It is not every day that sees a piece of draft legislation with an official price tag of €18-32 billion put out for consultation. And the paper itself looks like a record-breaker, containing a draft Regulation running to 117 clauses, 16 annexes and a mammoth 1,200 pages in all.
It has taken five years to get this far, with a White Paper the main previous output in 2001. Thanks to recurrent delays in publication of the consultative document, a formal proposal from the Commission has now slipped well into the autumn. With a further hiatus next year due to the enlargement of the EU and elections to the European Parliament, the legislation seems unlikely to be on the statute book before 2006.
The slippage has angered environmental groups, who hit out at the Commission on 7 May for showing a lack of leadership in getting the reforms under way. Their fear must be that further delays will play into the hands of the chemical industry, whose fierce lobbying against the proposals has rattled Commissioners and probably national governments as well.
A step change
Environment Commissioner Margot Wallström has described the package as radical, and for once this is not hyperbole. Now that the scope of the legislation is clearer, it can be put into perspective by comparing it what has gone before.
The EU has two statutory regimes on chemicals. Under one which came into being in 1981, some 300 new substances per year have latterly been tested and notified to the authorities before being placed on the market.
Under the second, which kicked off in 1994, 140 of the 100,000 'existing' substances which were on the EU market before 1981 have been placed on priority lists for risk assessment and, if necessary, control. Barely a dozen have since completed the process - and it was this failure to deliver results which prompted the present reforms.
The two regimes have been operated by the joint efforts of the European Chemicals Bureau - which has a staff of around 40, some of whom are occupied on other duties - and national authorities.
The numbers game
Under the proposed Regulation, the separation between new and existing substances will end. Instead, there will be one unified system for the Registration, Evaluation and Authorisation of Chemicals (REACH).
Largely as outlined in the White Paper, the new policy will kick off with registration of an estimated 30,000 chemicals placed on the EU market in quantities of more than one tonne per year per manufacturer. This will be done in three phases spread over eight years.
What the White Paper left obscure was how intermediates - chemicals used solely in synthesising other chemicals - and polymers would be dealt with. These have had special treatment under existing legislation, and essentially are not counted among the 100,000 substances which were on the market before 1981.
The exact numbers of these two groups of substances are not known. But estimates gathered by the Commission from industry indicate that there are 70-120,000 intermediates and around 70,000 polymers in use.
Suggestions that these chemicals should be brought under REACH have provoked vehement opposition from chemical manufacturers, as well as concern from national governments. The worry has been that including them all would not only drive up costs to industry but swamp the new system with registration data and distract attention from prioritising the most hazardous substances for risk assessment and control.
The Commission has responded by lightening the proposed registration requirements for some intermediates and polymers and exempting others altogether.
On 7 May, Enterprise Commissioner Erkki Liikanen said that at least 55% of the 70,000 polymers will be exempted from REACH - and the remainder treated with 'a very light touch'. Registration of polymers will not be required until 12 years after the legislation enters into force, except for those classified as 'dangerous'.
On intermediates, the Commission estimates that 40,000 will be exempted from REACH - the so-called 'non-isolated' intermediates which are never removed from the process in which they are manufactured. Another 40,000 used on the site of manufacture or supplied to up to two others will be subject to limited registration requirements, with no additional testing required for the vast majority.
The net result appears to be that full registration dossiers will be required for some 30,000 substances and perhaps 30,000 intermediates which are placed on the market, together with limited dossiers for some 40,000 other intermediates and 30,000 polymers - around 130,000 in total.
To these registrations can be added reports which downstream users will need to submit where their use of a chemical is not covered by the manufacturer's registration, or where their risk management measures differ from the manufacturer's recommendations.
Next comes the evaluation stage - at least, it had been expected to be a stage. However, under the draft Regulation, two evaluations will be carried out by national authorities.
In one, the 'standard' evaluation, they will examine any proposals for testing made by registrants or downstream users in order to meet specified data requirements.
The second will be a 'priority' evaluation in which the authorities themselves will examine whether the data requirements have been met and, if necessary, order the submission of further information.
There are two further elements to REACH. Authorisation of uses of the most hazardous chemicals is expected to apply to between 1,500 and 4,000 substances. And proposals for production, marketing and use restrictions are likely to emerge from the evaluation process, although the number of substances potentially affected is unknown.
A new chemicals agency
This system will require a major expansion of the existing chemicals management infrastructure - at both EU and national level.
At EU level, the Commission has proposed the establishment of a European Chemicals Agency (ECA) based, like the present Bureau, at Ispra in Italy. The Commission believes that it will need around 300 staff and cost €380 million over its first 11 years.
National capacities will also need to expand - probably by the same order of magnitude. However, as an internal Commission paper pointed out, Member States' workload 'will depend to a large extent on how many non-mandatory functions they take on' - a crucial observation.
The new Agency will have a long list of tasks. It will run a database in the pre-registration phase of REACH. It will receive companies' registrations, checking them for completeness and entering them on a database, as well as reports from downstream users. It will be at the heart of exchanges of information on registrations and authorisations and decisions relating to them, and help prepare decisions on substance restrictions. And it will maintain a classification and labelling inventory.
Beyond this there are numerous technical and co-ordination functions. These include the preparation of guidance and software tools, co-ordination of enforcement projects and drafting minimum enforcement criteria.
Most of the ECA's costs will be paid by industry. Last year, a study for the Commission by consultants Deloitte & Touche estimated that fees for registration and evaluation might run to €7,100 per substance, and fees for authorisation to €148,000 - though with significant variations possible around these figures depending on the number of substances entering the system and the rate at which authorisations proceed. The revenues will be split between the Agency and national authorities.
Delays and deadlines
So there will be a vastly expanded chemicals management infrastructure - assuming, of course, that the necessary chemists and toxicologists, currently in short supply, can be found. But will this deliver one of the regime's key objectives - to accelerate the process of risk management?
One obvious point is that the wrangling over the legislation has already caused a serious delay. If the Regulation enters into force in 2006, the registration deadlines for substances marketed in quantities of more than 1,000, 100-1,000 and 1-100 tonnes per year per company would be 2009, 2012 and 2017 - four years later than the White Paper proposed in the first two cases, and, with an extra year slipped in by the Commission for the last group, five years.
Experience suggests that most registrations will be made very close to the deadlines - so the first evaluations are a long way off.
Registration should not be confused with a requirement to submit full data packages. Registration dossiers will receive a 'completeness check' from the ECA. This will not be 'an assessment of the quality or the adequacy of any data', but rather a check - which may well be automated - to 'ascertain that all the elements' required are present.
What the draft Regulation leaves uncertain are the grounds on which registration may be rejected - and a chemical banned from the EU market.
It will be for national authorities to decide whether further information is required to complete a registration. However, the draft implies that refusing registration where information which can only be generated by testing is absent from a dossier will not be on the cards.
For this reason, the Regulation appears to provide a positive incentive for companies not to invest in full data packages before registration - and certainly does not support any expectation that many chemicals will be refused registration on this score.
Decisions on testing
The effect will be to push many decisions on data gaps into the evaluation stage. Here, too, there are important questions about deadlines and the extent of mandatory requirements.
On 'standard' evaluations, national authorities will be required to prepare draft decisions on companies' testing proposals two years after the registration deadlines for chemicals in the over 1,000 and 100-1,000 tonnes per year groups - most likely by 2011 and 2014.
There are, however, no deadlines for substances in the 1-100 tonnes group, nor for polymers or, except where they meet criteria of 'very high concern', intermediates not placed on the market.
On 'priority' evaluations, matters would be left entirely to the discretion of national authorities. The draft says that they 'may' evaluate companies' dossiers with a view to requesting further information. Only once they begin would a deadline apply - 12 months to prepare a draft decision.
What will be the subject matter of authorities' draft decisions? It will not be an evaluation of a company's risk assessment. Rather, it will be a decision on further testing and, where there is more than one registrant or downstream user who must conduct the same tests, a proposal on which should be responsible for them as part of the effort to minimise testing.
And why is it a draft decision? Because the Commission wants national authorities to act in a harmonised fashion, a draft decision will - after being passed to the registrant for comment - be circulated to the other 14 authorities. If any of them disagree with the proposal, the matter will be referred to a committee of Member States' experts at the ECA. And unless they are unanimous, the matter will then go to a further committee of Member States' representatives chaired by the Commission.
And the duration of this process? From the day that the draft decision is passed to the registrant, up to 195 days are allowed before the Commission must prepare its draft decision for referral to the committee of national representatives - which presumably will not meet straight away.
This process has been proposed, it should be noted, for decisions on testing which will run into many thousands. Add to that anything between many months and a couple of years for the tests to be carried out, and it is evident that few complete data packages will be available much before the middle of the next decade - the main exceptions being 1,300 or so high production volume chemicals for which test data are being generated under programmes run by the OECD and International Council of Chemical Associations.
So far, then, the scope for companies to play the system, the absence of deadlines for important parts of the evaluation process, the discretion afforded national authorities as to whether to initiate priority evaluations, and the cumbersome nature of decision-making on testing raise serious doubts about whether a more streamlined regulatory process will be delivered. Experience from the current programmes for evaluating 'existing' substances and plant protection products is not encouraging about the capacity of the EU's machinery to overcome such concerns.
Similar concerns are likely to be aired over the proposed procedures for authorisation and marketing use and restrictions.
Substances will automatically be subject to authorisation if they meet the 'very high concern' criteria laid down in the Regulation. They may also qualify for authorisation where a Member State has other evidence that they are PBTs or vPvBs or give rise to an 'equivalent level of concern' by virtue of endocrine-disrupting or other properties.
Decisions on whether a substance should enter the authorisation process will be taken by the committee of Member States' representatives. Proposals will come to it either directly from the ECA or indirectly from a Member State.
In the latter case, the initiating country's dossier will be circulated to other Member States, and if any disagreements are not resolved the matter will go to the Commission, which will prepare a proposal for decision by the committee - a process taking 105 days in all.
Authorisation decisions for substances placed on the market will rest with the Commission. Before then, however, there will be a lengthy process.
The procedure will begin when a substance is entered on an annex to the Regulation. Manufacturers, importers or downstream users will then have a period to apply for authorisation. The period is not specified in the draft, but is likely to be many months so as to allow companies time to prepare risk assessments, a socio-economic case, and a review of alternative substances or technologies.
Applications will go to the ECA, which will invite comments from third parties on alternative substances or technologies. It will have ten months to prepare a draft opinion, including a risk assessment of the uses covered by the application and its own socio-economic analysis.
Applicants will then have two months to comment on the draft opinion, and the ECA a further two months to consider their comments. Finally, the Commission will have three months to submit a draft authorisation decision to the committee of Member States' representatives.
At an absolute minimum, therefore, this process can be expected to take two years, but more probably longer.
Restrictions on substances
The second route for regulating substances will be through a 'restrictions process'. EU legislation restricting the marketing and use of dangerous substances dates back more than 20 years, but the new regime will extend to manufacture as well.
The process will be used to tackle substances posing unacceptable risks to humans or the environment which need to be addressed across the EU. It can be started either by the Commission requesting the ECA to prepare a dossier setting out the case for action, or by a Member State submitting a dossier to the ECA.
In the latter case, the Agency will appoint rapporteurs on risk assessment and socio-economic analysis to check whether a dossier meets a specification laid down in an annex. If they judge that it does not, they must inform the Member State within 45 days. A further 30 days will be allowed for the dossier to be brought up to scratch, otherwise the procedure will terminate.
Once completed, dossiers will be published on the Agency's website. Interested parties will have three months to comment.
Within a further six months, opinions on the suggested restrictions will have to be completed by two ECA committees on risk assessment and socio-economic analysis. If these diverge 'significantly', the latter committee will have up to 90 days more to review its opinion.
The finished opinions will then go to the Commission, which will have three months to submit a proposal to the committee of Member States' representatives.
From the point at which dossiers are published by the ECA, therefore, the process will last up to 15 months, not including the time it takes the committee to reach its decision. Add in the time that it takes for the ECA to draft a dossier and it is unlikely to be much less than two years.
Centralisation v delegation
Restrictions and, in particular, authorisations will not have to wait on the completion of risk assessments, but could be proposed at any time. Given that the ECA will not be operational until around 2008, however, it will be comfortably into the next decade before much is seen of either.
The procedures may seem long drawn out, but it is at least fair to compare them with what is done now. Marketing and use Directives generally take 4-5 years to get on the statute book, so halving the time would be a considerable advance - if deadlines are adhered to.
Nevertheless, it is worth asking whether the procedures could be truncated. According to Liz Surkovic of the Chemical Industries Association, their complexity could be much reduced if the Commission had been brave enough to propose empowering the ECA to take decisions which will instead be made through the Member State consultation and committee procedures.
The chemical industry has, indeed, been arguing for the Agency to be given this role in order to make the processes less complex. It also sees, no doubt, that a more centralised arrangement would depoliticise decision-making under REACH and make it easier to influence the outcomes.
Alternatively, it might be argued that the process could be simplified by delegating more authority to Member States - especially on registrations, where they could be left to take decisions on testing while following tightly drawn guidance which sought a consistency of approach.
Because of its ambition, REACH poses questions more starkly than what has gone before about the right balance between centralisation and delegation in chemicals regulation. So far, Member States have signalled that they want to retain influence over the process - but it remains to be seen whether they will be happy with the balance struck by the Commission, or indeed whether they will stick to that line once the regime's resource implications sink in.
The last word can go to national chemicals regulators who met last November to discuss the implications of REACH.
Key success factors for the new regime, they concluded, will include high-quality registration dossiers which minimise national authorities' work, clear criteria for prioritising substances for evaluation and risk management, and adequate resources for enforcement - on which there is currently an 'enormous backlog'.
The answers are not all in the Commission's hands - but those that it has provided so far seem unlikely to convince.