The EU-US MRAs on conformity assessment (part 1): Fortress Europe?

In my last post, I examined a claim made by Nick Clegg that MRAs on conformity assessment are needed by exporters from third countries if consignments are not to be ‘impounded and checked at the EU border’:


I pointed out that China does not in fact have an MRA with the EU on conformity assessment, and explained how it is able to satisfy the EU’s product legislation requirements in the same basic way as an EU manufacturer does, either through a manufacturer’s Declaration of Conformity or, where third party certification is needed, through local test facilities, which are in either a subsidiary or subcontracting relationship with an EU-based Notified Body.

In this series of posts, to complete the picture as it were, I look at the second country named by Clegg as having an MRA with the EU and show that, while there are indeed two such EU-US MRAs in existence, they play only a marginal role in helping US manufacturers to export to the EU. The story of the first of these, the 1998 ‘Agreement on mutual recognition between the European Community and the United States of America’, is of considerable interest in its own right, as much effort was invested in attaining the agreement, and great hopes were attached to it, but in the end it is generally considered to have been a partial failure. There exist detailed accounts of the negotiations by Schaffer (2002) 1 and Deveraux  (2006) 2, which give valuable insights into the differences between US and EU regulatory structures, as well as into the conduct of trade negotiations. In this post, I try to explain the events that led in May 1989 to a joint US-EU commitment to open discussions on mutual recognition of conformity assessment.

As is the case with the EEA (see Origins of the EEA), the stimulus for the EU-US MRA can be traced to the Single European Act of December 1985, with its objective of creating a single market by 1992, along the lines of the program that had been set out by the Commission in its June 1985 White Paper ‘Completing the Internal Market’. Central to the program was the ‘new approach to technical harmonization and standards’, which had been endorsed by a Council Decision on 7 May 1985. Under the so-called ‘new approach’, which had in fact been employed already in the 1973 Low Voltage Directive, but was now to be the primary means used to attempt the harmonisation of product rules, only the essential safety requirements (or other requirements in the general interest) were to be laid down in Community legislation, while detailed technical specifications were to be contained in standards developed by the standardisation organisations CEN and CENELEC.

Simplified procedures and the harmonised standards

The standards would be voluntary, with no compulsion upon manufacturers to employ them, but their use would provide a presumption of conformity with the legislation. In general, where manufacturers did not use the standards, there was a greater obligation upon them to have a specimen of the product examined, tested and certified by a third party body, to ensure that the essential safety requirements were met. To take the Simple Pressure Vessels Directive 87/404/EEC as an example, it being the first of the new wave of product safety directives, no EC type-examination of a prototype was obligated if the harmonised standards were used, although it was still necessary to submit the design and manufacturing schedule to an approved body for examination and certification:


In contrast, where the harmonised standards were not used, or used only in part, an EC type-examination was essential:


Toy Safety Directive 88/378/EEC, which followed next, did not in general necessitate the involvement of an approved body for toys manufactured in accordance with the harmonised standard (except ‘where appropriate’), but an EC type-examination was required if the standards were not used in full.

The Commission, in its Communication to the Council of 24 July 1989 (COM(89) 209), confirmed that under the New Approach, third party intervention would be required when the harmonised standards were not employed in full, but a ‘simplified procedure’ when they were:


And indeed the same distinction had already been made in the New Approach Council Resolution of 7 May 1985, referred to above, as may be gleaned from the following extract (Annex II, VIII, 3.1), where it is stated that the designated certification bodies will have to intervene where harmonised standards have not been used in full:


Nevertheless, for some products associated with high risk and hazard, the involvement of notified bodies (as they became known) was required whether the harmonised standards were employed or not. This was the case for Category II and III PPE (Personal Protective Equipment) under Directive 89/686/EEC, and for all types of Active Implantable Medical Devices (AIMD, a product category which includes heart pacemakers, for example) under Directive 90/385/EEC. 3

The harmonization of laws

The main purpose of the product safety Directives that were enacted from 1987 onwards was to harmonise the diverse national regulations so as to create a single market for these products by the end of 1992. The first of them, the Simple Pressure Vessels Directive:


had Article 100 of the EEC Treaty as its legal basis:


Article 100 provided for the Council, acting by unanimity, to issue Directives for the approximation of national laws so as to further the establishment or functioning of what was then still called the Common Market:


That there was a need for harmonisation through the ‘approximation’ of national laws, if a single market was to be attained, is made clear in a following recital of the Preamble to the Simple Pressure Vessels Directive, which refers to the disparity in national laws:


Therefore, it was argued, these disparate national laws had to be harmonised to ensure the ‘free movement’ of simple pressure vessels:


By ‘free movement’ is meant that if a simple pressure vessel is marketed legally in one Member State, then it can be marketed legally in every other Member State. Where this situation pertains, it can be said that there exists a single market for the product, so that a single design specification will be sufficient to satisfy the regulatory requirements in every Member State.

The purpose behind the creation of a single market, as stated in the Commission’s seminal 1985 White Paper was that Community manufacturers would be able to benefit from economies of scale and increase their international competitiveness:



It seems to me that the only home market of ‘continental proportions’ in 1985 was that of the United States, from which it seems to follow that by ‘international competitiveness’ was probably meant in particular competitiveness vis-a-vis the United States.

The Single European Act and qualified majority voting

By the time that the next New Approach product safety Directive (88/378/EEC) on the Safety of Toys was issued on 3 May 1988, the legal basis had changed from Article 100 EEC to Article 100a EEC:



The difference had been wrought by the Single European Act, which had come into force on 1 July 1987, and which had incorporated the ambition to complete the internal market by 31 December 1992, as well as to remove all internal frontiers:


Then Article 18 had provided for qualified majority voting in the Council for harmonisation measures which would help bring about this completion of the internal market by the end of 1992:


Although this revolutionary new article was tempered somewhat by Article 100a(4), and by declarations by the British and by the French that the Luxembourg compromise (which allowed in effect for a national veto where vital national interests were at stake) was still standing, the reality was that the ability of a Member State to resist the will of the majority had been greatly weakened. In an influential article published in the Yale Law Journal in 1991, and entitled ‘The Transformation of Europe’, J. H. H. Weiner, in a discussion of the implications of the new Article 100a wrote that the ‘Foundational equilibrium’ of the Community ‘seems to be shattered’, and that Member States were:

now in a situation of facing binding norms, adopted wholly or partially against their will, with direct effect in their national legal orders. 4

According to Weiner (p. 2458), the move to majority voting was the key to the success of the 1992 strategy, and Article 100a was the ‘single most important provision’ of the Single European Act:


Craig and de Burca state in ‘The Evolution of EU Law’ that ‘the new Article 100a … was the provision used for the enactment of much of the single market legislation post-1986’ and, so far as I am aware, it was the legal basis for all the harmonising product safety legislation. 5

Progress towards the completion of the internal market was so rapid that at the end of the European Council in Hanover on 27-28 June 1988, just short of a year after the Single European Act came into force, the Presidency was able to declare (p. 164) that the process was irreversible:


Over one third of the measures programmed in the 1985 White Paper had by then been agreed, and the European Council noted (p. 164) that it was the new voting procedures that had much such progress possible:


In the same Bulletin of the European Communities, published in September 1988, the Commission claimed that the credibility of the Community venture had been established in the eyes of the world. For evidence, it pointed to the interest shown by its near neighbours in a closer association:


It is striking that in the Bulletin’s 22-page long chapter on External Relations, two and a half pages are devoted to relations with EFTA and the EFTA states, and just four lines to the United States:


At point 2.4.24 (page 127), under the heading ‘United States protectionism’, it is reported that the European Parliament ‘highlighted the enormous, real economic costs that American protectionist measures in the industrial sector might entail’:


Meanwhile, in the United States, fears were growing of European protectionism.

Reaction to the 1992 program in the United States

On 13 October 1988, a month after the above Bulletin was published, the Ways and Means Committee of the US House of Representatives asked the US International Trade Commission (USITC) to investigate the possible impact on US economic activities of the 1992 single market program:


The USITC report, entitled ‘The Effects of Greater Economic Integration within the European Community on the United States’ was published in July 1989. It presented the US foreign policy view of the 1992 program, in its broader aspect, as being twofold. It welcomed an open and prosperous Europe, but feared that national industrial interests that were threatened by the opening up of the European market, could endeavour to protect themselves by erecting barriers to imports from outside Europe:


An article in Business Week on 1 August 1988 entitled ‘Laying the Foundation for a Great Wall in Europe’ 6 was followed on 22 August by one in the New York Times called ‘A Fortress Europe in 1992?’ by Robert Hormats, a former Assistant Secretary of State for Economic and Business Affairs. Another article in the New York Times on 23 October 1988 entitled ‘The Growing Fear of Fortress Europe’ quoted the US Trade Representative Clayton K. Yeutter as saying that ‘What’s important is that Europe not become free traders internally and protectionist externally’.

European counter responses

According to the Chicago Tribune (20 October 1988), both Lord Cockfield the Internal Market Commissioner and Willie de Clerq the Commissioner for External Affairs protested that it was the United States that was protectionist and not the European Community:


In less pugilist vein, the Commission released an Information Memo on 19 October 1988 to provide reassurance that ‘the Community will seek a greater liberalization of international trade: the 1992 Europe will not be a fortress Europe but a partnership Europe’. The Commission pointed out that there would be a single market for non-Community companies as well as Community ones, since there would be one frontier rather than twelve and only one set of rules, standards, tests and certification procedures:


Of relevance to current suggestions that the UK should pay for access to the single market is the following undertaking in the same document for ‘free access to 1992 benefits for firms from countries whose market is already open’:


Conformance with GATT

The Information Memo provided further reassurance that the development of the Community’s external economic policy would (p. 2) ‘take place in harmony with the Community’s existing international obligations’, including GATT. Article 3 of the GATT provides that imported goods shall be treated in the same way as domestic products, with Article 3(4) applying this principle to product laws, regulations and requirements:


The GATT Standards Code

At this time (1988), the GATT Uruguay Round was in progress. The Tokyo Round (1973-79) had ended with an ‘Agreement on Technical Barriers to Trade’, commonly known as the GATT Standards Code, distinguishing it from the WTO TBT Agreement, which was signed at the end of the Uruguay Round in 1994 and which revised, strengthened and extended the Standards Code. In accordance with Article 3(4) of the GATT, shown above, Article 2(1) of the Standards Code provided that imported products would receive no less favourable treatment than domestic products in relation to technical regulations and standards.

Article 5(1) went on to provide that where a positive assurance was required by central government bodies that products did in fact conform to technical regulations and standards, the procedures in place for imported products would be no less favourable than those for domestic products (the first two of six points being shown below):


Difficulties for exporters to the EC

It will be realised that putting imported products on a level with domestic products does not put foreign manufacturers on a level with domestic manufacturers. The exporter might well have to send a specimen of the product to the importing country for type approval, with attendant costs and delays, especially in the case of larger and weightier products, which could have to be sent by ship rather than air.

In late 1988, according to the USITC Report (p. 6-14), the Commission released a draft proposal for an EC policy-wide policy on the testing and certification of manufactured goods. It was published on 24 July as COM(89) 209, ‘A Global Approach to Certification and Testing’. Unlike COM(85) 19, the Commission’s 1985 communication proposing a ‘new approach’ to technical harmonisation and standards, which made no mention of external relations, COM(89) 209 included a whole chapter on ‘External Aspects of the Global Approach’. The Commission confirmed the Community’s adherence to the GATT Standards Code, which indeed was its ‘starting point’ in this area (p. 44):


The USITC Report acknowledged (p. 6-15) that the Commission had confirmed its commitment to the Standards Code, but pointed out that the new system could cause difficulties for US suppliers, since they might be forced to have their products tested in Europe by notified bodies:


Mutual recognition of tests and certificates

It was presumably a recognition of such difficulties as these that led to Article 5(2) of the Standards Code, which provided that the Parties ‘ensure, whenever possible’ that their central government bodies accept test results and certificates issued by bodies in the territories of other Parties, so long as these gave sufficient assurance of conformity. It was admitted that prior consultations might be necessary before a ‘mutually satisfactory understanding’ could be reached:



In a commentary on the Standards Code in the World Trade Law Journal (May-June 1980), R. W. Middleton (a former Assistant Secretary-General of the International Standards Organisation) suggested (p. 209) that this second paragraph of Article 5 was of ‘major significance’ and might ‘in the future become of key importance in the development of work on technical barriers to trade’.

The Commission stated (COM(89) 209, p. 45) that the Community, in order to fulfil its responsibilities under Article 5(2), was prepared:

to conclude agreements for mutual recognition of tests, reports, certificates and marks

provided that three conditions were met. First, the technical competence of the non-Community body should be assured, normally by being evaluated according to the EN 45000 series of European standards, which were themselves based on the international ISO/IEC Guides as well as on the work of ILAC (International Laboratory Accreditation Conference). Second, the benefits accruing to each side from such an agreement should be ‘equivalent’:


Third, the agreement should be limited to the testing, inspection and certifying activity of bodies that have been specially designated for the purpose.

Further, the Commission pointed out (p. 46) that such agreements would fall under the Common Commercial Policy, and in particular Article 113 EEC, which provided for the Council to authorise the Commission to open the necessary negotiations between the Community and third countries. Moreover, it pointed out that the obligation of Member States to accept products lawfully marketed in other Member States was not consistent with access to the single market being granted by bilateral agreements between individual Member States and third countries. It did not say explicitly that existing bilateral agreements on conformity assessment would have to cease, but this would seem to be a logical consequence for sectors which were subject to EC-wide regulation. The USITC Report did (p. 6-15) draw this conclusion, that such bilateral agreements ‘would become null and void.’

‘an outpouring of concern’

The USITC published a follow-up report in March 1990, which covered developments from 1 January 1989 onwards. It noted that government and private sector experts were in agreement that ‘the EC’s policy towards conformity assessment may ultimately have more significance on U.S. industry than the standards themselves.’ The EC’s move towards more uniform procedures was ‘heartily welcomed’ by most, but at a public comment period of hearings held by the Interagency Task Force in September 1989, there was (p. 6-24) ‘an outpouring of concern by U.S. manufacturers and testing labs alike’. U.S. producers feared that ‘the time and money associated with going through EC laboratories’ would ‘diminish their competitiveness vis-a-vis EC-based rivals.’ The many questions that were raised by U.S. business between July 1989 and February 1990 included (p. 6-24, and n. 51 on that page):


High-level meetings and the launch of the MRA

Berkeley Professor David Vogel summarises how the 1992 program led to American fears of a Fortress Europe, and pressure on the EC to allow American firms to participate in standards setting (an issue which I am not covering here), and certification:



He goes on to describe how negotiations between the U.S. Secretary of Commerce Robert Mosbacher and Internal Market Commissioner Martin Bangemann led to agreement to pursue mutual recognition agreements on conformity assessment, which were regarded as ‘extremely important’ at the time:


The EC/US Joint Communiqué of 31 May 1989 confirms that the two sides had agreed in principle to open exploratory discussions on mutual recognition of tests and certificates:


At the third Mosbacher-Bangemann meeting in June 1991, it was agreed that any mutual recognition agreements would entail designated U.S. conformity assessment bodies receiving EC notified body status:


 Subsiding fears

By then, however, fears of a Fortress Europe had subsided. Miller and Wasserman, writing in 1992, first describe the earlier concerns (p. 408):


and then say that they had eased over the previous two years, after U.S. companies had seen how the new regulations had been implemented in practice, and with the primary source of reassurance being that U.S. exports to the EC had in fact risen sharply:


Gilles Grin, currently Director of the Jean Monnet Foundation, has pointed out (p. 136) that even in 1988 there were some who saw that the fall of internal barriers in Europe would be beneficial to American business, and indeed that the Community was only doing what the U.S. had been encouraging it to do for decades:


Deliberate overreaction?

Grin goes further and suggests (p. 137) that the American fears may have been a deliberate overreaction designed to influence the 1992 program in the direction they favoured:


Even the USITC, in a review of development in 1989 in its 1st follow-up report (p. 6-17) noted that some U.S. firms ‘appeared to be operating on a “worst case” scenario’, and suggested that so long as the EC were to provide ‘reasonable opportunities for acceptance of U.S. tests’, the new system could prove to be an improvement of the present fragmented one:


Of course, an emphasis on the worst case may be a matter of responsible caution rather than of deliberate exaggeration to influence events, but it is not inconsistent with the latter.


It is striking how precisely the Community’s policy with regard to conformity assessment and third countries was geared to compliance with the GATT and the 1979 Standards Code in particular. On the one hand, imported goods were to meet the same technical requirements and have their conformity with these assessed by the same procedures. But on the other, there was to be no liberty for Member States simply to recognise certificates from laboratories established in third countries. Rather, there would have to be mutual recognition agreements, and these would have to be negotiated by the Commission on behalf of the Community as a whole.

Second, we have seen indications of a potential for EU-US relations to be fiercely competitive, with occasional wars of words erupting between them.

Third, it is worth considering whether the concerns over potentially protectionist features of the New Approach product legislation were deliberately exaggerated in order to secure policy outcomes that favoured U.S. firms. Whether due to U.S. pressure or not, the Community did soon make a decision to allow testing to be subcontracted to third country laboratories, greatly allaying American concerns. I intend to show how this unfolded in my next post.



The photograph is of the European Council meeting at Luxembourg on 2-3 December, at which the agreement that was reached (p. 7) ‘after thirty hours of relentless discussion’, included ‘wider use of majority voting in order to achieve a single large Community market’ by the end of 1992. Margaret Thatcher and Geoffrey Howe are in the right foreground, if I am not mistaken.


  1. G. Schaffer, ‘Reconciling Trade and Regulatory Goals: The prospects and limits of new approaches to transatlantic governance through Mutual Recognition and Safe Harbor Agreements’, Columbia Journal of International Law 29 (2002-3)
  2. C. Deveraux ‘The Mutual Recognition Agreements’ in ‘Case Studies in US Trade Negotiation’ by C, Deveraux, R. Lawrence, M. Watkins (Washington DC: Institute for International Economics). Another version with references can be found here.
  3. Before issuing an EC type-examination certificate under the AIMD Directive 90/385/EEC, the notified body has to carry out appropriate inspections and necessary tests in either case, whether to verify that the harmonised standards have been adhered to or, where the standards are not employed in full, to verify that the essential requirements of the Directive are met (Annex 3, 4.1 and 4.2).
  4. J. H. H. Weiler, ‘The Transformation of Europe’ The Yale Law Journal, Vol. 100, No. 8, Symposium: International Law. (Jun., 1991), pp. 2403-2483, at page 2462.
  5. P. Craig, G. de Burca, ‘The Evolution of EU Law’, 2nd ed. (Oxford University Press, 2011) p. 56
  6. Miller and Wasserman, ‘Trade Relations between the European Community and the United States’, Boston College International and Comparative Law Review, Vol. 15, No. 2 (1992) p. 408, n. 75. Link.

Leave a Reply

Your email address will not be published. Required fields are marked *