65157 POVERTY THE WORLD BANK REDUCTION AND ECONOMIC MANAGEMENT NETWORK (PREM) Economic Premise OCTOBER 2011 • Number 68 JUN 010 • Numbe 18 The WTO and the Doha Round: Walking on Two Legs Bernard Hoekman The Doha Round of the World Trade Organization (WTO) negotiations has been ongoing for 10 years, and given political cycles in major countries, there is not much hope for a rapid conclusion. The topics on the table are important, and in prin- ciple there is enough substance for all countries to gain from an agreement, but, unfortunately, too much emphasis has been placed on gains through market access alone. The Doha Round is about much more than market access. Concluding the talks arguably requires greater recognition of the value of trade policy disciplines that will be part of any agreement. The WTO is not just a market access negotiating forum; it is also a multilateral umbrella through which governments can agree on rules of the game for other trade-related policies. Given the slow progress of the Round, greater emphasis could be put on leveraging existing WTO bodies to enhance the transparency of nontariff measures, address regulatory concerns that impede liberalization of trade in services, and launch a dialogue on domestic economic policies that can create negative spillover effects for trading partners. The Doha Development Agenda (DDA) negotiations, under- WTO is not delivering on its “legislative� function—the devel- way for 10 years, are in a state of paralysis as a result of disagree- opment of new global rules of the game for national trade poli- ments between major players on the extent of new liberaliza- cies that generate negative spillovers. tion commitments, especially for nonagricultural products. A number of observers have called on policy makers to ac- Efforts to use the upcoming eighth WTO Ministerial Confer- knowledge failure, terminate the talks, and start a process of ence in December 2011 to partially “harvest� results in areas of defining a new negotiating agenda that includes issues of great- particular relevance to the least-developed countries (LDCs) er salience to businesses. Others call for a shift in negotiating and other low-income countries—such as duty-free, quota-free techniques and practices so as to prevent a small group of coun- access for LDCs and an agreement on trade facilitation—failed tries (or a group of small countries) from blocking agreement earlier this year. It has become increasingly clear that prospects among the largest trading nations. The utility of such recom- for successfully winding up the talks in the near future are dim. mendations is limited at best. The subjects that are on the table The deadlock is costly. Assessments of the market access di- in the DDA—agricultural trade policies, manufactured goods, mension of what has been negotiated to date suggest that the and services—will need to figure into any multilateral trade ne- DDA could generate a global welfare (real income) boost of gotiation. The problem that is holding up agreement is not some US$160 billion (Laborde, Martin, and van der Mensbrug- blocking behavior by small countries. The source of the dead- ghe 2011). This significantly underestimates the value of an lock that has prevailed since 2008 is disagreement among a agreement, because continued paralysis also means that the small number of large players on market access. This is not 1 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise something smaller countries can do much about. What is need- when potential new investment, competition, and procure- ed is a “critical mass� of the larger players to improve their offers ment disciplines were taken off the table. Since 2004, the nego- on market access—defined both as reductions in applied barri- tiations have centered primarily on a traditional market access ers to trade and locking-in policies through binding WTO com- and rules agenda (including disciplines on agricultural support mitments. policies). This agenda offers potential gains for all WTO mem- In what follows, this note argues that the Doha Round (and bers, both in terms of lower barriers on goods and services ex- the WTO more generally) should not be assessed primarily on ports and from a reduction in uncertainty regarding possible the basis of the extent to which agreements reduce applied lev- increases in levels of import protection—through greater tariff els of protection. The WTO is not just a marketplace in which bindings, reductions in the average level of bound tariffs (the countries exchange liberalization commitments; it is a vehicle so-called ceiling tariffs that governments commit not to ex- through which governments agree on rules of the game for ceed), and specific commitments for services.1 policies, and the institution through which implementation is Average tariff levels today are much lower than just a decade monitored and negotiated rules and commitments are en- ago, and far below the averages that prevailed in the 1980s. forced. These rule-setting and enforcement dimensions of the Quantitative import restrictions have largely disappeared. The WTO are very important for firms engaged in trade because last (2008) proposals under active discussion in the DDA they reduce uncertainty regarding the competition conditions would reduce the world average bound tariff for agricultural firms will confront when exporting or investing. Uncertainty products from 40 to 30 percent and from 8 to 5 percent for can be an important source of market entry and operating nonagricultural goods. Average applied farm tariffs faced by de- costs, and result in less investment and job creation. veloping country exporters would fall from 14.2 to 11.5 per- The political cycle in several major countries (such as China cent, and those on their exports of manufactures from 2.9 to and the United States) make it unlikely that the negotiations 2.1 percent. The reductions in applied tariffs are beneficial to will be concluded before the end of 2013. This creates an op- exporters and consumers, but do not appear to add up to a lot— portunity for WTO members to identify a forward-looking after all, if DDA only generates less than a 1 percentage point process and launch a work program to discuss policy matters cut in the average tariff on manufactures, this clearly will not do that are not part of the DDA. Much has changed during the much to lower prices of the goods concerned or enhance the decade of the Doha Round. The sustained high economic ability of exporters to compete in foreign markets. growth rates in large emerging markets—most notably China— Much criticism has been based on the results of global simu- have made these countries much more important as markets lation models that suggest the net real income gains from any and sources of competition. The world has moved from a situa- politically feasible DDA outcome are likely to be small in the tion characterized by low food prices to one where prices are aggregate: as mentioned above, what was on the table in 2008 expected to remain substantially higher on average than they would generate “only� US$160 billion in additional income as have been during recent decades, as well as more volatile. Great- a result of lower trade barriers. This is not insignificant and er demand for food and natural resources could potentially compares well to what was achieved in previous rounds (Mar- bring on a more activist use of trade-related policies that have tin and Messerlin 2007). Whether one regards this number as negative pecuniary spillovers on trading partners. All of these significant or not, this numerical lens misconstrues a critical developments call for multilateral cooperation to determine function of WTO negotiations. These negotiations are not pri- rules of the game for food, natural resource and climate-related marily about reducing applied levels of protection, but center trade policies, and to strengthen the monitoring and transpar- on establishing trade policy rules and reducing uncertainty ency-related activities of the WTO. through a “lock-in� of policies and binding of tariff rates, either Launching a discussion of some of these issues in working at, or much closer to, applied levels. The benefits of this dimen- groups under the auspices of existing WTO committees would sion of WTO negotiations are ignored in models simply be- ensure that time is not lost while market access negotiations in cause economists cannot quantitatively assess these features.2 the Doha Round continue. The results of the deliberations The quantitative analyses also tend to underemphasize the could feed into an eventual Doha Round conclusion, but more fact that although tariffs are generally already low on average, realistically would aim to define an agreed upon set of follow- therefore limiting the aggregate effect of further reductions, on activities that would be pursued under WTO auspices. the formula-based negotiation modalities that have been devel- oped will effectively eliminate all tariff peaks in Organisation Moving Away from the “Market Access for Economic Co-operation and Development (OECD) coun- Metric� tries. The focus should therefore be on what happens to prod- Negotiators have been working for almost 10 years to define a ucts and sectors where tariffs are much higher than average—ag- negotiating set. The contours of this set were narrowed down ricultural products, textiles, and footwear. The same is true of over time, especially following the 2003 Cancun ministerial, agricultural support policies in OECD countries, which gener- 2 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise ate costs for consumers that are not large enough to concern subsidization, more emphasis is needed on explaining why them greatly, and negatively affect only a relatively small pro- such ceiling bindings are valuable. Agricultural protection and portion of economic agents in countries with a comparative subsidies in OECD countries have reduced the amount of food advantage in specific products. But for the affected groups— that is traded internationally and led to greater instability of such as farmers in Brazil or coastal fishermen in West Africa— world prices, with large negative spillover effects on developing what is on the table matters much more than what is inferred countries, whether exporters or importers. Disciplines on the from looking at the reduction in average tariffs. ability of governments to use import or export barriers to insu- Tariff bindings—and more generally negotiated disciplines late domestic markets, and hence make world markets thicker, and restrictions on the ability of governments to use certain would be a major source of welfare gain for developing coun- policies—reduce the uncertainty that is inherently associated tries (Martin and Anderson 2011). with engaging in international trade. Exporters confront more Leveraging Critical Mass uncertainty than do firms that operate only on their domestic market. National transactions and contracts can be enforced in Major stakeholders in the negotiations have stressed that more national courts; there are no borders where goods may be held market access concessions are needed for any Doha deal to be up in customs; there is no exchange rate risk to worry about; acceptable. The contours of any deal to do more to lower ap- and so forth. The fixed costs of getting goods into a foreign mar- plied barriers to trade and agricultural support need to be pur- ket are higher than those associated with domestic transac- sued by the large players on a critical mass basis. A key feature of tions. Anything that can lower the costs associated with export- critical mass agreements is that they need not involve all of the ing will both benefit existing exporters, and, more importantly, WTO membership. Instead, they imply agreement among the encourage new exporters. As foreign market entry costs fall, large players, with the associated benefits extended to all WTO more firms will be able to start exporting to new markets. The members (that is, those who are not part of a deal are allowed to associated expansion of exports along this so-called extensive “free ride�). Such an approach is nothing new for the WTO; in margin of trade will boost economic welfare and growth.3 practice, negotiations under the General Agreement on Tariffs Trade barriers may be prohibitive for a firm—a 50 percent and Trade (GATT) were always limited to those countries with tariff will be hard to overcome for most firms: negotiations that the greatest interest in a particular area or set of products, with result in lower tariffs matter. But if tariffs are already at 5 per- whatever was eventually agreed upon being extended to all cent—and the average applied tariff in many countries today is members as a result of the most favored nation (MFN) rule. often around or below that figure—variability/uncertainty in The threshold for agreement has tended to be around 90 per- the taxes and regulatory regimes that apply in a market can be cent, that is, some 90 percent of the trade involved in an area or of much greater concern to firms, and have a much greater ef- set of products needed to be between the participating coun- fect in impeding firms’ investment in export activity and pene- tries. A recent example of a critical mass agreement is the Infor- tration of new markets. This is a key reason why trade rules mation Technology Agreement, but tariff negotiations in earli- matter—even if the associated tariff and other policy commit- er GATT rounds also conform to this rule of thumb (Hoekman ments do no more than establish a ceiling on the level of dis- and Kostecki 2009). crimination that foreign products may confront in a given mar- To date, efforts to extend what is on the table on market ac- ket (Francois 2001; Handley and Limão 2011). cess have centered on sectoral approaches and proposals for Advocacy for the Doha Round (and the WTO more gener- trade in goods. Developed countries with already low average ally) needs to center more on the effects of the negotiated rules tariffs have argued that they have little left with which to nego- and policy disciplines. Selling or criticizing the Round on the tiate and induce emerging market countries to significantly basis of simulated estimates of real income gains or export lower their applied tariffs. This argument neglects the fact that growth resulting from the application of market access formu- concessions need not be limited to merchandise tariffs—they lae misses much of the story. The complete ban on agricultural can involve agricultural policies, the procedural rules affecting export subsidies would be a major step forward, for example, antidumping, and others. Other elements of the DDA offer sig- and cannot be quantified by estimating the impact of removing nificant scope for countries to expand the level of their com- extant subsidies—especially in a period where high prices have mitments; services is one such area. Services negotiations have greatly reduced the prevalence of their use. The ban is signifi- been sidelined for much of the post-2001 period, in part be- cant because if world prices fall in the future, the decline can- cause of a decision that services talks would commence in full not trigger an increase in export subsidies. Maximum allowed force only after a deal on agricultural and nonagricultural mar- levels of domestic agricultural support (subsidy ceilings) would ket access modalities was concluded. fall by 70 percent in the European Union and 60 percent in the Trade and investment in services is inhibited by myriad pol- United States, based on 2008 modalities. Again, instead of icy barriers that are more restrictive than those applying to stressing how much a deal will reduce the actual amount of trade in goods. Moreover, the extent to which applied policies 3 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise are locked in through binding WTO commitments is limited ticularly important from a welfare and growth perspective, be- (Gootiiz and Mattoo 2009). This matters for a number of rea- cause new varieties of goods and services account for a large sons, but most important is that the productivity, and thus part of the potential gains from liberalization. This is a feature competitiveness, of both goods and services firms depends on of the DDA that is rarely sufficiently emphasized in discussions access to low-cost and high-quality producer services such as of what is on the table. It is not only the effect of a given reduc- telecommunications, transport, finance, and distribution. Ser- tion in trade costs on existing trade flows that generate benefits, vices have assumed added significance in the aftermath of the more important is that agreements that lower trade costs will 2008 financial crisis. Because services account for most nonla- generate new trade. bor costs of production, action to improve the efficiency of ser- Defining the Future Path to Be Pursued vices must be a major policy focus in deficit countries, comple- menting policies to switch the pattern of expenditures and There are many systemically important issues that the DDA reduce net consumption. Expanding domestic consumption does not address. One reason for concluding the DDA as rap- and investment in surplus countries must also focus on servic- idly as possible is to be able to move on to address these other es—including social and health insurance services, pension significant issues. Indeed, a precondition for successful conclu- fund/asset management, and so forth. Expanding the scope for sion of the talks is likely to be agreement to engage in efforts to international trade and investment in services can help support cooperate in areas that are currently off the table. Some of these the required structural changes. issues are well known and have given rise to tensions and dis- The market access outcome of the DDA would be greatly putes, for example: biofuel subsidies and other types of “green� enhanced if a critical mass of the 15–20 or so largest WTO industrial policy measures; the possible use of carbon border members were to agree to bind current levels of openness. The adjustment as part of domestic climate change mitigation pro- associated reduction in uncertainty would be valuable to firms grams; export taxes on inputs to support domestic downstream and encourage greater investment (Hoekman and Mattoo industries; and export restrictions on food products as part of 2010). In addition, if these countries could negotiate a package an effort to insulate domestic markets. Other important and of liberalization commitments organized around clusters of currently off-the-table issues include discrimination in govern- services that are critical to business users and the smooth func- ment procurement; restrictions on foreign ownership of assets tioning of the global economy—such as logistics and supply (natural resources, real estate, enterprises in sensitive sectors); chain management—they could significantly enhance the rele- and allegations of anticompetitive behavior by multinationals vance of the DDA to global business. or state-owned enterprises. Another DDA area that is of great potential importance There is also an important agenda revolving around increas- from a market access perspective is trade facilitation. The costs ing the transparency of WTO member policies, including non- created by inefficient trade facilitation—both monetary, and, tariff measures and what members do in the context of prefer- more importantly, those resulting from delays and uncertainty ential trade agreements. The financial crisis revealed major gaps associated with clearance and regulatory compliance—can be in the available information on trade and investment policies. greater than the cost of paying tariffs on the affected imports. WTO notification requirements are often not satisfied on a Recent trade literature has documented the importance of timely basis, if at all. In some areas—trade finance, for example— trade costs as a determinant of whether firms export; that ex- there are no global databases on flows and prices. Very little is porters tend to be among the most productive firms; and that known about applied government procurement practices. the productivity effect of greater trade—deriving from both im- There are no comprehensive depositories of information on ports and exports—is an important driver of overall economic nontariff measures applied by WTO members. Concrete ac- growth. Most firms do not export, and those that do often sell tions to enhance both monitoring and analysis of trade and in- into only a few markets. Major factors explaining this include vestment policies and their effects—including the extent to lack of information, difficulties in obtaining credit, and the which countries use policies to discriminate in favor of national various costs associated with entering each new export market. firms and specific trading partners in the context of preferential Trade clearance and associated regulatory compliance require- trade agreements—will help the WTO fulfill its role of sustain- ments are elements of such market entry costs that impede ing an open and nondiscriminatory multilateral trading system. smaller firms from participating in export activities. Space constraints prevent a substantive discussion of the is- A trade facilitation agreement that reduces such costs will sues that WTO members arguably need to come to grips with. expand trade along what trade economists call the “intensive� The main point here is to recognize that there are various issues and “extensive� margin. The first of these refers to greater ex- that concern all WTO members and that call for multilateral ports of products that are already being shipped to a given mar- cooperation and agreement on the rules of the game that ket; the second describes new exports—either new markets or should be followed to maintain an open trading system. Agree- new products. An expansion along the extensive margin is par- ing on a process to address these matters, or, at the very least, to 4 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise define where/how they are best addressed, will provide assur- new subjects to the agenda may eventually help conclude the ances that issues of interest to all WTO members will be ad- Round, although a good case can be made that there is already dressed in the future. For example, systematic exchange rate more than enough on the table, if more is worked out on the undervaluation is a matter that some observers have argued services front and the appropriate weight is given to the value of needs to be addressed through WTO rules so that the WTO binding policies as opposed to only actual liberalization of ap- dispute settlement mechanism can be used to determine in- plied policies. The main point, however, is that if the talks con- stances where a member should be allowed to impose trade tinue to drag on for some time, the hiatus provides an opportu- barriers on imports originating in a country that has been nity to launch discussions under auspices of the relevant WTO found to engage in deliberate undervaluation. At the moment, committees on subjects such as those noted above. The oppor- the WTO does not provide this possibility, other than GATT tunity cost of waiting for the DDA to conclude is increased sub- Article XV,4 which delegates to the International Monetary stantially if it means delaying discussions on systemically im- Fund the task of determining whether a country is using ex- portant matters that are not currently on the DDA agenda and change rate intervention to “frustrate the intent of the provi- require cooperative solutions. sions of the GATT.� There are good conceptual and practical Conclusion reasons why the WTO does not include disciplines in this area, and compelling arguments why efforts to go down this path are Concluding the Doha Round is important in itself and for sus- likely to do much more harm than good in terms of sustaining taining the cooperation that has resulted in the current open multilateral cooperation.5 Whatever one’s views, however, clar- rules-based multilateral trade regime. Continued paralysis is ifying what is and what is not subject to rules and what are per- costly for the system because it prevents progress on the legisla- missible policies to promote investment in/production of trad- tive side—the negotiation and agreement on rules of the game ables is important in defining the boundaries of the WTO. in new as well as old areas that are important for global markets’ A major element of any future agenda is to further reduce operation. The WTO offers a multilateral umbrella under barriers to trade and investment in services and more generally which the major trading powers can agree on how to manage address the effect of regulatory policies in segmenting markets, and support the needed process of “global rebalancing.� Using including so-called nontariff measures. Given that countries the WTO to map out rules of the game is likely to be much may have legitimate concerns about the effects of liberalization more productive than the pursuit of unilateral policies to deal because of inadequate or the absence of regulation, the post- with policy externalities—both in terms of supporting the Doha Round action agenda should include developing mecha- needed structural transformation and in maintaining an open nisms through which WTO members can engage each other on trading system. regulatory policies affecting the contestability of markets. Abstracting from the need to put in place mechanisms to The WTO could do much more to offer effective mecha- support a process of building trust and understanding on how to nisms through which members can learn from each other on address the market-segmenting effects of domestic regulation, how to design and implement regulatory systems that support moving forward arguably does not require fundamental changes greater trade while attaining underlying regulatory objectives. to the WTO negotiating process. There has been much discus- Regulators, trade and economic affairs officials, the business sion in this regard about the Single Undertaking: the notion that community, and other stakeholders need to work together to nothing is agreed until everything is agreed. This is clearly a fac- assess current policies and options for improving regulation in tor that can slow down the process of getting to yes. One solution a specific area and determine how cooperation between regula- that is often proposed is a greater reliance on plurilateral agree- tors could facilitate more trade. Instituting a parallel process ments that bind only those countries willing to sign on, who that does not involve negotiations but that instead focuses on then may decide not to accord the benefits of what has been the substance of regulation (or the effects of a lack of appropri- agreed to nonsignatories. The main reason to consider plurilater- ate regulation) could help countries improve regulatory out- als is to avoid free riding—an issue that arises if some large coun- comes and facilitate an expansion in trade. Such processes tries do not want to join. However, this is not the source of the should extend to regular, systematic discussion and multilater- current deadlock—the problem is that some large countries want al scrutiny of preferential trade agreement implementation, more than other large countries are willing to offer.7 with the goal of identifying good practices that could be widely The Single Undertaking also has benefits—it enhances the adopted by WTO members. Creating such mechanisms for ex- legitimacy of any negotiated outcome. But it does imply an op- change and learning can help avoid a recurrence of the DDA portunity cost if agreements in specific areas must wait for an experience with the Singapore issues and help prepare the overall deal agreement. If such areas also generate little in the ground for future negotiations on services.6 way of reciprocity value—that is, the issue is not something that Moving forward to discuss new issues of common interest trading partners care much about—carving them out of the need not wait for the conclusion of the Doha Round. Adding Single Undertaking will not come at the cost of taking negotiat- 5 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise ing chips off the table that could have been used to link to other Article XV) dealing with exchange rates—it will be very diffi- issues. Alternatively, if the gains and costs of agreement in a spe- cult to objectively assess to what extent a country is undercut- cific area are balanced, a carve-out also comes at little cost from ting its trade policy commitments to liberalize access to its mar- a linkage perspective. The best example of such an issue of the kets and/or is subsidizing its exports. There are many other first type is duty-free, quota-free access for LDCs, because this objectives that may underpin an active exchange rate manage- is an action that does not entail any reciprocity by the LDCs. ment policy that have nothing to do with seeking to circumvent An example of the second possibility is trade facilitation. Since trade policy commitments. For a detailed analysis and discus- inefficient trade facilitation generates mostly socially wasteful sion, see Staiger and Sykes (2010). costs—as opposed to rents or government revenues—moving 6. For an elaboration of these arguments, see Hoekman and forward on trade facilitation is important from an economic Mattoo (2010). The Singapore issues refer to transparency in welfare perspective, and would come at low cost from a “link- government procurement, investment policy, competition pol- age foregone� perspective, because most of the benefits accrue icy, and trade facilitation. In 1997, the WTO established work- to the countries that take actions to improve facilitation. In- ing groups for each of these subjects to determine whether to deed, in an area such as trade facilitation, given that most of the launch negotiations in these areas. No agreement could be benefits accrue to the countries that pursue reforms, govern- reached in the cases of procurement, investment, and competi- ments should simply do so rather than incur the opportunity tion policies (see Hoekman and Kostecki [2009]). costs of waiting for a deal to be struck at the WTO. 7. Plurilateral agreements differ from critical mass agreements in that the latter apply on a MFN basis—that is, they permit free About the Author riding by those that are not part of the critical mass. Bernard Hoekman is Director of the International Trade Depart- References ment at the World Bank, Washington, DC. The views expressed are personal and should not be attributed to the World Bank. Broda, C., J. Greenfield, and D. Weinstein. 2010. “From Groundnuts to Glo- balization: A Structural Estimate of Trade and Growth.� NBER Working Notes Paper 12512. Francois, J. 2001. “Trade Policy Transparency and Investor Confidence: Some 1. See Martin and Mattoo (forthcoming) for a recent compre- Implications for an Effective Trade Policy Review Mechanism.� Review of International Economics 9 (2): 303–16. hensive assessment the state of play in the DDA. Gootiiz, B., and A. Mattoo. 2009. “Services in Doha: What’s on the Table?� 2. For a more extensive treatment of some of these arguments, Journal of World Trade 43 (5): 1013–30. see Hoekman, Martin, and Mattoo (2010). Handley, K., and N. Limão. 2011. “Trade and Investment under Policy Uncer- 3. Recent empirical research has shown that the indirect pro- tainty: Theory and Firm Evidence.� Mimeo, University of Maryland. Hoekman, B., and M. Kostecki. 2009. The Political Economy of the World ductivity effects associated with opening markets to new im- Trading System. Oxford University Press. ported varieties of goods account for 10 to 25 percent of the Hoekman, B., and A. Mattoo. 2010. “Services Trade Liberalization and Regula- typical country’s per capita income growth (Broda, Greenfield, tory Reform: Redesigning International Cooperation.� World Bank Policy and Weinstein 2010). On the export side of the equation, the Research Working Paper 5517, Washington, DC. Hoekman, B., W. Martin, and A. Mattoo. 2010. “Conclude Doha: It Matters!� magnitude of the productivity gains from reducing trade costs World Trade Review 9 (3): 505–30. come from the expansion of trade along the extensive margin, Laborde, D., W. Martin, and D. van der Mensbrugghe. 2011. “Implications of driven by a process of intraindustry adjustment in which the the Doha Market Access Proposals for Developing Countries.� World Bank Policy Research Working Paper 5697, Washington, DC. less productive firms exit and the more productive ones expand Martin, W., and K. Anderson. 2011. “Export Restrictions and Price Insulation (see Redding [2010] for a survey of the recent literature). during Commodity Price Booms.� Policy Research Working Paper 5645, 4. See http://www.wto.org/english/docs_e/legal_e/gatt47_01_e World Bank, Washington, DC. .htm. Martin, W., and A. Mattoo. Forthcoming. Unfinished Business? The WTO’s Doha Agenda. Washington, DC: CEPR and the World Bank. 5. The level of the exchange rate is not a policy instrument on Martin, W., and P. Messerlin. 2007. “Why Is It So Difficult? Trade Liberaliza- which a government can make specific commitments. It is en- tion under the Doha Agenda.� Oxford Review of Economic Policy 23 (3): dogenous, and will reflect a mix of fiscal and monetary policies. 347–66. Whether a government is engaging in deliberate undervalua- Redding, S. 2010. “Theories of Heterogeneous Firms and Trade.� Princeton University and CEPR, http://www.princeton.edu/~reddings/papers/het- tion is inherently a subjective exercise that requires judgment. firmstrade_080110.pdf. Even if this assessment if left to the International Monetary Staiger, R., and A. Sykes. 2010. “Currency Manipulation and World Trade.� Fund—as is required by the relevant WTO provision (GATT World Trade Review 9 (4): 583–627. The Economic Premise note series is intended to summarize good practices and key policy findings on topics related to economic policy. They are produced by the Poverty Reduc- tion and Economic Management (PREM) Network Vice-Presidency of the World Bank. The views expressed here are those of the authors and do not necessarily reflect those of the World Bank. The notes are available at: www.worldbank.org/economicpremise. 6 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise