – August 2019
73
produced in accordance with EU rules and standards, and produced without taking
account of the EU rules was considered as significant. While such calculations cannot be
extrapolated to a whole category of products, they give an indication of the possible
detriment of the presence of rogue traders.
The evidence described above shows that e-commerce allows for a growing flow of
consumer products (both those falling under the Directive and those falling under
harmonised legislation) to enter the EU market including unsafe and recalled products,
with traders and products being often untraceable. While these problems also do occur in
the ‘offline’ environment, they are more relevant in the online environment which allows
products to enter the EU market without having a relevant economic operator in the EU.
It can therefore be concluded that on balance, the emergence of e-commerce has
negatively affected the effectiveness of the Directive in terms of enforcing the general
safety and traceability requirements, but also with respect to effective market
surveillance by the Member States.
5.2.8 Impact of new technologies in the effectiveness of the Directive
The development of new technologies is also putting into question the effectiveness of
the Directive as a safety net for products, aspects and risks not covered by product
harmonisation legislation.
As stated in the section on Relevance, the uncertainties regarding the applicability of
the Directive to new technologies, such as to new risks, have also an impact on its
effectiveness. A good example can be the follow up of a Safety Gate/RAPEX notification
from Iceland regarding a smartwatch for children242
. The Icelandic authority argues that
this product would not cause a direct harm to the child wearing it, but lacking a minimum
level of security, it can be easily used as a tool to have access to the child. As one of the
product’s intended function is to keep children safe through localisation, a consumer
would expect that it would not pose security threats to children that may affect their
safety by potentially being tracked and/or contacted by anyone. As measures regarding
this product were notified to Safety Gate/RAPEX, authorities in the rest of Member
States took follow-up actions.
However, as the surveys for the GPSD Study showed, many of those authorities were
unsure if the Directive applied to such risks due to the lack of explicit provisions in that
regard. The lack of explicit provisions and/or guidelines with regard to the safety of new
technologies has a negative impact on the effectiveness of the Directive. Furthermore,
almost half of the interviewed Member States authorities reported that they currently do
not carry out inspections on products incorporating new technologies, which could also
be due to such lack of clarity. The report on the safety and liability of new technologies243
242
Safety Gate/RAPEX notification from Iceland published in the EU Safety Gate’s website
(A12/0157/19)
243
Report on the safety and liability implications of Artificial Intelligence, the Internet of Things and
robotics COM/2020/64 final
74
also concluded that EU product safety legislation would need to be further clarified for
legal certainty.
The example of consumer products using new technologies illustrates both the strengths
and the weaknesses of the general safety requirement of the Directive. It confirms the
advantage of a general requirement that products need to be safe independent from the
technology used, i.e. the safety requirement being technology-neutral. However, it also
has shown its weakness in that certain key definitions, such as “safe product” and
“product”, which are broad and unspecific to apply in a wide range of situations, can be
ambiguous in the context of new technologies, and therefore create practical difficulties
for the application of the Directive, which reduce its effectiveness.
5.2.9 Effectiveness of the Food-Imitating Products directive
The number of Safety Gate/RAPEX notifications of food imitating products is small.
Between 2013 and 2019, a total of 114 notifications (less than 20 per year on average)
that relate to food imitating products, as table 5 shows. Moreover, it seems that the
product category “Food-imitating products” was only used up to 2015; afterwards, the
products have been categorised according to their use (cosmetics, clothing, etc.), and
aspects related to the imitating nature of the product were incorporated in the risk
assessment of the product itself (but not in a systematic manner by all Member States).
Table 5: Notifications to Safety Gate/RAPEX related to food-imitating products
Product category Year Total
2013 2014 2015 2016 2017 2018 2019
Cosmetics 3 1 28 1 33
Decorative articles 1 1 4 17 23
Food-imitating products 26 12 8 46
Other 2 2
Stationery 2 2
Toys 1 4 3 8
Total 27 13 8 4 1 36 25 114
Source: Safety Gate/RAPEX
To address this issue, in 2019 the Commission circulated a questionnaire to Member
States on the implementation of the FIPD. Its outcome showed that there are very
different interpretations of the FIPD (and notably whether food-imitating products should
be banned per se or be subject to a risk assessment). These divergences rendered the
FIPD ineffective in practice, as there is not a harmonised approach for this category of
products in the EU.
During a meeting of the Consumer Safety Network organised in November 2019,
Member States agreed that the development of guidelines would not enhance the
effectiveness of the FIPD and legal changes might be needed to address the diverging
interpretations of this directive.
75
5.2.10 Conclusion on effectiveness
The Directive appears overall to have met its objectives of ensuring a high level of safety
of consumers, while ensuring an effectively operating internal market for goods. Overall,
the Directive has reached a reasonable uniformity in its implementation and the set
of its rules makes it effective.
However, certain provisions of the Directive have proven to affect negatively its
effectiveness, resulting in a continuing influx of new unsafe products on the market. This
relates, as highlighted above, to the lack of a mechanism to arbitrate disputes on risk
assessment, challenges on the traceability of unsafe products, lack of resources of market
surveillance authorities, as well as the impact of e-commerce and new technologies on
the effectiveness of the Directive. Furthermore, more clarity could be sought with regards
to product recalls and other key definitions. The current divergences of interpretation of
the FIPD among Member States has also resulted to be inefficient, so actions in that
respect should be pursued.
5.3 Efficiency
5.3.1. Analysis of costs
The current costs of compliance with the Directive are directly accruing to businesses
and market surveillance authorities, and indirectly to consumers in the form of costs of
consumer goods and taxes.
5.3.1.1 Costs for businesses
The Directive applies fully to consumer products for which no specific EU harmonised
legislation exists. It does not apply to industrial/professional products. While the
Directive is also applicable to harmonised consumer products to the extent that there are
no specific provisions with the same safety objective in the EU harmonised legislation,
the significance of this ‘residual effect’ of the Directive depends on several factors, most
notably on the extent to which EU harmonised legislation reflects the same level of
protection. As the residual effects on manufacturing and distribution of harmonised
products are in any case expected to be minor compared to the effects in the area of non-
harmonised products, this assessment focuses on the latter. In other words, the following
assessment only considers the costs of compliance with the Directive for manufacturers
and distributors of non-harmonised consumer products.
Following this approach, the estimation of business costs is based on the following steps
and detailed in Appendix 2:
 Step 1: Estimation of EU companies’ total annual turnover from the production
and/or sales of non-harmonised consumer products in the EU - Based on this
approach, the total EU turnover from non-harmonised products in the selected
sectors amounts to EUR 773 billion for EU manufacturers, EUR 750 billion for
EU wholesalers and EUR 581 billion for EU retailers.
 Step 2: Deduction of extra-EU export – This results in an estimation of an annual
EU turnover related to non-harmonised products of EUR 655 billion for EU
76
manufacturers, EUR 707 billion EUR for EU wholesalers and approx. EUR 485
billion EUR for EU retailers.
 Step 3: Deduction of industrial and professional products - As a result, the total
annual EU turnover of EU companies from non-harmonised consumer products is
estimated at EUR 1032 billion EUR.
 Step 4: Derivation of empirical estimates for companies’ product safety-related
costs – These estimates take into account activities such as managing product
safety, testing for product safety, and product recalls.
 Step 5: Extrapolation of EU companies’ annual costs related to the Directive
including business-as-usual costs that occur also in absence of regulation.
 Step 6: Deduction of business-as-usual costs and extrapolation of EU companies’
annual compliance cost related to the Directive
As indicated in table 6, the estimated costs for EU businesses to comply with the
Directive amount to EUR 1.1 billion per year, of which EUR 343 million accrue to EU
manufacturers, EUR 321 million to EU wholesalers and EUR 439 million to EU retailers.
Table 6: Estimated annual cost for businesses to comply with the GPSD, by company size
class, in million EUR
Cost by company size Total costs
Company size (employees) From 0 to 49 50 – 249 250 or more All size categories
Total of manufacturing 79 101 163 343
Total of wholesale 118 81 122 321
Total of retail 232 44 163 439
Total 428 226 448 1 102
Source: GPSD Study, based on company costs survey and Eurostat data.
With regards to SMEs, companies with less than 50 employees are estimated to have
GPSD-related costs (after business-as-usual costs such as costs related to general due
diligence activities have been subtracted) of approx. 428 million EUR per year, and
companies with 50 to 249 employees are estimated have GPSD-related costs of approx.
226 million EUR per year (see Table 6 above). Accordingly, SMEs account for 59% of
the total of GPSD-related compliance costs in the EU, in line with their overall share in
the market. It should be noted that due to the relatively high number of EU SMEs that
engage in wholesale and (particularly in) retail sectors compared to manufacturing
sectors (and compared to large EU companies which are more engaged in manufacturing
activities), GPSD-related measures that impact on the distribution chains of non-
harmonised consumer products can be expected to have a higher aggregate impact on EU
SMEs than measures that impact on manufacturers.
5.3.1.2 Costs for authorities
Assessing the costs for market surveillance authorities to comply with the Directive is
complicated due to organisational differences across Member States. Market surveillance
systems for consumer products differ in the extent to which market surveillance is
conducted by authorities with broader or with narrower sectoral responsibility, as well as
77
to the degree of centralisation of the administration of each country. These organisational
features affect how market surveillance of non-harmonised consumer products is
organised, and in some cases the share of staff working on market surveillance of non-
harmonised products is unknown.
To overcome this limitation, the estimate of market surveillance costs has been based on
comprehensive staff data for 20 EU Member States obtained through interviews and
surveys. Consequently, the estimation for total EU27 staff-related costs for market
surveillance of non-harmonised consumer product amount to approximately EUR 122
million per year. Of this amount, EUR 14 million accrue in countries where
responsibility for market surveillance is centralised and EUR 108 million in countries
where responsibility for market surveillance is (partly) delegated to or competence of
sub-national administrations (for more info see Appendix 2). Non-staff related costs of
market surveillance activities for non-harmonised consumer products in the EU can be
considered negligible (EUR 0).
5.3.2. Analysis of benefits
The main benefit of the Directive is the consumer protection and its consequential
reduction of consumer detriment. The GPSD Study provided an analysis of detriment
due to product-related injuries and fatalities in the EU (see Annex 4). The analysis
concluded that the preventable detriment suffered by EU consumers and society due to
product-related accidents (and related injuries and pre-mature deaths) can be estimated at
EUR 11.5 billion per year. In addition to this injury related detriment, the consumers
also suffer financial costs of a total value of EUR 19.3 billion for 2019 arising from the
fact they have purchased unsafe products that they would not have purchased if they
knew these products are unsafe While it is not possible to estimate the detriment suffered
by EU consumers and society that is avoided by EU product safety legislation, including
the Directive, it is reasonable to assume that in absence of the general safety requirement,
the standards referenced and other provisions under the Directive, the detriment suffered
due to product-related accidents and financial costs for consumers would be substantially
higher, thereby outweighing the related costs for companies, market surveillance
authorities and consumers.
Furthermore, the Directive has contributed with additional benefits:
 Increased consumer trust: as reflected in Figure 10, there has been an increase of
consumer trust from 65% in 2008 to 78% in 2016;
 Increased business revenue (e.g. due to increased reputation /brand value);
 Improved quality / lifecycle of products;
 Better information on unsafe products/measures taken by authorities provided through
the Safety Gate/RAPEX (more than 25,000 notifications since the establishment of the
system);
 Better supply chain management due to traceability of products;
 Greater legal certainty;
 Lower operational risk for businesses;
 Deterrent effect on rogue traders;
78
 More level-playing field among businesses;
 Better functioning of the EU Single market;
 Higher level of protection of the environment due to reduction of unsafe products that
also have environmental impacts (e.g. lead in PVC, siloxanes, Nonylphenol);
 Better access to the market in non-EU/EEA countries.
2.5.1. 5.3.3. Balance and distribution between costs and benefits
About nine in ten respondents that had an opinion considered the costs due to product
safety requirements of the GPSD to be at least “moderately proportionate” to the
resulting benefits. Close to six in ten respondents that had an opinion even found these
costs to be “largely proportionate” or “very proportionate”.
This largely positive assessment is consistent with the analysis of compliance costs (see
table 7). A large part of costs related EU product safety legislation for consumer products
are considered as business-as-usual costs (BAU), i.e. costs that companies would incur
anyway (i.e. even in absence of product safety legislation, for example because these
costs relate to their due diligence procedures). Compliance costs due to the safety
requirements of the Directive that exclude business-as-usual costs are therefore limited,
compared to the benefits the Directive brings, including in terms of contributing to the
functioning of the internal market.
79
Table 7: Assessment of costs and benefits
Type Assessment of cost/benefits
Costs
Companies’
compliance costs
Consumer product safety-related compliance costs are estimated at 0.59% of turnover for
manufacturing sectors and 0.14% for retail and wholesale services sectors. Subtracting
costs that companies would incur anyway (i.e. even in absence of product safety
legislation, e.g. because these costs relate to due diligence), the estimated costs for
businesses to comply with the GPSD amount to EUR 1.1 billion per year, of which EUR
343 million accrue to EU manufacturers, EUR 321 million to EU wholesalers and EUR
439 million to EU retailers
Member States’
costs for market
surveillance costs
Total EU27 staff-related costs for market surveillance of non-harmonised consumer
product amount to approximately EUR 122 million per year. Of this amount, EUR 14
million accrue in (smaller) countries where responsibility for market surveillance is
centralised and EUR 108 million in (often larger) countries where responsibility for
market surveillance is (partly) delegated to or competence of sub-national administrations
Benefitsa)
Better information
on unsafe products
through Safety Gate
/RAPEX
In the period 2005 to 2019, a total 25 850 publicly available notifications were
transmitted through Safety Gate/RAPEX, including 25 051 notifications concerning
products with serious risks. In a 12 months period 2019/20, the analysed notifications
affected some 41.8 million items in total
A better functioning
internal market
The aim of free movement of (non-harmonised) goods within the internal market has
been achieved. There were only few cases where Member States prohibited or hindered
the import of products from other Member States that had been certified in line with EU
product safety law, and these cases all related to specific harmonised legislation but not to
the GPSD. There is no indication that Member States try to stop imports from other
Member States for reasons of their insufficient level of safety. Standardisation has
contributed to the uniform application of product safety law in the Member States. So far,
a total of 80 standards were referenced under the GPSD
Increased consumer
trust
Consumer trust in product safety in the EU has shown a slight increase over time, with
the proportion of consumers agreeing that essentially all non-food products in their
country are safe (or that only a small number are unsafe) increasing from 65% in 2008 to
78% in 2016, before decreasing again to 70%. The largest increase (9 percentage points)
occurred between the 2014 and 2016 surveys, before returning in 2018 to slightly above
the 2014 level
Reduced occurrence
of products
presenting health and
safety risks &
reduced number of
accidents/injuries
caused by unsafe
products
Based on data from the European Injury Database (IDB) an estimated 11 million
product-related injuries, in which consumers visited a hospital emergency department
due to the injury, occur in the EU each year. The related detriment is estimated at EUR
76.6 billion per year. This is the sum of detriment caused by non-fatal product-related
injuries, and the cost of premature death due to fatalities caused by mechanisms relevant
for product safety (such as tools, strangulation, electric current, or fire) occurring outside
of work-related locations. The preventable detriment suffered by EU consumers and
society due to product-related accidents can be estimated at EUR 11.5 billion per year. In
addition, consumers also suffer financial costs of a total value of EUR 19.3 billion per
year arising from the fact they have purchased unsafe products that they would not
have purchased if they knew these products are unsafe. It is reasonable to assume that in
absence of the general safety requirement of the GPSD, and the standards referenced
under the Directive, detriment suffered due to product-related accidents and financial
costs would be considerably higher.
Source: GPSD Study
5.3.4. Conclusion on efficiency
This evaluation therefore concludes that the costs of the Directive are proportionate to
the benefits it brings.
However, it should be highlighted that several of the factors that affect the effectiveness
of the Directive may also influence its efficiency:
80
 Complexity of the legal framework for product safety, in particular the co-
existence of two systems of rules on market surveillance;
 Difficulties on enforcement by market surveillance authorities the provisions of
the Directive, in particular due to challenges posed by E-commerce and
traceability;
 Lack of a mechanism to solve differences in risk assessment of authorities in
different Member States;
 Outdated/unclear terms and concepts used in the Directive, in particular with
regards to new technologies;
 Delays in notification of unsafe products through Safety Gate/RAPEX;
 Challenges related to the standardisation process;
 Suboptimal effectiveness of product recalls.
Consequently, actions aimed to address those issues would contribute to further enhance
the efficiency of the Directive.
5.4 Coherence
5.4.1 Internal coherence of the Directive
This evaluation did not identify discrepancies or inconsistencies between the
provisions of the Directive. This has been backed by stakeholders interviewed for the
preparation of the GPSD Study. However, certain stakeholder have signalled that certain
provisions of the Directive lack clarity. In particular, Art 5(1) of the Directive, that sets
obligations for producers to provide necessary information for tracing the origin of a
product, is considered to be very vague and subject to multiple interpretations. This
vagueness has led to slight differences in implementation in Member States and a lack of
certainty for economic operators.
5.4.2 Coherence with product harmonisation legislation and market
surveillance
Article 1.2 of the Directive provides that the provisions of the Directive “shall apply in so
far as there are no specific provisions with the same objective in rules of Community law
governing the safety of the products concerned”. This definition of the scope of the
Directive as lex generalis ensures its coherence with product harmonisation legislation,
as there are no overlaps between the Directive and a specific instrument. This role of
“safety net” is considered by stakeholders consulted for this evaluation as an
unquestionable element and cornerstone for the protection of consumers in the EU, as
well as an essential element for the free movement of goods in the Single Market. The
interlinks of the Directive with sectorial legislation has also been acknowledged by the
Court of Justice of the EU244
.
Nevertheless, despite the full coherence of the Directive with product harmonisation
legislation, some stakeholders state that there are in practice some uncertainties when
244
Judgment of the Court (Eighth Chamber) of 30 April 2009. Lidl Magyarország Kereskedelmi bt v
Nemzeti Hírközlési Hatóság Tanácsa. Case C-132/08
81
determining the applicable legislation to a specific category of products. However, as the
GPSD Evaluation and IA shows, the reasons for these practical challenges are not
necessarily linked to the scope of the Directive, that by definition plays a residual role for
harmonised products, but rather by the sometimes unclear scope of specific pieces of EU
product harmonisation legislation or by the lack of guidelines or supporting documents to
provide clarity to businesses.
An additional aspect that can contribute to such uncertainty is the existence of
discrepancies between some of the definitions of the Directive and definitions of the
product harmonisation legislation, in particular legislation following the New
Legislative Framework (NLF) set by Decision 768/2008/EC245
(see Table X). While the
divergence of those terms is not significant, a lack of alignment has a negative impact on
the coherence of the EU product safety framework.
Table 8: Comparison of key concepts in the Directive and Decision No 768/2008/EC
Concept Directive 2001/95/EC (GPSD), Article 2 e-h Decision No 768/2008/EC, Annex
I, Article R1, points 3, 6, 14, 15
Producer "producer" shall mean:
(i) the manufacturer of the product, when he is
established in the Community, and any other person
presenting himself as the manufacturer by affixing to
the product his name, trade mark or other distinctive
mark, or the person who reconditions the product;
(ii) the manufacturer's representative, when the
manufacturer is not established in the Community or,
if there is no representative established in the
Community, the importer of the product;
(iii) other professionals in the supply chain, insofar as
their activities may affect the safety properties of a
product;
‘manufacturer’ shall mean any
natural or legal person who
manufactures a product or has a
product designed or manufactured,
and markets that product under his
name or trademark;
Distributor "distributor" shall mean any professional in the
supply chain whose activity does not affect the safety
properties of a product;
‘distributor’ shall mean any
natural or legal person in the
supply chain, other than the
manufacturer or the importer, who
makes a product available on the
market;
Recall "recall" shall mean any measure aimed at achieving
the return of a dangerous product that has already
been supplied or made available to consumers by the
producer or distributor;
‘recall’ shall mean any measure
aimed at achieving the return of a
product that has already been made
available to the end user;
Withdrawal "withdrawal" shall mean any measure aimed at
preventing the distribution, display and offer of a
product dangerous to the consumer.
‘withdrawal’ shall mean any
measure aimed at preventing a
product in the supply chain from
being made available on the
market.
Source: GPSD Study
245
Decision No 768/2008/EC of the European Parliament and of the Council of 9 July 2008 on a common
framework for the marketing of products, and repealing Council Decision 93/465/EEC (Text with EEA
relevance OJ L 218, 13.8.2008, p. 82–128
82
It also needs to be signalled that Decision 768/2008/EC is not yet aligned with
Regulation 1020/2019 on market surveillance, that has incorporated new definitions such
as “fulfilment service provider”. The Decision is also currently under evaluation.
The respondents have pointed out that the interlink between the Directive and
legislation on chemicals, notably Regulation (EC) No 1907/2006 (REACH), sometimes
lacks clarity. For instance, the term “product” is not used in REACH, as it refers instead
to “articles”, while it uses “substances” for a chemical element and its compounds. Some
respondents suggested that this can cause confusion, when the measures refer to,
respectively a “dangerous product” (Article 2(c) and (g), Article 8(1)(e) and (f), Article
10(2)(a) and (d), Article 13(3) GPSD) and a “dangerous substance” (mainly Title VIII
REACH). In particular, the question of whether the presence of a dangerous substance
would always lead to the finding of a dangerous product pursuant to the Directive was
raised. This issue has been tackled in the revision of the guidelines on the operation of
the Rapid Alert System RAPEX in 2018. The revised guidelines set out that “[t]he risk
level of a product may be considered to be serious if it contains a chemical substance
either banned or in a concentration above the limit established by European legislation.
Therefore, in cases where measures are taken against products containing a chemical
substance subject to a restriction contained in EU Legislation, a notification may be
submitted without a detailed risk assessment” 246
. By presuming the serious nature of the
risk when a product contains a substance either banned or in a concentration above the
established limit, the work of market surveillance authorities is facilitated and this
ensures the consistency of measures among Member States.
Due to the fact that chemical limits in sectorial legislation often concern specific
categories of products, it is important to correctly classify a product, which can prove
difficult for borderline cases. The consistency of chemical limits for products that present
similarities but fall under different legislations (e.g. toys and childcare articles) can thus
be raised, as well as the possibility to proceed by analogy when establishing a risk
assessment for the products not covered by such restrictions.
In relation with food contact materials, the Evaluation observed that these products
generally fall under the purview of food law247
, however, some food contact materials
may require an evaluation in line with the General Product Safety Directive. This may
have implications that are not comprehensively addressed in the legal framework e.g.
regarding which authority is responsible for a specific problem and also whether to notify
the safety issue through Safety Gate/RAPEX or the Rapid Alert System for Food and
Feed.
246
Commission Implementing Decision (EU) 2019/417 of 8 November 2018 laying down guidelines for
the management of the European Union Rapid Information System ‘RAPEX’ established under Article 12
of Directive 2001/95/EC on general product safety and its notification system.
247
Food Contact Material Framework Regulation (EC) 1935/2004 and the General Food Law Regulation
(EC) 178/2002
83
Regarding market surveillance, as already stated above, there are two different systems
established by the EU legislation: one applicable to product harmonisation legislation
as established in Regulation 2019/1020 on market surveillance and the one set by the
Directive for non-harmonised products. The existence of these two systems exacerbates
the risk of weakening the general coherence of the product safety framework. The
majority of respondents of the OPC confirmed having experienced problems with the
divergence of rules between harmonised and non-harmonised products. Furthermore,
some stakeholders stated that the implementation of two different set of rules is not even
justified; a consumer organisation provided the paradoxical example that there are
different market surveillance provisions for a baby doll bed that for an actual baby bed.
This has also been considered by market surveillance authorities as a major challenge, as
the investigation and enforcement powers granted to them are different depending on
each system. The maintenance of both systems also entails internal administrative
costs and inefficiencies for Member States.
5.4.3 Coherence with standardisation policies
The evaluation identified that there is a divergence between the standardisation
procedures set by Regulation (EU) No 1025/2012248
and the Directive, as the latter
counts with an additional step. However, this divergence is justified and corresponds to
the fact that the Directive does not contain specific safety requirements in the legal text,
so it requires an additional step for the setting up of such safety requirements.
However, the inter-links of both instruments can result in some inefficiencies. In line
with this Regulation, after the decision on safety requirements by the GPSD Committee,
the decision on the standardisation request is adopted by a different Committee, the
Standardisation Committee, which is also the responsible for standardisation procedures
in case of harmonised products. The involvement of two different Committees with
Member States representatives is considered by the different actors involved to be not
efficient and burdensome. Consequently, this has an impact on the coherence of the
standardisation process of the Directive with the general framework provided by
Regulation (EU) No 1025/2012. This is an area that could be subject to administrative
simplification.
5.4.4 Coherence with consumer protection and product liability legislation
Some stakeholders have mentioned the lack of consistency between consumer rights
under the Directive and other EU consumer protection framework with respect to
redress options. For example, the Unfair Commercial Practices Directive 2005/29/EC
248
Regulation (EU) No 1025/2012 of the European Parliament and of the Council of 25 October 2012 on
European standardisation, amending Council Directives 89/686/EEC and 93/15/EEC and Directives
94/9/EC, 94/25/EC, 95/16/EC, 97/23/EC, 98/34/EC, 2004/22/EC, 2007/23/EC, 2009/23/EC and
2009/105/EC of the European Parliament and of the Council and repealing Council Decision 87/95/EEC
and Decision No 1673/2006/EC of the European Parliament and of the Council Text with EEA relevance -
OJ L 316, 14.11.2012, p. 12–33
84
(UCPD)249
could provide remedies for the victims of unfair commercial practices in cases
involving the purchase of unsafe products, such as a right to compensation. Furthermore
the Consumer Sales Directive 1999/44/EC250
, as well as the newly adopted Sale of Goods
Directive 2019/771/EU251
, could provide remedies for victims of non-conforming
products, such as a right to a termination of contract. The General Product Safety
Directive does not award such rights to consumers, which may not only limit the
involvement of stakeholders in monitoring product safety, but also weaken the coherence
of the consumer protection framework against unsafe products. However, the recently
adopted Directive on representative actions for the protection of the collective interests of
consumers252,
includes Articles 3 and 5 of the General Product Safety Directive in the list
of consumer legislation, a breach of which will entitle the qualified national entities to
bring forward a representative action in case such breach harm or may harm the
collective interests of consumers.
Directive 85/374/EEC concerning the liability for defective products253
lays down
common rules for strict liability (i.e. "liability without fault") of producers for damage
caused by defective products at European Union level. It allows parties that have been
injured by defective products to claim financial compensation for death, personal injuries
or for damage caused to an item of property intended for private use with a threshold of
500 EUR. In order to successfully claim compensation, the injured person has to prove
the damage, the defect and the causal relationship between defect and damage. The
evaluation of the Product Liability Directive254
, published in May 2018, concluded on its
persistent correspondence to the general expectations the public can have from product
safety legislation, such as the General Product Safety Directive. The EU product safety
rules describe the safety levels that products placed on the EU market must meet. In turn,
they represent the safety levels for these products that an injured person is entitled to
249
As amended by Directive (EU) 2019/2161 of the European Parliament and of the Council of 27
November 2019 amending Council Directive 93/13/EEC and Directives 98/6/EC, 2005/29/EC and
2011/83/EU of the European Parliament and of the Council as regards the better enforcement and
modernisation of Union consumer protection rules, which added a new Article 11a to the UCPD on redress
rights for consumers. This will be applicable as from 28 May 2022.
250
Directive 1999/44/EC of the European Parliament and of the Council of 25 May 1999 on certain aspects
of the sale of consumer goods and associated guarantees, OJ L 171, 7.7.1999, p. 12–16
251
Directive (EU) 2019/771 of the European Parliament and of the Council of 20 May 2019 on certain
aspects concerning contracts for the sale of goods, amending Regulation (EU) 2017/2394 and Directive
2009/22/EC, and repealing Directive 1999/44/EC - PE/27/2019/REV/1 - OJ L 136, 22.5.2019, p. 28–50
252
Directive (EU) 2020/1828 of the European Parliament and of the Council of 25 November 2020 on
representative actions for the protection of the collective interests of consumers and repealing Directive
2009/22/EC, OJ L 409, 4.12.2020, p. 1–27
253
Council Directive 85/374/EEC of 25 July 1985 on the approximation of the laws, regulations and
administrative provisions of the Member States concerning liability for defective products, as modified by
Directive 1999/34/EC of the European Parliament and of the Council of 10 May 1999 (OJ L 210, 7.8.1985,
p.29 and JO L 141 4.6.1999, p. 20)
254
COMMISSION STAFF WORKING DOCUMENT SWD(2018) 157 final - Evaluation of Council
Directive 85/374/EEC of 25 July 1985 on the approximation of the laws, regulations and administrative
provisions of the Member States concerning liability for defective products
85
expect under the Product Liability Directive. Moreover, the General Product Safety
Directive states in its recital (36) that its provisions should not affect victims’ rights
within the meaning of the Product Liability Directive, which remains an autonomous
legal regime. However, the evaluation of the PLD also noted that as technological
changes will bring about corresponding changes in the economy, the relevance of the
PLD to these new needs will have to be ensured.255
5.4.5 Coherence with E-commerce rules
The E-commerce Directive256
establishes the general legal framework for electronic
commerce in the EU. The obligations set out apply, inter alia, to online sellers of
products and services or online advertisers, as long as they are providers of an
information society service that fall within the scope of that Directive.
Under the E-commerce Directive, Member States cannot impose a general obligation on
online intermediaries to monitor the content or a general obligation to actively seek the
facts or circumstances indicating illegal activity.
Intermediary service providers carrying out hosting activities may benefit under certain
conditions from an exemption of liability for illegal information provided by third parties
using their networks or illegal activities initiated by third parties. However, the liability
exemption is subject to specific conditions. It only applies if the intermediary service
providers have no actual knowledge or awareness of the illegal activity or information
hosted or, upon obtaining such knowledge or awareness (for instance by a ‘sufficiently
precise and adequately substantiated’ notice257
), they act expeditiously to remove or to
disable access to it. If hosting service providers do not fulfil these conditions, they are not
covered by the liability exemption and thus can be held liable for the content they host.
While the E-commerce Directive does not define the concept of illegal information or
activity, based on the General Product Safety Directive as well as product harmonisation
legislation, this concept can also cover the offer of unsafe products. In that sense, it can
be established that the E-commerce Directive is coherent with the Directive.
However, stakeholders have signalled in the different consultations that the principles
mentioned above related to the prohibition of general monitoring and the liability system
make difficult the implementation of the provisions of the Directive. Participants at the
workshop on online marketplaces (see Annex 13) agreed that online marketplaces play
an essential role in the supply chain, and that more responsibilities and cooperation with
market surveillance authorities would help to effectively control the safety of products
sold through their platforms. This issue has been partially tackled by the recent proposal
255
Ibidem, page 49 and ff.
256
Directive 2000/31/EC of the European Parliament and of the Council of 8 June 2000 on certain legal
aspects of information society services, in particular electronic commerce, in the Internal Market (E-
commerce Directive) (OJ L 178, 17.7.2000, p. 1).
257
In Case C-324/09, L'Oréal vs. eBay, the European Court of Justice clarified that the relevant question
relating to the conditions for benefiting from a liability exemption was whether eBay was aware of facts
and circumstances from which the illegal activity was apparent (paragraphs 120 to 123).
86
of the Commission for a Digital Services Act258
, a comprehensive set of new rules,
which regulate the responsibilities of digital services that act as intermediaries, including
online marketplaces. In particular, the proposal sets out clear due diligence obligations
for online intermediaries such as a harmonised notice and action mechanism,
transparency obligations, measures to deter rogue traders, cooperation with competent
authorities and an obligation for online marketplaces to gather information on the identity
of traders using their platforms.
5.4.6 Coherence with wider EU policy
The evaluation shows that there are no coherence problems of the Directive with wider
EU policy and priorities. However, some of the stakeholders in the consultations of the
GPSD Study mentioned a few areas where frictions remain. Those issues cannot be
considered as coherence problems of the Directive per se and they cannot be solved by a
single instrument. However, they are worthy to be mentioned below.
As explained in the effectiveness section, the Directive is considered to be effective with
regards to its objective to allow the free movement of goods within the Single Market.
However, stakeholders mentioned that there might coherence problems with regard to
mutual recognition. Stakeholders indicated their uncertainty as to whether a particular
product, and its producers or distributors, could enjoy the free movement of goods, if
there was an option for the Member States to adopt voluntary national certification
systems. According to stakeholders, various Member States have such voluntary national
certification systems in place for various products. Whilst the presence of such
certification systems does not make the market prohibitive for products from outside this
Member State, as there is no obligation to comply with the potential additional safety
requirements, it may make the market more difficult to succeed in without that
certification mark, in particular for SMEs.
With regard to sustainability, several stakeholders highlighted that in their view the
emerging circular economy implied the need for legal certainty with regard to reused,
refurbished and repaired products, and in particular in relation with who within the
circular supply chain should bear the responsibility for the safety of the re-sold products.
In the context of the Sub-group on new technologies and product safety, several
stakeholders mentioned that for harmonised products, concepts such as ‘substantial
modification’ already provided clarity in this respect, and such provisions were missing
for non-harmonised products.
2.5.2. 5.4.7 Conclusion on coherence
It results from this evaluation that overall, the Directive shows a high degree of
internal coherence, as there are no major contradictions among its provisions, as well as
external (i.e. with other pieces of EU legislation), playing a key role in the Single Market
framework. Further coherence however could be sought through more clarity.
258
COM(2020) 825 final Proposal for a Regulation of the European Parliament and of the Council on a
Single Market For Digital Services (Digital Services Act) and amending Directive 2000/31/EC
87
As regards coherence with other pieces of legislation, this evaluation has detected a
major shortcoming regarding the coherence of market surveillance provisions.
While rules for harmonised and non-harmonised are clear and do not overlap between
each other, in practice the coexistence of two systems has an impact on the coherence of
the legal framework, and has implications for market surveillance authorities.
5.5. EU added value
The GPSD Study shows that both authorities and companies/business associations see on
average a “significant added value”. While other stakeholders provide a somewhat less
positive assessment, they consider the Directive to provide still more than “moderate
added value”. A considerable number of respondents across all stakeholder groups found
the Directive to even have “very significant added value”; namely 42 of 141 stakeholders
had this opinion. In contrast, only 9 respondents to our surveys found the GPSD to have
“no” or only “minor” added value. SMEs were also slightly less positive than other
business stakeholders.
The reasons for such positive assessment are clear. Action at EU level is important to
protect consumers and to ensure obligations on the EU businesses evenly. Furthermore,
common rules established under the Directive ensuring rapid circulation of
information among Member States on unsafe products are essential to protect consumer
and ensure a fair internal market. Products freely circulate across the Internal Market.
When an unsafe product is identified in a certain Member State it is very likely that the
same product could be found in other Member States too. This shows the clear cross-
border effect of product safety, which cannot be sufficiently achieved by the Member
States’ individual actions because they cannot ensure cooperation and coordination by
acting alone.
Since the adoption of the Directive there have been no procedures in the EU courts
related to national measures in the area of the health and safety of products that fall under
its scope of application.
The Directive has also contributed to economies of scale with regard to market
surveillance. The Directive allows the performance of joint market surveillance action
and exchange of sensitive information.
Furthermore the Safety Gate/RAPEX system established by the Directive allows faster
and more efficient circulation of information between Member States, ensuring a level
playing field. The circulation of information through the Safety Gate/RAPEX allows to
spread information on unsafe products among both Member States Authorities and
consumers.
At international level, the common set of provisions established by the Directive has
also allowed the EU to be stronger in promoting high level of safety at a more global
level, very important nowadays with the increasing circulation of goods via online
selling.
Repealing the Directive would lead to fragmentation and differing levels of
consumer protection. The safety of consumer products is so intrinsically linked to the
88
internal market that such fragmentation of rules would create a very critical obstacle for
the internal market and for the European consumer.
Therefore, this evaluation concludes that the EU added value of the Directive is
undeniable. Nevertheless, it needs to be recognised that the added value of the Directive
and its harmonious application in the Member States could be further improved by
clearer rules and/or guidance documents.
6. CONCLUSIONS AND LESSONS LEARNT
The role of the Directive as a cornerstone of consumer safety and the functioning of the
Single Market is uncontested. Its objectives remain fully relevant, and its EU added value
cannot be denied. The role of safety net of the Directive remains essential for consumer
protection, as it provides a legal basis aimed to ensure that no dangerous products end up
in the consumer’s hands. The establishment of the Rapid Alert System for dangerous
non-food products under the Directive has proven to be a success, and consequently
expanded via Regulation 765/2008 first and Regulation 2019/1020 afterwards to other
categories of products, such as professional products, and to other risks, such as risks to
the environment, security and other risks to other interests to the Union.
However, this evaluation has exposed a number of factors that put into question the
extent to which some provisions of the Directive contribute to the proper achievement of
the goals that it pursues.
In the first place, the growth of E-commerce has negatively influenced the relevance and
effectiveness of the Directive. As repeatedly stated through this report, the Directive
applies to all consumer products regardless if they are sold in brick-and-mortar shops or
online. However, it can be taken as a lesson learnt that the lack of explicit provisions of
the Directive to address the specificities of online selling, in particular the appearance of
new actors in the online supply chain, has negatively affected to the safety of EU
consumers as well as the level-playing field for compliant EU businesses. New rules
regarding E-commerce, in particular the Digital Services Act, are likely to contribute to
partially address this issue. However, explicit provisions directly related to E-commerce
and product safety could be envisaged.
The rapid development of new technologies is also questioning the relevance of some of
the key concepts of the Directive. The appearance of some new risks linked to
connectivity, the applicability of the Directive to software updates and downloads as well
as the evolving functionalities of AI-powered products raise the question whether the
Directive is clear enough to provide legal certainty for businesses and protection to
consumers. While legislative measures should be sought in product harmonised
legislation to address the particularities of these technologies, this evaluation has also
showed a need to ensure that the Directive remains as the safety net for all consumer
products, including those incorporating new technologies.
The evaluation has also identified the problems for the coherence of the EU legal
framework and the existence of two different sets of rules on market surveillance for
89
harmonised and non-harmonised products, that have also clear implications for the
effectiveness and efficiency of the Directive. An alignment between both frameworks, as
well as in the definitions of the Directive with product harmonisation legislation, should
be sought.
Finally, it is also apparent from this evaluation that it would be necessary to fine-tune
some of the provisions to improve the effectiveness of the Directive. In particular,
legislative changes or further actions are needed to improve the effectiveness of product
recalls. There is a need for mechanism to arbitrate disputes between Member States
regarding risk assessments. The traceability system under the Directive and the
resources constraints of market surveillance authorities are challenges that make it
difficult to effectively control the safety of products, and consequently need to be tackled
to ensure a proper protection of consumers and functioning of the Single Market.
Provisions of the FIPD on food-imitating products are currently not enforced in a
harmonised manner among Member States, so a solution to address this challenge is
needed.
90
Appendix 1: Procedural information
1. LEAD DG, DeCIDE PLANNING/CWP REFERENCES
 Lead DG: Directorate General for Justice and Consumers (DG JUST)
 Decide/Planning: PLAN/2019/6283 Review of the general product safety directive -
Proposal for a regulation on general product safety.
 REFIT (Evaluation)
2. ORGANISATION AND TIMING
 An Inter-service Steering Group oversaw the process to ensure coherence and
comprehensiveness with the Commission’s overall responsibilities and activities in
related policy areas.
 This GPSD ISSG held 5 meetings times (one informal meeting on 14/02/2020 and four
formal meetings on 12/06/2020, 08/10/2020, 18/11/2020 and 07/12/2020). DG JUST
consulted the ISSG on the different steps of this initiative: Roadmap/Inception Impact
Assessment, Consultation strategy, open public consultation questionnaire, the study
underlying the evaluation and impact assessment (ISSG provided comments on all study
steps and reports) and finally on the draft evaluation.
 Publication in EUROPA of the Roadmap on the evaluation, 30 June 2020
 Launch of the Open public consultation, 30 June 2020 - 6 October 2020 (14 weeks)
3. EXCEPTIONS TO THE BETTER REGULATION GUIDELINES
This Evaluation has been conducted back-to-back to the Impact assessment.
4. CONSULTATION OF THE RSB (IF APPLICABLE)
This Evaluation report is one of the annexes to the Impact Assessment report, on which the RSB
has been consulted.
5. EVIDENCE, SOURCES AND QUALITY
Studies commissioned or supported by the European Commission
 Study to support the preparation of an evaluation of the General Product Safety
Directive as well as of an impact assessment on its potential revision, Civic
consulting, December 2020
 Study for the preparation of an Implementation Report of the General Product
Safety Directive, Civic consulting, July 2020
 Study on the assessment of the opportunities for increasing the availability of EU
data on consumer product- related injuries, European Commission’s Joint
Research Centre’s, May 2020
 Behavioural Study on strategies to improve the effectiveness of product recalls,
LE Europe, December 2020
 Survey on consumer behaviour and product recalls effectiveness, April 2019
 Implementation of the new Regulation on market surveillance: indication of
origin, VVA Europe, May 2015
91
External Expertise
 Consumer Safety Network (CSN)
 Sub-group on Artificial Intelligence, connected products and other new
challenges in product safety to the Consumer Safety Network
Selective bibliography
 Bernstein A. (2013), ‘Voluntary Recalls’, University of Chicago Legal Forum, 1:
394 ff., available at: http://chicagounbound.uchicago.edu/uclf/vol2013/iss1/10
and Jacoby J. (1984), ‘Perspectives on Information Overload’, Journal of
Consumer Research
 OECD (2020)- ‘E-commerce in the time of COVID-19’, available at
http://www.oecd.org/coronavirus/policy-responses/e-commerce-in-the-time-of-
covid-19-3a2b78e8/#biblio-d1e705
 OECD (2018), ‘Enhancing Product Recall Effectiveness Globally’, available at
https://www.oecd-ilibrary.org/science-and-technology/enhancing-product-recall-
effectiveness-globally_ef71935c-en
Other Sources
 Eurostat
 European Injury Database (IDB)
 Safety Gate/RAPEX
 WHO CHOICE
92
Appendix 2: Analytical methods
This Appendix shows the analytical methods additionally used for the preparation of the
Evaluation. For more general information about the Analytical methods, please see
Annex 4.
5.1. Estimation of costs of compliance with the Directive for EU businesses
(for efficiency criterion)
Firstly, there is a need to calculate the estimation of the baseline market size, i.e. the total
turnover of EU businesses from manufacturing and/or selling non-harmonised consumer
products in the EU259
, before analysing company level compliance cost data, and
extrapolating it to EU level, based on the estimated baseline market size. The analysis is
structured according to six steps:
Step 1: Estimation of EU companies’ total annual turnover from the production and/or
sales of non-harmonised consumer products in the EU
Based on NACE industry codes and sector descriptions, those manufacturing sectors
were identified (NACE Rev. 2, B-E), wholesale services sectors and retail sectors
(NACE Rev. 2, G) in which consumer products are produced and/or sold, i.e. sectors that
clearly focus on the production and sales of industrial products were excluded. Sectors
related to motor vehicles have been excluded, in line with the focus on non-harmonised
consumer products. While retail sale can be assumed to be largely related to consumer
products (although retailers may also sell to professional users, and may sell services),
the wholesale and manufacturing in the listed areas clearly also contain
industrial/professional products, an issue considered in Step 3 below. To arrive at the
share of non-harmonised products produced and/or sold in these sectors, it was applied
the estimate provided in the 2017 EU impact assessment for the new Market Surveillance
Regulation, which estimated that about 54% of products circulating within the European
Single Market are harmonised products and 46% are non-harmonised products260
.
Step 2: Deduction of extra-EU export
To calculate the net turnover for non-harmonised consumer products that are only sold in
the EU, the share of extra-EU exports from the total turnover of EU companies was
deducted. The calculation is based on an approximation of sector-specific export shares.
The extra-EU trade by enterprise characteristics data provided by Eurostat do not exactly
match the sector classification of turnover data by enterprise size class. Therefore the
259
All estimates in this section refer to the EU27 as of 2020.
260
SWD (2017) 466 final PART 2/4 Commission Staff Working Document Impact Assessment
Accompanying the document COM (2017) 795.
93
extra-EU export shares of manufacturing, wholesale and retail sectors were approximated
on the basis of those sectors261
. The estimated extra-EU export shares of manufacturing,
wholesale and retail sectors were subtracted from the annual turnover of EU companies
with non-harmonised products in the selected sectors.
Step 3: Deduction of industrial and professional products
The EU turnover derived in Step 2 was corrected by the percentage shares of turnover
that can be attributed to the production and/or sales of consumer products in
manufacturing, wholesale and retail sectors. For this purpose, a different dataset was
drawn, namely the final consumption expenditure of households by consumption
purpose262
. There was a correction for the share of harmonised products, and arrived at
an estimate for total household consumption of non-harmonised products. For the
following analysis it was assumed that this consumption of non-harmonised consumer
products is equivalent to the total turnover from non-harmonised consumer products sold
by EU retailers. The estimated retail turnover from non-harmonised products indicated
before was adjusted accordingly, and the resulting amount was allocated between the
three enterprise size classes. Due to data limitations, the same methodology could not be
applied for manufacturing and wholesale sectors263
. For manufacturing and wholesale
sectors, it was estimated the share of turnover that can be attributed to consumer products
on the basis of the share of “consumer-oriented” wholesale services in total wholesale
services. It is assumed that the same share reflects the portion of consumer products
produced and/or sold by manufacturers. Based on this approach, the total annual EU
turnover of EU companies from non-harmonised consumer products was calculated.
Step 4: Derivation of empirical estimates for companies’ product safety-related costs on
the basis of survey responses
In the GPSD study, businesses were asked to indicate staff time used for managing
product safety, testing for product safety, recalls and other consumer product safety
related activities. Respondents were asked to consider all costs for ensuring product
safety of both harmonised and non-harmonised consumer products (excluding
pharmaceuticals, medical devices or food), as the identification of costs for non-
harmonised products only was not considered to be feasible. In addition to staff
requirements, companies were asked to provide estimates for other costs to comply with
safety requirements for consumer products (e.g. costs for external legal advice, costs for
261
These sectors are: “Manufacture of textiles, Manufacture of wood and of products of wood and
cork, except furniture; manufacture of articles of straw and plaiting materials”, “Manufacture of paper and
paper products”, “Manufacture of computer, electronic and optical products”, “Manufacture of electrical
equipment”, “Manufacture of furniture”, “Wholesale trade, except of motor vehicles and motorcycles”, and
“Retail trade, except of motor vehicles and motorcycles”.
262
Eurostat, Final consumption expenditure of households by consumption purpose (COICOP 3
digit) [nama_10_co3_p3].
263
Eurostat data do not allow to extract “pure” consumer products for manufacturing and wholesale
sectors, i.e. final products that are consumed by households.
94
external safety testing, costs for certification of safety of products etc.)264
. The cost
estimates provided by the respondents also include business-as-usual costs, which would
incur even in absence of product safety regulation (see Step 6). These estimates were
used to estimate companies’ annual regulatory compliance costs in Euro terms. The
calculation of Euro-denominated costs for staff was based on the EU’s (weighted)
average wage for the business economy, which in 2019 was 27.50 Euro per hour265
. To
account for overhead costs, a 25% mark-up was added to staff-related costs.
Subsequently, the costs for each company were related to the EU turnover for consumer
products, i.e. companies’ annual cost resulting from activities to comply with safety
requirements for (all) consumer products as a share of the related turnover.
Step 5: Extrapolation of EU companies’ annual costs related to the Directive including
business-as-usual costs that occur also in absence of regulation
For each enterprise size class, the empirical median values for companies’ relative
product safety-related costs (Step 4) were multiplied by the annual turnover of EU
companies that can be attributed to the production and/or sales of non-harmonised
consumer products in the EU (Step 3). The results of this calculation still include
business-as-usual costs.
Step 6: Deduction of business-as-usual costs and extrapolation of EU companies’ annual
compliance cost related to the Directive
In the GPSD Study, businesses were asked to indicate the share of the total product
safety-related costs that they would incur anyway (i.e. even in absence of product safety
legislation, e.g. because these costs relate to due diligence), hereafter referred to as
business-as-usual costs, BAU. These estimates reflected the self-assessment of the
companies that are part of the sample, and are therefore subjective in nature. However, as
concerns differences between manufacturers, on the one hand, and wholesalers and
retailers, on the other, it was considered that the estimates to be in line with expectations
and a credible basis for the final step of the assessment. The empirical median values of
these shares were applied to the product safety-related cost estimates derived in Step 5.
Excluding business-as-usual costs, compliance costs of EU companies were obtained that
can be attributed to non-harmonised consumer products, i.e. the costs for businesses to
comply with the Directive.
5.2. Estimation of costs of compliance with the Directive for Member States
(for efficiency criterion)
The estimation of market surveillance authorities’ staff-related costs related to market
surveillance activities for non-harmonised consumer products in the EU was based on the
following three steps:
264
Business stakeholders were asked to estimates average costs per month in EUR.
265
Labour cost for LCI (compensation of employees plus taxes minus subsidies), provided by
Eurostat.
95
Step 1: Identification of authorities’ annual FTEs for market surveillance activities
related to non-harmonised consumer products
The number of full time equivalent (FTE) staff for market surveillance of consumer
products as provided in the country research was used. Where the available country
estimates related to the market surveillance of non-harmonised consumer products, this
figure was directly used in the calculation. Where estimates related to the total staff for
market surveillance of both harmonised and non-harmonised consumer products, staff
was allocated according to the 54%/46% ratio for harmonised/non-harmonised products
circulating within the European Single Market to derive an estimate for related market
surveillance activities266
. It should be noted that a share of 46% in staff time for market
surveillance of non-harmonised consumer products is 12 percentage points higher than
the empirical median share indicated by authorities for activities devoted to non-
harmonised products in the stakeholder survey (34%), potentially causing an estimate at
the higher end of authorities’ actual costs that can be attributed to market surveillance
activities for non-harmonised consumer products. For seven countries, no information on
staff numbers was available at all.
Step 2: Approximation of annual FTEs for market surveillance activities related to non-
harmonised consumer products for countries for which data was not available
For the seven countries, for which no staff data was available (Croatia, Germany,
Hungary, Italy Slovenia, Slovakia, and Spain) it was estimated the number of FTEs on
the basis of the data for the remaining 20 Member States. To account for institutional
differences with regard to the level of centralisation, it was considered that two clusters
of countries, in line with the characteristics of the respective market surveillance systems
as described above: Cluster 1: responsibility for market surveillance is centralised (no
sub-national administrations involved); Cluster 2: responsibility for market surveillance
is (partly) delegated to or competence of sub-national administrations, in line with the
administrative structure of the country.
To derive estimates for the number of FTEs per million population for Slovenia and
Slovakia (more centralised market surveillance), the sample median of 3.5 FTEs per
million population was applied. To derive FTE estimates for the number of FTEs per
million population for Croatia, Germany, Hungary, Italy and Spain (more decentralised
market surveillance), the sample median of 4.6 FTEs per million population was applied.
266
As mentioned before, the 2017 EU impact assessment for the new Market Surveillance Regulation
estimated that about 54% of products circulating within the European Single Market are harmonised
products and 46% are non-harmonised products. See SWD (2017) 466 final PART 2/4 Commission Staff
Working Document Impact Assessment Accompanying the document COM(2017) 795.
96
Step 3: Calculation of annual staff costs for market surveillance activities related to non-
harmonised consumer products
In the final step, it was calculated that the EUR equivalent of the estimated number of
staff required for market surveillance of non-harmonised consumer products by
multiplying the number of FTEs per million population by:
 The size of population for each country (in million);
 The number of person-hours per year (1 720) 267
; and
 The average wage of 28.00 EUR, which corresponds to the EU27 average wage
of “administrative and support service activities” (18.70 EUR) and “professional,
scientific and technical activities” (37.30 EUR) for 2017 (latest figure available in
Eurostat database).
5.3. Methods for other supporting estimations (for effectiveness criterion)
Other supporting estimations include the analysis of data from the rapid alert system.
Data from Safety Gate/RAPEX was used for the analysis of the baseline situation and the
related problem analysis. For this purpose, a full dataset covering the years 2005 to 2019
was retrieved. The dataset consisted of a total of 25 850 notifications that are publicly
available. The dataset included 25 051 notifications concerning products with serious
risks, 738 notifications of products with other risk levels, and 61 other types of alerts.
This dataset was merged with a second dataset covering notifications in the period 2011
to 2019, which included complementary (not publicly available) data.
267
Following EU Horizon 2020 guidelines, one person year corresponds to 1 720 person-hours per
year. See, e.g. the H2020 Programme: User's Guide for the Personnel Costs Wizard.
97
Appendix 3: Evaluation questions
Relevance
To what extent the initial objectives of the GPSD correspond to the current needs?
To what extent is there a need to clarify concepts set out in the GPSD, such as
“product”, “safe product”, “serious risk” and “placing on the market”?
How well adapted is the GPSD to online sales?
How well adapted is the GPSD to challenges posed by new technologies, such as
cybersecurity risks in relation to safety, self-evolving products and stand-alone
software or emerging safety issues in the post-market phase of the product?
How well is GPSD adapted to increased level of direct imports towards the EU?
How well adapted is the GPSD to environmental issues with health impact? In
particular, how this health impact is considered by taking into account the
assessment done under REACH related to chemicals?
Effectiveness
To what extent does the GPSD meets its objectives of achieving a high level of
consumer protection through the reduction of unsafe products and contributing to
the functioning of the Single Market? Which are the main elements that have
contributed to this? Is there anything missing?
To what extent has the market surveillance system established by the GPSD (in
particular the Rapid Alert System for dangerous non-food products) been effective?
How has the development of e-commerce affected the effectiveness of the GPSD?
How has the development of new technologies, such as Artificial Intelligence, Internet
of Things and connected devices, affected the effectiveness of the GPSD?
How effective has been the development and use of the standards supporting the
implementation of the GPSD?
How well is GPSD adapted to ensure efficient corrective actions are taken, in
particular recalls?
How well is GPSD adapted to ensure effective market surveillance?
Are there any aspects/means/actors that render certain aspects of the Directive more
or less effective than others (including product recalls), and if there are, what lessons
can be drawn from this?
What are, if any, the consequences or effects (either positive or negative) that were
not originally planned?
98
Efficiency
What are the regulatory (including administrative) costs of the GPSD for the
different actors involved (Member States authorities, businesses, consumers) and for
the society overall? In particular, what is the economic cost for businesses to comply
with the GPSD?
What are the benefits of the GPSD for the different actors involved (Member States
authorities, businesses, consumers) and for the society overall?
To what extent are these costs proportionate to the benefits?
What factors influenced the efficiency of reaching the objectives which the GPSD sets
out?
Coherence
Are there any discrepancies and/or inconsistencies between the provisions of the
GPSD?
Are there overlaps and/or complementarities between the GPSD and any other Union
legislation with similar objectives, in particular regarding market surveillance,
product harmonisation legislation, including horizontal legislation on chemicals and
food contact materials legislation, standardisation, consumer protection law and
product liability, and also other union legislation such as the E-commerce Directive?
To what extent is the Directive coherent with wider EU policy, such as rules on free
movement of goods, mutual recognition, customs, competition, industrial policy,
sustainability (environmental protection) and trade?
EU added value
What is the added value of the GPSD compared to what could reasonably have been
expected from Member States acting at national level?
What would be the most likely consequences of withdrawing the GPSD? How would
it affect the functioning of the Single Market and the health and safety of consumers?
99
Appendix 4: Implementation of the Food Imitating Products
Directive (FIPD)
Council Directive 87/357/EEC (the Food-Imitating Products Directive) was adopted to
address the lack of harmonisation amongst national measures trying to ensure product
safety of products ‘appearing to be other than they are’. Before the adoption of the
Directive, there were legal provisions or regulations in force in several Member States
concerning certain products which, appearing to be other than they are, endanger the
safety or health of consumers. However, these provisions differed in content, scope and
field of application. In particular, these provisions concerned in certain Member States all
products which resemble foodstuffs while not being such whilst in other Member States
they concern products likely to be confused with foodstuffs, especially confectionery.
This situation created significant barriers to the free movement of goods and unequal
competitive conditions within the Community without ensuring effective protection for
consumers, especially children.
The products covered by the FIPD possess, pursuant to Article 1(2), a ‘form, odour,
colour, appearance, packaging, labelling, volume or size’ that consumers, especially
children, could confuse with foodstuffs, and endanger health of safety of consumers
(Article 1(1)). The fact that these products imitate foodstuffs could then lead to
consumers putting such products in their mouths, sucking or ingesting them, which could
be dangerous. This led the European legislator to provide that Member States should take
all measures necessary to prohibit the marketing and introduction of such products on the
market268
through the above-mentioned Directive. The justification for the adoption of
this measure was twofold: to improve consumer protection, especially the protection of
children, as well as to ensure a level-playing field on the Internal Market of such
products. The latter goal aimed at eliminating barriers to the free movement of goods that
could imitate other products, but which would not create serious risks to consumer
protection.
While most Member States have implemented the Food-Imitating Products Directive into
national legislation as in the Directive, without additional provisions, there are
differences in interpretation. Some authorities perceive products in this category as
dangerous per se, whilst others are of the opinion that any serious risks need to be proven
through an appropriate risk assessment procedure. Indeed, the Food-Imitating Products
Directive has been adopted before the GPSD, which sets out the principle of the necessity
of risk assessment of the product safety before taking appropriate measures against
products and some Member States started to apply to food-imitating product the GPSD
logic while others stayed at the primary interpretation of the Food-Imitating Products
Directive as a ban of these products. This had then led to differences in the national
assessment whether a particular food-imitating product should be prohibited from the
market. However, the number of Safety Gate/RAPEX notifications of food-imitating
268
Article 2 Food Imitating Directive.
100
products is small. Between 2013 and 2019, a total of 114 notifications that relate to food-
imitating products269
. The table below shows the product categories for these
notifications in the period 2013 to 2019. The number of notifications of food-imitating
products is fairly small – up to 36 notifications out of the approximately 2 000
notifications annually. The number has varied significantly over the years, from 1 to 36
notifications annually.
Table 9: Notifications to Safety Gate/RAPEX related to food-imitating products
Product category Year Total
2013 2014 2015 2016 2017 2018 2019
Cosmetics 3 1 28 1 33
Decorative articles 1 1 4 17 23
Food-imitating products 26 12 8 46
Other 2 2
Stationery 2 2
Toys 1 4 3 8
Total 27 13 8 4 1 36 25 114
Source: Safety Gate/RAPEX
The table shows that the product category “Food-imitating products” was only used up to
2015. Afterwards, the products have been categorised according to their use (cosmetics,
clothing, etc.). This seems to indicate that a change of practice has occurred in the
Member States to remove the overlap between the category “food-imitating products”
and other product categories. Most of these notifications makes reference to the Food-
Imitating Products Directive in the description of the risk and include a statement like
“The product does not comply with the Food-Imitating Products Directive.” Apparently,
many authorities ban a food imitating product because the Food-Imitating Products
Directive directly bans such products without the need for a risk assessment. However,
this raises questions regarding the proportionality of such measures.
The vast majority of the notifications related to food-imitating products (87%) mentions
or includes choking in the description of the risk associated with the product, presumably
because the product is or contains small parts. The second-most common risk type is
“chemical” (12%).
There is little evidence available regarding the adverse effect of food-imitating products.
A 2011 opinion by the Scientific Committee on Consumer Safety concluded that “Few
cases of accidental ingestion of food-resembling or child-appealing products are reported.
This may be due to the lack of sufficient registered information to discriminate these
269
These are identified in different ways, and some cases meet several of the criteria at the same time: The
parameter “Category” includes “Food imitating products” (46 notifications); The parameter “Product”
includes the text “imitat” (6 notifications); The parameter “Description” contains the text “imitat” (46
notifications); The parameter “Risk” contains the text “imitat” (57 notifications). Cases that were identified
using the filtering term “imitat” have subsequently been reviewed manually to remove cases that did not
refer to food (e.g. notifications related to “leather imitation”, “imitation of gun”, etc.)
101
types of products. Data from poison centres and scientific literature on accidental
ingestion of cosmetics or liquid household products suggest that the majority of such
ingestions result in mild gastrointestinal effects. […] The weight of evidence from
accidental ingestion of cosmetics suggests that there is a low risk of acute poisoning in
either children or the elderly. For household products, there is a slight increase of a more
serious outcome”. From the opinion, which focused on chemical consumer products
resembling food and/or having child-appealing properties’, it appears that these food-
imitating products rarely represent serious or high risks. However, the opinion also
concludes that “there is a lack of specific data on accidental ingestion from consumer
products resembling food and/or having child-appealing properties”.
It can therefore be concluded that while a majority of Member States seems to apply the
provisions of the Food-Imitating Products Directive only in cases where the risks are
serious, there are also countries that consider products in this category as dangerous per
se. In other words, the legal framework for food-imitating products is applied differently
in different countries.
102
Annex 6: Effects of the COVID-19 crisis in the context of the
policy options and their expected impacts
The analysis of the impact of COVID-19 in the Study underlying the GPSD Evaluation
and Impact assessment is based on macroeconomic data and a series of interviews
conducted with companies active in (also) non-harmonised consumer products. The
Study explored the companies’ views on the policy options and potential effects of
COVID-19 in this respect, as well as expectations concerning relevant long term,
structural changes.
In terms of structural changes that would need to be considered for changes in the EU
legislative framework in general, interviewees expected:
 A change in consumer behaviour towards more quality products that are also
more eco-friendly;
 Energy efficiency will be an important topic and people will likely buy more
sustainable and environmentally friendly products;
 Promoting the reuse, refurbishment and recycling of used products;
 A stronger focus on hygiene is expected, and the safety of these products will
have a more prominent role for consumers.
Furthermore, in the short-term, interviewees expected practical difficulties in
conducting market surveillance for authorities due to COVID-19 restrictions. An
interviewee noted that COVID-19 had led to a reduction in controls. Accordingly, this
interviewee considered that the need for a clearer GPSD has increased as it would help
conducting market surveillance more effectively in the ‘new normal’ with less visits of
inspectors.
Interviewees also expected in the medium-term to long-term that reduced public budgets
(due to potential austerity measures after the pandemic) would mean that the downward
trend in market surveillance capacities in Member States that they had noted after the
financial crises would continue. According to their view, this would increase the need for
a less complex legal framework and a (resulting) more efficient market surveillance,
more efficient recall procedures, and increased support through EU programmes.
Concerning the impacts of the options for a possible revision of the GPSD, interviewees
emphasised the overall impact of COVID-19 on the baseline situation, i.e. the
increased importance of e-commerce, including with third countries, which was expected
to put additional demands on authorities in terms of online market surveillance. Safety of
products sold online is therefore expected by interviewees to become more important,
which would make Options 3 and 4 (and especially the suggested changes regarding
online sales and online marketplaces) more relevant. All interviewees stressed the
importance of having a common set of rules in the EU, and of reducing administrative
burdens, e.g. to explore differences in legislation between countries. Interviewees also
emphasised that good guidance would help in decision-making of companies.
The Study concludes that the COVID-19 crisis has increased the need for reducing
existing, and avoiding additional administrative burdens, while the growing
103
importance of online sales channels has emphasised the need to close related
legislative and enforcement gaps.
104
Annex 7: SMEs Test
The results of the SMEs test, based on the data from the GPSD Study and other
consultation activities, are summarised in the following table:
(1) Preliminary assessment of businesses likely to be affected
All businesses handling consumer non-harmonised products are likely to be affected, including SMEs.
For SMEs, the estimated annual costs to comply with the GPSD today (after subtraction of
business-as-usual costs) are EUR 428 million per year (companies < 50 employees) and EUR 226
million per year (companies 50 to 249 employees). The median value for consumer product safety-
related costs in proportion of the total annual turnover appears to decrease with the company’s
size/turnover. This is likely due to scale effects. This general pattern is confirmed by SMEs’ replies to
the business stakeholder survey in the GPSD Study. Accordingly, SMEs account for 59% of the total
of GPSD-related compliance costs in the EU, in line with their overall share in the market. (See
section 5 of the Impact Assessment Report, baseline scenario)
(2) Consultation with SMEs representatives
EU SMEs and SMEs associations have been consulted during the preparation of the Evaluation and
Impact assessment. For the GPSD Study, considerable efforts were made to reach out to businesses,
including SMEs and their representatives. The survey questionnaires were widely distributed amongst
SMEs and other business stakeholders as follows: questionnaires were sent to more than 1000 SMEs
and other businesses and more than 300 relevant business associations (including UEAPME,
BusinessEurope, Digitaleurope, EMOTA, EuroCommerce, etc) in all EU27 Member States (plus UK).
Only relatively low number of direct responses were received from SMEs in the GPSD Study (6
survey responses from SMEs, 2 interviews with SMEs, 37 survey responses from business
associations, which often also represent SMEs (e.g. Toy industry, furniture
Industry, etc.). However, no response was received from business associations that only or specifically
represent SMEs.
Business associations representing SMEs also provided feedback on the GPSD Roadmap/Inception
Impact Assessment and replied to the Open Public Consultation (16 replies received from SMEs and
68 from business associations).
Consultation results:
In the survey conducted for the GPSD Study SMEs did by large provide similar assessments to
companies in general. However, SMEs responding to the survey only reported ‘minor’ additional
costs due to differences in the safety requirements in Member States that are caused by differences in
the national implementation of the GPSD (e.g. regarding traceability requirements). In contrast to
larger companies, none of the SME respondents indicated ‘moderate’ or ‘significant’ additional costs
due to differences in the national implementation of the GPSD. A possible reason is that larger
companies are more likely to operate in several EU Member States than SMEs, and therefore
experience relevant legislative differences more often.
Also notable is that in the assessment of the effectiveness of options, as well as concerning the
benefits they bring, business associations often provided considerably lower assessments than
companies, including SMEs, especially regarding policy options involving legislative change.
Specific views of SMEs on the impact of COVID 19 crisis include:
 Regarding the question of how the COVID-19 pandemic has in any way affected how product
safety in companies is safeguarded and any related supply chain issues, an SME respondent
mentioned that testing for new devices is taking longer due to delays related to the pandemic.
105
Such delays for example in supply chain functioning can have a particularly detrimental effect
on business activity of SMEs.
 A SME respondent also stated that the pandemic is likely to have an impact on long-lasting,
structural changes of the 'new normal' in behaviour and economy that are significant for
SMEs. For example, people are expected to buy more sustainable and environmentally
friendly products.
 Regarding the question of how the COVID-19 crisis affects the policy options and their
expected impacts, one SME respondent stressed that it was a bad time to be bringing in
additional regulations for businesses. The interviewee mentioned that many businesses are
really struggling, and that this is especially the case for SMEs. He continued that these
challenges, including for SMEs, also arise from changing consumer habits.
 However, one SME respondent also argued that better clarification of rules would reduce
costs significantly. Cost reductions would take place on the sales level, i.e. less explanation to
buyers and insurance companies, and the design and manufacturing level, i.e. less explanation
of product safety requirements to engineers. But the SME respondent also stated that one-off
costs would generally increase for all regulatory options except for Option 1 due to
familiarisation costs occurring in the organisation, i.e. explaining the new rules internally and
externally. And these costs are especially significant for SMEs, according to the interviewee.
(3) Measurement of the impact on SMEs
The Impact assessment comes to the following conclusions regarding the impacts on SMEs of
the different options (see Section 6 of the Impact assessment Report):
Option 1: No significant firm-level impacts are to be expected due to the implementation of Option 1
for SMEs. Option 1 does not foresee any legislative changes to the GPSD and therefore no new
requirements but also no possibilities for more costs efficiencies for SMEs.
Option 2:
Total costs for SMEs in the EU27 in the first year of implementation are estimated at EUR 21
million. They would fall in subsequent years down to EUR 16.6 million.
Estimated benefits for SMEs linked to cost savings, that are currently caused by differences in the
national implementation of the GPSD and would be partly solved if a the new instrument is a
Regulation under the Option 2, would amount to EUR 34 million annually, compared to the baseline.
Idem for Options 3 and 4 which already foresee the choice of Regulation as legal instrument replacing
GPSD.
SMEs generally estimate that a revision of the product safety requirements of the GPSD according to
Option 2 would bring a variety of at least ‘minor’ to ‘moderate’ benefits270
. At the same time
Option 2 would impose additional adjustment (e.g. familiarisation cost) as well as compliance
costs on SMEs271
, in particular for manufacturers. Data show that SMEs would likely face relative
higher compliance costs than large companies from the implementation of the proposed policy
measures.
Even though the relative cost increases are generally higher for SMEs, the net impact on SMEs
overall costs depends on the benefits that can result from a revised GPSD aligned to the market
270
Significant benefits due to improved quality/lifecycle of products and a deterrent effect on rogue traders, relatively
strong benefits are increased consumer trust, better supply chain management due to improved traceability of products
and better access to the market in non-EU/EEA. These areas are seen as benefits that SMEs assess to be ‘moderate’ to
‘significant’. This is also the case for lower operational risks for businesses and easier compliance with product safety
requirements. By contrast, SMEs considered several benefits to be less than ‘moderate’, including a more level playing
field among businesses and greater legal certainty.
271
This is particularly the case for SMEs that (voluntarily) decide to install and operate customer registration systems.
Similarly, mandatory elements for product recalls (product description with a photograph, description of risk,
instructions on what to do, link to a recall website and free phone number or online service for queries) would increase
the cost of SMEs that have put unsafe consumer products to the market.
106
surveillance rules and traceability requirements in Regulation (EU) 2019/1020. We expect the SMEs
could save some of the costs that currently arise from inconsistencies in the implementation and
enforcement of the GPSD across the EU. Taking into consideration these benefits and the fact that
the changes in SMEs’ costs from Option 2 are very small, we expect that the overall net effect from
Option 2 on SMEs’ costs is rather low and therefore unlikely to affect SMEs’ operations. This is
also confirmed by cost data above.
Option 3:
Total costs for SMEs in the EU27 in the first year of implementation are estimated at EUR 111
million. They would fall in subsequent years down to EUR 99.9 million.
Estimated benefits for SMEs linked to cost savings that are currently caused by differences in the
national implementation of the GPSD and would be partly solved by the choice of Regulation and
would amount to EUR 34 million annually, compared to the baseline.
As concerns the benefits for SMEs, the GPSD Study shows that small companies generally estimate
that Option 3 would bring a variety of at least ‘minor’ to ‘moderate’ benefits, especially due to its
deterrent effect on rogue traders and better detection of unsafe products. However, Option 3 is
considered by small companies as less beneficial when it comes to reducing legal complexity or
making compliance with product safety requirements easier for SMEs. In the case of medium-sized
companies, Option 3 is seen as a suitable contribution to an increased level-playing field among
businesses and to have a significant benefit linked to reducing the occurrence of unsafe products and
for contributing to a better functioning of the EU internal market. Finally, moderate benefits are
expected regarding the potential to increase business revenue or consumer trust.
Even though the relative cost increases are generally higher for the SMEs, the impact on SMEs
overall costs is still considered moderate when measured against the benefits that would result from a
greater level of regulatory harmonisation and reduced regulatory complexity through the choice of a
Regulation. The changes in SMEs costs are estimated to be limited and Option 3 would not be
expected to affect operations considerably272
.
Option 4:
Total costs for SMEs in the EU27 in the first year of implementation are estimated at EUR 187.1
million. They would fall in subsequent years down to EUR 166.3 million.
Estimated benefits for SMEs linked to cost savings that are currently caused by differences in the
national implementation of the GPSD and would be partly solved by the choice of Regulation and
would amount to EUR 34 million annually, compared to the baseline.
Overall, under options 2 to 4, the effects of additional compliance costs will have a larger relative
cost impact on SMEs than on large companies. Even though the relative cost increases are higher for
SMEs, the impact on SMEs overall costs is still considered moderate when measured against the
benefits that would result from a greater level of regulatory harmonisation. The changes in SMEs
costs are small and implementation of any of the options would not be expected to significantly affect
SMEs.
Figures 1 and 2 below show that SMEs assess Option 3 as addressing the best the challenges for
product safety and bringing the highest additional benefits among all the options:
272
This consideration is also true for specific information obligations, such as the obligation for actors across the
online supply chain to provide all safety information online that is also required to be provided with a product in 'brick
and mortar' stores, and the related obligation for online platforms to make sure that third-party sellers, such as SMEs,
provide this information. We expect these costs to be relatively minor for companies selling consumer products on
these platforms, including SMEs.
107
Figure 1: In your view, to what extent would Option [...] effectively address the following
challenges for product safety? – Average across all challenges, by business stakeholder group
(Source GPSD Study)
Figure 2: Where do you see the greatest additional benefits that would result from the
implementation of Option […]? – Average across all benefit categories, by business stakeholder
group (Source GPSD Study)
Options 3 and 4 are seen as providing most benefits by companies overall, as well as SMEs as specific
sub-group, which assessed these benefits on average between ‘moderate’ and ‘significant’. While
SMEs consider Option 2 to also bring more than ‘moderate’ benefits, companies as a whole and
business associations are more sceptical, with an assessment on average across benefit categories of
less than ‘moderate’ benefits regarding Option 2. In general, business associations tend to see less
benefits than companies and SMEs. Note however, that the sample of SMEs responding to the survey
is very small, so that results have to be interpreted with care.
4) Assess alternative options and mitigating measures
The SMEs and micro-SMEs are not exempted from any of the obligations foreseen under the
different options analysed in this Impact Assessment. Indeed, EU product safety legislation does not
allow for "lighter" regimes for SMEs since a consumer product must be safe whatever the
characteristics of its supply chain to meet the general objective of product safety and consumer
protection. However provisions are foreseen in the EU legislation e.g. to facilitate access for SMEs to
EU safety standards including those adopted under the GPSD (see Article 6 of Regulation (EU)
1025/2012). (See Section 5 of the Impact Assessment Report)
1 2 3 4 5
Option 1
Option 2
Option 3
Option 4
No change in Very significant
benefitsat all
Businessassociations Companies SMEs
108
Annex 8: Effectiveness of recalls of consumer products
1. CONTEXT
1.1. Prevalence of recalls in the EU
Product recalls are one of the most common measures to mitigate the risks posed by
dangerous products in the EU273
. Among the over-2000 measures reported each year to
the Safety Gate/RAPEX274
, about half concern recalling products from consumers. A
total of six thousand recalls have been notified in the system from 2013 to 2019, with an
overall increase of approximately 35% (see Figure 1). These figures are likely an
underestimation, as not all recalls in a country are necessarily notified at EU level275
.
Figure 1: Number of recalls registered in the Safety Gate/RAPEX in 2013-2019
The increasing trend can to a large extent be attributed to the increase in the number of
recalls concerning motor vehicles, which grew by a factor of more than 3 from 159
recalls in 2013 to 507 (i.e. 48%) in 2019. Apart from motor vehicles, the five most
frequently recalled product categories according to Safety Gate/RAPEX alerts were toys;
clothing and textiles; electrical appliances and equipment; lighting equipment; and
childcare articles and children equipment.
273
By the term “recall” we refer to the process aimed in particular at achieving the return of a dangerous
product that has already been supplied to consumers, initiated directly by the producer or distributor of
the dangerous product, or ordered by authorities.
274
https://ec.europa.eu/consumers/consumers_safety/safety_products/rapex/alerts/repository/content/pages/r
apex/index_en.htm
275
As regards products posing a less than serious risk, notification is encouraged but not mandatory
in the case of voluntary measures taken against products covered by the GPSD and in the case of both
voluntary and compulsory measures taken against products subject to EU harmonised legislation. In
addition, Member States are not required to notify corrective measures in cases where the effects of the
product risk cannot go beyond the territory of the Member State.
0
200
400
600
800
1000
1200
2013 2014 2015 2016 2017 2018 2019
109
1.2. Legal provisions
Under the GPSD, producers and distributors have to inform immediately the competent
national authorities when they become aware that a product they have placed on the
market is dangerous, giving details of the actions taken to prevent risks to consumers276
.
Producers have the primary responsibility for the safety of products they place on the
market. If necessary to avoid risks to the health and safety of consumers, they have the
obligation to take appropriate action (including withdrawing the dangerous product from
the supply chain, warning consumers or, as a measure of last resort, recalling products
that have already been supplied to consumers)277
.
Distributors have to act “with due care” and must not supply products which they know
are unsafe. They also have to cooperate in the action taken by producers and competent
authorities to avoid the risks and pass on information on product risks278
.
Member States authorities have the power to order a recall or to coordinate or, if
appropriate, to organise it together with producers and distributors279
. In general,
authorities shall encourage and promote voluntary actions by producers and distributors,
including where applicable by the development of codes of good practice. However, if
voluntary action is unsatisfactory or insufficient, they shall order the measures or
organise them themselves280
.
There are no specific rules at the EU level on how a recall should be organised. The
GPSD only states that "recall shall take place as a last resort, where other measures
would not suffice to prevent the risks involved, in instances where the producers consider
it necessary or where they are obliged to do so further to a measure taken by the
competent authority” and that it “may be effected within the framework of codes of good
practice on the matter in the Member State concerned, where such codes exist”.
In fact, few Member States have established such codes of good practice or guidelines on
recalls (see table 1 below). In addition, most consist of a very brief description of the
recall process with little further requirements.
Table 1: National guidance documents on recalls
Country Website
Austria https://www.sozialministerium.at/Themen/Konsumentenschutz/Produktsicherheit/Gefaehrliche-
Produkte-und-Rueckrufe.html
Belgium https://economie.fgov.be/fr/themes/qualite-securite/securite-des-produits-et/rappel-dun-produit-
ou-autre
Denmark https://www.sik.dk/erhverv/produkter/vejledninger/generelle-vejledninger-om-
produkter/tilbagetraekning-og-tilbagekaldelse-produkter
Finland https://tukes.fi/en/products-and-services/dangerous-products
France A guide on product recalls is being developed
Germany https://www.baua.de/DE/Themen/Anwendungssichere-Chemikalien-und-
276
GPSD Art 5 (3).
277
See GPSD Art 5 (1), (b) of the third subparagraph, and last paragraph.
278
GPSD Art 5 (2).
279
GPSD Art. 8(1)(f)(ii).
280
GPSD Art. 8(2).
110
Produkte/Produktsicherheit/Rueckrufmanagement/Handlungsempfehlungen.html
https://www.baua.de/DE/Themen/Anwendungssichere-Chemikalien-und-
Produkte/Produktsicherheit/Rueckrufmanagement/Rueckrufmanagement_node.html
Norway https://www.dsb.no/lover/produkter-og-forbrukertjenester/veiledning-til-forskrift/veileder-om-
meldeplikt-ved-farlige-produkter/
Sweden https://www.konsumentverket.se/for-foretag/produktsakerhet/salt-farlig-vara/
In the absence of EU-wide rules on recall procedure, communication or remedies, each
Member State follows its own approach, with some common elements (e.g. in most
countries, voluntary recalls are more common than mandatory ones), but also diverging
requirements.
The increase in the number of product recalls over time and the fact that most recalls take
place voluntarily can be considered as an indication that the GPSD has contributed to
making recalls more widely used as a corrective measure. On the other hand, the lack of
minimum EU-wide requirements regarding recall communication, remedies or
monitoring has been repeatedly reported as a significant shortcoming, with negative
impacts on consumer safety and on level-playing field for businesses.
2. ASSESSMENT OF RECALL EFFECTIVENESS
The effectiveness of product recalls varies considerably depending on factors such as
customer traceability and product type. Recalls tend to be considerably more effective, if
affected consumers can be identified and contacted directly (e.g. because the product was
registered, bought online or delivered to customer’s home). Recall effectiveness also
increases with product price and expected lifespan and decreases with product age. In the
automotive sector, up to 100% success rates have been reported, thanks in particular to
mandatory motor vehicle registration281
. Likewise, Samsung’s recall of over 4.6 million
Galaxy Note7 phones resulted in 90% return rate within four months and a further 7%
within 7 months thanks to the sending of over 23 million alerts and push notifications
to the company’s customers and a software update that reduced battery capacity up
to 0%282
.
Success rates tend to be much lower for cheaper products and when it is not possible to
reach out to affected consumers. In general, the proportion of products successfully
recovered from consumers remains low, as recognised by a recent OECD report283
. For
instance, one Member State indicated that the return rates for recalled products rarely
exceed 10%, except when products have been purchased online284
. Another national
authority estimated that around 80% of products that have relatively low value and short
lifespan remain in consumers’ hands285
.
Recall participation also depends on consumers’ characteristics. Socially disadvantaged,
relatively young and less safety-conscious consumers are less likely to both participate in
recalls (especially if such participation is time-consuming) and to register their products
281
German Federal Motor Transport Authority (KBA) presentation at the Interrnational Product
Safety Week, 11/10/2010, https://youtu.be/S-RGd4jVhvQ, iTWire (2021). Car manufacturers complete
99.9% of Takata airbag recall, https://www.itwire.com/automotive/car-manufacturers-complete-99-9-of-
takata-airbag-recall.html.
282
OECD (2018), Measuring and maximising the impact of product recalls globally, p. 9.
283
OECD (2018), Enhancing Product Recall Effectiveness Globally, p. 5.
284
Idem, p. 17.
285
European Commission, 2021, Behavioural study on strategies to increase the effectiveness of product recalls.
111
(making them more difficult to contact in case of a recall)286
.The EU-wide societal cost of
recalled products remaining in consumers’ hands have been estimated at approximately
€378 million in 2019 due to healthcare costs, productivity losses and losses of quality of
life.287
The consequences of delayed and badly managed recalls are also exemplified by the
deaths and injuries caused by products such as faulty Takata airbags (estimated to have
cause 35 deaths and 300 injuries worldwide288
) and Fisher-Price rock ‘n play baby
sleepers (associated with 59 baby deaths in the US289
).
3. REASONS FOR LIMITED RECALL EFFECTIVENESS
3.1. Many consumers are not aware of product recalls
The main obstacle to recall effectiveness is the difficulty of reaching out to the owners of
recalled products, which means that many EU consumers are simply not aware that they
own a recalled product.
3.1.1. Limited direct communication with consumers
There is a general agreement that direct communication with consumers (e.g. via email,
telephone, SMS or connected devices) is more effective not only in reaching affected
consumers but also in encouraging consumer response compared to indirect methods
such as press releases or recall announcements published on companies’ and authorities’
websites. In an online experiment carried out by the European Commission 72% of
respondents who were presented with a direct recall notification acted on it compared to
31% of respondents who saw a generic notification with the same description of hazard,
action to take and remedy. Based on these figures, the EU-wide cost savings from using
direct recall communication have been estimated at €73 million in 2019, i.e. a fifth of the
overall estimated cost of recall ineffectiveness290
.
Effectiveness of generic vs direct recall notification
286
Idem.
287
Idem.288
https://www.consumerreports.org/car-recalls-defects/takata-airbag-recall-everything-you-need-to-
know/
288
https://www.consumerreports.org/car-recalls-defects/takata-airbag-recall-everything-you-need-to-know/
289
https://www.washingtonpost.com/gdpr-
consent/?next_url=https%3a%2f%2fwww.washingtonpost.com%2fbusiness%2f2019%2f10%2f17%2fstudy-
concludes-design-rock-n-play-other-infant-sleepers-led-deaths%2f
290
European Commission, 2021, Behavioural study on strategies to improve the effectiveness of
product recalls.
112
Source: Online experiment
Direct contact is also seen as the most effective communication channel by industry
stakeholders291
and indicated as the preferred communication method by all consumer
groups292
.
Effectiveness of different communication channels (sum of replies: ‘effective’ and ‘very effective’)
Source: Online industry survey
These findings are corroborated by actual recall monitoring data from the US Consumer
Product Safety Commission (CPSC), which found that direct recall alerts result in an
291
Idem.
292
European Commission (2019). Survey on consumer behavior and product recalls effectiveness, p.
20, available at:
https://ec.europa.eu/consumers/consumers_safety/safety_products/rapex/alerts/repository/tips/Product.Reca
ll.pdf.
113
average return rate of 50% compared to 6% for joint press release issued by the CPSC
and the recalling company293
.
Yes, authorities interviewed for the GPSD Implementation Study recognised that
identifying and directly reaching out to affected consumers is a challenging task. The
Directive does not contain any requirements on consumer traceability. As for national
guidance documents, Finnish and Belgian ones require that affected consumers be
contacted directly, whenever feasible, but in most countries there are no similar
requirements.
For most recalled products, customer data is not available and even in situations when it
is available, it is not always used to reach out to the owners of recalled products because
of data protection concerns.
3.1.1.1.Suboptimal use of product registration schemes
When a consumer registers a product, he or she provides information (e.g. an email
address) that personally links them to this product and allows for direct contact in case of
product recalls or safety warnings. Yet, apart from motor vehicles (whose registration
with public authorities is mandatory), registration schemes are only available for few
higher-value product categories like domestic electronic appliances and communication
devices294
. In addition, even in these sectors actual registration rates tend to be rather
low. In a recent consumer survey, declared registration rates were 37% for
communication devices, 33% for domestic electrical appliances, 24% for childcare
articles and 8% for toys295
.
The main barriers to product registration include:
 No link between product registration and safety: Research on existing EU
registration schemes indicated that product safety is in general not highlighted in the
analysed registration schemes and the main benefit of registration is framed in terms
of general customer support or marketing. Only 4 out of the 40 analysed schemes
mentioned product safety as a sole or one of the benefits of product registration in the
invitation to register296
. As a consequence, consumers seldom see the safety benefit of
registering their products. Only 40% of consumers across the EU indicated they are
aware of the possibility to register their products for safety purposes297
. In a more
recent consumer survey, the most frequently-reported reasons for not registering a
product were not knowing that registration was possible (42%) and not seeing the
293
CPSC (2017), CPSC Defect Recall Data Carol Cave Deputy Director, Office of Compliance and
Field Operations July 25, 2017, available at: https://www.slideshare.net/USCPSC/cpsc-recall-
effectiveness-workshop-recall-data. (The US CPSC classifies a case as a ‘recall alert’, if the company is
able to contact 95% of affected consumers using direct notification channel, in which case no press release
is required).
294
European Commission, 2021, Behavioural study on strategies to improve the effectiveness of
product recalls.
295
Idem
296
Idem
297
European Commission, 2019, Survey on consumer behaviour and product recalls
effectiveness. Final Report
https://ec.europa.eu/consumers/consumers_safety/safety_products/rapex/alerts/reposit
ory/tips/ Product.Recall.pdf
114
benefit of registration (25%)298
. Likewise, in consumer focus groups, participants
associated product registration with warranty or after-sales support but did not make
a link between registration and safety299
.
 Data protection concerns: While the majority of 40 analysed registration schemes
did envisage the use of registration data for safety notifications or general product
support in their privacy/data notices, in all but two cases, the notices also mentioned
marketing/after-sales communications in addition to safety communications. In
addition, only in a minority of cases consumers were given the choice to opt in or out
from marketing communications300
. In the online industry survey, half of respondents
said that they used customer information received through product registration to
send marketing information.
Use of customer information obtained through product registration (N=23)
Source: Online industry survey
At the same time, concerns about how their personal data will be used is a major
concern for consumers. In the EU focus groups, a key reason participants provided
for not registering their products was that they felt uncomfortable about providing
personal data, which they feared would be used for targeted advertising (profiling)
and other marketing purposes301
. Likewise, in a US survey, 59% of respondents were
concerned about unwanted communication from the company after registering a
product and 79% said they would be more likely to register products if companies
were prohibited from contacting consumers for non-safety-related issues302
.
298
European Commission (2021), Behavioural study on strategies to improve the effectiveness of
product recalls.
299
Idem
300
Idem
301
European Commission, 2021, Behavioural study on strategies to improve the effectiveness of
product recalls.
Schoettle, B., Sivak, M. (2015). Consumer Preferences Regarding Product Registration. (Report No.
UMTRI-2015-26), available at:
91%
52%
48%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
To contact customers in
case of safety issues, e.g.
a product recall
To offer after-sale
service and support
To send information
about new products and
special offers
115
The possibility to use product registration data for marketing purposes has been
limited in some third countries. For instance, the US mandatory product registration
card for childcare products needs to provide clear reassurance to consumers that the
information provided will only be used in case of safety alerts and recalls303
.
Likewise, the UK code of practice recommends that consumers should be asked to
consent to their contact details being recorded exclusively for product safety
purposes, without having to opt in or out of marketing communication304
.
 Effort required: While most products are registered (with the manufacturer) after
the actual purchase, the point of sale has been identified as the key moment during
which consumers can be prompted to register their products305
. In an online
experiment carried out by the European Commission, considerably more respondents
clicked on an invitation to register a product when they saw it at the check-out (45%)
than when they saw it after completing the purchase, either as part of the packaging
(14%) or as a general registration campaign (10%)306
. The amount of information that
needs to be retrieved and/or filled in can also act as a deterrent. In the online
experiment, when the product information was pre-filled (mimicking for instance QR
code scanning), 87% of respondents completed the registration compared to 63%
when they needed to fill in the information themselves307
. Finally, time limits for
registration or the requirement to provide a proof of purchase may make it more
difficult or even impossible to register many products, in particular those received as
gifts or bought second-hand308
.
3.1.1.2. Other sources of customer data not routinely used for recalls
Data allowing to link customers to specific purchases is also routinely collected by
companies through other sources. While customarily intended for marketing promotion,
loyalty programmes and other data (e.g. digital receipts or delivery records) held by
retailers can also enable identification of consumers in case of product recalls. A third
source of consumer data that could be used for recall purposes is the consumer
information provided in the context of online purchases. Purchasing a product directly
from the online seller implies that customer’s contact details are is registered
automatically and the online seller can hence easily use the information provided in the
event of a product recall. When the purchase takes place through an online marketplace,
depending on its business model, the marketplace can either notify consumers directly or
request the sellers to do so in the event of a recall campaign. Similarly, the data held by
http://deepblue.lib.umich.edu/bitstream/handle/2027.42/116020/103219.pdf?sequence=1&isAllowed=y303
S
ee:https://www.cpsc.gov/Regulations-Laws--Standards/Rulemaking/Final-and-Proposed-Rules/Consumer-
Registra-tion-of-Durable-Infant-or-Toddler-Products/
303
See:https://www.cpsc.gov/Regulations-Laws--Standards/Rulemaking/Final-and-Proposed-
Rules/Consumer-Registra-tion-of-Durable-Infant-or-Toddler-Products/
304
UK Department for Business, Energy & Industrial Strategy (BEIS), 2018, Supporting better
product recalls: Code of practice on consumer product safety related recalls and other corrective actions
https://www.bsigroup.com/en-GB/pas7100-supporting-better-product-recalls/
305
Notes from EU expert workshop on recall effectiveness of 23/10/219, p. 3, U.S. Consumer
Product Safety Commission (CPSC), 2017, Transcript of Recall Effectiveness Workshop, 25 July 2017, pp.
21-29. https://www.cpsc.gov/Recall-Effectiveness.
306
European Commission, 2021, Behavioural study on strategies to improve the effectiveness of
product recalls.
307
Idem
308
Idem
116
payment card providers or insurance companies could also be used to inform consumers
about relevant recalls.309
However, economic operators are hesitant about using customers’ information collected
for non-safety purposes in the event of a recall because of a possible legal uncertainty
about the compliance with the General Data Protection Regulation.310
Even though
safety-conscious companies were likely overrepresented in the survey311
, 28% of
respondents in an online survey carried out by the European Commission indicated that
they do not use customer data (collected e.g. through loyalty schemes, online sales,
digital receipts etc.) to contact customers in case there is an issue with their product312
.
Do you use customer data collected for other purposes (e.g. loyalty schemes, online sales, digital
receipts etc.) to contact customers in case there is a safety issue with their product? (N=147)
At the same time, consumers may provide “dirty data” (e.g. fake contact details or an
email address that they do not regularly check) when signing up for loyalty programmes
to avoid receiving marketing information. A possibility for consumers to provide
separate contact details only for the purpose of safety notifications has been put forward
as a good practice to increase the usefulness of loyalty programmes for recall
purposes313
.
3.1.2. No comprehensive sources of public information to consumers
In most cases, it will not be possible to directly reach all consumers affected by a recall.
In such situations, using a multitude of communication channels is recommended as the
best approach to increase the visibility of the recall message and appeal to different
309
Notes from EU expert workshop on recall effectiveness of 23/10/2019, pp. 2-5.
310
Notes from EU expert workshop on recall effectiveness of 23/10/2019, p. 2.
311
Mostly companies interested in product safety might have participated, especially that the survey
was promoted in the Safety Gate weekly reports.
312
European Commission, 2021, Behavioural study on strategies to improve the effectiveness of
product recalls.
313
Notes from EU expert workshop on recall effectiveness of 23/10/2019, p. 4.
117
consumer categories314
. These channels may include company's website, social media
and newsletters, paid announcements in newspapers, TV or radio as well as point-of-sale
information. Social media ads have been identified as particularly cost-effective not only
in targeting specific audience but also in activating the word of mouth, people being able
to share recall information with their contacts who they think may be using a recalled
product315
. At the same time, online communication channels may be less effective in
reaching older and less Internet-savvy consumers316
.
At the moment, there are no rules on public sources of recall information to consumers in
the GPSD. Likewise, in most EU/EEA countries, the recalling company has no legal
obligation to put the recall notice on their website, social media, newsletter or retail
outlet. In response to the consultation on the GPSD Roadmap/Inception Impact
assessment, it was highlighted that some countries demand printed advertising as part of
the recall communication, while in other countries, this is reportedly not the case.
Even though safety-conscious companies were likely overrepresented in the survey317
,
less than one in three respondents in an online survey carried out by the European
Commission indicated that their recall procedure envisages the use of newspaper articles
to encourage recall participation and slightly over a half said the same for company’s
social media318
.
314
Bond, C., Ferraro, C., Luxton, S., & Sands, S. (2010). Social media advertising: An investigation of
consumer perceptions, attitudes, and preferences for engagement. In P. Ballantine, & J. Finsterwalder
(Eds.), Proceedings of the Australian and New Zealand Marketing Academy (ANZMAC) Conference
2010 -'Doing More with Less' (pp. 1 - 7). University of Canterbury.
315
U.S. Consumer Product Safety Commission (CPSC), 2017, Transcript of Recall Effectiveness
Workshop, 25 July 2017, pp. 7-12, 32-4, https://www.cpsc.gov/Recall-Effectiveness.
316
European Commission, 2019, Survey on consumer behaviour and product recalls effectiveness.
Final Report https://ec.europa.eu/consumers/consumers_safety/safety_products/rapex/alerts/repository/tips/
Product.Recall.pdf
317
Mostly companies interested in product safety might have participated, especially that the survey
was promoted in the Safety Gate weekly reports.
318
European Commission, 2021, Behavioural study on strategies to improve the effectiveness of
product recalls.
118
Does your recall procedure envisage the use of one of the following channels to encourage customer
participation in a product recall? (N=21)
In addition, not all authorities engage in spreading recall information on their website
and/or social media (in addition to notifying recalls to the Safety Gate/RAPEX319
).
Identified national central recall portals are listed in the table below320
.
Table: National central recall portals
Country Website
Austria https://www.ages.at/produktwarnungen/ (also a related app)
Bulgaria https://kzp.bg/opasni-stoki-v-bulgaria?f_category=0&f_year=2020&f_search= (mandatory
recalls)
https://kzp.bg/novini/dekatlon-balgariya-eood-predpriema-dobrovolni-merki-vav-vrazka-s-
ustanovena-opasnost-pri-izpolzvane-na-detski-shorti-za-bord-500kid-olaian (voluntary
recalls)
Denmark https://www.sik.dk/farlige-produkter
Estonia Estonia: https://www.ttja.ee/et/tarbijale/ohutus
Finland https://marek.tukes.fi
319
However, not all recalls need to be notified to Safety Gate/RAPEX. As regards products posing a
less than serious risk, notification is encouraged but not mandatory in the case of voluntary measures taken
against products covered by the GPSD and in the case of both voluntary and compulsory measures taken
against products subject to EU harmonised legislation. In addition, Member States are not required to
notify corrective measures in cases where the effects of the product risk cannot go beyond the territory of
the Member State.
320
In addition, specialised portals for motor vehicles exist in Finland
(https://recall.trafi.fi/#vclass=&mark=&model=) and the Netherlands
(https://terugroepregister.rdw.nl/Pages/Terugroepregister.aspx)
119
France https://www.economie.gouv.fr/dgccrf/securite/avis-rappels-produits
Germany www.rueckrufe.de
Hungary https://fogyasztovedelem.kormany.hu/#/veszelyes_termekek_
Iceland https://www.neytendastofa.is/neytendur/solubonn-innkollun-voru/
Ireland https://www.ccpc.ie/consumers/product-safety/product-recalls/
Latvia https://www.ptac.gov.lv/lv/jaunumi?category%5B103%5D=103
Luxembourg https://portail-qualite.public.lu/fr/alertes.html
Malta https://mccaa.org.mt/Section/Content?contentId=4407
Norway https://farligeprodukter.no/
Poland http://publikacje.uokik.gov.pl/hermes3_pub/
Romania https://anpc.ro/categorie/44/retrageri-voluntare-de-produse
Slovenia https://www.gov.si/zbirke/seznami/nevarni-proizvodi/
Sweden https://www.konsumentverket.se/aktuellt/aterkallelser-av-varor/
3.2. Consumers fail to respond to recalls
The extent to which a product recall will be successful and prevent harm depends also on
whether consumers respond, once they become aware that a product they own is being
recalled.
An EU-wide survey on recall effectiveness by the European Commission found that over
a third of consumers (35%) did not react to a recall that was relevant to them: 31%
continued using the product with extra caution, while 4% took no action whatsoever321
.
The corresponding figures in a most recent consumer survey in 10 EU countries were
24% and 13%, respectively322
. Lack of consumer responsiveness was also pointed out by
several MSAs interviewed for the Implementation Study.
Consumers’ propensity to respond to recalls depends on several external factors.
Product characteristics, such as product value, expected lifespan, age and type, all play
a role. Low product value typically decreases consumers' motivation to participate in a
recall323
. The US CPSC data show that return rates increase with the price of the
product324
. Likewise, in a recent survey by the European Commission, self-reported
recall participation rates ranged from 73% for motor vehicles and 63% for furniture to
39% for clothing and footwear and 31% for children’s toys325
. MSAs interviewed for the
Implementation Study also highlighted the importance of product value. In particular,
recalls of low-priced products from Asia, distributed on open-air markets, Asian shops or
online marketplaces, were reported to be very ineffective. On the contrary more
consumers are inclined to return expensive products such as cars326
. Consumers will also
be more motivated to respond to a recall, the newer the product and the longer they
321
European Commission (2019). Survey on consumer behavior and product recalls effectiveness, p.
20, available at:
https://ec.europa.eu/consumers/consumers_safety/safety_products/rapex/alerts/repository/tips/Product.Reca
ll.pdf.
322
European Commission, 2021, Behavioural study on strategies to improve the effectiveness of
product recalls.
323
OECD (2018), Enhancing Product Recall Effectiveness Globally, p. 17.
324
CPSC (25th
July 2017), Recall effectiveness workshop meeting minutes, p. 41.
325
European Commission, 2021, Behavioural study on strategies to improve the effectiveness of
product recalls.
326
Implementation Study.
120
expect it to last.327
Furthermore, consumers tend to underestimate the risks associated
with leisure products (e.g. for sports, recreational vehicles) as they are often linked with
positive emotions328
. In addition to product characteristics, cconsumers’ socio-economic
status also plays a role, disadvantaged and younger consumers being less likely to
respond to a recall.329
However, consumers’ propensity to respond to a recall also depends on the recall internal
characteristics: the clarity and persuasiveness of recall communication, the ease of the
recall process and the attractiveness of the remedies offered.
3.2.1. Unclear recall communication
The way in which recall announcements are formulated and presented can affect
consumers' understanding, perception of risk and motivation to act.
3.2.1.1.Content and layout of recall notices not specified in the EU
Lengthy and complex recall notices may lead to information overload and
disengagement, especially among lower-educated and time-poor consumers. It is
therefore important for recall announcements to use plain and concise language, avoiding
legal terms and jargon. The inclusion of a product picture and further product identifiers
(as well as a clear visual indication of where to find them on the product) can also help
consumers immediately determine if they own the recalled product. Finally, the inclusion
of colour and graphic elements may also help draw consumers’ attention to the recall.330
A number of jurisdictions outside the EU defined the main elements that need to be
included in a recall notice, with the aim of making them clearer and more salient. A
standardised template for recall notices has been set out in Australia and the UK (see
below). In the EU, elements that need to be included in a recall notice are specified in the
Finish and Norwegian guidance documents but in most countries there are no
requirements in this regard.
Australian and UK template for recall announcements
327
European Commission, 2021, Behavioural study on strategies to improve the effectiveness of
product recalls.
328
U.S. Consumer Product Safety Commission (CPSC); XL Associates; Heiden Associates, 2003.
329
Idem, European Commission (2019). Survey on consumer behavior and product recalls
effectiveness, p. 20, available at:
https://ec.europa.eu/consumers/consumers_safety/safety_products/rapex/alerts/repository/tips/Product.Reca
ll.pdf.
330
OECD report, Notes from EU expert workshop on recall effectiveness of 23/10/2019, p.
121
In the focus groups carried out by the European Commission, participants preferred
shorter paragraphs with clear subheadings over a single body of text with the same
information. They also appreciated the inclusion of product picture. The red frame
received more mixed feedback. While some participants considered it as “eye-catching”,
others felt that it was exaggerated and made the notice look like spam email. However,
eye-catching imagery might work better for indirect communication.331
Moreover, stakeholder interviews suggested that defining key elements to be included in
all recall notices would be beneficial for the companies too who would have clarity as to
what information is required and in what format.332
3.2.1.2.Downplaying perceived risk
Consumers’ perception of the likelihood and severity of the risk posed by a recalled
product is another important factor influencing their decision to participate in a recall333
.
It is therefore crucial that recall notices clearly explain the hazard associated with the
recalled product, avoiding euphemisms (such as 'overheating' or ‘thermal event’ instead
of 'fire hazard') as well as any other terms that may decrease consumers' perception of
risk (such as 'voluntary' or 'precautionary')334
. Some third jurisdictions have banned this
kind of terms in their guidance documents (Australia, South Africa, UK).
In focus groups carried out by the European Commission, participants felt that a good
recall message needs to show a sense of urgency, underlining the risk to the consumer.
The phrase “voluntary product recall” in the heading was viewed as to weak, self-evident
331
European Commission, 2021, Behavioural study on strategies to improve the effectiveness of
product recalls.
332
Idem
333
CPSC, XL Associates and Heiden Associates (2003), ‘Recall effectiveness research: a review and
summary of the literature on consumer motivation and behavior’, p. 17 ff, available at:
https://www.cpsc.gov/s3fs-public/RecallEffectiveness.pdf.
334
Idem
122
or confusing because participants were unsure to whom “voluntary” referred or what it
meant. 335
At the same time, the analysis of existing recall announcements showed that more than
half of them (35 out of 55) used terms that could downplay consumers’ perception of
risk, for example:
‘voluntary/precautionary recall’ or ‘voluntary replacement programme’ (27 cases)
‘in rare cases’/’in specific conditions’ (20 cases)
highlighting that there have been no reported injuries (2 cases).336
3.2.2. Burdensome recall procedure
If participating in a recall is costly and does not outweigh the compensation proposed,
this will serve as a disincentive to respond to a recall. The costs associated with recall
participation may include financial costs (e.g. of shipping back the product), opportunity
costs/loss of time, required effort, loss of product use etc. In a recent consumer survey,
recall process taking too much time and effort was the second-top reason for not
responding to a recall (after the product being cheap)337
.
Some of the best practices to minimise consumer effort identified through European
Commission’s mapping of existing recall campaigns and expert workshop include:
the company picking up any bulky items (like washing machines) from consumer’s
home or arranging for in-home repair,
offering a pre-paid postage, if the product needed to be sent back,
allowing customers to return the product in any shop that supplies the product (rather
than the one where they purchased the product) or at a neighbourhood collection
point,
accepting the return without a proof of purchase (which consumers are unlikely . to
keep, especially in the case of cheaper purchases).338
The setting up of a free hotline or other two-way communication mechanism with
consumers to answer any queries about a recall was also identified as a good practice339
.
Furthermore, it was suggested that in some cases, the recalled products could be disposed
of or repaired by consumers themselves instead of being returned to the recalling
company340
.
335
European Commission, 2021, Behavioural study on strategies to improve the effectiveness of
product recalls.
336
Idem.
337
European Commission, 2021, Behavioural study on strategies to improve the effectiveness of product
recalls.
338
Idem
Notes from EU expert workshop on recall effectiveness of 23/10/2019, pp. 8-9.
339
Idem.
340
Notes from EU expert workshop on recall effectiveness of 23/10/2019, pp. 8-9.
123
3.2.3. Insufficient remedies
Consumers’ likelihood to respond to a recall will also depend on the attractiveness of the
compensation offered.341
Tardy or insufficient remedies may reduce consumers’
propensity to act upon a recall.
At EU level, the General Sales Directive guarantees the right to repair, replacement or
refund in case of product non-conformity for a minimum period of 2 years after the
delivery of the goods. However, many products are recalled after the expiry of the
obligatory guarantee period.
The European Commission’s mapping of existing recall campaigns found that in some
cases the description of the remedy in the recall notice was unclear or missing. In
addition, sometimes the remedy was seen as inadequate. This included cases when
consumers were offered a discount to buy a new product from the same company rather
than a proper remedy such as a repair, replacement or refund (detected in two US
campaigns) or when the remedy involved providing consumers with free replacement
parts for self-repair but the instructions were not very user-friendly and the process itself
was time-consuming.342
In consumer focus groups, some participants mentioned that they would like a
compensation for the period during which they cannot use the product.343
3.2.4. Behavioural biases
Behavioural biases may also negatively impact consumers' propensity to return a recalled
product. The box below includes a comprehensive overview of such biases identified by
the OECD.
Consumer behavioural biases applied to product recalls344
Information overload: If recalls contain too much information or
consumers feel overwhelmed with information on recalls, they may
disengage and not take action, especially if they are also time poor. With
the growing number of product recalls in countries, consumers may suffer
from "recall information" fatigue, and may not pay attention to the recall
alerts that are relevant to them.
Framing effects: Consumers are influenced by how information is
presented. Presenting an option in a certain way may induce consumers to
evaluate the choice from a particular reference point. For example,
consumers are less likely to respond to voluntary recalls if the potential
hazards are not clearly stated. They are also less likely to follow
instructions, including the steps they should follow to return the product, if
such instructions are presented in a complex and lengthy message.
341
OECD (2018), Enhancing Product Recall Effectiveness Globally, p. 21.
342
European Commission, 2021, Behavioural study on strategies to improve the effectiveness of
product recalls.
343
European Commission, 2021, Behavioural study on strategies to improve the effectiveness of
product recalls.
344
OECD (2018), Enhancing Product Recall Effectiveness Globally, p. 34.
124
Inertia: With an inherent fear of the unknown, when consumers face
complex products or a bewildering array of choices, they may ignore
possible choices or choose not to choose. Consumers may also rely on
simple “rules of thumb” to avoid change or are guided by the values,
actions and expectations of a particular society or group.
Endowment effect: Consumers often demand much more to give up an
object than they would be willing to pay to acquire it. A consumer’s value
of a product increases when it becomes part of their endowment, so if the
perceived inconvenience associated with returning a recalled product
outweighs the compensation (i.e. return, refund or replacement),
consumers are less likely to return it. This is because naturally humans
tend to be loss averse, even if it is in relation to a recalled product.
Over-optimism: Consumers tend to think that they are more likely to
experience an outcome that is better than the average expected outcome.
This may cause them to miss or ignore warnings, or negative messages.
Consumers may also be more inclined to keep using unsafe products,
particularly if these products have been used for a long time without injury
caused to them or to someone they know.
Time inconsistency: Consumers may make choices that are not consistent
across time periods due to conflicts between short-term urges and long-
term interests.
125
Annex 9: Cooperation activities under GPSD
The GPSD is one component of a broader consumer product safety framework. Apart
from legislative texts (which also include harmonised legislation), it comprises non-
legislative initiatives aimed at advancing the safety of consumer products, via cross-
border and international cooperation, as well as cooperation with stakeholders.
1) Cross-border cooperation
The Consumer Safety Network and RAPEX contact point network
The Consumer Safety Network (CSN) is the expert group gathering product safety
authorities of EU/EEA countries to discuss about ongoing and emerging product safety
issues.
The RAPEX contact point network gathers all national RAPEX contact points to discuss
questions related to the operation of the rapid alert system Safety Gate- RAPEX.
These networks are important fora for authorities to discuss about the challenges they
face and exchange best practices. They also offer the possibility to foster cooperation
with stakeholders: in December 2018, a CSN workshop was organised on the
cooperation between national product safety authorities and consumer organisations.
Similarly, in November 2019, a CSN workshop took place on the cooperation between
authorities and online marketplaces.
Coordinated activities on the safety of products (CASP)
The European Commission helps market surveillance authorities (MSAs) responsible for
the enforcement of non-food consumer product safety across Europe work together, pool
resources and share best practices. It does so by organising coordinated market
surveillance and coordinated activities on the safety of products (CASP). These projects
provide EU/EEA authorities with the financing tools needed to jointly test products,
determine their risks and take the necessary measures against any identified dangerous
products in order to keep all European consumers safe.
CASP projects can focus on the analysis of a single product or a group of products
(product specific activities) or on the exchange of best practices on market surveillance
(horizontal activities).
The MSAs work on resulting recommendations, which are converted into communication
material to be disseminated among economic operators and consumers. At the end of
each project cycle, the CASP Closing Event presents the outcomes of the projects and
launches a European wide communication campaign, contributing to greater awareness
about product safety.
Over the last 15 years, the European Commission has co-financed more than 50 activities
with a total EU budget around 27 million EUR since the start of the activities. Most
coordinated actions have resulted in the identification of a significant number of
126
dangerous products, with non-compliance rates around 20%345
, leading to consequent
notifications in the Safety Gate/RAPEX for 14 categories of products.
2) International cooperation
Cooperation with China
In recent years, about half of the alerts on Safety Gate have given China as the dangerous
product’s country of origin. This is mainly due to the volume of Chinese
products imported into Europe. Since 2006, the European Commission has had formal
cooperation on product safety with the Chinese authorities, which follow up on
dangerous products from China notified in the rapid alert system and report to the
Commission about the follow-up measures they take. Cooperation with this major
producer country helps improve product safety “at source”, so that fewer dangerous
products reach European markets.
Information exchange with Canada
Given the increasingly global nature of supply chains, cooperation is also key with
countries that have the same or similar products on their market. This allows for more
timely identification of emerging product safety issues and enables EU Member States to
better detect dangerous products. Exchanges of information on dangerous consumer
products started with the Canadian authorities in 2019 under the EU-Canada
Comprehensive and Economic Trade Agreement (CETA). The exchanges aim at
helping EU Member States better target their enforcement efforts and identify
emerging product safety risks.
Multilateral cooperation
Given the global nature of most product safety challenges (e.g. the safety of products
sold online, the safety of artificial intelligence and connected products), it is paramount
that the EU position is voiced in international discussions on these subjects to better
protect its consumers. The Commission is actively participating in the OECD and
UNCTAD Working Groups on Consumer Product Safety. The European Commission
has also become an ex-officio board member of the International Consumer Product
Health and Safety Organisation (ICPHSO) board. ICPHSO is an international, not for
profit and neutral forum for product safety stakeholders, which offers regular occasions
for product safety stakeholders to meet, discuss about major global challenges and share
good practices in this field.
3) Cooperation with businesses to go above and beyond minimum legal
requirements
Product Safety Pledge
In June 2018, four online marketplaces (AliExpress, Amazon, eBay and Rakuten France),
signed a Product Safety Pledge. They were later joined by Allegro, CDiscount,
Wish.com, Bol.com and eMAG.
345
Joint Action reports repeatedly indicate that these high rates of non-compliance were not necessarily
representative for the market, as non-random samples were taken and often samples were tested where a
visual inspection had suggested possible deficiencies.
127
The conclusion of this Pledge was facilitated by the European Commission with the
objective of increasing the safety of products sold online by third-party sellers through
online marketplaces. This initiative, which is the first one of its kind in the product safety
area, sets out specific voluntary actions in 12 different areas by online marketplaces that
go beyond what is already established in EU legislation. As part of the Pledge, signatory
online marketplaces have committed to report to the European Commission every six
months on the actions taken to implement the Product Safety Pledge, with the inclusion
of key performance indicators.
EU Product Safety Award
In 2019, the European Commission launched the first EU Product Safety Award,
rewarding businesses that go the extra mile to protect consumers, beyond the minimum
requirements laid down in EU law.
128
Annex 10: Standardisation procedure
The GPSD requires producers to put only safe products on the market. This requirement
may be difficult to apply because of the lack of a common benchmark on what
constitutes a safe product. Standards can respond to this need and they play an important
role in EU product safety law. In the framework of the GPSD, they facilitate market
access and ensure the safety of products.
The European Commission can request the European Standardisation Organisations
(ESOs) to develop standards to support product safety legislation. In the case of the
GPSD, the Commission first needs to issue a Commission decision on safety
requirements to be met by the standard and then issue a standardisation request (mandate)
to the ESOs to develop the standard. When the requested standard is developed, the
Commission checks the compliance of the standard with the safety requirements. If the
outcome is positive, the reference of the standard is published in the Official Journal of
the European Union. Products compliant with the standards referenced in the OJ EU are
presumed to be safe. This helps businesses to know what to comply with and the market
surveillance authorities by providing a benchmark to assess safety of a product.
Ultimately they contribute to safer products on the market for the benefit of consumers.
The procedure is represented in the following scheme:
The overall process can be quite long and burdensome. However, the standardisation
process must strike a balance between speed and the quality of the outcome, thus, of the
standard.
The process under the GPSD includes one step more than the procedure applied in
relation to harmonised standards. The reason is that the harmonisation directives contain
essential safety requirements on which standards can be based. In the case of the GPSD,
its wide coverage calls for specification of the safety requirements for a specific product,
which then serves as a guideline for the work of the European standardisation bodies.
There is room for improvement especially as regards Step 2. The GPSD Study found that
there seems to be room to streamline the process that currently requires the involvement
of two different committees. This appears to duplicate work, and leads to inefficiencies,
as the members of the two committees are not necessarily the same.
STEP 1.
Commission
issues a Decision
to set safety
requirements
•Preliminary work
to plan Decision
•With input from
Member States
(involving the
Consumer Safety
Network), industry,
ESOs and
consumer
organisations
•Vote by MS in
GPSD Committee
•Commission
adopts Decision
STEP 2.
Commission
issues a formal
mandate to ESOs
to develop
standard
•Commission issues
a standardisation
e uest ‚ a -
date to ESOs
•Formal
consultation of
stakeholders
(including ESOs)
•Vote by MS in
Standardisation
Committee
•Adoption & notifi-
cation of request
•ESO
accepts/rejects
STEP 3.
Development of
standard by ESO
compliant with
safety
requirements
•Programming of
the standardi-
sation process
•Drafting of
standard
•Public enquiry
•Refining the draft
standard
•Ratification and
publication of the
standard
•Submission of
references to
Commission
STEP 4.
Commission
issues a Decision
about the
referencing of
the standard
•Verification that
standard is in line
with safety
requirements and
vote by MS in
GPSD Committee
•Commission
adopts Decision
recognising
standard /
reference to stan-
dard published in
the Official Journal
of the EU
Lack of
common EU
benchmark
on what
constitutes
„safe“ for a
specific
product, for
which certain
risks have
been
established
Common EU
benchmark
(EN
Standard)
established
and applied
by
producers
Annex 11: Summary of replies to the open public consultation
Q1 In your view, to what extent are current EU safety rules for non-food consumer
products covered by the GPSD adequate to protect consumers?
A large majority of respondents expressed that current EU safety rules for non-food
consumer products covered by the GPSD could be improved in specific areas to be more
adequate to protect consumers (71%). Nearly one in four respondents held that the
current rules were fully adequate, whereas only a small minority considered them not to
be adequate at all (6%), see the following figure.
Total no. of respondents: 214, Single-choice question
Q2 Are you aware of any problems related to the implementation of EU safety rules
for consumer products covered by the GPSD?
When asked about problems related to the implementation of safety rules for products
covered by the GPSD, respondents most commonly expressed that rules were not adapted
to online trade (39%) and that the rules were not appropriately enforced (39%). More
than a third also considered the rules not to be adapted to new technologies (36%) and
perceived legal definitions as not sufficiently clear or outdated (35%). Slightly less than a
third of respondents (30%) reported that roles and obligations of different economic
operators were not appropriately defined and that there were difficulties for consumers to
report unsafe products. Lastly, approximately a fifth of respondents regarded as
problematic that there were no specific requirements for product recalls (22%) or listed
other issues (23%) (see the following figure).
Fully
adequate
23%
Could be
improved in
specific areas
71%
Not adequate
at all
6%
130
Total no. of respondents: 205, Multiple-choice question
In their comments, respondents detailed their views regarding problems with the
implementation of the GPSD. In line with ‘online trade’ being a major concern to the
safety of products, many respondents referred to problems in this area, often specifically
referring to online trade with third countries. In this context, the “current lack of
responsibility and acknowledgment of online marketplaces' role in the supply chain” was
mentioned by a large number of respondents. A related issue that was frequently
indicated concerned customs controls of consignments from non-EU countries to
consumers. Often, comments were framed as suggestions for improvement, i.e. it was
suggested to ensure better customs controls of these consignments.
Many stakeholders considered that the GPSD was not properly enforced or implemented
effectively, with a typical example being a statement that the “current framework is not
implemented effectively (over 500 MSAs in Europe with no minimum standard for recall
or takedown notices)”. Respondents referred to differences in risk assessment, and
considered resources of market surveillance authorities to be insufficient, including
regarding testing. Finally, a number of respondents raised specific issues, such as:
 The lack of data on injuries and accidents, and the need that accident and incident
alert systems must be adjusted to work in practice;
 Limitations in scope of the directive, such as the lack of services covered
(including financial services and online gambling services with a high addictive
potential), and counterfeiting;
 Limitations of the concept of “safe product”, which was considered to not always
meaning that the product is safe for specific vulnerable groups;
 The absence of specific rules on the safety of child appealing products;
 Lack of adaptation to new technologies, including the lack of coverage of
software as “product” (see also next question).
23%
22%
30%
30%
35%
36%
39%
39%
OTHER
THERE ARE NO SPECIFIC REQUIREMENTS FOR
PRODUCT RECALLS
DIFFICULTIES FOR CONSUMERS TO REPORT UNSAFE
PRODUCTS
THE ROLES AND OBLIGATIONS OF DIFFERENT
ECONOMIC OPERATORS ARE NOT APPROPRIATELY…
LEGAL DEFINITIONS SUCH AS PRODUCT , SAFE
PRODUCT OR PLACING ON THE MARKET ARE…
RULES ARE NOT ADAPTED TO NEW TECHNOLOGY
PRODUCTS (E.G. CONNECTED DEVICES)
PRODUCT SAFETY RULES ARE NOT APPROPRIATELY
ENFORCED
RULES ARE NOT ADAPTED TO ONLINE TRADE
131
Q3 Do you think that the safety of products involving new technologies is adequately
regulated?
Almost half the respondents considered the safety of products involving new technologies
to be not adequately regulated (47%), with only 18 % stating the opposite. The other 35%
did not know (see the following figure).
Total no. of respondents: 227, Single-choice question
Q4 When incorporated into a physical product, software can malfunction and cause
a safety issue. When considering whether a product is safe, should the definition of a
product in the GPSD specifically encompass also the software incorporated into it?
When asked whether the definition of a product in the GPSD should specifically
encompass software incorporated into the product, the majority of respondents agreed,
even in case the software is downloaded after the product has been sold (56%).
About a quarter of respondents considered that only software already installed into the
product when sold should be included. Only a small minority answered that the definition
should not encompass software (6%) or preferred an “other” option (12%), see figure
below.
Yes
18%
No
47%
Don't know
35%
132
Total no. of respondents: 211
Respondents could provide comments, and again explained their positions. Those that
argued that all software should be covered (even if downloaded after the purchase of a
product) typically provided one or more of the following arguments:
 Safe software is as essential as safe hardware. One cannot function without the
other;
 Software can change the properties and hence the safety of a device. The GPSD
must have a safety net function here;
 Modern goods are made from software and hardware, and both of them can cause
safety issue. It is important that both are covered;
 Encompassing software into the product definition would eliminate ambiguities.
In contrast, there were also several respondents that did not see any need for changes to
the GPSD in this respect, arguing:
 Physical product can cause harm and should be designed to be safe, not software;
 The definition of products currently used in the GPSD is up to date and broad and
flexible enough to reflect new developments and challenges;
 A risk-based approach should be taken with new technologies. If there are gaps,
targeted legislation is the most effective way to address risks.
Finally, comments considered how responsibility should be allocated in case software
would cause a product to be unsafe. There was no consensus in this respect. Rather, the
answers reflected the different perspectives of the respondents: Those that saw a strong
role of the manufacturer argued that the safety of the product must take into account
software designed to operate the product, even if the software is installed afterwards, and
that a product should have safety functions that minimise the risk of malfunction of added
software. In contrast, others suggested that the duties of the manufacturer of a device
should not cover safety risks posed by a standalone software installed later, over which
the manufacturer has no control. Some respondents also argued that if users installed
software, manufacturers or distributors should not be responsible for its safety unless they
had instructed or authorised this.
12%
6%
26%
56%
Other
No
Yes, but only when software is already installed
into the product when sold
Yes, also when software is downloaded into the
device after it has been sold
133
Q5 How important do you think it is that products that could be modified via
software updates/downloads or machine learning are required to remain safe
throughout their lifetime?
Almost all respondents considered a requirement for products that could be modified via
software updates/download or machine learning to remain safe to be very important
(72%) or rather important (24%). A mere 3% regarded the requirement as rather
unimportant, while none of the respondents considered it not to be important at all (see
the following figure).
Total no. of respondents: 217, Single-choice question
Q6 Products incorporating AI applications can evolve via machine learning and
other techniques, even after they have been acquired by consumers, potentially
posing safety risks. In your opinion, at which moment of the lifecycle of the product
should manufacturers have safety obligations?
A clear majority of respondents favoured safety obligations for manufacturers of products
incorporating AI applications at the design stage and also during the lifecycle of the
product (75%), whereas only 9% of respondents expressed that the obligations should be
limited to the design stage. 16% preferred an “other” solution (see the following figure).
Total no. of respondents: 213, Single-choice question
Very important Rather
important
Rather
unimportant
Not at all
important
72%
24%
3% 0%
75%
9% 16%
OBLIGATIONS AT THE DESIGN STAGE
AND ALSO DURING THE LIFECYCLE OF
THE PRODUCT
OBLIGATIONS ONLY AT THE DESIGN
STAGE
OTHER
134
In their comments, respondents mostly reiterated their views without providing additional
details. Some respondents that agreed to safety obligations of manufacturers during the
lifecycle of the product, elaborated on limitations to this principle. According to their
view the level of safety during the lifecycle should be compared to the level of risk with
human control (not “zero risk”), should exclude “substantial modifications” or be limited
to the “foreseeable period of use of the product”. Finally, there were also respondents that
saw a need for regulating AI horizontally, and not in the GPSD.
Q7 Have you experienced any product safety incident within the last 5 years?
A clear majority of respondents did not experience any product safety incident within the
last 5 years (65%), with the share of affirmative answers being 35%, see the following
figure).
Total no. of respondents: 175, Single-choice question
If yes, how did you buy the product?
Among the 62 respondents that answered yes to Q7 the largest group had bought the
product online from a seller based in the EU (27%), followed by each 24% of
respondents that had acquired the product from a physical shop or online from a seller
based outside the EU. About one fifth of respondents stated that they had obtained the
product online but were unsure where the seller was based and 5% gave no answer (see
the following figure).
Multiple-choice question
No
65%
Yes
35%
5%
19%
24%
24%
27%
No answer
Online but I am not sure where the seller was
based
Online from a seller based outside the EU
From a physical shop
Online from a seller based in the EU
135
Q8 Have you experienced any lack of information linked to safety when buying
products online?
A small majority of the respondents reported that they had experienced a lack of
information linked to safety when buying products online (51%), while the other half did
not share this experience (49%), see the figure below.
Total no. of respondents: 185, Single-choice question
If yes, what was this lack of information linked to?
Of those 94 respondents that answered yes to Q8 a clear majority considered the lack of
information to be linked to missing contact of the producer of a product bought online to
report a safety incident to (65%) and/or to not receiving enough warnings or instructions
concerning the product sold online (63%). Slightly less than half of the respondents
perceived that there was no place to report a product safety incident on the website as
problematic in this respect (44%), while a quarter of the respondents considered other
factors to be relevant. Another 1% gave no answer (see the following figure).
Multiple-choice question
Q9 Online marketplaces enable companies to sell to EU consumers but, according to
EU rules, they do not have direct legal obligations for the safety of products hosted
No
49%
Yes
51%
1%
26%
44%
63%
65%
No answer
Other
No place to report a product safety incident on the
website
Not enough warnings or instructions concerning the
product sold online
No information available to contact the producer of
the product bought online to report a safety incident
136
on their platform by sellers. Are you aware of any problems this regime would bring
about?
The majority of respondents expressed that they were aware of problems associated with
online marketplaces having no direct legal obligations for the safety of products hosted
on their platform by sellers (53%). However, almost half of the respondents indicated the
opposite (47%) (see the following figure).
Total no. of respondents: 209, Single-choice question
Q10 What should be the role of online marketplaces as regards the safety of
products offered on their website?
When asked about the role that online marketplaces should play regarding the safety of
products offered on their websites, the most commonly supported notions were that they
should remove dangerous products listed on their website when notified (77%), that
online marketplaces should prevent the appearance of dangerous products, including their
reappearance once they have been removed (66%) and that they should inform sellers of
their obligation to comply with EU rules on products (64%). More than half of the
respondents agreed that online marketplaces should inform consumers when a dangerous
product has been removed from the marketplace (55%). A slightly lower number of
respondents thought that online marketplaces should do a cursory check on all products
offered on their website to identify products that likely do not comply with safety rules
(42%). Several respondents also indicated the option “other” (14%), see the figure below.
No
47%
Yes
53%
137
Total no. of respondents: 221, Multiple-choice question
Nearly all detailed comments provided by respondents concerned suggestions for
additional obligations for online marketplaces. In many cases, respondents considered
that due to the active role of the marketplaces in facilitating transactions, platforms
should have the same responsibilities as other importers/distributors/traders (selling
online or in brick-and-mortar shops). Respondents suggested a large variety of possible
additional obligations for online marketplaces, including:
 Taking reasonable efforts/work more proactively to prevent appearance of
dangerous products;
 Using technology (such as AI) to swiftly identify and delist unsafe products;
 Checking all products offered in order to identify non-compliant/recalled
products;
 Verifying the presence of the CE mark and the absence of obvious safety
concerns;
 Implementing testing requirements, especially for products falling under the
WEEE, Battery and Packaging Directives;
 Keeping a notification form available for suspected cases of unsafe products;
 Informing market surveillance authorities immediately of unsafe products
identified;
 Cooperating with enforcement authorities;
 Informing consumers who have previously bought a product that was later taken
down following a valid request for takedown;
14%
42%
55%
64%
66%
77%
Other
Online marketplaces should do a cursory check on all products
offered on their website to identify products that likely do not
comply with safety rules
Online marketplaces should inform consumers when a dangerous
product has been removed from the marketplace
Online marketplaces should inform sellers of their obligation to
comply with EU rules on products
Online marketplaces should prevent the appearance of dangerous
products, including the reappearance of dangerous products'
listings once they have been removed
Online marketplaces should remove dangerous products listed on
their websites when notified to them
138
 Not sending unsafe products to consumers when they have already been ordered
but are not sent yet;
 Recalling unsafe products from consumers and destroying the stored products;
 Identifying repeat offenders which would prevent them to re-list dangerous
products on the marketplace;
 Being liable for products they sell.
A minority of respondents suggested, however, that responsibilities and obligations
imposed on online marketplaces should be proportionate (or effective and feasible) and
take into account the size and scope of the provider in question. They also considered that
obligations should not handicap SMEs competing against bigger and more established
companies who are better placed to overcome related regulatory burdens.
Finally, several respondents suggested that the scope of checks to be conducted by
marketplaces should include not only unsafe products, but also counterfeit products in
general, and life animals/pets sold on platforms, to safeguard that only registered animals
are offered by traceable sellers (with a specific reference to illegal puppy trade).
Q11 What are the main challenges for enforcement?
When asked about the main challenges for enforcement half of the respondents
considered as problematic that Member States’ authorities did not have enough resources
(49%), followed by the difficulty of taking enforcement actions against economic
operators outside the EU (46%). Other challenges included that not enough control
checks are carried out, including by customs (29%), that there is not enough cooperation
between market surveillance authorities in the EU (27%), and that these authorities assess
product risk differently (19%). All other answer items were indicated by 18% or less of
respondents.
139
Total no. of respondents: 207, Multiple-choice question with maximum 3 choices
In their comments, respondents referred to the above listed challenges in detail. It was
suggested that the listed enforcement issues would need to be addressed through GPSD
reform and through making available more resources at national level for enforcement
and better controls, including customs controls. Several stakeholders referred to
Regulation (EU) 2019/1020 on market surveillance (which is applicable in the
harmonised product sectors), indicating that it provided relevant means of enforcement.
Other suggestions included that:
 National authorities needed more resources or information on products coming
from outside the EU to ensure effective checks;
 The GPSD should be converted into a regulation in order to rule out national
differences in implementation;
 Powers should be specified and harmonised, and enforcement be improved
through joint action by several Member States.
Q12 Do you think that products covered by the GPSD should only be placed on the
EU market if there is an economic operator established in the EU who is responsible
for product safety purposes?
A large majority of respondents considered that products covered by the GPSD should
only be placed on the EU market if there is an economic operator established in the EU
18%
12%
14%
18%
18%
19%
27%
29%
46%
49%
Other
Dangerous products are difficult to trace
Fragmentation of the market surveillance legislation
between harmonised and non-harmonised products
Me e States autho ities a ot take effe ti e a tio s
online ( e.g. mystery shopping, restrict access to the…
Me e States autho ities la k spe ifi po e s: they
cannot impose efficient sanctions on economic…
Me e States autho ities assess p odu t isks
differently
Not enough cooperation among market surveillance
authorities in the EU
Not enough control checks carried out, including by
customs
Enforcement actions against economic operators outside
the EU are difficult
Me e States autho ities do ot ha e e ough
resources
140
responsible for product safety purposes (70%), with the other 30% expressed the opposite
view (see the following figure).
Total no. of respondents: 186, Single-choice question
Q13 Are you aware of any issue where additional competences of the European
Commission for the enforcement of product safety rules could improve the safety of
consumers?
Close to two thirds of respondents indicated that they were not aware of issues where
additional enforcement competences of the European Commission could improve the
safety of consumers (64%). However, more than a third of the respondents suggested the
opposite (36%), see the figure below.
Total no. of respondents: 170, Single-choice question
Q14 Should the system of product traceability be reinforced in the GPSD so that
products can be better traced if there is a safety issue?
A large majority of respondents agreed that the system of product traceability should be
reinforced in the GPSD (82%), while only 18% of respondents did not regard this as
necessary (see the following figure).
No
30%
Yes
70%
No
64%
Yes
36%
141
Total no. of respondents: 171, Single-choice question
Q15 Do you experience problems with the divergence of rules between harmonised
and non-harmonised products?
Almost one in three respondents reported having experienced problems with the
divergence of rules between harmonised and non-harmonised products (30%), while 23%
stated the opposite. Almost half of the respondents expressed that they did not know
(47%) (see the following figure).
Total no. of respondents: 178, Single-choice question
Q16 Products which resemble foodstuff, while not being such, have a separate
regime (Council Directive 87/357/EEC). This has given rise to different
interpretations on whether such products are dangerous in itself or not. Should
these products keep having a separated regime or be incorporated into the general
product safety legal instrument?
A large majority of respondents expressed that products which resemble foodstuff should
be incorporated into the general product safety legal instrument (69%), whereas the other
31% felt that the regime should remain separate.
No
18%
Yes
82%
Yes
30%
No
23%
Don't know
47%
142
Total no. of respondents: 155, Single-choice question
No, it should
remain
separate
31%
Yes, it should
be
incorporated
69%
143
Annex 12: Stakeholders opinions on the benefits of the
different options
The survey conducted in the context of the GPSD Study shows the following
stakeholders’ views on the additional benefits for businesses resulting from the different
options:
Additional benefits for businesses resulting from the implementation of Option 1
Additional benefits for businesses resulting from the implementation of Option 2
144
Additional benefits for businesses resulting from the implementation of Option 3
Additional benefits for businesses resulting from the implementation of Option 4
145
1 2 3 4 5
Increased businessrevenue (e.g.due to increased
reputation/brand value)
Lower operational risk for businesses
Easier compliance with product safety requirementsfor
SMEs
Better supply chain management dueto improved
traceability of products
Reduced legal complexity
Better information on unsafe products/ measurestaken
by authoritiesprovided through Safety Gate /RAPEX
Greater legal certainty
No change in Very significant
benefitsat all additional benefits
Other stakeholders
Authorities
Companies/ Businessassociations
146
Annex 13: Minutes from the EU workshops addressing the sale
of illegal goods online
The workshops were co-organised by DG CNECT and DG JUST, 8, 10, 13 and 17 July
2020, as part of a broader engagement with stakeholders and evidence collection strategy
for the Digital Services Act package as well as the revision of the General Product Safety
Directive.
The objective of the workshops was to gather up-to-date information on the state of play
concerning the main challenges in addressing the sale of illegal goods online. It focused
in particular at measures and good practices from marketplaces and the cooperation with
authorities and responsible third parties. Panellists and participants – which included
online marketplaces, retail associations, consumer organisations, national market
surveillance authorities as well as representatives from the European Commission - were
invited to share their experiences and engage in a discussion on potential new policy and
regulatory measures.
The event was made of four separate online sessions:
Session 1: Sellers and products identification mechanisms, 8 July 2020 - The first session
was focused on the information online marketplaces are currently gathering on their
sellers. Online marketplaces started with a short overview of practices in identifying their
business sellers and product listings on their platforms. Most online marketplaces
specified that business sellers are required to submit background information (e.g.
company name, VAT number, address, etc.) before being admitted to sell.
Overall, all participants agreed on the importance of having transparency as regard
business traders. Some participants highlighted that more should be done in this context,
especially when it comes to sellers established outside the EU and therefore not always
covered by EU rules. Some stakeholders considered that more cooperation with
authorities in Member States could also help identifying rogue sellers.
Session 2: How to tackle dangerous goods and product safety issues online: notice and
action procedures and the role of the Safety Gate/RAPEX - The first part of this session
concerned best practices on notice and action procedures to tackle dangerous goods,
including notices from authorities, consumer associations, consumers and other actors.
Generally, all participants agreed that a harmonised notice and action procedure would
facilitate the fight against dangerous products online. Some participants highlighted that
often notices are not accurate enough and online marketplaces have difficulties in
identifying the dangerous products notified. In this regard, many participants called for a
minimum information requirement for notices. Online marketplaces also stated that filters
are not entirely reliable and that such tools should always be accompanied by human
review and notice and action mechanisms.
The second part of the session concerned Safety Gate/RAPEX. In this regard, a number
of investigations carried out by consumer organisations, retail associations and market
surveillance authorities were also presented, with results on the number of dangerous
products available online raising clear concerns. Marketplaces are taking some action,
such as periodically checking Safety Gate/RAPEX (as they have committed in the
Product Safety Pledge). Some participants pointed out, the information in the Safety Gate
147
only shows only part of the issue and more needs to be done in this regard. Some
remedies were proposed by national authorities, such as establishing an obligation to
cooperation with market surveillance and custom authorities. Some participants also
suggested to have an API interface to Safety Gate/RAPEX which would then be linked to
online marketplaces and allow them and consumers to have real-time information on
product safety.
Session 3: What other measures and challenges for keeping consumers safe from
dangerous goods online? – The session focused on other preventive measures that
marketplaces can take to ensure that no dangerous product is placed on the market. Three
main aspects were mentioned by participants. First, the importance of data, that in many
cases is not provided by the seller, making enforcement very difficult. Second, online
sales and product safety are global issues, therefore international cooperation is key to
address these challenges. Thirdly, many participants mentioned the issue around
traceability, and how it needs to be enhanced so dangerous products sold online can be
correctly identified and corrective measures can be enforced by both platforms and
authorities. The challenge of reappearance of dangerous products already removed was
also addressed, although not specific measures or solutions were mentioned by
participants.
Session 4: Consumer law and online marketplaces, 17 July 2020 -The main focus of this
session was to address content that is illegal because it constitutes a violation of
applicable EU consumer law.
The session started with a short presentation held by DG JUST on the relevance of EU
consumer law for a) online marketplaces regarding their own activities and content; b) the
business users of online marketplaces; and c) online marketplaces in their capacity as
hosts of their business users.
The discussion then zoomed in on third-party content and the measures that online
marketplaces are taking to prevent activities that violate applicable EU consumer law.
Online marketplaces specified that their objective is to create trust on the platform, both
for consumers and sellers. They further stated that sellers are in charge of their own
compliance, but that they are responsible to give them the means to be able to be
compliant with EU law.
Some participants flagged that the main problem with EU consumer law is the lack of
resources and enforcement.
Cooperation was also mentioned by many participants as being the key to ensure a
coherent enforcement of EU consumer law. According to many participants, all the actors
in the supply chain should work together to raise awareness around consumer rules.
148
Annex 14: Minutes from the EU Workshop on strategies to
maximise the effectiveness of product recalls
On 23 October 2019, a workshop has been organised by the European Commission to
discuss the strategies to maximise the effectiveness of product recalls. The participants of
the workshop included regulators from around the world, representatives of international
organisations (OECD, UNCTAD), consumer organisations, industry and academics.
The aim of the workshop was to take stock of existing market practices and regulatory
approaches, and identify possible new avenues to maximise recall effectiveness. As a
next-step, the most promising ideas could be tested through an EU behavioural study, the
results of which should also be of relevance for other stakeholders and jurisdictions.
The workshop was divided into three thematic sessions, focusing on i) Strategies to
facilitate direct consumer contact, ii) Strategies to increase consumer response to recalls
and iii) Roles and responsibilities in the recall process. The last part of the workshop was
limited to regulators in order to discuss possible next steps.
1. Session 1: Strategies to facilitate direct consumer contact
The objective of this session was to:
 get a comparative overview of strengths and weaknesses of various methods of
identifying the owners of recalled products, and
 brainstorm on best ways to address the barriers that prevent consumers from
sharing their contact details for safety notification purposes.
PRESENTATIONS
UK Electrical Safety First talked about strategies to improve product registration.
According to Electrical Safety First research, only 1 in 3 people register electric products,
because they 1) fear unwanted marketing communications 2) think it takes too much
effort and time 3) do not see the benefit. The charity has launched several awareness-
raising campaigns to encourage registration of electrical devices and domestic
appliances346
. Further avenues to encourage product registration include 1) separating
marketing from safety notifications, 2) standardising and simplifying registration
material, 3) point-of-sale registration and 4) technological solutions (barcodes/QR/RFID,
connected devices).
Decathlon explained that the company uses different strategies to communicate recalls to
its customers – from general displays in stores and on the website to personalised letters
and e-mails. Phone calls and SMS are also used in case of small-scale recalls. The
experience shows that direct communication is the most effective strategy.
The main source of consumer data is a voluntary loyalty programme, which varies across
different countries, and allows Decathlon to contact affected consumers directly. The
346
Among others: https://www.electricalsafetyfirst.org.uk/product-recalls/product-registration/,
https://www.electricalsafetyfirst.org.uk/what-we-do/campaigns/nan-knows-best/,
http://www.whitegoodsafety.com/
149
company also offers voluntary registration at the moment of purchase, but for the time
being this possibility is limited to bicycles, which are the highest-risk product.
The company noted that even if they do not post recall announcements on social media,
consumers may do it themselves and even share incomplete and incorrect information.
Therefore, the best option is to share recall information directly, or through sports
community groups, which additionally encourages word-of-mouth.
The Finnish Safety and Chemicals Agency (Tukes) focused on how economic operators
should use loyalty cards data for recalls in the light of the EU personal data protection
rules.
They started by recalling an incident of toxic olives causing botulism that occurred in
Finland in 2011. In addition to vast media attention, most of the affected consumers were
contacted directly due to accessibility to personal data in customer loyalty registers.
Nowadays Tukes's guidance on product recalls explicitly recommends that if a company
has a comprehensive customer register available, contacting people directly is the most
effective way to notify them about a product recall. Also, according to the European
Commission's survey, consumers would prefer direct personal communication regarding
recalls.
However, following the entry into force of the EU General Data Protection Regulation
(GDPR) in May 2018, economic operators have become much more cautious about using
loyalty programmes data to reach out to consumers in case of a recall. In an ideal
scenario, recalls should be defined as the initial purpose of personal data processing in
customer registers. This way, when consumers sign up for a loyalty scheme or make a
purchase online, they would explicitly agree to be contacted in case of safety issues.
However, even in the absence of such explicit consent, the use of customers' personal
data can be justified. To make sure that companies do not to delay safety measures
because of the GDPR concerns, Tukes, the Office of Data Protection Ombudsman, and
the Finnish Commerce Federation are currently preparing a memorandum on the issue.
However, some practical questions still remain. This is especially important as GDPR is
part of the European legislation, and interpretations need to be harmonised throughout the
EU.
The representatives of the Commission mentioned that the Commission is looking into
developing a hands-on guidance on how personal data should be handled in the recall
process. She confirmed that the recommended default approach, in line with the General
Data Protection Regulation, is for businesses to include in their privacy policy the
possibility to contact customers to inform them about safety notifications.
GROUP DISCUSSIONS
GROUP 1: Increasing product registration rates
There was a general consensus that the best way to reach out to consumers with recall
information is to contact them directly and that increasing product registration rates
should therefore be encouraged.
The participants explored major barriers to registration both from a consumer and a
business perspective, and suggested possible solutions.
150
Most consumers simply do not understand the interest of registering their products, and
do not make the link between registration and safety. This is especially the case for lower
value products (Australian representative mentioned the cut-off threshold of 50 AU$). It
is therefore important to clearly communicate about the benefits of product registration.
There is also a widespread lack of trust that the data will not be used for marketing
communication. The US product registration card for childcare articles provides a
statement that the information will only be used in case of safety alerts and recalls.
However, it was mentioned that even that could be further strengthened (e.g. by adding
that any other use would be subject to liability).
At the same time, consumers are not a heterogeneous group; they have different risk
perceptions and attitudes. The more risk-averse are also more likely to register their
products. The challenge is to reach out to the rest. Engaging more vulnerable consumer
groups is a particular challenge. Local messengers can be helpful in this regard (for
instance churches in New Zealand).
The participants agreed that the registration process should become as seamless as
possible. One option to explore is standardising and simplifying registration material to
only capture the minimum necessary information. Technological solutions - such as
mobile QR code scanning (already very popular in China and Southeast Asia) or even a
dedicated product registration app – would also minimise the effort required from the
consumer.
The timing of registration matters too. It was noted that the key moment during which
consumers can be prompted to register their products is at the point-of-sale. The CPSC
evaluation has shown that unless the registration card is filled out (and assisted)
immediately after the purchase, there is little chance that it will ever be. While point-of-
sale registration may be difficult in physical stores with high customer flow, this is quite
straightforward in the case of online purchases. The idea of encouraging registration upon
the receipt of the product was also put forward. The UK mentioned their pilot project in
cooperation with Whirlpool, testing registration at delivery.
As for the business side, it was noted that economic operators may be reticent to highlight
the link between registration and safety so as to not to suggest even the slightest
possibility of their product not being safe. It may be therefore more appropriate and
effective for the message to come from the government agency.
It was also mentioned that if economic operators were obliged to offer the possibility of
product registration to consumers, especially at the point/moment of sale, this would
greatly increase registration rates. The role of retailers in the process should be carefully
analysed: while they are not the responsible party, they could greatly facilitate
communication.
Beyond product registration, it was noted that customers' data is generated also during
other touch points of the consumer journey: most products are paid by card; big items are
usually delivered and complex items installed. Yet, this data is not being used for safety
purposes because of concerns about personal data protection.
Finally, it was mentioned that being able to reach the affected consumers directly does
not guarantee that he or she will react. In the case of New Zealand Takata recall, even
very intense direct contact (up to 6 letters) did not guarantee a 100% response (despite
free repair and very serious risk). New Zealand is therefore shifting from the carrot to
151
stick approach (“if you fail to respond to the recall, your vehicle will be not pass the
compulsory periodic inspection").
GROUP 2: Other methods of identifying affected consumers
To start with, participants were asked about their personal experiences when subscribing
to loyalty programs (to which programmes and why they were more likely to subscribe
to) and how familiar they were with the concept of "dirty data" (which can be for instance
fake contact details or an email address that is not checked by the user).
There was a general agreement that this depends on the trust towards the brand or
website. Some said they prefer to create specific email addresses for this very purpose,
fearing to be overwhelmed with marketing emails. This is in line with different surveys
showing that consumers all over Europe tend to provide “dirty data” for loyalty
programme registration: almost a third of people tend to give an email address that they
do not regularly check.
The benefits of providing accurate data when signing up to such programmes are not
really obvious to consumers (the main motivation being financial), while they see the
immediate adverse effects of appearing on more marketing listings and receiving more
advertising. This issue is important, as it might undermine the effectiveness of direct
safety notifications.
A good practice could be for consumers enrolling in a loyalty programme to have the
possibility to provide a specific email address/phone number to receive safety
notifications only (similarly to what is suggested for product registration).
The need for the consumer to be “pulled” rather than “pushed” was also stressed when
discussing how best to encourage consumers to share their personal data for safety
purposes. Most consumers do not see the benefit of giving out their data. Communicating
on these issues appears essential. Participants stressed that even though loyalty
programmes can be helpful, they should not be the only means of generating customers'
data for safety purposes.
The potential role of financial institutions in contacting affected consumers was also
discussed. In some cases, it is indeed possible for bank service providers to identify
consumers who have purchased a recalled product when the payment has been done with
a credit/debit card. Some participants shared experiences with such use of financial
institutions in recalls:
- An extra benefit of banks directly communicating recall information to consumers
is that consumers tend to have more trust in these messages.
- Handling of personal data is one of consumers' major concerns when third parties
are involved in the recall process.
- The emergence of “FinTech” (financial technology companies) should also be
considered when looking at ways to facilitate consumer identification, taking into
account that customers can have virtual credit cards for purchases online, or can
have multiple credit cards per account.
Participants also discussed how online marketplaces can play a facilitating role, taking
advantage of the channels and systems they have already put in place to communicate
with both consumers and sellers. Two different models exist: some platforms ask the
152
sellers of recalled products to notify affected consumers, others send the recall message
themselves.
The benefits linked to connected products were also stressed. When a connected product
itself is subject to a recall, this technology can be used to warn consumers or, if they fail
to act, switch off the product or reduce its performance. Connected products may also be
able to “monitor” other products. For instance, some smartphones can identify the
charger that is plugged to them and could alert consumers in case this charger is recalled.
In the food sector, a connected fridge could identify recalled items placed into it.
It was also emphasised that consumers may sometimes not react to a recall alert, even
when properly informed. In such situations, different solutions could be explored:
- the role of voice assistants to remind consumers of the need to act;
- the possibility to remotely switch off the unsafe product;
- asking consumers who failed to react to a recall whether they have stopped using
the product or have disposed of it (such information could be asked e.g. in recall
notifications sent by online marketplaces).
Finally, the issue of products that are no longer in the hands of the initial buyer (e.g. gifts,
second-hand products) was discussed. A good practice could for recall notification to
include into the instruction “If you have given or sold this product to another person,
please forward them this message for their safety”.
2. Session 2: Strategies to increase consumers' response to recalls
The aim of this session was to explore how recall communication and procedure can be
improved to enhance consumer response.
PRESENTATIONS
UK Office for Product Safety and Standards presented the interim results of the UK
government research into consumer attitudes towards product safety. The study revealed
that only 17% of consumers consider product safety a priority when purchasing a
product, and 23% had registered a product they recently purchased. Overall, the results
highlight that consumers assume products are regulated and safe, and are therefore not
motivated to register their products. The study has also identified different consumer
profiles, indicating that the communication should be adapted to reach each segment:
"aware", "busy families", "second-hand shoppers", "less connected", "quick to repair",
"buy to last", "latest quality".
Based on the research results and recent Whirlpool tumble dryer recall, some of the most
important elements to include in a recall notice include word "safety" in the title,
photograph of the product, clear description of the hazard and instructions on what to do
as well as highlighting the incentive (e.g. free replacement). Finally, methods for
communicating to hard-to-reach consumers should be considered.
In March 2018, the UK government adopted a Guidance on Recalls (PAS 7100), which
instructs manufacturers, importers and distributors on how to effectively plan, manage
and monitor product recalls. The second part aims at regulators and their role in
monitoring incidents and supporting businesses. The guidance will be reviewed and
updated, based on learnings from different product recalls.
153
The Australian Competition and Consumer Commission presented the findings from
Australian research into the Takata airbag recall. The Takata recall is the biggest
compulsory recall in Australia's history, where 1 in 4 vehicles was affected by defective
airbag which had caused 26 deaths worldwide. A communication campaign “A faulty
airbag can kill you” was launched to alert consumers of the hazard. The research based
on this recall case study focused on identifying the link between recall effectiveness and
1) different communication and engagement strategies, 2) consumers' socio-economic
characteristics and 3) behavioural biases. The findings indicate the importance of
generating initial recall awareness through mass media, and following up with direct
notifications methods (with priority given to two-way channels). It is also crucial to reach
out to the local areas where affected consumers live and communicating in the language
they speak. Trusted messengers, such as community organisations, are important to get
the message across.
The French General directorate for competition, consumer policy and fraud control
(DGCCRF) presented the actions taken by the French government to improve recall
effectiveness after the 2018 Lactalis infant milk formula crisis. In this particular case,
voluntary measures were not enough, so the Minister of Economy and Finance ordered a
mandatory recall. The recall faced several effectiveness issues, both on the producer level
(failure to correctly identify affected batches, to take effective measure and to withdraw
all recalled products from the supply chain) and on retail level (some recalled products
still on sale). As a follow-up to the Lactalis crisis, new legal remedies were adopted both
in the food law (October 2018) and in the economic law (May 2019): 1) self-test reports
showing environmental contamination to be immediately forwarded to authorities 2)
increased accountability of producers and retailers during withdrawals and recalls 3)
centralised public recalls website. The latter will cover food and feed as well as non-food
consumer products (except for pharmaceuticals and medical devices). The website will be
linked to a mobile application – “Signal Conso” – that allows consumers to report any
concerns (safety as well as economic and commercial).
GROUP DISCUSSIONS
GROUP 1: Content and format of a recall notice
The group reflected on how to improve the content and visual presentation of recall
notices to make them more eye-catching and persuasive.
The experts agreed that the message should be short, straight to the point and attention-
grabbing. Some argued that terms like 'important safety notice/warning' may actually be
preferable to 'recall' in the headline. The risk needs to be communicated in a transparent
and understandable way, avoiding "sugar-coating" and jargon. If relevant, the message
should be translated into different languages so that it reaches consumers in their own
language. Since "a picture is worth a thousand words", product identification information
should be accompanied by high-quality visual material. The idea of using a recognisable
recall symbol was also mentioned.
It was noted that standardising the content and layout of recall announcements has the
potential to make them simpler and more easily recognisable. EBay mentioned their
positive experience with recall messages harmonisation. However, there may also be
downsides to standardisation: if recall notices become too similar, they may also be less
154
eye-catching and some products have a very specific target audience). It was agreed that
some key elements and ground rules, applicable to all recall notices, should be
standardised and compulsory.
Moreover, recall messages could include certain behaviourally-informed 'nudges' and
persuasion techniques, such as social proof, loss aversion, reciprocity, personalisation.
For instance, appealing to social norms (i.e. highlighting that the majority of people
engage in and/or approve of certain behaviour) has proved effective in other policy areas.
If using statistics, the numbers need to be real, yet somehow surprising to the public
(typically higher than expected). In cases when recall response rates are low, mentioning
dynamic statistics (“increasingly many”) could be an option. Personalisation (e.g. adding
a person's name to the message) has also been shown to encourage action but it is only an
option when the consumer can be contacted directly. In addition, opportunity cost (i.e. the
costs of not having corrective measures in place) could be highlighted in communication
to help enhance consumers trust in the whole product safety system.
There was a broad agreement that setting up a centralised government recall portal can
provide a useful single access point for consumers. However, this should not remove the
responsibility from businesses to spread the message through all possible channels.
GROUP 2: Communication channels
The group agreed that the most appropriate communication channels to reach out to the
affected consumers will depend on a number of factors.
The age of the target audience is of crucial importance. While Facebook and Instagram
are key to reach younger consumer groups, traditional media such as TV and radio
remain the preferred communication channels for older consumers.
Language, ethnicity and culture were identified as other important aspects when
elaborating a recall campaign. In addition, some groups may require very targeted
communication approach, e.g. through religious communities. Cultural differences also
play a role. As an example, a recall carried out at the same time and with the same
communication material in a European country and in an Asian country resulted in very
different response times.
In addition, geography and how easy it is for consumers to return a recalled product also
needs to be taken into account. An example of a producer sending help to remote
communities to repair the product was mentioned. Such approach would also be helpful if
the product is too heavy to be brought/sent back. Another solution would be to ask
consumers to destroy the product themselves and send only a proof of destruction to
receive a remedy.
Regardless of the specificities of the case and characteristics of the target audience, direct
communication with individual consumers is always more effective than blanket
communication campaigns. According the US CPSC research, direct contact results in an
average 50% return rate, compared to 6% in case of generic press releases.
It was also highlighted that the channels should not be separated from the message itself.
A recall notice should be formulated in a way that is attractive and understandable for the
target audience. Otherwise, it will not motivate action. Consumers should get clear
information on the risk at hand and instructions on what to do with the product. If
155
information is presented in a clear and transparent way, it may actually boost consumers'
trust in the recalling company.
Examples of communications to catch consumers' attention and motivate response
include using animals or icons as recall ambassadors (e.g. the US campaign with a cat
riding a unicorn).
Finally, it was mentioned that the communication capacities will depend on the size of
the company and its place in the supply chain. Bigger companies are usually more aware
of the importance of recalls and have dedicated departments to deal with them. SMEs and
start-ups face more challenges.
GROUP 3: Recall procedure and remedies
The main topic was how to increase consumer response to recalls, focusing on the recall
procedure itself. While acknowledging that the “one-size-fits-all” approach is not
possible, participants recognised that all steps of the recall process should be as simple as
possible and the remedies offered should be sufficient from a consumers' perspective.
Often, consumer inaction is due to the procedure to follow being too complex or difficult
to understand, while the recall should be as easy as buying the product.
To start with, a recall notice should be attractive and easy to understand for the consumer,
providing clear information that a product he/she bought is affected, what steps to take
and where to get more information. It is crucial to adopt the consumers’ language in the
recall communication. For instance, a recent IKEA recall in Sweden was published in
Swedish, English and Arabic to reach also immigrant families. Effective communication
also means using pictures in addition to text.
Participants emphasised that there is still no means of identification for many recalled
products, and suggested that some identification requirements should be mandatory. At
the same time, it was noted that multiple model names and codes in the same recall
notification may be confusing. The use of online tools would be one solution to allow
consumers to easily ascertain if their product is affected by a recall.
The setting up of a hotline to answer to consumers’ questions on a recall was also
identified as a good practice, even though it might not be possible for SMEs due to the
costs involved.
Participants agreed that guidance would be a good idea to help businesses navigate the
recall procedure, especially those that do not perform recalls on a regular basis. Having a
timeframe for different steps in the recall procedure, following the UK example, would
help (even though, again, the need for a case-by-case approach should be stressed); the
same goes for standard responses. Overall, economic operators expressed their preference
for a coordinated approach to recalls at the EU level.
As for deadlines, the earlier a recall is published the better. However, this does not mean
that the information about previous recalls should be removed. To increase recall’s speed,
IKEA uses "risk communicators", who send targeted messages early in the stage of the
recall process. They also keep a communication channel for consumers for any queries.
Typically, businesses face a response peak (highest number of returns) in Month-2 (i.e.
after 8 weeks), then they start “re-launching” the recall.
When defining the remedies, it is important to adopt a holistic approach, taking account
of both safety and environmental considerations. The type of product and the risk at stake
156
will also have an impact. Having large numbers of consumers bringing or sending back a
product may not always be the most efficient and sustainable solution. It was noted that
for large items it may be preferable for the economic operator to pick up the recalled
product themselves (even though for SMEs this may be a challenge). For low-value
products, providing the information "danger: dispose of it" may be the best approach. In
some cases self-repairs by consumers (under instruction) may be feasible, and could also
help minimise consumers' fear of losing the product (which is a major barrier to product
returns). The question is what burden can be placed on the consumer to return the product
and under which circumstances and what type of products can be left in the consumers'
hands (for self-repair or disposal).
Another issue is how to take account of reactions other than returns. It would be very
helpful to introduce a feedback mechanism, whereby the consumer could indicate what
they did in response to a recall (e.g. "I disposed of the product, I repaired it myself, I had
it repaired by the manufacturer, I sent//brought it back to the manufacturer/retailer, I no
longer own the product").
It was suggested that public authorities should play a more active role in communicating
recall information to consumers as "trusted messengers". Official letters by the authorities
with a reference to the recalling company can help boost consumers' trust and increase
response rates. Authorities should also provide support to businesses in designing the
most efficient recall remedies (businesses often go for the least expensive solution). In
addition, in case the measures taken by economic operators are not sufficient, the
authorities should step in and support businesses in designing a better second recall
solution to avoid "recall fatigue". Finally, participants pointed to the need for
strengthened cooperation between authorities from different jurisdictions.
Among the challenges on the consumer side, participants mentioned specific socio-
demographic characteristics (e.g. consumers with disabilities or living in remote areas)
and different risk perceptions. For instance, informed consumers may feel less risk-
averse. Some people may also try to repair a dangerous product on their own, the fear of
loss being an important barrier to recalls. Therefore, it is important to adopt a flexible
approach both to the communication and to the remedies.
It was also highlighted that recalls can be particularly difficult if other people can be
harmed by the unsafe product and not necessarily the buyer himself (e.g. cases where the
product generates environmental pollution or poses a risk of fire). Incentivising people to
take action in such cases may be particularly challenging.
Financial incentives were considered useful, but not appropriate for all products, which
again calls for a case-by case approach. The US CPSC recommends a “bounty” (refund
of the product's initial price plus a little extra) in well-justified cases. Financial incentives
and fines should be used in particularly high-risk cases when the product needs to be
removed from homes (e.g. strollers) or when other means do not work. Financial
incentives should be used intelligently and not too often so that consumers do not get
used to "picking the best offer". The case of a recall where the financial incentive
increased during the second wave was brought up; the risk could be that in future recalls,
consumers would wait to act until the incentive has increased, which would be counter-
productive.
Another way to increase consumers' propensity to reply to a recall is to minimise the
effort required from them. This could include product’s collection by the businesses
157
themselves, repair outside of business hours, or “proposed default dates” for
repair/collection.
Hasbro mentioned that in the framework of their programme to return old toys for
recycling, consumers have two possibilities, i.e. to print a free-shipping document or
leave the toys at a collect point at neighbourhood collection points (retailers). The latter
solution, which is similar to systems used to recycle medicines and batteries, has been
found to be more effective. It was, however, noted that this solution may not be feasible
for products bought online, if the retailer has no physical shop.
3. Session 3: Roles and responsibilities in the recall process
The objective of this session was to discuss the roles and responsibilities of different
actors involved in a recall process and how to improve their cooperation to optimise
recall effectiveness.
PRESENTATIONS
US Consumer Product and Safety Commission (CPSC) and Health Canada gave a joint
presentation on the benefits of regulators' proactive involvement in recall communication
and recall coordination across jurisdictions. The CPSC shared their extensive experience
of collaborating with the recalling firm on defining the corrective action plan before
recall announcement. In particular, the authority and the recalling firm negotiate the
language of a joint press release and other recall communications. The CPSC is also
closely involved in monitoring the effectiveness of the recall (e.g. on the basis of monthly
progress reports from the recalling firm). Moreover, product safety authorities in the US,
Canada and Mexico have a long-lasting structured cooperation, which includes
coordinating important voluntary recalls across jurisdictions. Around a third of the 250-
300 recalls taking place in Canada each year are done jointly with the US. The benefits
include increase in consumer participation, leverage from other regulators, information
sharing, strengthening cross-border protection of consumers, consistent messaging and
reducing the burden on the industry. Health Canada explained their criteria for classifying
recalls depending on the level of hazard and corresponding timelines for communicating
recall information. Health Canada's Recalls and Safety Alert website provides single
access point to all recall information. It also contains guidance for businesses on how to
complete an effective voluntary recall.
Goodbaby International explained the company's internal product incident management
procedure and its 24/24/24 commitment to product safety. A product recall may be
triggered by 1) critical incidents, 2) injury reports and/or 3) test failures. The analysis
stage includes full understanding of the risk, possible solutions as well as implications in
terms of logistics and project management. Depending on the severity of the case, a
decision is made to either continue safety campaign and prepare a product recall or close
the campaign and transfer the case to 8D/quality management. The implementation phase
consists in determining and developing the most effective overall procedure to "go live"
with the recall, The legal authority is also updated at this stage. The activation phase
includes recall communication, administering the solution to all products in the market
and storage as well as monitoring and reporting on the response rates. In the final stage,
158
the safety campaign is completed, authorities notified of the outcome and lessons
gathered for future incidents.
Amazon described company's product recall programme. In line with the voluntary
commitments made in the Product Safety Pledge (June 2018), authorities' recall websites
and the EU Safety Gate are monitored on a daily basis and dangerous product listing are
removed within 2 working days. The same deadline is applied to responding to authority
contacts. Advanced search options are used to identify all affected products (e.g. in case a
batch number is missing, image search may be used). Confirmed listings are immediately
removed from sale and multilingual key-word based filters are implemented to identify
and prevent re-listing. Tailored messages are sent to suppliers and, if required by the
recall type and scope, to individual customers. Physical inventory is removed from the
supply chain by returning the products to the supplier or destroying them.
SGS Digital Trust Services talked on behalf of the TIC Council about the impact of
cybersecurity risks linked to Internet of Things (IoT) devices and the implications for
consumers, manufacturers and regulators. IoT vulnerabilities usually affect all Member
States (as compared e.g. to food) and may pose a wide range of risks to consumers
(including fatal safety risks), businesses and the society as a whole. The vulnerabilities
may not be transparent for the manufacturers and the users or there may be technical or
other challenges to fixing and patching them. The actions to be taken by consumers and
manufacturers were discussed, including the possibility of reducing or switching off the
devices capacity to prevent incidents. The implications for the regulation, surveillance
and enforcement work were also examined.
GROUP DISCUSSIONS
GROUP 1: Roles and responsibilities of authorities
The Commission announced that a coordinated activity (CASP) on recalls will be
organised next year and there are still some places available for EU market surveillance
authorities to join.
Different authorities shared their experiences and practices regarding product recalls.
From the discussions held, it can be concluded that there is no systematic approach to
market surveillance authorities’ involvement in the recall process, either before it is
launched (e.g. economic operators checking with the authorities whether and how the
recall should be carried out) or in monitoring the effectiveness of measures taken by the
businesses.
In some jurisdictions, such as the Australia, New Zealand, the UK and the US, specific
rules and guidance documents have been issued to support economic operators
(particularly smaller ones) in carrying out a successful recall.
The jurisdictions that are more actively engaged in supporting the businesses in the recall
process emphasised that this has a positive impact on the timeliness and effectiveness.
This is confirmed by the comparative research carried out by the OECD Secretariat.
The role of consumer associations was touched upon as well. Recalls may be triggered by
test reports from consumer associations, even though different level of requirements of
tests performed by authorities, economic operators or consumer associations may
complicate the process. Consumer associations tend to look at levels of safety that go
159
beyond legal regulations, so the authorities may not be able to use them as a direct basis
for corrective measures.
The possibility of joint recalls was also mentioned, based on the example of Canada and
the US that have put in place a well-running system of bilateral and even trilateral (with
Mexico) recalls. The benefits include greater consistency in the timing of the recall and
the remedies offered to consumers. These joint recalls are performed on a voluntary
basis: the company performing the recall must agree to this.
The discussion concluded with the suggestion that whenever economic operators are in
doubt about how to conduct a recall, they should always contact the authorities.
GROUP 2: Roles and responsibilities of businesses
This Group discussed what businesses could do to make product recalls easier for
consumers and increase their effectiveness.
It was noted that bigger manufacturers tend to have more control of their distribution
channels and may thus be in a better position to manage recalls on their own. Distributors
usually do cooperate (depending on supply chain and market conditions). Several
participants agreed that there is a continuous flow of information between manufacturer
(technical knowledge) and distributor (local market knowledge) to agree on the best way
to perform recalls. It was mentioned that manufacturers may prefer to take over the
consumers’ mailing list, but distributors may be reluctant to make these available because
of data protection concerns and because such lists are an important marketing tool.
Participants noted that traceability can be an issue if the supply chain is long and
complex.
It was mentioned that authorities may need to step in, in particular in small countries or in
case of SMEs that are not able to control their distribution channels. For small
companies, a recall can be extremely challenging- not just because of the reputational
damage but also because of its financial impact. It was suggested that a possible solution
could be to have insurances or emergency funds.
Participants representing online marketplaces noted that they are a different actor than
“retailers” or “distributors”, as they do not have access to physical products. It was also
mentioned that online marketplaces take different approaches to recalls, some companies
taking control of the process and approaching the buyers directly and others requesting
the sellers to reach out to their customers. An issue is that some sellers are not limiting
their sales to one marketplace, and that they have no obligation to report back about recall
results.
Regarding communication, participants noted that guidelines would be beneficial for both
businesses and consumers. However, such guidelines should remain general because each
and every recall is different, and there is a need for flexibility.
The importance of monitoring the effectiveness of each recall action was clearly
recognised. IKEA mentioned their internal monitoring system per country (small EU
countries tend to be very fast to return while e.g. Japan is very slow). Businesses should
create return statistics, as well as Q&A for internal use (not all the information gathered
is needed by consumers).