The United Kingdom will confirm its departure from the European Union on 31st January 2020. As part of its independent trade policy, the government has committed to improve access to UK markets for the poorest countries. This note sets out three ways it can do so: expanding duty-free market access while avoiding piecemeal trade agreements that undermine Africa’s own trade integration efforts; using an alternative framework for those trade agreements it does negotiate with developing countries; and supporting a “back-to-basics” multilateral negotiation at the World Trade Organization that could help to rebuild confidence in that institution and thus protect the interests of small and vulnerable countries. After a brief review of the background and context, it sets out specific proposals in each of these areas.
As the United Kingdom prepares to complete its exit from the European Union, it is developing elements of a post-Brexit trade policy. In the short run, a priority for British policymakers has been to minimize trade disruptions, primarily by replicating as much as possible the relations that the country currently has with the rest of the world as a member of the single market. That means trying to negotiate dozens of trade continuity agreements around the world, as well as creating policies to maintain preferential access for developing countries. In a welcome step, the government approved legislation more than two years ago to ensure duty-free, quota-free market access continues for the least developed countries (LDCs).
Once past the transition, the political declaration setting out the framework for the future relationship with the EU confirms the UK will not remain in a customs union and will develop an independent trade policy. As it does so, there is ample scope to improve on the policies to which it is party as an EU member, including in its approaches to using trade as a development tool. In this context, British policymakers have emphasized the goals of maintaining access for the poorest countries, supporting regional integration in Africa and improving current levels of market access for those countries, and ensuring the consistency of these policies with World Trade Organization rules. 
This note builds on previous CGD publications that recommend that post-Brexit trade policy for development should:
be rooted in a broadly liberal trade policy that eliminates barriers to trade to benefit British consumers and the economy as a whole, while opening opportunities for developing country exporters, and
extend duty-free, quota-free market access, with flexible rules of origin, to more poor countries in Africa and elsewhere.
After providing a brief review of the background and context framing the options in this area, this note will first, flesh out the second recommendation to expand duty-free, quota-free access to more countries—without running afoul of WTO rules on nondiscrimination and without having to negotiate fragmented economic partnership agreements that undermine African integration efforts. Second, in situations where a regional or bilateral trade agreement could be useful, the note will suggest an alternative to the EU framework that better respects partner countries’ policy autonomy and development priorities. Third, it will present proposals developed more fully elsewhere (Elliott 2020) for a multilateral negotiation at the WTO that could help to rebuild confidence in that institution and thus protect the interests of small and vulnerable countries.
Background and Context
Before turning to a discussion of key elements of trade preference programs, it is useful to recall that they are needed because developed countries’ trade policies so often discriminate against developing country exports. For many developing countries, the tariffs in their major export markets are often highest on agricultural products and labor-intensive manufactures, such as clothing and footwear, where they have a comparative advantage. For example, while the overall (trade-weighted) average tariff in the EU is 3.2 percent, it is two to three times as high for imports of agricultural and fishery products, and clothing. Without trade preferences, Kenyan agricultural exports would face an average tariff of 8.5 percent in the EU market, while Mauritius’ nonagricultural exports would face a tariff of 13.6 percent.
Many industrialized countries’ also have policies with tariff escalation as a prominent feature. This means they import raw commodities duty-free, but impose higher tariffs on more processed products, which discourages the development of value-added activities in the commodity-exporting country. Tariff escalation is a particular problem for many African countries that remain relatively dependent on commodity exports. Two-thirds of sub-Saharan African exports are oil, other mineral resources, or metals that the EU imports duty-free when they are in raw form, but taxes at rates of 4–5 percent or more when they undergo processing. Another 15 percent of sub-Saharan African exports are agricultural products, many of which face high barriers, including tariff escalation. For example, the EU imposes no duty on cocoa beans but has an 8 percent tariff on cocoa powder. The EU restricts chocolate imports with a complicated set of tariffs and quantitative restrictions. Many fruit and vegetable exports face opaque and confusing tariffs that vary by season.
The Generalized System of Preferences (GSP) and other preference programs are generally a second-best alternative to eliminating barriers across the board that discriminate against developing countries because preferences create their own distortions, along with the opportunities (Crawfurd et al., 2017; Mendez-Parra 2017, pp.8-9). The UK has moved in a first-best direction by announcing a temporary tariff schedule in the case of a no-deal Brexit that eliminates tariffs on goods not produced in the UK, including footwear, bicycles, and a range of agricultural products.
Since these products face relatively high most-favored-nation (MFN) tariffs under the EU schedule, however, eliminating them will result in the erosion of trade benefits for developing countries currently exporting to the UK on preferential terms under the EU’s GSP or Everything But Arms (EBA) programs, or as parties to free trade agreements. The temporary tariff schedule recognizes the potential for preference erosion by maintaining duties on some high preference margin products, including bananas, fish, sugar, and clothing. That is a highly inefficient way, however, to transfer resources to developing countries, and it would be better to aim for as few exceptions as possible and provide aid for trade and adjustment assistance to help countries facing preference erosion problems.
Still, as long as countries maintain tariffs, there will be a role for preference programs to offset discrimination against poorer countries. And British policymakers can improve on the EU model—particularly by doing more to support the need for deeper regional integration in Africa.
Maintaining Preferential Market Access for Poor Countries while Supporting African Integration
A core WTO principle is the MFN principle, which requires members to provide the treatment provided to their “most favored” partner to all other WTO members. That is, WTO members are not supposed to discriminate based on the origin of the goods they import. But there are exceptions, notably including the 1979 Enabling Clause. That provision permits WTO members to provide better than MFN treatment to promote economic growth and global integration in developing countries. Within that group, the WTO recognizes United Nations-designated LDCs as being particularly vulnerable and therefore deserving of even more expansive preferential treatment. Since adoption of the Enabling Clause, developed countries have created GSP programs to reduce or eliminate tariffs on designated imports from eligible developing countries, and most have created special preference programs for LDCs.
The original idea behind the Enabling Clause was that trade preferences should be “generalized” across developing countries, with more generous treatment reserved for LDCs. But that principle eroded over the years as the EU and the United States finetuned their programs to suit national priorities. The US supplemented its GSP program with regional programs providing more generous benefits for certain groups of countries in Latin America, notably Haiti, as well as for Africa under the African Growth and Opportunity Act. But the US remains the only developed country not to offer duty-free, quota-free market access to all LDCs. The regional programs cover around two-thirds of LDCs, but not those in Asia, and Washington had to get a WTO waiver to shield it from legal challenges.
In addition to a GSP program, the EU originally had special trade preferences for former colonies and commonwealth countries in Africa, the Caribbean and the Pacific (ACP) under the Cotonou Agreement. But faced with WTO rulings against elements of that program, the EU opted not to request a waiver. The EU now has a three-tiered GSP program:
standard GSP, which excludes upper-middle-income developing countries and limits product coverage for competitive low-middle-income exporters
the EBA program, which provides duty-free, quota-free market access on all commercial products for LDCs
the GSP+ program, which provides less than EBA but better than GSP access for countries that have relatively small and undiversified exports and meet certain standards with respect to protecting human rights and the environment, and promoting good governance (annex A)
In many ways, the EBA program is a model for providing the broadest possible access for LDCs. The conundrum for the EU was that these tiers left a number of Cotonou beneficiaries with either no preferential access (mainly upper-middle-income countries) or with more limited preferences under the standard GSP. The approach of negotiating economic partnership agreements (EPAs) to replace Cotonou preferences and provide similar access for non-LDCs in Africa (and elsewhere) has been highly controversial and not particularly successful.
The EPAs have thus far avoided a legal challenge because they have been notified to the WTO as interim agreements under Article XXIV of the General Agreement on Tariffs and Trade, another of the authorized exceptions to the MFN principle (Elliott and Temprano Arroyo 2019). While the EPAs have preserved ACP countries’ one-sided preferential access to the EU market in the short run, they are intended to be reciprocal and to cover many issues beyond tariffs in the longer run. Many of the provisions are unpopular, however, and some countries have signed on only because they fear losing preferential access to the EU market (Jones and Marti 2009).
The EPA approach has been particularly divisive in Africa because the hodgepodge of preferential arrangements has complicated negotiations originally designed to be undertaken with regional economic communities. This is undermining the development of “regional production networks” that a recent report from the Organisation for Economic Co-operation and Development, done in conjunction with the African Union Commission (AUC/OECD 2019, p. 17), emphasized as one of the three sets of core policies that Africa needs to pursue for growth.
The joint report notes that regional integration in Africa is important to increase economies of scale that could help attract foreign direct investment and make local firms more competitive. The report also emphasizes how regional integration can promote intra-African trade, which is particularly important for small and medium enterprises looking to diversify products and markets and scale up (ibid., pp. 30-31). Yet the EU’s EPA approach has arguably set back rather than advanced regional integration in Africa, with the result being a patchwork of unilateral preferences with LDCs, bilateral steppingstone EPAs with a few non-LDCs from regions with little support for the EPAs (central and western Africa), and interim EPAs with regional communities in eastern and southern Africa that are being provisionally applied.
The UK will face a similar dilemma in deciding how to improve market access for Africa without seeking a WTO waiver, which it is unlikely to do given other trade priorities and limited negotiating resources. A WTO ruling in response to India’s challenge to the EU’s GSP+ program suggests an alternative way forward, however. While the WTO overruled a specific element of GSP+, it essentially gave a green light to the overall EU approach by finding that trade preferences for developing countries could vary as long as “similarly situated” countries receive the same tariff treatment (see annex A). That suggests that a post-Brexit UK policy could provide duty-free, quota-free access to more countries than just the LDCs—and thereby maintain most African exporters’ access to its market—as long as it uses objective criteria to identify the beneficiaries. Such a policy could mitigate the need for ad hoc arrangements with non-LDC African countries, such as the EPAs that the EU has pushed despite extensive criticism.
An EBA+ Approach to Trade Preferences that also Supports Regional Integration in Africa
Currently, most African countries have duty-free, quota-free access to the EU market, or something very close to it under the EBA program, EPAs, or other arrangements. But, as noted, that hybrid approach is complex, inconsistent with the goal of continental-level integration, and unpopular in many African countries, as well as with much of the development advocacy community in the UK. Given the mix of low- and middle-income countries in Africa, some sort of hybrid approach is inevitable without a request for a WTO waiver like the United States has for the African Growth and Opportunity Act. Still, it need not be as divisive and complex as what the EU has done.
Given that 46 of the 55 countries in Africa are either LDCs or low-middle-income countries, a simpler approach should begin by extending the current duty-free, quota-free scheme for LDCs to other low- and low-middle-income countries (LICs and LMICs). One such approach is in Mendez-Parra (2017), which proposes that the UK should develop a two-tier post-Brexit preference program that would be available to LDCs, other low- and low-middle-income countries, and small island developing states (SIDS) that are not high income. The extension of duty-free, quota-free market access to the SIDS appears to have been motivated primarily by a desire to protect Mauritius from preference erosion for its clothing, sugar, and tuna exports. But the proposal would also add 17 other upper-middle-income SIDS that may not really need these preferences.
Creating two tiers of benefits also adds complexity that may not be necessary or helpful. Countries in the above categories that meet certain vulnerability criteria would receive preferential access similar to that currently received by LDCs under the EBA program. Relatively larger and more competitive LMICs—such as India, Indonesia, and Vietnam—would be placed in a second tier with far more limited benefits. This limitation would be necessary to preserve benefits for the poorest countries currently eligible for the EBA, and it would likely make the proposal more politically palatable. Mendez-Parra suggests that the vulnerability criteria determining eligibility for the more generous tier of benefits would be similar to those under the EU’s GSP+ program, which exclude larger and more diversified exporters. But the paper is not clear on which countries would be affected—including Bangladesh, which does not meet the current GSP+ vulnerability criteria.
A simpler, more poverty-focused program is possible, which I call EBA+. First, given the openness to a broader MFN liberalization signaled in the temporary tariff schedule that the government released in case of a no-deal Brexit, a two-tiered program may not be necessary. If maintained in the post-Brexit schedule, those tariff cuts will open access for some key products to all developing countries, including footwear exporters such as Indonesia and Vietnam.
Second, the definition of vulnerability for purposes of setting eligibility criteria for the duty-free, quota-free program could be simpler and more transparent. Similar to a Norwegian program for smaller LMICs—those with a population smaller than 75 million—the UK could use smallness and low incomes as the key criteria for EBA+ preferences. A figure around 70 million has the advantage of avoiding “thresholds effects” because no low- middle-income countries fall in the range between 55 million and 80 million.
In the EBA+ scheme proposed here, all LDCs, regardless of population, as well as other low- and low-middle-income countries with populations smaller than around 70 million would be eligible for duty-free, quota-free market access. Income rather than population should then serve as the criterion for graduation from the program. So, for example, if Tanzania, with 55 million people in 2018, passes the population limit before it reaches upper-middle-income country status, it would remain eligible. The scheme should also retain the flexible rules of origin available to current EBA beneficiaries, improving them where possible. It should also ensure that other administrative requirements are as simple and transparent as possible.
The proposed EBA+ would maintain or improve market access for a number of low-middle-income African countries, without the burden of having to negotiate EPAs: Cameroon, Côte d’Ivoire, Ghana, Kenya, Morocco, Tunisia, and Zimbabwe. Based on the World Bank’s list of LICs and LMICs in 2018, this proposal would add only three countries outside Africa that are not currently receiving better than GSP preferences in the EU—Micronesia, Tajikistan, and Uzbekistan—and slightly improve preferences for three GSP+ countries that meet the EBA+ criteria—Bolivia, Krygyzstan, and Mongolia. There are also six EU free trade agreement (FTA) partners—El Salvador, Honduras and Nicaragua, Moldova, Ukraine, and Papua New Guinea—that would be eligible for this scheme, but all except Moldova and Ukraine have signed or initialed trade continuity agreements with the UK. British policymakers in conjunction with their counterparts in the FTA countries would have to decide how to move forward after the transition.
Table 1 lists the non-LDCs that would be eligible for full market access under the EBA+ proposal, along with the type of preferences that they currently enjoy in the EU market. The table also shows larger and richer developing countries that could lose some preferential access under this proposal, compared to what they have now under EU arrangements. That is the issue to which I now turn.
Countries Left Out
The EBA+ option does, however, leave out some African countries that Mendez-Parra identifies as being relatively dependent on existing EU preferences—either because they are upper-middle-income countries or because they are beyond the population threshold. Of these, the UK has already agreed or initialed trade continuity agreements with the Eastern and Southern African grouping—including Mauritius and Seychelles—and the South African Development Community EPA group—including South Africa and Namibia. But it is still negotiating with parties to EuroMed agreements, including Algeria and Egypt. Other African countries that do not meet the EBA+ eligibility criteria are Equatorial Guinea, Gabon, and Botswana, because they are upper-middle-income countries, and Nigeria, because its population is over the threshold. But these are resource exporters that typically make little use of preferences.
Outside of Africa, Pakistan, and the Philippines could lose GSP+ preferences for which they are currently eligible, unless British policymakers opt to retain that program or make other arrangements. India, Indonesia and Vietnam currently receive only standard GSP in the EU and would not be eligible for the expanded duty-free, quota-free scheme, also because their populations are too large. But, as noted, if the UK goes ahead with a more liberal MFN tariff schedule, they would still see some benefits.
British policymakers could opt to add another tier of unilateral preferences for less vulnerable developing countries to mitigate these effects. If they decide to create a version of the EU’s GSP+ program, they might consider an alternative that uses preferences to support and reward countries contributing to global public goods, such as those hosting large numbers of refugees. The UK will also have to decide whether to retain a select number of EPA-like arrangements for the relatively richer or larger countries, such as Pakistan, that currently receive relatively generous preferences from the EU.
Both Mendez-Parra and the EBA+ proposal have merits—in particular that they extend duty-free, quota-free access to several non-LDC African countries and thus avoid the need for EPAs in eastern, central and western Africa where they have proved difficult. The EBA+ option, however, would be simpler and would retain the focus on poorer countries.
Developing a More Development-Friendly Template for Free Trade Agreements
The EPAs that the EU has negotiated in Africa and elsewhere assume that deep integration is the ultimate goal and that the template should be largely the same for developing countries as for more developed economies, at least in the longer run. The interim and steppingstone EPAs are primarily a vehicle to maintain preferential access for the signatories in the short to medium run. But they envision an eventual outcome of broadly reciprocal obligations with commitments in areas well beyond border barriers. Both assumptions, which the United States shares in its negotiations of FTAs, are questionable. In both cases, the efforts aim to get partner countries to adopt regulatory approaches in covered areas that are similar to those of the EU or the US. And the partners with no alternatives, such as the EBA, sign because they have no choice.
Developing countries push back on many of the demands in EU and US FTAs and EPAs because they are not convinced these demands are in the country’s interest, given their size and level of development (Elliott 2019b). Strong intellectual property laws in countries with little domestic innovative activity to protect mostly just result in rent transfers from poor countries to richer ones. The threat of being hauled before an independent arbitration panel and potentially having to pay millions of dollars in compensation can mean that investor-state dispute settlement mechanisms deter vulnerable countries from regulating in the public interest. And broader regulatory harmonization is neither democratic nor economically efficient when countries’ priorities or tastes and preferences differ—to wit, the European rejection of US exports of chlorine-washed chicken or hormone-fed beef.
So, in cases where they may be useful, what kind of trade agreement would be more appropriate for developing countries? Overall, it would be preferable to focus on tariffs and other border measures, along with nondiscriminatory treatment of foreign investors, perhaps with modest exceptions for industrial policy measures. When it comes to services, UK negotiators should prioritize capacity building for developing country policymakers over pushing for access to sectors where regulatory institutions may not be ready to cope with an influx of foreign companies (Messerlin 2009). Especially for poorer developing countries, but perhaps for others as well, negotiators should leave out most of the chapters dealing with domestic regulatory issues relating to Intellectual property, sanitary and phytosanitary standards, labor, and the environment. Exporters would still have to meet British product standards, but British negotiators would not try to force harmonization of those standards to suit their own preferences. Capacity building to promote better regulatory processes, including verification of compliance, would be more useful for developing country partners.
A more streamlined approach might also be more politically palatable. The recent debate in the United States over the renegotiated North American Free Trade Agreement—now called the US-Mexico-Canada Agreement—underscores the growing public concerns with the potential for such agreements to intrude too far beyond the border. Before agreeing to vote on that agreement, the Democratic Party majority in Congress insisted on removing or amending provisions that had become standard in US trade agreements, including strong patent protections for pharmaceuticals and the investor-state dispute settlement mechanism for foreign investors. Similar issues had triggered earlier public protests in Europe during negotiations with Canada and, until President Donald Trump suspended them, the Transatlantic Trade and Investment Partnership talks between the EU and the US.
This approach does not mean that all WTO+ issues should be ignored, and there is a growing demand for trade agreements to support the provision of global public goods—especially climate change mitigation. Thus, there would still need to be mechanisms to address violations of international norms when trade is implicated, for example, allowing countries to block imports of counterfeit trademarked products or those produced with forced labor or other egregiously exploitative labor practices. Trade agreements also need to reflect the accumulating evidence that a climate crisis is already upon us. But in situations involving global public goods, multilateral approaches will be far more effective and should be the priority. Trade agreements can play no more than a supporting role.
Overall, using trade agreements with smaller, weaker countries that have little bargaining power to pursue harmonization of standards as desired by the more powerful country is unlikely to be either effective or equitable. For the most part, trade agreements should be about trade again.
Applying the Template to the African Continental Free Trade Area
Embracing a more development-friendly FTA template should make it easier to eventually negotiate a bilateral trade with the African Continental Free Trade Area (AfCFTA) group—when they are ready, and if such a bilateral agreement is still appropriate at that time. If the multilateral system is bolstered and operating effectively (see below) and the AfCFTA has succeeded in spurring growth and development on the continent, one could hope that, by the time both African and British policymakers are ready to negotiate a bilateral trade agreement, it will no longer be useful.
That is possible because it is looking as though it may be a rather long time before the AfCFTA is fully implemented and the Africans are ready to negotiate on a continental basis with anyone. As of October 2019, all the countries except Eritrea had signed the AfCFTA but barely half the signatories had ratified it. The long history of the many, overlapping regional economic communities across the continent does not give much reason for confidence that the AfCFTA will rapidly achieve its goals of eliminating tariffs and other internal barriers to trade. Thus, the goal of concluding a trade agreement with the AfCFTA to replace what will likely still be a mix of unilateral preferences and subregional FTAs—hopefully less complex than what the EU has—is likely to be something that is only possible in the long run.
In the meantime, the UK’s goal should be to do no harm to the AfCFTA efforts, as the EU approach has arguably done. That is, avoid FTAs or EPAs or other arrangements that divide rather than help unite the continent. On the more proactive side, British policymakers should ensure that the rules of origin in whatever preference programs or FTAs it negotiates support the goal of continent-wide integration. The EBA rules, which are easier to meet for key sectors such as clothing and fish products, along with global cumulation—covering beneficiaries of all preferential arrangements—would be the best way to achieve this. Rules of origin that allow cumulation only among African countries would be a second-best approach, but it risks raising production costs and reducing benefits relative to what could be achieved with global cumulation.
UK Leadership to Support the Multilateral System
The WTO is currently under enormous strain and the UK is a small, open economy that relies more heavily on the rules-based system than larger trading partners, such as the EU and the US (Jones and Copeland). For that reason, it should join with other similarly placed countries and push for a new multilateral negotiation to help rebuild confidence in the WTO’s ability to function. For this proposal to have any chance of success, there will have to be a restoration of leaders in the United States that are more supportive of multilateralism. But other countries, including major emerging markets, need to be willing to cooperate as well. Not only that, to serve the confidence-building goal, it is important for all WTO members to contribute and, for this particular negotiation, that there be as little special and differential treatment for developing countries as possible.
Given the strains—and the goal of minimizing special and differential treatment—this negotiation should take a back to basics approach and focus on relatively simple goals in the short run. The key elements in the proposed negotiation would entail all members:
binding their tariffs (including LDCs),
converting all tariffs to ad valorem equivalents,
eliminating nuisance tariffs (say, < 2 percent), and
setting de minimis thresholds for low value shipments on which duties will not be assessed (say, < $100).
The General Agreement on Tariffs and Trade (GATT), which was incorporated into the WTO, was created as part of the post-World War II economic system to help unwind the tangle of tariffs that had exacerbated the Great Depression. One of the core obligations of this agreement calls on members to bind their tariffs at levels they negotiate with other members. Once bound, countries cannot raise tariffs above those levels except under exceptional circumstances as defined in the agreement (Jackson 1991, p. 40). Developed country members generally bind most or all of the tariffs and most of them have applied tariffs that are close to those bound levels. Special and differential treatment allows developing country members, and LDCs in particular, to retain more flexibility in their tariff schedules and to bind fewer tariff lines. This flexibility, however, comes at the cost of increased uncertainty about the future level of those tariffs. Under this proposal, developing countries might retain tariffs at relatively high levels, but they would at least agree to bind them at those levels.
Non ad valorem tariffs insulate domestic producers from global price changes, making them relatively more trade-distorting. With an ad valorem tariff, the duty is calculated as a percentage of the value of the imported product. So, for example, a 10 percent tariff on a $100 item equals $10. With a specific tariff, which is the most common non ad valorem tariff, the duty is an amount per a set quantity of imports. For example, a tariff set at $10 per ton of imported wheat would be equal to an ad valorem tariff of 10 percent when the price of wheat is $100 per ton, but if the price falls to $50 per ton, the relative tariff cost rises to 20 percent. If a specific tariff is expressed in terms of the local currency, traders also face an exchange rate risk with specific tariffs. Some countries also use mixed or compound tariff formulas that are more complex and often opaque, further increasing uncertainty regarding their ultimate impact. Other tariffs use technical specifications that can be incredibly complex, as with EU tariffs on some food products.
Low tariffs, typically in the range of 2 to 3 percent, are often called nuisance tariffs because they cost more to collect than they raise in revenue. These tariffs are generally too low to provide any protective effect and eliminating them would save governments money and allow them to concentrate scarce resources on more important tasks, such as reducing drug trafficking or customs fraud. Analysis of tariff schedules from the WTO reveals around 2,000 tariff lines that G20 countries have bound at 2 percent or less. The European Union, United States, and China account for more than half of those, with Australia, Canada, Japan, and Korea accounting for most of the rest. The value of trade covered by those tariff lines in 2016 was almost $1 trillion, not a trivial amount. Some of that trade occurs on preferential terms, either under unilateral trade preference programs, such as the Generalized System of Preferences, or as part of bilateral or regional trade agreements. Even in those cases, however, elimination of nuisance tariffs would have benefits by reducing paperwork associated with rules of origin that establish eligibility for preferential access.
Similar to eliminating nuisance tariffs, many countries set de minimis thresholds for low-value shipments on which it can cost more to assess duties than to forgo them. The Trade Facilitation Agreement, which aims at reducing unnecessary costs associated with trade, calls on WTO members to set de minimis thresholds below which import duties would not apply, but it suggests nothing about the desirable level of such thresholds. The Trade Facilitation Agreement also explicitly excludes value-added and other internal taxes from application of this provision. This is consistent with GATT Article III, which requires member governments to treat imports the same as domestic products for purposes of domestic regulation and internal taxes.
Setting de minimis thresholds for low-value shipments could lower trade transactions costs—paperwork and brokerage fees—beyond the savings associated with not having to pay the import duty. It would also save government resources and it would allow customs authorities to focus on higher-value and riskier shipments. Tax fairness arguments support the Trade Facilitation Agreement approach of focusing on import duties and not VAT or other internal taxes. But doing so would also reduce the benefits from reducing paperwork and lowering the brokerage fees associated with getting products through customs.
Is this agenda politically feasible? Analysis of available data shows that developed and developing countries alike would have to contribute under these proposals. With respect to tariff bindings, about half of all WTO members have bound all of their tariff lines and roughly two-thirds have bound 95 percent or more of the tariffs in their schedules. Those with binding levels below 95 percent are overwhelmingly developing countries. Two-thirds of WTO members also already use only ad valorem tariffs. For more than half of the others, non ad valorem tariffs constitute less than 2 percent of their tariff lines. Developing and developed countries alike use non ad valorem tariffs, but they are particularly a problem in developed countries with high levels of protection for agriculture. On nuisance tariffs, developing countries tend to have relatively higher tariffs and, if the developing country members of the G20—other than China—are representative, they have few, if any, bound tariffs at low enough levels to qualify as nuisance tariffs. With respect to de minimis thresholds, data compiled by the Global Express Association for 70 countries plus the European Union shows that only six countries, all developing, had no de minimis threshold at all. Another 29 countries, including both developed and developing, set thresholds at less than or equal to $50, and 12 had de minimis levels between $50 and $100. There were 23 countries, at varying income levels, with thresholds over $100.
In sum, developing countries would contribute relatively more in terms of binding tariffs and developed countries would be most affected by the elimination of nuisance tariffs, while countries at all income levels would be affected by the proposals to convert tariffs to ad valorem forms and set de minimis thresholds for low-value shipments. In general, these are low-cost reforms that countries at all levels of development should be able to implement. The one possible exception is the last item, on de minimis thresholds, which does require a certain level of capacity in the customs service to guard against false invoicing and illicit shipments. At the same time, exempting low-value shipments from the paperwork associated with levying tariffs frees up resources to focus on riskier and higher-value shipments.
The point of negotiating a back-to-basics package of measures would be as a confidence-building measure to help restore the WTO’s role at the center of the global trading system. It would not have a major impact on market access around the world. So, would it be worth it? This package would build on the Trade Facilitation Agreement and would help to fill in the remaining gaps in the shallow integration agenda. But it would also address 21st Century issues by lowering transactions costs in ways that facilitate trade across global value chains and for small and medium enterprises. And that should be of as much interest to American, European, and Japanese firms as it is to firms from emerging markets and developing countries.
This package would also help to increase transparency, predictability, and reduce the cumulative costs of goods that often cross borders multiple times as they move across globally fragmented supply chains. And it would reduce some of the fixed costs of trade, thereby providing disproportionate benefits for small businesses. Eliminating nuisance tariffs and setting de minimis thresholds at higher levels than currently exist in many countries would also save governments time and money and allow them to focus scarce resources on more valuable or riskier shipments.
The main goal, however, is to restore confidence in the rules-based trading system. That is something where the post-Brexit UK policymakers can and should take a leading role.
The EU’s trade policies for development have many positive elements, but they fall short in key areas, particularly with respect to Africa. As it develops an independent, post-Brexit trade policy, the UK can become the model for using trade to promote development, as well as lowering costs for consumers.
Improving on the situation should begin with an improved duty-free, quota-free preference program that is simpler, covers more countries, and is more development-friendly. The EBA+ program recommended here would provide duty-free, quota-free market access to all current LDCs, as well as other relatively small low-income and low-middle-income countries—those with populations of less than 70-75 million people.
The UK can also do better in situations where policymakers decide that EPA-like trade agreements are necessary to maintain market access for current preference beneficiaries that are not eligible for the new EBA+ program. Both in Africa and more broadly, British policymakers should use a different template for trade agreements with developing countries that recognizes their diversity and puts development objectives ahead of commercial goals. Such agreements should focus on more narrowly trade-related issues and reject recent trends towards regulatory harmonization as a key goal of trade agreements.
Finally, as a (relatively) small, open economy, the UK shares an interest with most developing countries in having a strong, rules-based multilateral trading system. After Brexit, it should join with other similarly situated countries to take the lead in pushing for a multilateral negotiation that could help rebuild confidence in the WTO. The focus should be on a back to basics agenda that seeks to make trade more simple, transparent, and predictable and to which all WTO members, regardless of level of development, agree to contribute.
For annexes and references, please view the linked pdf file.
 In particular, see CGD (2010) on global trade preference reform; Crawfurd, Mitchell and Anderson (2017) on post-Brexit British trade and development policies; and Timmis and Mitchell (2019) on EU trade policy towards Africa.
 World Trade Organization et al. (2018). Also see the discussion of the trade component of CGD’s Commitment to Development Index here, https://www.cgdev.org/commitment-development-index-2018#CDI_TRA, accessed 8 January 2020; see also Elliott (2019b) for more on how US tariffs discriminate against poor people both at home and abroad.
 Article XXIV, which authorizes members to negotiate bilateral or regional free trade agreements under certain circumstances, is a growing exception to the MFN rule. Elliott and Temprano Arroyo (2019) explain these exceptions in more detail.
 See Elliott (2009) for a similar proposal.
 British policymakers could also improve generally on the ad hoc and uncertain approaches to LDC graduation that is the default for most preference providers; see Elliott (2019a).
 See Crawfurd et al. (2017), Timmis and Mitchell (2019) and Elliott (2009) for specific proposals in these areas.
 Morocco and Tunisia are small LMICs and would be eligible for the duty-free, quota-free scheme.
 Armenia and Sri Lanka could also lose GSP+ preferences as they recently moved into the upper-middle-income category and thus would not be eligible for the EBA+ proposed here.
 See Elliott and Temprano Arroyo (2019) for an analysis of how GSP could be used to incentivize and support livelihood opportunities for refugees and their host communities.
 This is based on Elliott (2020), which details each of the four items recommended for negotiation.
 There are other urgent reform needs to restore WTO credibility, most notably fixing the dispute settlement system, but multilateral negotiation is a function that has been neglected for some time and desperately needs a boost. Updating the rules would also help lessen the pressures on dispute settlement.
 Note that this proposal would not prevent countries from applying tariffs at levels below where they are bound, as many developing countries currently do.
 A note prepared by the WTO Secretariat during the Doha Round of trade negotiations provides definitions and examples of the different non ad valorem formulations. The note was prepared for the negotiations on agricultural liberalization and it also provides information on the distribution of non ad valorem tariffs on agricultural products across countries and products. See WTO (2004).
 India initially also challenged the incentives relating to labor rights and the environment, but later withdrew that part of the complaint.