WPS5361 Policy Research Working Paper 5361 Aid for Trade Building on Progress Today for Tomorrow's Future Bernard Hoekman John S. Wilson The World Bank Poverty Reduction and Economic Management Network International Trade Department & Development Research Group Trade and Integration Team July 2010 Policy Research Working Paper 5361 Abstract Since 2005, donors and development agencies have (through improved market access, investment flows, increased the overall value of aid for trade and put in and knowledge transfers); deeper engagement with the place several mechanisms to channel such aid and to private sector--a key source of information on what ensure that it targets national priorities. This paper works and what does not; a stronger focus on improving reviews recent trends in the allocation of aid for trade the "behind the border" policies that affect the efficiency and analyses of its effectiveness. It identifies a number of key services sectors and help determine firm-level of opportunities for concerted action to enhance the competitiveness; and a stronger focus on monitoring and impact of aid for trade initiatives, including greater evaluation of results. involvement by middle-income countries in the initiative This paper--a product of the International Trade Department, Poverty Reduction and Economic Management Network; and Trade and Integration Team, Development Research Group--is part of a larger effort to monitor and evaluate the effectiveness of aid for trade programs and to identify concrete actions to leverage multilateral cooperation in this area. 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 JSWilson@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 Aid for Trade: Building on Progress Today for Tomorrow's Future Bernard Hoekman and John S. Wilson* The World Bank Keywords: Aid for trade, official development assistance, multilateral cooperation, trade preferences, trade in services, regional integration JEL codes: F13, F35, O19, O24 * A first draft of this paper was presented at a G-20/World Bank conference in Busan, Korea, June 4, 2010. It draws on a work program supported by the DFID-World Bank Global Trade and Financial Architecture project and the Trade Costs and Facilitation project supported through the Multi-Donor Trust Fund on Trade. We are grateful to Shahrokh Fardoust, Arancha González Laya, Ann Harrison, Alan Winters, and Ernesto Zedillo for helpful comments on the conference draft, to Elisa Gamberoni and Richard Newfarmer for valuable inputs, and to Marco Antonio Martinez Del Angel, Alberto Portugal Perez and Benjamin J. Taylor for excellent research assistance. The views expressed are personal and should not be attributed to the World Bank. 0 Aid for Trade: Building on Progress Today for Tomorrow's Future Bernard Hoekman and John S. Wilson The World Bank Introduction Aid for trade is financial and technical assistance that facilitates the integration of developing countries into the global economy through initiatives that expand trade. By furthering economic growth and development, the benefits of aid for trade are shared by all trading nations. This includes not only the poor in least developed and other low-income countries, but also citizens in middle income countries and those in the most developed nations of the globe. Trade benefits all nations. Examples of aid for trade, as defined by the OECD1, include the financing of transportation and logistics infrastructure (infrastructure is the largest share of official development assistance in aid for trade), assistance to help firms conform to international product standards, capacity building in border management, and implementation of projects that connect rural producers to markets. Aid for trade also spans measures to assist workers, producers and communities in adjusting to changes in trade policies or the terms of trade (e.g., as a result of erosion of trade preference programs). The global initiative on aid for trade was launched at the 2005 G8 meeting in Gleneagles, Scotland.2 Since 2005, donors and multilateral development banks have increased the overall value of aid for trade and put in place several mechanisms to both channel such aid and to ensure that it reflects and addresses national priorities. The commitment to aid for trade has been re- iterated repeatedly by major donors at global aid for trade review meetings hosted by the WTO in 2007 and 2009 and in G-8 communiqués. The G-20 Summit in London in April 2009 included a statement of continued support for implementation of the commitments made on aid for trade 1 Although the general understanding of "aid for trade" is more or less universal, there are some technical differences across multilateral organizations and governments as to which types of development assistance qualify as such. The World Bank, for instance, takes a much narrower approach to identifying lending that is "trade- related" and does not include infrastructure lending under this umbrella unless it is part of a project that specifically addresses, either directly or indirectly, the cross-border movement of goods. On the other hand, the World Bank includes funds lent to countries on non-concessional terms as "trade-related lending," whereas the official OECD definition of "aid for trade" includes concessional lending only (as required by the terms of Official Development Assistance). For the purposes of this paper "aid for trade" refers to the official OECD definition, unless noted otherwise. See World Bank (2009) for a discussion of differences in OECD and World Bank approaches. 2 See http://en.g8russia.ru/docs/16.html. At the December 2005 WTO Ministerial in Hong Kong, a new WTO Aid for Trade Task Force was created to provide recommendations to the WTO Director-General on how to best "operationalize" aid for trade. The Ministerial Declaration also included explicit references to the importance of aid for trade to assist least developed countries (LDCs) "to build the supply-side capacity and trade-related infrastructure that they need to ... implement and benefit from WTO Agreements and more broadly to expand their trade. WTO "Hong Kong Ministerial Declaration," doc wt/MIN(05)/W/3/Rev.2, 18 December 2005. 2 by members.3 Delivering on these commitments is particularly important in the current global economic situation: aid for trade that results in improvements in productivity of firms and farmers in poor developing countries can both assist countries in recovering from the crisis and enhance longer-term growth and development prospects. By design, there is no central entity or global financial coordination mechanism that takes the lead on or is the focal point for delivering aid for trade.4 Instead, aid for trade is supplied through existing country-based allocation mechanisms by bilateral donors and international development agencies. The country-centric approach helps ensure that aid targets priorities identified by governments. However, the recipient country-cum-donor community-centric focus of the initiative also limits the potential impact of the enterprise in that more can be done to involve other actors in the delivery and assessment of aid for trade. Important vehicles used to raise awareness and monitor progress in delivery of aid for trade by donors are the Enhanced Integrated Framework (EIF) for trade-related technical assistance to the least developed countries (LDCs) and regional and global aid for trade reviews co-organized by the WTO.5 Since 2005, significant progress has been made by bilateral donors in implementing aid for trade commitments and by developing countries in identifying aid for trade priorities. However, there is still insufficient awareness and understanding in the broader development community of what the aid for trade initiative entails and how it works. There is also very limited data and analysis on the impact of aid for trade on the ground. This paper reviews recent trends in the delivery of aid for trade, its allocation and analyses of impact and effectiveness. It then identifies a number of opportunities for concerted action to enhance the impact of aid for trade initiatives. Developing mechanisms and concrete programs to transfer resources from middle income countries (investment, knowledge) as well as the private sector could do much to enhance the effectiveness of aid for trade in supporting trade and employment growth in low-income developing countries. 1. Why Aid for Trade Matters A key rationale for launching the aid for trade initiative was that firms in many developing countries may be unable to benefit from existing and prospective market access opportunities that the trading system or specific countries/regions offer ­ such as preferential (duty-free, quota- free) market access.6 Poor quality infrastructure and high trade and other operating and transactions costs in particular act to block many of the advantages of reduced barriers to trade 3 "The Global Plan for Recovery and Reform," http://www.londonsummit.gov.uk/resources/en/PDF/final- communique. 4 In contrast to other areas recently identifies as priorities for development assistance at a global level ­ such as the Global Agricultural and Food Security Program (GAFSP) established in 2009 with earmarked funding of $1 to $1.5 billion to scale-up agricultural assistance targeted to the food security of low income countries ­ donors decided there was no need for such a mechanism in the trade area. 5 The main objective of the EIF is to assist LDC governments in identifying trade projects that can be considered in the overall process of defining aid allocation priorities at the national level. There have been two global reviews to date, in 2007 and 2009. 6 See Prowse (2006) and Hoekman (2007) for a discussion of the genesis and rationales for the multilateral aid for trade initiative. 3 achieved in international and bilateral market access talks. A major feature of most aid for trade is that it is aimed at lowering costs and enhancing the productivity of firms in recipient countries. By focusing on boosting investment in infrastructure and complementary measures to create the preconditions for improved access to higher quality, lower cost public and private services, aid for trade can help countries to capture more of the benefits of existing market access opportunities. The need to deliver on aid for trade is heightened in the current economic environment. There are at least three reasons for this: Trade is a powerful mechanism to help countries overcome the shock of the crisis. Given the lack of progress in bringing the WTO Doha Development Round to closure, delivery on aid for trade would provide an important signal that the major players in the world economy recognize the importance of taking actions to expand trade. Aid for trade can help countries diversify into new markets and products--helping poor countries benefit from the emergence of a multi-growth pole world economy. Aid for trade, allocated effectively, can improve productivity in recipient countries by lowering costs and enhancing competitiveness, thereby enhancing growth prospects. Trade is a channel for poor countries to recover from the downturn. As economic activity and demand recovers from the financial crisis, consumers and enterprises in importing countries can be expected to be even more sensitive to prices of the goods and services they buy than before. Aid for trade that supports measures to improve the competitiveness of countries with weak trade capacity is therefore important. Moreover, as fiscal and monetary stimuli are gradually withdrawn, aid for trade can help maintain demand for goods and services and attract investment in tradable activities. Thus, aid for trade can provide a boost to developing countries during a period when they sorely need it. 7 It can also help reduce pressures for protectionism and increase support for trade reforms in developing countries, further expanding trade prospects by helping to keep markets open globally. Aid for trade can help increase diversification. Trade openness gives rise to risks as well as benefits. The recent crisis was exceptional in being truly global in scope: all countries were negatively affected. However, the crisis also illustrated once again that more diversified economies do better than those that rely on just a few products or markets as the source for their foreign exchange. Diversification can help reduce output volatility (Haddad, Lim and Saborowski, 2010). Many low-income countries are not well diversified ­ in part because of high trade and other costs that aid for trade can help reduce. Aid for trade can enhance productivity in low-income countries. There is a long-standing debate regarding developing countries' capacity to effectively absorb increased aid flows. By focusing on lowering trade and other transactions costs and improving the productivity of the 7 Even considering increased aid flows and commitments over the past several years, the World Bank estimates that developing countries confronted a financing shortfall of between US$270 billion and US$700 billion in 2009. External financing needs for developing countries are likely to increase because of the fallout of the crisis. 4 economy as a whole, allocating assistance to enhance trade capacity can help avoid negative competitiveness spillovers generated by aid inflows such as Dutch disease and pressures for real appreciation. As Reis and Farole (2010) note, the post-crisis "competitiveness policy framework" should tackle the priorities of aligning macroeconomic incentives (e.g.: trade barriers, real exchange rates and labor market policies), reducing at-the-border and behind-the- border trade costs, and overcoming government and market failures (e.g., shortages in trade finance, stimulating technology diffusion, improving product standards). Aid for trade can help low-income countries address this agenda ­ without targeting specific industries or potentially distorting policies to support product-specific investments. It can do so by improving trade policy coordination; trade facilitation, skill formation, trade-related infrastructure; and administrative procedures (Cali and te Velde, 2008). 2. Trends in Aid for Trade What is aid for trade? The OECD compiles statistics on official development assistance (ODA) in support of trade. These data distinguish between the following broad categories of support: (1) technical assistance for trade policy and regulations, (2) productive capacity building (including trade development), (3) trade-related infrastructure, and (4) trade related adjustment. Examples of support to trade policy and regulatory reform include projects at the country level to harmonize regulations to international norms. Capacity building and trade development include projects to assist in diversification of exports. Trade-related infrastructure projects include roads, ports, and telecommunications network investments. Trade adjustment assistance involves aid to help with costs associated with trade liberalization, including tariff reduction and preference erosion, for example. According to the data reported by the OECD, some 25 percent of ODA was directed toward aid for trade in 2008, and about 35 percent of aid that donors and governments allocated to particular sectors.8 Bilateral donors provided low-income countries, including LDCs, with about US$15.6 billion in aid for trade in 2008. This amounted to some 40 percent of the total US$39 billion in concessional aid for trade commitments in 2008. The LDCs received about one-fourth of aid for trade commitments. Donors provided about half of aid for trade commitments to middle income countries, mostly from bilateral sources. The supply of aid for trade has increased over the 2002/2005-2008 period by 21 percent in real terms. Low-income countries saw their share of total aid for trade increase from 44 to 54 percent, while 59 percent (US$ 4.7 billion) of the additional funds went to Sub-Saharan Africa (OECD and WTO, 2009). It is important to note that the OECD definition of aid for trade that is used here is a very broad measure of trade-related assistance and therefore overstates the overall magnitude of aid for trade. It includes all financing of infrastructure with the exception of water and sanitation projects. As infrastructure accounts for a large share of total ODA expenditures, this inflates the aggregate numbers for aid for trade. The reason for the use of a wide definition is 8 This "sectoral allocable aid" excludes funds for debt relief, administrative costs and budget support, as well as resources that are allocated to support trade finance. The G20 mobilized a collective US$250 billion effort to support trade finance during the crisis. Access to such finance is an important determinant of the costs of trade and the ability of exporters to operate. 5 that it is very difficult to distinguish to what extent specific forms of infrastructure support trade as opposed to non-tradable activities.9 Trends in aid for trade declined in absolute terms through 2002, after which it rose, reflecting renewed donor interest in growth and developments such as the launch of the Doha Development round (Figure 1). Even so, aid for trade has not kept pace with either total development assistance or that portion allocated to particular activities. Multilateral providers of assistance ­ aid that is channeled through IDA and the regional development banks ­ on average allocated a far higher proportion of their concessional aid for trade assistance to low-income countries than do bilateral donors. Some 93 percent of every aid for trade dollar goes to low-income countries ($6.6 billion of a total of $7.1 billion in assistance ­ Figure 2). However, bilateral donors provided 46 percent of their aid for trade to low-income countries. This highlights the importance of multilateral concessional lending for trade ­ and the urgency from an aid for trade perspective ­ of successfully completing the replenishment of the International Development Association's concessional fund for low-income countries (IDA-16).10 Figure 1: Aid for Trade, 1995-2008 (constant, 2008 USD Million) ODA Commitments (constant 2008 US$ million) 200000 150000 Total ODA excluding actions related to 100000 debt 50000 0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 ODA excluding actions related to debt Aid for trade Source: OECD CRS database. According to the OECD's most recent comprehensive report on aid for trade, Asia is the largest recipient of aid for trade. Aid to Africa has been closing in year-by-year in second place. In 2007, Asia received US$10.7 billion, over half of which went to Central and South Asia. Although the volume of aid for trade funds destined for Asia remained stable from 2002 to 2007, the region's share of total aid for trade funds dropped from 50 percent in the 2002-2005 period to 42 percent in 2007. This is due, in part, to Africa's increasing share of global aid for trade funds. The region received US$9.5 billion in 2007, representing 38 percent of total aid for trade funds, up from 30 percent in the baseline period. Flows to all other regions were significantly smaller. 9 It should also be noted that the OECD/WTO numbers exclude development assistance provided outside of the framework of the OECD Development Assistance Committee (DAC), and thus do not cover assistance provided by countries such as China. 10 Note: When references are made to CRS data, USD figures are in 2008 constant terms, whereas statistics attributed to OECD/WTO (2009) are in 2006 constant terms. 6 Latin America received US$2 billion and Oceania received US$1.6 billion in this period. Europe received the least, at US$1.2 billion, and was the only region to register a decrease in aid for trade funds from the baseline period to 2007 (OECD/WTO, 2009). Multilateral providers of assistance ­ aid that is channeled through IDA and the regional development banks ­ on average allocated a far higher proportion of their concessional aid for trade assistance to low-income countries than do bilateral donors. Some 93 percent of every aid for trade dollar goes to low-income countries (about $6.6 billion of a total of $7.1 billion in assistance). Bilateral donors gave 46 percent of their aid for trade to low-income countries. Figure 2: Aid for Trade by recipient group, bilateral vs. multilateral donors Multilateral Donors Bilateral Donors (excluding the EC) Not Not (including the EC) Allocated Allocated Middle by by Income, Income, Income, 4% 46% 3% Low Income, Low Middle 46% Income, Income, 93% 42% Note: Commitments in 2008 to low income (IDA eligible) countries. Source: OECD CRS database. The increased focus on the trade agenda by developing countries is also reflected in an expansion in trade-related activities and investments by the World Bank Group. A recent review of World Bank country assistance strategies (CASs) found that trade is now on the agenda of the majority (65 percent) of the Bank's clients (Strachan, 2009). These CASs identify trade as an important priority and present assistance programs with a clear focus on one or more of the following thematic areas: regional integration, export diversification, trade facilitation, and market access. This is translating into increased operational support, through ESW, lending, and in some cases, technical assistance to help countries achieve their medium term objectives. World Bank trade- related lending ­ using the World Bank's narrower definition vis-à-vis aid for trade ­ more than doubled since 2002, rising to some $1.4 billion in 2008 from about $550 million in 2002 (Figure 3). Concessional lending to the public sector has increased by more than half (World Bank, 2009). However, the trend in terms of number of projects and countries with trade operations has 7 been declining in recent years, illustrating that expanding aid for trade continues to require high- level attention by policymakers. Figure 3: Trends in World Bank Trade Lending, 2001­2009 Source: SAP/Business Warehouse. Notes: Trade components are defined by thematic codes assigned to IDA and IBRD projects. The increase in value of lending in 2009 comes from a $2.125 billion Western Europe-Western China International Transit Corridor Project. The rise in aid for trade has occurred against the backdrop of success in reducing import tariffs and removal of other traditional barriers to trade. This is true even considering the long stalled Doha negotiations at the WTO. As formal trade barriers have been eliminated for a significant portion of global trade, countries have focused on other impediments to trade flows ­ both through domestic and collective action. Global trade reform and capacity building is increasingly anchored in an agenda to minimize trade transaction costs to further leverage comparative and competitive advantages. This shift in the global trade agenda has been accompanied by a significant increase in aid for trade assistance from bilateral donors and multilateral institutions. As discussed further below, achieving better targeted ­ and effective ­ aid for trade would benefit from a less ad hoc multilateral framework for coordination. It is clear, however, that there is a large supply of aid for trade assistance, the majority of which is provided by multilateral institutions and G20 donor countries. The G-20 is well placed to lead in this regard. In 2007, of the top fifteen non-institutional donors of aid for trade ODA, eight are G20 members, including the European Communities (OECD and WTO, 2009). The G20 therefore has an opportunity to provide strong and visible global leadership, in partnership with multilateral institutions and developing countries, to shape the aid for trade agenda going forward. Commitments to sustain and grow aid for trade commitments at recent summits have been encouraging,11 but there is a need for a more direct and visible approach in ensuring concrete action plans on aid for trade to help drive the development agenda forward as global recovery continues. 11 See "Global plan for recovery and reform," communiqué issued at the close of the G20 London Summit, April 2, 2009. At: http://www.londonsummit.gov.uk/resources/en/PDF/final-communique. 8 3. Does Supply of Aid for Trade Match Demand? The distribution of aid for trade is as important as overall amounts. There are a number of different perspectives on the question of whether the supply of aid for trade aligns with the demand and need for aid. One approach is to focus on analysis of poverty reduction strategy papers (PRSPs) to evaluate whether and how countries are integrating trade policy and institutional reforms into development plans. A United Nations Development Program (UNDP) study that reviewed 72 PRSPs found that 85 percent included one or more components devoted to trade (Kosack, 2008). This marks a significant increase from previous analyses ­ a 2000 study found only about 25 percent of completed PRSPs had a section relating to trade. Moreover, 52 of the 72 PRSPs included in the latest UNDP study related trade policies to poverty profiles. This development, among other more specific differences across various iterations of PRSPs, suggests that countries are increasingly considering linkages between trade and poverty reduction. The same can be inferred from the survey of World Bank Country Assistance Strategies mentioned earlier. One of the first attempts to evaluate the balance between supply and demand based on empirical evidence and data was undertaken by Gamberoni and Newfarmer (2009). They find that supply of aid for trade has responded to demand in the sense that countries that are most in need of aid for trade ­ as measured by trade capacity and performance ­ tend to receive relatively more assistance. Subsequent analysis that builds on and extends the methodology developed by Gamberoni and Newfarmer has focused on the relative impact of hard versus soft infrastructure investments, aiming to obtain a better understanding where aid for trade funds may be best spent in advancing capacity building goals. For example, Portugal-Pérez and Wilson (2010a) construct indicators of trade capacity from a set of primary variables that measure the availability and quality of trade-related infrastructure and regulation (e.g., the fixed line network, quality and capacity of ports, airports, rail, and roads, governance and corruption, costs and time to clear trade consignments, various indicators of the business and investment climate, etc.). Using factor analysis, these variables are condensed into four specific factors that capture distinct features of the trade environment. Two of these indicators are related to the "hard dimension" of trade capacity ­ Information and Communications Technology (ICT) and Physical Infrastructure ­ and the other two are measures of the "soft" dimension of trade capacity: a Business Environment Trade Indicator and a Border Management and Customs Efficiency Indicator. Martinez and Wilson (2010) use these four factors to create a measure of the demand (need) for aid for trade. The authors regress actual supply of aid for trade against this measure and find that most of the countries that have the greatest need are close or above the predicted line, indicating an approximate match between supply and demand (Figure 4). Moreover, the results are consistent in the sense that countries with the lowest scores on the trade capacity indicator (associated with higher values of the Index), receive higher levels of aid for trade. However, it is also clear that there is a lot of variance around the trend, and that many countries are receiving less support than these various indicators of need suggest would be appropriate. 9 Figure 4: Matching Demand with Supply of Aid for Trade Potential Demand for AFT, 2005 Controlling for GDP per Capita, Rule of Law and Trade Facilitation Index AFT Disbursements as share of GDP MNG .03 KGZ GHA MWI .02 MOZ NIC VNM ETH GEO ZMB .01 MDA BIH MLI GMB LKA BEN UGA T ZA PHL AR M BOL PRY T HA KEN M DG TUN SLV NAM I DN AZE CM KAZ R BGD J OR PER HND MAR UKR ECU E MYS J CHNAM BRA IND H COL CRIGTM MUS MEX BWA ARG RV DZ A T UR PAK ZW 0 CHL ZAF PAN URY VEN -.01 1 1 .5 2 2.5 3 3.5 TF Index Average AFT Bottom Performers Top Performers Source: Authors Calculations based on indexes from W ilson & Portugal (2010), World Bank and ITC Source: Martinez and Wilson (2010). 4. Aid for Trade: Impacts and Effectiveness There is an extensive literature analyzing the relationship between aid and economic growth. The analytical methods employed in these studies and the results are subject to significant debate. The literature provides a mixed picture as to whether there is a positive relationship between aid and growth.12 There may be many reasons for this. It may be due to the type of aid delivered (for example, purely humanitarian vs. policy change driven) or reflect differences in absorptive capacity in developing countries. 13 One factor that can explain a lack of a positive relationship between aid and growth is aid-induced appreciation of real exchange rates--with aid inflows inducing Dutch disease.14 A comprehensive review of this literature is beyond the scope of this paper. It is useful, however, to outline, in brief, the complexities in analyzing and understanding 12 See Rajan and Subramanian (2005) for a survey and new assessment; Cali and te Velde (2009) for a synthesis of the extant literature. 13 See Radelet, Clemens, Bhavnani (2006), OECD (2006). 14 The effect is well known: aid flows may be used to finance expenditures of non-tradable goods and services, leading to a rise of their relative price with respect to tradable goods and thus, to a real appreciation of the exchange rate. This reduces the competitiveness of the exporting sector, as resources are transferred from the tradable to non- tradable sectors and drives up wages and other input costs. Estimates of whether aid induces a Dutch disease phenomenon can vary widely. Much depends on assumptions about the marginal productivity of additional aid and public expenditures, the complementarities between public and private capital, and the degree of flexibility of labor costs and other key resources. See e.g., Radelet, Clemens and Bhavnani (2006). 10 the relationships between aid, trade, and growth. Debate continues, in particular, on the causality between aid and trade. 15 Until the late 1990s a large share of ODA was tied to trade in the sense that procurement of goods and services financed by aid was tied to sourcing from the donor country. Any positive trade-aid relationship, therefore, could be due to policy decisions made in donor countries. Many researchers have indeed found strong links between foreign aid and donor exports.16 Causality could also run the other direction--from trade to aid--insofar as donors allocate aid to those countries that they have the strongest trade ties with.17 Analyses that test for the direction of causality generally conclude that it depends on the pair of donor and recipient countries.18 Whatever the precise channels, the results suggest a positive relationship between aid and trade. In light of the commitments and action to increase aid for trade funding, questions as to how aid for trade specifically helps to improve the trading performance of developing countries ­ and how effective taxpayer funding is in attaining aid for trade objectives ­ have gained increased prominence. This is especially true in a post-crisis environment characterized by a much tighter fiscal situation in all donor countries. Bilateral donors and international development agencies are actively engaged in efforts to go beyond simple monitoring of the flows and allocations of aid for trade to an assessment and/or analysis of the impact of aid for trade. Evaluation is critical to discovering ways to improve the effectiveness of development assistance ­ and aid for trade is no exception. Evaluation can occur at several levels: do countries in need get it (the question asked above), are programs taken as a whole effective in expanding output and reducing poverty (programmatic evaluations), are projects achieving their stated goals, say in expanding electric power (project evaluation), and are outcomes different than in comparable situations without the project or different than they would have been in the absence of project interventions (impact evaluation). Measuring the impact of aid for trade is challenging. In part this is because of data limitations. Many projects may not have information on defined baselines against which impacts can be assessed. Trade-related development projects often lag behind best practice in not being designed to allow rigorous ex post evaluation of impacts. A major challenge is that often standard impact evaluation methods cannot be applied to aid for trade because the assistance takes the form of general budget support. Frequently it will be difficult to disentangle the impact of aid for trade projects and programs on welfare, income, and equity (e.g., distributional effects). Much of the assessment of aid for trade to date has been at an aggregated level, focusing on whether trade performance of countries and indicators of trade capacity have improved. What is 15 Suwa-Eisenmann and Verdier (2007) provide a comprehensive review of this topic. 16 For example, Nilsson (1997) observes for trade between the EU and recipient countries that $1 of aid generated $2.6 of exports from donor to recipient for the period 1975 to 1992. Wagner finds that increasing aid to a country by 1 % increases the donor exports to the recipient by 1.33 %. Other researchers have explored additional links that may exist between the donor and recipient that may lead to additional trade, such as political or economic considerations (Lloyd et al., 2000). Nelson and Silva (2008) is a recent analysis that obtains much smaller number using a fixed effects gravity model estimation. 17 Morrissey, 1993; Osei et al., 2004. 18 Lloyd et al. (2000) as well as Arvin et al. (2000). 11 needed is more detailed analysis of the impact of specific aid for trade interventions on the ground, which in turn will depend on identifying new ways to support long-term investment in micro-level trade cost and outcome data. A recent OECD review of project evaluations for trade-related development assistance projects found that project documents often had insufficiently clear measurable objectives (OECD 2006). Quantitative baselines or benchmarks that would allow ex post assessments of the degree of improvement in specific measures of trade performance or trade capacity were frequently not included. This is an important finding in itself because it implies that donors and beneficiaries have to do a better job in identifying objectives. The OECD report concludes that, in half of the evaluations, trade-related assistance contributed to raising awareness of the importance of trade and knowledge of trade issues, while helping to strengthen country dialogues on trade policy. Major project weaknesses that were identified included inadequate needs assessments, weak project management and governance, a lack of integration into an overall trade strategy or development program, weak links to poverty reduction, inadequate donor coordination, and inadequate communication to, and expertise in, field missions. A 2006 evaluation by the Independent Evaluation Group of World Bank trade projects and programs found that in general trade-related adjustment loans performed better than other adjustment loans (86 percent satisfactory versus 78 percent for non trade loans), while trade- related investment loans performed slightly worse (69 percent versus 72 percent satisfactory) (Independent Evaluation Group, 2006). A follow up review found that in 2007, more than 85 percent of projects were evaluated to have had moderately satisfactory, satisfactory or highly satisfactory outcomes. These generally performed better than non-aid for trade projects (World Bank, 2009). More programmatic forms of evaluation use cross-country data on the effects of increasing aid for trade in specific areas. Given that aid for trade is targeted at specific types of activities and interventions a more precise identification strategy can be employed to assess the magnitude of effects and direction of causality. Cali and te Velde (2008, 2009) analyze the effects of various categories of aid for trade and find that aid for trade facilitation reduces the cost of trading. A US$1 million increase in aid for trade facilitation projects is associated with a 6% reduction in the cost of packing goods, loading them into a container and transporting the consignment to the port of departure and loading them on a vessel or truck. They also demonstrate that aid for trade allocated to infrastructure results in an expansion of exports, especially in the mining and manufacturing sectors, with effects being the greatest in Africa where infrastructure is weak. However, aid for trade that is allocated to productive capacity (as opposed to infrastructure or facilitation) has no statistically significant effect on exports. Helble, Mann, and Wilson (2009) undertake a similar analysis, focusing on the effects of trade development assistance (productive capacity building), trade policy assistance, and infrastructure assistance on bilateral trade flows, using a gravity regression framework. The findings suggest there are very high marginal returns to aid for trade targeted at trade policy and regulatory reform projects: $1 dollar of aid for trade targeted at trade policy and regulatory reform is estimated to 12 increase trade flows by some $700. While aid allocated in this area will encounter diminishing returns, this type of analysis suggests that the rate of return to aid for trade can be very high. As noted, impact evaluation is still an incipient endeavor in the aid for trade field ­ work of this type is far more limited than in health and other fields of development assistance. A recent example is Brenton and von Uexkull (2009), who undertook an impact evaluation for export development projects targeted on specific export products. They found that such projects: (a) have coincided with, or predated, stronger export performance in the targeted commodities; (b) have had a greater impact on export growth for products with initially high export levels than on those with low export levels (although this may be because technical assistance is directed towards industries that are already set to take off); and (c) were likely to be more successful if they addressed specific market failures or policy shortcomings in activities in which the country had a long-run capacity for global competitiveness (as was the case in Rwanda's donor- supported strategy to move into the high quality, specialty end of the coffee market). They conclude that, done well, export development programs can succeed: cut flowers had been a growing export industry in Uganda for a decade when an export development program was started in 2003. Following the program, export value almost tripled within one year. Although other Ugandan exports also rose strongly at this time, cut flowers significantly increased their export share. In the case of Mongolia, a traditional exporter of wool products, exports had declined and lost share in the export portfolio in the late 1990s and early 2000s. After the implementation of an Export Development program in 2003, exports of wool products entered a steady growth path, outperforming overall export growth in 2005.19 Taken together, the available literature tends to validate central Paris Principles: aid for trade can be effective, provided that countries own the program and incorporate trade objectives thoroughly into their development strategies. Nearly all bilateral and multilateral organizations are working to improve effectiveness, but not all have recent, comprehensive evaluations of their programs.20 With more than 40 bilateral and multilateral agencies involved in trade-related technical assistance, the scope for learning from each other is great.21 5. Challenges and Priorities Looking Ahead Ensuring timely and continued disbursements of existing aid for trade commitments to developing countries should be the first priority to guarantee the uninterrupted implementation of ongoing aid for trade programs, thereby helping developing countries mitigate some of the effects of the economic crisis and benefit more fully from the ongoing recovery in trade. We 19 See also Freund and Pierola (2010) for some anecdotal discussion of the effects of aid on export diversification in Peru. 20 Four important bilateral donors have undertaken evaluations of aid for trade programs relatively recently: USAID, DFID, SIDA, and the Netherlands. 21 The OECD is the primary focal point for efforts to enhance monitoring and evaluation of aid for trade projects and programs. Donors involved in providing assistance for trade-related analysis or programs include the International Trade Centre (Geneva), the United Nations Conference on Trade and Development, the United Nations Development Program, the World Bank, the Enhanced Integrated Framework, the Food and Agriculture Organization, the United Nations Industrial Development Organization, the World Customs Organization, the World Intellectual Property Organization, as well as regional development banks and many bilateral donors. 13 argue, in what follows, that there are a number of areas where collective action by the G20 can enhance the effectiveness of aid for trade as an instrument to promote inclusive growth. 1. Leveraging investments in infrastructure: The services "software" agenda An increasing number of countries identify infrastructure as a regional priority, as revealed by the self-assessment questionnaires carried out for the OECD-WTO (2009) report. As noted above, infrastructure is the largest category of aid for trade: infrastructure projects account for about 54% of the global aid for trade portfolio. Recent research has found evidence on the potential gains to investment in hard infrastructure, including improved export performance (Francois and Manchin, 2008). There is also evidence of a significant potential for reduced trade transaction costs and increased consumer welfare from investment in infrastructure, such as new ports (Abe and Wilson 2009). Investment in infrastructure may also have a greater impact in countries with lower per capita income in terms of generating a higher marginal impact on export performance (Portugal-Perez and Wilson 2010a). Investment in infrastructure must be accompanied by measures that reduce trade costs (Hoekman and Nicita, 2010) and by appropriate regulation ­ for instance, policies that promote competition in transport services and improvements in border management.22 The quality of public and private services can be an important determinant of the size of the payoffs to improvements in hard infrastructure. More generally, the efficiency, variety and costs of services inputs are critical for the competitiveness of firms and farmers because they represent an important share of the total costs of production. Being able to compete in international markets is increasingly determined by access to low-cost and high- quality producer services such as telecommunications, transport and distribution, and finance. Policies that raise operating costs or preclude innovation therefore can be very detrimental to the performance of the national economy. Policy reforms that revolve around increasing the contestability of services markets, facilitating new entry and the supply of new service products are also cheap in financial terms-- they often do not require massive investments in hardware. Developing countries tend to have more and higher barriers to services trade and investment, as shown by the negative correlation between GDP per capita and the restrictiveness of services trade and investment policies as measured in Gootiiz and Mattoo (2009) (Figure 5). Removing such restrictions can generate substantial benefits, leading to lower cost and higher quality producer services for firms and farmers in these countries. Global outsourcing and integration into international value chains increasing depend on having access to a variety of services. An increasing body of research demonstrates that reforms in services sectors has a positive effect on the productivity of both foreign ­and locally owned manufacturing firms that use services inputs (see Francois and Hoekman, 2010 for a recent survey of the literature). 22 Raballand and Teravaninthorn (2009) find that a lack of competition in trucking in West and Central Africa results in higher transport prices and lower quality of services compared to more contestable Africa markets. 14 Figure 5: Services trade restrictiveness Index ETH 80.0 IND ZWE 60.0 IRN QAT PHL IDN MYS DRC VNM CHN PAN KWT MWI BGD EGYTUNLBN NPL THA OMN 40.0 YEM LKAJOR NAM BLR DZA VEN SAU MOZ TZA URY BHR LSOCMR UKR CRI FRA KEN BOL BWA MEX CIV MAR KOR CAN CHL UZB HND ALB TUR HUN ITAEL DNK RWA ZMBNGA UGA KHM MLI SEN PAK PRY ARM BRA RUS PRT B GRCAUT DEU GHA ESP COL ARG CZEJPNUSA PERKAZ ZAF NZLSWE GBR KGZ MNG ROM AUS 20.0 MDG NIC GTMDOM BGR GEO MUS LTU IRL FIN POLTTO NLD ECU 0.0 4 6 8 10 12 Log of GDP per capita, 2007 Note: GDP per capita (constant 2000 US$); 102 countries Source: Gootiiz and Mattoo (2009) + A noteworthy feature of the pattern of services trade and investment policies is that landlocked countries apply more restrictive policies than coastal countries. This appears particularly true in the air transport and telecom sectors, in which landlocked countries have no inherent disadvantage (Borchert, Gootiiz, Grover, and Mattoo 2010). While there are many reasons why being landlocked might lead to lower availability of services and higher prices, restrictive policies contribute to the poor performance in services sectors, beyond the handicap imposed by geography. This suggests that supporting policy reforms to enhance the contestability of "backbone" services in landlocked countries could be a priority area for aid for trade. To date, however, much of the aid for trade effort has put the emphasis on support for hard infrastructure and improving productive capacity. Less has been done to improve the services- related policies and regulation that help determine the efficiency of (cost of using) infrastructure networks. This is one area where the support and leadership of the G-20 can make a difference. There are two dimensions to this: (i) ensuring that aid for trade assistance includes an adequate focus on pro-competitive regulation and other policies that affect the functioning of producer service markets (Hoekman and Mattoo, 2007); and (ii) doing more to provide access to the knowledge and experience on these matters that exists in the middle-income, emerging market members of the G20, as well as developed economies. Focusing explicitly on improving the operation and efficiency of services sectors is important in itself from a development perspective, but is also important from a global perspective. As argued by Claessens, Evenett and Hoekman (2010) and Hoekman and Messerlin (2010), rebalancing the world economy ­ reducing large current account surpluses and deficits ­ will require improvements in productivity (competitiveness) and domestic absorption in deficit and surplus 15 countries, respectively. In practice this cannot be achieved through monetary, fiscal and exchange rate policies alone ­ it will require changes in the structure of economies, including action to increase the availability, variety and quality of services inputs. 2. Expanding South-South integration through trade reform and market access Another area where the G-20 can provide important leadership is through expanded market access ­ especially for the least developed countries ­ led by reform in middle income countries to expand trading opportunities in a south-south context. This would provide an opportunity to low-income economies both to expand trade, and, as importantly, help them diversify across a larger number of markets. South-South trade has been growing rapidly in recent years as a result of high rates of economic growth that were achieved by many developing countries. The BRIC countries, for example, have an import share of 12 percent (2008) compared to just 6 percent in 1996. Meanwhile, high- income countries' share of import demand decreased from 81 percent in 1992 to 72 percent in 2008 (Haddad and Hoekman 2010). Significant trade barriers remain in many of the dynamic emerging markets. The emphasis in policy forums such as the WTO has been on developed country market access conditions, including achieving duty-free, quota-free access for the LDCs and addressing key constraints that reduce the value of preferential access such as rules of origin. While this is important, it arguably represents a missed opportunity for low-income developing countries that confront high barriers against exports in middle-income countries. Fugazza and Vanzetti (2008) compare the potential effects of the removal of barriers on South- South trade with the gains from developed country liberalization and from regional free trade areas within Africa, Asia and Latin America using a general equilibrium model, GTAP. Their simulations indicate that the opening up of northern markets would provide annual welfare gains to developing countries of $22 billion. However, the removal of South-South barriers has the potential to generate gains 60 per cent larger. The results imply that giving greater emphasis to removing barriers between developing countries could give a significant boost to trade with low- income countries. Overall, research suggests that whereas traditionally the bulk of South­North trade flows were in less sophisticated sectors with fewer learning opportunities; this may not be the case today, particularly among the dynamic Asian economies. Klinger (2009) studies the composition of South­South as opposed to South­North trade in recent years to consider whether the South as a market provides developing countries with greater opportunities to transform their productive structures and move to more sophisticated export sectors than the Northern market does. His results show that for many developing countries, including in Africa and Central Asia, exports within the South are more sophisticated and better connected in the product space than exports to the North, whereas the opposite is true for the faster-growing economies of Asia and Eastern Europe (excluding the Commonwealth of Independent States). Klinger also finds that the primary source of cross-country variation in export sophistication and connectedness is between northbound rather than southbound export baskets. 16 Post-crisis projections are that middle-income markets will grow more rapidly than those of high-income countries. The emergence of multiple growth poles in the South offers low-income countries an opportunity to diversify both across markets and products given that developing country consumers have differentiated preferences and demand. Moreover, increased South- South trade reduces the exposure of developing countries to possible prolonged slow growth markets in Europe, Japan, and the United States. It also mitigates risk associated with increased market openness and trade-led growth through product and good diversification effects, as mentioned above. South-South trade has already increased at the extensive and the intensive margins. Exports of LMICs to BRICs rose from 7 percent of the total in 2000 to 12 percent in 2008 (Figure 6). The average value of a transaction from LMICs to BRICs increased 444 percent during the 1996- 2008 period, while that from LMICs to HICs rose only 180 percent. However, developing countries still export substantially fewer varieties than high-income or even middle-income countries ­ there is therefore great scope for further diversification (Figure 7). Middle-income emerging markets also are a source of knowledge and FDI ­ which in turn can drive additional trade growth in low-income economies. Harnessing these opportunities is in part a function of putting in place the appropriate policies ­ including removal of market access barriers. If all OECD countries were to offer 100 percent DFQF, exports of the LDCs could increase by up to $2 billion more than they would under a 97 percent scenario (Bouët et al., 2010). But these export gains would be greater still if major middle-income nations were to offer DFQF access to LDCs--by up to $5 billion ­ reflecting higher tariffs in these countries. To be effective, such improved market access needs to be accompanied by liberal rules of origin and related administrative requirements. Figure 6: Developing Countries Account for Figure 7: Southern Countries Still Export an Increasing Share of World Trade Fewer Varieties Than Northern Ones (percent share of world trade) (average number of exported varieties at 6-digit HS) Source: Haddad and Hoekman (2010), drawing on UN-COMTRADE data. Note: LICs=low-income countries; LMICs=lower middle-income countries; UMICs=upper middle-income countries; and HICs=high-income countries. 17 3. Supporting regional cooperation and integration of markets: Capacity building Although much (most) of the aid for trade agenda is national in scope, there has been a recent rise in the demand for assistance to support regional integration. One factor driving this is a recognition that key constraints to a country's competitiveness may lie outside its borders. This is most directly the case for landlocked countries. There are a number of common priorities for regional integration in areas such as transport infrastructure, road corridors, energy and water, and trade facilitation. Efforts to integrate neighboring markets for goods, services and factors of production (workers, investment) can help stimulate South-South trade by reducing trade costs and allowing economies of scale to be realized. Much of the agenda here revolves around initiatives to lower transactions and operating costs for firms on both side of the border. Lowering such costs in a cooperative (joint) manner does not give rise to the types of welfare- reducing trade diversion that can arise from preferential reduction of tariffs: lower trade costs benefit all trade partners ­ they facilitate trade with the rest of the world as well as with neighbors.23 There is evidence of the benefits of strengthened regional cooperation. The Asia Pacific Economic Cooperation (APEC) agenda on trade provides one example. Increased transparency can have a positive impact on trade and welfare. Based on a computable general equilibrium model, Abe and Wilson (2008) find that trade in APEC countries would increase by 11 percent and global welfare would expand by $406 billion by reform aiming at raising transparency to the average level in the region. The simulations suggest that most of the increase in welfare would take place in member economies undertaking reform. Among the reformers, the GDP of Vietnam, Thailand, Russia, and the Philippines would increase approximately 20 percent. There is also evidence to suggest that reform in some of the poorest regions of the world could generate substantial benefits. Improvements in trade facilitation indicators in Africa, for example, to cut trade costs half-way to the best performer in the region for Ethiopia is roughly equivalent to a 7.6% average cut in tariffs faced by Ethiopian exporters in export markets (Portugal-Perez and Wilson, 2009). Cooperation at the regional level poses specific challenges in that the costs and benefits of projects can be very asymmetrical, with most of the required investments (and thus costs) accruing to a country that gets relatively little benefit from the investment. As this can greatly reduce support for regional projects that are critical to landlocked developing countries, one rationale for aid for trade is to increase the incentives for joint action in areas where benefits are distributed asymmetrically across countries. For example, landlocked developing countries in Africa, in which more than a quarter of the continent's population lives, face a substantial competitive disadvantage due to high trade costs (Djankov et al, 2006; Raballand and Teravaninthorn, 2009; Arvis and Raballand, 2010; Arvis et al. 2010). These countries also tend to have lower levels of foreign direct investment. Portugal-Perez and Wilson (2010b) explore the relationship of trade costs and incoming FDI into developing countries, including landlocked ones. Preliminary estimates show a negative relationship between trade costs and FDI in a North-South context. Indeed, most incoming FDI 23 As has been discussed extensively in the literature on regionalism, it is important that policy not target an expansion in intra-regional trade per se as a policy objective. What matters is to reduce barriers to trade generally, and regional agreements can help do so ­ especially for land-locked countries. 18 in developing countries finance operations entailing the transport of goods across borders, as in extracting industries, or in industries exporting goods intensive in low-skilled labor. In that context, they argue that domestic trade costs can be seen as a tax on operations and have an impact on FDI attractiveness. Landlocked developing countries are particularly affected as they tend to have higher export costs than their coastal neighbors. For these states, domestic problems are multiplied by those prevailing in transit/coastal countries through negative spillover effects. But there are also externalities for coastal countries: a nationally focused strategy often will not be sufficient to maximize trade and growth opportunities if neighboring markets are ignored. Policy reforms and actions that can lead to significant improvement of the business environment and attract investment are of a public good nature: the associated outputs are non-excludable (it is difficult to prevent countries who may not have contributed to its provision from using it) and non-rival in consumption (use by a neighboring country does not affect the supply or quality of the good); hence the need for a collective action solution at a regional level. The need for regional cooperation is understood by all stakeholders. However, the range of available instruments to support regional projects and cooperation is limited. This results in the under-provision of financing/assistance for multi-country trade-related projects (Hoekman and Njinkeu, 2010). Weak capacity of existing regional secretariats and pro-reform civil society groups, and the diffuse nature of the benefits of existing integration mechanisms for the private sector have also resulted in a poor implementation track record. Moving the regional integration agenda ahead requires addressing frontally the political economy of regional cooperation and coordination by increasing the incentives for implementation. This requires engagement on different fronts, with a reward/incentive scheme that targets all relevant actors ­ national governments, sub-national entities, and non-state actors. Dedicated funds to support regional cooperation, covering both "software" (regulatory institutions, policy changes) and hardware (infrastructure to support cross-country flows of goods, services and people) could help to fill the gap that currently exists. A concerted focus on identifying and financing regional projects that would help to address the national priorities could also help overcome resistance to beneficial regional market integration (beneficial in the sense of helping to attain the competitiveness objective). A practical way forward would be for a greater proportion of donor funds for aid for trade to be allocated to regional projects and programs.24 Most regional and multilateral institutions already have trust funds through which such resources could be channeled. 4. Harnessing the private sector as a source of knowledge, capital and information Given the broad nature of the aid for trade agenda ­ encompassing areas from border management to regulatory reform and infrastructure investment ­ there is a sizeable number of stakeholders involved from both the public and private sectors. As such, there is great scope to 24 While proposals for earmarked funds are controversial, as earmarking can be inconsistent with aid effectiveness (the activities for which funding is earmarked may not be a priority in individual countries), the creation of a mechanism that earmarks an overall amount for trade does not need to imply that countries must identify trade as a priority; it simply provides greater credibility to countries that if they decide that trade projects are a priority, development assistance will be available. 19 make effective use of public-private partnerships that capitalize on private sector expertise in prioritizing areas for reform and identifying potential solutions. Different models have proliferated at the national level, an example being national trade facilitation associations that work to connect stakeholders in the public and private spheres in order to carry out work at a broad national level, or in specialized areas such as border management reform (e.g., TradeNet of Singapore and Tradelink of Hong Kong, China).25 These networks serve as important platforms for developing national strategies and action plans for reform, in addition to providing stakeholders with a mechanism for coordination and harmonization of policy measures across industries and sectors. More generally, the private sector is already undertaking numerous initiatives to address concrete problems or to leverage ongoing investments to enhance development impacts. Examples are the Global Facilitation Partnership, a network of logistics companies and international agencies working on trade and transport (see http://www.gfptt.org/), growth corridor initiatives that are supported by Yara International in Ghana, Malawi and Tanzania, the Business Action for Africa network which has various focused initiatives such as the alliance for Improving Customs Administration in Africa, and the Private Investors for Africa coalition. Many countries have also put in place annual or more frequent mechanisms through which political leaders meet with the domestic private sector to discuss concerns and identify priorities for (joint) action. Much more can and should be done to harness the knowledge and information that exists in the private sector, both as a source of data on constraints to trade and policies or factors that needlessly increase costs of trading, and as a source of potential solutions to specific problems. Greater sharing of information on such initiatives and learning about what works and what does not would enhance the visibility of such efforts and boost the role of the private sector in the broader aid for trade program.26 5. Bolstering monitoring and evaluation of the effectiveness of aid for trade Effective monitoring of delivery of aid for trade and the extent to which it responds to national priorities as defined by recipient governments is important to allow accurate assessments and evaluation of outcomes. Most donors monitor and evaluate their aid for trade programs in accordance with generic evaluation guidelines or with specific guidelines for themes and sectors falling under aid for trade (see OECD/WTO Donor Questionnaire). More learning could be generated by applying, whenever possible, the kind of impact-evaluation methods now widely used in the evaluation of poverty, health and education projects. The essence of these methods consists of using control groups to benchmark the improvement in the 25 See UNESCAP,"Trade & Investment Issues, Aid for Trade and Public-Private Partnerships" E/ESCAP/CMG(4/I)/2, 3 July 2007. 26 The World Bank is developing a new Public-Private Partnership on Aid for Trade Facilitation as a platform for an exchange of information and learning in the area of trade facilitation. The project will design and implement practical and achievable trade facilitation projects that lower trade costs by addressing the lack of capacity of developing countries to rapidly move goods and services across borders. A central focus of the work will be to improve the "software" of trade logistics and border management to complement and enhance hard infrastructure investments. In addition, the partnership will leverage private sector expertise in producing real-time trade performance data, which may be used to encourage policy-oriented trade facilitation reform. 20 performance of individuals "treated" by particular programs. Clearly, not all trade-related programs can be amenable to such "treatment-effect" methodologies. The easiest are export- promotion programs that target individual firms..27 Even in the limited areas where their application is relatively straightforward, applications of "clinical" impact evaluation methods to trade-related programs have so far been limited. They provide no evaluation of spillover effects--even though spillovers are key to the justification of public intervention. Moreover, like all clinical impact evaluations, they have uncertain "external validity", as what works in one setting may not work in another. Notwithstanding these and other caveats that have been extensively discussed in the literature (Rodrik 2008), they offer a valuable tool to understand what works and what does not. In particular, when carefully thought out, they can help identify which components of assistance programs work best. That is, beyond their contribution to general accountability, they have the power to generate useful knowledge to renew the factual basis on which to base policy advice and donor practice. Investment in data and analysis should span work at both the macro and micro levels. An agenda in this area must center on a framework for rigorous evaluation of aid-for trade-projects, empirical research on aid impact evaluation, drawing on macro datasets from the CRS OECD databases and micro-data from projects that are implemented by development agencies. Country and/or regional analyses of aid for trade effectiveness are needed to assess how types of aid for trade funds are spent in relation to their returns, as measured by increased trade flows, lower trade costs, etc. Data on trade costs could be collected from a variety of sources, including trade support institutions, customs authorities, and international transport companies. Detailed data will be needed to assess policies related to specific aid for trade interventions ­ e.g. support for industrial upgrading, certification of firms (ISO 9000), or technical assistance for transport logistics (Wilson, 2010). At the country- and regional level, the private sector has an important role to play monitoring and evaluation, as directly affected and concerned firms, industries and associations have access to information regarding the effectiveness of efforts to address specific trade-related bottlenecks and constraints. 6. Moving the Agenda Forward The G-20 is uniquely positioned to support specific actions to expand global trade. The fragile economic recovery ­ combined with the need to strengthen the international trading system in support of sustainable and inclusive growth and employment ­ place the aid for trade initiative at the forefront of policy importance. In addition to delivering on the commitments made in Gleneagles and Hong Kong on expanding aid for trade flows, there are four strategic themes that a G-20 "Action Agenda on Aid for Trade and Development" might support: 1. Establish a platform for capacity-building and transfer of knowledge to improve regulation of producer services and the operation of network infrastructure. A coordinated program of assistance and knowledge exchange that includes active involvement of middle income G20 countries could do much to increase the rate of return 27 See Volpe and Carballo (2008) for an evaluation of Peru's export-promotion program. 21 on aid for trade investments in hard infrastructure by creating a mechanism that will focus on strengthening capacity to put in place the associated complementary "software" inputs ­ policies, pro-competitive regulation, etc. ­ that are critical to realize both social (equity) objectives and improve the efficiency of use of network infrastructure. This is an agenda that goes beyond leveraging investments in infrastructure. It encompasses producer and business services more generally. An important factor that explains lack of progress in negotiations aimed at liberalization of trade and investment in services is uncertainty and concerns regarding the possible consequences of making market access commitments. Establishment of a forum that is aimed at substantive discussion and analysis of the impacts of liberalization and specific regulatory policies and policy changes could do much to build a common understanding of where there are indeed large gains from liberalization (Feketekuty, 2010; Hoekman and Messerlin, 2010). There is no existing institution that has an obvious comparative advantage in playing this role. One option could be to pursue a consortium approach, in which a number of policy institutes, international organizations and networks of regulators (such as the International Competition Network) from around the world combine to provide the needed knowledge resources and deliver the suggested services, working with or through a central hub entity that would be created. Such a "knowledge platform" would need a governance structure in which donor governments and other funders are represented. A possible model is the one that was used to establish the Global Development Network. 2. Complementing the financial aid for trade provided by high-income countries with market access reform by middle income G20 members to lower barriers to exports from poor countries so as to expand south-south trade. Extending duty-free, quota-free access for LDCs to all G20 members, with minimal exceptions, would constitute a concrete initiative that would directly promote the trade and development prospects of the poorest countries in the world. It is an initiative that is completely at the discretion of countries in that it can be done at the stroke of a pen. It would come at very low "cost" in terms of additional imports given that the production and trade structures of the LDCs and the G20 countries have little overlap and that LDCs are in any event very small suppliers. Any DFQF initiative would need to be accompanied with liberal rules of origin and rules of cumulation--as has now been documented extensively, restrictive rules of origin can greatly reduce the effectiveness of DFQF programs. Concrete solutions to the rules of origin constraint have been developed by several importing countries and can be emulated by middle-income G20 members (see Elliott, 2010). 3. Creation of a new "public-private aid for trade partnership" to leverage the dynamism in the private sector for strengthening trade capacity in the countries that are recipients of aid for trade. Given the high payoffs from improving trade facilitation ­ encompassing areas from border management to regulatory reform and adoption of modern ICT technologies ­ such a partnership might focus initially on capitalizing on private sector expertise in prioritizing areas for reform and identifying potential solutions, while leveraging the coordinating capacities of governments and/or multilateral donor institutions. More generally, the focus could be on highlighting initiatives that are being 22 implemented by firms and associations and sharing information on good practices and effective models of public-private partnerships that aim to leverage trade opportunities. 4. Joint action to provide dedicated financial support for a concerted program of monitoring and evaluation of aid for trade anchored in systematic data collection and research. All donors and recipients recognize the importance of monitoring and evaluation and analysis of trade outcomes and performance. The OECD is taking the lead in coordinating efforts to share the results of monitoring and evaluation by donors and agencies and to learn from experience. There is, however, no dedicated funding to ensure consistent cross-country collection of data on trade outcomes and their determinants on a comparable basis. While much knowledge is generated through ex post evaluation of projects, this does not result in datasets that allow for benchmarking of countries and tracking of performance over time. A concerted effort is needed to ensure that data are collected to allow the impacts of policy reform efforts and interventions to be compared across countries and over time. This will require agreement among governments and agencies on the specific indicators for which data should be compiled. Candidates include measures of trade costs ­ e.g., clearance/waiting times; the number of times that trucks are stopped along transport corridors; rejection or inspection rates of consignments at borders; trade diversification; and trade and investment policies (e.g., restrictiveness of service sector policies and the prevalence and intensity of nontariff measures). References Abe, Kazutomo and John S. Wilson. 2008. "Trade, Transparency, and Welfare in the Asia Pacific," Journal of International Economic Studies 12(2): 35-78. Abe, Kazutomo and John S. Wilson, 2009. "Weathering the Storm: Investing in Port Infrastructure to Lower Trade Costs in East Asia," World Bank Policy Research Working Paper 4911. Arvin, M., Cater, B. and S. Choudry. 2000. "A causality analysis of untied foreign assistance and export performance: the case of Germany," Applied Economic Letters, 7:315-319. Arvis, Jean-Francois and Gael Rabbaland. 2010. The Cost of Being Landlocked. Washington DC: World Bank, forthcoming. Arvis, J.F. and others. 2010. Connecting Landlocked Countries to Markets. Washington DC: World Bank, forthcoming. Borchert, Ingo, Batshur Gootiiz, Arti Grover, and Aaditya Mattoo. 2010. "Landlocked or Policy Locked? The Effect of Services Policy Restrictiveness on Service Sector Performance," World Bank: unpublished manuscript. Bouët, A., D. Laborde, E. Dienesch and K. Elliott. 2010. "The Costs and Benefits of Duty-Free, Quota-Free Market Access for Poor Countries: Who and What Matters?," Washington DC: Center for Global Development Working Paper 206. Brenton, Paul, and Erik von Uexkull. 2009. "Product Specific Technical Assistance for Exports-Has it been Effective?" The Journal of International Trade & Economic Development, 18(2): 235-54. Bruhn, Miriam and David McKenzie, 2008. "In pursuit of balance: randomization in practice in development field experiments," World Bank Policy Research Working Paper 4752. Cali, Massimiliano and Dirk Willem te Velde, 2008. "Towards a Quantitative Assessment of Aid for Trade," London: Commonwealth Secretariat. 23 Cali, Massimiliano and Dirk Willem te Velde. 2009. "Does Aid for Trade Really Improve Trade Performance?," London: Overseas Development Institute. Claessens, C., S. Evenett and B. Hoekman. 2010. "Rebalancing the Global Economy: A Primer for Policymakers," VOXEU column. Collier, P. and D. Dollar 2002. "Aid allocation and poverty reduction", European Economic Review, 46(8): 1475-1500. Djankov, S., Freund, C. and S. Pham Cong, 2006. "Trading on time", Policy Research Working Paper 3909, The World Bank. Elliott, K. and others. 2010. "Open Markets for the Poorest Countries: Trade Preferences That Work," Report by CGD Working Group on Global Trade Preference Reform. Washington DC: Center for Global Development. Feketekuty, G. 2010. "Needed: A New Approach to Reduce Regulatory Barriers to Trade," at http://www.voxeu.org/index.php?q=node/5208. Francois, J. and B. Hoekman. 2010. "Services Trade and Policy," Journal of Economic Literature, forthcoming (CEPR Discussion Paper 7616). Francois, Joseph and Miriam Manchin, 2007. "Infrastructure and Trade." World Bank Policy Research Working Paper 4152. Freund, Caroline and Martha Denisse Pierola. 2010. "Export Entrepreneurs: Evidence from Peru," World Bank, mimeo. Fugazza, Marco and Vanzetti, David. 2008. "A South-South Survival Strategy: The Potential for Trade among Developing Countries," World Economy, 31(5): 663-84. Gamberoni, E., and Richard Newfarmer, 2009. "Aid for Trade: Matching Supply and Demand," World Bank Policy Research Working Paper 4991. Gootiiz, B. and A. Mattoo. 2009. "Services in Doha: what's on the table?" World Bank Policy Research Working Paper 4903. Haddad, M. and B. Hoekman. 2010. Trading Places: International Integration after the Crisis," in O. Canuto and M. Giugale (eds.) The Day after Tomorrow: economic policy challenges for developing countries in the post-crisis world, Washington DC: World Bank. Haddad, Mona, Jamus J. Lim and Christian Saborowski. 2010. "Trade Openness Reduces Growth Volatility When Countries are Diversified," World Bank Policy Research Working Paper 5222. Helble, Matthias, John Wilson and Catherine Mann, 2009. "Aid for Trade Facilitation," World Bank Policy Research Working Paper 5064. Hoekman, Bernard. 2007. "Aid for Trade: Helping Developing Countries Benefit from Trade Opportunities," in D. Njinkeu and H. Cameron (eds.), Aid for Trade and Development. New York: Cambridge University Press. Hoekman, B. and A. Mattoo. 2007. "Regulatory Cooperation, Aid for Trade and the GATS," Pacific Economic Review, 12(4): 399-418. Hoekman, B. and P. Messerlin. 2010. "The EU and Rising Economic Powers: Focus on Services Markets," Europe's World, forthcoming. Hoekman, B. and A. Nicita, 2010. "Assessing the Doha Round: Market access, transactions costs and aid for trade facilitation, The Journal of International Trade & Economic Development, 19(1): 65-80. Hoekman, B. and D. Njinkeu. 2010. "Aid for trade and export competitiveness: New opportunities for Africa," in Export Supply Response Capacity Constraints in Africa. Nairobi: African Economic Research Consortium. Independent Evaluation Group. 2006. Assessing World Bank Support for Trade, 1987­2004. Washington: World Bank. 24 Klinger, Bailey (2009) "Is South-South Trade a Testing Ground for Structural Transformation?" UNCTAD, Policy Issues in International Trade and Commodities Study Series #40. Kosack, S. 2008. "Trade for Poverty Reduction: The Role of Trade Policy in Poverty Reduction Strategy Papers," United Nations Development Program, New York. Lloyd, T., M. McGillivray, Morrissey, O. and R. Osei, 2000, `Does Aid Create Trade? An Investigation for European Donors and African Recipients', European Journal of Development Research, 12(1): 107­23. Management Systems International and the University of Pittsburgh, 2010. "Evaluation of US Trade Capacity Building Assistance (2001-2009). Forthcoming. Martinez, Marco and John S. Wilson, "Aid for Trade: Revisiting Supply and Demand," World Bank, mimeo. Morrissey, Oliver, 2006. "Aid or Trade, or Aid and Trade?" Australian Economic Review, 39: 78-88. Nelson, D and S.J. Silva (2008) "Does Aid Cause Trade: Evidence from an Asymmetric Gravity Model' WP 2008/21 University of Nottingham Leverhulme Centre. Nilsson, L. 1997, EU and Donor Exports: the Case of the EU countries, Essays on North-South Trade, Lund: Lund Economic Studies (chapter 3). OECD. 2006. "Trade-Related Assistance: What do recent Evaluations Tell Us?" Paris: OECD (http://www.oecd.org/dataoecd/31/45/37624158.pdf). OECD, 2008. "Monitoring Aid for Trade: a Background Note." OECD Policy Dialogue on Aid for Trade, www.oecd.org/trade/aftdiaolgue2008. OECD and WTO. 2008 Aid for trade at a glance 2007 : first global review World Trade Organization ; Organization for Economic Co-Operation and Development, Geneva/Paris. OECD and WTO. 2009 Aid for trade at a glance 2009 : maintaining momentum World Trade Organization & Organization for Economic Co-Operation and Development, Geneva/Paris. Osei, Robert; Oliver Morrissey; and Tim Lloyd. 2004. "The nature of aid and trade relationships." The European Journal of Development Research, 16(2):354­374. Portugal-Perez, Alberto and John S. Wilson 2009. "Why Trade Facilitation Matters to Africa," World Trade Review 8:3 379-416. Portugal-Perez, Alberto and J. Wilson. 2010a. "Export Performance and Trade Facilitation Reform: Hard and Soft Infrastructure," World Bank Policy Research Working Paper 5261. Portugal-Perez, Alberto and John S. Wilson. 2010b. "Trade costs and FDI", mimeo World Bank. Prowse, Susan. 2006. "Aid for Trade: A Proposal for Increasing Support for Trade Adjustment and Integration," in S. Evenett and B. Hoekman (eds.), Economic Development and Multilateral Trade Cooperation. Washington DC: Palgrave-McMillan and World Bank. Rajan R.G. and A. Subramanian (2005) "What Undermines Aid's Impact on Growth?" IMF WP/05/126. Radelet, Steven, Michael Clemens, and Rikhil Bhavnani. 2006. "Aid and Growth: The Current Debate and Some New Evidence." In P. Isard, L. Lipschitz, A. Mourmouras,and B. Yontcheva (eds.), The Macroeconomic Management of Foreign Aid: Opportunities and Pitfalls. Washington, DC: IMF. Reis, José Guilherme and Thomas Farole, 2010. "Exports and the competitiveness agenda: policies to support the private sector," in O. Canuto and M. Giugale (eds.) The Day after Tomorrow: economic policy challenges for developing countries in the post-crisis world, World Bank, forthcoming. Strachan, Yolanda. 2009. "Trade in World Bank Country Assistance Strategies," Washington DC: World Bank (International Trade Department, August). Suwa-Eisenmann, A., and T. Verdier. 2007. "Aid and Trade," Oxford Review of Economic Policy 23(3):481-507. 25 UNCTAD. 2005. Trade and Development Report, 2005. United Nations publication. New York and Geneva. UN Millennium Taskforce on Trade, 2005. Trade for Development. New York: UN Millennium Project. Wilson, John S. 2010. "Aid for Trade: What We Know and Need to Know about Effectiveness, Monitoring, and Evaluation." (mimeo). Wilson, J.S., C. Mann and T. Otsuki. 2005. Assessing the potential benefit of trade facilitation: A global perspective. The World Economy, 28(6). World Bank. 2009. Unlocking Global Opportunities: The Aid for Trade Program of the World Bank Group. Washington DC: World Bank. World Trade Organization. 2005. "Hong Kong Ministerial Declaration," WT/MIN(05)/W/3/Rev. 2, 18 December. 26