Regulating Chemical Substances under REACH: The Choice … · cation and labelling was introduced...

Regulating Chemical Substances under REACH:The Choice between Authorization and Restriction

and the Case of Dipolar Aprotic Solvents

Lucas Bergkamp and Nicolas Herbatschek

The REACH Regulation establishes several chemicalregulatory regimes, which operate, by and large, asstand-alone, but ostensibly complementary pro-grammes. The two key REACH programmes for direct‘command and control’ regulation of chemical risk are‘restriction’ and ‘authorization’. In the case of sub-stances of very high concern, both restriction andauthorization are available as risk management mea-sures. Because REACH fails to establish an indepen-dent, coherent and unbiased framework for chemicalrisk assessment and policy analysis of these alterna-tive regulatory options, their deployment has beenfraught with difficulties. This article reviews theREACH provisions governing the restriction andauthorization programmes, and the differences, simi-larities and interrelations between them. In the secondpart, the problems arising in the application of the tworegimes are illustrated with reference to the case ofdipolar aprotic solvents. This case study demonstratesthat the most appropriate regulatory instrument maybe a regime other than REACH. The third part setsforth some recommendations to improve current prac-tice and move towards a predictable, reasonable andbalanced REACH application.

INTRODUCTION

This article reviews key issues associated with the regu-lation of chemical substances under the REACH Regu-lation,1 and illustrates these issues with particularreference to the recent case of dipolar aprotic solvents.The REACH Regulation establishes two stand-aloneregimes for the regulation of chemical risk – authoriza-tion and restriction – which differ in importantrespects.2 To coordinate between these two regimes, it

provides no more than a few rudimentary rules. Theissue arises how the choice between authorization andrestriction is, and should be, made. In regulatingchemical risk, regulators do not have discretion in allcases to pursue either authorization or restriction. Thischoice is available only with respect to substances ofvery high concern (SVHCs), and then only if chemicalrisks are unacceptable and not adequately controlled.Conversely, if a substance is not an SVHC (e.g., a sub-stance posing only physico-chemical risk), authoriza-tion is not an option and restriction is the only way toregulate its unacceptable risks. The literature on thisissue is scant,3 and this article seeks to fill the void. Inaddition, beyond the four corners of REACH, the mostappropriate regulatory measure to address a specificchemical risk may be available under another chemicalrisk regulatory programme. The recent experience withregulatory initiatives regarding dipolar aprotic solventsprovides insights into the evolving Commission prac-tice in selecting regulatory instruments.

This article first discusses how REACH fits into thegeneral framework of EU chemical legislation. Itanalyzes, compares and examines the interrelationsbetween the restriction and authorization programmes,including with respect to imports. Specific attention ispaid to the question of how REACH coordinates the tworegimes and contemplates the choice between the twoinstruments. We also assess the implications of generalprinciples of EU law on the choice of regulatory instru-ment. The next part highlights these issues with refer-ence to the case of dipolar aprotic solvents. This partdiscusses the early steps in the procedure, the lessonslearned and the move toward a coherent approach.Applying the relevant legal concepts to this case, weidentify the most suitable regulatory approach for thesesolvents. The final part presents the conclusions and afew recommendations.1 Regulation 1907/2006/EC of 18 December 2006 Concerning the

Registration, Evaluation, Authorization and Restriction of Chemicals(REACH), Establishing a European Chemicals Agency, AmendingDirective 1999/45/EC and Repealing Council Regulation (EEC) No793/93 and Commission Regulation (EC) No 1488/94 as well asCouncil Directive 76/769/EEC and Commission Directives 91/155/EEC, 93/67/EEC, 93/105/EC and 2000/21/EC, [2006] OJ L396/1(‘REACH’).2 See L. Bergkamp (ed.), The European Union REACH Regulation forChemicals: Law and Practice (Oxford University Press, 2013).

3 See, e.g., Milieu Ltd, Technical Assistance Related to the Scope ofREACH and Other Relevant EU Legislation to Assess Overlaps(European Commission, 2012); L. Bergkamp, n. 2 above; D.Drohmann and M. Townsend (eds.), REACH: Best Practice Guide toRegulation (EC) No 1907/2006 (Hart, 2013); European ChemicalsAgency (ECHA), Workshop on the Candidate List and Authorisationas Risk Management Instruments (ECHA, 2009).

bs_bs_banner

Review of European Community & International Environmental Law

RECIEL 23 (2) 2014. ISSN 2050-0386 DOI: 10.1111/reel.12083

© 2014 John Wiley & Sons Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.

221

REACH AUTHORIZATIONAND RESTRICTION

The REACH Regulation is vast and convoluted. In addi-tion to the Regulation itself, which runs to approxi-mately 120 pages of dense text, there is a large numberof annexes totalling some 400 pages, implementing leg-islation – in particular on testing methods4 and aclosely related Regulation on the classification, label-ling and packaging of substances (CLP Regulation).5 Toassist the authorities and industry with implementa-tion, a series of guidance documents have been issuedby the European Chemicals Agency (ECHA).6 ECHA’sappeal board7 and the European courts have alreadyissued a fair number of decisions on the interpretationof REACH.8 A body of literature on various aspects ofREACH is emerging,9 including comparative analyses10

and examinations of technical aspects,11 and the first

books exclusively on REACH have been published.12

Further, several reports have been published in connec-tion with the review and evaluation of the Regulation.13

We do not attempt to provide an overview of the entireRegulation, but instead we focus on the broader policycontext in which REACH operates, its objectives andthe key programmes it establishes. We then discussREACH’s key regulatory programmes, authorizationand restriction, in more detail.

REACH AND OTHER EUCHEMICALS-RELATEDLEGISLATIONThe EU has had chemicals legislation for decades. In2013, the Commission listed a total of 156 separate leg-islative instruments that address some aspect of chemi-cal use, release or control.14 This collection of EUinstruments deals not only with chemical substances inbulk and chemical products, but also with chemicals inair, water, soil and waste.15 There are substance-specificregulations for mercury, polychlorinated biphenyls

4 Regulation 440/2008 of 30 May 2008 Laying Down Test MethodsPursuant to Regulation (EC) No 1907/2006 of the European Parlia-ment and of the Council on the Registration, Evaluation, Authorizationand Restriction of Chemicals (REACH), [2008] OJ L142/1.5 Regulation 1272/2008/EC of 16 December 2008 on Classification,Labelling and Packaging of Substances and Mixtures, [2008] OJL353/1.6 <http://echa.europa.eu/support/guidance>.7 <http://echa.europa.eu/about-us/who-we-are/board-of-appeal/decisions>.8 Summaries can be found at: <www.reachpsforum.eu>.9 See, e.g., L. Bergkamp and J.C. Hanekamp, ‘The Draft REACHRegime: Costs and Benefits of Precautionary Chemical Regulation’,11:5 Environmental Liability (2003), 1; L. Bergkamp and T. Smith.‘Perfect Storm in the EU: Tidal Wave of Product Regulation’, Execu-tive Counsel (2004), 9; E. Fisher, ‘The “Perfect Storm” of REACH:Charting Regulatory Controversy in the Age of Information, Sustain-able Development and Globalization’, 11:4 Journal of Risk Research(2008), 541; M. Bronckers and Y. Van Gerven, ‘Legal RemediesUnder the EC’s New Chemicals Legislation REACH: Testing a NewModel of European Governance’, 46:6 Common Market Law Review(2009), 1823; F. Fleurke and H. Somsen, ‘Precautionary Regulationof Chemical Risk: How REACH Confronts the Regulatory Challengesof Scale, Uncertainty, Complexity and Innovation’, 48:2 CommonMarket Law Review (2011), 35; L. Bergkamp, G. Michaux and N.Herbatschek, ‘Nanotechnology Regulation in Europe: From REACHand Nano-Registries to Cosmetics, Biocides and Medical Devices’,11:1 Nanotechnology Law and Business (2014).10 See, e.g., A.D.K. Abelkop, J.D. Graham, L.R. Wise and Á. Botos,‘Regulating Industrial Chemicals: Lessons for US Lawmakers fromthe European Union’s REACH Program’, 42:11 Environmental LawReporter (2012), 11042; V. Heyvaert, ‘Regulating Chemical Risk:REACH in a Global Governance Perspective’, in: J. Eriksson, M.Gilek and C. Rudén (eds.), Regulating Chemical Risks, Europeanand Global Challenges (Springer, 2010), 217; J.T.O. Reilly, ‘WhatREACH Can Teach Us about TSCA: Retrospectives on America’sFailed Toxics Statute’, 1:1 European Journal of Risk Regulation(2010), 40; O. Renn and E.D. Elliott, ‘Chemicals’, in: J.B. Wiener,M.D. Rogers, J.K. Hammitt and P.H. Sand (eds), The Reality ofPrecaution: Comparing Risk Regulation in the United States andEurope (Taylor & Francis, 2011), 231.11 See, e.g., G. Schoeters, ‘The REACH Perspective: Toward a NewConcept of Toxicity Testing’, 13:2–4 Journal of Toxicology and Envi-ronmental Health (2010), 232; G. Schaafsma et al., ‘REACH, Non-testing Approaches and the Urgent Need for a Change in Mind Set’,53:1 Regulatory Toxicology and Pharmacology (2009), 70; D.

Romano, T. Santos and R. Gadea, ‘Trade Union Priority List forREACH Authorisation’, 65:1 Journal of Epidemiology and CommunityHealth (2011), 8.12 See L. Bergkamp, n. 2 above; D. Drohmann and M. Townsend, n.3 above.13 Eurostat, The REACH Baseline Study: Five Years Update(Eurostat, 2012); PricewaterhouseCoopers (PwC), Review of theEuropean Chemicals Agency (ECHA) (PwC, 2012); Gaia, REACHContribution to the Development of Emerging Technologies (Gaia,2012); Milieu Ltd, Implementation and Enforcement of Restrictionsin Member States (Milieu Ltd, 2012); Centre for Strategy and Evalu-ation Services (CSES), Interim Report: Impact of the REACH Regu-lation on the Innovativeness of the EU Chemical Industry (CSES,2012); Milieu Ltd, PACE and Risk and Policy Analysts, InspectionsRequirements for REACH and CLP (Milieu Ltd, 2011); CSES,Interim Evaluation: Functioning of the European Chemical Marketafter the Introduction of REACH (CSES, 2012); Milieu Ltd, Techni-cal Assistance Related to the Scope of REACH and Other RelevantEU Legislation to Assess Overlaps (Milieu Ltd, 2012); Risk andPolicy Analysts and Okopol, Technical Assistance to Prepare theCommission Report on Operation of REACH (Risk and Policy Ana-lysts, 2012); Risk and Policy Analysts, Review of the RegistrationRequirements for Polymers and 1 to 10 Tonnes Substances (Riskand Policy Analysts, 2012); DHI, Okopol and Risk and Policy Ana-lysts, Assessment of Health and Environmental Benefits of REACH(DHI, 2012); Joint Research Centre, Scientific Technical Support onAssessment of Nanomaterials in REACH Registration Dossiersand Adequacy of Available Information (Joint Research Centre,2012).14 European Commission, Report from the Commission to the Euro-pean Parliament, the Council, the European Economic and SocialCommittee and the Committee of the Regions in Accordance withArticle 117(4) of REACH and Article 46(2) of CLP, and a Review ofCertain Elements of REACH in Line with Articles 75(2), 138(2), 138(3)and 138(6) of REACH, Staff Working Document SWD(2013) 25.15 E. Stokes and S. Vaughan, ‘Great Expectations: Reviewing 50Years of Chemicals Legislation in the EU’, 25:3 Journal of Environ-mental Law (2013), 411, at 415–417.

LUCAS BERGKAMP AND NICOLAS HERBATSCHEK RECIEL 23 (2) 2014

© 2014 John Wiley & Sons Ltd

222

http://echa.europa.eu/support/guidance

http://echa.europa.eu/about-us/who-we-are/board-of-appeal/decisions

http://echa.europa.eu/about-us/who-we-are/board-of-appeal/decisions

http://www.reachpsforum.eu

(PCBs)16 and other substances.17 The EU has also regu-lated the production, transportation, import andexport, and use of hazardous chemicals, such aschlrofluorocarbons (CFCs).18 Our focus is on EU legis-lation addressing environmental and health and safetyrisks associated with chemicals in bulk – not chemicalsin solid products or in air emissions or waste waterdischarges.

The initial core of EU chemical legislation was built on,in chronological order: classification and labelling ofdangerous substances; the notification of substances,with the exception of existing substances;19 classifica-tion and labelling of mixtures of substances;20 a specificgovernment-driven risk assessment programme;21 andrestrictions on the marketing and use of dangerouschemicals adopted by the Commission through thecomitology process.22

In recent times, however, EU chemical legislation haschanged dramatically. The adoption of the REACHRegulation implied a paradigm shift in the governanceof chemical risks in Europe.23 A new system of classifi-cation and labelling was introduced by the Classifica-tion, Labelling and Packaging Regulation.24 In additionto these core chemical laws, the EU adopted other leg-islation aimed at regulating chemical risk. First, EUoccupational health and safety legislation deals specifi-cally with chemical agents.25 Pursuant to this directive,EU-wide indicative and binding occupational exposurelimit values and biological limit values can be estab-lished. Second, specific products made up of chemicals,such as biocides, pesticides, pharmaceuticals and cos-metics, are subject to product-specific legislation.Third, chemical restrictions, including maximum con-centration values and outright bans, have also beenincluded in regulations concerning products such as

automobiles,26 electronics,27 toys,28 medical devices29

and cosmetics.30 In all these cases, the EU legislaturechose to retain and deploy separate regulatory instru-ments independent of REACH. In addition, the Com-mission is examining selected issues, such as endocrinedisruptors,31 nanotechnology32 and combined effects,33

and may adopt additional legislation outside ofREACH.

Prior to REACH, EU chemical legislation was primarilyrisk-based, and distinguished between risk assessmentand risk management.34 Subject to limited exceptions,such as pharmaceutical products,35 the burden of proofwith respect to the presence of chemical risk was placedon the authorities: they had to establish that a chemicalsubstance created an actual risk and thus should beregulated.36 The process of assessing the risks of exist-ing substances was perceived as too slow.37 To addressthese shortcomings, REACH transferred the burden ofproduction of information on chemicals to the compa-nies that manufacture and distribute them.38 In addi-tion, REACH also reverses the initial burden of proof39

of safety by requiring that manufacturers and importersensure that the risks associated with the uses of theirchemicals are ‘adequately controlled’. As a safety net,where an unacceptable chemical risk is found toremain, restrictions can still be imposed.40 In addition,reflecting the precautionary principle, certain hazard-ous substances can now be subjected to authorization,which is intended to phase out and substitute thesesubstances.41

16 Directive 96/59/EC of 16 September 1996 on the Disposal of Poly-chlorinated Biphenyls and Polychlorinated Terphenyls (PCB/PCT),[1996] OJ L243/31.17 For a review, see Milieu Ltd, n. 3 above.18 Regulation 1005/2009/EC of 16 September 2009 on Substancesthat Deplete the Ozone Layer, [2009] OJ L286/1.19 Directive 67/548/EEC on the Approximation of the Laws, Regula-tions and Administrative Provisions Relating to the Classification,Packaging and Labeling of Dangerous Substances, [1967] OJL196/1.20 Directive 1999/45/EC Concerning the Approximation of the Laws,Regulations and Administrative Provisions of the Member StatesRelating to the Classification, Packaging and Labelling of DangerousPreparations, [1999] OJ L200/1.21 Regulation 793/93/EC on the Evaluation and Control of the Risks ofExisting Substances, [1993] OJ L84/1.22 Directive 76/769/EEC of 27 July 1976 on the Approximation of theLaws, Regulations and Administrative Provisions of the MemberStates Relating to Restrictions on the Marketing and Use of CertainDangerous Substances and Preparations, [1976] OJ L262/201.23 See F. Fleurke and H. Somsen, n. 9 above; E. Fisher, n. 9 above.24 Regulation 1272/2008/EC, n. 5 above, Article 50.25 Directive 98/24/EC of 7 April 1998 on the Protection of the Healthand Safety of Workers from the Risks Related to Chemical Agents atWork, [1998] OJ L131/11.

26 Directive 2000/53/EC of 18 September 2000 on End-of lifeVehicles, [2000] OJ L266/34.27 Directive 2011/65/EU of 8 June 2011 on the Restriction of the Useof Certain Hazardous Substances in Electrical and Electronic Equip-ment, [2011] OJ L174/88.28 Directive 2009/48/EC of 18 June 2009 on the Safety of Toys, [2009]L170/1.29 Directive 93/42/EEC of 14 June 1993 Concerning Medical Devices,[1993] OJ L169/1.30 Regulation 1223/2009/EC of 30 November 2009 on CosmeticProducts, [2009] OJ L342/59.31 L. Bergkamp and N. Herbatschek, ‘Key Concepts and Scope’, in:L. Bergkamp, n. 2 above, 40, at 53–55.32 See L. Bergkamp, G. Michaux and N. Herbatschek, n. 9 above.33 See L. Bergkamp and N. Herbatschek, n. 31 above, at 47.34 Regulation 793/93/EC on the Evaluation and Control of the Risks ofExisting Substances, [1993] OJ L84/1.35 Directive 2001/83/EC of 6 November 2001 on the Community CodeRelating to Medicinal Products for Human Use, [2001] OJ L311/67.36 B. Hansen, ‘Background and Structure of REACH’, in: L.Bergkamp, n. 2 above, 17, at 21–22.37 Ibid., at 17–18.38 This is the ‘no data, no market’ concept. REACH, n. 1 above, Article5.39 As A. Abelkop and J.D. Graham, Regulation of Chemical Risk:Lessons for TSCA Reform from Canada and the European Union(forthcoming) argue: ‘[T]he EU processes for the production of dataand determinations of safety for specific uses entail an iterativeprocess of shared burdens.’40 REACH, n. 1 above, Article 68.1.41 Ibid., Article 55.

RECIEL 23 (2) 2014 REGULATING CHEMICAL SUBSTANCES UNDER REACH


223

REACH OBJECTIVESAND REGIMESThe REACH Regulation followed a review of the pre-existing EU chemicals legislation. Based on this review,the Commission concluded that the prior legislation didnot provide ‘a high level of protection’ as required bythe Treaty on the Functioning of the European Union(TFEU).42 A major problem identified by the Commis-sion was ‘the general lack of knowledge about the prop-erties and the uses of existing substances’, also referredto as the ‘data gap’.43 This knowledge gap was attribut-able to the fact that under the Dangerous SubstancesDirective only new substances, not existing ones, had tobe tested and notified.44 Consequently, the authoritiesdid not possess much data on ‘existing substances’.

On the drawing board, REACH was intended as a risk-based chemical control programme.45 Hazards wouldcontinue to inform decision making, but the idea wasthat risk management decisions would be based onactual risk, as opposed to hazard or potential risk.While hazard, in theory, is conceived as an intrinsicproperty of a substance that is independent of the quan-tity present and the physical form of the substance,46

risk is an actual probability of adverse effects occurringin a specific situation and the magnitude of such effectsat specific doses and levels of exposure.47 Risk is quan-tifiable to some degree or within a certain range. Com-pared to hazard-based regulation, the advantage ofregulating risk is that measures can specifically targetthose factors in a causal chain that lend themselves bestto regulatory interventions, thus generally decreasingthe costs and increasing the benefits of such interven-tions. As designed and implemented, however, REACHis a mix of hazard- and risk-based regulation. Forinstance, regulation may be based solely on hazard oron the use of proxies or surrogates for exposure or risk(or factors influencing exposure or risk), rather thanactual, proven risk.48 A hazard-triggered regime, autho-rization has proven to be the most controversial andunwieldy REACH programme.49

A main feature of the REACH Regulation is the cre-ation of a comprehensive, albeit unintegrated, systemof government oversight, subject to qualifications andexemptions, over all ‘existing’ and ‘new’ substancesthroughout their entire life cycle, including design andproduction, industrial and consumer use, and recyclingand reuse.50 Consequently, its scope is vast; subject tolimited exceptions, REACH applies to all bulk chemi-cals used in industrial processes and to chemicalspresent in mixtures and products ranging from clean-ing products, and paints to clothing, furniture and elec-trical appliances. REACH is complicated; it is not oneregulatory programme, but a collection of ‘stand-alone’but ostensibly complementary programmes, includingregistration, evaluation and authorization of chemicalsubstances, as well as generally binding restrictions onchemical substances and supply chain information.Importantly, REACH established a new Europeanagency – the European Chemicals Agency (ECHA) –which is responsible for receiving registration dossiers,managing the huge chemical databases resulting there-from, and issuing guidance to assist producers, import-ers and other authorities in implementing REACH.51

ECHA is also involved in the restriction and authoriza-tion processes.52 In addition to ECHA, the Commissionand Member State authorities are in charge ofadministering various parts of REACH, includingauthorization and restriction, creating a patchwork ofgovernment powers, procedures and oversight,53 whichonly further adds to REACH’s complexity.

The four main programmes established by REACH –registration, evaluation, authorization and restriction –can be characterized as follows. The registration pro-gramme involves the analysis and submission to ECHAof information on a given substance, and is required forsubstances manufactured or imported in a volume of atleast 1 tonne per year per manufacturer or importer.54

Evaluation, which may focus on a dossier55 or a sub-stance,56 is the review of that (and possibly additional)information by the authorities, which may result inqueries and orders to provide additional information.Authorization is a permitting programme for specificuses of certain dangerous substances (SVHCs) and isintended to cause ‘substitution’ (replacement) of thosesubstances by safer ones.57 Restrictions are generallybinding requirements that restrict the manufacture,

42 Treaty on the Functioning of the European Union, [2010] OJC83/47 (‘TFEU’), Article 191.43 European Commission, White Paper, Strategy for a Future Chemi-cals Policy, COM(2001) 88, at 6.44 Directive 79/831/EEC Amending for the Sixth Time Directive67/548/EEC on the Approximation of the Laws, Regulations andAdministrative Provisions Relating to the Classification, Packagingand Labelling of Dangerous Substances, [1979] OJ L259/10, Article5.45 See generally R.E. Lofstedt, ‘Risk versus Hazard: How to Regulatein the 21st Century’, 2:2 European Journal of Risk Regulation (2011),149.46 C.J. van Leeuwen, ‘General Introduction’, in: C.J. van Leeuwen andT.G. Vermeire (eds), Risk Assessment of Chemicals: An Introduction(Springer, 2007), 1, at 3.47 Ibid.48 See L. Bergkamp and N. Herbatschek, n. 31 above, at 52.49 See also R. Lofstedt, ‘The Substitution Principle in Chemical Regu-lation: A Constructive Critique’, 17:5 Journal of Risk Research (2013),

543. Cf. A.D.K. Abelkop and J.D. Graham, ‘Principles and Tools ofChemical Regulation: A Comment on “The Substitution Principle inChemical Regulation: A Constructive Critique” ’, 17:5 Journal of RiskResearch (2013), 581.50 REACH, n. 1 above, Annex I, sections 0.3 and 5.2.2.51 Ibid., Article 77.52 Ibid., Articles 58.3, 59, 64 and 69–72.53 L. Bergkamp and D.Y. Park, ‘The Organizational and Administra-tive Structures’, in: L. Bergkamp, n. 2 above, 23.54 REACH, n. 1 above, Article 6.1.55 Ibid., Articles 40–43, 49–51 and 53–54.56 Ibid., Articles 44–50 and 52–54.57 Ibid., Article 55.



224

placing on the market and/or use of chemicals, either inbulk or as used in products, and is aimed at managing atEU-level chemical risks that are unacceptable becausethey are not adequately controlled.58 In addition to thefour programmes, REACH sets forth rules on notifica-tion of SVHCs in articles,59 supply chain communica-tion60 and obligations of downstream users.61

REACH’s main programmes operate largely indepen-dently of each other. This implies, for instance, thatsubstance evaluation is not a prerequisite for eitherauthorization or restriction. Logically, a risk assess-ment and analysis of available regulatory instrumentsshould precede additional regulation of the substanceconcerned since a thorough evaluation of the risksarising from the uses of a substance, the risk manage-ment measures implemented in practice and the regu-latory options informs the decision as to whether anyfurther measures are necessary.62 REACH, however,does not prescribe any such sequence. Substance evalu-ation, which falls short of full risk assessment and cost-benefit analysis of possible regulatory measures, isoptional, but if it is conducted, it may trigger the autho-rization or restriction procedure.

To summarize, REACH is aimed at reducing the risksassociated with chemical substances over their entirelife cycle. To this end, it establishes several stand-aloneprogrammes, including authorization and restriction. Abasic principle is that companies manufacturing andusing chemicals must ensure safe use and prevent unac-ceptable risks.63 The authorization and restriction pro-cedures are important, yet distinct processes forensuring chemical safety, raising the question what roleeach of these regimes plays. The following sub-sectionsdiscuss the two regimes in more detail.

AUTHORIZATIONThe authorization regime is intended to phase outcertain hazardous substances for suitable alternatives,either chemical or technological.64 Once a substance issubject to authorization, it may not be placed on themarket for a use or used on its own, in mixtures orincorporated into an article unless the use has beenspecifically authorized.65 Manufacturers and users ofSVHCs listed in Annex XIV must apply for authoriza-tion by a deadline imposed pursuant to REACH, andmust cease use of the chemical by the ‘sunset’ date

unless authorization has been granted or an applicationfor authorization is pending.66

Substantive RequirementsThe Commission is in charge of making decisions onauthorization applications pursuant to a procedure(discussed below).67 Authorization should be granted ifthe registrant demonstrates that: (i) the risks are‘adequately controlled’, provided that the substanceconcerned is not a ‘non-threshold’ substance or a per-sistent, bioaccumulative and toxic (PBT), very persis-tent and very bioaccumulative (vPvB), or ‘equivalentconcern’ substance;68 or (ii) the substance providessocio-economic advantages that outweigh the risks andno ‘suitable alternative substances or technologies’ areavailable.69 The burden of proof is on the applicant. Toforce the substitution of the substance concerned,authorization is subject to review and potential with-drawal70 by the Commission. A review may be initiatedat any time – for example, if new information on suit-able alternatives becomes available.71 Downstreamusers may use a substance for an authorized use pro-vided they have obtained an authorization or obtain thesubstance from a company to which an authorizationhas been granted, and stay within the conditions of thatauthorization.72 They must also inform ECHA,73 so thatthe authorities are fully aware of who is using sub-stances requiring authorization.

The authorization regime involves three steps, eachwith their own criteria and levels of involvement ofvarious authorities and stakeholders. The three phasesare: listing of the substance on the Candidate List;74

listing it on the list of substances subject to authoriza-tion – REACH, Annex XIV (the ‘Authorization List’);75

and the authorization of the use itself.76 The last phaseagain consists of several steps, as discussed below. Eachstep is subject to the specific procedural and substan-tive requirements set out in the REACH Regulation.77

The Candidate List is primarily a list of substances to beconsidered for listing on Annex XIV. Additionally,listing also triggers communication requirements forsuppliers of chemical substances and articles.78 Basedin part on listing rendering these informational require-ments applicable, some authorities have taken the posi-

58 Ibid., Article 68.1.59 Ibid., Article 7.2.60 See, in particular, ibid., Articles 31–34.61 Ibid., Articles 37–39.62 L. Bergkamp and M. Penman, ‘Conclusions’, in: L. Bergkamp, n. 2above, 427.63 Ibid., Articles 14.6 and 37.5.64 Ibid., Article 55.65 Ibid., Article 56.1.

66 Ibid.67 Ibid., Article 64.8.68 Ibid., Article 60.2.69 Ibid., Article 60.4.70 Ibid., Article 61.3.71 Ibid., Article 61.2.72 Ibid., Articles 56.1–2.73 Ibid., Article 66.74 Ibid., Article 59.75 Ibid., Article 58.76 Ibid., Article 60.1.77 Candidate Listing, for instance, is regulated in ibid., Articles 60–64.78 Ibid., Articles 7.2, 31.1(c) and 31.2(b).



225

tion that substances may be listed on the Candidate Listeven if there is no intent to move them to Annex XIV.79

Any use of the Candidate List for purposes other thanlisting for authorization would be questionable,however, since Candidate Listing is embedded in theauthorization listing process, and not a completelyseparate process that could serve other purposes. TheECHA is the primary decision maker with respect to theCandidate List,80 and other authorities can provideinput.81 Listing on both the Candidate List and AnnexXIV is based on hazard.82 Once a substance is listed onAnnex XIV, the authorization requirement applies tothe substance.83 At an applicant’s request, the Commis-sion must grant a renewable time-limited authorizationif either risks are adequately controlled (the ‘adequatecontrol route’)84 or the socio-economic benefits out-weighs the risks and there is no suitable alternative (the‘socio-economic route’).85

Certain substances and uses of substances are exemptfrom REACH authorization.86 Where an exemptionapplies to the whole authorization regime, as a generalrule, the exemption should be deemed to apply to boththe Candidate List and Annex XIV. With respect tointermediates, however, the Commission has adopted aposition that substances used solely as intermediatesmay be added to the Candidate List and the Authoriza-tion List based on a category approach – for example, ifthe intermediate can be used as a substitute for a sub-stance subject to authorization.87 Authorization doesnot apply to the use of substances outside the EU, suchas the incorporation of substances into articles, whichare subsequently imported into the EU.88

ProcedureThe authorization regime imposes a very substantialregulatory burden on companies. The preparation of anauthorization application takes substantial time andresources. Under the socio-economic route, the appli-cant must demonstrate that there are no suitable alter-natives.89 If there are suitable alternatives, only the

adequate control route is open and the applicant mustcommit to a substitution plan.90 The outcome of theprocedure is uncertain as any third party may suggest asuitable alternative and the authorities have discretion.

The authorization itself does not have an expirationdate, but is subject to review by a deadline determinedby the authorization on a case-by-case basis.91 Such areview may result in the withdrawal of the authoriza-tion or the adoption of a further review deadline.92 Ifcircumstances have changed or new information onsubstitutes becomes available, the Commission mayreview and, based on the results, suspend and withdrawthe authorization at any time,93 which further adds tothe uncertainty.

Exemptions from AuthorizationAlthough REACH provides for the possibility ofexempting specific uses from authorization obliga-tions,94 authorities have been reluctant to entertain anyexemptions. The only exemption adopted so far is theone for phthalates in the immediate packaging ofmedicinal products.95 This exemption, which was notrecommended by ECHA, was approved by the REACHCommittee at the proposal of the Commission.96

In several cases, exemptions from authorizationrequested by companies have been denied.97 Theprocess of listing of trichloroethylene illustrates theissues that have arisen with respect to such requests.The requested exemption was for use in surface treat-ment. ECHA stated that, once a ‘binding OccupationalExposure Limit’ (bOEL) has been adopted, it will assesswhether such a bOEL could be considered the bindingminimum requirement for controlling risks to workers’health;98 the existing EU workers protection legisla-tion99 could be the legal basis for the exemption. A dis-cussion unfolded about the question of what level ofregulation of a use should be considered a solid legalbasis for granting an exemption from authorization forthat particular use. Specifically, the issue was whether

79 See, e.g., European Commission, n. 14 above, at 72.80 ECHA makes the final decisions. REACH, n. 1 above, Articles 59.6and 59.8. However, the Commission decides if ECHA’s MemberState Committee cannot reach unanimous agreement. Ibid., Article59.9.81 Input can be provided by stakeholders and Member States. Ibid.,Article 59.82 Ibid., Article 57.83 Ibid., Article 56.1.84 Ibid., Article 60.2.85 Ibid., Article 60.4.86 See L. Bergkamp and N. Herbatschek, n. 31 above, at 75–81.87 European Commission, ‘Roadmap on Substances of Very HighConcern, Communication to COREPER/Council’ (6 February 2013),at 8.88 See N. Herbatschek, L. Bergkamp and M. Mihova, ‘The REACHProgrammes and Procedures’, in: L. Bergkamp, n. 2 above, 82, at133.89 REACH, n. 1 above, Article 60.4.

90 Ibid., Article 62.4(f).91 Ibid., Article 60.8.92 Ibid., Article 61.1.93 Ibid., Article 61.2.94 Ibid., Article 58.2.95 Regulation 143/2011/EU of 17 February 2011 Amending AnnexXIV to Regulation (EC) No 1907/2006 of the European Parliamentand of the Council on the Registration, Evaluation, Authorization andRestriction of Chemicals (‘REACH’), [2011] OJ L44/2.96 ECHA, ‘Recommendation of 1 June 2009 for the Inclusion of Sub-stances in Annex XIV’.97 See, e.g., ECHA, ‘ “Responses to Comments” Document for 4,4'-diaminophenylmethane (MDA) (EC No. 202-974-4)’ (1 June 2009), at13–26.98 ECHA, ‘Recommendation of the ECHA of 20 December 2011 onthe Inclusion of Substances in Annex XIV’.99 Directive 98/24/EC, n. 25 above; Directive 2004/37/EC of 29 April2004 on the Protection of Workers from the Risks Related to Expo-sure to Carcinogens or Mutagens at Work, [2004] OJ L229/23.



226

the existence of framework legislation that authorizesthe adoption of specific risk management measures issufficient, or whether the specific implementingmeasure should also be in place. It is unclear, however,why that would be relevant; if, under the alternativeregulatory framework, there is no specific measure inplace yet, it could be adopted under that framework.

Role of Authorization RegimeAs of 1 April 2014, there are 151 substances on theCandidate List and 22 substances on Annex XIV. By2020, the Commission contemplates that ‘all relevant,currently known SVHCs’ be included on the CandidateList.100 Note that SVHCs may be subjected to theREACH authorization regime based solely on hazard;101

no showing of actual exposure and risk is required.Substance evaluation does not have to precede theauthorization listing procedure. Further, REACH doesnot require that SVHCs be added to Annex XIV. TheCommission thus has discretion in making this deci-sion, with REACH only providing some prioritizationcriteria to guide the decision, including PBT or vPvBproperties, wide dispersive use and high volumes.102

Since the authorization programme is indiscriminatewith respect to chemical uses in the listing procedure, isnot risk-based and imposes high administrative cost, itis not necessarily the most appropriate measure forregulating any particular SVHC. For instance, if no suit-able substitutes are likely to exist in the near future, orthere already is pressure for substitution, authorizationis unlikely to result in actual substitution and maytherefore not be the most suitable regulatory option.Even worse, where the available substitutes are moreharmful than the substance subject to authorization,the environment and human health may be positivelyharmed.103 If the concern is limited to one or a few uses,authorization may be overly broad, as it affects all uses.In some cases, the Commission takes these limitationsof REACH authorization into account, as it did in rela-tion to cobalt104 and dipolar aprotic solvents, asreviewed in this article. In this respect, the so-called‘risk management option’ (RMO) analysis, discussedbelow, plays a useful role.

RESTRICTIONThe REACH restriction programme is intended tomanage risks that are not adequately addressed by the

other provisions of the REACH Regulation, includingthose on authorization.105 Under the restrictionsregime, the manufacture and use of chemical sub-stances, as well as their presence in products, can besubjected to generally binding limitations and condi-tions, including complete prohibitions.106

Restrictions adopted pursuant to the previous chemicallegislation and REACH are set forth in Annex XVII, andtypically set conditions on the manufacture, placing onthe market or use of certain dangerous substances,preparations or articles. Restrictions are adopted oramended pursuant to a procedure that involves theMember States,107 ECHA108 and the Commission,109 andallows for input from stakeholders.110 Like authoriza-tion, the process proceeds on the basis of a structureddossier, which must meet certain quality standards.111

Compared to the pre-REACH legislation, the procedurefor the adoption of restrictions has become more com-plicated and several substantive requirements havebeen added. Below, we first discuss substantive require-ments and then procedure.

Substantive RequirementsA restriction is defined as ‘any condition for or prohi-bition of the manufacture, use or placing on themarket’.112 Such a condition or prohibition may targetany use of a substance on its own, in a mixture or in anarticle,113 with limited exceptions.114 While authoriza-tion applies only to the placing on the market and use ofsubstances in the EU, a restriction may be imposed onthe manufacturing and use of substances, and on thepresence of substances in articles, including importedarticles.115 A restriction may involve a complete ban, aconcentration limit equal to the detection limit (e.g.,0.1% by weight), a maximum concentration or migra-tion limit, a limitation to industrial use, etc. In terms ofscope, a restriction may be limited to a particular use ofa substance, a group of substances or the manufactur-ing and all uses of a substance. Further, a specific ana-lytical method to determine compliance with therestriction may be imposed.116

100 See European Commission, n. 14 above, at 72.101 REACH, n. 1 above, Article 57.102 Ibid., Article 58.3.103 J.D. Graham and J.B. Wiener (eds), Risk versus Risk: Tradeoffs inProtecting Human Health and the Environment (Harvard UniversityPress, 1995).104 Regulation 348/2013/EU of 17 April 2013 Amending Annex XIV toRegulation (EC) No 1907/2006 of the European Parliament and of theCouncil on the Registration, Evaluation, Authorization and Restrictionof Chemicals (REACH), [2013] OJ L108/1, recitals 11 and 12.

105 European Commission, REACH in Brief (European Commission,2007), at 13.106 Ibid.107 REACH, n. 1 above, Article 69.4.108 Ibid., Articles 69–72.109 Ibid., Articles 69.1 and 73.110 Ibid., Articles 69.6 and 71.1.111 See ibid., Annex XV.112 Ibid., Article 3.31.113 Ibid., Article 67.1.114 Ibid., Articles 2.1–2, 67.1–2 and 68.1.115 See N. Herbatschek, L. Bergkamp and M. Mihova, n. 88 above, at147.116 See, e.g., Communication in the Framework of the Implementationof Regulation (EC) No 1907/2006 of the European Parliament and ofthe Council concerning the Registration, Evaluation, Authorizationand Restriction of Chemicals (REACH), [2012] OJ C142/8.



227

All restrictions are listed on Annex XVII. Pursuant toREACH, a restriction must be adopted if there is an‘unacceptable risk to human health or the environment. . . which needs to be addressed on a Community-widebasis’.117 ‘Unacceptable risk’ is not defined, whichimplies that the authorities have discretion in constru-ing this concept. REACH provides that, if the Commis-sion, ECHA or a Member State ‘considers’ that a risk isnot ‘adequately controlled’, a restriction proposal(Annex XV dossier) must be prepared.118 Annex XVstipulates that such a proposal must meet the rulesapplicable to the Chemical Safety Report (CSR),119 spe-cifically those used in connection with assessingadequate control in the registration and authorizationprocedures.120 This suggests that the term ‘unacceptablerisk’ may encompass all inadequately controlled risks,but these terms may not be entirely synonymous asREACH seems to distinguish ‘not adequately con-trolled’ from ‘unacceptable risk’. There is an additionalcondition, however, before a restriction may beadopted: it must be necessary that the unacceptable riskidentified be addressed at the EU level.121 This conditionis intended to ensure that the subsidiarity principle isrespected, which implies that REACH restrictions maynot be imposed with respect to risks that can better beaddressed at the national level, such as risks associatedwith specific processes in plants located in specificMember States,122 risks arising only in one or a smallnumber of Member States or highly variable risks thatare dependent on local conditions.123

Substantively, REACH does not limit the nature anddesign of restrictions. Although REACH references sub-stitution only in relation to authorization, a restrictionlikewise has substitution effects since some uses, orways in which a substance is used, are prohibited andthus will have to be substituted for. Depending on howa restriction is designed, such substitution pressure canbe very strong; for example, like authorization, arestriction may provide for a time-limited exemptionfor a particular use.124 In other words, assuming thatthere is evidence of risks that are not unacceptable andadequately controlled, a restriction can be designedin a way that closely resembles the results of REACHauthorization.

Under the REACH restriction regime, any restrictionproposal must cover information on alternatives,including information on the risks to human health and

the environment related to the manufacture and use ofthe alternatives,125 and establish that a restriction isindeed the most appropriate measure to address therisk concerned.126 To show that it is the most appropri-ate measure, an assessment is required of the effective-ness, practicality and enforceability of a restriction.127

Thus, although an unacceptable risk is found to exist, arestriction may not be the most appropriate measure toaddress this risk. Other measures could include harmo-nized classification and labelling or an occupationalexposure limit under workers’ safety law. The Commis-sion has some discretion in selecting the most effectiveand proportionate measure from two or more availableoptions.128

Finally, REACH requires that a restriction decision‘take into account the socio-economic impact of therestriction, including the availability of alternatives’129

and recommends that a Member State conduct a socio-economic analysis before submitting a restriction pro-posal.130 This information should be part of the AnnexXV dossier.

ProcedureThere are two procedures for the adoption of restric-tions: a regular procedure131 and a fast-track proce-dure.132 The regular procedure can be initiated by arequest from the Commission to ECHA to prepare arestriction proposal,133 by ECHA for any substance onthe Authorization List after the sunset date,134 or by aMember State.135 The final decision is adopted by theCommission in a comitology procedure following publicconsultation and opinions from ECHA’s advisory com-mittees.136 Restrictions are adopted by the Commissionpursuant to the regulatory comitology procedure withscrutiny.137 In its decision, the Commission may departfrom the opinions of ECHA’s Committees if it providesa detailed explanation.138 This explanation should beattached to its decision.139 The fast-track procedure maybe used to adopt a restriction with respect to carcino-genic, mutagenic or teratogenic/repro-toxic (CMR)substances in consumer products or uses.140 The proce-

117 REACH, n. 1 above, Article 68.1.118 Ibid., Articles 69.1–2 and 69.4.119 Ibid., Annex XV, section I.120 See also ECHA, ‘Guidance for the Preparation of an Annex XVDossier for Restrictions’ (June 2007), at 49.121 REACH, n. 1 above, Article 68.1.122 See ECHA, n. 120 above, at 51.123 See generally TFEU, n. 42 above, Article 5.124 See, e.g., REACH, n. 1 above, Annex XVII, entry 18a, secondcolumn, section 6(b).

125 Ibid., Annex XV, section II.3.126 Ibid.127 Ibid.128 See, e.g., CJEU, Case T-368/11 Polyelectrolyte Producers Groupand others v. Commission, [2013] ECR II-0000, at paragraph 88.129 REACH, n. 1 above, Article 68.1.130 Ibid., Article 68.1, first sub-paragraph, Annex XV, section II.3 andAnnex XVI.131 Ibid., Articles 68.1 and 69–73.132 Ibid., Article 68.2.133 Ibid., Articles 69.1 and 69.3.134 Ibid., Article 69.2.135 Ibid., Article 69.4.136 Ibid., Articles 70–72.137 Ibid., Article 68.1.138 Ibid., Article 73.3.139 Ibid., Article 73.1.140 Ibid., Article 68.2.



228

dure does not involve ECHA’s committees and islimited to the regulatory comitology procedure withscrutiny.141

Role of RestrictionThe restriction regime is designed to impose generallyapplicable restrictions on the manufacture and use ofsubstances. Unlike authorization, it is not limited toSVHCs.142 Further, while authorization requires use-specific applications that must be submitted by eachcompany,143 the restrictions imposed by REACH applydirectly and require nothing other than compliancewith the specific conditions stated in Annex XVII.Restriction has a long history dating back to the Mar-keting and Use Directive.144 This means that there issignificant experience with operating this programme.However, as discussed below, the allocation of admin-istrative burdens under the restrictions process maycorrelate negatively to Member State authorities’ incli-nation to propose restrictions, in particular in light ofthe other available regulatory pathways, includingauthorization. Like authorization requirements, restric-tions are enforced by the Member States.145

THE CHOICE BETWEENAUTHORIZATION, RESTRICTIONAND OTHER EU CHEMICAL LAWSAuthorization and restriction are the two ‘commandand control’ regimes for regulating chemical risk underREACH. The substantive standards and procedures forthe application of each of these instruments differ insignificant ways. For instance, the restriction regimerequires that the Commission prove that a substanceposes an unacceptable risk to human health or the envi-ronment,146 and the restriction must be tailored to thatparticular risk, and it may cover any stage in the lifecycle of a substance.147 In comparison, authorization ishazard-triggered, a low standard, limited to SVHCs asdefined,148 and prohibits any use except as authorized.Once a substance has been listed on Annex XIV, allplacing on the market and uses of that substance aresubject to authorization, and substitution is encour-aged, regardless of whether there is any exposure orrisk to human health or the environment.149

The significant differences between the programmessuggest that the choice between the two has substantialramifications. Remarkably, the EU authorities have not

effectively engaged with this issue. In a 2011 report onthe operation of REACH and the CLP Regulations,ECHA noted that ‘it is crucial to develop sufficientcommon understanding’ on when to initiate the autho-rization and restriction processes so as to enhance ‘theeffective use of resources in authorities and industry,legal clarity and predictability’.150 Unfortunately, theCommission did not advance this discussion in thecontext of the REACH implementation review. In its2013 general report on REACH, the Commission issilent on the issue151 and the Staff Working Documentnotes merely that that ‘there may be cases where it isappropriate to launch the two processes (authorizationand restriction) in order to target different uses’.152

Further, it should not be forgotten that REACH is notthe only framework for regulating chemical risk. Theremay be other laws or regulations that could offer analternative to a measure issued pursuant to REACH.

Coordination between Authorization andRestriction under REACHUnfortunately, REACH does not provide a set of rules toguide the choice between restriction and authoriza-tion.153 However, REACH is not entirely silent on therelation between the two regimes. It is clear, forinstance, that the two programmes are not designed asmutually exclusive, which raises the issue as to whenboth may be applied to the same substance. REACHattempts to enhance predictability and consistency ofregulation through the following provisions:

• A restriction cannot interfere with the authorizationregime. Once a substance is added to the Authori-zation List, a restriction can no longer target risksresulting from the hazard for which it has been sub-jected to authorization, insofar as such risks arisefrom the use of the substance on its own, in amixture or its incorporation in an article.154

• On the other hand, a restriction may cover risksresulting from other hazards or from the presenceof the substance in an article, as well as any otherrisks.155 The idea is that a restriction could supple-ment the authorization regime to cover importedarticles not subject to authorization.156 REACHtherefore requires that ECHA consider whether

141 Ibid.142 Ibid., Article 68.1.143 Ibid., Article 56.1.144 Directive 76/769/EEC, n. 22 above.145 See also Milieu Ltd, n. 3 above.146 REACH, n. 1 above, Article 68.1.147 Ibid., Articles 67.1 and 68.1.148 Ibid., Article 57.149 Ibid., Article 55.

150 ECHA, The Operation of REACH and CLP (ECHA, 2011), at 32.151 European Commission, General Report on REACH, Report fromthe Commission to the European Parliament, the Council, the Euro-pean Economic and Social Committee and the Committee of theRegions in Accordance with Article 117(4) of REACH and Article46(2) of CLP, and a Review of Certain Elements of REACH in Linewith Articles 75(2), 138(2), 138(3) and 138(6) of REACH, COM(2013)49.152 See European Commission, n. 14 above, at 71–72.153 See ECHA, n. 3 above.154 REACH, n. 1 above, Article 58.5.155 Ibid., Article 58.6.156 Ibid., Articles 56.1 and 69.2.



229

risks resulting from a substance subject to authori-zation should be restricted once the sunset date haspassed.157

• An authorization may not be granted if it wouldconstitute a relaxation of a restriction.158 Further, asubstance must be taken off the Authorization Listif all uses have been prohibited under REACH orother EU legislation,159 as any authorization wouldbe a relaxation of such a prohibition. Implicitly, thisprovision is an acknowledgment that restrictionmay also have strong substitution effects.

• An exemption from authorization may be estab-lished based on existing restrictions. If chemicalrisks are ‘properly controlled’ under existing specificEU legislation imposing minimum safety require-ments, an exemption from authorization for the sub-stance concerned may be granted.160 A REACHrestriction or a restriction under other EU legislationcould meet the condition of minimum safety, andthus justify an exemption from authorization.

From a policy perspective, the significant differencesbetween authorization and restriction and the generalprinciples of EU law (discussed below) demand that ineach case the question be answered whether authoriza-tion would be more appropriate than alternative regu-latory instruments to address a particular risk. Indeed,the Member States and the Commission increasinglyconduct an assessment of available risk managementoptions before a substance is added to the CandidateList.161 The Commission has significant discretion onwhether and when to list a substance on the Authoriza-tion List; it may decide not to list a substance recom-mended for listing by ECHA if it believes thatrestriction would be a more appropriate instrument.162

Although the Commission left open the possibility thatsome of these substances will also be subject to restric-tions,163 no clear policy on such double regulation hasyet been articulated. The key policy questions still needto be answered. First, in which cases could authoriza-tion target only some of a substance’s uses, whilerestrictions would address other uses that are exemptfrom authorization? With respect to imported products,double regulation surely needs to be considered: ifauthorization applies to a substance’s use in domesticmanufacture of products, the question arises whether arestriction should also be imposed with respect toimported products. Second, with respect to the sameuse, could authorization address some specific risks

(e.g., by listing a substance only for human hazard onthe Authorization List) while restrictions target otherrisks (e.g., environmental risks) associated with thatuse? Are there realistic scenarios where such risk-baseddifferential regulation would be appropriate? Third, isdouble or differentiated regulation appropriate? Most,if not all, that can be achieved through authorizationcan also be achieved through restriction. The questiontherefore arises whether authorization should play thedominant role in the regulation of SVHCs that the leg-islature may have had in mind.164,165 Given thesubstantial administrative costs associated with autho-rization, there would have to be clear, proven advan-tages. If the same results can be achieved throughrestriction, it is likely to be the less restrictive measureand double regulation can be avoided. An issue likely toarise in this context is the effect on substitution pres-sure. Again, a restriction, which can impose a substitu-tion plan and provide for time-limited exceptions, is notnecessarily inferior to authorization in this respect.Whether review of individual substitution plans, asrequired in connection with authorization, will haveadded value is an open question.

Coordination with Other Chemical LawsREACH operates independently from other chemicalregulatory programmes, subject to limited exclusions toavoid double regulation.166 For instance, active sub-stances used in biocides and plant protection productsare regarded as ‘being already registered’ and thusexempt from registration, but not from the otherREACH programmes.167 The idea is that overlapsbetween regulations may result in inconsistencies oreven conflicts, or in unnecessary regulatory burdens.The pre-existing regulations for which REACH exemp-tions have been established, which are more limited inscope and more specific than REACH, are deemed toaddress the related risks adequately. The wording of theexemptions differs. In some cases, the exemption doesnot reference the particular regulations concerned (e.g.,transport of dangerous chemicals168); in other cases, theexemption refers to material as defined by a particularlaw (e.g., waste169), or a product falling within the scopeof a particular law170 or ‘in the finished state, intended forthe final user’.171

Restrictions continue to be imposed under other legis-lation of both pre- and post-REACH origin. The Direc-tive on the restriction of the use of certain hazardoussubstances in electrical and electronic equipment

157 Ibid., Article 69.2.158 Ibid., Article 60.6.159 Ibid., Article 58.7.160 Ibid., Article 58.2.161 See European Commission, n. 87 above, at 8–9.162 An example is cobalt compounds, which were recommended forlisting by ECHA but not listed by the Commission. Regulation 348/2013/EU, n. 104 above, recital 12.163 See European Commission, n. 14 above, at 72.

164 REACH, n. 1 above, Article 57(a)–(e) versus (f).165 See also L. Bergkamp and N. Herbatschek, n. 31 above, at 60–62.166 Ibid.167 See also Milieu Ltd, n. 3 above.168 REACH, n. 1 above, Article 2.1(d).169 Ibid., Article 2.2.170 E.g., ibid., Article 2.5(a).171 Ibid., Article 2.6.



230

(RoHS Directive),172 which imposes maximum concen-tration levels on six chemicals in electrical and elec-tronic equipment, was adopted prior to REACH, andthere are now plans to expand the number of chemicalsregulated under it.173 Likewise, the Directive on end-of-life vehicles (ELV Directive),174 which aims at makingvehicle dismantling and recycling more environmen-tally friendly including by restricting the use of fourchemicals in vehicles,175 is a pre-REACH regime.

After the adoption of REACH, the EU legislature con-tinued to adopt product group-specific chemicalrestrictions for products such as biocides,176 plant pro-tection products,177 toys178 and cosmetics,179 and a pro-posed regime for medical devices is pending.180 Theseregulatory measures have been adopted on an ad hocbasis as an integrated framework for regulatingproduct-related chemical risks does not exist. There is aserious question as to whether this is an appropriatepractice since the scientific and policy analysis con-ducted by the EU legislature before such restrictions areadopted is generally less rigorous than the REACHRegulation requires. For instance, when the EuropeanParliament recently adopted restrictions on CMRs andendocrine disrupters in medical devices, it did not haveany risk assessment report available. In general,REACH offers a better framework for imposing therestrictions sought by the EU legislature. If the intent isto circumvent the demands imposed by the REACHrestrictions procedure, the issue arises how the qualityof decision making on chemical restrictions outside theREACH context is guaranteed. There is reason forconcern since no general framework for science-baseddecision making exists outside of REACH.181

Under current EU law, EU authorities thus may have achoice of regulatory frameworks to manage risks asso-ciated with the manufacture or use of a chemical sub-stance. The available regulatory frameworks may differsubstantially in terms of scope (e.g., a concentrationlimit for a substance under the RoHS Directive willapply only to electronics). In addition, the specific tools

available under each of these frameworks may alsodiffer from one instrument to the next (e.g., a restric-tion on a specific technological use in multiple productscan be imposed under REACH, but not under the RoHSDirective). If a chemical risk is best regulated throughone specific tool, the lack of availability of that specifictool under a regulatory framework may render thatregulatory framework less appropriate. Further, theprocedure for adopting a chemical risk managementmeasure also varies between regulatory frameworks; insome cases, a legislative amendment would be requiredfor a particular restriction (e.g., a labelling requirementunder the medical devices legislation). These and otherconsiderations should come into play when the authori-ties decide on the pathway for regulating a particularchemical risk.

A policy choice of regulatory framework arises not onlyin relation to chemical restrictions targeting specificproduct groups, but also with respect to chemical risksin the workplace. Since substances in the workplacefrequently give rise to a need for risk management andREACH requires that such risks be identified andmanaged,182 this issue is bound to come up regularly.Indeed, in the case of occupational exposure, the samerestriction (e.g., a binding occupational exposure limit)could be adopted either under REACH or under theworker safety legislation. Unlike for product-relatedchemical risk, the EU has established an integratedoverarching framework for addressing occupationalrisks arising from chemicals. The decision makers andapplicable procedures and legal and practical aspects,however, may differ significantly. The following consid-erations may be relevant to the choice between regula-tion under REACH or regulation under the EUoccupational health legislation:

• Specificity and level of integration of frame-work: The EU occupational safety legislation pro-vides a framework that has been designedspecifically for the purpose of offering integratedprotection of the health and safety of workersexposed to occupational risk, including risks arisingfrom chemical agents. If a substance poses onlyoccupational risk, the occupational legislation gen-erally is the most suitable and appropriate regula-tory framework. This appears to be also theCommission’s position.183

• Procedure: Under REACH, the draft restrictionmust go through a lengthy procedure at ECHAbefore the Commission adopts the final decision bycomitology in the form of a regulation.184 Under theEU occupational legislation on chemicals at work

172 Directive 2011/65/EU, n. 27 above.173 Umweltbundesamt, Study for the Review of the List of RestrictedSubstances under RoHS2 (Umweltbundesamt, 2014).174 Directive 2000/53/EC, n. 26 above.175 The Directive restricts the following chemicals: mercury, cadmiumand hexavalent chromium. Ibid., Article 4.2(a).176 Regulation 528/2012/EU Concerning the Making Available on theMarket and Use of Biocidal Products, [2012] OJ L167/1.177 Regulation 1107/2009/EC Concerning the Placing of Plant Protec-tion Products on the Market and Repealing Council Directives 79/117/EEC and 91/414/EEC, [2009] OJ L309/1.178 Directive 2009/48/EC, n. 28 above.179 Regulation 1223/2009/EC, n. 30 above.180 European Commission, Proposal for a Regulation of the EuropeanParliament and of the Council on Medical Devices, and AmendingDirective 2001/83/EC, Regulation (EC) No 178/2002 and Regulation(EC) No 1223/2009, COM(2012) 542.181 For a discussion on the role of science in REACH, see N.Herbatschek, L. Bergkamp and M. Mihova, n. 88 above, at 165–170.

182 E.g., REACH, n. 1 above, at Annex I, section 5.1.1.183 ECHA, ‘Minutes of the 21st Meeting of the Committee for Socio-economic Analysis, 10–12 December 2013’, SEAC/M/21/2013FINAL, at 6.184 REACH, n. 1 above, Articles 68–73.



231

(Directive 98/24), a draft decision prepared by theCommission is first assessed by the Scientific Com-mittee on Occupational Exposure Limits (SCOEL)and then goes through the legislative procedure foradoption of a directive,185 which, once adopted,must be transposed into national law. This proce-dure may take more or less time than a procedurefor the adoption of a REACH restriction, dependingon the facts and circumstances.186

• Effect, subsidiarity and legitimacy: A REACHrestriction is directly effective, while a measure pur-suant to Directive 98/24 has to be transposed by theMember States.187 The occupational law’s decision-making process and transposition, however, maybetter reflect the EU’s governance philosophy andthe subsidiarity principle188 than a REACH restric-tion. It may also enhance the measure’s legitimacyand the Member States’ ‘buy-in’ or ownership.Further, the transposition process might advancethe Member States’ understanding of what effectiveimplementation in their territories requires.

Important issues about the relationship betweenREACH and other chemicals-related laws have arisenin connection with Article 58.2 of REACH, which allowsexemptions from the authorization process. In theprocess of listing trichloroethylene on Annex XIV, forwhich there already is an indicative occupational expo-sure level (iOEL; see below), the question came up howREACH relates to worker protection legislation.189 TheCommission’s Directorate-General Environmentargued that worker protection legislation cannot serveas a basis for an exemption because it is not efficientlyimplemented.190 With the exception of Romania, theMember States agreed with the Commission andapproved the listing of trichloroethylene on Annex XIVwithout exemptions.191 Even though the listing wasagreed without exemptions this case is likely to influ-ence future exemptions under REACH authorization.In a 2013 report, the Commission recognized the factthat there is limited experience with Article 58.2exemptions.192 The Commission made specific refer-ence to the workers protection legislation, echoing thetrichloroethylene situation.193 Indeed, in these andother cases, the Commission should grant exemptionsfrom authorization, as appropriate.

Another example of overlap between REACH andsector-specific legislation that would require exemp-tions from REACH authorization is the case ofcadmium. If cadmium, currently on the CandidateList,194 is subsequently listed on Annex XIV withoutexemptions, it would not be available for any use unlessauthorization is granted unless, perhaps, one of theexemptions for cadmium granted under sector-specificlegislation such as the RoHS Directive applies. Wouldcompanies be able to invoke these exemptions, orwould they need to request authorization? Recognizingthis overlap between RoHS exemptions and REACHauthorization, the Commission has emphasized a needto consider exemptions from REACH authorization.195

It is to be hoped that, going forward, the authorities willadopt a more practical view on exemptions from autho-rization. This would do justice to the intent of theREACH legislation: REACH provides for possibleexemptions because some uses are, or can be, subjectedto efficient risk management measures pursuant to EUsector-specific or other regulation, including nationalregulation.

The Choice between Restrictionand AuthorizationBecause REACH is a collection of several stand-aloneregulatory regimes that loosely hang together, one andthe same risk can be targeted by two or more evalua-tive regimes at the same time (e.g., in connection withsubstance evaluation and Candidate Listing) and twoor more risk management measures (e.g., restrictionand authorization). This can occur even before a sub-stance evaluation has been completed, or a risk assess-ment and policy analysis have been conducted. Putdifferently, REACH does not provide an adequateframework for making risk management decisions.Logically, risk assessment and policy analysis shouldprecede any proposal for subjecting a substance torestrictions or authorization. Further, specific rulesand criteria on how the authorization and restrictionsregimes relate are not set out in the REACH Regula-tion. A robust framework for the selection of theappropriate risk assessment procedure and risk man-agement measure to address a specific risk would helpto structure the choice between the various toolsoffered by REACH and other legislation, and make theprocess more predictable.

Authorities have only recently started to focus on thequestion of whether the authorization or restrictionregime is the most suitable regulatory response to aparticular chemical risk.196 Undoubtedly, this debatewill continue for some years to come. The experiencethus far has shown that authorization is not necessarily

185 Directive 98/24/EC, n. 25 above, Article 3.4.186 Cf. ECHA, n. 183 above.187 Directive 98/24/EC, n. 25 above, Article 3.5.188 TFEU, n. 42 above, Article 5.189 See N. Herbatschek, L. Bergkamp and M. Mihova, n. 88 above, at159.190 This is non sequitur reasoning: if there is a problem with theimplementation of the occupational safety legislation, this issueshould be addressed but cannot serve as a justification for imposingmore requirements under other legislative regimes.191 Regulation 348/2013/EU, n. 104 above.192 See European Commission, n. 14 above, at 12.193 Ibid.

194 <http://echa.europa.eu/web/guest/candidate-list-table>.195 See European Commission, n. 14 above, at 12.196 See ECHA, n. 3 above.



232

http://echa.europa.eu/web/guest/candidate-list-table

the preferred approach for all substances that meet thecriteria for being classified as SVHCs or that are listedon the Candidate List or Annex XIV. There are severalaspects of the authorization and restriction regimesthat are relevant to making an informed choice.

First, in some cases, authorization is less comprehen-sive and effective than restriction. Authorization mustbe related to the property for which a substance waslisted on Annex XIV197 and cannot target all propertiesof a substance and the related risks.198 Although alluses of a substance listed on Annex XIV are subject toauthorization, including uses that do not expose theenvironment or human beings to the risks associatedwith the properties for which the substance was listed,the conditions attached to an authorization may notaddress the risks associated with other intrinsic prop-erties, even if these other risks are not adequately con-trolled.199 Thus, if a substance is listed on Annex XIVfor repro-toxicity, the authorization may be aimed atreducing the risks arising from repro-toxicity, but notrisks associated with other properties of the same sub-stance, such as liver toxicity. If a substance has mul-tiple dangerous properties, only one or some of whichcould be addressed in the authorization process,restriction may be a more effective option.200 Put dif-ferently, a REACH restriction can impose conditionstargeting properties of substances and risks arisingfrom uses of substances that cannot be covered underthe authorization regime. This flexibility of the restric-tions regime can be a major advantage with respect tocertain chemical risks, such as release of a chemicalfrom a product into the environment or residues of asubstance in an article.201

Second, in terms of administrative cost, authorization ismuch more costly and burdensome than generallybinding restrictions, if the results achieved are similar.This is so because each manufacturer or user of thesubstance concerned must file an individual applica-tion, which is an onerous process, and authorization issubject to expiration and possible renewal.202

Third, for authorization, the availability of alternativesis not considered at all when substances are listed; it isconsidered only late in the process in relation to indi-vidual authorization applications.203 Consequently,substances may be subjected to the expensive and bur-densome authorization programme, even though theycould more efficiently be handled under the restrictionsregime, for instance, when the exceptions to a restric-tion can be identified upfront for all entities concerned.In connection with authorization, REACH does notrequire that the availability of substitutes be consid-ered, and a positive finding be made that authorizationis the preferred regulatory measure, before substancesare added to Annex XIV.

Fourth, authorization decreases legal certainty for thecompanies concerned, which are already exposed to thestigmatizing effects of the Candidate List.204 Onceissued, authorization can be withdrawn at any point intime if new information becomes available or circum-stances change,205 subject only to the general propor-tionality principle.206 The interest of legal certainty andincentives for investment, however, would tend tosuggest that authorizations should have a minimumvalidity period during which they cannot be reviewed,altered, or withdrawn.

Fifth, the authorization regime is ambiguous. Article56.1 of REACH stipulates that ‘a manufacturer,importer or downstream user shall not place a [AnnexXIV] substance on the market for a use or use it himselfif that substance is included in Annex XIV’, unlessauthorization for that use has been granted or anexemption applies. The problem is that the concept of‘placing on the market for a use’ (emphasis added) isambiguous; the manufacturer sells the substance to acustomer, and the customer decides how the substanceis used. If the substance is intended for export, theapplication of the provision becomes irrational.

Sixth, authorities may impose a restriction only if theycan demonstrate that there is an unacceptable risk thatneeds to be addressed on a EU-wide basis, while theyneed to demonstrate only that a substance is an SVHCon the Candidate List to add it to the Authorization List.As discussed above, an assumption that SVHCs arebetter regulated through authorization is erroneous.

Seventh, in some cases a substance may be subjected tomultiple regulatory measures, but the timing of adop-tion of each regulatory measure has an impact. Arestriction under REACH or another EU law adoptedprior to authorization could justify a general exemption

197 REACH, n. 1 above, Article 60.2. See also ECHA, n. 3 above, at35.198 REACH, n. 1 above, Article 56.1.199 Note the discrepancy here: all uses of a listed substance aresubject to authorization, irrespective of whether a specific useexposes the environment or human beings to the risks associatedwith the properties for which the substance was listed, but authoriza-tion must be granted if the risks associated with these properties areadequately controlled, even if the use concerned involves otheruncontrolled risks not associated with such properties.200 Restrictions may cover the full range of risks that are notadequately controlled. REACH, n. 1 above, Article 67.1.201 C. Schulte, L. Tietjen, A. Bambauer and A. Fleischer, ‘Five YearsREACH: Lessons Learned and First Experiences. I – An Authorities’View’, 24:31 Environmental Sciences Europe (2012).202 REACH, n. 1 above, Articles 60.8 and 61.

203 Ibid., Articles 60.4–5 and 61.2(b).204 L. Bergkamp, ‘Does REACH Present a Business Opportunity?’, in:L. Bergkamp, n. 2 above, 396, at 400.205 REACH, n. 1 above, Article 61.206 Ibid., Article 61.3.



233

from authorization. Once a substance is included in theAuthorization List, however, the adoption of a restric-tion of the same substance under REACH is subject toconditions. Further, REACH provides that a generalexemption from authorization based on existing EU lawmay be adopted upon inclusion of a substance on theAuthorization List,207 but this rule does not apply if suchEU law is adopted afterwards. The Commission couldarguably amend the Authorization List accordingly, butthere is no requirement for the Commission to do so.

In practice, differences with respect to jurisdiction, pro-cedure and administrative burdens exercise substantialinfluence on the choice between authorization andrestriction. In particular, the issue of administrativeworkload seems to be a consideration in decidingwhether to pursue authorization or restriction. 208 Com-pared to dossiers for Candidate Listing, very fewrestriction dossiers have been proposed so far.209 Areason for this may be that the administrative burdenassociated with the preparation of a restriction dossierfalls on the generally short-staffed Member Stateauthorities, which may have difficulty compiling allinformation necessary to relate registered uses with endproducts (articles). In addition, Member States may notall have the resources necessary to conduct an analysisof risk management options, an analysis of substitutesand a socioeconomic analysis – all of which arerequired for the preparation of a restriction dossier.210

More than any other factor, Member States’ reluctanceto shoulder these administrative burdens may skew theprocess towards authorization.

From a substantive law perspective, assuming proveninadequately controlled risk arising from an SVHC,both authorization and restriction may be able toachieve the same regulatory ends. In fact, it is hard tothink of an authorization requirement that cannot sub-stantively be translated into a restriction. As a matter oflaw, authorization differs from restriction with respectto the assignment of the burden of proof of adequaterisk control (which is imposed on regulated entitiesversus the government, respectively) and the additionalconditions for authorization of PBT/vPvB and non-threshold substances that ‘socio-economic benefits out-weigh the risk to human health or the environmentarising from the use of the substance’ and ‘there are nosuitable alternative substances or technologies’.211 Pro-cedurally, a restriction, like authorization, couldrequire a substitution plan, but it could probably not

provide for case-by-case assessment of such plans.Whether these differences limit the choice betweenauthorization and restriction, however, depends on thespecific facts of each case. In many cases, restrictionmay serve as an effective and equivalent alternative forthe authorization of an SVHC.

If the risks associated with an SVHC are adequatelycontrolled in reality (which is not necessarily reflectedin registration dossiers), there is no ‘unacceptable risk’and restriction is not an option. This standard is met ifthe risk characterization ratio212 exceeds 1 and, thus,actual exposure levels exceed the levels that correspondwith the ‘derived no effect level’ (DNEL) or ‘predictedno effect concentration (PNEC).213 In other words, itrequires ‘[e]vidence . . . that implemented risk manage-ment measures . . . are not sufficient’.214 If the risk char-acterization ratio does not exceed 1 and evidence ofinsufficient risk management is not available, authori-zation is the only regulatory option. This raises thequestion of whether such regulation is necessary sinceall risks are adequately controlled. The EU legislaturealready answered this question, determining thatauthorization may be imposed also where risks are con-trolled. Under these circumstances, however, all usersof the SVHC concerned will receive authorization,215

except if the SVHC is a PBT/vPvB or a ‘no threshold’substance.216 The end result could be similar under an‘authorization-resembling’ restriction approach, if itwere permitted. The main difference would be that arestriction would impose one generally applicable set ofterms and conditions, while authorization wouldimpose terms and conditions that could, to someextent, be individualized (e.g., with respect to the dura-tion). This analysis exposes a deficiency in the wayREACH has been devised: if the relative benefits of suchindividualization are outweighed by the relative disad-vantages, a generally binding restriction may be thepreferred option, but REACH does not empower the EUto impose it absent unacceptable, uncontrolled risk.

To conclude, the choice between authorization andrestriction is such a convoluted and skewed processbecause REACH suffers from fundamental designflaws. It does not establish an independent, coherentand unbiased framework for chemical risk assessmentand policy analysis of alternative regulatory options.Instead, it establishes separate, diverging, chieflyhazard-based procedures that employ inconsistent

207 Ibid., Article 58.2.208 See ECHA, n. 3 above, at 44 and 53.209 As of 27 June 2014, more than 160 dossiers for Candidate Listingand less than 30 restriction dossiers have been prepared. See<http://echa.europa.eu/addressing-chemicals-of-concern/registry-of-intentions>.210 See N. Herbatschek, L. Bergkamp and M. Mihova, n. 88 above, at152.211 REACH, n. 1 above, Article 60.3–4.

212 Ibid., Annex XV, under II.3.213 Ibid., Annex I, under 6.4.214 Ibid., Annex XV, under II.3.215 Authorization would have to be granted to each applicant pursuantto the adequate control route. Ibid., Article 60.2.216 If the SVHC concerned is a PBT/vPvB or a ‘no threshold’ sub-stance, the end result would be that uses will be authorized only iftheir socio-economic benefits outweigh their risks and no suitablealternative is available. Ibid., Article 60.4.



234

http://echa.europa.eu/addressing-chemicals-of-concern/registry-of-intentions

http://echa.europa.eu/addressing-chemicals-of-concern/registry-of-intentions

safety standards.217 It is no wonder that REACH has notconsistently produced predictable, balanced and rea-sonable results.

RMO AnalysisThe prioritization of SVHCs is the first step in theauthorization process.218 There are currently over 1,500substances that meet the SVHC criteria based on theirclassification.219 ECHA indicated that it will be able toprocess no more than 20–30 substances per year. Thus,this process raises issues of prioritization for listing onthe Candidate List. Due to the allocation of powers, theprocedural specificities and the politics, the prioritiza-tion process has been plagued by a lack of transparencyand predictability. In response to pressure to make theSVHC prioritization process more transparent, theCommission launched an initiative to better coordinatethe Member States’ intentions in 2012.220

As part of this initiative, the Commission recommendedin its SVHC Roadmap to 2020 that an analysis of riskmanagement options be completed before a formal pro-posal for inclusion on the Candidate List is submit-ted.221 The overall aim of the SVHC Roadmap is todefine the approach to prioritizing substances for inclu-sion on the Candidate List. Pursuant to REACH, onlyCMRs, PBTs, vPvBs and ‘substances of equivalentconcern’, such as endocrine disruptors, may beincluded on the Candidate List.222 The RMO is intendedto identify the best regulatory option to manage therisk, either through REACH or outside of it.223 In thiscontext, the main purpose of the RMO analysis wouldbe ‘to either document that the authorization route issuggested for an (potential) SVHC or that there arespecific reasons for an SVHC which would overrulegoing the authorisation route to achieve its substitu-tion’. This analysis could lead to the conclusion that ‘adifferent risk management route (e.g. restriction oraction under a different legislation)’ or ‘no action’ is thepreferred option. The Roadmap acknowledges thatRMO analysis224 should be at the heart of any prioriti-zation decision of a competent authority.225 However,the Commission has also stated that a total of 440 sub-stances should be listed by 2020, of which 80 sub-stances should be assessed by the end of 2014.226

Unfortunately, this political commitment creates a riskthat, despite the move towards RMO analysis, unsounddecisions will be made under time pressure.

An RMO analysis is intended to assist in determiningwhether a substance is a ‘relevant SVHC’ in the contextof the SVHC Roadmap. Relevant SVHCs are substancesthat: (i) meet the criteria for Candidate Listing; (ii) areproduced and/or used in Europe in relevant quantities;(iii) have been registered for non-intermediate uses;and (iv) do not pose solely a risk that is not adequatelycontrolled and should thus be subjected to a REACHrestriction.227 For reasons that are unclear, the prioriti-zation criteria for Annex XIV listing (PBT/vPvB prop-erties, wide dispersive use and high volumes) play norole in this process. The main purpose is to determinewhether risk management measures are necessary, andif so, ‘to identify the most appropriate instrument toaddress concerns’,228 which may be authorization,restriction, another measure or a combination thereof.Thus, RMO analysis involves policy analysis. A moreexpansive version of policy analysis is required in con-nection with a restriction proposal, as discussed above.

Although not required by REACH, RMO analysis is des-tined to become standard practice in connection withCandidate Listing.229 In some cases, RMO analysisappears to be more than a formality (e.g., the Commis-sion has rejected proposed authorization of certaincobalt compounds on the ground that restriction wouldbe more appropriate),230 but the current RMO analysisprocess leaves much to be desired. For instance, thedocumentation generated in the process is made avail-able only to Member State authorities and the Commis-sion; only a ‘public version’ of the conclusions ispublished.231 There is no transparency and no publicconsultation, except if the evaluating competentauthority decides otherwise. The implementation of theSVHC Roadmap will address these concerns only par-tially.232 Further, from a substantive perspective, theRMO analysis falls far short of fledged policy analysis –for instance, it does not involve an analysis of the costsand benefits of various policy options.233 Political con-siderations may come into play,234 and are able totrump RMO analysis. In August 2012, the Commissionasked ECHA to prepare 37 dossiers for substances to belisted on the Candidate List by December 2012.235 Due

217 A.D.K. Abelkop, Á. Botos, L.R. Wise and J.D. Graham, ‘How CanREACH Be Improved?’, in: L. Bergkamp, n. 2 above, 390.218 REACH, n. 1 above, Article 59.219 See N. Herbatschek, L. Bergkamp and M. Mihova, n. 88 above, at152.220 Ibid., at 154.221 See European Commission, n. 87 above.222 REACH, n. 1 above, Article 57.223 See European Commission, n. 87 above, at 3.224 RMO analysis is coordinated by ECHA, while the Commissionhandles the policy issues. ECHA, ‘SVHC Roadmap to 2020: Imple-mentation Plan’ (9 December 2013), at 15.225 See European Commission, n. 87 above, at 3.226 Ibid., at 14.

227 Ibid., at 8–9.228 See ECHA, n. 224 above, at 5.229 See European Commission, n. 87 above.230 Regulation 348/2013/EU, n. 104 above, recitals 11–12.231 See ECHA, n. 224 above, at 18.232 Ibid., at 17–19.233 See European Commission, n. 87 above, at 9.234 A. Hanschmidt, M. Lulei and A. Paetz, ‘Five Years REACH:Lessons Learned and First Experiences – An Industry’s View’, 25:19Environmental Sciences Europe (2013), at 6.235 See N. Herbatschek, L. Bergkamp and M. Mihova, n. 88 above, at154.



235

to the short timeline, ECHA did not manage to produceRMOs for these substances and Annex XV dossierswere published without proper analysis, contrary to theCommission’s own recommendation.236 Both MemberStates and other stakeholders expressed disappoint-ment with this deficiency.237 The future will tell whetherthis was a one-off incident.

A major problem with the current RMO process is thatit is a rudimentary version of policy analysis with atendency to be conclusory, rather than analytical, and aslant towards authorization. A recent case238 shows howbasic and unenlightening RMO analysis can be. Thislack of thoroughness is at least in part due to the factthat stakeholders are not consulted, which is a seriousproblem if the information required for a sound analy-sis is not publicly available. No serious attempt is madeto compare risk management options, and to the extentany such analysis is produced, it is unable to informsound decision making. If RMO analysis is to be a foun-dation for rigorous policy making, the authoritiesshould at a minimum determine and document in detailwhat risk is to be addressed, what the objectives of riskmanagement should be, whether any regulatorymeasure is required to achieve those objectives, and ifso, what the costs and benefits are of the availableoptions, and which of these options would be the mostappropriate measure and why. The ‘no action’ optionshould be seriously considered, where appropriate. Ifaction is required, a comparison of possible regulatorymeasures should be based on cost-benefit analysis,which could build on the criteria provided under AnnexXV. These criteria include: effectiveness;239 practical-ity;240 monitorability;241 and a favourable socio-economic assessment.242 Other relevant considerationsinclude: risk/risk tradeoffs;243 regulatory consistency(e.g., how would a new measure compare in terms ofstringency to another comparable existing measure);and uncertainties (e.g., are the risk profiles of substi-tutes fully understood?).

The Role of General Principles of EU LawAlthough neither REACH nor other EU legislationimpose a robust general framework for choosing themost appropriate regulatory instrument to address aspecific chemical risk, provisions of general EU law are

relevant to this choice. Specifically, the general prin-ciples of EU law define the boundaries within whichthis choice has to be made. The most relevant generalprinciples include: the principle of equal treatment; theproportionality principle; and the obligation to statereasons. We will deal with each of these in turn.

Under the principle of equal treatment,244 persons inthe same situation must be treated in the same wayunless a different treatment is objectively justified (andnot arbitrary).245 This principle also prohibits differentsituations from being treated similarly without objec-tive justification.246 REACH requires that each sub-stance be analyzed and tested, as necessary, todetermine its hazards and risks, irrespective of howsimilar the substance is to another substance that hasalready been analyzed and tested.247 Under certain con-ditions, REACH permits that substances be groupedinto a category if they are structurally similar and havesimilar intrinsic properties. In some cases, however, theequality principle requires that substances be treated asa group, even if they do not meet the criteria for acategory. For example, when considering CandidateListing, ECHA should assess very hazardous substancesthat are deemed substitutes as a group at the same time.In its SVHC Roadmap, the Commission has adopted acategory approach that covers intermediates, whichwould otherwise not be considered.248 Further, theequality principle requires that substances that aresimilarly positioned be treated equally (e.g., substancesthat are substitutable and have similar properties). If,say, three substances are substitutable and have similarhazardous properties, a regulatory measure targetingonly one of them would unfairly impose regulatoryburdens on the manufacturer or users of that particularsubstance, thus providing an unfair advantage to themanufacturers or users of the other two substances.

The proportionality principle249 requires that ‘measuresadopted by Community institutions do not exceed thelimits of what is appropriate and necessary in order toattain the objectives legitimately pursued by the legis-lation in question; when there is a choice betweenseveral appropriate measures recourse must be had tothe least onerous, and the disadvantages caused mustnot be disproportionate to the aims pursued’.250 Themeasure must thus be: suitable or reasonably likely to

236 Ibid., at 154.237 International Lead Association, ‘Commission Undermines Credibil-ity of REACH Candidate List’ (21 August 2012).238 Denmark, ‘Annex XV Report – Proposal for Identification of aSubstance as a CMR 1A or 1B, PBT, vPvB or a Substance of anEquivalent Level of Concern – Sodium Peroxometaborate’ (February2014), at 24.239 REACH, n. 1 above, Annex XV, section 3(i).240 Ibid., Annex XV, section 3(ii).241 Ibid., Annex XV, section 3(iii).242 Ibid., Annex XV, section 3.243 See J.D. Graham and J.B. Wiener, n. 103 above.

244 Charter of Fundamental Rights of the European Union, [2000] OJC364/1, Article 20.245 ECJ, Case 20/71, Sabbatini v. European Parliament, [1972] ECR345.246 E.g., ECJ, Case 14/59, Sociétés des Fonderies de Pont-à-Mousson v. High Authority, [1959] ECR 215.247 REACH, n. 1 above, Articles 6.1 and 10.248 See European Commission, n. 87 above, at 8. Intermediates areexempt from the REACH authorization regime. REACH, n. 1 above,Article 2.8(b).249 TFEU, n. 42 above, Article 5.4.250 ECJ, Case C-15/10, Etimine, [2011] ECR I-06681, at paragraph124.



236

achieve its purpose; necessary to achieve its purpose;and the least detrimental.251 In the case of three similar,substitutable substances, for instance, a more onerousregulatory measure applied to only one or two of thesesubstances would not be the least detrimental and thusbe disproportional. The proportionality principleapplies not only to the choice between authorization,restriction or other regulatory measures, but also to thesubstantive and procedural requirements of any suchmeasure.

Decisions must state the reasons on which they arebased.252 The statement of reasons

must show clearly and unequivocally the reasoning of theinstitution which adopted the measure, . . . enable thepersons concerned to understand the full significance of andthe reasons for the measure at issue in order to enable themto safeguard their rights and . . . enable the EU judicature toexercise its powers of review of legality.253

The required extent and detail of the statement ofreasons is a function of the measure, its context and theapplicable legal rules.254 Where the persons concernedwere involved in the process by which a measure hasbeen adopted, the requirement to state reasons may becircumscribed as they may have had an opportunity tolearn about the reasons through their involvement.255

This obligation also requires that the decision maker,where relevant, explain why a particular measureshould be preferred over alternative measures.

ConclusionsThe REACH Regulation sets forth only a few provi-sions aimed at coordinating between authorizationand restriction. Likewise, the relationship betweenREACH and other chemical laws is not addressed inany significant detail. Specifically, REACH does notrequire policy analysis or a structured cost-benefit (orsimilar) analysis of alternative approaches to regulat-ing chemicals, except, to some extent with respect toproposed restrictions.

To reduce this gap, the Commission has developed theRMO analysis, which is aimed at confirming that theright regulatory tool is used to address the risks associ-ated with a specific SVHC. RMO analysis might helpsomewhat to ensure that the most appropriate regula-tory measure is applied to a particular substance or agroup of substances. However, the process is limited asit does not include policy analysis or cost-benefit analy-sis of alternative regulatory measures. Moreover, it is

biased towards authorization, as the default assump-tion is that authorization is the most appropriatemeasure.

THE CASE OF DIPOLARAPROTIC SOLVENTS

The problems arising under REACH in relation to thechoice of the most suitable regulatory instrument canbe illustrated with reference to dipolar aprotic solvents.These are powerful solvents that are used in specialtyapplications, such as certain polymerization reactions,which require polar solvents that do not donate protonsand thereby react. They can dissolve relatively largequantities of organic salts and are essential for chemicalreactions involving strong bases, which would reactwith protic solvents. This group of solvents includesNMP,256 DMAc257 and DMF;258 their properties andapplications are summarized below.

The case of dipolar aprotic solvents highlights thetrouble authorities may have in identifying the mostsuitable regulatory instrument in the absence of anoverarching framework for policy analysis and instru-ment selection. It also demonstrates that the ‘one sub-stance at a time’ regulatory approach does not workwell in cases where substances can substitute eachother. Eventually, however, the authorities identified acoherent, consistent and logical way forward for thisgroup of substances. As discussed below, the first pro-posed solution, authorization, appeared not to be themost suitable way to address risks arising from thesesubstances.

CHEMICAL PROPERTIESAND USESThe three dipolar aprotic solvents share several impor-tant similarities, including those regarding chemicalproperties, uses, volumes, classifications and risk pro-files. The polar nature of these substances enables themto act as a combined solvent and reaction catalyst.259 Inaddition, their high boiling point allows reactions torun at high temperature, without the need to operateunder pressure.260 Applications of these substances arelimited to industrial processes.261 Common uses includethe manufacturing of pharmaceuticals, agrochemicals,fine chemicals, coatings and fibres.262 DMAc, for

251 P. Craig and G. De Burca, EU Law: Text, Cases and Materials, 5th

edn (Oxford University Press, 2011), at 526.252 TFEU, n. 42 above, Article 296.253 E.g., CJEU, Case T-368/11, Polyelectrolyte Producers Group andothers v. Commission, [2013] ECR II-0000, at paragraph 101.254 ECJ, Case C-380/03, Germany v. Parliament and Council, [2006]ECR I-11573, at paragraphs 107–108.255 Case C-15/10, n. 250 above, at paragraph 116.

256 N-Methyl-2- pyrrolidone.257 N,N-Dimethylacetamide.258 N,N-Dimethylformamide.259 ECHA, ‘Background document for N,N-Dimethylacetamide(DMAC)’ (29 November 2012).260 Ibid.261 Ibid., at 6.262 Ibid., at 3–6.



237

instance, is used as solvent in a polymerization reactionof fibres such as acrylic and meta-aramid. The annualvolume of these substances that is estimated to beimported and/or manufactured in the EU rangesbetween 10,000 and 100,000 tonnes a year (t/y) forDMF, 10,000 and 50,000 t/y for NMP, and 11,000 and19,000 t/y for DMAc.263 All of these substances are highvolume and have been registered by the first REACHregistration deadline in December 2010. All three sub-stances are classified as toxic for reproduction, category1B.264 DMAc and DMF are also classified for acute tox-icity, category 4 (harmful if inhaled or in contact withskin);265 DMF as eye irritant, category 2;266 and NMP asspecific target organ toxicity – simple exposure (‘STOTSE’) (may cause respiratory irritation), category 3, andeye and skin irritant, category 2.267 All three substanceshave similar uses and risk profiles. In industrial appli-cations, they pose some occupational risk, which hasbeen addressed under EU worker safety legislation.They do not pose environmental risk,268 however, andthere is no evidence of any relevant exposure of con-sumers, and thus no consumer risk.269

Based on these similarities, in many applications, thesesubstances are interchangeable.270 They pose similarrisks, which, by and large, are limited to potential occu-pational exposure. The situation is different when itcomes to substitution of a dipolar aprotic solvent withanother substance that does not belong to this group.There are no generally accepted substitutes for thesesubstances in most applications, including in the pro-duction of fibres.271 Dimethyl sulfoxide is considered tohave higher permeability, thermal instability and lesssolvating power for polymer. Moreover, its decomposi-tion products cause an unpleasant odour.272 Sulfolanehas a higher melting point and thus is often operation-

ally impracticable.273 Acetone and acetonitrile have lesssolvating power.274 The lack of adequate substitutes insome applications should inform the choice betweenpossible regulatory tools.

REGULATORY MEASURESAPPLICABLE TO DIPOLARAPROTIC SOLVENTSBefore the recent regulatory activity under REACH, thethree solvents were already regulated pursuant to EUlegislation. The existing regulatory measures applicableto dipolar aprotic solvents are summarized below.

• Harmonized classification and labelling: As notedabove, these solvents are subject to similar harmo-nized classification and labelling. Under the pre-REACH legislation, they were similarly classified.275

• REACH restrictions: Annex XVII of REACH pro-vides that substances classified as toxic to reproduc-tion, which includes the three dipolar aproticsolvents, may not be supplied to the general publicif the individual concentration of the substance isequal to or exceeds a concentration threshold abovewhich a mixture including the substance is regardedas hazardous.276 The concentration threshold forDMAc and NMP is 5%.277 No specific concentrationthreshold has been specified for DMF, whichimplies that it is subject to the generic concentra-tion limit of the Dangerous Preparation Directive of0.5%,278 which will go down to 0.3% under the CLPRegulation as of 1 June 2015.279 Suppliers have tolabel these substances and mixtures visibly, legiblyand indelibly with the statement ‘restricted to pro-fessional users’.280

• Indicative OELs: The three substances are subjectto iOELs under EU occupational health and safetylegislation.281 These limits relate to presence in air,and are expressed in ‘milligrams per cubic metre ofair at 20°C and 101,3 KPa’ (mg/m3) and ‘parts permillion by volume in air (ml/m3)’ (ppm). There are

263 ECHA, ‘Draft Results of the Fifth Prioritization of the SVHCs on theCandidate List with the Objective to Recommend Priority Substancesfor Inclusion in Annex XIV’ (24 June 2013), at 4; ECHA, ‘Draft Resultsof the Fourth Prioritization of the SVHCs on the Candidate List withthe Objective to Recommend Priority Substances for Inclusion inAnnex XIV’ (20 June 2012), at 26 and 33.264 Regulation 1272/2008/EC, n. 5 above, Annex VI.265 Ibid.266 Ibid.267 Ibid.268 Organization for Economic Cooperation and Development(OECD), ‘SIDS Initial Assessment Report for 13th SIAM on N,N-dimethylacetamide (DMAC)’ (November 2001), at 4.269 The ECHA background documents on DMF and DMAc, and theNetherlands’ Annex XV restrictions report on NMP, provide informa-tion on consumer exposure to the three substances concerned. Therewere only four notifications to ECHA on the presence of these sub-stances in articles above 0.1% and ‘no consumer use of articles’. Thissuggests that there is no consumer exposure, if the small number ofnotifications cannot be explained by the exemption for uses alreadycovered by a registration dossier. ECHA, ‘Data on Candidate ListSubstances in Articles’, found at: <http://echa.europa.eu/information-on-chemicals/candidate-list-substances-in-articles>.270 See, e.g., ECHA, n. 259 above, at 11.271 Ibid.272 Ibid.

273 Ibid.274 Ibid.275 Directive 67/548/EEC, n. 19 above, Annex I.276 REACH, n. 1 above, Annex XVII, entry 30.277 Regulation 1272/2008/EC, n. 5 above, Annex VI.278 REACH, n. 1 above, Annex XVII, entries 28–30, column 2, junctoDPD, Annex II, Part B, Table VI.279 Regulation 1272/2008/EC, n. 5 above, Annex I, Table 3.7.2, fourthline, third column.280 REACH, n. 1 above, Annex XVII, entry 30.281 Directive 98/24/EC, n. 25 above; Directive 2000/39/EC of 8 June2000 Establishing a First List of Indicative Occupational ExposureLimit Values in Implementation of Council Directive 98/24/EC on theProtection of the Health and Safety of Workers from the RisksRelated to Chemical Agents at Work, [2000] OJ L142/47; Directive2009/161/EU of 17 December 2009 Establishing a Third List ofIndicative Occupational Exposure Limit Values in Implementation ofCouncil Directive 98/24/EC and Amending Commission Directive2000/39/EC, [2009] OJ L338/87.



238

http://echa.europa.eu/information-on-chemicals/candidate-list-substances-in-articles

http://echa.europa.eu/information-on-chemicals/candidate-list-substances-in-articles

eight-hour time-weighted average (TWA) limits andshort-term limits. The applicable limits for the threesolvents at issue are as follows (note that nationallimits may be lower282): NMP: 40 mg/m3 and10 ppm (eight hours TWA) and 80 mg/m3 and20 ppm (short term);283 DMAc: 36 mg/m3

and 10 ppm (eight hours TWA) and 72 mg/m3 and20 ppm (short term);284 and DMF: 15 mg/m3

and 5 ppm (eight hours TWA) and 30 mg/m3 and10 ppm (short term).285

• Legislative pressure for substitution: Because thesolvents are classified as toxic for reproduction,they must be substituted under both the Directive2010/75 and Directive 98/24.286 Pursuant to Direc-tive 2010/75, CMR substances, such as the threesolvents, must be substituted ‘within the shortestpossible time’. 287 Under Directive 98/24, employersare required to eliminate risks or reduce them to aminimum, with a preference for substitution.288

In short, the dipolar aprotic solvents are classified asrepro-toxic and covered by iOELs pursuant to EU occu-pational health and safety regulation. In addition, thesesubstances are earmarked for substitution undersolvent and occupational safety regulations.

EARLY STEPS IN THEPROCEDURE: DIVERGINGAPPROACHESDespite the existing regulatory measures and theabsence of any assessment of the adequacy thereof,efforts were launched to regulate dipolar aprotic sol-vents under the REACH Regulation. These efforts werefuelled by the classification of these substances asrepro-toxic under the CLP Regulation.289 The firstactions under REACH were listing on the CandidateList.290

Thus, in the initial stages of regulation pursuant to theREACH Regulation, these substances proceeded along

a similar regulatory path. ECHA earmarked all threesolvents for authorization, but then their regulatorypaths diverged. The divergence resulted from a pro-posal submitted by the Netherlands to impose restric-tions on NMP.291 The Dutch proposal involved areduction of the current occupational exposure limits to5 mg/m3 (eight-hour TWA) with peak exposures below10 mg/m3 (15 minute STEL).292 These limits would bemandatory, rather than indicative, as is currently thecase under the Directive 2000/39. In addition, compa-nies would be required to deploy preventive measuresto avoid dermal exposure in professional and industrialsettings.293 Further, manufacturers and industrial andprofessional users of NMP would have to be able todemonstrate at the request of the local authorities thatthey comply with these limits and obligations by main-taining an exposure monitoring programme.294 Thenext step in the procedure is the preparation andrelease of opinions by ECHA’s Risk Assessment andSocioeconomic Analysis Committees; the deadline forthe final opinions is September 2014.295

In addition to this proposal, the Netherlands submittedtwo proposals for ‘reclassification’ of NMP andDMAc.296 These proposals seek to withdraw the specificconcentration limits for classification of mixtures con-taining these substances as repro-toxic category 1B.Such withdrawal would result in the generic concentra-tion limit of 0.3% for repro-toxic category 1B297 becom-ing applicable to mixtures containing NMP andDMAc298 as of 1 June 2015.299 In the case of NMP, thislower concentration limit is intended to effectivelyeliminate the use of NMP in consumer applications, as‘[t]he use of NMP in concentrations below 0.3% willhave no functionality in the present consumer applica-tions’.300 Finally, Italy has signaled its intention tosubmit a restriction proposal for DMF in articles andfor professional uses by 16 January 2015.301

No analysis has been done of the effects of the proposeddiverging measures on the relative competitiveness ofthe three solvents and the potential for market distor-

282 Directive 98/24/EC, n. 25 above, Article 3.3.283 Commission Directive 2009/161/EU, n. 281 above.284 Directive 2000/39/EC, n. 281 above.285 Directive 2009/161/EU, n. 281 above.286 Directive 98/24/EC, n. 25 above.287 Directive 2010/75/EU of 24 November 2010 on Industrial Emis-sions (Integrated Pollution Prevention and Control), [2010] OJ L334/17, Article 58.288 Directive 98/24/EC, n. 25 above, Article 6.2.289 Regulation 1272/2008/EC, n. 5 above, Annex V, Table 3.1, entries606-021-00-7, 613-286-00-2 and 616-011-00-4.290 ECHA, ‘Inclusion of Substances of Very High Concern in theCandidate List (Decision of the European Chemicals Agency)’ (20June 2011); ECHA, ‘Inclusion of Substances of Very High Concern inthe Candidate List (Decision of the European Chemicals Agency)’ (19December 2011); ECHA, ‘Inclusion of Substances of Very HighConcern in the Candidate List (Decision of the European ChemicalsAgency)’ (18 December 2012).

291 See <http://echa.europa.eu/restrictions-under-consideration/-/substance/4506/search/+/term>.292 Regulation 1272/2008/EC, n. 5 above.293 The Netherlands, ‘Annex XV Restriction Report: Proposal for aRestriction – N-Methylpyrrolidone (NMP)’ (9 August 2013).294 Ibid.295 REACH, n. 1 above, Articles 70 and 71.1.296 The Netherlands, ‘CLH Report: Proposal for Harmonised Classifi-cation and Labelling – 1 – methyl-2-pyrrolidone’ (August 2013).297 Regulation 1272/2008/EC, n. 5 above, Annex I, Table 3.7.2, fourthline, third column.298 REACH, n. 1 above, Annex XVII, entry 30, column 3, section 1,second paragraph, second indent.299 Regulation 1272/2008/EC, n. 5 above, Articles 59.7 and 62.300 See The Netherlands, n. 296 above, at 12.301 ECHA, ‘Registry of Current Restriction Proposal Intentions’, foundat: <http://echa.europa.eu/registry-of-current-restriction-proposal-intentions/-/substance/5202/search/+/term>.



239

http://echa.europa.eu/restrictions-under-consideration/-/substance/4506/search/+/term

http://echa.europa.eu/restrictions-under-consideration/-/substance/4506/search/+/term

http://echa.europa.eu/registry-of-current-restriction-proposal-intentions/-/substance/5202/search/+/term

http://echa.europa.eu/registry-of-current-restriction-proposal-intentions/-/substance/5202/search/+/term

tion or disruption. Although REACH does not explicitlyrequire such analysis in connection with listing forauthorization, sound policy making in accordance withgeneral principles of EU law presupposes an under-standing of such market impacts. It is entirely conceiv-able that even minor differences in regulatorytreatment could have significant effects in the marketsfor interchangeable substances. Any such substitutioneffects may result in adverse impacts on the environ-ment and health and safety (e.g., where companiesswitch to a product that causes more serious adverseeffects). This suggests that a systematic process isrequired to ensure that such substances are assessedjointly before policy action is proposed. The Commis-sion’s RMO analysis approach should be adapted toincorporate this feature.

The current situation with respect to the regulation ofthe three solvents pursuant to the REACH authoriza-tion and restrictions regimes is as follows. DMAc iscovered by: ECHA’s 4th Annex XIV Recommendation,302

but its inclusion has been put on hold by the Commis-sion;303 a reclassification proposal;304 and REACH,Annex XVII, entry 30. DMF is: included in ECHA’s 5th

Annex XIV Recommendation;305 is covered byREACH, Annex XVII, entry 30; being considered forfurther restriction under REACH with a proposal to besubmitted by January 2015;306 and covered by a reclas-sification proposal.307 NMP is: not included in an AnnexXIV Recommendation; covered by REACH, AnnexXVII, entry 30; the subject of a reclassification pro-posal; and being considered for further restrictionunder REACH. Thus, all three solvents were initially ondiverging regulatory paths. The authorities realized,however, that this is a problem.

TOWARDS A CONSISTENTAPPROACHThe initial situation created a risk of inconsistent treat-ment of three very similar substances. While a restric-tion proposal is pending for NMP, DMAc and DMF arebeing reviewed for possible authorization. This diver-

gence is not the result of an objective policy decision; itis due to the fact that various authorities have the powerto initiate measures under REACH, and the significantoverlap between risk regulatory programmes underREACH and under other chemicals-related legislation.

When the facts were established, however, authoritiesbecame uncomfortable with the inexplicable incoher-ence in proposed regulatory approaches for these threesolvents.308 They seem to have realized that divergencewould have significant risk management consequencesand greatly affect the relative demand for each of thethree substances causing market distortion and pos-sible disruption, without there being any rational expla-nation. These commercial consequences may well havemade the authorities wary of proceeding along theinitial paths.

The initial incoherent approach is also questionablefrom policy and legal perspectives. From a policy per-spective, in light of the similarities in terms of chemicalproperties, uses and risk profiles, there would appear tobe no justification for treating the three solvents differ-ently. Any differential treatment would cause unjusti-fied distortions in the markets for these solvents byimposing significant additional costs for the industryon the use of DMAc and DMF in comparison to NMP.The protection of human health and the environmentwould not in any way be advanced by such differentialtreatment, and may be positively harmed – forinstance, due to unfavourable substitution effects.309 Inthe absence of any other justification for such differencein treatment, the need for another, coherent approachbecame apparent.

From a legal perspective, the general principles of EUlaw strongly suggest that the risks arising from the sol-vents should be addressed through one coherent, con-sistent regulatory measure that avoids unjustifieddistinctions between them. Under the principle of equaltreatment, any treatment favouring or disfavouring oneof the substances would not be justified by relevantdifferences. The proportionality principle would argueagainst the use of authorization for any of the threesubstances since authorization would not have any ben-efits, but the cost for industry would be high. If arestriction is sufficient to control the risks posed by anyof the three substances and pressure for substitutionexists, subjecting one or more of these substances toauthorization would, by definition, not be the leastonerous and thus be disproportional.

If the authorities were to proceed with diverging regu-latory approaches for dipolar aprotic solvents, manu-

302 ECHA, ‘Recommendation of the European Chemicals Agency of17 January 2013 for the Inclusion of Substances in Annex XIV toREACH (List of Substances Subject to Authorisation)’.303 European Commission, Draft Commission Regulation AmendingAnnex XIV to Regulation (EC) No 1907/2006 of the European Par-liament and of the Council Concerning the Registration, Evaluation,Authorization and Restriction of Chemicals (REACH) (2013), recital11.304 <http://echa.europa.eu/harmonised-classification-and-labelling-previous-consultations/-/substance/766/search/+/term>.305 ECHA, ‘Recommendation of the European Chemicals Agency of 6February 2014 for the Inclusion of Substances in Annex XIV toREACH (List of Substances subject to Authorisation)’.306 See ECHA, n. 301 above.307 ECHA, ‘Registry of Intentions’, found at: <http://echa.europa.eu/registry-of-submitted-svhc-intentions/-/substance/2443/search/+/term>.

308 See European Commission, n. 303 above, recital 11.309 E.g., for technical reasons, a company that switches to anothersolvent to avoid the authorization process may need to use more ofthe substitute solvent to achieve the same level of production.



240

http://echa.europa.eu/harmonised-classification-and-labelling-previous-consultations/-/substance/766/search/+/term

http://echa.europa.eu/harmonised-classification-and-labelling-previous-consultations/-/substance/766/search/+/term

http://echa.europa.eu/registry-of-submitted-svhc-intentions/-/substance/2443/search/+/term



facturers, importers and users would have a cause ofaction. They could challenge a Commission decision tosubject one or two of the solvents to authorizationbefore the European court. In this case, no appeal couldbe made to the ECHA Board of Appeal, which haslimited jurisdiction.310 If no action is brought beforethe European courts, the invalidity of any such deci-sion could be invoked in enforcement actions beforenational courts.

THE LEAST ONEROUS AND MOSTAPPROPRIATE MEASUREEU law requires that the three solvents be subjected tothe same or a very similar regulatory measure. Thisraises the issue as to what the most appropriate, leastonerous measure is. The analysis of this issue proceedsin two steps. First, authorization should be compared torestriction. Second, if restriction is the least onerousmeasure, it has to be determined whether a REACHrestriction or a restriction pursuant to other legislationis most appropriate.

With respect to the choice between authorization andrestriction, our analysis above shows that, in general,authorization is more onerous for industry. Authoriza-tion requires individual applications, which are bur-densome, limited in time and subject to withdrawal; ifthere already is pressure for substitution,311 there wouldappear to be no justification for the added burden asso-ciated with authorization. Authorization also consumesmore government and industry resources than restric-tion since it involves an individual permitting pro-gramme.312 In the case of the three solvents, the riskconcerned is occupational exposure. This is the kind ofrisk that can be addressed effectively through exposurelimits and protection measures. Such general measuresare less onerous than company-specific and use-specific authorizations.

In-depth policy and legal analysis supports the conclu-sion that restriction is the preferred approach. First, itis plausible that an RMO analysis313 for the three sol-vents, which includes, among other things, an assess-ment of whether there is a risk that is not adequatelycontrolled and needs to be addressed at EU level314 andwhether there already is pressure for substitution pro-

vided by specific EU legislation, would confirm that arestriction or a binding occupational exposure limit isthe most appropriate option.

In this case, authorization is less effective than restric-tion because authorization can address only any risksassociated with the specific property for which thesolvent concerned would be listed (repro-toxicity). Con-sequently, in practice, authorization often has to becomplemented with restrictions for risks that may notbe considered in the authorization procedure. Restric-tions, on the other hand, are not so limited and canaddress the full range of risks. There are only a fewspecific uses that are likely to result in exposure, andthese are all in industrial settings, thus posing occupa-tional risks. Occupational risks lend themselves well tocontrol through work place safety measures.

Authorization is not necessary to achieve proper riskcontrol and substitution and would have unintendedadverse consequences for competition and the internalmarket. There is no indication that existing EU legisla-tion does not already impose minimum requirementsthat properly control the risks, at least in most uses.Authorization is not necessary because existing EU lawson worker safety and solvent emissions already fullyprotect workers and require substitution of the sol-vents. Additionally, substitution is not realistic in theabsence of any less hazardous suitable alternatives.315

The adverse effects of subjecting the substances toauthorization would be serious, predominantly, but notexclusively, in the economic and commercial area. EUmanufacturers of certain products that rely on the sol-vents concerned, such as man-made polymer fibres,would be put at a competitive disadvantage comparedto non-EU manufacturers of the same products,without any benefit for human health. The unintendedadverse consequences of authorization would hit smalland medium-sized enterprises particularly hardbecause they would not be able to afford the resourcesnecessary to obtain authorization and would be de factoexcluded from using these solvents. Further, if authori-zation results merely in imposing exactly the same riskmanagement measures that are already in place, therewould be no additional health or environmental benefitwhile administrative cost would mushroom. In otherwords, authorization lacks any added value.

The case of these solvents is comparable to the cobaltcompounds, which the Commission, despite ECHA’srecommendation, decided not to add to Annex XIV andfor which it recommended restriction.316 Cobalt com-pounds have a similar profile to NMP, DMAc and DMF:

310 REACH, n. 1 above, Article 91.1.311 Any other regulation may create pressure for substitution, but itmay be hard to quantify such pressure or compare it to the pressurethat would be generated by authorization.312 Fees are charged to companies applying for authorization, whichshifts part of the cost to industry.313 An RMO analysis was conducted for DMAc, but did not includerestriction as a risk management option. UK Statement forDimethylacetamide, in: ECHA, ‘Minutes of the 27th Meeting of theMember State Committee (MSC-27)’ (December 2012), at 33 (‘UKStatement’).314 REACH, n. 1 above, Article 68.1.

315 See UK Statement, n. 313 above. See also J.D. Graham and J.B.Wiener, n. 103 above; R. Lofstedt, n. 49 above; A.D.K. Abelkop andJ.D. Graham, n. 49 above.316 Regulation 348/2013/EU, n. 104 above, recital 12.



241

high volume, and posing occupational risks.317 Thus, therisks posed by the solvents concerned could be effec-tively managed through restrictions.

The proportionality principle suggests that, as in thecase of cobalt compounds, a binding occupationalexposure limit or a restriction of exposure to the sol-vents at issue, for which only a few specific uses arelikely to result in exposure, is a more appropriateregulatory measure than authorization. Conversely, forthe following reasons, authorization would be mani-festly disproportional:

• Suitability of the measure: The authorizationregime is aimed at proper risk control, substitutionand good functioning of the internal market. Autho-rization would impose strict risk control measures,but for most, if not all, uses, existing regulatory andindustry measures already impose proper risk con-trols. Substitution of these solvents is not realisticsince all existing suitable substitutes are also classi-fied as very hazardous. As far as possible, pre-existing legislation already provides incentives forsubstitution – that is, the development of less haz-ardous suitable alternatives.318 As discussed above,subjecting these solvents to authorization wouldimpede the functioning of the internal market.Adding these solvents to Annex XIV would thus notbe suitable for achieving the objectives of the autho-rization regime. On the other hand, authorizationwould further stigmatize these solvents and affectdemand, but these commercial effects are notincluded in an RMO analysis.

• Necessity of the measure to achieve itspurpose: As discussed above, authorization wouldnot be necessary to achieve its purpose, sinceworkers are properly protected under EU occupa-tional health and safety law; consumers and theenvironment are not at risk; and Directives 2010/75and 98/24 already provide pressure forsubstitution.

Based on these arguments, further specific restrictionsare to be preferred over authorization. However,restrictions can be imposed pursuant to REACH orother chemicals-related legislation. As the risks posedby the three solvents concerned are occupational, twomain types of possible measures should be considered:REACH restrictions including occupational measures,and binding occupational exposure limits and othermeasures pursuant to occupational legislation. Ameasure under the latter, a specific worker health andsafety framework, is the most appropriate regulatorymeasure to address the risks associated with these sub-stances. Consequently, with a restriction proposal cur-

rently being considered for NMP, no further regulatoryaction is justified in the interim to include DMAc andDMF on Annex XIV. Rather, if a revised OEL is themost appropriate way to address the risks concerned,all three solvents could be subjected to a revised OELunder occupational safety legislation.

With respect to the most appropriate way to imposefurther specific restrictions, additional REACH restric-tions would appear to be superfluous given the existingoccupational exposure controls. As discussed above,indicative exposure limits under the worker safety leg-islation are already in place; if these limits are found tobe too lax, they could be amended, and be made man-datory, to ensure a high level of worker protection. TheEU occupational legislation requires also that compa-nies take all appropriate measures to eliminate orreduce chemical risks.319 There is no objective reason toprefer REACH restrictions over these occupationalexposure limits, and the applicable procedures andeffects associated with the two types of measures pointin different directions. If a substance poses only risk forworkers, the occupational safety law provides a moreappropriate framework for protective measures. Unlesspersuasive evidence to the contrary is provided, the EUshould conclude that the risks posed by dipolar aproticsolvents can be adequately addressed through EU occu-pational safety legislation.

If there is such persuasive evidence, it does not followthat all dipolar aprotic solvents and all of their usesshould be subjected to REACH authorization, which isaimed at phase-out and substitution.320 For example,assuming that there are risks, they would arise only incertain specific applications and uses, and not in otherapplications and uses. If this appears to be the case, theuses that raise only occupational risks could beaddressed through occupational exposure limits, andthe uses that cause other risks could be subjected torestriction, which is aimed at managing identified risks.Any residual uses that cannot be managed properlythrough restrictions could possibly be considered forauthorization. REACH contemplates this type of situa-tion, and allows uses already subject to proper riskcontrol to be exempt from authorization.321 Only if itappears that the uses of dipolar aprotic solvents posesuch diverging risks, the combination of occupationalexposure limits, restrictions and, maybe, authorizationmight constitute the least onerous and most appropri-ate set of measures.

THE CURRENT SITUATIONAlthough there are clear signs that the authorities aremoving towards a coherent approach, there is no final

317 See, e.g., ECHA, ‘Background Document for Cobalt Dichloride’(20 December 2011).318 Directive 2010/75/EU, n. 287 above, Article 58.

319 Ibid., Articles 5–6.320 REACH, n. 1 above, Article 55.321 Ibid., Article 58.2.



242

resolution as of April 2014. While the Commissionrecently adopted a clear position, ECHA appears toremain ambiguous.322 In December 2013, the Commis-sion decided not to include DMAc in its draft amend-ment to Annex XIV. It provided the following reasonsfor its decision:

[I]n view of the similarities of the two substances [DMACand NMP], both in the intrinsic properties and in the indus-trial applications, and in order to ensure a consistent regu-latory approach for both, the Commission considers itappropriate to postpone the decision on the inclusion ofDMAC in Annex XIV until the Agency submits to the Com-mission the opinions of the Committees for Risk Assessmentand Socio-economic Analysis on NMP in accordance withArticle 72 of Regulation (EC) No 1907/2006.323

Based on this clear signal from the Commission, onewould expect ECHA to adjust its approach to conformto the Commission’s position. On legalistic groundsthat are not entirely persuasive, however, ECHAincluded DMF in its Fifth Recommendation for autho-rization. It provided the following justification:

The MSC opinion notes that also other polar aprotic solventsthan DMF, namely DMAC and NMP, are currently consid-ered for potential further regulatory action under theREACH Regulation. ECHA agrees with the view expressedin the MSC opinion that it is not appropriate for ECHA toassess the pertinence of other regulatory risk managementinstruments in the Annex XIV recommendation, which isone step in the authorisation process. Considering thatDMAC is included in ECHA’s 4th Annex XIV recommenda-tion and given that the outcome of the ongoing restrictionprocess for NMP is not known, ECHA has included DMF inthis 5th recommendation to enable a consistent approach.324

It could be that ECHA is just very cautious, does notwant to prejudge any policy outcomes, and thereforeleaves all options open. It would be helpful, however,for the policy makers to be clear about where theprocess is going.

In its restrictions proposal, the Netherlands argues thata REACH restriction is to be preferred over authoriza-tion, as the authorization process is ‘costly and time-consuming both for industry as for authorities’ andcreates ‘large uncertainty to industry regarding the con-tinuation of their business because an authorisationrequest will only be given for a limited period of time’.325

This analysis is consistent with the facts, and is sound.The Netherlands’ report, however, goes on to reject theoption of reducing the exposure limits under the occu-pational legislation, invoking three arguments, all ofwhich are weak. First, the SCOEL would use a methodof deriving an OEL that differs from the REACH meth-odology. No argument is presented by the Netherlands,however, to the effect that the SCOEL method would be

inferior to the REACH method. Second, the reportnotes that ‘harmonised implementation of [a revised]indicative OEL by all Member States is not guaran-teed’.326 It fails to mention, however, that if the OEL ismade binding, this issue would be resolved. Third, theNetherlands observes that ‘the OEL by definition onlyprotects workers from the risks following inhalatoryexposure’ – not from risks following dermal exposure.Under the occupational legislation, however, additionalrisk management measures to reduce dermal exposurecan be required.327

CONCLUSIONS ANDRECOMMENDATIONS

The analysis of the REACH restriction and authoriza-tion regimes presented in this article has revealed asignificant gap in the regulatory framework. To a largeextent, the various REACH regimes operate as stand-alone programmes. REACH does not require a struc-tured policy analysis of alternative approaches toregulating chemicals. It does not require, or establish aframework for, comparison of alternative regulatorymeasures to determine which one is the least onerousand most appropriate, as required by general principlesof EU law. In addition, REACH contemplates that regu-lation proceeds ‘one chemical at a time’. This meansthat it is up to the authorities, on a case-by-case basis,to decide whether any analysis of alternative regulatoryinstruments is necessary, and if so, which substancesshould be included in such analysis, what optionsshould be reviewed and how it should be conducted. Toremedy this design defect, the EU needs a flexible, inde-pendent, coherent and unbiased framework for riskassessment of single or appropriately defined groups ofsubstances and efficient cost-benefit analysis of alter-native regulatory options.

In an attempt to remedy this gap to some extent, theCommission has developed the RMO analysis. Thisprocess is intended to ensure that the right regulatorytool is used to address the risks associated with a spe-cific SVHC. RMO analysis is aimed at preventing theuse of unnecessarily costly or otherwise inefficientinstruments and reducing the risk of a measure vio-lating general principles of EU law. Indeed, RMOanalysis will help to ensure that the most appropriateregulatory measures are applied for a particular sub-stance or a group of substances. The RMO process,however, does not properly reflect the EU’s basic prin-ciples of transparency and consultation, and proceedswith limited involvement of and disclosure to stake-holders. Moreover, RMO analysis is currently done

322 See, e.g., the inclusion of DMF in ECHA, n. 305 above.323 See European Commission, n. 303 above, recital 11.324 See ECHA, n. 305 above, Annex I, 3.325 See The Netherlands, n. 296 above, at 14.

326 Ibid.327 Directive 89/391/EEC on the Introduction of Measures to Encour-age Improvements in the Safety and Health of Workers at Work,[1989] OJ L183/1, Article 16.1; Directive 98/24/EC, n. 25 above.



243

only in connection with Candidate Listing, while somelimited version is conducted in connection withrestriction proposals. Policy analysis, including cost-benefit analysis, however, should be applied system-atically and comprehensively before any regulatorymeasures are proposed.

The proper role of REACH authorization, which is notwell defined by the EU legislature, is at the heart ofRMO analysis. Authorization is not necessarily the leastonerous and most appropriate measure for risks asso-ciated with the manufacture and use of an SVHC. As ageneral rule, SVHCs should be subjected to REACHauthorization only if that is the best regulatory instru-ment in light of the specific properties, uses and risksposed by the substances concerned and their substi-tutes. Listing for authorization cannot be automatic andbased entirely on hazard. Instead, a choice betweenauthorization, restriction, a measure under other legis-lation (such as an occupational exposure limit) or nomandatory regulation at all should be made on the basisof a sound comparison of the costs and benefits (includ-ing strengths and weaknesses) of these tools in the spe-cific context. Pursuant to general principles of EU law,the least restrictive effective instrument should be used.Although the EU has left the issue of double regulation(i.e., authorization and restriction of the same sub-stance) open, given the substantial administrative costsassociated with authorization and the lack of clear,proven advantages of authorization over generallybinding restrictions (including with respect to substitu-tion), there is a question as to whether and, if so, when,such double regulation will be the least restrictivemeasure. More generally, the REACH authorizationregime illustrates the pitfalls of hazard-based regula-tion. Once the role of authorization has been properlydefined, it may well become clear that authorization isjustified only in exceptional cases.

The case of dipolar aprotic solvents was explored toillustrate these issues, and the possible solutions withinthe context of REACH and other legislation. Dipolaraprotic solvents have similar properties, classifications,industrial uses and exposure and risk profiles. Never-theless, the authorities initially launched diverging pro-cedures to further regulate these chemicals withoutcomprehensive risk assessment or comparative cost-benefit analysis. Subsequently, the Commission recog-nized that this uncoordinated process had to berectified and started pursuing a coherent, structuredapproach for all three solvents.

An RMO analysis for dipolar aprotic solvents wouldlikely show that the least onerous and most appropri-ate regulatory instrument for this group of substancesis not authorization and not even restriction based onREACH. Rather, all known risks associated with thesesolvents can be handled under occupational legisla-tion. In light of the specific concerns about these sol-

vents, which are limited to the potential for workerexposure, appropriate occupational exposure limitsand possibly other work place measures would be themost targeted effective measure; the existing indicativelimits should be reviewed in light of the evidence and,if necessary, they could be made binding and lowered.The Commission appears to have reached this conclu-sion, but ECHA’s position remains somewhat ambigu-ous in light of its recent recommendation that DMF belisted for authorization. Since no justification has beenprovided for singling out DMF, and no such justifica-tion is likely to exist, the Commission should ensurethat the coherent approach it proposed is followed byall actors involved. Occupational protection measurescan adequately address the concerns associated withthe use of dipolar aprotic solvents. There is a risk,however, that the choice of regulatory instrument isskewed by politics and ‘turf battles’ between authori-ties. Within the Commission, Directorate-GeneralEnterprise is responsible for REACH restrictions,while Directorate-General Employment handles occu-pational safety measures. The EU leadership shouldensure that any internal conflicts, if they arise, do notaffect the public interest in regulation based on objec-tive policy analysis.

Based on the analysis presented in this article, we offerthe following recommendations to the various bodiesinvolved with regulating chemical risk arising fromSVHCs. These recommendations can be implementedthrough reinterpretation of the REACH Regulation,without amendments being necessary, and throughchanges in administrative practices.

First, the Commission should confirm that the Candi-date List is a part of the authorization procedure andshould not be used for purposes other than adding sub-stances to Annex XIV. In other words, if there is nointent to consider an SVHC for listing for authorization,it should not be added to the Candidate List.

Second, substances should be prioritized for assess-ment based on a set of criteria that is indicative of thedegree of risk arising from their actual use. The REACHcriteria for prioritizing substances for AuthorizationListing (hazardous properties, the degree of dispersive-ness of use and the volumes) are a good starting pointfor drafting such criteria.

Third, before procedures are launched to regulatechemical risk arising from an SVHC, a fledged riskassessment and policy analysis, including cost-benefitanalysis, should be conducted. As part of this process,substitutes for the substance to be regulated are iden-tified, and an assessment is made as to whether one ormore of the substitute substances should be regulatedalongside the target substance as a group or category.Risk/risk trade-offs are then examined, and the avail-able regulatory options identified and compared in



244

terms of their costs and benefits. In addition to theCommission, experts from ECHA’s advisory bodiescould play a role in this process. The objective is toselect the regulatory option that best achieves thedesired regulatory ends at lowest costs. This procedureis transparent, and public consultations are held, asappropriate, throughout the process. To implementthese recommendations, it may be possible to reformthe current RMO analysis by converting it into a stand-alone procedure, expanding it to include risk assess-ment and policy analysis, and eliminating the bias infavour of authorization. Once a regulatory pathway hasbeen selected, the work already done should be fed intothis process to avoid duplication of efforts.

Fourth, in order to enable the authorities to select thebest regulatory tool for managing risks associated withSVHCs, REACH’s provisions on restrictions should beinterpreted to require no more than a showing of unac-ceptable risk if recommended risk management mea-sures are not applied (and that EU-wide action isrequired). The current interpretation eliminates therestrictions option in cases where it could be the pre-ferred regulatory instrument.

Fifth, in order to assist decision makers, a comprehen-sive, multidisciplinary study should be conducted onthe costs and benefits of authorization and of ‘doubleregulation’ (authorization plus restriction) of the samesubstance. The current understanding of the pros andcons of these options is incomplete. A better under-standing can help to better define the role of authoriza-tion in the spectrum of regulatory instruments.

Sixth, if the sole risk arising from an SVHC is occupa-tional risk, regulation of that risk should proceed pur-suant to EU occupational health legislation, unless thislegislation does not promote substitution and substitu-tion is deemed to be a regulatory end. Under the EUoccupational health legislation, specific specializedprocedures have been established for addressingoccupational risks in an integrated fashion. Unless

there are good reasons for doing so, REACH should notinterfere with these processes.

Seventh, if the risks arising from an SVHC are product-related rather than occupational, on the other hand,risk assessment and policy analysis should proceed pur-suant to REACH. At the EU level, there is no integratedframework for assessing and managing product-relatedchemical risk, which implies that outside of the REACHcontext, chemical restrictions may be adopted on thebasis of limited ad hoc procedures, which underminesthe quality of decision making. In the risk assessmentand policy analysis programme under the REACHumbrella, the available regulatory options can be iden-tified and assessed. These options include the REACHrestrictions programme and product-specific regula-tions, such as RoHS.

Once these recommendations are implemented, theregulation of SVHCs will become a more transparent,cooperative and objective process. More generally, oursuggestions are intended to cause the EU to moveaway from a dogmatic-theoretical approach to regulat-ing SVHCs to a science-based, pragmatic process.These improvements, in turn, will help to ensure thatregulatory decisions under REACH are in the publicinterest.

Lucas Bergkamp is a partner in the law firm of Hunton& Williams, which provides advice on regulations suchas REACH. He also heads the European RegulatoryPractice.

Nicolas Herbatschek is an associate in the law firmHunton & Williams.

The authors thank Adam Abelkop, John Graham and ananonymous reviewer for comments on previous drafts ofthis article. The opinions expressed in this article are thepersonal views of the authors – not of Hunton & Wil-liams or its clients. Any errors or omissions remain thesole responsibility of the authors.



245

Regulating Chemical Substances under REACH: The Choice … · cation and labelling was introduced...

Documents

Transcript of Regulating Chemical Substances under REACH: The Choice … · cation and labelling was introduced...