Open Plurilateral Agreements, InternationalRegulatory Cooperation and the WTO

Bernard HoekmanEUI and CEPR, Italy

Charles SabelColumbia Law School, New York

AbstractSustained high growth in many developing countries (‘the rise of the rest’) combined with long-standing World Trade Organi-zation (WTO) working practices hamper the ability of the WTO to perform its routine functions and paralyze efforts to adaptto new circumstances. For want of an alternative, preferential trade agreements have taken up some of the slack in addressingdifferences in domestic regulation of product safety, environmental and social conditions, but these are exclusionary and inef-ficient from a global perspective. In this article, we argue that a new type of agreement based on open plurilateral coopera-tion offers better prospects for groups of countries to explore and develop their potential common interests on regulatorymatters, while safeguarding core aspects of their national regulatory sovereignty and increasing the possibility of regeneratingthe WTO from within.

Policy Implications• A new instrument – the open plurilateral agreement (OPA) – should be created within the WTO to encourage issue or

domain specific regulatory cooperation outside of trade agreements.• Open plurilateral cooperation allows groups of countries to explore and develop their potential common interests on regu-

latory matters.• Open plurilateral agreements offer a mechanism to reduce the trade costs created by international differences in regula-

tion while ensuring national democratic control and accountability.• The determination of regulatory equivalence for purposes of membership of open plurilateral agreements should be

linked to provision of capacity building services.

Rapid growth in developing economies (‘the rise of the rest’)combined with long-standing working practices hampersthe ability of the World Trade Organization (WTO) to per-form its routine functions and paralyzes efforts to adapt tonew circumstances, above all the decreasing importance oftariffs and the increasing importance of regulatory differ-ences as impediments to trade. Domestic regulation of pro-duct safety, environmental protection and competitionpolicy are increasingly addressed in preferential trade agree-ments (PTAs), typically among regional trading partners(e.g., the Comprehensive and Progressive Agreement forTranspacific Partnership – CPTPP) or between a trading areaand a key partner (e.g., the Comprehensive Economic andTrade Agreement between Canada and the EU) (see e.g.,D€ur et al., 2014; Hofmann et al., 2018). Palanco and Sauv�e(2018) describe and discuss the coverage of regulatorycooperation chapters in recent trade agreements. Suchagreements often address matters covered by the WTO butwith additional rigor – so-called WTO-plus provisions – andmay also cover matters previously unaddressed by the WTO

– WTO-extra provisions, to use the terminology of Hornet al. (2010). PTAs by design are discriminatory. Often theyare exclusionary as well in that they are not open to anycountry that wants to join them. PTAs are thus as much ormore a device for opting out of the current system as foraugmenting it. Because they are broad trade deals – morelimited variants of the regime embodied by the WTO – PTAsare as likely to be challenged politically as instruments of anexcessive and unwanted globalization as the WTO itself.Despite, indeed often because of their innovative ambitions,PTAs in practice do little to extend the reach of multilateralrulemaking to urgent new topics while rendering the institu-tional foundations of global trade more fragile.In parallel to the expansion in use of PTAs, extensive reg-

ulatory coordination occurs outside the WTO, in sector-orproduct-specific international regimes (see Overdevest andZeitlin, 2018; Pacheco et al., 2018; Zeitlin and Overdevest,2019). In the case of trade in some environmentally sensitivegoods, such as forest products and palm oil, concerns thatcommerce in these goods does not further threaten

Global Policy (2019) doi: 10.1111/1758-5899.12694 © 2019 University of Durham and John Wiley & Sons, Ltd.

Global Policy

Research

Article

endangered species, contribute to deforestation or under-mine the rights of native communities have led to theemergence of national laws barring, for example, import ofwood harvested in violation of various legal requirements.Such arrangements have a direct bearing on the ability offirms to export products to markets where such regulationhas been adopted. The resulting regimes differ from PTAs inthat they are not discriminatory and not designed to beexclusionary. The burden of coordinating consistent imple-mentation and enforcement of many similar yet subtly dif-ferent national norms and procedures arising within thesepatchwork regimes is great; and the strains of bearing thatburden, together with many other signs of fragility, rein-force, if only indirectly, the sense we can no more count onthe spontaneous self-organization of states facing commonthreats to protect core values in global commerce than wecan rely on the comprehensive, top-down response of PTAs.

This article extrapolates from successful cases of far-rang-ing or deep international regulatory cooperation in domainssuch as sustainable forestry and trade in timber, food safetyand civil aviation to propose a novel vehicle – the openplurilateral agreement (OPA) – for groups of countries todevelop potential common interests in discrete regulatorydomains within the WTO, while safeguarding key aspects oftheir national regulatory sovereignty. To avoid the pitfalls ofPTAs and patchwork regimes, OPAs will be formally orga-nized as multilateral agreements. We argue that an OPA is avaluable regulatory instrument in and of itself; that the WTOhas multiple potential roles – ranging from transparency ser-vices and technical assistance to dispute resolution – to playin establishing such an institutional facility; and that doingso can support the regeneration of the WTO from within.The proposal is at once a concrete project for reform and aheuristic for understanding the mismatch between currentdevelopments in trade governance and how conceptually tocome to grips with it.

The article is organized as follows. The following sectionintroduces the main features of what we understand whenusing the term OPA. The third section discusses briefly howchanges in the organization of production have made regu-lation – and importantly regulation requiring deep and con-tinuing cooperation to establish the equivalence of tradepartners’ regimes – an increasingly prominent issue in traderelations. We then present case studies of two kinds ofregimes that most closely approximate OPAs, starting withthe bottom-up development of the regime to regulate tradein forest products. This type of regime is standalone andincompletely integrated. Only recently has there been a pro-liferation of the institutions for mutual monitoring andreview that would make for effective implementation andenforcement of its norms. Its development illustrates thepossibilities but also the difficulties of achieving cooperationwithout any overarching institutional support of the kindinclusion in a reformed WTO could provide. To better under-stand the operation of governance based on regulatoryequivalence, we consider an alternate approximation toOPAs: bilateral regimes that are deeply integrated throughjoint review and deliberation. The case studies here are of

food safety and civil aviation, typical of the kinds of arrange-ments, based on treaties or executive agreements, that haveemerged ad hoc in response to the manifest need for close,regular collaboration in the absence of generally acceptedmethods for institutionalizing it.In the fourth section we take up the blockages in the

WTO. For analytic purposes, we distinguish immediate orproximate drivers of obstruction having to do with changesin the external environment and WTO working practicesfrom structural causes having to do with the very constitu-tion of the WTO itself as the embodiment of a particularunderstanding of how to regulate global trade. Among theproximate causes figure the redistribution of power fromthe advanced to developing countries, the backlash againsttrade as one emblem of a threatening globalization in manyof the advanced countries, and the understanding of con-sensus rules that has come to be habitual in WTO decision-making. Of the fundamental causes, most significant issurely the assumption that trade is self-evidently a globalproblem, and as such should be regulated by a single, com-prehensive multilateral agreement, adopted and periodicallyrenewed by consensus of the members acting as a sover-eign body. From this it followed that the only internationaltrade agreements compatible with the WTO itself are eitherthose that effectively exclude other members but reflectcomprehensive ambitions and ‘sovereign’ legitimation on asmaller scale – PTAs, covering ‘substantially all trade’ of sig-natories as homuncular versions of the WTO – or narrower,domain-specific agreements that can be nested in the WTObecause they are subject explicitly or implicitly to consensusrequirements of the body of members as a whole. Twooptions currently exist, at least formally, for such domain-specific agreements among a subset of the WTO member-ship: plurilateral agreements (PAs) allowing signatories todiscriminate against non-parties if all WTO members assentto their registration, and critical mass agreements that, incontrast, do not require consensus because they do not dis-criminate against non-parties. The former are practicallyirrelevant because of the consensus requirement, the latterbecause they create insurmountable free rider problems, asnewcomers have the trade benefits accorded to the originalparties without undertaking the regulatory adjustmentsrequired of the initiators.The fifth section turns to OPAs as an alternative form of

smaller group cooperation and discusses how the WTO canhelp to make OPAs truly open and how OPAs can open theWTO. Precisely because OPAs authorize and encourage aprocess of consensus building and transformation that is(initially) partial, involving some but not all of the members,and are embedded in the routine operation of the WTO,rather than accorded the legitimating dignity of deliberationin a ‘constituent’ or ‘sovereign’ bargaining round, they seemto offend the very constitution of the WTO. We argue to thecontrary. Because OPAs are provisional, placing regulatorycooperation under continuing democratic control, they arerespectful of national sovereignty and thus deeply consis-tent with the legitimating principles of the WTO. Yet theyalso offer a mechanism for the generation of rich and

© 2019 University of Durham and John Wiley & Sons, Ltd. Global Policy (2019)

Bernard Hoekman and Charles Sabel2

reliable norms to guide commerce and, when necessary,direct adjudication. In regulatory affairs, as in advanced con-tracting among jointly innovating firms, the very process ofdetermining whether each party’s efforts is serving the jointpurpose clarifies that purpose, helps the parties betterachieve it, and allows each to assess whether the other hasthe intention and capacity to do so. Put another way, andconnecting our proposal for OPAs to a current of thought ininternational relations that reaches back to the work ofChayes and Chayes (1993, 1998) on the new sovereignty,the very articulation of norms touches off a process of sociallearning that helps the actors adhere to them, or under-stand early on – before much harm is done – that a coun-terpart will or cannot. The sixth section concludes.

Open plurilateral agreements

We conceptualize OPAs as agreements authorizing regula-tory authorities in a given domain to cooperate with peersin member countries to facilitate export of goods and ser-vices produced in their jurisdiction to other member mar-kets, subject to critical, continuing determination by theregulators themselves and their national political oversightbodies, that products and production processes meet evolv-ing national standards. Cost reductions can be achievedwhen regulatory authorities accept, upon mutual examina-tion, a determination by their counterparts that an expor-ter’s products conform to the importer’s standards,eliminating redundant compliance checking. Ultimately,going much further, authorities in OPA member countriesmay upon additional mutual scrutiny recognize that, differ-ences in their respective standards or conformance testsnotwithstanding, their systems produce equivalent decisions,making approval by one tantamount to approval by theother, subject again to continuing verification that this isthe case.

Open plurilateral agreements differ from PTAs in threeways. First, PTAs are comprehensive trade deals and thusentail cross-issue linkages. A commitment in domain A may,as part of an encompassing bargain, be compensated byconcessions in domain B. OPAs are domain-specific: commit-ments or cooperation in one domain are not bargainingchips in negotiations about policies in others. Second, PTAsare enduring, detailed agreements, fixing the terms of tradefor the foreseeable future, subject only to presumably mar-ginal periodic adjustments. OPAs establish frameworks forcontinuing reciprocal review of existing regulatory standardsand their implementation, and joint evaluation of potentialalternatives and adaptions to new developments. Putanother way, OPAs entrench particular values more deeplythan PTAs, but treat the precise expression of those valuesin regulatory rules as more easily contestable and corrigible.Because they treat standards as values in themselves,expressive of deep and abiding social and political commit-ments, rather than counters in periodic rounds of trade bar-gaining, and because they institutionalize ongoing review ofthe interpretations and elaboration of those commitments,OPAs make it easier for nations to assert distinctive aspects

of their sovereignty than PTAs. They also make it easier forpolities, wary of elites that have proven inattentive to reper-cussions of globalization, to hold regulatory authorities andtheir political overseers to account.The third difference between PTAs and OPAs concerns

the conditions under which countries not party to the origi-nal agreements can subsequently join. To conform to therules of the WTO, PTAs must, as noted, be comprehensive.To accede to a PTA a new member must agree to all theterms of the elaborate compromise struck by the originalsignatories. In practice this makes most PTAs closed agree-ments.Because they are domain-specific, accession to an OPA

requires a narrower and more limited commitment than aPTA. A candidate member must only undertake to meet theregulatory requirements established by OPA members asthese apply to the particular class of goods or services cov-ered by the agreement. Insofar as members of an OPArequire only equivalent performance – not identical proce-dures or institutions – in conformance testing, standard set-ting and enforcement in each domain, they permitcandidate members to produce the required regulatory out-come by the process best suited to their own traditions andconditions. In further contrast to a PTA, accession to an OPAcan be achieved stepwise, with candidate members estab-lishing the equivalence of their methods in one phase ofthe regulatory process, or one or another product of a par-ticular class, then another, and another, so that tradeexpands and collaboration deepens even when full equiva-lence of regulatory systems is a distant goal. For all thesereasons we call this type of plurilateral agreement open.What types of policy areas and regulation lend them-

selves to OPAs? And of the set of potential OPAs, which areappropriate to pursue under the aegis of the WTO asopposed to a sector- or issue-specific basis independent ofthe WTO? These questions hark back to policy discussionsand academic debates in the 1990s regarding the bound-aries of the WTO. Prompted by suggestions to link trade(market access) to non-economic objectives, these debatesmost frequently focused on the desirability of making tradeconditional on labor and environmental standards, but con-sidered the advisability of linking trade and competition pol-icy as well. This ‘trade and . . .’ discussion led to variousproposals for criteria to justify inclusion of a policy area inthe WTO.Hoekman and Kostecki (2009) suggested a simple rule of

thumb for decision-making in such cases: whether an issuehas a direct bearing on the contestability of markets forgoods and services, or, more bluntly, affects market access.Maskus (2002) proposed a policy area should be linked totrade if it is: (1) ‘trade-related enough’, in the sense of gen-erating trade impacts that imply significant negative interna-tional externalities; and (2) can be characterized bycoordination failures that prevent countries from realizingshared interests through adoption of common (interna-tional) standards, making the use (the threat) of trade sanc-tions an efficient instrument to induce cooperation. Otherstook the view that a broader focus was needed, including

Global Policy (2019) © 2019 University of Durham and John Wiley & Sons, Ltd.

Regulatory Cooperation and the WTO 3

whether non-pecuniary, negative externalities could bereduced by eliminating differences in regulation specifyinghow a product is produced, for example, ensuring produc-tion is environmentally sustainable or that it conforms tointernational labor standards.

Strong resistance by many governments to consider disci-plines on social and environmental policies within the WTOshifted efforts in this direction to PTAs. As a result, in thecase of the EU and the US, PTAs incorporate provisions onnon-trade policy areas. Most notably, the 2015 EU tradestrategy (Trade for All) calls for EU trade and investment pol-icy to support and promote EU values and standards enu-merated in the Lisbon Treaty relating to human rights, laborrights, environmental, health and consumer protection, sus-tainable development, economic governance (corruption),and the EU’s broader objectives of protecting people andthe planet. Thus, to a significant extent the reality today isthat major trading powers make market access conditionalon compliance with national or international regulatorynorms that apply in their domestic markets. Further exam-ples are the EU timber regime discussed below, the EU dataadequacy regime and the judgment by the CJEU that theInvestment Court System, which replaced investor-State dis-pute settlement (ISDS) arbitration of investment-related dis-putes, must respect (be consistent with) EU values. Suchregimes are piecemeal and often idiosyncratic, with import-ing nations imposing different standards (or enforcing stan-dards in different ways), with confusingly different, andoften unpredictable effects on the ability of exporters to sellgoods and services. OPAs offer a mechanism to graduallymultilateralize such regimes.

International regulatory cooperation: The future ison the way

Regulation has become more prominent in trade in partsimply because earlier rounds of trade negotiations dramati-cally reduced tariffs. In trade between advanced countriestariffs in most goods are now at nuisance levels in that col-lection costs may exceed the revenue they produce. Regula-tions become more conspicuous as perceived barriers totrade because other restrictions have been reduced orremoved altogether.

At the same time, deep and now familiar changes in theorganization of production have put new burdens on regu-lators domestically and increased the need for collaborationwith capable foreign peers. The pace of innovation hasincreased with the advent of the knowledge economy. Glob-alization responds to and further accelerates the changes.Final producers, no longer limited by the technologicalchoices of their internal production units, as they were inthe days of vertical integration, collaborate with capablesuppliers to combine novel technologies in new productsrapidly. More rapid innovation increases the risk of introduc-ing latent hazards: hidden defects that go unnoticed whenproducts initially are approved for sale. The regulator’s tradi-tional emphasis on fixing, ex ante, conditions for marketoperators, and assuring compliance with them proves too

limited. Ex ante or pre-marketing review and approval isbeing complemented by greatly increased attention to post-post-approval or ex post monitoring of the performance ofthe good or service in the market. This change is marked bythe diffusion of requirements, mandated by public regula-tion and by private standards enforced by contract amongsupply chain partners, to report breakdowns in control thatthreaten the safety of products or production processes, totrace these incidents to their source and root cause, and totake corrective and preventative action. The just-in-time dis-ciplines that enable smooth operation within supply chainsalso equip firms to detect and report such breakdowns.The need for an increased, and increasingly fine-meshed

international regulatory cooperation follows directly. In aworld where firms source inputs globally and potentiallyhazardous products are globally distributed, regulatory sys-tems must encompass relevant trade partners to be effec-tive. National regulatory systems that provide early warningof possible dangers in the goods traded among them, andrespond quickly to breakdowns, must have confidence inone another’s oversight of the linked producers. Amidstrapid change, that confidence can only be maintained byongoing scrutiny of partners’ regulatory practices in particu-lar domains and joint investigation and, perhaps mutualadjustment, when differences and failures are detected – allwithout jeopardizing the right of unilateral withdrawal fromagreements that is the continuing prerogative of sover-eignty.At its best, such international collaboration allows regula-

tors routinely to reevaluate their own performance in the lightof others’ successes and failures, improving on the ability ofany national regime in isolation effectively to defend the val-ues entrusted to it. When this is so, reductions in trade costsor increased market access and improvements in regulatorycontrol of markets – reflecting shared values and respect forpersisting differences – can be complementary, not conflict-ing. Moreover, given that trade negotiations are often decriedtoday as a technocratic or producer-interest driven conspiracyagainst national values and the popular will, the continuingand searching mutual examination by regulatory equivalencecreates novel possibilities for transparent and publiclyaccountable decision-making, and thus for reconciling sover-eign self-determination with the stepwise extension of eco-nomic exchange and regulatory cooperation.The pressure for deep international regulatory cooperation

of this kind is felt most urgently where the risks of latent haz-ards, introduced inadvertently by technological change or byactors seeking to profit from deception, are the greatest. Asimportant are national preferences that are reflected in regu-lations pertaining to the process of production at home andin foreign markets from which products are sourced. Exam-ples include compliance with international labor standardsagreed in the ILO, national data protection and privacy orenvironmental sustainability norms. These pressures canresult in stand-alone institutional arrangements under two,contrary sets of conditions. If exporting countries and import-ing countries have sharply different interests that also divergewithin each group, the result is a piecemeal regime or regime

© 2019 University of Durham and John Wiley & Sons, Ltd. Global Policy (2019)

Bernard Hoekman and Charles Sabel4

complex of the kind observed in forestry and palm products.If, on the contrary, exporters and importers have largely over-lapping interests – for example, trade is mostly intra industryof the type that occurs with value chains, with each countryproducing components or variants of what the other pro-duces – the result can be a bilateral agreement to seek regula-tory equivalence. We take up these cases in turn.

The problems and possibilities of building an ad hocregime

Deforestation accelerates climate change, destroys irreplace-able habitats, and dislocates communities of first peoples andother forest dwellers. It was almost self-evident that protec-tion of forests by limiting illegal logging and trade in illegallylogged timber would figure prominently in the reform agendathat emerged as concern for the environment first crestedpolitically in the early 1990s. Regime building in this domainhas proceeded by fits and starts and in bits and pieces. Thecloser the institutional set up gets to being fully operational,the clearer it becomes that its effectiveness will depend onthe kind of deep and continuing regulatory cooperationincreasingly central to (but buried within) PTAs and moreclearly exposed in the bilateral arrangements to be describednext. Whether the current arrangements can formalize therequired mutual monitoring and review is an open question.At the least developments in regulation of trade in illegal log-ging suggest that it may be possible, under favorable condi-tions, to create a plurilateral regime with many of thecharacteristics of an OPA outside (though in conformity with)the WTO, but that the coordination costs of such ad hoc solu-tions are likely to be dauntingly high. The potential gains fromcreating an institutional scaffolding to reduce those coststhrough economies of scale and scope in institution buildingare correspondingly great.

Early efforts to secure a legally binding, comprehensiveconvention on forest preservation failed at the 1992 EarthSummit in Rio. In the aftermath, the World Wide Fund forNature (WWF), a British NGO, recognizing the immediatelimits of public action, convened stakeholders representingthe private sector, environmental organizations, and vulnera-ble communities – grouped in separate ‘chambers’ and bal-anced between the Northern and Southern hemispheres –to establish the Forest Stewardship Council (FSC). The FSC isan organization for setting voluntary standards for forestpreservation and a system for demonstrating compliancewith them. Broad, continuing international discussion underthe auspices of the World Bank, the UN IntergovernmentalForum on Forests and the G8 of the need for extensivereform of forestry governance in timber exporting countriesin addition to voluntary standards and certification embold-ened the EU to unilaterally advance the Forest Law Enforce-ment, Governance and Trade (FLEGT) action plan in 2003.Under pressure from NGOs, some exporters and new legisla-tion in the US (the Lacey Act), that measure was reinforcedby an EU Timber Regulation (EUTR) a decade later. FLEGTcreated preferential or expedited access to the EU singlemarket for countries – all developing – that cooperate with

the EU in fostering sustainable forestry and deterring illegallogging. The EUTR creates important incentives for theextension of these cooperative efforts, and for compliancewith the obligations arising under them by imposing poten-tially substantial penalties on firms that introduce timberinto the EU from non-cooperating exporters, or that fail toexercise due diligence in ascertaining that the products theycommercialize in fact meet the agreed requirements.Under FLEGT, wood-exporting countries are invited to

enter bilateral voluntary partnership agreements (VPAs) withthe EU. These are in fact binding agreements. A party to aVPA agrees to review and periodically revise its current for-estry law and administration with national stakeholders toachieve the aims of FLEGT and international agreements towhich it is a signatory; it must also establish a correspond-ing timber legality assurance system (TLAS) – under theoversight of an independent auditor, but increasingly withcivil society participation – to ensure the revised legalrequirements are respected from harvesting to export. Forits part the EU, to induce agreement to the terms ofthe VPA, facilitates access for FLEGT-licensed timber to itsmarket and supports (with the help of donor institutionswith the requisite expertise) the capacity building of publicand private actors needed for compliance. A joint commit-tee of representatives of the EU and the partner countrymonitor implementation of the agreement and settle dis-putes not so intractable as to require arbitration. The jointcommittee may also recommend adjustments to the agree-ment, for example in relation to provision of capacity-build-ing services. That the agreements are jointly governed andprovide technical and adjustment assistance assuaged con-cerns about the WTO conformity of the VPAs and helpedwin the assent of developing countries, whose oppositionhad earlier frustrated negotiation of a global forest conven-tion.The US Lacey Act amendment and the EUTR produced a

cascade of responses that, as intended, increased the scopeand grip of the new protections. As of 2018, seven countriesincluding Cameroon, the Central African Republic, andLiberia in Africa; Guyana and Honduras in Latin America andIndonesia and Vietnam in South-east Asia had agreedVPAs with the EU. Formal negotiations or preparations fornegotiations are underway with eight more, similarly dis-tributed geographically. If all these negotiations succeed,VPA countries will account for close to 80% of global tradeand EU imports of tropical wood products. Other advancedcountries, including Australia and South Korea have enactedoutright prohibitions or legislation with deterrent effect onthe import of illegally harvested wood. The leading transna-tional private forest certification organizations, such as theFSC, have adjusted their standards to meet EUTR require-ments.Increasing participation in the emergent regime has not

been matched by advances in effective administration andenforcement. On the contrary, both VPA partner countriesand counterpart authorities in the EU and its member stateshave struggled to comply with their obligations under theagreements, though the pace of progress seems to have

Global Policy (2019) © 2019 University of Durham and John Wiley & Sons, Ltd.

Regulatory Cooperation and the WTO 5

increased recently. It took Indonesia, the first country toexport FLEGT-licensed timber, nine years from the start ofnegotiations with the EU to establish a reliable process fordoing so; in Ghana, which is expected to follow suit shortly,implementation has taken longer still – and these are theleaders in their cohort. Within the EU, member states wereslow to establish competent authorities (CAs), which areresponsible under the EUTR for investigating ‘substantiatedconcerns’ regarding trade in illegal timber reported by thirdparties, typically civil society organizations; even where CAsexisted, they were often slow to develop procedures for ver-ifying operators’ compliance with their due-diligence respon-sibilities. It took four years and a combination ofinfringement proceedings against stragglers, additional guid-ance and revisions in the regulation itself to bring essentiallyall EU member states into compliance. Only recently hasthere been an increase in enforcement activities rangingfrom assessments of due-diligence systems and requests forcorrective action to the imposition of fines.

These delays and difficulties should not be surprising. Asdevelopment banks and donor organizations know too well itis fiendishly difficult to build state capacity in ministries innational capitals; it is still harder to build capacity in theremote provinces – far from central authority, often governedby local satraps whose livelihood is trafficking of various kinds– where illegal logging, and its concealment by the genera-tion of false documents, is often a mainstay of the economy.Problem solving in such settings, as is typically the case withregulation of latent harms, proceeds in steps, with successive,apparent solutions repeatedly revealing new facets of the ini-tial difficulty until root causes can be exposed and addressed.Not uncovering unexpected difficulties in the prolonged, ini-tial phases of such investigation is more likely a sign of self-deception than precocious success. The source of the EU’s dif-ficulties in complying with its own requirements is no mysteryeither. When the EUTR came into effect, the member stateshad no national systems for making the determinations of thelegality of newly marketed wood products expected of expor-ters under the regulation. Such systems had to be constructedfrom scratch within the EU even as the definitions of legalityand procedures for ascertaining conformity were taking shapein the VPA discussions. Here too reports of rapid, frictionlesscompliance, had there been any, should have triggered suspi-cion, not self-satisfaction.

Recent assessments of the FLEGT Action Plan by the Euro-pean Court of Auditors in 2015 and an independent evalua-tion team a year later make clear that further progress inimplementation and enforcement will depend on intensifiedcooperation of the actors within the EU and between themand the actors in the VPAs taken separately and together.The new initiatives include joint assessments of progresswith the partner countries to establish multi-year roadmapsfor implementation and monitoring; creation of tools forearly detection of problems; and, in partnership with expert,outside organizations, programs to extract and make avail-able for general use practical lessons of experience underthe VPAs in areas such as shipment testing and timbertraceability.

These initiatives mesh with and reinforce an explosion ofinstitutionalized cooperation induced by the growing aware-ness of failures in implementation and enforcement. TheEuropean Commission, for example, convenes the CAs fivetimes a year in a EUTR/FLEGT Expert Group to review EUTRenforcement, make authoritative determinations in disputesbetween producer countries and CAs, and pool informationwith the aim of defining best practices and issuing guidance.The CAs have on their own established an informal enforce-ment network, with a confidential website for sharing inspec-tion reports, and notifications of substantiated concerns ofillegality. Sub-groups of CAs have begun to map supplychains in high-risk countries and assess the adequacy of com-panies’ due-diligence procedures. This cooperation extends tojoint training and inspections; it is increasingly formalized inregular, regional exchanges among CAs from, for example,the Nordic and Baltic countries and the Mediterranean. Finally,both the Commission and the CAs cooperate more and moreclosely with regulators in the Timber Regulation EnforcementExchange (TREE), a network connecting public officials, civilsociety organizations, and business representatives in the EU,US, Australia, Canada, and East Asia along with officials fromInterpol to review enforcement developments in high-riskregions. TREE discussions, for example led the US Departmentof Justice to introduce EUTR due diligence requirements intothe consent decree entered with a firm alleged to haveimported illegal timber, although these requirements are notanticipated in the amended Lacey Act.The end of the coordination problems is not in sight. To

note one of many pressing issues: the EUTR encouragesoperators to rely on qualified, private legality standard set-ters and verification organizations such as the FSC, whileassigning continuing responsibility for ascertaining the relia-bility of private systems. It is unclear how far that responsi-bility extends; at the limit, ad absurdum, the operator wouldhave to make its own legality checks to confirm the validityof the audit system’s certifications, rendering the latterpointless. Enforcement of the EUTR is the responsibility ofmember state CAs and courts. Even if CAs follow commonguidance with respect to this and other questions of inter-pretation, courts disagree in their interpretation of regula-tions, exposing operators to what seems to them,understandably, arbitrary differences in liability (see AsserInstitute, 2017; IKEA,2018; Nathan et al., 2018). Somemethod of harmonizing judgements as well as guidance willhave to be found, or even well intentioned actors will des-pair of participating. Given the continuing and in manyrespects unforeseeable changes in supply chains and envi-ronmental understanding and other imponderables, prob-lems of this type will be unending, or, rather, continue inever new forms. ‘Stabilizing’ the regime will require furtherdevelopment of the institutions of continuous coordinationnow proliferating of the kind that would be provided by anOPA, as we will describe below. If they were to succeed,these cooperative measures would likely transform the for-est protection efforts into a novel, plurilateral variant, out-side the WTO, of the bilateral mutual-equivalence regimesdiscussed next.

© 2019 University of Durham and John Wiley & Sons, Ltd. Global Policy (2019)

Bernard Hoekman and Charles Sabel6

Regulatory equivalence in trade: bilateral regimes

To see how OPAs can be established and how their gover-nance might function we look to two close, bilateral approx-imations. First, the steps leading to the Food Safety SystemsRecognition Arrangement – a declaration of regulatoryequivalency – signed by the Food and Drug Administration(FDA), the Canadian Food Inspection Agency, and theDepartment of Health of Canada. Second, the agreement on‘cooperation in the regulation of civil aviation safety’through the Bilateral Aviation Safety Agreement (BASA)entered into by the US and EU in 2011. The process ofestablishing equivalence depends crucially on exactingreciprocal scrutiny of regulatory ends and means – and thuspresumptively affords heightened possibilities for domesticreview and accountability in the partner countries. The gov-ernance of the eventual bilateral arrangement further insti-tutionalizes mutual scrutiny, preserving the sovereign rightto disengage from the regime in whole or part while routin-ing possibilities for ever deeper and broader cooperationwithin the scope of the agreement.

In the US, outbreaks of food-borne illness transmitted byleafy greens (especially dangerous because often eaten raw)led California wholesalers in 2006 to create a regime – con-tractual, but enforced by a state inspectorate – requiringgrowers to conduct a hazard analysis of the critical controlpoints (HACCP) review of their farms. This aims at identify-ing where pathogens could enter the production process,and proposing and testing methods of avoiding or mitigat-ing those risks. The Food Safety Modernization Act (FSMA)of 2010 codified this regime, extended it to many moreproducts under the jurisdiction of the FDA and establishedprocedures for responding to breakdowns in controls. Asthe US was modernizing its food safety legislation, Canadawas doing the same: like the FSMA, the Safe Food for Cana-dians Act (SFCA), passed in 2012, mandated HACCP controlsfor the entire supply chain, incident reporting and traceabil-ity; and like the FSMA the SFCA anticipated close coopera-tion with the regulators in key trading partners. For bothcountries, the determination proceeded in two steps: first, acareful desk review of the partner’s standards and proce-dures to ascertain how general organizational goals aretranslated into specific standards and routines, and to verifythat the routines are routinely followed. Then observationby field teams of the partner’s audit of a range of food-pro-cessing plants and reference laboratories to understand howexperience on the ground is translated into decisions anddocuments.

The FDA’s desk review began with the development of anInternational Comparability Assessment Tool (ICAT) forassessing the robustness of a trading partner’s food safetysystem in ten domains such as inspections and responses tooutbreaks of food-related illness. A capable partner authorityis expected to conduct ‘periodic self-assessments and qual-ity assurance reviews’ of its inspection and other programsto ‘determine areas or functions . . . that need improvement,to develop improvement plans and to establish timelines forimplementing improvements;’ similarly a robust food-safety

system is expected to include periodic review of enforcementactions ‘to assess areas in need of improvement or correctiveaction,’ and update ‘policies and practices based on findings.’The ICAT review also included presentation by Canadian offi-cials at the national and provincial levels of case studies that,starting with source documents such as audit reports, docu-mented the chain of decision-making in particular productrecalls and enforcement actions against firms, allowing the USreviewers to determine whether the information generated bythe Canadian food-safety system was effectively used to serveits expressed goals. This extended desk review was then com-plemented by weeks of site visits in which a pair of three-member, interdisciplinary US teams (one in the West ofCanada, one in the East) prepared to shadow Canadian inspec-tors in various plants by reviewing their training records, thenobserved the actual inspection of processing facilities, withattention to the records consulted and interactions with keymanagers. The entire process is meticulously described in areport in which the FDA reviewers recommend ‘a positive find-ing of system recognition’ – current FDA lingo for a determina-tion of regulatory equivalence.As a second example illustrating the trajectory and gover-

nance mechanisms of bilateral, sectoral regulatory collabora-tion based on continuing mutual scrutiny, consider the US EUBASA. Civil aviation is, like pharmaceuticals, among the mostrigorously and successfully regulated industries: passengerfatalities per 100 million passenger-kilometers flown globallyin commercial air transport fell from 0.8 in 1960 to 0.08 in1980 and 0.03 in 1990; since then it has ranged between 0.05and 0.01 (see ICAO, 2017). The International Civil AviationOrganization (ICAO), formed under the Chicago Conventionon International Civil Aviation in 1944, provides the frame-work for international regulation in the sector. ICAO estab-lishes a ‘mutual acceptance’ regime in which the certificationby one signatory that equipment or flight crews under itsjurisdiction meet ICAO standards is accepted by other signa-tories. Mutual acceptance does not amount to agreement ofregulatory equivalence for three reasons. First, ICAO onlyestablishes minimum standards; more demanding jurisdic-tions such as the US, EU, Japan, China, Brazil and Canada insiston more rigorous ones. Second, innovation outpaces thecapacity of ICAO’s 191 member states to establish new stan-dards, so, for example, it may be impossible to certify designsfor next generation equipment under ICAO. Third, even whenstandards are available and acceptable to all parties, thecapacity and willingness to engage in conformance testingvaries greatly among the signatories, and some will notaccept the certifications of the others (see Jennison, 2013; Eis-ner and Parker, 2016).For all these reasons, states insistent on higher civil avia-

tion standards enter into bilateral agreements with peers.Beginning in 1996, the US negotiated BASA’s: formal frame-work agreements in which the partner countries by treaty orexecutive order authorize their respective air safety authori-ties (called technical agents or TAs) to examine and revieweach other’s practices and treat those found equivalent ascommon technical implementation procedures (TIPs) for

Global Policy (2019) © 2019 University of Durham and John Wiley & Sons, Ltd.

Regulatory Cooperation and the WTO 7

certification purposes. Of these agreements the EU-US BASA,covering principally the airworthiness of equipment fromdesign to manufacture and maintenance, as well as confor-mity to environmental standards, is the most comprehensiveand developed, with a formal governance structure for reso-lution of disputes and possible extension of the agreementto additional areas of air safety (see Jennison, 2013).

Like the ICAO, the EU-US BASA establishes the principleof mutual acceptance, but in a distinctive sense: each partyinsists on compliance with its own, distinct standards, butagrees to rely ‘to the maximum extent practicable’ on theother party’s certifications that its products or services do so(see EU-US BASA, 2011, Annex 1, 3.2.4). Formally this meansthat an Airbus designed in the EU must be certified by theTA in the EU, the European Air Safety Agency (EASA) asmeeting the airworthiness or ‘type’ requirements of the TAin the US, the Federal Aviation Administration (FAA), andthat a maintenance station in the US repairing Airbus equip-ment must be certified by the FAA as meeting EASA stan-dards.

In practice and by design, this kind of reciprocal accep-tance of compliance certifications shades into collaborationand de facto recognition of the equivalence of many stan-dards. With regard to maintenance stations, each authoritydetermined before entering the agreement that the other’sbasic system for quality control and reporting was equiva-lent to its own, and listed separately in the maintenanceannex to the BASA a small number of ‘special conditions’:‘requirements [in the relevant regulation of either party] thathave been found, based on a comparison of the regulatorymaintenance systems, not to be common to both systemsand are significant enough that they must be addressed’(see EU-US BASA, 2011, Annex 2X.5). Thus EASA can certify arepair facility in the EU performing work on US equipmentas meeting FAA requirements only if it complies with EASA’sstandards and meets the special conditions defined by theFAA (see EASA/FAA, 2011).

Similarly, in certifying the design of new aircraft types, theauthorities first determined the equivalence of their respec-tive methods of ascertaining an organization’s qualificationto produce reliable aircraft designs and a manufacturer’scapacity to maintain a reliable quality control system andthen provided for the exceptional cases where equivalencecannot be presumed. For example, early in the design pro-cess, when encountering novel, unregulated design ele-ments (also called ‘special conditions’) the FAA and EASA,separately or together, can issue new standards maintaining‘a level of safety equivalent to that established in the [exist-ing] regulations’; likewise, either authority can waive theobligation to conform with a particular certification require-ment when differences are thought to be inconsequential,or find that different design features or test methodsachieve an ‘equivalent level of safety.’ Only if a difference inthe scope and stringency of requirements does not fallunder one of these exceptions will one of the authoritiesfind a ‘Significant Standards Difference’ and (if further dis-cussion does not resolve the issue) declare the difference a‘validation item,’ meaning that it will test for itself whether

the other party has made an adjustment that meets its stan-dard (see Eisner and Parker, 2016). Regulatory cooperationin the sense of enlarging the scope of reciprocal acceptancealso extends to consideration of deep changes in the natureof design standards – for example, from specifications ofpermissible equipment to specifications of the level of per-formance the proposed equipment must attain – that reflectthe broad changes in the organization of production andthe overall context of regulation set out above.The governance structure created by the Agreement is

designed to encourage resolution of disputes arising undercurrent arrangements but also to extend regulatory coopera-tion. As its name indicates, the Bilateral Oversight Board,including representatives of the TAs (and, for the EU, repre-sentatives of the member state regulatory authorities withcontinuing air safety jurisdiction), reviews progress underthe BASA and sets the agenda for further reform. It is thefinal arbiter of disputes and has explicit authority to approvenew, domain-specific annexes. A Certification OversightBoard, composed of representatives of the TAs with exper-tise in airworthiness certification and environmental testing,and a Joint Maintenance Coordination Board, with TA repre-sentatives whose expertise is in repair, coordinate the tech-nical discussions between the authorities in their respectivedomains and whenever possible resolve disputes arisingfrom those discussions, referring only intractable ones to theBilateral Oversight Body. This ensures that disputes are nor-mally resolved at relatively low levels of the administrativestructure, by persons likely to have deep knowledge of theissues, rather than by higher authorities with limited under-standing of current practices. If a party, after the fruitlesspursuit of a remedy, loses confidence in a class of approvalsissued by the other, it suspends acceptance of only thatkind of approval, without disturbing the remainder of theagreement. Because of this severability, each authorityunderstands that the other could indeed act on a particu-larly vehement objection to a test or standard without fearof precipitating a political crisis; the credibility of this threathas a deterrent effect that reduces the chances that thepower of partial suspension being exercised.Are civil aviation and food safety exceptional cases, not

examples of general tendencies in the development of regu-lation and regulatory cooperation? A recent study cautionsagainst generalization, arguing that the role of ICAO as aninternational standard setter and the high degree of ex postliability for aviation accidents – the near certainty that negli-gence will be uncovered and heavily sanctioned – make thecase unusual, if not singular (see Eisner and Parker, 2016).Industry-specific organizations with the authority to set mini-mum standards for their members and to frame the agendafor further reform are today pervasive. The Codex Alimentar-ius plays this role globally for food safety (helping to diffuseHACCP-based regulation), as does the International MaritimeOrganization in maritime safety, and the International Con-ference on Harmonization in the area of pharmaceutical reg-ulation. Such organizations do not displace national,bilateral or (mega-)regional regulation any more than ICAOdisplaces the FAA, EASA or the results of their cooperation

© 2019 University of Durham and John Wiley & Sons, Ltd. Global Policy (2019)

Bernard Hoekman and Charles Sabel8

under the EU-US BASA. In all these cases the ‘global’ stan-dard setter, whatever its actual scope, crystalizes consensus,exposes new ideas and initiatives to informed criticism, gen-eralizes successes and calls attention to egregious cases ofnon-compliance with minimal norms. These organizationsare a common feature in the current regulatory landscape,not a distinctive outcropping that can explain the outcomein civil aviation or any other particular sector.

Ex post liability too is becoming commonplace, as thespread of incident reporting systems, including the obliga-tion to trace serious defects to their source, make it harderand harder to escape liability for negligence: faults are regis-tered in the course of production and failure, especiallyrepeated and systematic failure, to take corrective actionmakes it difficult to disclaim liability. Imminent changes,suggested again by the trajectory of civil air regulation,could well make liability all but inescapable. Through the1970s improvements in air safety largely resulted frominvestigation of aircraft accidents. As the number of acci-dents declined, reports of incidents – out-of-control eventsthat could lead to accidents – spurred further improve-ments. As the number of incidents declines ICAO nowargues that emergent hazards are best detected by continu-ous, real-time monitoring of engines and aircraft (see Rata-jczyk, 2011; Jennison, 2013). This example, too, could bemultiplied.

Nonetheless, the cases of bilateral cooperation presentedmight be exceptional on other grounds. However much theymight ramify – aircraft safety involves design, manufactur-ing, maintenance and operations; food safety spans farms,processing, retail distribution – these domains may be dis-crete or bounded in the sense that they do not implicatewhat we might think of roughly as core sovereign preroga-tives such as the monopoly control of force or the power oftaxation. Regulation in other areas might, however, lead tosuch a ‘sovereigntist’ ascent, the anticipation of which mightstop efforts at mutual accommodation through regulatoryequivalence before, or as they begin. A case in point mightbe banking and other financial services, where, as the recentfinancial crisis showed, failures of regulation in one jurisdic-tion can be rapidly transmitted to others, with catastrophicconsequences. Within the EU, for instance, efforts to estab-lish regulatory equivalence in banking may have touchedoff such an upward progression. To list only the most promi-nent reforms under discussion: regulatory equivalencerequires equivalent systems of deposit insurance, of resolu-tion of insolvent banks, of common, risk-adjusted capitalreserve requirements (which must be compatible with BaselIV international standards now under construction) and of a‘single rule book’ for evaluating the underwriting practicesof systematically important and ordinary banks. Perhapsthese regulatory domains are so intertwined that they canonly be managed by a single, central authority, so thesearch for mutual adjustment through regulatory equiva-lence results in the creation of a de facto banking or finan-cial-markets sovereign, if not a new sovereign tout court? Ifthere were many such domains regulatory equivalencewould remain marginal and the choice might be, in fact,

between a global technocracy and more or less democrati-cally accountable megaregional ones.The main counter consideration is that under uncertainty,

regulatory homogeneity is unacceptably risky. Imposing uni-form rules (for example, treating sovereign debt as riskless)is a recipe for disaster when, as all but inevitably happens,some of the uniform rules are inapplicable or unintentionallyinduce risky behavior themselves. By the same token, uni-form requirements choke off the exploration of diverse, con-textualized responses to local conditions and with them thepossibilities of mutual learning that is indispensable to rapidadjustment under uncertainty. Such considerations are verymuch in evidence in the EU debate on banking reform, ascounterweights to arguments for centralization and unifor-mity. The Chair of the Supervisory Board of the SingleSupervisory Mechanism for European Banks (European Cen-tral Bank), which helps create the framework for and moni-tors the results of banking supervision in Member States,cautioned against ‘pursuing a one-size-fits-all supervision’ infavor of an approach that ensures ‘consistency across insti-tutions and supervision tailored to credit institutions’ speci-ficities’ to ‘accommodate banking diversity’ – considered tobe ‘very desirable from a financial stability perspective’ (seeNouy, 2015). How to institutionalize a framework maintain-ing consistency while allowing and learning from diversity –the stuff of regulatory equivalence – remains an open ques-tion in this and other domains. In this case, at least, theapparent exception accords with the rule.Under the EU’s leadership, international regulatory coop-

eration in data privacy may well develop in analogous ways,with important spillovers in competition law. These exam-ples, together with the extensive but incomplete changes inthe governance of trade in forest products discussed earlier,are too numerous, and each too important in its right, to beoutliers. They are more likely to be forerunners. For now, wesimply regard them as illustrative of the broad changesreshaping regulatory cooperation as regulation becomesmore and more central to trade while minimizing, and ofteneffectively avoiding engagement with the WTO. We nextturn to the WTO and the question whether cooperation ona plurilateral basis akin to the examples discussed aboveoffers prospects for revitalizing the organization.

The WTO: Stuck in a rut?

The WTO struggles to perform routine, core functions suchas dispute resolution and the provision of current, completeand correct information on national trade policies. Its abilityto fulfill its mandate is in question, including as a convenerof trade negotiations. Members failed to conclude the firstround of multilateral trade negotiations launched underWTO auspices in 2001, the Doha Development Agenda. Thefailure of the Doha round in turn has precluded discussionof a new work program that includes the regulatory subjectsincreasingly central to polities and international business.Instead, this was taken up by many countries in the contextof PTAs (see, for example, Bollyky, 2017; Chase and Pelk-mans, 2015; and Mumford, 2014, 2018).

Global Policy (2019) © 2019 University of Durham and John Wiley & Sons, Ltd.

Regulatory Cooperation and the WTO 9

The WTO has been deadlocked by shifts in the power bal-ance among its members, by working practices that magnifythe paralyzing effects of those shifts, and by constitutionallimits on the acceptable forms of international cooperationwithin the global trade regime. The new, unstable balanceor imbalance of power will persist, but recent developmentssuggest the constitutional understanding of the WTO maybe interpreted more permissively, to allow precisely thekinds of deep international regulatory cooperation amonggroups of countries, without the prior consent of all mem-bers, that could lead by means of OPAs and other instru-ments to the elaboration of norms and procedures whichcould in time become broadly consensual.

The redistribution within the WTO of power, from theadvanced countries that presided over the creation of theorganization to the countries with rapidly growing econo-mies that now contend for influence within it, is commonknowledge. Developing and emerging economies havecome to account for almost half of the value of global mer-chandise trade and global GDP. Many rapidly advancingcountries no longer suffer the hegemony of the foundingpowers, such as the US; China openly challenges it. Nor isthe former hegemon eager to reassert authority within themultilateral system of its own creation. While growth ininternational trade, cross-border investment and knowledgeflows have supported major reductions in global povertyand increased real incomes of hundreds of millions of peo-ple in low-income countries, these developments have coin-cided with a decline in relatively unskilled employment inmanufacturing in OECD member countries. For many, tradehas become a potent symbol of a collection of alarmingand apparently uncontrollable technological, economic andpolitical changes. In virtually all advanced countries, broadcoalitions of parties and civil society groups on the left andthe right oppose trade agreements and the WTO as visibleembodiments of globalization.

To the burdens on decision-making created by theassertiveness of rising powers and the hesitations of the for-mer hegemon must be added the constraints that resultfrom WTO decision-making conventions, especially the work-ing practice that decisions are to be taken by (unanimous)consensus, and the convention that developing countriesare to be accorded special and differential treatment (Ber-telsmann Stiftung, 2018). The first permits any WTO memberto veto initiatives and block efforts that seek to go beyondthe issues agreed to comprise the (stalled) Doha Develop-ment Agenda. The second has allowed advanced develop-ing countries to offer less than full reciprocity in tradenegotiations and the application of certain WTO rules. Suchsystematically lop-sided arrangements are no longer accept-able to many high-income countries, first and foremost theUS. Many developing countries view special and differentialtreatment as a basic feature of the bargain underpinningthe WTO.

Decision by consensus is especially revealing of the self-understanding of the WTO because it holds sway in disre-gard of the formal admissibility of other procedures. Major-ity voting is allowed under the WTO’s governing statues,

but voting does not occur. To vote, it is widely held, wouldundermine the legitimacy of WTO decisions. Countries largeand small rely on the consensus practice as a guarantee thatthe results of negotiations are acceptable to them, ensuringthe ‘ownership’ of the WTO by members and their polities.Thus legitimated, consensus decision-making is also used toblock activities that have nothing to do with negotiations,such as setting the agenda of committee meetings or pro-posals to discuss a trade policy-related matter that is notcovered by a WTO agreement or not part of the DohaDevelopment Agenda. The result has been that new rulemaking has largely migrated to PTAs, rather than being con-ducted under the auspices of the WTO.The same insistence that substantial additions or depar-

tures from existing rules – lawmaking – requires (unani-mous) consensus among the members sharply limits theforms of international cooperation in trade that the WTOrecognizes as compatible with membership in the generalbody of the organization. In addition to PTAs, two optionscurrently exist for groups of WTO members to collaborateon a policy area. In the first option, critical mass agreements(CMAs), a group of countries agrees to specific policy com-mitments that are inscribed into their WTO schedules andapply on a non-discriminatory basis to all WTO members. Inthe second, plurilateral agreements (PAs) (under Art II:3WTO), a club agrees to undertake specific commitments in apolicy area and restricts the benefits of cooperation to clubmembers.The major example of a CMA is the Information Technol-

ogy Agreement (ITA) abolishing tariffs on information tech-nology products. The ITA has 82 participants, including the28 EU member states, and has increased global trade sub-stantially in electronic products. CMAs have also been con-cluded for services sectors – an example is a ReferencePaper on regulation of basic telecommunications whose sig-natories commit to apply specific regulatory disciplines on anondiscriminatory basis to all WTO members. Talks on apossible Environmental Goods Agreement spanning the EUand 17 other WTO members may result in another CMA.The disciplines established by CMAs apply only to the

subset of WTO member countries that sign on to them; butthe benefits extend on a most-favored-nation basis to allmembers, including those that do not participate. Becausecommitments are implemented on a nondiscriminatory basissuch agreements do not require consensus for incorporationinto the WTO – those WTO members that decide to join aCMA can simply inscribe the provisions of the agreementinto their schedules of commitments under the GATT and/orthe GATS, as appropriate (Hoekman and Mavroidis, 2017).The two key features of CMAs are that they are open – thepresumption is that any WTO member can participate if itdesires to, whether as part of the group that initially agreesto pursue cooperation on a matter, or after the establish-ment of an agreement – and that they are applied on aMFN basis. OPAs, we will see, are a special type of CMA.CMAs differ from the second type of agreement in that

the benefits of PAs can be limited to signatories. The mainexample of a PA is the Government Procurement

© 2019 University of Durham and John Wiley & Sons, Ltd. Global Policy (2019)

Bernard Hoekman and Charles Sabel10

Agreement (GPA). Because WTO PAs can apply in a discrimi-natory manner, their incorporation into the WTO is subjectto the approval of all WTO members, including those thatdo not join. Art. X.9 WTO stipulates that the Ministerial Con-ference of the WTO may decide to add a new PA to theexisting ones ‘exclusively by consensus’. Thus, the bar foradoption of new PAs is very high, as consensus allows anyone member to block the incorporation of a new agree-ment.

PTAs, CMAs and PAs among groups of WTO members allrespect the consensus principle – the juri-generative capaci-ties of the members of the WTO in their unanimity – but doso in different ways. PTAs honor the principle by emulation,adopting it as their own: They are formed by a comprehen-sive consensus of the parties that mirrors in the scope ofthe topics it covers, but not the extent of its membership,the ideal of the WTO itself – a consensus within a consen-sus. CMAs are exempted from application of the consensusprinciple but only because they only create new facilities fornonparties, and never limit existing ones. As a result, theyare not deemed to require assent of the sovereign body ofmembers. The situation of PAs is the reverse. They do limitthe rights of nonparties, and therefore are subject to con-sensus requirements so onerous they are seldom met.

The grip of the consensus principle has been importantbut was never complete. There are some recent signs that itis loosening. As a result, the WTO to some extent alreadyincorporates elements of regulatory cooperation on the linesof OPAs, and developments in the last several years suggestsubstantial, further opening. For example, the establishedWTO agreements on Technical Barriers to Trade (TBT) andSanitary and Phytosanitary (SPS) measures call on membersto base their product regulation on international standardswhere these exist. Since international standard setting hassome of the aspects of mutual interrogation and continuousadjustment characteristic of OPAs, the familiar TBT and SPSprovisions can be thought of as a roundabout obligation toincorporate the output of OPAs into member regulationwhen possible. The obligation is not binding. WTO membersare free to use national product standards as long as thesedo not restrict trade unnecessarily and are applied on anondiscriminatory basis. If international standards are usedthe presumption is that these cannot be contested. In thesame vein, provisions in the TBT agreement and the GATSencourage WTO members to establish mutual recognitionagreements (MRAs). MRAs concluded between two or moreWTO members must be open to participation by additionalmembers who meet requirements: closed MRAs are explic-itly prohibited – again a key feature of any OPA.

The December 2017 Ministerial Conference in BuenosAires announced a more direct challenge to the conventionof consensus. Different groups of WTO members launcheddiscussions on four new areas: on policies to assist micro,small and medium-sized enterprises (MSMEs); e-com-merce (digitial trade); investment facilitation; and domesticregulation of services. WTO members that joined thesegroups demonstrated that the consensus working practicecannot be used to prevent groups of countries discussing

issues of common interest. These initiatives offer hope formultilateral cooperation on the model of the OPA, and withit the possibility of the regeneration of the WTO as an insti-tution. The decision of 76 WTO members, including China,the EU and the US, at the 2019 annual meetings of theWorld Economic Forum in Davos to launch negotiations onrules to govern e-commerce and associated cross-borderdata flows illustrates this may be the shape of things tocome.

Why and when pursue OPAs in the WTO?

An important question is whether a regulatory OPA – or anundertaking to explore the possibilities of agreement – con-cluded outside the WTO counts more credibly (to a skepticalcivil society) as a new, stand-alone form of cooperation thanthe same agreement concluded inside the WTO. The answerto this question depends in large part on the trade effectsof a given domestic policy. There must be a market accessor trade cost dimension associated with a regulatory mea-sure that is applied on a nondiscriminatory basis (it is clearthat discriminatory trade policy measures are appropriatelythe focus of the WTO, and cooperation to address theeffects of such measures may take the form of critical massplurilateral agreements).In our view, the potential domain for WTO OPAs spans

any area of domestic regulation that directly affects tradecosts (market access), including a ban on access if domesticregulatory norms are not satisfied. It also covers instanceswhere regulation indirectly raises trade costs because firmsincur market-specific compliance costs that are in partredundant because national regulation in different countrieshas similar goals and regimes are equivalent.As discussed previously, international regulatory coopera-

tion that is linked to market access or reduces trade costsoccurs outside the WTO on a bilateral, plurilateral or multi-lateral basis. We do not argue that this is inappropriate orthat such instances of cooperation need to be brought intothe WTO. What we do argue is that the WTO can provide aplatform to support multilateralization of bilateral or regio-nal/plurilateral cooperation that determines access to mar-kets or reduces trade costs. Insofar as extant regulatorycooperation agreements do not have mechanisms to sup-port accession by third countries; do not have a mandate orstructure to foster openness and transparency to outsiders;or do not provide for technical assistance for non-partiesinterested in joining an arrangement there is a potential rolefor the WTO to help multilateralize cooperation.Open plurilateral agreements can help WTO members

articulate the norms and procedures to address the increas-ingly important question of regulatory diversity in ways thatrespect their sovereignty without relying on the paralyzingconvention of consensus. For OPAs to flourish they must betruly open – accessible to countries who may lack theresources and expertise to meet regulatory requirementsestablished by OPA members or to defend their interests indisagreements over what regulatory arrangements count asequivalent. The WTO can provide valuable assistance in the

Global Policy (2019) © 2019 University of Durham and John Wiley & Sons, Ltd.

Regulatory Cooperation and the WTO 11

form of technical support and facilitation of dispute resolu-tion. In doing so it can build on the substantial foundationsrecently created by the Trade Facilitation Agreement (TFA).The TFA, concluded in 2013, is the first multilateral agree-ment negotiated under WTO auspices. It comprises a set ofagreed good practices in the area of trade facilitation. Acentral feature of the TFA is to help individual membersadopt the good practices. It incorporates innovative mea-sures for providing technical support that could be adaptedto the related task of helping members join OPAs. Underthe TFA developing and least-developed countries experi-encing difficulties meeting implementation deadlines notifythe Committee on Trade Facilitation – the secretariatbody established to administer the agreement – and pre-pare a self-evaluation of the causes and remedies of theproblems. The Committee then appoints an expert groupwith five highly qualified members; the Group reviews theself-assessment of incapacity and accepts or amends itsfindings; and the Committee upon final review approves aplan of technical support. Just such a procedure could beused to determine whether and in what form the WTOshould provide technical support to members struggling tojoin particular OPAs. Analytically the TFA has an importantaffinity with the OPA: membership in both is open to allWTO members, but on condition that candidates demon-strate the equivalence of their regulatory measures and pro-cedures in a particular domain with those of the currentregime.

Because of this affinity, the expert groups created underthe TFA could serve as a template or building block for cre-ating a dispute resolution mechanism to address differencesbetween OPA members as a body and a candidate memberas to whether the latter’s standards and institutions in therelevant domain meet the equivalence test or on the pathto meeting it. Such judgements necessarily fuse normativeand technical criteria. Review of decisions by an expertgroup would help distinguish the most nearly technicalaspects of the difficulties and suggest concrete measures toaddress them. At a minimum such review, and theresponses it prompts, would help clarify the reasons for per-sistent objections and implicitly – in time, perhaps explicitly– set standards for the depth and extent of reason givingexpected as justification of such decisions. This type ofreview, together with the joint, ‘internal’ review of requestsfor membership that would presumably be undertaken as amatter of course by the existing parties to an OPA wouldensure that the agreements are as open as they claim to be.Cumulatively the decisions of such expert bodies would cre-ate a kind of meta-standard establishing tests to determinewhether the breadth and depth of the review of a candi-date’s formal procedures and institutional practices is suffi-ciently searching and attentive to the possibility thatdivergent means can achieve shared ends to count as legiti-mate.

Expert groups on the model of what is embodied in theTFA could, in addition to dispute resolution, play an impor-tant role in ensuring that OPAs fulfill their promise of open-ing the WTO to emergent, potentially consensual regulatory

norms and novel procedures. Expert groups reviewing dis-putes in different OPAs covering the same general domain– clubs with different standards and members – would bein an excellent position to identity possible commonalitiesin their approaches or suggest measures to bridge differ-ences. At the least, the circulation of these expert groupswould encourage the more general exchange of ideasamong overlapping OPAs. At best, it would help crystallizeconsensus when views and practices do begin to converge.The regular operation of the WTO would thus enable groupsof parties to trade on terms that reflect their distinctive val-ues while achieving consensus when possible through thisexploration of difference.

Conclusion

Regulation is now a prominent feature of trade policy andincreasingly is being considered in the design of tradeagreements (Mumford, 2018). Prominent recent examplesare the CPTPP, the USMCA, and EU trade agreements withCanada and Japan. To a significant extent, these PTAs codifygood regulatory practices and procedures that haveemerged from OECD and APEC deliberations. They are per-missive of deeper forms of regulatory cooperation of thetype discussed in this article but do little to encourage it.Differences in regulation across jurisdictions for tangible andintangible products may impede trade by generating redun-dant transactions costs or segmenting markets (OECD,2017). Access to markets may be linked to (conditional on)achievement of regulatory standards, either international –e.g., CITES – or specific national norms and values that pre-vail in an importing market – e.g., EU norms for imports oftimber; cross-border data flows; and investor-State arbitra-tion. OPAs offer a potential mechanism to support deepercooperation in a specific area without undercutting nationalsovereignty and the ability to regulate, and gradual multilat-eralization of such cooperation.Until recently, WTO members have been unable to

engage in deliberations to extend WTO disciplines to coverregulatory policy areas that are not or incompletely coveredby existing agreements. One reason for this are WTO work-ing practices – notably consensus and special and differen-tial treatment for developing countries (BertelsmannStiftung, 2018) – that have impeded the pursuit of coopera-tion among groups of WTO members to address issues ofcommon concern. This constraint was relaxed at the 2017Ministerial Conference in Buenos Aires, which saw thelaunch of plurilateral discussions on both old and new sub-jects, potentially opening the door to OPAs. Four ‘joint initia-tives’ were launched in Buenos Aires, one of which (on e-commerce and cross-border data flows) was transformedinto a plurilateral negotiation group in early 2019. Insofar asthese initiatives result in OPAs among groups of WTO mem-bers they offer the prospect of reducing regulatory compli-ance costs for firms operating internationally whileenhancing the ability of regulatory agencies to attain soci-etal objectives more efficiently. Providing a forum wherecountries can consider cooperation in a specific area of

© 2019 University of Durham and John Wiley & Sons, Ltd. Global Policy (2019)

Bernard Hoekman and Charles Sabel12

regulation is valuable. Having a (WTO) standard form fornew, domain-specific regulatory cooperation agreements(OPAs) can substantially cut transaction costs of cooperationfor countries while offering a credible prospect of gradualmultilateralization. This is not a strong feature of the formsof regulatory cooperation observed in PTAs (which are gen-erally closed) or those pursued, often bilaterally, on a sec-toral basis, some of which we have discussed in this paper.

Acknowledging OPAs as WTO-conforming could assistcountries to reduce the international trade costs of regula-tory heterogeneity – for example in areas such as data pri-vacy and protection, rules of origin, competition policy andinvestor-State arbitration – and help regenerate the WTO asa forum for mutually beneficial cooperation on trade-relatedpolicies. Put starkly, the WTO has failed as a forum forpreparing the (re-)negotiation of comprehensive trade deals.Its operation reflects the consensus of its members, whiledoing too little to shape it, or more modestly, even to aidmembers in the reconsideration of their positions. This couldchange if the WTO not only recognizes the legitimacy ofOPAs – something that is currently the subject of livelydebate among WTO members – but becomes a clearinghouse for comparing and diffusing their results and gradualexpansion of membership. In principle, the WTO can providea platform to support gradual multilateralization of OPAsthrough provision of technical support to candidate mem-bers and resolving disputes between OPAs and countiesseeking to accede to them. Through the formation of OPAswith different geographic scope and substantive reach,groups of WTO member countries could both test and re-elaborate alternative regulatory standards and designs forinstitutional cooperation. The growth or decline of OPAswould demonstrate the attractiveness to newcomers ofthose that survive this winnowing. In this way OPAs, operat-ing under auspices of the WTO would be dynamic and flexi-ble vehicles by which member states come to reconsidertheir particular regulatory commitments and institutionalhabits in light of the experience of like-minded others. Themost successful approaches could serve as the startingpoints for generalization and codification in internationalstandards or international framework agreements of variouskinds.

A WTO hospitable to OPAs would no longer be hostageto the consensus of its members, but be a partner in articu-lating it. A WTO that admitted of OPAs would create fora inwhich members could address commercially urgent, politi-cally sensitive problems of regulatory coordination withoutembedding them in comprehensive deals spanning manypolicy areas, concluded in secrecy, and typically shieldedfrom democratic accountability. Continuing reviews ofagreements struck in such fora would bring to light and rec-oncile varying interpretation of norms before they result inconflicting enforcement strategies. By embracing OPAs theWTO would reduce the dangers of fragmentation fromwithin and the risks of creating, at the margins of the globaltrade order, ad hoc, patchwork – and therefore typicallyfragile – regimes to deal with important collective actionproblems.

Clearly not all types of regulatory cooperation are appro-priate for consideration in OPAs under WTO auspices. Forexample, OPAs are less suitable to address the negative spil-lover effects of explicitly discriminatory policies that give riseto terms of trade effects, as free rider concerns will arisethat may preclude small group cooperation. Many regulatorypolicy areas that give rise to trade costs do lend themselvesto OPAs, however, as free riding concerns either do not ariseor can be addressed. The scope for OPAs may be largerthan is perceived by economic policymakers.In our view formal recognition of OPAs as a new kind of

arrangement, entailing no obligation to agree in the end tocooperate and the severability of eventual agreements offersan avenue for regenerating the WTO by assisting countriesto reduce the trade costs of regulatory differences whileimproving regulatory outcomes. Not pursuing this routewould be to miss an opportunity to revitalize the WTO. Per-haps more important is the scope that OPAs offer to partici-pants and stakeholders to multilateralize sectoral regulatorycooperation initiatives of the type discussed above that areproliferating outside the ambit of trade agreements, and tosome extent are reflected in recent PTAs. Such arrange-ments tend to be closed or lack explicit mechanisms to con-sider, let alone encourage, participation by additionalcountries. OPAs can help support the gradual multilateraliza-tion of international regulatory cooperation initiatives thataddress matters that are of potential interest to many coun-tries, by helping to enhance transparency, encourage mutualreview, support learning, and provide a platform tostrengthen regulatory capacity in countries where needed tobe able to participate in a given OPA.

NotesWe are grateful to the referees, the editor, Robert Basedow, ThomasBollyky, Marco Bronckers, Paola Conconi, Philipp Genschel, SeanHeather, David Levi-Faur, Nuno Lim~ao, Patrick Low, Aaditya Mattoo, Pet-ros Mavroidis, Peter Mumford, Julie Nind, Dani Rodrik, Richard Stewart,Robert Wolfe, Jonathan Zeitlin and participants in the WTO Trade Dia-logue, Updating Trade Cooperation: An Economic View, Geneva, 11December 2018 for helpful comments and suggestions on earlier drafts.The project leading to this article received funding from the EuropeanUnion’s Horizon 2020 research and innovation program under grantagreement No 770680 (RESPECT).

1. Open plurilateralism as an organizing concept is an element of theGovernment of New Zealand’s trade policy strategy. See Vitalis(2018).

2. See OECD (2017) for a review of the significance of such costs. Ifcompliance with regulatory standards pertaining to the process ofproduction is a necessary condition to imports to be permitted –e.g., a ban on slave labor or a requirement that timber be harvestedin a sustainable manner – non-compliance may imply that tradecosts are prohibitive in practice: trade is not permitted.

3. Issue linkage is a key feature of trade agreements and efforts toextend their coverage to nontrade issues such as domestic regula-tion. See e.g., Conconi and Perroni (2002); Lim~ao (2005, 2016),Maggi (2016).

4. Just how closed PTAs can be is demonstrated by the United King-dom’s difficulties withdrawing from the EU: all the commercially fea-sible alternatives to membership entail continuing, deep

Global Policy (2019) © 2019 University of Durham and John Wiley & Sons, Ltd.

Regulatory Cooperation and the WTO 13

engagement with EU regulators – at odds with the reassertion ofnational sovereignty that motivated Brexit – and reestablishingdeep regulatory cooperation after even a brief interruption wouldbe prohibitively costly

5. See Rollo and Winters (2000) and the contributions to Bhagwati andHudec (1996) for discussions of trade and non-trade policy linkagesand their desirability.

6. Traditional ISDS was replaced by Investment Court System (ICS) thatconditions resolutions of investor-State disputes on the right of gov-ernments to regulate in pursuit of policy objectives such as publichealth, safety, environment, public morals and the promotion andprotection of cultural diversity. The ICS replaces temporary adhoc tribunals with a permanent court made up of a Tribunal of FirstInstance (operating under auspices of a relevant EU trade agree-ment, such as CETA) and an Appeal Tribunal staffed by professionaland independent adjudicators who are appointed for long terms ofoffice by both parties and are subject to a strict code of conductand required to open hearings to the public; publish documentssubmitted during cases and permit interested parties to intervenein proceedings and make submissions. See http://ec.europa.eu/trade/policy/in-focus/ceta/ceta-explained/index_en.htm#read-more. The CJEU found that the ICS cannot call into question the level ofprotection of a public interest that led to the introduction of regula-tions affecting an investor in the EU. See https://curia.europa.eu/jcms/upload/docs/application/pdf/2019-04/cp190052en.pdf.

7. See Sabel et al. (2018) and Hoekman and Sabel (2019) for additionaldiscussion and references to the literature.

8. This section follows the accounts in Overdevest and Zeitlin (2015,2018) and Zeitlin and Overdevest (2019).

9. Japan has imposed a system of green public procurement with gov-ernment registration of compliance with a voluntary code of con-duct (Zeitlin and Overdevest, 2019).

10. The following draws is based on the FDA’s presentation of ICAT, athttp://www.fda.gov/downloads/Food/InternationalInteragencyCoordination/UCM331177.pdf.

11. FDA, Report of the Systems Recognition Assessment of Canada, ND,at http://www.fda.gov/downloads/Food/InternationalInteragencyCoordination/UCM503960.pdf

12. EU-US BASA, International Agreements Council Decision of 7 March2011 concerning the conclusion of the Agreement between theUnited States of America and the European Community on cooper-ation in the regulation of civil aviation safety, at http://eur-lex.europa.eu/resource.html?uri=cellar:64d9e1a2-633c-4e91-bbf5-053e5ab1b432.0010.02/DOC_1&format=PDF

13. See EASA/FAA, Technical Implementation Procedures for Airworthi-ness and Environmental Certification between the Federal AviationAdministration of the United States of America and the EuropeanAviation Safety Agency of the European Union, at: https://www.faa.gov/aircraft/air_cert/international/bilateral_agreements/baa_basa_listing/media/EUTIP.pdf

14. A fatal crash on 10 March 2019, the second in 5 months, led theEASA on 12 March 2019 to suspend all flight operations and com-mercial flights of Boeing Model 737-8 MAX and 737-9 MAX air-planes. At the time the EASA – and other regulators – groundedthe 737 MAX, the FAA held that the 737 MAX was airworthy, illus-trating the severability of the regime. In the event the FAA followedsuit a few days later based on initial findings of the accident investi-gation.

15. See generally on EU banking reform: Ferran, 2014; Zeitlin, 2016).16. For in-depth discussion of this area, see Ferran (2014, 2018) and

Moloney (2018).17. See Bertelsmann Stiftung (2018) for an in-depth discussion.18. Some WTO provisions specify consensus as the decision-making

rule, for example, Art. X:9 on amendments to include new Annex 4plurilateral agreements. Art. IX WTO specifies that if voting occurs,unanimity is required for amendments relating to general principles

such as non-discrimination; a three-quarters majority for Interpreta-tions of provisions of the WTO agreements and decisions on waiv-ers; and a two-thirds majority for amendments relating to issuesother than general principles. Where not otherwise specified andconsensus cannot be reached a simple majority vote is sufficient.Art. X provides that a member cannot be bound by a vote on anamendment that alters its rights or obligations and that it opposes.In such instances, the Ministerial Conference may decide to requestthat the member concerned withdraw from the WTO or to grant ita waiver.

19. The term plurilateral is sometimes used to describe all three ofthese possibilities, giving rise to potential confusion. For in-depthdiscussion of WTO-consistent cooperation among subsets of WTOmembers see Lawrence (2006) and Hoekman and Mavroidis (2015,2017).

20. See Gnutzmann-Mkrtchyan and Henn, 2018 for estimates of the glo-bal trade impact of the ITA.

21. Eighty-two WTO members have signed the Reference Paper. Seehttps://www.wto.org/english/tratop_e/serv_e/telecom_e/telecom_e.htm.

22. Negotiations commenced in July 2014. See https://www.wto.org/english/tratop_e/envir_e/ega_e.htm.

23. See, for example, Wijkstr€om and McDaniels (2013) and Mavroidis(2016) for discussion of the SPS and TBT agreements.

24. Participation in these groups spans a broad cross-section of themembership. The EU participates in all four groups. The US is partof one (e-commerce). China was a sponsor of three of the fourgroups – it initially decided not to join the group on e-commerce.However, it decided to join the group when negotiations werelaunched in early 2019. India as well as many African countriesdecided not to be part of any of the groups. Independent ofwhether a WTO member is a sponsor/supporter of a group, deliber-ations of the groups are open to all.

25. See https://www.ft.com/content/3a8b7458-1fe5-11e9-b2f7-97e4dbd3580d.

26. See Hoekman (2016) for an in-depth discussion and assessment ofthe TFA.

27. The TFA differs from the OPA in that membership in the former isultimately obligatory, even if on a schedule that respects the candi-dates’ capacities and idiosyncrasies; and subgroups of membersmay not form an alternative trade facilitation regime.

28. Existing examples include ISEAL and the Global Food Safety Initia-tive (GFSI). ISEAL is a meta-standard setter for organizations thatdefine sustainability standards and establish procedures for certify-ing compliance with them; the GFSI plays a similar role for privateorganizations developing certification programs for food safetymanagement systems. See ISEAL (2014) and GFSI (nd, version 7.2).OPAs can emulate the role that these types of institutions play intheir respective domains in policy areas where such cooperation isnot as inclusive or open, that is, not designed with a view to fostergradual multilateralization of international cooperation.

29. Indeed, even a conflictual policy area such as subsidies arguablylends itself at least to some extent to an OPA. For example, thetype of review and reasoning that is part of any OPA is needed inthe subsidies context given that their potential salience in address-ing market failures and achieving social objectives makes it muchharder to write down hard rules. The design of any rules and theirimplementation must be informed by an understanding of theunderlying goals (as elucidated by the government using them),and neutral assessments of their effects.

References

Asser Institute. (2017) ‘EU Timber Regulation: From an Operator'sPerspective’ [online]’[online]. Available from: https://www.asser.nl/DoingBusinessRight/Blog/post/towards-a-due-diligence-jurisprudence-

© 2019 University of Durham and John Wiley & Sons, Ltd. Global Policy (2019)

Bernard Hoekman and Charles Sabel14

the-eu-timber-regulation-s-requirements-in-courts-by-wybe-th-douma[Accessed 31 May 2019].

Bertelsmann Stiftung. (2018) Revitalizing Multilateral Governance at theWorld Trade Organization[online]. Available from: https://www.bertelsmann-stiftung.de/en/publications/publication/did/revitalizing-multilateral-governance-at-the-world-trade-organization/ [Accessed 31 May2019].

Bhagwati, J. and Hudec, R. (eds.) (1996) Harmonization and Fair Trade:Prerequisite for Free Trade?. Cambridge, MA: MIT Press.

Bollyky, T. (2017) ‘The Role of Regulatory Cooperation in the Future of theWTO’, RTA Exchange. Geneva: International Centre for Trade andSustainable Development and Inter-American Development Bank.

Chase, P. and Pelkmans, J. (2015) This Time It's Different: Turbo-chargingRegulatory Cooperation in TTIP. Special Report No. 110. Brussels:CEPS.

Chayes, A. and Chayes, A. (1993) ‘On Compliance’, InternationalOrganization, 47(2), pp. 175–205.

Chayes, A. and Chayes, A. (1998) The New Sovereignty: Compliance withInternational Regulatory Agreements. Cambridge, MA: HarvardUniversity Press.

Conconi, P. and Perroni, C. (2002) ‘Issue Linkage and Issue Tie-in inMultilateral Negotiations’, Journal of International Economics, 57(2),pp. 423–447.

D€ur, A., Baccini, L. and Elsig, M. (2014) ‘The Design of InternationalTrade Agreements: Introducing a New Dataset’, Review ofInternational Organizations, 9(3), pp. 353–375.

Eisner, N. and Parker, R. W. (2016) ‘Learning from Experience: ThreeCase Studies in International Regulatory Cooperation’, ReportPrepared for the European Commission[online]. Available from:https://doi.org/10.2139/ssrn.2780605 [Accessed 31 May 2019].

European Aviation Safety Agency and the Federal AviationAdministration. (2011) ‘Aviation Safety Agreement between theUnited States and the European Community’, Available from: https://www.faa.gov/aircraft/repair/media/EASA_EU_roadshows.pdf[Accessed 31 May 2019].

EU-US BASA. (2011) ‘International Agreements Council Decision of 7March 2011 concerning the conclusion of the Agreement betweenthe United States of America and the European Community oncooperation in the regulation of civil aviation safety‘, Official Journalof the European Union 291, Available from: http://eur-lex.europa.eu/resource.html?uri=cellar:64d9e1a2-633c-4e91-bbf5-053e5ab1b432.0010.02/DOC_1&format=PDF. [Accessed 9 November 2011].

Ferran, E. (2014) ‘Financial Supervision’, in D. M€ugge (ed.), Europe andthe Governance of Global Finance. Oxford: Oxford University Press,pp. 16–34.

Ferran, E. (2018) ‘Regulatory Parity in Post-Brexit UK–EU FinancialRegulation: EU Norms, International Financial Standards or a HybridModel?’, in K. Alexander, C. Barnard, E. Ferran, A. Lang and N.Moloney (eds.), Brexit and Financial Services: Law and Policy, Oxford:Hart Publishing, pp. 1–16.

Global Food Safety Initiative (GFSI). ‘Requirements for the Managementof Certification Programs‘, version 7.2[online]. Available from: https://www.mygfsi.com/files/Technical_Documents/BR_V7.2/Benchmarking_Requirements_v7-2_Part2.pdf [Accessed 31 May 2019].

Gnutzmann-Mkrtchyan, A. and Henn, C. (2018) ‘Peeling Away the Layers:Impacts of Durable Tariff Elimination’, Journal of InternationalEconomics, 115, pp. 259–276.

Hoekman, B. (2016) ‘The Bali Trade Facilitation Agreement andRulemaking in the WTO: milestone, mistake or mirage?’, in J. Bhagwati,P. Krishna and A. Panagariya (eds.), The World Trade System: Trends andChallenges. Cambridge, MA: MIT Press, pp. 149–192.

Hoekman, B. and Kostecki, M. (2009) The Political Economy of the WorldTrading System. Oxford: Oxford University Press.

Hoekman, B. and Mavroidis, P. (2015) ‘Embracing Diversity: PlurilateralAgreements and the Trading System’, World Trade Review, 14(1), pp.101–116.

Hoekman, B. and Mavroidis, P. (2017) ‘MFN Clubs and SchedulingAdditional Commitments in the GATT: Learning from the GATS’,European Journal of International Law, 28(2), pp. 387–407.

Hoekman, B. and Sabel, C. (2019) ‘In a World of Value Chains: WhatSpace for Regulatory Coherence and Cooperation in TradeAgreements?’, in B. Kingsbury, R. Stewart, P. Mertensk€otter and T.Streinz (eds.), Megaregulation Contested: Global Economic Orderingafter TPP. Oxford: Oxford University Press.

Hofmann, C., Osnago, A. and Ruta, M. (2018) ‘The Content ofPreferential Trade Agreements’, World Trade Review. https://doi.org/10.1017/s1474745618000071

Horn, H., Mavroidis, P. and Sapir, A. (2010) ‘Beyond the WTO? AnAnatomy of EU Free Trade Agreements’, The World Economy, 33(11),pp. 1565–1588.

IKEA. (2018) EU Timber Regulation: From an Operator's Perspective.Presentation to the Commission Expert Group on EU TimberRegulation and the Forest Law Enforcement, Governance and Trade(FLEGT) Regulation (main group)[online]. Available from: http://ec.europa.eu/transparency/regexpert/index.cfm?do=groupDetail.groupMeetingDoc&docxml:id=25128 [Accessed 31 May 2019].

International Civil Aviation Organization (ICAO). (2017) ‘Outlook for AirTransport to the Year 2025’ (ICAO Circular Cir. 313 AT/134). Availablefrom: https://www.icao.int/sustainability/Documents/C313_Outlook_En.pdf [Accessed 17 June 2019]

ISEAL. (2014) Setting Social and Environmental Standards ISEAL Code ofGood Practice Version 6.0[online]. Available from: https://www.isealalliance.org/sites/default/files/resource/2017-11/ISEAL_Standard_Setting_Code_v6_Dec_2014.pdf [Accessed 31 May 2019].

Jennison, M. (2013) ‘The Future of Aviation Safety Regulation: New US-EU Agreement Harmonizes and Consolidates the TransatlanticRegime, but What is the Potential for Genuine Regulatory Reform?’,Air & Space Law, 38(4–5), pp. 333–350.

Lawrence, R. (2006) ‘Rulemaking amidst Growing Diversity: A ‘Club ofClubs’ Approach to WTO Reform and New Issue Selection’, Journal ofInternational Economic Law, 9(4), pp. 823–835.

Lim~ao, N. (2005) ‘Trade Policy, Cross-border Externalities and Lobbies:Do Linked Agreements Enforce More Cooperative Outcomes?’,Journal of International Economics, 67(1), pp. 175–199.

Lim~ao, N. (2016) ‘Preferential Trade Agreements’, in K. Bagwell and R. W.Staiger (eds.), The Handbook of Commercial Policy, vol 1B.Amsterdam: North-Holland, pp. 279–367.

Maggi, G. (2016) ‘Issue Linkage’, in K. Bagwell and R. Staiger (eds.), TheHandbook of Commercial Policy, vol 1B. Amsterdam: North Holland,pp. 513–564.

Maskus, K. (2002) ‘Regulatory Standards in the WTO: Comparing IntellectualProperty Rights with Competition Policy, Environmental Protection, andCore Labor Standards’,World Trade Review, 1(2), pp. 135–152.

Mavroidis, P. C. (2016) The Regulation of International Trade: The WTOAgreements on Trade in Goods, vol 2. Cambridge, MA: MIT Press.

Moloney, N. (2018) The Age of ESMA. Governing EU Financial Markets.Oxford: Hart Publishing.

Mumford, P. (2014) ‘Regulatory Coherence: Blending Trade andRegulatory Policy’, Policy Quarterly, 10(4), pp. 3–9.

Mumford, P. (2018) ‘Regulatory Cooperation’, in J. Drake Brockman andP. Messerlin (eds.), Potential Benefits of an Australia-EU Free TradeAgreement: Key Issues and Options. Adelaide: University of AdelaidePress, pp. 149–167.

Nathan, I., Chen, J., Hansen, C. P., Xu, B. and Li, Y. (2018) ‘Facing theComplexities of the Global Timber Trade Regime: How do ChineseWood Enterprises Respond to International Legality VerificationRequirements, and What are the Implications for RegimeEffectiveness?’, Forest Policy and Economics, 92, pp. 169–180.

Nouy, D. (2015) Presentation of the First ECB Annual Report onSupervisory Activities at the European Parliament's Economic andMonetary Affairs Committee. European Central Bank, Brussels[online].Available from: https://www.bankingsupervision.europa.eu/press/

Global Policy (2019) © 2019 University of Durham and John Wiley & Sons, Ltd.

Regulatory Cooperation and the WTO 15

speeches/date/2018/html/ssm.sp180326.en.html [Accessed 31 May2019].

OECD (2017) International Regulatory Cooperation and Trade:Understanding the Trade Costs of Regulatory Divergence and theRemedies. Paris: OECD.

Overdevest, C. and Zeitlin, J. (2015) ‘Forest Law EnforcementGovernance and Trade (FLEGT): Transnational dynamics of an EUexperimentalist regime’, in J. Zeitlin (ed.), Extending ExperimentalistGovernance? The European Union and Transnational Regulation.Oxford: Oxford University Press, pp. 137–168.

Overdevest, C. and Zeitlin, J. (2018) ‘Experimentalism in TransnationalForest Governance: Implementing EU Forest Law EnforcementGovernance and Trade (FLEGT) Voluntary Partnership Agreements inIndonesia and Ghana’, Regulation & Governance, 12(1), pp. 64–87.

Pacheco, P. G., Schoneveld, G., Dermawan, A., Komarudin, H. and Djama, M.(2018) ‘Governing sustainable palm oil supply: Disconnects,complementarities, and antagonisms between state regulations and privatestandards’, Regulation &Governance. https://doi.org/10.1111/rego.12220

Palanco, L. and Sauv�e, P. (2018) ‘The Treatment of RegulatoryConvergence in Preferential Trade Agreements’, World Trade Review,17(4), pp. 575–607.

Ratajczyk, M. (2011) ‘Regulatory Framework for a ‘Performance-Based’Approach to Air Safety Management in the European Union’, Air andSpace Law, 36(6), pp. 401–426.

Rollo, J. and Winters, L. A. (2000) ‘Subsidiarity and GovernanceChallenges for the WTO: The Examples of Environmental and LabourStandards’, The World Economy, 23(4), pp. 561–576.

Sabel, C., Herrigel, G. and Kristensen, P. (2018) ‘Regulation underUncertainty: The Coevolution of Industry and Regulation’, Regulation& Governance, 12(3), pp. 371–394.

Vitalis, V. (2018) ‘National Interest Analysis for the Comprehensive andProgressive Agreement for Trans-Pacific Partnership’, Statement tothe Foreign Affairs, Defence and Trade Committee, May 3[online].Available from: https://www.parliament.nz/resource/en-NZ/52SCFD_EVI_76583_999/0886a670ec6aa58cfc5e47172141ad4ab3e66cbc [Accessed3 May 2019]

Wijkstr€om, E. and McDaniels, D. (2013) ‘Improving RegulatoryGovernance: International Standards and the WTO TBT Agreement’,Journal of World Trade, 47(5), pp. 1013–1046.

Zeitlin, J. (2016) ‘EU Experimentalist Governance in Times of Crisis‘, WestEuropean Politics 39(5): pp. 1073–1094.

Zeitlin, J. and Overdevest, C. (2019) ‘Experimentalist Interactions: The EUForest Law Enforcement, Governance and Trade (FLEGT) Initiativeand the Emerging Transnational Timber Legality Regime’, draft paperpresented at the International Biennial Conference of the EuropeanUnion Studies Association, Denver, Colorado, 9–11 May.

Author InformationBernard Hoekman is Professor and Director, Global Economics at theRobert Schuman Centre for Advanced Studies, European University Insti-tute in Florence, Italy where he also serves as the Dean for ExternalRelations, and a CEPR Research Fellow.

Charles Sabel is Professor of Law and Social Science at Columbia LawSchool. Previously he was Ford International Professor of Social Scienceat MIT. Current projects include research on experimentalist solutions toglobal problems, the transformation of US administrative law in the faceof uncertainty, and new models of economic development.

© 2019 University of Durham and John Wiley & Sons, Ltd. Global Policy (2019)

Bernard Hoekman and Charles Sabel16