WPS8579 Policy Research Working Paper 8579 Services Globalization in an Age of Insecurity Rethinking Trade Cooperation Aaditya Mattoo Development Economics Development Research Group September 2018 Policy Research Working Paper 8579 Abstract Decades of services trade negotiations have produced a liberalizing commitments. Third, insulating domestic con- plethora of rules and commitments but limited real lib- sumers from international market failure is a precondition eralization. One reason is a form of “negotiating tunnel for further liberalization in many services sectors, and the vision,” which has led to a focus on reciprocal market open- relevant international bargain needs to be an exchange of ing rather than on creating the regulatory preconditions regulatory commitments by exporters in return for market for liberalization. This paper makes four points. First, cur- access commitments by importers. Fourth, such bargains rent trade disciplines are a useful but inadequate restraint create a risk of exclusion for nonparticipants that can and on regulatory protection. Second, proposed disciplines should be addressed. The paper illustrates these arguments on domestic regulation would add value but would not drawing upon recent developments relating to data privacy, solve problems with the application of existing trade law financial services, labor mobility, and competition policy. and could create a hold-back problem in securing new This paper is a product of the Development Research Group, Development Economics. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/research. The author may be contacted at amattoo@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Services Globalization in an Age of Insecurity: Rethinking Trade Cooperation Aaditya Mattoo* Keywords: services trade, regulatory cooperation, GATS, trade agreements JEL codes: F13, F15 *World Bank, 1818 H Street NW, Washington DC 20433. Email: amattoo@worldbank.org. I would like to thank Steve Charnovitz, Andrew Crosby, Bernard Hoekman, Patrick Low, Hamid Mamdouh, Juan Marchetti, Ricardo Melendez, Pierre Sauve, Harsha Singh, Sherry Stephenson and the participants in the E15 Expert Group on Services Trade for helpful discussions and comments on an earlier version of the paper. Work on this paper is supported in part by the Multi-Donor and Umbrella Trust Funds for Trade and Development, and by the Strategic Research Program. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author and do not necessarily represent the views of the World Bank. Services Globalization in an Age of Insecurity: Rethinking Trade Cooperation Decades of services trade negotiations have produced a plethora of rules and commitments but little real liberalization. One reason is the risk of market failure in globalized services markets, manifested in data privacy violations, financial instability, monopolistic practices and illegal immigration. National regulators’ inability to address these international market failures leads to trade protection as a form of regulatory precaution. Services trade negotiations have yielded a poor harvest in part because they focused solely on reciprocal market-opening rather than on creating the regulatory preconditions for liberalization (Hoekman, 1996). Greater international cooperation involving regulators may advance liberalization and produce more meaningful agreements. Services trade negotiators have not ignored domestic regulation, but seen it primarily through the lens of securing access to markets (Lim and Meester, 2014). Thus, the goal has been to ensure that the presence of prudential regulation or the absence of pro-competitive regulation in importing countries does not become a trade barrier. Where market failure due to informational problems – e.g. in areas like financial and professional services – prompts national regulators to impose licensing, qualification and other requirements, trade rule-making has sought to ensure that these requirements do not unduly burden foreign providers. Where market failure due to monopolies – e.g. in network-based services like communications and transport – allows incumbent firms to frustrate entry and competition, international rules have required national regulation to ensure fair access to essential facilities. There are two problems with this market access-centered approach. The first is that existing international trade rules and commitments are hard to enforce and have uncertain value. It has always been difficult to strike a balance between allowing space for the legitimate use of domestic regulation while preventing its protectionist abuse (Krajewski, 2014). As discussed below, an importer-only approach to trade disciplines risks allowing either less regulatory discretion than is politically unacceptable, or more regulatory discretion than is consistent with predictable market access. The second problem with this approach is that it does not facilitate new market opening and international commitments by helping national regulators deal with international market failure. A country will be reluctant to open its financial markets unless it is confident that it can protect its consumers and prevent financial instability, or open its transport and internet-based services markets if it is afraid that the gains from liberalization will be appropriated by international oligopolies. In some cases, such as the supply of services through locally incorporated subsidiaries, the importing country can in principle deal unilaterally with market failure because the provider is in its jurisdiction. But doing so requires adequate regulatory capacity and could lead to higher costs of trade by fragmenting markets (e.g. by requiring local capital adequacy or local servers). In other cases, such as cross-border banking, transport or data-processing 2  services, addressing market failure efficiently is not possible without the cooperation of the regulator in the exporting country. Regulatory externalities exist in goods trade as well, but there is an important reason why regulatory cooperation is not just an “add-on” in services but a precondition for further liberalization. In goods trade, a country could liberalize trade policy and still apply technical regulations at the border. The intangibility of services and the simultaneity of production and consumption makes post-production and pre-consumption inspection difficult. An inability to ensure ex ante compliance by producers located abroad with nationally desired regulations translates into a reluctance to liberalize. Traditional regulatory cooperation has focused on alleviating the costs of regulatory heterogeneity for producers, due to the diseconomies of producing to different standards for different markets. The solution to such market-fragmentation is the harmonization or mutual recognition of national regulations, which the European Union has pursued with significant success in goods and more limited success in services (Pelkmans, 2003). We highlight instead the impact of non-compliance by foreign producers with domestic standards to protect domestic consumers. Even identical national regulations do not address the problem remains that foreign regulators are legally mandated to protect the interests of their own consumers and not those of foreign consumers. European competition authorities punish cartels of European firms if they exploit European consumers but not if they exploit foreign consumers. Indian data laws protect the privacy of Indian citizens, not foreigners. An alternative form of consumer-centric international cooperation can help deal with this jurisdictional problem. When negative externalities are transmitted via exports of services, regulators in exporting countries would assume the obligation to protect the interests of foreign consumers in return for importing countries committing to allow access to their markets. These exporter commitments need not be in the context of trade agreements but could be secured in other existing or new fora. Essentially, a source country obligation to ensure conformity with destination country standards would be an integral part of a trade deal. Importing country regulators would thus be reassured that exporting country regulators would cooperate to protect their consumers’ privacy, financial security, and well-being in general from the consequences of international market failures. Greater market opening and commitments may be forthcoming if international negotiations did not just require regulators to tie their hands and allow market access unconditionally on an MFN basis, but also an opportunity to secure assistance from source countries to deal with problems they cannot solve on their own. This is not to deny that there are other impediments to progress, such as the power of vested interests. But addressing the legitimate reasons for holding back 3    from liberalization and making commitments may allow a more productive focus on negotiating away the barriers to trade. In the subsequent sections, this paper makes the following arguments: I. Services trade negotiations have not led to much liberalization. II. Current trade disciplines are a useful but inadequate restraint on regulatory protection. III. Proposed trade disciplines on domestic regulation add value but do not solve existing problems and can create a new hold back problem. IV. Third, insulating domestic consumers from international market failure is a precondition for further liberalization in many services sectors, and the relevant international bargain needs to be an exchange of regulatory commitments by exporters in return for market access commitments by importers. V. Such bargains are already being made or could plausibly be made in the areas of data privacy, financial services, labor mobility, and competition policy. VI. But such bargains create a risk of exclusion for non-participants that can and should be addressed. I. Services trade negotiations have not delivered much liberalization Services have been negotiated multilaterally as well as in regional, plurilateral and bilateral contexts. The first multilateral services negotiations took place in the context of the Uruguay Round which resulted in the conclusion of the General Agreement on Trade in Services (GATS) in 1995. A second round of multilateral negotiations on services was incorporated in the Doha agenda, made some progress but has not been concluded. In parallel, an increasing number of preferential trade agreements have included services, beginning with the North American Free Trade Area (NAFTA) in the early 1990s and most recently and ambitiously, in the Trans-Pacific Partnership (TPP), which eventually became the Comprehensive and Progressive Agreement for Trans-Pacific Partnership (CPATPP). Consider how GATS commitments, the Doha offers, and the services commitments in the TPP or CPTPP compare with applied policies. We rely on a measure of the restrictiveness of a country’s policy regime for any subsector-mode, the “services trade restrictions index” (STRI), which has the weakness of being subjective but the virtue of being simple, transparent and robust (Borchert et al. 2014).1 We focus on five broad sectors: telecommunications, transport, financial services, distribution and selected professional services. Borchert et al. (2011) compared the Uruguay Round commitments in services of 92 countries and the best offers that many of these countries made in the Doha negotiations with their applied                                                              1 It is notoriously difficult to measure policies affecting services trade because of their variety and complexity (see, for example, Hoekman (1996) and Deardorff and Stern (2008)). 4    trade policies. Figure 1 shows that the services restrictiveness index of the Uruguay Round commitments was slightly higher than that of the Doha offers, while the index of the Doha offers was much higher than that of applied policy. In fact, Uruguay Round commitments were on average 2.3 times more restrictive than applied policies and Doha promised greater security of access to markets but very little additional liberalization. Figure 1: Uruguay Round Commitments, Doha Offers and Actual Policy Source: Borchert, Gootiiz and Mattoo (2011), drawing on the World Bank services trade restrictions database, http://iresearch.worldbank.org/servicetrade.htm. Consider now the TPP (or the CPATPP). Figure 2 shows that the TPP commitments improved upon the Doha round offers, but for no country was the average openness implied by TPP commitments greater than that offered by currently applied policies in the sectors we cover (Gootiiz and Mattoo, 2017). Thus, the TPP enhanced transparency and policy certainty because parties’ services commitments cover more trading partners, more sectors and were in some cases closer to applied policies than their commitments under previous agreements. But since the TPP 5    commitments seldom go beyond countries’ applied policies, the explicit liberalization resulting from the agreement was limited to a few members and a few areas.2 Figure 2: Services Trade Restrictiveness Indices for TPP Countries Doha Offers, TPP Commitments and Applied Policy Overall STRI 100 Services trade restrictiveness index 75 50 25 0 BRN MYS CHL SGP MEX VNM CAN USA PER JPN AUS NZL Doha TPP Policy Note: For Brunei, applied policy information is missing. Absense of a bar means the score is zero Note 1: BRN =Brunei, MYS =Malaysia, CHL = Chile, SGP= Singapore, MEX= Mexico, CAN= Canada, VNM = Vietnam, USA =United States, PER = Peru, JPN= Japan, AUS = Australia; NLZ = New Zealand. Note 2: The country STRI is an aggregate score of indices for services sectors covered, the country STRI comprises STRIs of five major sectors, financial, telecom, retailing, transportation, and professional services. These five sectors contain different subsectors and the relevant modes. Source: Gootiiz and Mattoo (2017), drawing on the World Bank TPP database, http://iresearch.worldbank.org/servicetrade/TPPservices.htm.                                                              2 However, in comparing TPP commitments with the policy status quo, our measures may underestimate the liberalizing impact of the TPP in two ways. First, the TPP negotiating process could already have induced changes in applied policies over the course of the negotiations; for example, Mexico’s telecommunications policy was reformed in 2013, and Malaysia carried out several reforms between 2009 and 2013 in financial services, telecommunications, and professional services. Second, some of the TPP commitments specify liberalization several years after the completion of the agreement, for example by Vietnam in telecommunications and retail services and by Malaysia in the financial sector. 6    II. Current trade disciplines are a useful but inadequate restraint on regulatory protection In this section, we briefly review challenges in interpreting and implementing existing trade disciplines, and how difficult it is to strike a balance between allowing regulatory freedom and preserving the value of market access commitments. II.1 Challenges in interpreting key trade disciplines The three pillars of the GATS, and most other services trade agreements, are the provisions on Most-Favored-Nation (MFN) Treatment (GATS Article II), National Treatment (GATS Article XVII) and Market Access (GATS Article XVI). These valuable provisions seek, respectively, to prevent discrimination between trading partners, in favor of domestic providers, and the use of quantitative restrictions. Certain limitations of the key disciplines are well recognized. For example, it is well known that the MFN obligation under GATS allowed for one-time exemptions, and that the National Treatment and Market Access obligations in GATS-type agreements apply only in sectors included in member’s schedule and there too can be subject to limitations.3 It is not, however, as well appreciated that neither these rules, nor other supporting provisions, adequately address national regulatory discretion in implementing policies, especially through the issuing of licenses. MFN and National Treatment cover in principle both de jure and de facto discrimination, i.e. both explicitly different treatment and effectively different treatment.4 However, virtually all the exemptions listed under the MFN obligation and limitations scheduled under the National Treatment obligation pertain only to explicit discrimination. The fact that WTO Members have not sought legal cover for de facto discriminatory practices does not imply that such practices do not exist. Rather Members are reluctant to determine unilaterally whether and where they discriminate in effect, and to confess to the existence of such discrimination in their schedules. The burden of identifying and establishing the existence of discrimination, therefore, falls on other interested parties, i.e. as exporters and exporting countries. Determining whether a measure is consistent with the MFN or NT is a challenge. It hinges, first of all, on establishing whether the imported services are "like" each other. The wider the definition of "likeness," the greater will be the set of measures that are inconsistent with Articles                                                              3  Other NAFTA-type of agreements take a negative-list approach with regard to sectoral coverage, implying that national treatment applies to all sectors except those which are explicitly specified.  4 The elaboration in Article XVII:2 and XVII:3 of the National Treatment provision left no doubt that the term “treatment no less favorable” should be interpreted to mean not only de jure discrimination but also de facto discrimination. The fact that there was not similar elaboration in the MFN provision created some doubt about its scope. But the Appellate Body in the Bananas Dispute upheld the Panel's conclusion that the provision prohibits both forms of discrimination, even though it differed on the basis for this conclusion went further than the Panel in clearly stating that its conclusion was not limited to the case under consideration. 7    II and XVII. If a doctor is a doctor, a regulation that imposed greater qualification requirements on a doctor trained in Country A than on a doctor trained in Country B would violate Article II, or on a doctor trained abroad than on a domestic doctor would violate Article XVII. On the other hand, the narrower the definition of "likeness" the more measures will conform to Articles II and XVII. If a doctor trained in country A is not "like" a doctor trained in country B, then the treatment of the doctor trained in country A would have to be compared with the treatment of a doctor trained in a country with similar training standards, and the additional qualification requirement could then be found consistent with Article II. Crucially, the national treatment (Art XVII) and MFN (Art II) obligations would have limited force if likeness is in the eye of the regulator. The market access (Article XVII) obligation differs from the other two provisions in that it prohibits only de jure quotas but not de facto quotas. So, it is only to be expected that the schedules under Article XVII list only explicit quantitative restrictions. Crucially, a Member is free today to implicitly restrict the number of service providers, for example by varying qualification standards – as the Japanese body responsible for regulating the accountancy profession reportedly used to do in order to limit the annual flow of qualified chartered accountants – or by imposing prohibitive entry fees –as the Zambian telecommunications ministry used to do in order to prevent foreign providers from establishing their own international gateways. II.2 Challenges in implementing key trade disciplines Whether there is de facto discrimination or a de facto quota depends entirely on how policy is implemented. The key mechanism for allowing entry in services is the license.5 The license is defined broadly to cover any decisions or administrative actions, including authorizations or approvals, which allow a service provider to enter the market and provide a service. In this broad sense, virtually every service provider in virtually every country needs a license to operate – from an accountant and an architect to a bank and a builder. While licenses are almost always required, they are almost never issued in a non-discretionary manner. The limited evidence available from the World Bank Services Trade Restrictions Database suggests that licensing criteria are usually made public, reasons for denial are often provided, but the fulfillment of publicly stated criteria does not automatically lead to a license being issued (Figure 3). Figure 3: Transparency, accountability and predictability                                                              5 GATS Article XXVIII defines a “measure” to be “any measure by a Member, whether in the form of a law, regulation, rule, procedure, decision, administrative action, or any other form.” It is convenient to distinguish, as the GATS does, between “measures of general application” and other measures. Measures of general application include laws, regulations and rules (which inhabit the GATS schedules) as well as standard administrative procedures. Other measures include decisions and administrative actions, which are one-off acts implementing measures of general application.   8    Source: World Bank Services Trade Restrictions Database. Note: The database covers 102 countries. “Miss” denotes missing data. Regulators cross the world and across services sectors clearly enjoy significant latitude in making judgments of “likeness” in the decision of whether and how to issue a license, as well as in determining the number of licenses they issue. This regulatory discretion imposes a cost on trade. For example, see Table 1 for estimates of the “regulatory tax” paid by foreign professionals already licensed to practice in other jurisdictions in order to practice in the United States – one of the more open economies in the world. This tax reflects the costs for professionals of taking tests of competence, requalification, retraining and fulfilling other licensing requirements, such as working for a period of time in underserved areas. Some of this tax may be legitimate and necessary to ensure the desired quality of a service but the problem is that there is at present no way of establishing whether that is indeed the case or whether some elements of the tax are in fact discriminatory. Table 1: Foreign professionals pay a large regulatory tax in order to practice in the US Source: Mattoo and Mishra (2006) 9    GATS and other agreements are not blind to the fact that discretionary licensing can weaken other disciplines. 6 GATS Article VI on Domestic Regulation has already put in place certain disciplines. Thus, Article VI:1 requires that in sectors where a member has made specific commitments, all measures be administered in a “reasonable, objective and impartial manner.” Article VI:2(a) mandates the institution of judicial, arbitral or administrative tribunals for the objective and impartial review of and, where justified appropriate remedies for, administrative decisions.7 But recognizing the inadequacy of these restraints, Article VI:4 mandates negotiations to develop disciplines necessary to ensure that “qualification requirements and procedures, technical standards and licensing requirements do not constitute unnecessary barriers to trade in services.” But these negotiations have only produced an outcome for the pilot exercise on accountancy, and those too remain in legal limbo. 8 At the most recent WTO Ministerial Conference, 29 WTO Members (counting the European Union as one) issued a statement reaffirming their commitment to advancing negotiations on the basis of recent proposals.9 The proposed disciplines deal with streamlining the administration of measures; enhancing transparency of not just of the measures but also of the process through which they are developed including the opportunity to comment; ensuring measures are based on objective and transparent criteria and the procedures are impartial and not unduly burdensome; and flexibility and support for implementation by developing countries. A smaller group of members also propose the application of the “necessity test,” i.e. that measures are not more burdensome than necessary to ensure the quality of the services.10 In sum, given the currently applicable provisions, it is evident that the discretion retained by regulators creates scope for the implementation of de facto discrimination or de facto quantitative restrictions through the issuing of licenses. II.3 Security of market access vs regulatory flexibility Apart from the discretion inherent in the implementation of policy, WTO Members have taken two routes to retaining regulatory discretion in specific areas: either they have made no commitments in the relevant area; or they have made commitments but insisted on the creation                                                              6  The decision to issue a license is in some ways analogous to the decision to issue a government procurement contract. In both cases, the challenge for international disciplines is to limit discretion which is open to protectionist abuse. Therefore, certain provisions of the WTO’s Government Procurement Agreement, particularly those relating to ex ante and ex post transparency, as well as the challenge procedures are potentially relevant to services.  7 There are also procedural disciplines regarding informing applicants about the status and decisions (Article VI:5), and establishing procedures to verify the competence of foreign professionals (Article V:6). 8 See e.g. the Disciplines on Domestic Regulation in the Accountancy Sector, which were adopted by the Council for Trade in Services on 14 December 1998. 9  Joint Ministerial Statement on Services Domestic Regulation, Ministerial Conference Eleventh Session, Buenos Aires, 13 December 2017, WT/MIN(17)/61, The recent proposals are set out in WTO document WT/MIN(17)/7/Rev.2, 13 December 2017, and are believed to resemble closely the still confidential draft disciplines developed during unfinished plurilateral negotiations on the Trade in Services Agreement.  10  The necessity test has been proposed by Chile; Hong Kong SAR, China; the Republic of Moldova; Peru; New Zealand and Switzerland.  10    of exceptions provisions, such as the Annex on Movement of Natural Persons, the Prudential Carve Out in financial services, and Exceptions under Article XIV for Privacy, etc. The Annex on Movement of Natural Persons states that “The Agreement shall not prevent a Member from applying measures to regulate the entry of natural persons …provided that such measures are not applied in such a manner as to nullify or impair the benefits accruing to any Member under the terms of a specific commitment.” The Annex on Financial Services, in the so called “prudential carve out” states that “Notwithstanding any other provisions of the Agreement, a Member shall not be prevented from taking measures for prudential reasons…Where such measures do not conform with the provisions of the Agreement, they shall not be used as a means of avoiding the Member’s commitments or obligations under the Agreement.” Finally, Article XIV on General Exceptions states that “nothing in this Agreement shall be construed to prevent the adoption or enforcement by a Member of measures necessary to secure compliance with laws or regulations which are inconsistent with the provisions of this Agreement including those relating to the protection of privacy of individuals in relation to the processing and dissemination of personal data…” These attempts to accommodate regulatory concerns strike an awkward and uneven balance between allowing regulatory freedom to purse certain objectives and preserving the value of market access commitments. But it is far from clear how much security is offered to the regulator or the exporter by the coexistence of commitments and largely untested exceptions provisions. And one of the two will inevitably be disappointed by any specific interpretation. The key questions is: can and should the appropriate balance between permissiveness and restraint be set by a WTO panel? Or should the balance be struck by Members through a cooperative process, with more symmetric commitments by importing and exporting countries? And should this cooperative process be pursued bilaterally, regionally or multilaterally? III. Limitations of proposed regulatory disciplines The proposed disciplines on domestic regulation, as reflected in the accountancy disciplines, WTO members’ submissions, and emerging from plurilateral fora like the TISA, have valuable elements. But the proposed disciplines do not adequately solve existing problems with the implementation of key GATS disciplines and may create new problems. First, proposed disciplines do not resolve the MFN and National Treatment conundrum identified above relating to the difficulty of establishing whether measures are de facto discriminatory. The problem is that new disciplines on domestic regulation under GATS Article VI:4 are seen as additional to the disciplines contained under the MFN, market access and national treatment provisions – i.e. applying to the right-most segment in Figure 4. Thus, for example, the “necessity test” is envisaged as analogous to the necessity test contained in the Agreements on 11    Technical Barriers to Trade (TBT) and on the Application of Sanitary and Phytosanitary Measures (SPS), as a further requirement on measures that have already passed the non- discrimination test that they not be more burdensome than necessary to achieve a legitimate objective. Such a necessity test may be premature because the primary need is not for a National Treatment plus or an MFN plus discipline, but to find a means of establishing whether regulatory requirements imposed on foreign services or service providers are discriminatory. That is, when a foreign doctor from country A is asked to undergo an examination and one from country B is asked to undergo an examination and three years of residency, is that inconsistent with MFN and/or national treatment? Denying the regulator any freedom to make a distinction is politically unsustainable; conceding to the regulator infinite freedom to make distinctions would make a mockery of GATS disciplines. The only reasonable way of making this determination is to ask some variant of the question: is the difference in treatment necessary?11 Figure 4: The Current View on the Scope of GATS Disciplines Second, the incorporation of a necessity test for non-discriminatory measures (as envisaged in GATS Article VI:4 and the draft accountancy disciplines) could be seen as unduly intrusive (Delimatsis, 2014). In situations where the level of attainment of an objective does not increase with stringency of a measure, it is easy to establish that a less stringent measure would achieve the objective. But in most cases, the level of attainment increases with stringency. The European Court of Justice could apply a proportionality test, weighing the benefits of higher attainment against the costs in terms of lost trade, but the WTO system does not represent a                                                              11 Some might baulk at this idea because GATT jurisprudence has established that a necessity test cannot be read into the national treatment (or MFN) provisions. But Article XVII:3 may offer an acceptable route to a similar assessment when it notes that “Formally identical or formally different treatment shall be considered to be less favorable if it modifies the conditions of competition in favour of services or services suppliers of the Member compared to the like services or service suppliers of another Member.” Either way, the unavoidable question is whether difference in treatment is necessary, not whether a non-discriminatory measure is necessary.    12    similar political consensus. In any case, countries currently seem unwilling to countenance rules targeting measures which do not discriminate in any way but are deemed to be too stringent. Finally, if such disciplines were incorporated into the GATS, the result could be an additional “hold-back” problem. In the goods trade world, where national treatment is a general obligation, quotas are prohibited, and tariffs are bound, strict disciplines on technical barriers targeted the main remaining source of impediments to trade without creating an incentive to retain protection. In services, where specific commitments do not yet cover all sectors or measures (as is evident from the discussion in Section I), premature stringency of regulatory disciplines that kick in when a country makes specific commitments, could inhibit the willingness to make binding commitments even to provide market access and national treatment. Reaching for the first best of deep goods-like disciplines could thus make even the second best of across-the-board basic disciplines hard to attain. IV. Regulatory cooperation as a precondition for liberalization It is convenient to begin by describing the more conventional producer-centric case for regulatory cooperation, as a complement to liberalization. That will help set the stage for the alternative consumer-centric case for cooperation, as a precondition for liberalization. IV.1 Producer-centric regulatory cooperation to address regulatory heterogeneity Regulatory heterogeneity arises when requirements differ across countries because of either differences in institutions (leading typically to “horizontal” differentiation e.g. in legal services) or differences in social preferences (leading to “vertical differentiation” e.g. stringency of financial or privacy regulation). The traditional case for regulatory cooperation arises from the fact that regulatory heterogeneity segments international markets in a way that prevents the exploitation of economies of scale in production. This is illustrated in Figure 5 drawn from De Bruijn et al. (2008). For example, since each East African country has its own regulatory requirements for services professionals, compliance costs cannot be spread out over provision of professional services in other East African countries but must be incurred separately in each market. De Bruijn et al. (2008) estimate that EU stock of FDI could increase by 20-35 percent if regulatory heterogeneity was reduced as a result of a common services regulation directive. Such regulatory heterogeneity cannot be addressed by imposing traditional trade disciplines because the problem is not due to protectionism or anti-competitive intent. But there is nevertheless an economic cost of such heterogeneity because each country is independently choosing its regulations without considering the negative impact on foreign producers and hence on competition. Therefore, there are potential gains from international cooperation where each 13    country trades-off the benefits of maintaining different nationally optimal regulations against the benefits of integrating markets through some form of regulatory convergence. Figure 5: The Costs of Regulatory Heterogeneity and the Gains from Regulatory Cooperation Source: De Bruijn et al. (2008). In some cases, regulatory cooperation could be far reaching and lead to harmonization or mutual recognition, which would eliminate the costs of regulatory heterogeneity for firms and liberate them from the uncertainty of discretionary licensing (Trachtman, 2014). In other cases, regulatory cooperation could be valuable even if it only involves greater mutual understanding of how regulatory discretion in each jurisdiction will be exercised because that too would lend predictability to commitments. IV.2 Consumer-centric regulatory cooperation to address international externalities The alternative case for regulatory cooperation arises because regulators in the jurisdiction of the service exporter do not consider the consequences of market failure for consumers in the jurisdiction of the service importer. Even if regulations in each country were identical, this problem would still exist. This is because the regulator in country A who has the power to control the actions of the provider in country A is not mandated to look after the interests of consumers in country B; and the regulator in country B, who is mandated to look after the interests of consumers in country B, does not have adequate control over the actions of the provider in country A. Such externalities matter profoundly in today’s world, where security is a growing concern in multiple dimensions closely related to services trade: financial internationalization and financial insecurity; digital trade and informational insecurity; labor mobility and insecurity about illegal immigration; and increasing concentration in global services markets and insecurity about monopolistic exploitation. 14    The shared element in each of these examples is a regulatory or policy externality transmitted from one jurisdiction to another through services exports. The inadequacy of financial regulation in one country can affect consumers and financial stability in other countries to which its financial institutions export services. Weak data protection in a country which exports data- processing services can compromise the privacy of citizens of other countries. Imperfect policing and emigration checks in a country whose individuals travel abroad to provide consultancy services could undermine law and order in other countries. Poor regulation of hospitals and universities in one country can hurt the health and human capital of foreign citizens who visit for treatment or education. As noted above, regulatory externalities exist in goods trade as well, but a country can liberalize trade policy and still apply technical regulations at the border. The intangibility of services and the simultaneity of production and consumption make post-production and pre-consumption inspection difficult. Since governments cannot ensure compliance with desired regulations ex ante, they are reluctant to liberalize. Harmonization and mutual recognition are not a solution to the problem. Thus, in services, trade restrictions can become a form of regulatory precaution, a second-best response to market imperfections. Only when a country is assured that imported services are adequately regulated will it liberalize and give up the right to protect. But in what form is this reassurance to be secured? IV.3 Destination-specific exporter regulatory commitments Conventional trade negotiations and rule-making are primarily concerned with reciprocal liberalization of import policy (Figure 6). Accordingly, rules and commitments focus on tying the hands of importers: tariffs are bound; quotas are prohibited or restrained; discrimination against imports and trading partners is prohibited or restrained; and there may be further disciplines on importing country product standards - e.g. the requirement in goods that they must be “necessary” to achieve a legitimate objective. For the most part, trade rules do not concern themselves with exporter disciplines or commitments. The rare examples in goods include prohibitions or restraints on export subsidies, quotas and agricultural assistance. Furthermore, reciprocal trade liberalization is in some cases complemented by producer-centric regulatory cooperation, designed to address the impact of regulatory heterogeneity through harmonization and mutual recognition. This asymmetric structure of trade rules, which focus entirely on rules and commitments on importing countries and none (or very few) on exporting countries, does not create a natural home for consumer-centric regulatory cooperation. The result is an unwillingness on the part of importing countries to give up protection, or regulatory discretion, or both. The solution may be mutually binding commitments by exporting and importing countries. The former would make 15    regulatory commitments to look after the interests of consumers in importing countries and in return the latter would make commitments to allow access to their markets – as represented by the diagonal line in Figure 6. Figure 6: Exporter regulatory commitments in return for importer market access commitments V. Destination-specific exporter regulatory commitments in practice Here are four examples that reflect recent developments and illustrate the heterogeneity of approaches to cooperation across modes of delivery and sectors. Trade in digital services, financial services, and labor mobility highlight issues arising, respectively, in trade through cross-border delivery, commercial presence and presence of natural persons, while competition policy is a cross-cutting issue. Developing countries have a particularly strong interest in trade in digital services and labor mobility as exporters of services and in financial services and competition policy as importers of services. In the former cases, they would seek to make regulatory commitments to secure market access, whereas in the latter cases, they would see to make market access commitments conditional on regulatory commitments from exporting countries. V.1 Regulatory commitments to ensure liberal trade in digital services and free data flows12 Governments are taking different approaches to regulating personal data collected by private enterprises. One of the most ambitious efforts is the European Union’s new General Data Protection Regulation, with wider scope and stronger enforcement than in the earlier Data Protection Directive, which took effect in May 2018 (European Union, 2018). The EU                                                              12 This section draws on Mattoo and Meltzer (2018). 16    conception of privacy as a fundamental human right reflects its own history and cultural trajectory which other countries might not share. To ensure that the personal data of EU citizens is not abused abroad, data can be transferred out of the EU only under certain conditions. One is that the country meets the EU’s “national adequacy” requirement by enacting a national privacy law essentially equivalent to that of the EU. So far only five countries have received a positive national adequacy finding (Argentina, Uruguay, Israel, New Zealand, and Canada (commercial organizations)). A national law imposes the same standard on all firms in the country, regardless of where they sell, at home or abroad. That could adversely affect poorer countries. Prematurely stringent privacy laws could hurt the development of markets by inhibiting the flow of information. For example, the reporting of personal credit histories is critical to consumer credit, and privacy laws could create significant asymmetries of information and affect the efficiency of markets (Kitchenman, 1999). If a country’s national law fails the EU adequacy test, as happened in the case of India, their firms will be required to use either Binding Corporate Rules (BCRs), designed for multinational companies to move data globally, or Standard Contractual Clauses (SCCs) for each business deal. Both instruments require the levels of protection, oversight and access for individuals that would be offered in the EU. Both also require a data controller or processor, who can be held liable for breach, to be established in an EU Member State. Both routes are costly and time-consuming. The requirement of a presence in the EU increases costs and limits the benefits of seamless cross-border digital trade, especially for smaller firms. A survey in India of the impact of the earlier, less-stringent EU Data Protection Directive revealed that the BCR process took over six months, and 90 percent of the respondents used SCCs which also involved a complex process and took on average more than 3 months (NASSCOM-DSCI, 2013). As many as two-thirds of the surveyed services exporters claimed a significant loss of business opportunities because of the requirements. Is it possible to satisfy the EU’s legitimate needs without obliging other countries to accept nation-wide EU standards or to incur the substantial compliance costs associated with SCCs and BCRs? The EU-US Privacy Shield Agreement In 2016, the United States and the EU concluded the Privacy Shield – an arrangement that the EU Commission has deemed “adequate” under the Data Directive – thereby enabling the transfer of personal information from the EU to businesses in the United States participating in the Privacy Shield (EC 2017). The Privacy Shield replaced the EU-U.S. Safe Harbor framework, which in 2015 the CJEU found did not provide an adequate level of privacy protection due to an 17    absence of rights of redress for EU citizens with respect to government access to their data (Schrems v. Data Protection Commissioner 2015). Under the Privacy Shield, U.S. companies through an industry body or individually self-certify to the U.S. Department of Commerce that they will protect personal data consistent with the Privacy Framework, which includes the Privacy Shield Principles (Privacy Shield Framework 2018). These Principles largely reflect the key elements of the EU Data Directive (U.S. Department of Commerce 2016). The main ones are: to give European data subjects notice that a US entity is processing their data; to provide choice including whether to opt out of providing personal information; accountability for any onward transfers to third parties of personal information; to take reasonable and appropriate steps to protect personal data from loss or misuse; to process personal data only for purposes the organization intends to use it; to give European data subjects access to their personal information and the ability to correct, amend or delete inaccurate information; and to enforce the principles and to give European data subjects access to affordable enforcement mechanisms. U.S. businesses are required to publish their privacy policies, and the Privacy Shield gives the U.S. Federal Trade Commission jurisdiction over such businesses should they breach their own policy. In addition, the United States provides various means of redress for people, whose personal data has been compromised, including a direct complaint to the business or a complaint to the Department of Commerce. Also, under the Privacy Shield, the United States has agreed to establish an ombudsperson to address complaints about government agencies’ access to personal information from the EU on national security grounds. Despite the upgrading of privacy protection under the Privacy Shield compared with Safe Harbor, the stability of the arrangement is in doubt. The first annual review of the Privacy Shield by the Commission and the Department of Commerce identified a number of concerns on the EU side that the operation of the Privacy Shield is not providing a sufficient level of privacy protection. In particular, the EU expressed concern over inadequate access by Europeans to redress mechanisms, including failure to appoint officials at the State Department Ombudsperson. Such an agreement with the EU would give participating US firms big advantages over existing options. First, unlike in the case of BCRS and SSCs, the firms would not be required to establish a costly presence in the EU because the assessment of conformity with EU standards would take place at home by domestic regulators. Second, unlike in the case of national adequacy, firms would not be obliged to adopt more stringent and more costly standards for data involving transactions at home or with countries less demanding than the EU. The Provisions on Data Flows in the Trans-Pacific Partnership What was originally the TPP provision and eventually became the CPTPP provision on data flows, contained in Chapter 14 on Electronic Commerce, has been widely and justifiably 18    described as far-reaching. While commitments to allow data flows are implicit in GATS commitments, it helps to have greater clarity through the CPTPP’s explicit obligation under Article 14.11.2 that “Each Party shall allow the cross-border transfer of information by electronic means, including personal information, when this activity is for the conduct of the business of a covered person.” In addition, the commitments on data flows are not limited to the agreements services commitments. Similarly, even though data localization requirements could arguably fall foul of the GATS national treatment obligation because they could be regarded as de-facto discriminatory, it helps to have an explicit obligation in Article 14.13.2 that “No Party shall require a covered person to use or locate computing facilities in that Party’s territory as a condition for conducting business in that territory.” The remarkable innovation in the CPTPP is that, even though there is an exceptions provision similar to that in the GATS (TPP Article 14.11.3), the CPTPP breaks new ground in creating obligations on data destination countries to prevent fraud and deception and protect personal information. In particular, Article 14.8.2 on Personal Information Protection requires that “each Party shall adopt or maintain a legal framework that provides for the protection of the personal information of the users of electronic commerce.” And Article 14.8.3 stipulates that “Each Party shall endeavor to adopt non-discriminatory practices in protecting users of electronic commerce from personal information protection violations occurring within its jurisdiction.” Such reciprocal obligations on data source and destination countries are a perfect example of the type of regulatory cooperation that is needed to reassure data source countries that their commitments to openness will not place their consumers at the mercy of indifferent foreign regulators. Moreover, the existence of such shared obligations also reduces the needed scope for unilateral action by source countries under the exceptions provisions and, therefore, creates greater security of access for exporters. It remains to be seen how parties will interpret and implement Article 14.8.3 stipulation to “endeavor to adopt non-discriminatory practices in protecting users of electronic commerce,” and whether all consumers and contracts will be adequately covered regardless of jurisdiction in which they are located. Nevertheless, the TPP approach to data flows may well be a model for the form of regulatory cooperation that would induce wider and deeper commitments in services trade. The close link between unrestricted data flows and regulatory cooperation was also evident in the decision to exempt the financial services industry from the TPP rules safeguarding the cross- border flow of data. Treasury Secretary Jacob Lew was reported to have defended the US insistence on this exclusion, stating that “One of the issues here is the requirements of our regulators in terms of ... what they need to have their prudential reviews of financial institutions. We can’t give away something that our financial regulators would need here in the United States" (Law360 2016). In future negotiations, if countries were to promise to “extradite” data needed for prudential reviews to the concerned regulator, then there would be less need for local data storage and the exclusion of financial services from the data flow obligation. 19    V.2 Regulatory commitments to facilitate efficient trade in financial services More than in other sectors, the consequences of financial liberalization depend on the regulatory and supervisory framework. But in internationalized markets, it is not just regulation at home that matters but also the quality and objectives of regulators in trading partners. If financial regulators in each jurisdiction are either not capable or not inclined to consider the impact of their actions in other jurisdictions, foreign trade and investment can create risks. An aversion to assuming those risks can lead to a reluctance to open markets. There is already a long history of international cooperation to develop shared standards for financial regulation. A central issue has been capital requirements, which specify the amount of loss-absorbing equity financing financial institutions must maintain. Many countries think of even the strengthened requirements under the Basle III accord on capital as a floor, and are inclined to impose tougher capital requirements. In this respect, Basle III reflects a more general issue with international standards in services. In the few instances where they exist, they are viewed as necessary rather than sufficient conditions. Their fulfillment in the home jurisdiction does not guarantee unimpeded rights of entry and operation in other jurisdictions. These issues are illustrated by the difficulties that confronted even the relatively open US and EU financial markets in the currently dormant Transatlantic Trade and Investment Partnership (TTIP). After the financial crisis, a key element of the reconfiguration of the financial systems in the United States under the Dodd-Frank legislation involved limiting the leverage of banks and other financial institutions, imposing restrictions on proprietary trading, and developing resolution procedures for firms facing solvency. European financial institutions feel burdened by some of these developments; e.g. new rules ended exemptions under which US subsidiaries of European banks like Barclays and Deutsche Bank could operate with relatively little capital because their global parents were regarded as well-capitalized. A commitment by the parent regulator to make capital outside the United States available in certain adverse states of the world may have helped sustain the exemption. The EU Council sought to address both market access and regulatory issues within the TTIP, calling for a common framework that was “binding on all regulators and other competent authorities” and a “common frameworks for prudential regulation” (Johnson and Schott, 2013, p 2). In contrast, the US Trade Representative, Michael Froman was willing to deal with market access issues in the TTIP but declared “that nothing we do in a trade agreement should undermine the ability of regulators on both sides to regulate in the public interest” and would have preferred to pursue regulatory cooperation within existing fora (Johnson and Schott, 2013).13 For example, the Federal Deposit Insurance Corporation (FDIC) and the Bank of England have agreed on how to deal with situations in which a global megabank were to fail.                                                              13 These fora include the Basle Committee for Banking Supervision, the Financial Markets Regulatory Dialogue (US-EU), the International Organization of Securities Commissions, the International Association of Insurance Supervisors, the Financial Stability Board, and the G-20. 20    Regardless of where and how regulatory cooperation takes place, both parties clearly recognize its importance in sustaining open markets. Even within Europe, there is a growing tension between the increasingly integrated markets and absence of an integrated supervisory system. Consider the example of the European Free Trade Area (EFTA) market, in which a “single passport” allows a Bank in one EFTA country to establish branches in other member countries without being subject to effective local regulation. Under this provision, Iceland’s Landsbanki established Icesave branches in the United Kingdom and the Netherlands. The failure of the bank in 2008 revealed that UK and Dutch deposits in the branches were not adequately covered by deposit insurance. The EFTA court eventually ruled that Iceland was not responsible for paying deposit insurance after the collapse. The systemic implications for cross-border banking are still being felt. More generally, even within Europe, there is a realization that integrating banking markets requires the development of Europe-wide mechanisms for supervision by the European Central Bank, a common deposit-insurance scheme, and a resolution regime for failing cross-border banks, including bankruptcy and restructuring procedures. Moreover, the credibility of the mechanisms depends on the power to enforce of the European Court of Justice. At the heart of the Brexit financial market negotiations, is the difficulty UK regulatory authorities face in credibly assuming “source country regulatory commitments” by only promising to adhere to European regulatory standards but not accepting the jurisdiction of the ECJ. Each of these issues is also relevant to trade in financial services involving developing countries, but a slightly broader issue is particularly relevant in North-South relations. While it is generally held that the presence of foreign financial institutions has been a stabilizing force when financial crises originated in developing countries, foreign presence may not have been an entirely benign influence during the recent crisis which originated in industrial countries. Haas (2014), in a review of the literature, shows how the crisis was propagated eastwards by Western banks reducing the credit supply in emerging Europe (faster than domestic banks did).14 However, it is notable that foreign banks that took part in the Vienna Initiative, a public-private coordination mechanism to guarantee macroeconomic stability in emerging Europe, were somewhat more stable lenders.15 The Initiative specifically sought to limit the negative fallout from nation-based uncoordinated policy responses to the global crisis and to avoid a massive and sudden deleveraging by cross-border bank groups in emerging Europe. Even though the initiative was put in place during the crisis, the existence of such a mechanism does provide a guarantee                                                              14 Crisis transmission to Latin America was more severe in countries where foreign banks were lending across borders rather than through subsidiaries. Subsidiaries that were funded locally instead of through the international wholesale markets or through their parent banks were particularly stable credit sources. 15  The Vienna Initiative was launched at the height of the first wave of the global financial crisis in January 2009. It brought together all the relevant public and private sector stakeholders of EU-based cross-border banks active in emerging Europe. See http://vienna-initiative.com/.  21    against “financial nationalism” by exporting countries during crises (“British loans for British firms”) which can have a chilling effect on liberalization. It is evident that some forms of cross-border regulatory commitments or obligations are necessary to maintain open financial markets. The commitments need not take place within the context of trade negotiations but could take place in other fora. Whether these parallel efforts need to result in legally binding agreements will depend on how much trust there is between regulatory institutions in different countries, as well as on how much need there is for regulators to respond flexibly to different and changing financial conditions in different countries. Multilaterally, efforts in this area could be facilitated by greater coordination between, on the one hand, the IMF and the WTO, which help guarantee states’ financial openness, and, on the other hand, institutions like the Bank for International Settlements and the Financial Stability Forum (with ideal expanded membership), which deal with financial regulation. V.3 Source country obligations in bilateral labor agreements to facilitate labor mobility Progress in negotiations on the movement of individual service providers (referred to as mode 4 in the GATS) has become almost a precondition for more meaningful developing country participation in the process of reciprocal market opening. However, it is proving extremely difficult for a number of countries to make any “concessions” in this area. How can we make mode 4 a positive outcome rather than a millstone for the services negotiations? First, countries need to recognize that simply asserting that mode 4 is about trade in services and not about migration cannot dispel deep-rooted fears raised by the entry of foreign individuals. These fears must be acknowledged and addressed. One way forward may be to consider reciprocal obligations on mode 4, drawing upon the experience of a few relatively successful bilateral and regional agreements (Saez, 2013). The inclusion of labor mobility in the framework of a trade agreement implies that obligations are assumed by host countries alone, to provide market access on an MFN basis regardless of conditions in source countries. In contrast, the assumption of obligations by source countries also is a key element of regional trade agreements (e.g. APEC) that have facilitated mobility of the skilled, and of bilateral labor agreements (e.g. between Spain and Ecuador, Canada and the Caribbean, Germany and Eastern Europe) that have slightly improved access for the unskilled. Source country obligations include pre-movement screening and selection, accepting and facilitating return, and commitments to combat illegal migration. In effect, cooperation by the source can help address security concerns, ensure temporariness and prevent illegal labor flows in a way that the host is incapable of accomplishing alone – and constitute a service for which the host may be willing to pay by allowing increased access. In the current GATS framework, when a country makes a market access commitment, it is obliged to grant a fixed level of access every year in the future regardless of domestic economic 22    conditions. In contrast, bilateral labor agreements allow host countries to vary the level of access depending on the state of the economy. One example is the bilateral agreement between Germany and certain Eastern European countries, under which the mutually agreed and enforced quota on temporary migrants increased (decreased) by 5 percent for each one percentage point decrease (increase) in the level of unemployment. It may be desirable to consider shared GATS commitments along these lines, which allow necessary flexibility albeit in a cooperative, transparent, predictable and objectively verifiable manner, and would be a big improvement over the opaque, unilaterally implemented economic needs tests that infest GATS schedules. V.4 Source country obligations on pro-competitive regulation Anti-competitive practices in international markets that fall outside the jurisdiction of national competition authorities have been an issue in maritime, air transport and communication services, but are now provoking concern in new digital services in search, advertising, communication and distribution. The current GATS provision on anticompetitive practices (Article IX) provides only for information exchange and consultation. Meaningful international cooperation on the enforcement of competition policy would reassure “importing” countries facing jurisdictional constraints or limited enforcement capacity that the gains from liberalization will not be appropriated by international cartels. The problem is not hypothetical. Fines of around $1 billion or more were imposed by the UK’s Financial Conduct Authority, the US’s Commodity Futures Trading Commission and Swiss regulators on the world’s biggest banks – Barclays, JPMorgan Chase, Royal Bank of Scotland, Citigroup and Credit – for manipulating foreign exchange markets.16 The rigging apparently took place through information sharing and coordinated trading. In 2013, the European Commission fined eight international financial institutions, including Barclays, Deutzche bank, RBS and Societe Generale, a total of €1.71 billion for participating in illegal cartels in markets for financial derivatives covering the European Economic Area (EEA). In 2012, large banks also struck deals resulting in large fines which helped them to avoid prosecution for manipulation of the Libor benchmark interest rate. In 2010, the European Commission fined 11 air cargo carriers a total of nearly €1 billion for operating a worldwide cartel which affected cargo services within the European Economic area (EEA).17 The carriers coordinated their action on surcharges for fuel and security without discounts over a six-year period. In 2013, the European Commission opened formal antitrust proceedings against 14 of the world’s biggest container shipping firms in Europe and Asia. The                                                              16 “Five banks set to pay more than $6bn to settle forex manipulation claims,” Financial Times, 12 May 2015. 17 Several known airlines are among the 11 undertakings fined, namely Air Canada, Air France-KLM, British Airways, Cathay Pacific, Cargolux, Japan Airlines, LAN Chile, Martinair, SAS, Singapore Airlines and Qantas – with Lufthansa (and its subsidiary Swiss) receiving full immunity from fines under the Commission's leniency program, as it was the first to provide information about the cartel. 23    Commission was concerned that the public announcement of new prices enabled companies to signal future price intentions to each other, which led to higher prices for container liner shipping transport services on routes to and from Europe. The fact that anticompetitive practices are being carried out in the EU and US markets despite their powerful competition authorities raises concern about the practices of these multinational banks, airlines and shipping lines in other countries. Could it be the case that actions by the EU and US authorities provide a global public service by acting against cartels? Unfortunately, there are limits to both when and how action is taken. If anti-competitive effects are only felt outside their jurisdiction, there is no basis for action under current national laws. A notable aspect of the European Commission’s fines imposed on the air cargo carriers was that all carriers were granted a 50 percent reduction on sales between the EEA and third countries because part of the harm of the cartel fell outside the EEA. International cooperation could help overcome the limitations of national bodies, especially in small countries, but is not easy to accomplish. For example, South Africa is a significant investor across southern Africa and is one of the biggest foreign investors in Zambia. For the most part, South African investment has been welcome but there are also concerns that South African companies may abuse their market power, to the detriment of local producers and consumers. The Zambian Competition Commission (ZCC) is not able to deal effectively with accusations against South African companies of anticompetitive behavior in Zambia because: it often does not have the jurisdiction to deal with companies that operate in Zambia but are not locally incorporated, or whose actions do not fall under Zambian Competition law for some other reason; it does not have the ability to enforce its decisions, even where it does have jurisdiction to issue judgments; it is often unable to obtain the information necessary to investigate the activities of a foreign company from the home jurisdiction of that company; and despite its competent and motivated staff, it lacks the resources to conduct the detailed empirical investigations required to address effectively allegations of anticompetitive behavior. One possibility would be for Zambia to condition opening its market to South African firms on a commitment by the South African authorities to investigate anticompetitive behavior by South African companies in Zambia, or to assist the local authorities in doing so. In principle, it would be in South Africa’s interest to provide such reassurance. However, Section 3 of South Africa’s Competition Act states that the Act “applies to all economic activity, within, or having an effect within, the Republic” (own italics). Still the South African Act does allow South Africa’s agencies to help to a limited extent: foreign agencies investigate behavior that has a South African as well as regional impact; share information but only if the companies concerned agree to this; through technical assistance. In the longer term, an option is to form a regional competition agency to which national competition agencies could forbear jurisdiction in specific circumstances, as in the case of the European Commission within Europe. For example, to save costs, in May 2000, St Lucia, 24    Dominica, Grenada, St Vincent and the Grenadines, and St Kitts and Nevis set up, with World Bank support, the Eastern Caribbean Telecommunications Authority (ECTEL), the first regional telecommunications authority in the world. Although the member countries retained their sovereign power over licensing and regulation, ECTEL provides technical expertise, advice and support for national regulations. Apart from the economies of scale in establishing a common regulator, there are at least three other advantages. It promotes the development of harmonized and transparent regulation in the region, allows for greater independence (and hence credibility) in regulatory advice, and enhances bargaining power in negotiations with incumbents and potential entrants. In fact, there is evidence that the creation of ECTEL, along with other reforms, prompted a decline in the price of a daytime call to the United States between 24 and 42 percent in these countries. More generally, the United States and the European Union that own and control many services multinationals, could at least begin by ending the exemption of collusive practices whose effects are felt outside their jurisdiction from the scope of their competition law. More ambitious, but perhaps not very realistic, would be to create a right for foreign consumers to challenge anti- competitive practices by services firms in the national courts of countries whose citizens own or control these firms- giving to consumers rights analogous to those which firms enjoy under the WTO rules on intellectual property and government procurement. VI. Regulatory commitments can create a risk of exclusion Regulatory cooperation along the lines proposed here will inevitably be among a sub-set of countries at least initially, because not all countries will be able to make or extract credible commitments. That may naturally lead to patterns of trade based on mutual trust rather than comparative advantage. Data, investment and people may travel between jurisdictions that have been able to reassure each other about their standards, rather than to other non-conforming jurisdictions. Of course, in some cases, such as the enforcement of competition policy, regulatory cooperation can generate positive externalities for third countries and be a global public good. But even in this case, as we saw above, there is a risk that regulators primarily address anticompetitive behavior that has an adverse impact within their own jurisdictions rather than in third countries. VI.1 Allowing the excluded countries access to recognition agreements GATS Article VII on Mutual Recognition Agreements dealing with recognition, attempts to strike a difficult balance. On the one hand, it is permissive and allows space for regulatory cooperation. Thus, Article VII:1 recognizes that “a member may recognize the education or experience obtained, requirements met, or licenses or certifications granted in a particular country” as part of an agreement or autonomously. 25    The remaining paragraphs of Article VII seek to ensure that this freedom is not abused. Article VII:2 requires a Member who enters into a mutual recognition agreement (MRA) to afford adequate opportunity to other interested Members to negotiate their accession to such an agreement or to negotiate comparable ones. More importantly, Article VII:3 stipulates that a Member must not grant recognition in a manner which would constitute a means of discrimination between countries in the application of its standards or criteria for the authorization, licensing or certification of services suppliers, or a disguised restriction on trade in services.18 Recognition, unilateral or through an MRA, amounts to an acceptance of likeness vis-à-vis certain suppliers, it also defines a standard of treatment vis-à-vis other suppliers and provides others with a potentially valuable foothold. Unfortunately, WTO Members tend to notify recognition agreements under the exemption for preferential agreements contained in GATS Article V. These agreements include the one establishing the European Union, agreements between the European Union and neighboring countries, and the Closer Economic Relations Treaty between Australia and New Zealand. The tendency of Members to notify MRAs under the “closed” Article V rather than “open” Article VII may reflect an attempt to share these gains on a limited reciprocal basis by avoiding the obligation to extend recognition more widely. Ideally, participants in MRAs would agree not to impose restrictive rules of origin and extend their benefits to all providers that have met comparable standards. VI.2 Avoiding the tyranny of harmonization by allowing destination-specific commitments As countries strive to become candidates for recognition agreements by improving their domestic standards, they may confront a difficult choice if standards are not “separable”. Separable standards are destination specific. For example, the Philippines may train some nurses who are going to work in the United States to a higher (or different) standard than those who are going to work domestically. Non-separable standards are those which for technological or, more usually, legal reasons are origin specific and must be identical for all destinations. Examples are economy-wide prudential regulations in financial and professional services, and the “national adequacy findings” stipulated in the EU’s privacy directive. The efforts by the Philippines (and to an extent, India) to respond to the EU privacy directive illustrate the dilemma for developing countries. On the one hand, if they choose not to enact laws deemed adequate, they could be shut off from participation in the large EU market. In the absence of such laws and given the weakness of local legal and regulatory systems, it might be difficult for private firms in developing countries to emulate United States firms like Microsoft and credibly commit to meet the required high standards. On the other hand, if they do enact stringent laws, it is unlikely that they could be made specific to trade with particular                                                              18 It is also relevant that Article VII:5 states that "wherever appropriate, recognition should be based on multilaterally agreed criteria" and requires Members to work towards the establishment and adoption of such criteria. 26    jurisdictions, and so the result could be an economy-wide increase in the costs of doing business. For instance, if private sector estimates generated in the United States are to be believed, information sharing saves the customers of 90 financial institutions (accounting for 30 percent of industry revenues), $17 billion a year ($195 per average customer household) and 320 million hours annually (4 hours per average customer household) (Glassman, 2000). In fact, the Philippines initially enacted national privacy legislation in order to ensure continued access to the EU data processing market. But the result was that many Philippines-based US firms found it difficult/costly to operate in the Philippines and suspended investment plans. Whereupon the Philippines government was obliged to reverse course. India too has apparently been struggling to find the right balance. This is not to suggest that there might not be good reasons to adopt high standards. However, the desired level of standards may differ across countries, and if trade is made conditional on the existence of “comparable” laws, then there might be a socially costly “race to the top.” For example, in financial services, universal “know-your-customer” laws enforced through blacklisting threats for non-complying jurisdictions, has forced many developing countries to adopt standards that deny access to financial services to poor households that cannot meet the burdensome documentation requirements. In general, it would therefore be desirable to allow recognition of standards in a way that allows separability, i.e. the standard has to be met only by services and services suppliers destined for the relevant market. Then, developing countries would not be obliged to trade-off access to markets against access to services for their own citizens. VII. Some final observations Finally, here are a few observations anticipating some concerns with the proposed course of action. Regulatory commitments do not mean freezing national regulation and depriving regulators of the flexibility to respond to changes in markets and technology. Regulatory cooperation will be a dynamic process because regulation is dynamic. Thus, the regulations emanating from Dodd- Frank were still being written years after the law was passed and some are now being amended. The Safe Harbor Framework served an important purpose, but was renegotiated as the Privacy Shield, which too may be amended. Besides, there is significant heterogeneity across services sectors, as the four examples presented in this paper reveal. The scope and form of regulatory commitments will differ across services sectors depending on the nature of the regulatory externality. There will also be differences across countries within the same services sector. We already have significant regulatory cooperation in certain areas (e.g. in financial services) but not in other areas (e.g. labor mobility). There are also similarities across services sectors, e.g. in the procedural aspects of regulation. 27    Thus, regulatory cooperation could also take the novel form of a trade facilitation agreement in services to address procedural impediments to trade, as has recently been proposed in the WTO. Regulatory cooperation has in some cases been more inclusive of developing countries (e.g. development of financial regulatory standards under Basle III) but less so in other cases (e.g. competition policy). Most relevant in the present context, regulatory cooperation has been ad hoc and rarely synchronized with trade negotiations. 19 This dichotomous approach internationally is mirrored by the lack of coherence nationally, reflected in the lack of coordination between ministries of commerce and sectoral ministries in the diagnosis of regulatory inadequacies and the sequencing of remedial action and liberalization. Developing countries need more support in diagnosing and remedying regulatory inadequacies that could undermine the benefits of liberalization at home and abroad. On the one hand, they need to be equipped to make credible commitments to conform to export country standards in areas like digital trade and labor mobility. On the other, they need support to extract commitments from industrial countries to extend the reach of their prudential and pro-competitive regulation to cover their providers serving developing country markets. Poorer developing countries would participate meaningfully in negotiations that offered an opportunity not merely to make binding commitments, but also to mobilize assistance for such regulatory reform. Participation in supply chains diminish the importance of regulatory impediments to trade, but regulatory impediments can still influence the distribution of gains within supply chains. The fragmentation of services into tasks has meant that it is no longer necessary for services exporters to jump through all the regulatory hoops and incur all the regulatory costs, such as those estimated in Table 1. Thus, an architect in Colombia can today produce a draft design or an accountant in India a draft tax return and send it digitally to their locally recognized counterparts in the United States who will ensure and certify conformity with local standards. However, if there are restrictions on entry into architecture and accountancy services in the United States, then US licensed professionals can extract a rent for their services and grab a larger than competitive share of the total value of the service. Even if there are no restrictions on entry, a similar problem arises if the US service providers do not really have a comparative advantage in providing the final certification services. Thus, regulatory cooperation that makes all stages of services production contestable remains relevant in a world of global supply chains. Can regulatory commitments be incorporated in a multilateral agreement? In general, we would expect regulatory cooperation among countries to proceed step-by-step in small groups, self- selecting into specific arrangements and gradually deepening them. As a first step, importing countries may specify conditions unilaterally and determine conformity unilaterally, but lend additional transparency and predictability to their requirements by listing them, e.g. as Additional Commitments under Article XVIII of the GATS. A further step could be for importing countries to recognize conformity assessment in specific exporting countries when                                                              19 Mattoo (2005), Hoekman and Mattoo (2013). 28    there is trust in enforcement even though norms diverge. In parallel, groups of countries could also make collective additional commitments when they converge in regulatory requirements – say, in a WTO Reference Paper. These steps could pave the way ultimately for mutually binding obligations on source and destination countries, which is one of the most innovative elements of the rules on data flows in the Comprehensive and Progressive Agreement for Trans-Pacific Partnership (CPTPP). Finally, it is of course true that, in some cases, regulatory concerns may be a smoke-screen for vested interests. But addressing legitimate regulatory concerns would help lay bare the protectionist interests and traditional reciprocal liberalization is the best antidote for the power of these interests. Regulatory cooperation may not be sufficient to deliver liberalization, but it is necessary. References Borchert, Ingo, Batshur Gootiiz and Aaditya Mattoo (2011), Services in Doha: What is on the Table?, in Will Martin and Aaditya Mattoo (eds.), Unfinished Business: The WTO’s Doha Agenda, CEPR and World Bank, London. Borchert, Ingo, Batshur Gootiiz and Aaditya Mattoo (2014), Policy Barriers to International Trade in Services: Evidence from a New Database, World Bank Economic Review, vol. 28(1), 162-188. Deardorff, Alan V. and Robert M. Stern (2008), “Empirical Analysis of Barriers to International Services Transactions and the Consequences of Liberalization”, Chapter 5 in: Aaditya Mattoo, Robert M. Stern and Gianni Zanini (eds.), A Handbook of International Trade in Services, Oxford University Press, pp. 169-220. De Bruijn, Roland, Henk Kox and Arjan Lejour (2008), Economic benefits of an Integrated European Market for Services, Journal of Policy Modeling, vol. 30, issue 2, pages 301-319. De Haas, Ralph (2014), The Dark and Bright Side of Global Banking: a (somewhat) cautionary tale from emerging Europe, EBRD Working Paper No. 170, London. Panagiotis, Delimatsis (2014), Who’s afraid of necessity? And why it matters?, Chapter 6 in Lim, Aik Hoe, and Bart De Meester (eds), (2014), WTO Domestic Regulation and Services Trade: Putting Principles into Practice, Cambridge University Press, Cambridge. EC (European Commission). 2017. EU – U.S. Privacy Shield – First annual Joint Review. Article 29 Data Protection Working Party, WP255. https://ec.europa.eu/newsroom/just/document.cfm?doc_id=48782 EU (European Union). 2018. GDPR Portal. https://www.eugdpr.org/ 29    Glassman, C. A. (2000), Customer Benefits from Current Information Sharing by Financial Services Companies, a study for the Financial Services Roundtable, December. Gootiiz, Batshur, and Aaditya Mattoo (2017), Services in the Trans-Pacific Partnership: What Would be Lost?, Journal of World Trade, Volume 51(5), pp. 787-816. Hoekman, Bernard (1996), Assessing the General Agreement on Trade in Services, in: Will Martin and L. Alan Winters (eds.), The Uruguay Round and the developing countries, Cambridge: Cambridge University Press, pp. 88-124. Hoekman, Bernard, and Aaditya Mattoo (2013), Liberalizing Trade in Services: Lessons from Regional and WTO Negotiations, with Bernard Hoekman, International Negotiations, vol. 18, pages 131–151. Johnson, Simon, and Jeffery J. Schott (2013), Financial Services in the Transatlantic Trade and Investment Partnership, Peterson Institute for International Economics Policy Brief No. PB13- 26, Washington DC. Kitchenman, W. F. (1999), “U.S. Credit Reporting: Perceived Benefits Outweigh Privacy Concerns,” The Tower Group. Krajewski, Markus (2014), Balancing legal certainty with regulatory flexibility, Chapter 5 in Lim, Aik Hoe, and Bart De Meester (eds), (2014), WTO Domestic Regulation and Services Trade: Putting Principles into Practice, Cambridge University Press, Cambridge. Lim, Aik Hoe, and Bart De Meester (eds), (2014), WTO Domestic Regulation and Services Trade: Putting Principles into Practice, Cambridge University Press, Cambridge. Mattoo, Aaditya, and Pierre Sauve (2003), Domestic Regulation and Services Trade Liberalization, Oxford University Press and World Bank, Washington, D.C. Mattoo, Aaditya (2005), Services in a Development Round: Three Goals and Three Proposals, Journal of World Trade, vol. 39, 1223-1238. Mattoo, Aaditya, and Deepak Mishra (2009), Foreign Professionals in the United States: Regulatory Impediments to Trade, with Deepak Mishra, Journal of International Economic Law, vol. 12, 435-456. Mattoo, Aaditya, and Joshua P. Meltzer (2018), International Data Flows and Privacy: The Conflict and Its Resolution, World Bank Policy Research Working Paper 8431. NASSCOM‐DSCI (National Association of Software and Services Companies ‐ Data Security Council of India) (2013.), Survey of the Impact of EU Privacy Regulation on India’s Services Exporters. 30    Pelkmans, Jacques (2003), Mutual Recognition in Goods and Services, Working Paper no. 16, European Network of Economic Policy Research Institutes. Privacy Shield Framework (2018), “Privacy Shield Overview.” https://www.privacyshield.gov/ProgramOverview Saez, Sebastian (ed.) (2013), Let Workers Move: Using Bilateral labor Agreements to Increase Trade in Services, World Bank, Washington, DC. Schrems v. Data Protection Commissioner, Court of Justice of the European Union, C362/14, 6 (2015). Trachtman, Joel P. (2014), Mutual recognition of services regulation, Chapter 7 in Lim, Aik Hoe, and Bart De Meester (eds), (2014), WTO Domestic Regulation and Services Trade: Putting Principles into Practice, Cambridge University Press, Cambridge. U.S. Department of Commerce. 2016. E.U.‐U.S. Privacy Shield Framework Principles, Issued by the U.S. Department of Commerce. https://www.privacyshield.gov/servlet/servlet.FileDownload?file=015t00000004qAg 31