WPS5517 Policy Research Working Paper 5517 Services Trade Liberalization and Regulatory Reform Re-invigorating International Cooperation Bernard Hoekman Aaditya Mattoo The World Bank Poverty Reduction and Economic Management Network International Trade Department & Development Research Group Trade and Integration Team January 2011 Policy Research Working Paper 5517 Abstract Trade and investment in services are inhibited by a range place the preconditions for future market opening. The of policy restrictions, but the best offers so far in the second proposal is for a new approach to negotiations Doha negotiations are on average twice as restrictive as in the World Trade Organization, with a critical mass of actual policy. They will generate no additional market countries that account for the bulk of services production opening. Regulatory concerns help explain the limited agreeing to lock-in applied levels of protection and pre- progress. This paper develops two proposals to enhance committing to reform of policies affecting foreign direct the prospects for both liberalization of services trade and investment and international movement for individual regulatory reform. The first is for governments to create service providers--two areas where current policy is most mechanisms ("services knowledge platforms") to bring restrictive and potential benefits from liberalization are together regulators, trade officials, and stakeholders to greatest. If these proposals cannot be fully implemented discuss services regulatory reform. Such mechanisms in the Doha time frame, then any Doha agreement could could identify reform priorities and opportunities for at least lay the basis for a forward-looking program of utilization of "aid for trade" resources, thereby putting in international cooperation along the proposed lines. This paper is a product of the International Trade Department , Poverty Reduction and Economic Management Network; and the Trade and Integration Team, Development Research Group. 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://econ.worldbank.org. The authors may be contacted at Bhoekman@ worldbank.org and 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 Trade Liberalization and Regulatory Reform: Reinvigorating International Cooperation* Bernard Hoekman (World Bank and CEPR) Aaditya Mattoo (World Bank) Keywords: services regulation, trade liberalization, international negotiations, trade agreements, development assistance, aid for trade, knowledge platforms, WTO, GATS JEL classification: F1, F2, F5, L80 * This paper draws on research that has been supported by the Global Trade and Financial Architecture project and the MultiDonor Trust Fund for trade. We are grateful to Paul Brenton, Nora Dihel, Abdel Hamid Mamdouh, JeanChristophe Maur and Sebastian Saez for inputs on the some of the elements of the proposal for a knowledge platform that is discussed in this paper, to Rolf Adlung and Patrick Messerlin for helpful comments and to participants in the Cordell Hull Institute meeting on services in Sydney, December 1112, 2010 and the Second Research Conference on Empirical Issues in International Trade and Finance, Indian Institute of Foreign Trade, New Delhi, December 1617, 2010. The views expressed are personal and should not be attributed to the World Bank. Introduction The efficiency of services industries has a powerful influence on economic growth. The productivity and competitiveness of both goods and services firms increasingly depends on access to lowcost and highquality producer services such as telecommunications, transport, finance and distribution. Telecommunications are crucial to the dissemination and diffusion of knowledge; transport services affect the cost of shipping goods and the movement of workers within and between countries; financial services influence the transformation of savings into investment and their allocation to the most productive uses; professional services such as accounting, legal, engineering and consulting services reduce transaction costs associated with the operation of financial markets and the enforcement of contracts, and are also channels through which process innovations are transmitted across firms. Retail and wholesale distribution services connect producers and consumers. Health and education services are key inputs into ­ and determinants of ­ the quality of human capital (Francois and Hoekman, 2010; Mattoo, Stern and Zannini, 2008). Services have assumed added significance from a broader global policy perspective in the aftermath of the 2008 financial crisis. There are strong political and economic imperatives that require adjustment in countries with major current account surpluses and deficits. Much of the policy debate on rebalancing revolves around fiscal, monetary and exchange rate policies. But structural reforms, including in services, also need to be part of the package. As services account for most nonlabor costs of production, action to improve the efficiency of services must be a major focus of policy in deficit countries, complementing policies to switch the pattern of expenditure and reduce net consumption. Expanding domestic consumption and investment in surplus countries must also involve a focus on services such as financial intermediation, insurance, and retail distribution and related logistics. Expanding the scope for international trade and investment in services can help generate the required structural changes. Recent studies such as Arnold et al. (2007, 2008, and 2010) and Fernandes and Paunov (2008) analyze the effects of allowing foreign providers greater access to services industries on services markets and on the productivity of manufacturing industries relying on services inputs. The results show a positive relationship between FDI in services and the performance of domestic firms in services and in manufacturing. Francois and Woerz (2008) find that increased import penetration in producer services has a positive effect on the skill and technology mix of exports in a panel of OECD countries, with greater openness in producer service sectors implying better export performance by skill and technology intensive industries. In last three decades, a variety of policy reforms have supported the rapid internationalization of production, consumption, trade and investment in services, reflecting increasing recognition that barriers to foreign entry and operations affect services performance. Given relative labor abundance in many developing countries, modern information technologies allow ever more cross border, "disembodied" trade in services to occur.1 But barriers to trade and investment remain prevalent. A recent World Bank research project has documented that services barriers in both 1 Determinants of trade in services include relative factor costs and trade costs (access to ­ and quality of ­ telecom networks) but also "technology transfer" costs. The latter can limit the feasibility of offshoring to locations that have the required institutions and capacity (Keller and Yeaple, 2009). 2 highincome and developing countries are higher than those for trade in goods, and that emerging markets have barriers that are on average substantially higher than OECD countries (Borchert, Gootiiz and Mattoo, 2010). Services are on the table in the WTO and in an increasing number of preferential trade agreements, bilateral and regional. However, the best offers in the Doha negotiations are on average still twice as restrictive as actual policy. It is in each country's interest to pursue reforms that increase the contestability of domestic services markets. The reciprocal exchange of liberalization commitments can help in achieving that outcome. There will be gains in terms of greater exports, but more important from an economywide perspective are the benefits that come from having more efficient domestic services industries. Limited progress in opening services markets through the WTO reflects diminished interest on the part of both governments and business. Governments have been reluctant to commit multilaterally because of fears that this may deprive them of the freedom to regulate, concerns that regulatory frameworks are not able to manage unrestricted entry and competition, and that the preconditions that need to be satisfied to allow greater trade ­ especially through movement of service providers ­ have not been put in place. Business interest has been limited because industrial country services markets are mostly open, except a few hardened pockets of protection (e.g. in transport and labor mobility), and developing countries are unilaterally liberalizing their markets. Many firms perceive the probability of re imposition of protection to be low, given growing twoway exchanges through both trade and FDI. There is also a sense of pessimism in the business community about whether the negotiations can deliver greater openness or even greater security of access in a way that is meaningful to their operations--in part because services regulation is not the focus of attention. In this paper we argue that new approaches are needed to enhance the prospects of negotiating meaningful commitments on services trade and improving regulatory policies affecting services markets. Market access negotiations need to be complemented by parallel processes that focus on improving services regulation and enforcement and putting in place the regulatory cooperation that may be needed to allow greater trade to occur. While services negotiations have addressed domestic regulation, the focus has been on ensuring that regulations do not become impediments to market access. The substance of regulation is left to the discretion of governments. Although we believe that it makes sense to limit trade agreements to the removal of discriminatory policies, in practice the "benign neglect" of domestic regulation implies that there are no assurances that liberalization will in fact be beneficial (increase national welfare) or that it will increase export opportunities for firms. Trade negotiators do not concern themselves about the adequacy of national regulation and enforcement institutions, or the need for international regulatory cooperation where there are regulatory externalities ­ e.g. in the exercise of prudential regulation in banking or procompetitive regulation in transport. What is needed is not to extend the General Agreement on Trade in Services (GATS) into the substance of domestic regulation but to put in place voluntary mechanisms in which the regulatory dimension of enhancing the performance of services industries is addressed cooperatively. Our premise is that all countries would participate more meaningfully in negotiations if they knew that the regulatory preconditions for benefiting from making binding commitments were in place. Once regulatory concerns and the cooperation needed to allow some types of trade to occur beneficially 3 has been identified, negotiations on market access could focus on the removal of explicit discriminatory policies affecting the movement of services providers ­ whether this involves legal entities (firms) through FDI or natural persons (through temporary crossborder movement ­ mode 4 of the GATS). The remainder of this paper is organized as follows. Section 1 briefly reviews where matters stand in the WTO and in major PTAs on services. Section 2 reviews a number of hypotheses that have been put forward to explain the limited traction to date. Section 3 makes a case that a key reason for slow progress at the WTO level in reducing barriers to trade and investment in services is that trade negotiations have inadequately addressed regulatory concerns. The outline of a possible mechanism to address such concerns and improve services regulation is sketched out. Section 4 discusses an approach to WTO services negotiations that could lead to a more ambitious outcome. Section 5 concludes. 1. Where are we on services trade negotiations? Negotiations on services were initiated in the GATT in 1986 with the launch of the Uruguay Round. They have also become prominent in the context of regional integration agreements, in particular the EU, but also in many more recently concluded preferential trade agreements. The WTO can potentially help to improve services performance by inducing countries to liberalize access to markets or to (pre)commit to doing so, thus increasing competition. The WTO disciplines on policies affecting trade in services, embodied in the GATS differ from those for trade in goods in important ways. Because trade in services can occur through different "modes" the GATS covers not just crossborder trade but also trade in services that takes place through the crossborder movement of consumers and suppliers (both temporary presence of individuals and longerterm via FDI). While the definition of trade is broader than in goods, the disciplines are more flexible. There is no general requirement to allow access to markets and to provide national treatment. Exceptions may be scheduled for national treatment on a sector specific basis. The same is true for the use of a limited set of specific policies that reduce market access. There is an extensive literature describing and analyzing the GATS--what follows limits attention to the sectoral coverage of the commitments made by WTO members. In an early assessment of the GATS, Hoekman (1996) concluded that the coverage of sectorspecific commitments on national treatment and market access was limited--in practice they implied that the GATS was an instrument through which a subset of policies that had already been implemented by Members on a unilateral basis were "lockedin". That is, the Uruguay Round did not deliver any actual liberalization. For many developing countries the coverage of specific commitments was well below 50 percent of all services and modes of supply. Adlung and Roy (2005) updated Hoekman's assessment and found very little progress in extending the coverage in the first 10 years of the GATS. An implication of the way commitments are made in the GATS is that they do not provide a full picture of applied policies, making it only possible to compare levels/coverage of commitments across countries. What is not possible is to infer how those commitments relate to actually applied policies. There is a requirement under Art. III:3 GATS to inform the Secretariat about all measures in scheduled sectors that significantly affect trade (there is no requirement to report on policies affecting 4 sectors that are not scheduled). Most members have not complied with this notification requirement. Thus, in the last 10 years Albania has notified some 120 measures under Art. III:3, as compared to just a handful by Canada, the EU and the US combined. Until very recently it was therefore not possible to assess the extent to which commitments implied a locking in of applied policies. A new database on services trade barriers compiled at the World Bank now allows such an assessment to be made. The database reveals that applied policies of 102 surveyed countries in a number of sectors are much more restrictive than in others, with mode 4 being the least open.2 Borchert, Gootiiz and Mattoo (2010) build on earlier work by Gootiiz and Mattoo (2009) and use the database to compare applied policies and the GATS commitments of the 93 WTO Members surveyed.3 They conclude that there is a very significant gap between applied and "bound" policies (the commitments) because most services liberalization around the world has been undertaken unilaterally. The policy commitments (bindings) made in the GATS during the Uruguay Round by the 93 WTO members surveyed are on average 2.3 times more restrictive than currently applied policies ­ i.e., countries could more than double their average levels of restrictiveness without violating their commitments. Figure 1 is a scatter diagram using this database where the location of each country reflects the overall restrictiveness of its services trade policies and its per capita income. A number of patterns can be seen. Highincome countries are clustered together at the bottomright, showing that they are quite open overall. There is much more variation in the restrictiveness of services policies for lowincome countries. Some of the poorest countries, like Ecuador, Ghana, Nigeria, Senegal, and Mongolia are remarkably open, while some of the most restrictive policies today are visible in the fast growing economies of Asia, including China, India, Indonesia, Malaysia, the Philippines and Thailand, as well as in the Middle East, including Egypt, Saudi Arabia and Tunisia. Figure 2 uses the same data but focuses on average levels of restrictiveness by sector and geographic region. It can be seen that there is a lot of variation across sectors, with some being subject to high discriminatory barriers in all regions (transport and professional services) whereas other sectors are much more restricted in some countries than others. Financial services and retail distribution tend to be much more open to foreign competition in high income countries than in many developing countries. 2 The database covers five sectors: financial services (banking and insurance), telecommunications, retail distribution, transportation, and professional services. In each sector, the survey covers the most relevant modes of supplying that service: cross border trade in services (mode 1) in financial, transportation and professional services; commercial presence or FDI (mode 3) in each services sector; and the presence of service supplying individuals (mode 4) in professional services. Results of the survey are summarized in an index of services trade restrictiveness (STRI). For each sector and mode of supply the openness of policy towards foreign suppliers is mapped on a 5point scale ranging from 0 (for no restrictions) to 1 (highly restricted), with three intermediate levels of restrictiveness (0.25. 0.50 and 0.75). Sector results are aggregated across modes of supply using weights that reflect judgments of the relative importance of the different modes for a sector. For example, mode 4 (temporary movement of suppliers) is important for professional services, but not for telecommunication, where mode 3 is the dominant mode of contesting a market. STRIs are aggregated using sector GDP shares as weights. 3 Nine of the countries in the database are not WTO members. 5 Figure 1: Restrictiveness of Services Trade Policies by GDP per capita, 2007 (102 countries) 100 Restrictiveness of services trade policy ETH 80 ZWE IND IRN 60 IDN PHL EGY THA MYS DRC PAN BHR CHN OMN SAU 40 NPL VNM BGD JOR LBN BW A LKA TUN LSO NAM DZA BLREN V MWI UGA MOZ TZAYEM ZAF FRA KEN MLI CRI MEX URY NGA CIV BOL CMR ITA UZB PRT KOR FINDNK BEL RWA MAR UKR TURRUS CHL 20 PAK ZMB KHM GHA HND SEN GTM ALB HUN CZE GRC ESP DEU AUT CAN MDG PRY PER MUS KGZ BRA COL KAZ ARG ROM NZL JPN SWE AUS GBR USA MNGARMDOM BGR LTU POL NLD IRL NIC GEO ECU TTO 0 0 10000 20000 30000 40000 GDP per capita, 2007 Note: GDP per capita, PPP (constant 2005 internat US$) Kuwait and Qatar excluded Source: Borchert, Gootiiz and Mattoo (2010). Figure 2: Restrictiveness of Services Trade Policies by Sector (102 countries) 80 Restrictiveness of services trade policy 20 40 0 60 SA R ME NA EA P Wo rld LA C A FR ECA OE CD Financia l Transportation Telec om Professional R etailing Source: Borchert, Gootiiz and Mattoo (2010). In 2000, a new round of WTO negotiations on services was launched, as called for by the Uruguay Round agreement. These negotiations were folded into the Doha round when it was launched in 2001. The talks aimed to increase the sectoral coverage of commitments. As offers made by countries are confidential it is difficult to accurately assess what additional market access has been offered. Borchert, Gootiiz and Mattoo (2010) use available public information on the offers by 62 WTO members that had been put on the table in the Doha Round to assess how much these would increase the coverage of the commitments of WTO members. Doha offers improve on GATS commitments, but at this stage the gap between offers and actual policy is still large: the best offers submitted so far improve on current GATS commitments by about 10 percent, but remain on 6 average twice as restrictive as actual policies (Figure 3) (if countries made no Doha offer, existing commitments are used). Figure 3: Existing commitments, Doha offers and actual policy by GDP per capita, 2010 (62 WTO Members that submitted Doha offers) 100 Services trade restrictiveness index IND BH R CRI U RY GTM TTO HN D GTM TU N LKA N TUC OL 80 KEN ID N EGY THA PAK TTO BH R IND N PHL NIC M AR N GAIDPRY C OL NIC LKA CH L MYSL CH BOL THA PAN MU S MU S PAN N GA 60 INDM AR DOM BRA TUR PAK DOM R TUR PE ID N PHL THA MEX FRA EGY C HN BRA PAN MYS MEX PR T KOR ITA BH R BEL DN K C HN BGR ARG OMN POL FIN 40 JOR TU N Z AF OMN KOR FRA SWE LKA Z AF ITA PE R PR T GRCESP DEU DN K BEL POL HU N JPN FIN AU T GBR CAN NLD IR L U SA JOR Z AF CRI CZE CAN KEN U RY MEX HU N CZE FRADEU AU T AUS N GA ITA JPN FIN SWE CH L LTU PR T GRCESP BEL DN K IR L GBR AU TNLD 20 PAKM AR ALB TUR LTU N HU KOR CZE NZL GRC DEU AUS CAN GTM HN D PE RBRA ARG JPN U SA PRY C OL ESP AUS SWE DOM POL GBR IR L NIC BOL MU S BGR LTU NZL NLD TTO 0 0 10000 20000 30000 40000 GDP per capita,PPP Restr ictive ness of GATS co mmitm ent Fitted values Restr ictive ness of DOH A Offers Fitted values Restr ictive ness of actual policy Fitted values STRI for 61 countries, excluding Qatar and 31 countries that did not submit offers Source: Borchert, Gootiiz and Mattoo (2010). Figure 4: Existing commitments, Doha offers and actual policy by region Binding gap, offer gap and applied policy for 93 countries 80 Services trade restrictiveness index 20 40 0 60 SAR GCC LAC AFR MENA EAP OECD ECA Actual Policy Offer gap (Doha Offer-Actual policy) Offer Improvement (Uruguay Round commitment-Doha Offer) Source: Borchert, Gootiiz and Mattoo (2010). As they stand today, Doha offers on services do not offer any liberalization of actual policy. Furthermore, two of the currently most protected areas in both industrial and developing countries, transport and professional services, are either not being negotiated at all or not with any degree of seriousness. In 2008 the chair of the Trade Negotiations Committee held a "signaling exercise" among a group of ministers, at the time that "modalities" in agriculture and nonagricultural market 7 access were being discussed. At the signaling exercise, participating ministers indicated that they might significantly improve their services offers. Clearly there is great scope to do so. Services in preferential trade agreements What about preferential trade agreements (PTAs)? Most PTAs negotiated since the early 1990s include provisions on services. Although many early PTAs did (and do) not go much beyond the GATS, more recent vintage agreements often have a higher level of ambition. An assessment by Roy, Marchetti and Lim (2007) concludes many of the trade agreements reported to the WTO since 2000 have a sectoral coverage that greatly exceeds the commitments the countries involved made in the GATS. This applies both to the existing GATS commitments and the offers put forward in the Doha round. In part this is because the approach to scheduling commitments differs from the GATS--a negative list approach is often used instead of a positive list. This also makes the agreements more transparent. Roy, Marchetti and Lim also conclude, however, that the substantive disciplines (rules) that are included in many of the agreements are similar to those in the GATS, i.e., the depth of the associated commitments often does not go much beyond what PTA members committed to under the WTO. Fink and Jansen (2009) provide another recent review the extent of services liberalization in PTAs negotiated in the 1990s and 2000s. They find that there is great variance across PTAs in terms of coverage of services and the depth of commitments, with more commitments made in sectors where countries have also made more extensive commitments in the GATS. Sensitive sectors such as health, transport and financial services as well as the movement of natural service suppliers (mode 4) tend to be subject to the fewest commitments. In areas where there are no WTO disciplines, there tend not to be rules in PTAs either--e.g., safeguards, subsidies, or procurement. The same is true as regards domestic regulation. Fink and Jansen (2009) argue that the rules of origin for services that are contained in many PTAs are mostly liberal, in that PTA benefits extend to nonmember firms that are established (have a commercial presence) and have substantial business operations in a PTA member. They conclude that such liberal rules of origin necessarily mean PTAs on services are multilateralizing in nature--i.e., building blocks for multilateralism. Dee and Sidorenko (2007) focus on Asian PTAs, asking whether recent agreements have tended to target regulatory restrictions that discriminate explicitly against foreigners, and if so, if these are the restrictions that matter most, in an economic sense. Focusing on banking, distribution, ports, professions, telecommunications, air passenger transport, and electricity generation, they conclude that the PTAs covering services focus on the less important services trade barriers. That is, they do not deal with the policies that result in the greatest markup of prices over costs (i.e., rents) ­ which tend to be regulatory policies that apply to all providers no matter their origin. To sum up, even though recent PTAs have wider sectoral coverage of services, they do not appear to have induced significant market opening. There are certainly exceptions ­ such as the liberalization of telecommunications in several Central American countries as a result of the CAFTA with the US ­ but there is also evidence that many PTAs have less than GATS commitments. Indeed, Adlung and Morrison (2010) conclude that there are GATSminus elements in about 80 percent of all PTAs. The benefits of PTAs relative to the GATS may be more in the area of improved cooperation and communication between officials on policies affecting the sectors and activities 8 that are covered by the PTAs. As discussed below, this can be a significant source of longerterm benefit in promoting improvements in regulation and identifying issues of concern and potential solutions. 2. Why such limited traction? In the case of goods trade, there is an extensive literature that identifies several possible motivations for governments to engage in trade negotiations. This includes first and foremost the terms of trade (market access) rationale: countries negotiate away the negative termsoftrade externalities that are created by the imposition of trade restrictions. Trade agreements can also potentially help governments implement reforms that enhance the contestability of services markets but that are opposed by politically powerful vested interests. This is because international trade agreements offer a way for breaking domestic deadlocks by mobilizing export groups to support reform. Another strand of economic literature argues that trade agreements offer a commitment mechanism to governments. By committing to certain rules that bind policies, i.e., "policy reform anchors," government may use international commitments to make domestic reform programs more credible. In principle these rationales should carry over to services in that groups that benefit from better access to export markets are induced to throw their weight behind import liberalization.4 The puzzle is that in practice they do not appear to be as strong as has historically been the case for trade in goods. In a number of papers ­ Hoekman and Messerlin (2000), Mattoo (2005), Hoekman, Mattoo and Sapir (2007) and Hoekman (2008)--we explore a number of hypotheses why the GATS may not be as effective as the GATT in helping countries to make commitments to bind policies and reduce levels of protection. One potential explanation is that there is less need in the services context for traditional reciprocitydriven market access negotiations. Insofar as inefficient service industries will generate costs for all downstream sectors, there is likely to be more pressure for unilateral reform than is the case for a tariff that protects a specific goods industry (as these have fewer economywide repercussions). In practice most reforms that have been implemented by both developed and developing countries have been autonomous. While this may help explain the significant reform in services since the late 1980s in most countries, barriers to trade and investment continue to characterize many service industries in both developed and developing countries. The difficulties experienced in the EU Single Market context in integrating European services markets and the burden imposed on firms in developing countries as a result of highcost services illustrates that this explanation has limited validity. Another potential reason why reciprocity may be less powerful in supporting reforms is that services exporters often face contrasting conditions of access: markets that are already open (and required no effort by firms to open) and markets that are almost irremediably closed. Thus, cross 4 The strategy pursued by China in the context of its accession to the WTO illustrates how trade negotiations can be used to good national advantage, with China committing to eliminate most restrictions on foreign entry and ownership, as well as most forms of discrimination against foreign firms. 9 border trade in services (e.g., business process outsourcing) or trade in tourism ­ two activities where many developing countries are net exporters ­ are generally not affected by restrictive policies. Conversely a large set of developing countries confront particularly high barriers for the one mode that is of export relevance to them ­ mode 4 ­ which in practice is for all intents and purposes not on the table in the WTO. A consequence is that the standard political economy of trade negotiation may break down: domestic opposition to reform and liberalization by incumbents cannot be counterbalanced by export interests seeking better access to foreign services markets. The prospective additional profits associated with better access to foreign markets for exporting firms may be much smaller than the rents/excess profits that are captured by sheltered incumbents in the countries concerned. Yet another reciprocityrelated argument is that reciprocity is less powerful in services because a policy reform that is made at the request of a trading partner is often automatically going to be of benefit to all other countries. This is because services policies often do not differentiate between firms operating on a market in the way that goods trade policies differentiate between imports from different sources. As noted, even the rules of origin in services agreements tend to be liberal (all firms based in and conducting substantial business operations in participating countries are usually eligible, not just those owned and/or controlled by citizens). An implication is that if countries engage in liberalization through PTAs, this may also open markets to providers from the rest of the world.5 In addition to PTAs, the existence of bilateral investment treaties (BITs) may play a role in diluting interest in trade negotiations. Protection of investment under BITs is more effective than what is available under the GATS as a result of access to investorstate arbitration and the possibility of monetary compensation for damages suffered as a result of government policies that violate the agreement. Many BITs that the US has negotiated provide for national treatment on a pre establishment basis across all sectors, as compared to the GATS where national treatment is generally only partial, depending on the coverage of a country's specific commitments. There have been some 120 servicerelated arbitrations under BITs since 1995, as compared to just a handful of WTO disputes (Adlung and Molinuevo, 2008). Probably one of the most important reasons for lack of negotiating traction is that governments and citizens may be concerned that liberalization (defined as nondiscrimination between foreign and domestic firms in the application of national policies) needs to be complemented by effective regulatory standards and implementing bodies. Regulation in services is pervasive and is driven by both efficiency and equity concerns (Mattoo and Sauvé, 2003; Mattoo et al., 2008). The characteristics of many services give rise to market failures. For example, the existence of natural monopoly or oligopoly is a feature of "infrastructure services" that require specialized distribution networks: roads and railways, airports, or cables and satellites for telecommunications. Regulation of the owners/operators of the networks can then enhance efficiency by preventing prohibitive 5 Note that this can explain why PTAs may not go much further than what countries have already decided to do unilaterally. Alternatively it can explain why business may be less interested in the WTO process: if they can use PTAs to open markets, this is effectively the same as multilateral reform. Thus, the extent to which PTAs do result in more open markets is a very important empirical question that should be the focus of much more research. 10 charges for access or interconnection to their established networks. Problems of asymmetric information are frequent in the services context. Buyers (consumers) confront serious hurdles in assessing the quality of service providers ­ e.g., the competence of professionals such as doctors and lawyers, the safety of transport services, or the soundness of banks and insurance companies. When such information is costly to obtain and disseminate and consumers have similar preferences about the relevant attributes of the service supplier, the regulation of entry and operations in a sector can increase welfare. In addition to efficiency justifications for regulation, governments may regulate to achieve equity objectives ­ e.g., ensuring access to services for disadvantaged regions, communities or households. Instruments to pursue equity objectives may rely on command and control ­ e.g., requiring a minimum number of poor households to be served, establishment of certain number of facilities in specific locations ­ or on the price mechanism. Examples of the latter are universal access funds that are competitively allocated to providers and used to cover the additional costs of service delivery.6 The GATS does not impose constraints on domestic regulation beyond its nondiscrimination and transparency provisions. It also does nothing, however, to help governments determine whether they have adequate national regulation in place and whether there is a downside risk associated with specific commitments.7 An example is a commitment to allow for foreign banks to open branches. Branch banking implies that the capital adequacy standards of the home country apply and that there may not be any restrictions on the ability of the parent bank to transfer capital out of the market it establishes branches in. If a government is concerned about this it could limit access of foreign banks to subsidiaries that must satisfy local capital adequacy requirements. But this presupposes that there is effective regulation and an effective regulator that considers the potential effects of different forms of liberalization. Another example is when there are significant rents associated with a certain policies. Regulatory agencies may argue that it is better that these accrue to domestic agents than to foreign firms, even if the latter are more efficient providers of services. The name of the game in trade negotiations is market access: this may easily result in a transfer of rents from local firms to foreign ones if the regulatory regime is not one that ensures markets are contestable. The implication is that broader regulatory reform is in many cases needed to ensure that welfare increases after liberalization. In general, improved prudential and procompetitive regulation will be necessary to deliver the full benefits of liberalization in sectors such as financial services; basic telecommunications and other networkbased services. Thus, attention should focus on strengthening and maintaining a robust capacity to identify, understand and design the domestic regulatory reforms that are needed to enhance the efficiency of services sectors. Indeed, a case can be made that the weight of empirical 6 Kenny and Keremane (2007) discuss the example of telecommunications, where countries such as Chile, Peru and Uganda have put in place systems in which private providers compete (bid) for performancebased subsidies that are conditional on providing services to targeted households/regions. This ensures that they reap some of the benefits of competition, while minimizing outlays for the government ­ the "reverse auction" process allows it to discover the true cost of service provision. 7 This is nothing new. The same applied to the GATT, which did nothing to engage contracting parties on what constituted a good trade policy (Hudec, 1987). However, this is a much more important issue in the services context as the adverse consequences of opening goods markets are limited to adjustment costs for the less efficient firms in the sectors concerned. 11 evidence implies that in many cases domestic regulatory reform may need to be put ahead of removing policies that discriminate against foreign firms.8 A final potential reason for limited traction is the disengagement of the business community noted in the Introduction ­ in sharp contrast to the Uruguay Round ­ because of their diminished incentives and low expectations. The interest in securing further liberalization has been diluted because industrial country services markets are mostly open, except a few hardened pockets of protection (e.g. in transport and labor mobility), and developing countries are unilaterally liberalizing their markets. The interest in securing further bindings has diminished because growing mutual interdependence­ the North's on outsourced services and the South's on foreign providers of transport, telecom, finance ­ is creating a selfenforcing equilibrium of openness with a reduced likelihood of policy reversal. It may also be that experience with the GATS has created a sense of pessimism in the business community about whether the negotiations can deliver greater openness or even greater security of access. The decision by trade Ministers in the Doha Round to focus first on negotiating modalities for agriculture and nonagricultural market access and only then turn to services may also have played a role.9 Services industries could (and did) interpret this as implying that their interests were not a priority. Whatever the reasons, the disengagement based on low expectations has been selffulfilling ­ because little is expected, little effort is expended and therefore little is achieved. 3. Addressing regulatory concerns How can we break out of the lowlevel equilibrium trap of low expectations and limited engagement? First of all, taking action to help countries define and attain regulatory objectives will help promote additional liberalization of market access and the realization of gains from greater trade and investment in services. It is necessary to recognize that in the case of services, countries confront a triad of reform goals: (i) enhancing competition/contestability of markets; (ii) ensuring effective regulation to deal with market failures (efficiency); and (iii) attainment of social/noneconomic objectives (equity). There are three instruments to attain these goals. 1. Unilateral reform. This appears to be slowing down, with most reform having been implemented in the 1990s, leaving a `hard core' of restrictions that benefit a variety of vested interests. 2. Market access negotiations. As noted above, trade agreements can help deal with political economy constraints (vested interests), but the process has not been very effective because regulation in general has been neglected. The neglect has muddied both domestic reform, reflected in arguments that a country or sector is "not ready", and undermined market access negotiations. 8 See for example Mattoo and Payton (2007) for specific examples in a lowincome country. 9 I.e., concrete negotiations on requests and offers in services would commence only once there was clarity on the contours of what would be on the table in these other areas. 12 3. Regulatory cooperation. This has been ad hoc and partial both in terms of substance (what has been covered) and participants (often bilateral and/or issuespecific). Moreover, it has also tended to be disconnected from what is on the table in trade negotiations. A concerted, international effort to pursue the third instrument could help enhance the effectiveness of the first two instruments. Two types of cooperation on regulatory matters are needed in support of more open services markets. First, international assistance for national (or regional) regulatory reform and strengthening of the regulatory institutions needed to achieve efficiency and equity objectives (such as widening access to services). This is not simply a matter of allocating (more) donor funding. More important is to diagnose prevailing policies (or the lack of policies), how they are being implemented and what the effects/impacts are and then to build on this diagnosis to devise strategies for reform. This is a process that takes time and one that will benefit substantially from information on the approaches and experiences of other countries that have (had) similar challenges. There is not necessarily any regulatory "best practice" for a sector or cluster of activities--in many cases there will be many options and countries need to figure out what approach may be most appropriate to their circumstances and needs. One mechanism that could help in the process of diagnosis and identification of a "menu" of regulatory options is to develop a "services knowledge platform"-- discussed below.10 Once a reform path has been defined implementation could be assisted by external assistance from development partners as part of the multilateral "aid for trade" initiative or regional programs. Note that the proposed type of cooperation will only generate additional market access (the main goal of the WTO process) as an indirect benefit ­ it is not the primary objective. The second type of cooperation that is needed is between regulators and is explicitly focused on expanding market access opportunities: aiming to address regulatory externalities that impede greater trade in services. The types of externalities that may arise will differ depending on the service activity. More cooperation on prudential regulation may be a precondition for trade in financial services and in informationbased services to occur-- for example, regulators may need to converge on a set of regulatory or data protection standards and establish that such standards are enforced. Competition agencies may need to have assurances that procompetitive regulations apply in partner markets to ensure that gains from liberalization are not appropriated by international oligopolies. Particularly important is cooperation between regulatory agencies in host and source countries on mode 4--the experience of a number of successful bilateral labor agreements demonstrates that this is a precondition for arrangements that expand the circular flow of people. 1. Fora for learning: Building knowledge platforms There are two specific dimensions to the broad challenge of national and regional regulatory cooperation and services policy reform: (i) addressing knowledge gaps ­ increasing information on regulatory experiences and impacts and identifying alternative options/good practices; and (ii) 10 This general concept was proposed by Thalwitz and Nabli (2010). 13 identifying the impact of ­ and the options for dealing with ­ the political economy constraints that impede the implementation of welfare improving reforms. A number of analysts including Geza Feketekuty (2010), one of the "fathers" of the GATS, have suggested that the negotiating process (that is, efforts to improve market access opportunities) by itself needs to be complemented by other approaches.11 Feketekuty argues that a mechanism is needed to share experiences regarding services regulation and reform, to generate information on the substance of regulation and enforcement in different countries, on what works and why, and what did/does not. An important function of such a mechanism is to bring together sectoral regulators and experts with trade officials and specialists. The former often do not think about trade and the trade implications of sectoral regulation, but are the "owners" of the policies that affect trade opportunities. The negotiation literature stresses that negotiators need to learn about the preferences and interests of other parties, as well as their own, and this is a process that that takes time. Negotiations invariably involve a complex process of interaction between domestic groups that result in an understanding of negotiating objectives/priorities. Learning is critical when it comes to the substance of policy rules--officials and stakeholders need to understand what the implications are of a given proposed rule and how it will impact on the economy. Matters are already complex when negotiations revolve around traditional trade policies such as tariffs and quotas. But they are an order of magnitude more complicated when it comes to services given that there is almost invariably a regulatory dimension. Establishment of a "knowledge platform" ­ a forum that is aimed at fostering a substantive, evidence/analysisbased discussion of the impacts of sectorspecific regulatory policies ­ could help build a common understanding of where there are large potential gains from opening markets to greater competition, the preconditions for realizing such gains, and options to address possible negative distributional consequences of policy reforms. Generating information on the impact and experience with reform programs that were pursued in other countries could help governments both assess prevailing policies and institutions in their own nations, and identify policy reform options. More specifically, such a forum could fulfill a number of roles. First, a mechanism through which information is generated on current services activity, prices and trade flows and prevailing regulatory policies. Better information on services policies and performance would help facilitate broad based discussion on what priority sectors are and where the key regulatory problems lie. Second, enhance knowledge of regulatory experiences and impacts in other countries, in the process identifying alternative options/good practices through collection and sharing of information on the factors underlying successful efforts to expand trade in services and the complementary policies that can be used to address market failures and distributional 11 This has become a more general theme in the literature and commentary on the trading system, building on arguments by international relations scholars--e.g., Abbott and Snidal (2000), Chayes and Chayes (1995) and Reddy and Sabel (2003). See Hoekman (2005). 14 concerns. Information and experiences from a range of countries can help ensure that regulations and standards that are adopted reflect local conditions and capacities for effective implementation. Third, by bringing together representatives of a range of countries (officials, regulators, private services suppliers) governments can discuss and learn about alternative approaches that have been pursued in practice to address the political economy constraints that may impede regulatory reform and constrain efforts to reduce barriers against foreign providers of services. No forum currently exists that has a global scope, bringing together sectoral/regulatory expertise (experts) and officials responsible for trade, investment and competition policies from highincome, emerging and other developing countries. No existing institution has an obvious comparative advantage in playing this role. However, many organizations have expertise and experience that could contribute to a process of "learning to learn" from country experiences. For example, the World Bank Group has substantial sectoral expertise in a number of services (e.g., transport, telecoms, finance, health and education) and has regular countrylevel policy dialogue on regulatory issues with line ministries and agencies through its wide network of country offices and fieldbased staff. There are organizations and networks that have already established processes and mechanisms through which trade and other government officials may interact with regulators ­ prominent examples include the OECD and APEC. A knowledge platform to identify good practices in regulation and services policies must be broad based. In practice a consortium approach will be needed, in which a number of policy institutes, international organizations and networks of regulators from around the world combine to provide the needed knowledge resources, working with or through a central hub entity. It is important to tap into knowledge across the globe for a specific sector or issue, including both developing and developed countries. International sectoral organizations such as the ITU (for ICT/telecoms); the IMF/BIS, IASB, IOSCO, and the Berne Union (for financial sectorrelated standards/regulation), the IOM/ILO (for migration and crossborder movement of people); and networks of sectoral regulators and related institutions (such as the International Competition Network) could be the focal points for specific activities. The same applies to entities such as APEC and the OECD, UNCTAD and WTO secretariats and business associations such as the Coalition of Services Industries that exist in a number of countries. Specific modalities of operation of a knowledge platform on services policies will need to be identified. To be effective and relevant a platform must be countryfocused, demanddriven, and action oriented. At the same time, the basic premise is that there is much to be learned from the experiences of countries around the world.12 In practice knowledge platforms may best be designed on a regional basis, linked to regional integration initiatives and regional institutions (such as the regional development banks). First steps could be to undertake a "mapping exercise" to identify existing international networks of regulators (regional or global) and related epistemic 12 In its new Knowledge Strategy, the World Bank recognizes the importance of SouthSouth learning and more generally the scope and need for developing country policymakers to learn from experiences obtained in other countries, with knowledge platforms being one instrument through which to do so. 15 communities, and to develop pilots to test the concept, focusing either on a just a few priority sectors or clusters for a given region and on tools that can be used to assess regulatory needs and the status quo. Examples of the latter include regulatory audits and impact assessments. The various regional fora and platforms could be linked to each other ­ through websites; sharing of tools and databases; etc. ­ and be brought together annually to exchange information and results. 2. International regulatory cooperation More broadly, as noted above, international regulatory cooperation could be a powerful catalyst for market access liberalization. Facilitating regulatory cooperation could help deal with apprehensions about liberalization on all modes. For example, in financial services, confidence in cooperation by the home country regulator could lead to openness to both commercial presence and crossborder trade. Similarly, in international transport services, confidence in the enforcement of homecountry competition law may increase the willingness to liberalize in importing countries. We focus here on the presence of natural persons or mode 4. Progress in these negotiations has become a precondition for more meaningful developing country participation in the process of reciprocal market opening. And 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 of all, Members need to recognize that simply asserting that mode 4 is about trade in services and not about migration cannot dispel deeprooted fears raised by the entry of foreign providers. These fears have to be acknowledged and addressed. One way forward may be to take a more cooperative and less antagonistic approach to mode 4, drawing upon the experience of a few relatively successful bilateral and regional agreements. The inclusion of labor mobility in the framework of a multilateral 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 mechanisms (e.g. APEC) that have facilitated mobility of the skilled, and bilateral labor agreements (e.g. between Spain and Ecuador, Canada and the Caribbean, Germany and Eastern Europe) that have to a limited extent improved access for the unskilled. Source country obligations include premovement 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. But provision of this "service" will require both action by regulatory authorities in source countries and confidence in the host country that the various preconditions are being implemented. Aidforservices trade to support implementation An effective knowledge platform will identify specific areas for policy reform and ways in which regulatory institutions and enforcement capacity need to be strengthened for liberalization to be beneficial. As mentioned, the main focus of a platform or forum is not services liberalization per se, but to focus on where it is necessary/desirable to bolster and improve national services regulation. A platform could however also be a vehicle to support the international regulatory cooperation that is needed to allow greater trade in services to occur ­ such as measures to provide a more effective 16 enabling framework for an expansion in mode 4trade. What matters of course at the end of the day is that actions are taken to implement the reforms that have been identified by governments as priorities. In many cases implementation will involve a need for investments: in training, in infrastructure, data and information systems, etc. The multilateral aid for trade initiative that was launched at the 2005 WTO ministerial in Hong Kong provides a mechanism through which lowincome developing country governments can obtain financial and technical assistance to implement regulatory improvements. To date most aid for trade assistance has focused on infrastructure investments and trade facilitation ­ not on regulation or service sector policies, but there is no reason why aid for trade would not be provided for services policy reforms is these are identified as priorities by developing country governments (Hoekman and Mattoo, 2007). Dealing with data gaps and improving transparency The types of cooperation discussed above should extend to ­ and benefit from ­ a concerted effort to enhance transparency of policies and address data lacunae. Statistics on services trade and investment flows are generally very limited, greatly constraining the scope for analysis of the impediments, determinants and impacts of such flows. For most countries services data are highly aggregated, often not available on an origindestination (bilateral) basis, and generally reported on a less timely basis than are data on trade in goods. Data on applied policies are not reported to or by the WTO on a comprehensive basis as a result of the positive list approach to scheduling in the GATS. As noted above, few WTO members have fully complied with notification requirements for scheduled sectors. In principle, this is a task that should be undertaken by and through the WTO. Ideally a final Doha package would include a commitment by WTO members to report on applied policies in scheduled sectors and give the WTO secretariat the mandate to collect and report data on all applied policies, whether scheduled or not. But independent of whether this can be achieved, better data are a precondition for assessing the impacts of policy and regulation, for monitoring and evaluation, as well as for effective negotiation. 4. A new approach to services negotiations It is possible to have two levels of ambition for services negotiations. One is to see international negotiations as a means to advance liberalization beyond unilateral levels. The other is simply to harvest unilateral liberalization. That latter view accepts that many of the constraints identified in the Section 2 cannot be overcome through a reciprocal exchange of policy commitments. We believe that the primary role of the WTO is to act as a lockin, harvesting mechanism. The reduction in uncertainty associated with binding applied levels of protection is very important even if often ignored by firms, analysts and policymakers. But we also believe that once legitimate regulatory concerns are addressed, it should be possible to take a bolder approach to the services negotiations, to break out of the lowlevel equilibrium of diminished business expectations and limited business engagement, and to negotiate a reduction in actual barriers to trade and investment. For the Doha Round to be brought to a successful conclusion, a final package that is balanced and commercially relevant will need to be agreed. Significant movement in the direction of liberalization 17 of applied policies will not be possible for many countries in the near term given the great diversity in regulation and regulatory capacity. In practice the process of learning, policy and regulatory reform, and strengthening of implementing bodies that is required will take many years. Therefore part of any Doha package should be a commitment to engage in a work program to create and support the operation of the types of mechanisms discussed in the previous section, and to accept that liberalization of services markets is a longerterm endeavor that will take time and is conditional on an appropriate regulatory and competition framework being in place. Such recognition will help the Doha Development Agenda live up to its name. But greater ambition will be needed on the market access dimension of services negotiations if the outcome of Doha is to be commercially relevant and thereby garner the needed political support by key industries to approve a negotiated deal in national parliaments. We have written elsewhere on the possible elements of a meaningful Doha deal on market access (Mattoo, 2005). To date offers have been very limited, and few signals have been given by WTO members that they are willing to use the GATS to open service markets to greater competition. A bolder approach involving a subset ("critical mass") of the major players on services could span three elements: A pledge not to impose any new restrictions, especially on cross border trade (mode 1) and FDI (mode 3), by inscribing binding language to this effect in the schedules of specific commitments in the GATS. Precommitments ­ inscribed again in each country's specific commitments ­ to implement reforms by a certain date (to be negotiated) in the future to liberalize trade in services, especially on FDI and in the transport sector;13 and Agreement to expand the scope for temporary movement of services suppliers with source country obligations, specified in a model schedule. A package spanning these three elements would offer benefits for established services exporters and to developing countries. The occupational shifts in importing countries and adjustment costs associated with increased trade in services are increasingly raising the specter of protectionism. This specter may never materialize because of the triumph of good sense. But it would still be wise to preempt protectionism, to ensure that any adjustment pressures created by international competition are dealt with through domestic assistance rather than by discrimination against foreign services providers. More important, fully lockingin applied policies would send a strong signal that governments are willing to use the GATS as a mechanism to enhance the predictability and transparency of policy. Recent research has demonstrated that there are substantial gains from services reform and that a large proportion of those gains are associated with FDI (Francois and Hoekman, 2010). Given that barriers to FDI in many services sectors are often high, this is an area where the WTO can have an impact. Precommitment to open a sector to FDI at a future date gives domestic incumbents and 13 International air and maritime transport is often not subject to GATS commitments. Commitments in this area are particularly important as countries cannot unilaterally reform international transport. The policies in other countries, industrial and developing, also limit competition. The de jure and/or de facto exclusion of transport would, therefore, be a serious omission from a "development round". 18 potential foreign entrants time to prepare and the government time to put in place regulatory improvements if these are deemed necessary. Gains cannot necessarily be realized by a mechanical opening up of services markets--there may be a need to put in place appropriate regulatory frameworks to ensure that FDI liberalization will not result in adverse consequences for national welfare (due, for example, to an abuse of market power made possible by entry barriers). A pre commitment approach towards opening sectors is likely to be a useful mechanism to address regulatory considerations explicitly: insofar as a country believes that opening should be delayed for reasons of regulatory weakness or reform needs, it would be forced to identify and articulate those needs and commit to addressing them within the negotiated time frame. The most stringent barriers to service trade are to the mobility of individual providers (mode 4). Such temporary movement offers arguably the neatest solution to the problem of how some forms of international migration are best managed, enabling the realization of mutual gains from trade while averting to a large extent social and political costs in host countries and brain drain from poor countries. Research has found that if industrial countries were to allow temporary access to foreign service providers equal to just 3 per cent of their labor force, the global gains would be over $150 billion ­ more than the gains from the complete liberalization of all trade in goods (Walmsley and Winters, 2006). These gains would be equally shared by the industrial and developing countries. Research on the effects of entry of skilled students and service providers demonstrates the powerful impetus this can have to innovation. As argued previously, a precondition for making progress in this area is regulatory cooperation. One possibility is that host countries commit under the GATS to allow access to any source country that fulfills certain prespecified conditions ­ along the lines of mutual recognition agreements in other areas. Even if these conditions were unilaterally specified and compliance determined unilaterally, it would still be a huge improvement over the arbitrariness and lack of transparency in existing visa schemes. Another issue is that 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 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 that existed between Germany and certain Eastern European countries before their accession to the EU, under which the quota on temporary migrants increased (decreased) by 5 percent for every one percentage point decrease (increase) in the level of unemployment. It would be desirable to consider GATS commitments along these lines, which allow necessary flexibility in a transparent, predictable and objectively verifiable manner. This would be a big improvement over the opaque economic needs tests that currently are prevalent in most GATS schedules. How can the cooperation and flexibility that is a key element of successful bilateral agreements be built into the GATS? One possibility may be to address the specificity of mode 4 by developing a model schedule (used previously to address the idiosyncrasies of maritime transport) which incorporates certain regulatory principles (used previously to address regulatory issues in telecommunications). A WTO Member could specify its market access commitments under Article XVI either in fixed absolute terms, as the US has done with respect to its specialty occupation program, or in terms of a variable quota that is linked to observable economic indicators such as the 19 rate of unemployment. A host country could list the requirements that source countries need to fulfill in order to qualify for the program such as credible premovement screening and selection, accepting and facilitating return, and commitments to combat illegal migration. Consistently with the GATS as a market access agreement, the source country obligations would not be subject to dispute settlement but nonfulfillment could result in a loss of access to markets. Denial of access could of course be challenge by the demonstration that conditions had been fulfilled. Eventually, when greater trust had been established between countries, it would be desirable to negotiate these conditions (and even establish a mechanism to certify their fulfillment) multilaterally rather than in an unequal, nontransparent and potentially labordiverting bilateral context.14 In these circumstances, the approach may well come to resemble plurilateral agreements such as already exist in APEC, where all participant countries would be required to fulfill certain conditions as both source countries (e.g. screening) and host countries (e.g. allowing admission). Such a reciprocal arrangement could allow freer mobility of service providers within the participant countries. Moving forward The Doha services negotiations initially followed the traditional GATT requestoffer approach, where each country requests concessions of trading partners and a process of bilateral talks ensue. Annex C of the Hong Kong ministerial declaration endorsed a shift away from bilateral requestoffer approaches towards a plurilateral approach. Rather than each country with an interest in expanding the coverage of GATS disciplines in a specific sector or mode of supply engaging in bilateral talks with all its major (potential) trading partners, the idea was to shift to a process in which the main demandeurs on a sector/mode would agree on a minimum "package" of commitments they would like to see, and then meet collectively to discuss these with the main "target" countries (Wolfe, 2007).15 An implication was that talks would revolve around a "critical mass" of countries, reducing 14 GATS Article VII protects recognitionrelated initiatives from potential challenges under MFN. It is potentially broad in scope in terms of the covered measures, and includes "criteria for authorization ... of services suppliers" and allows a Member to recognize "requirements met ... in a particular country." But it strikes an important balance. Any conditions that were specified and the determination of compliance could not be used to discriminate between countries or as a disguised restriction on trade. And it would be obliged to afford other WTO Members adequate opportunity to negotiate their accession to any such arrangement. 15 The plurilateral approach that was pursued in 2006 for services should be distinguished from the plurilateral agreements ("clubs") that are permitted under the WTO. Under the latter some WTO members agree to specific disciplines that apply only to signatories, and need not apply the resulting benefits to nonsignatories (see Lawrence, 2006). A plurilateral agreement can only be appended to the WTO on the basis of consensus (unanimity if there is recourse to voting--see Article X:9 WTO). Thus, the plurilateral option offers a mechanism for groups of WTO Members to agree to rules in a policy area that is not covered by the WTO or goes beyond existing disciplines as long as the Membership as a whole perceives this is not detrimental to their interests. The plurilateral approach in services is different. It is more akin to what has been called a critical mass approach to negotiating agreements: a group of countries that is large enough to internalize most of the benefits of cooperation on an issue agrees to certain commitments that then extend to all WTO members as a result of the MFN rule. Many observers have advocated greater use of critical mass approaches in the WTO (e.g., Martin and Messerlin, 2007; Gallagher and Stoler, 2009). 20 transactions costs while still ensuring that most of the gains from agreement would be internalized among those participating. Plurilateral talks were pursued among some 30 countries ­ mostly OECD members and the larger developing economies ­ starting in early 2006.16 Requests were tabled for legal; architecture/ engineering; computerrelated; postal/courier; telecommunications; audiovisual; construction; distribution; education; environmental; financial; maritime; air transport; and energy services, as well as crossborder trade; mode 3; mode 4; and MFN exemptions. Requests were generally less ambitious than those made in the earlier bilateral requestoffer process, which often had called for full market access and national treatment in sectors of export interest to the demandeur.17 Although progress was reportedly made as a result of the shift in negotiating modalities, with the countries that received requests actively responding/engaging in the process (Wolfe, 2007), the inability to move forward on agriculture and nonagricultural market access dominated the efforts of negotiators. A subsequent "signaling" exercise in mid 2008 that aimed at identifying what countries would be willing to offer revealed that ­ conditional on a deal on agriculture and NAMA modalities, which negotiators at the time thought was close ­ many countries were in a position to substantially improve offers. The Doha package that we have sketched out--including the mechanisms for regulatory reform and cooperation that were described in Section 3 and the measures to guarantee and further open access to services markets conditional on and linked to implementation of appropriate regulatory measures ­ is consistent with the plurilateral process. It will require a critical mass of the largest players (a "G25") to sign on to such a package for most of the benefits to be internalized. It is also consistent with the objectives regarding coverage of negotiations that were agreed by WTO Ministers in Annex C of the 2005 Hong Kong declaration. In the process of identifying the coverage of a Doha deal, consideration could at least be given to focusing the talks on clusters of service activities rather than only on the (sub) sectors that are defined in the GATS. The latter are based on a statistical classification that in practice is not always relevant to the interests of businesses that supply services. Given that most products are generated in and move through a supply chain, there will usually be a number of services that are critical for a firm or industry to be able to contest a market effectively. That is, firms may require a bundle of complementary services inputs to produce and sell their products. A shift towards negotiating on clusters of services activities, which together significantly impact on the profitability of an industry or group of industries (or, indeed, the feasibility of supplying a market), could help make the negotiating process more relevant to the needs of business. 16 Messerlin and van der Marel (2009) note that just eight countries (the EU, US, Brazil, Canada, China, India, Japan Korea and Mexico) account for over 80 percent of global production of services, close to the critical mass threshold that apliedin previous negotiations. 17 The decision to move to negotiations among subsets of countries and focus on a set of "minimum standards" for liberalizing commitments was premised on the idea that the dynamics would shift towards a Member having to justify its refusal to concede the threshold level rather than the focus of negotiations revolving around Members seeking to extract the maximum concessions from partners. 21 One reason for this is that a "cluster approach" can help identify regulatory policies that impose needless or excessive costs on firms and that may not be addressed by the disciplines imposed by Articles XVI and XVII of the GATS (market access and national treatment). However, the GATS permits Members to make "Additional Commitments" under Article XVIII. Insofar as a clusterbased negotiating process identifies additional disciplines that are important to make a set of commitments meaningful, these can be included through this provision. This in turn would help generate interest in and support for the process by the key stakeholders that are critical to provide the political support for any proposed Doha package. What are possible clusters? One that has already been identified by some WTO members are multi modal logistics and supply chain management, which would bundle distribution, storage/warehousing, and transport (including air cargo and maritime as well as road/rail) and related advisory services.18 This cluster would have the advantage of leveraging whatever the outcome will be of the trade facilitation talks, which are dealing with customs clearance related issues and transitrelated policies. Another possibility is a business process outsourcing cluster which would cover many of the services which are being electronically traded today, some of which do not even figure clearly in existing classifications (Mattoo and Wunsch, 2004). Other clusters could be defined on the basis of consultations with business associations. 5. Concluding remarks Services account for more than half of the GDP of almost all countries, and more than 75 percent in highincome countries. They are critical determinants of the competitiveness of firms in any country, no matter what its per capita income is. Poor countries will benefit from better, higher quality and cheaper services. All countries ­ whether rich or poor ­ have an interest in services as exportables as well. There is great scope to expand trade in services: the fact that crossborder trade in services is generally no more than 20 percent of all trade for a country is not just a result of the fact that many services can only be supplied through FDI. The low share of services is in part the direct result of high barriers to trade and investment. Trade agreements offer an instrument to reduce such barriers. WTO negotiations on trade in services have not made much progress. Part of reason for this is the decision by Ministers that the primary focus was to be on agriculture and NAMA, and that only once modalities were agreed for reducing barriers to trade in goods would attention center on services. This arguably was a mistake in that it resulted in much less attention being given to an area where the gains from reform are potentially greatest. But another part of the reason is that negotiating the liberalization of services is complicated. Adequate national regulation and international regulatory cooperation will often be necessary. A concerted effort is needed by governments to strengthen and improve service sector regulation and implementing institutions, as well as to cooperate with each other where there are significant regulatory externalities. Much of what remains to be done to remove barriers to trade in services in 18 See Communication from Australia, "Focusing plurilateral requestoffer negotiations: ideas on clustering," Council for Trade in Services Special Session, JOB/SER/33, November 2, 2010. 22 many developing countries will be conditional on such regulatory improvements. An important element of any Doha package on services should therefore be agreement to create mechanisms to promote procompetitive domestic regulatory reform. The aid for trade initiative has mobilized greater awareness of the need to consider traderelated priorities in the allocation of government expenditures and assistance provided by the donor community. Knowledge platforms can act as an instrument to increase the focus on the services reform agenda in developing countries in the provision of aid for trade. They can also help prepare the ground for future commitments on services in trade agreements and support an expansion in mode 4 trade. Although comprehensive liberalization of service markets in all 153 WTO members in the Doha round is neither possible nor at this point in time desirable, the largest services economies (a "G 25") can and should go further. There is both great scope and urgency for WTO members to demonstrate that greater ambition on services liberalization is possible. A binding commitment to lockin applied policies at current levels of openness and agreement to precommit to a significant reduction in discrimination against foreign suppliers for a set of key services sectors/clusters would do much to change the perception that the WTO cannot deliver on services. At the same time the proposed critical mass approach will ensure that those countries with weak regulatory systems can decide for themselves whether and when they are ready to participate in making commitments. The G25 group should have the regulatory capacity to ensure that liberalization will not affect the attainment of regulatory objectives. However, they may also need to pursue domestic regulatory reforms before opening up some services sectors to foreign competition, and will need to strengthen regulatory cooperation to facilitate greater trade in some services, especially those involving the movement of natural persons. A precommitment approach will allow such pre conditions to be put in place and ensure that there is an agreed timetable to open markets to greater competition. Explicitly recognizing that services liberalization cannot ­ and should not be ­ divorced from services regulation will do much to help harness the potential that trade agreements have to expand services trade and investment. References Abbott, K. and D. Snidal, (2000). "Hard and Soft Law in International Governance," International Organization 53(3): 421­56. Adlung, R. and M. Molinuevo (2008). "Bilateralism in Services Trade: Is there Fire behind the (BIT) Smoke?," Journal of International Economic Law 11(2): 365409. Adlung, R. and P. Morrison (2010). "Less than the GATS: `Negative Preferences' in Regional Services Agreements," Journal of International Economic Law 13(4): 110343. Adlung, R. and M. Roy (2005). "Turning Hills into Mountains? Current Commitments under the General Agreement on Trade in Services and Prospects for Change," Journal of World Trade, 39: 116194. Arnold, J., B. Javorcik and A. Mattoo (2007). "The productivity effects of services liberalization: Evidence from the Czech Republic," World Bank Policy Research Working Paper 4109. Arnold, J., A. Mattoo and G. Narciso (2008). "Services Inputs and Firm Productivity in SubSaharan Africa: Evidence from FirmLevel Data," Journal of African Economies, 17(4): 57899. Arnold, J., B. Javorcik, M. Lipscomb, and A. Mattoo (2010), "Services Reform and Manufacturing Performance: Evidence from India", CEPR Discussion Paper 8011 (see also http://www.voxeu.org/index.php?q=node/5662). 23 Borchert, I. B. Gootiiz and A. Mattoo (2010). "Restrictions on Services Trade and FDI in Developing Countries," World Bank, mimeo. Chayes, A and A. H. Chayes (1995). The New Sovereignty: Compliance with International Regulatory Agreements. Boston: Harvard University Press. Dee, P. and A. Sidorenko (2006). "The Rise of Services Trade: Regional Initiatives and Challenges for the WTO," in C. Findlay and H. Soesastro (eds.), Reshaping the AsiaPacific Economic Order. London: Routledge. Fernandes, A and C. Paunov (2008). "FDI in Services and Manufacturing Productivity Growth: Evidence for Chile," World Bank Policy Research Working Paper 4730. Feketekuty, G. (2010). "Needed: A new approach to reduce regulatory barriers to trade," http://www.voxeu.org/index.php?q=node/5208. Fink, C. and M. Jansen (2009). "Services provisions in regional trade agreements: stumbling or building blocks for multilateral liberalization?," in R. Baldwin and P. Low (eds.), Multilateralizing Regionalism. Cambridge: Cambridge University Press. Francois, J. and B. Hoekman (2010), "Services Trade and Policy," Journal of Economic Literature, 48(3): 64292. Francois, J. and J. Woerz (2008). "Producer Services, Manufacturing Linkages, and Trade," Journal of Industry, Competition and Trade, 8(3): 199229. Gallagher, P. and A. Stoler (2009). "Critical Mass as an Alternative Framework for Multilateral Trade Negotiations," Global Governance 15(3): 37592. Gootiiz, S. and A. Mattoo (2009). "Services in Doha: What's on the Table?," Journal of World Trade, 43(5): 1013­30. Hoekman, B. (1996). "Assessing the General Agreement on Trade in Services," in W. Martin and L.A. Winters (eds.), The Uruguay Round and the Developing Countries. Cambridge: Cambridge University Press, 88124. Hoekman, B. (2005). "Operationalizing the Concept of Policy Space in the WTO: Beyond Special and Differential Treatment," Journal of International Economic Law 8(2): 40524. Hoekman, B. (2008). "The General Agreement on Trade in Services: Doomed to Fail? Does it Matter?" Journal of Industry, Competition and Trade, 8: 295318. Hoekman, B. and A. Mattoo (2007). "Regulatory Cooperation, Aid for Trade and the GATS," Pacific Economic Review, 12(4): 399418. Hoekman, B. and P. Messerlin (2000). "Liberalizing Trade in Services: Reciprocal Negotiations and Regulatory Reform," in P. Sauvé and R. Stern (eds.), Services 2000: New Directions in Services Trade Liberalization, Washington D.C.: Brookings Institution, 487508. Hoekman, B., A. Mattoo and A. Sapir (2007). "The Political Economy of Services Trade Liberalization: A Case for International Regulatory Cooperation?" Oxford Review of Economic Policy, 23(3): 367 91. Hudec, R. (1987). Developing Countries in the GATT Legal System. London: Trade Policy Research Centre. Reissued in 2010 by Cambridge University Press. Keller, Wolfgang, and Stephen Yeaple (2009) "Global Production and Trade in the Knowledge Economy", CEPR Discussion Paper 7175. Lawrence, R. (2006). "Rulemaking Amidst Growing Diversity: A `club of clubs' approach to WTO reform and new issue selection," Journal of International Economic Law, 9(4): 82335. Mattoo, A. (2005). "Services in a Development Round: Three Goals and Three Proposals," Journal of World Trade, 39 (December), 122338. Mattoo, A. and L. Payton (eds.), (2007). Services Trade and Development: The Experience of Zambia., Washington D.C.: Palgrave and World Bank 24 Mattoo, A. and P. Sauvé (eds.) (2003). Domestic Regulation and Service Trade Liberalization, Washington, D.C.: The World Bank and Oxford University Press. Mattoo, A. and S. Wunsch (2004), PreEmpting Protectionism in Services: The WTO and Outsourcing, Journal of International Economic Law, 7: 765800. Mattoo, A., R.M. Stern and G. Zannini (eds.) (2008). A Handbook on International Trade in Services. Oxford: Oxford University Press. Martin, W. and P. Messerlin (2007). "Why Is It So Difficult? Trade Liberalization Under the Doha Agenda," Oxford Review of Economic Policy. 23: 34766. Messerlin, P. and E. van der Marel (2009). "Leading with Services: The Dynamics of Transatlantic Negotiations on Services," Groupe d'Economie Mondiale Policy Brief, Paris: SciencesPo. Reddy, S. and C. Sabel (2003). "Learning to Learn: Undoing the Gordian Knot of Development Today," Columbia University Law and Economics Working Paper No. 308. Roy, M., J. Marchetti and H. Lim (2007). "Services Liberalization in the New Generation of Preferential Trade Agreements: How Much Further than the GATS?," World Trade Review, 6(2): 145593. Thalwitz, M. and M. Nabli (2010). "Knowledge Platforms: A Conceptual View," World Bank, mimeo. Wolfe, R. (2007). "Can the Trading System Be Governed? Institutional Implications of the WTO's Suspended Animation," Centre for International Governance Innovation Working Paper 30 (Waterloo, Ontario). 25