This is an old revision of the document!
Joseph Stiglitz and Andrew Charlton, 2005, Fair Trade for All: How Trade Can Promote Development, Oxford: Oxford University Press
The Uruguay Round completed in 1994, establishing the World Trade Organisation (WTO) and drastically expanding the scope of multilateral law concerning global trading rules. In the decade since its completion, consensus has emerged that the round heavily favoured developed countries and many developing countries have suffered unexpected costs in the implementation of its provisions. The collapse of the WTO ministerial conference in Seattle in 1999 — where a new round of negotiations had been due to be launched — brought the rich world into this consensus and convinced developed countries that the WTO would have to move in a distinctly new direction if it were to convince its developing members that a new agreement was in their interests.
That consensus was formally announced at the next ministerial conference in Doha in 2001 in the form of the Doha Declaration. This conference launched the so-called 'Development Round', with developed countries promising to prioritise a rebalancing of the interests of developed and developing countries. However, developed countries as a bloc have reneged on those commitments, setting an agenda for Doha negotiations with only minimal concessions to development. This agenda focused particularly on the 'Singapore Issues', three subjects on which the WTO established working groups, and one which the WTO Goods Council was mandated to examine at the Singapore ministerial conference in 1996:
Developing countries, wary of entering another bad agreement after the failures of Doha, have blocked progress on developed countries' target areas. The US has largely responded to the stalemate by switching its attention to bilateral agreements with individual developing partners, a process which developing countries recognise maximise the US' power advantage.
This book attempts to outline what a true agenda for a development round should contain, based on a review of theoretical and empirical economics. The main proposals for reform of the trading system are detailed in Chapters 6 to 9. Chapter 13 examines the adjustment costs that, for some of the smallest and poorest countries, are likely to outweigh any potential benefits of multilateral liberalisation.
Trade liberalisation is supposed to deliver gains as resources are transferred from protected sectors in which a country does not have comparative advantage, to those sectors where it is more efficient and where it can export more successfully. But in developing countries the lack of resources (labour and other production inputs) available to new industries is not usually the constraint which prevents the growth of new export sectors. Developing countries have vast reserves of resources, particularly labour, which are already unemployed or underemployed. Thus trade liberalisation isn't required to 'free up' these resources for use in new industries. Unless complementary policies are used… liberalisation may, by removing the protection given to domestic industries, just leave the workers and other resources used in formerly protected industries idle in the short run. —p6
The proposition that international trade is welfare-enhancing is one of the most established in economics, dating at least to Adam Smith (1776) and David Ricardo (1816). It has certainly played a part in the development of many countries, from Britain through Japan, North America and Australia, and more recently many Asian economies. These examples tend to show that trade is necessary but insufficient for sustained development.
However, whilst economics textbooks discuss the difference between the binary states of autarky (no trade) and free trade, real countries have had to make much more subtle choices between an infinite number of slightly varying trade regimes representing different levels of overall liberalisation. Theory is much more equivocal when it comes to these choices.
Almost every country retains some form of tariff restriction, and yet the world has been moving progressively towards greater liberalisation since the GATT (immediately following the Second World War).
Some of the developed countries that have been the most ardent advocates of trade liberalisation have been somewhat duplicitous in their advocacy. They have negotiated the reduction of tariffs and the elimination of subsidies for the goods in which they have a comparative advantage, but are more reluctant to open up their own markets and to eliminate their own subsidies in other areas where the developing countries have an advantage. As a result we now have an international trade regime which, in many ways, is disadvantageous to the developing countries. —p12
A particular twist of economic theory that has been used by the developed countries is to pretend that unilateral liberalisation by developing countries in itself is good, regardless of whether its trading partners are also undergoing liberalisation relevant to its export interests. However, the great majority of benefits of free trade for a developing country come from free access to foreign markets for products which it can export competitively; so unilateral liberalisation brings few benefits (and many problems) — the benefits of its own liberalisation fall mainly to its trading partners.
The next section will examine the assertion that trade is good for welfare and good for growth with reference to a few country histories.
The modern East Asian growth phenomenon began with Japan in the immediate post-war period, growing to be the second-largest global economy within a few decades. Many neighbours successfully followed similar development models, first South Korea, Taiwan, Hong Kong and Singapore, and more recently Thailand, Indonesia, Malaysia and China.
East Asia successfully refuted two classic propositions of development, proving that:
There were important differences between the policy approach of different countries.
They also had much in common:
Most controversial has been the role of the state in the East Asian miracle. These countries followed some orthodox prescriptions and ignored others, leading both orthodox and unorthodox commentators to claim that their success hinged on policies that they advocate.
Whereas, against orthodox opinion:
Laissez-faire economists argue that the latter policies had no effect or were harmful, and that Asian success was a consequence of the former. Some even argue that growth would have been stronger without state intervention, although this is a “particularly unpersuasive counterfactual”3), as no country has ever achieved higher growth.
One thing is clear, that the Asian economies were highly successful without a strict adherence to the Washington Consensus and that controversy certainly remains about the effectiveness of different trade and industrial policies, and controls on capital flows in different contexts.
In this context it is inappropriate for the world trading system to be implementing rules which circumscribe the ability for developing countries to use both trade and industry policies to promote industrialisation. The current trend to force a narrow straightjacket of policy harmonisation on developing countries is simply not justified by the available evidence. —p17
Many Latin American countries were inspired by the experiences of the main belligerents of World War II, who all abandoned free market pretensions in favour of intensive state planning and coordination to achieve the massive increase in output required by the war effort. The economic success of Soviet Russia and the wartime economies stood in stark contrast to the largest free-market economic failure in history across Western Europe and the US in the decade prior to hostilities.
This practical experience was supported by development economists who argued that massive government intervention was required in a 'big push' to benefit from economies of scale and demand from workers in other industries. Government intervened to:
Obviously, this policy of favouring industries in which the country had no comparative advantage directly contradicts Ricardo's theory. Their critique of comparative advantage was that it's too static: it predicts a (static) welfare gain, but says nothing about economic growth or the way in which comparative advantage develops over time. The areas (particularly agriculture) in which developing countries had a comparative advantage seemed unlikely to benefit from as much productivity growth as industrial sectors: accepting their comparative advantage appeared to mean consigning themselves to a future of ever-increasing relative poverty.
[T]he theory of comparative advantage told South Korea, as it emerged from the Korean War, that it should specialise in rice. But Korea believed that even if it were successful in increasing the productivity of its rice farmers, it would never become a middle- or high-income country if it followed that course. It had to change its comparative advantage, by acquiring technology and skills. It had to focus not in its comparative advantage today, but on its long run, its dynamic comparative advantage. And government intervention was required if it was to change its comparative advantage. —p30
Latin American countries grew rapidly in the decades of import substitution (averaging 6 per cent across the region during the 1970s), but then one country after another fell into crisis in the early 1980s. They defaulted on their debt and suffered a decade without growth.
The neoliberal conclusion was that Latin America had failed due to its inferior policy: state intervention to support national industries had left them uncompetitive and inefficient, whilst permanent state support had lead to runaway inflation and unmanageable debt.
The opposing view argues that Latin America's failures had more to do with external shocks than with domestic policy. According to the South Centre, they simultaneously experienced:
Consequently, the blame lies with a combination of poor debt policy combined with unfortunate global circumstances. They borrowed heavily in the 1970s to avoid the global recession that followed the oil price shocks and were hit hard when the Federal Reserve raised interest rates to unprecedented levels. This reading of history is supported by the fact that all of the Latin American countries, regardless of whether they were experiencing significant problems with import substitution, were hit to a similar extent at a similar time — if the policy were to blame, a differential ability to survive the shock would be expected.
This alternative view suggests that it was Latin America's open capital markets rather than is relatively closed trade policy which led to the lost decade… LA's reliance on foreign capital flows and FDI are what made it particularly vulnerable to global economic shocks. —p22
In 1994, Mexico entered the North American Free Trade Agreement (NAFTA) with the US and Canada. This was an opportunity for laissez-fair orthodoxy to prove the wonderful benefits of free trade: Mexico had access to the largest economy in the world, with no serious transport or infrastructure limitations.
Ten years on, the results have been mixed. The benefits:
However, in many ways NAFTA has failed to live up to its advocates' claims:
Three lessons should be taken from the experience:
Aside from the traditional argument that trade brings a welfare gain through comparative advantage, there are four further channels through which liberalisation can bring costs and benefits:
Traditional arguments for liberalisation usually focus on welfare rather than growth — ie the long-run rate of growth will not be affected.
However, the assumptions which yield the result that free trade is superior to any intermediate range of trade restrictions (Samuelson, 2962) are restrictive and often inapplicable to developing countries. The standard argument is that liberalisation improves efficiency: it drives inefficient local producers out of business, and these resources are absorbed by efficient industries. However, this assumes that the economy is initially operating at full capacity — extremely rare in developing countries. If there is spare unused capacity, then rather than transfer resources from low- to high-efficiency industries, instead previously used resources will become unemployed. On the contrary, it is not necessary to shift resources in order to build export industry, merely to mobilise unused resources. Concern about increased unemployment is one of the greatest concerns in practice, particularly understandable where safety nets are weak. It's problematic, therefore, that the models that predict welfare gains blithely assume full employment.
The second key assumption is the existence of perfect risk markets. However, international markets are highly risky, and liberal trade policy increases exposure to that risk. For instance, in a closed domestic market, if some shock reduces production of a certain good, the price will rise (due to scarcity), mitigating the loss of revenue of that producer. If the market is global, prices will not be affected by domestic production — they are likely to remain constant, they may even fall for unconnected reasons. Under free trade, risk-averse firms may prefer low-efficiency production if it protects them from the risk of higher-efficiency activities.
Under quite plausible conditions, one can show that free trade is Pareto-inferior to autarky — everyone is worse off — which is just the opposite result to that of the conventional wisdom. —p26
A third assumption is in the effectiveness of the price mechanism. However, in developing countries markets are often absent, and the price mechanism fails. Where one industry depends on inputs from another not-yet-existing industry, liberalisation may enable circumvention of the problem by importing inputs, but many inputs are non-tradables, so both industries may need to be established simultaneously, but the price of the inputs and goods of each will not be known until after that establishment. Where scale economies exist, this situation may require a restriction of competition from foreign producers to overcome irremediable market failures.
Government revenue can also provide a rationale for trade taxes, in certain circumstances. The orthodox view is based on the Diamond-Mirrlees production efficiency theorem (1971): for a small open economy it is optimal for government to raise taxes on household demand rather than border taxes. But the assumptions are restrictive and don't hold in developing countries. Often trade taxes are far less costly to collect than other forms; where this is so they may be optimal. Recently, VAT has been pushed hard by the international institutions, but where large informal sectors exist beyond the reach of such taxes (as in all developing countries), these may be welfare-reducing relative to trade taxes.
Trade liberalisation also affects inequality: there are winners and losers. Theoretical models assert that everybody “could” be made better off under liberalisation if the winners compensated the losers — but this almost never happens. Liberalisation is therefore often seen as “unfair”. Such compensation, theoretically, is assumed costless, but in fact the costs of such redistributive schemes may well be prohibitive.
Another type of market failure is based on information externalities. Market information is scarce and highly costly in developing countries — for instance, it is normal that the only means of determining whether a certain business model is profitable is to try it (with one's own capital, given the failure of capital markets). The potential value of demonstrating a business model is profitable can be enormous, but very often it is impossible for the entrepreneur to capture more than a fraction of this value. If the model fails, the entrepreneur pays all of the costs; if it succeeds, copy-cat firms capture some if not most of the gains. There are compelling real-world examples: cut flowers in Colombia, apparel in Bangladesh, software in India — in each case the entrepreneur that demonstrated the profitability of the model received a miniscule fraction of industry profits.
[I]t is no great surprise that low-income countries are not teeming with entrepreneurs engaged in self-discovery. –Dani Rodrik, 2004, cited p29
Perhaps more important than the impact on welfare is liberalisation's impact on long-term economic growth. Perhaps liberalisation will increase growth:
These tend to emphasise the importance of knowledge, learning and human capital, but these are areas that often require extensive government intervention to push an economy towards comparative advantages that can capitalise on this long-term development.
The infant industry argument states that a new industry requires a period of 'learning' during which it cannot compete efficiently. It attempts to justify limited protection by government until firms in a new industry have completed this learning. Its critics argue that if the long-term profitability of the industry is larger than the short-term losses, then those firms should be able to finance their learning by borrowing.
There are two reasons that this criticism may be flawed. Firstly, asymmetric information leads to capital market failure.
Banks would have to be willing to lend to enable firms to sell below cost, in the hope that by doing so their productivity will increase so much that they will become a viable competitor. It should be obvious that such loans would be viewed, at best, as highly risky. (And banks would likely view themselves at a marked informational disadvantage in judging whether the firm's claims about its future prospects — in spite of large current losses — have credibility.) —p32
Secondly, if there are spillovers and some of the gains of learning are captured by rival firms (for instance, if employees leave to set up competitor firms), then the gains to the pioneer firm may not cover the cost of investing in their learning.
Once capital market imperfections are taken into account, then protection may be optimal, as Dasgupta and Stiglitz (1980) show. They argue, in particular, that protection may have advantages over other instruments, eg subsidies, when government has limited sources of income and limited ability to target.
An alternative approach is to create strong intellectual property rights that protect the pioneer firm both from foreign and domestic competition.4)
The economic growth literature has been successful in demonstrating the importance of some variables for economic development, including education, institutions, health and geography. However, the relationship between trade liberalisation and growth is much more controversial. –32
Many authors have attempted to establish a reliable relationship between greater openness and higher growth. A few studies in the early nineties claimed to have established the relationship, but they themselves contained careful qualifications, and subsequent scrutiny has persuasively shown that they should be interpreted with extreme caution.
In particular the studies were identifying the negative effects of macroeconomic imbalances, instability and geographic location and misattributing them to trade restrictions. —p34
Their conclusions were not, therefore, supported by the data. It should also be noted that research has focused on openness rather than liberalisation.
Certainly, the bullishness of voices like the IMF in assuming an established relationship between liberalisation and growth cannot be supported by empirical evidence (but then international trade is not within the IMF's remit, so there is no reason that it should understand trade policy).
On the right, claims that forthright unilateral liberalisation by developing countries will lead to welfare gains and economic growth are not supported by empirical evidence and often contrary to historical experience. The reforms they recommend, and in some cases coerce, “might in fact set development programmes back.”5)
On the left, a resistance to all calls for reform and an attempt to seek a complete answer to developing countries' economic problems in reform by developed countries may also be counterproductive.
Many in the left interpret the right's focus on liberalisation despite the lack of empirical evidence to support their case as evidence of the right's malevolence. This is unfair. The prescriptions are based more on the belief amongst right-wing economists that developing country governments are incapable of implementing anything but the simplest trade policy successfully.6) More complex regimes are also more likely to be corruptly implemented.
Alan Winters, Director of the Development Research Group at the World Bank:
“the application of second-best economics [ie activist trade policy] needs first-best economists, not its usual complement of third- and fourth-raters.” —p37
In the authors' view such concerns are real but overemphasised: look in particular to East Asia, where government officials proved highly competent in implementing effective and innovative activist policies. Instead of assuming government constraints to be unassailable, these constraints can be combated through good institutional design and international technical assistance.
The historical evidence that activist trade policy has always almost been used during the 'take-off period' of industrialisation, which Ha-Joon Chang has documented for all developed countries, still applies to the most recent industrialisers (although the timelines are a little more compressed than in historical examples). China began to grow rapidly in the large 1970s, but did not begin tentative trade liberalisation until the late 1980s; Indian growth increased in the early 1980s while tariffs were still rising in some areas and did not fall significantly until the early 1990s.
The General Agreement on Tariffs and Trade (GATT) came into force in 1948 between 23 developed countries. It was largely a response to the spiral of trade policies adopted by major economies during the Great Depression. Countries raised tariffs against their neighbours in a 'beggar-thy-neighbour' cycle. Each policy was rational in the absence of a cooperative agreement, but all policies together left all countries worse off. GATT sought to establish a transparent, reciprocal system of tariff limits, which would benefit all parties and help to protect the global economy from depression (assisted by the Bretton Woods institutions).
At its inception, the needs of developing countries were not considered. The agenda of GATT negotiations focused entirely on issues of relevance to trade between developed countries. The GATT underwent further rounds of negotiations (the most recent to complete was Uruguay in 1994), but up until the end of the 1970s there was little expansion in scope beyond goods trade between developed countries; instead, negotiations merely focused on agreed reductions in tariff levels by all parties. In the 1960s and 1970s, this tariff reduction was largely forced by the US in response to progressing European integration that threatened to cut the US out of European markets in favour of other European producers. Goods trade of interest to developing countries was mostly excluded (in particular, agricultural products and textiles fell outside of the GATT). As more developing countries gained independence it became possible for them to participate, but there was little motivation for them to do so:
Agriculture and textiles were covered by separate agreements, whose provisions were starkly different to those of the GATT. Under the Multifibre Arrangement (MFA) developing countries bargained bilaterally for textiles quotas. Developing countries as a bloc were actively discriminated against, as developed countries could export to one another without restriction. Massive tariff protection and subsidies of agriculture remained throughout the developed world.
In the 1980s, many developing countries began to take a more active role in trade negotiations, with the launch of the Uruguay Round in 1986, for several reasons:
The Uruguay Round marked a significant expansion in the trade-related agenda, including the following areas for the first time:
These issues were all of interest primarily to developed countries, although their offensive interests were now both developed and developing markets. As manufacturing falls to only 14 per cent of US GDP, there had been a danger that in the long term the GATT might have proved to have been more use to China than to its original authors. In return, the developing countries were promised greater goods access in areas of interest to them (again, particularly agriculture and textiles).
As the Uruguay Round was nearing completion, many developing countries were convinced that it was in their interests to complete because of the ambitious predictions of welfare gains made by international institutions:
In practice, these predictions proved to be over-optimistic, particularly concerning the gains to developing countries.
Indeed many of the poorest countries in the world will actually be worst off as a result of the round. Some estimates report that the 48 least developed countries are actually losing a total of $600 million a year as a result of… Uruguay. —p47
Sub-Saharan Africa is estimated to have lost around $1.2 billion — many of the net losers are amongst the poorest countries in the world.
One reason for the predictions' failures is that they didn't correctly model the implementation and detail of the agreement. They failed to predict the way in which the US would backload textiles commitments to reduce their value. Market access improvements, once again, reflected the priorities and interest of developed countries, with the following result:
[A]fter the implementation of Uruguay Round commitments, the average OECD tariff on imports from developing countries is four times higher than on imports originating in the OECD. —p47
The commitments made by developing countries in the Uruguay Round were unprecedented, and it is now clear that few countries had any idea of the cost of implementing this wide range of new commitments (the cost of doing so was also ignored in projections of the net benefits of the Round). Developed countries had offered financial and technical assistance in meeting these commitments, but crucially developing countries had made bound commitments, whereas developed countries had made only non-binding indications that they may be willing to assist, with the result that developing countries had to beg for money every time they struggled to meet a commitment to which they were legally committed.
Unsurprisingly, in the aftermath of Uruguay, developing countries became disillusioned with the agreement. The promised benefits failed to materialise, whilst commitments in areas such as intellectual property were having unanticipated and disastrous effects, partly due to the extreme way in which they were being interpreted by developed countries.
Although Uruguay had laid the basis for the inclusion of agriculture within the GATT, the focus had been on converting non-tariff barriers to tariffs (tariffication). This could be done in such a way that even after tariff reductions, applied tariffs did not fall (and water remained in many tariffs), the result being that Uruguay did not bring significant reductions in agricultural tariffs. Similarly with subsidies:
In 1986–8 farm subsidies were equivalent to 51 per cent of all OECD farm production, and fourteen years later, after the implementation of Uruguay commitments… they still accounted for 48 per cent[.] —p50
Even in industrial goods there was scope for further reduction — as noted earlier, tariffs on developed country exports remained many times higher than developing country exports (averaging 0.8 per cent compared to 3.4 per cent). These figures hide important peaks:
This type of tariff escalation discourages basic processing operations by creating enormous effective tariffs on value adding, and are acutely anti-development. Along with such peaks they are “manifestly unfair”.7)
Experience with the new areas of coverage illustrated that the agenda had been highly biased towards developed countries. GATS negotiations focused on sectors of strategic importance to developed countries (such as finance and business services), whereas services that could realistically be supplied by developing countries were neglected (in particular, the movement of workers has an enormous bias towards the very high-skilled). The Singapore Issues, in the way in which they were formulated, were also of interest only to developed countries.
The Doha Round began in 2001 with the Doha Declaration laying out an ambitious developmental work programme ahead. However, progress in the two years to the next ministerial meeting (Cancún, 2003) was minimal. US and EU offers on agricultural tariffs and subsidies fell far short of developing countries expectations; meanwhile, the US passed the 2002 Farm Bill, further increasing support to American farmers, seemingly a demonstration of America's total lack of commitment to the spirit of agricultural liberalisation. The US and EU pushed the Singapore Issues hard, despite their clear lack of development benefits. The EU and US both unilaterally introduced new tariff preferences for least developing countries (LDCs, a UN classification distinct from developing countries, which covers approximately 50 of the poorest countries in the world) — respectively 'Everything but Arms' (EBA) and the 'African Growth and Opportunity Act' (AGOA). However, despite the rhetoric, the tangible results of these schemes has been negligible at best: “Brenton (2003) concludes that trade in goods given preferences for the first time under the EBA in 2001 amounted to just two hundredths of one per cent of LDC exports to the EU in 2001.”8)
The failure to make progress and strong feelings that the US had reneged on its Doha Declaration and earlier commitments were exacerbated by the negative public reaction to US-WTO attempts to use TRIPS provisions to prevent African governments from sourcing cheap generic drugs to combat the AIDS pandemic. The effects of increasing US subsidies on West African cotton farmers drew further public attention.
In 2001, rich countries provided subsidies to their farmers which amounted to six times the amount of their development aid, respectively $311 billion and $55 billion… The African countries' demands, including that the subsidies be reduced and compensation be paid to African farmers, were swiftly rejected by the US. —p62
One consequence was the formation of unprecedentedly strong new negotiating blocs amongst developing WTO members: the G20 leading negotiations on agriculture, the G90 united in opposition to Singapore Issues and the G33 pressing for the ability of developing countries to protect their agricultural issues. Greater coordination brought greater strength, and the Cancún ministerial ended in failure when developing countries refused to accept further discussion of Singapore Issues or EU-US proposals on modalities for agricultural liberalisation.
In mid-2004, the EU and US were ready to make limited concessions in the hope of salvaging the round, and all Singapore Issues apart from trade facilitation were dropped: the round would focus on core issues in agriculture, services and goods. However, by the end of 2004, the Doha Round has made little progress, and has failed to live up to the ambition contained in the Doha Declaration.
The concept of a 'Development Round' poses a fundamental challenge to the GATT-WTO system. Throughout its history, the GATT negotiation process was motivated purely by self-interested bargaining. Each party participated only because it saw a benefit to itself, and lobbied only in favour of provisions that it saw as in its interests.
The concept of a 'Development Round' requires a fundamental departure from this approach. It requires the adoption of shared principles that can be used to judge whether any given provision is in the interests of “development”, whether or not it is in the individual interests of any party. As yet the WTO has made no serious attempt to address this issue — it has barely been discussed, and certainly no progress has been made towards agreeing a set of principles. But without such principles motivating a new approach it seems unlikely that the US-EU will be able to negotiate a 'Development Round' that also satisfies their concept of their rational self-interest — particularly if the Round also seeks to redress the unbalanced outcome of Uruguay. Following the historical GATT approach, the US-EU will refuse any genuine development agreement.
The authors propose four criteria by which to just a development round:
* Any agreement should be assessed in terms of its impact on development; items with a negative effect on development should not be on the agenda.
* Any agreement should be fair.
* Any agreement should be arrived at fairly.
* The agenda should be limited to trade-related and development-friendly issues.
Present and historical negotiations have not been lead by impact assessment. Where they are done, they are done by individual parties to inform their own negotiating positions only.
The WTO Secretariat should take on the function of providing impact assessment of proposals before they are negotiated, to determine whether they are development-friendly. Assessments should be done on a country basis, and should also take into account differential impacts on groups within each country, particularly vulnerable groups.
This assessment cannot be based on neoclassical models, which are pathologically ill-suited to developing economies:
The importance of adjustment costs also implies that WTO members should avoid bilateral agreements on the way to multilateral agreements, as these are likely to create short-term trade diversion.
WTO members also need to be honest about the fact that some provisions are redistributive rather than good, individually, for every single party. Complex theories of political economy are currently used to explain why countries are reluctant to liberalise unilaterally, but at least in many cases, it should be openly recognised that countries simply don't benefit from their own liberalisation, but only from the liberalisation of other markets.
Intellectual property rights are distinct. They increase the incomes of those in advanced industrial countries at the expense of those in less advanced industrial countries. Arguments that they lead to greater innovation are not backed by empirical evidence. Drug companies in particular develop drugs for developed markets, intellectual property rights in poor countries seem not to motivate them to greater research relevant to distinct health threats in these areas. In any case, if and when they are good for development they can be adopted unilaterally by developing countries.
The authors do not attempt to justify why a development agreement should be fair. There does seem to be a growing international consensus, in rhetoric at least, that it should.
Opponents of the principle argue that trade negotiations are voluntary, and therefore beneficial for all — if developing countries lost out from participation in the WTO, they would leave. There are problems with this argument:
It should be noted that, to the extent that countries are different, an agreement that applies uniformly to all countries may affect each country differently — it will not necessarily be fair. For this reason, it is important to adopt the following principle:
Any agreement that differentially hurts developing countries more or benefits the developed countries more… should be presumptively viewed as unfair… any reform [should] be progressive, ie… a larger share of the benefits [should] accrue to the poorer countries. This was almost certainly not true of the Uruguay Round. —p75–76
One key problem with this criterion is that agricultural subsidies pose massive costs for developed countries, so their elimination is likely to benefit the developed economies more than others. To address this, the only benefits and costs that are considered should be those that are granted to others (ie costs and benefits of reform for oneself should be ignored).
Fairness also requires that two failures of the dispute settlement system be rectified:
The EC, Japan and the US were complainants in almost half (143 of 305) of all bilateral disputes… between 1995 and 2002. By contrast the 49 members classified by the UN as Less [sic] Developed Countries did not bring a single challenge in that period.
A further difficulty in the application of fairness is whether to view WTO agreements in isolation from the actions of other multilateral institutions (asking that WTO agreements, in isolation, be 'fair'/progressive) or whether to attempt to ensure, through WTO agreements, that developing countries are treated fairly by 'the system' as a whole. For instance, if a certain developing country commits to a fair level of liberalisation under Doha and then, in the course of an IMF bailout, is forced into a further round of liberalisation at the insistence of the IMF, should the WTO be held responsible for the overall unfairness of the country's treatment? A more concrete example is the imposition of capital market liberalisation by the US through bilateral treaties. Capital market liberalisation increases economic volatility, which in turn raises the risk premium demanded by foreign investors. The IMF may also insist on “usurious”10) interest rates. If a government subsidises credit in order to bring interest rates down to a little over global market levels (in violation of WTO rules), should the government be held culpable and be subject to countervailing duties, or should it be regarded as simply correcting a distortion imposed by another part of the international system?
In the South, there is a tendency to view such multi-pronged attack as coordinated. Whilst such coordination may not exist, nevertheless the impact is often much the same as it would be if they were coordinated.
The high interest rates, tax policies, and trade liberalisation policies demanded by the IMF do exacerbate the adverse effects on developing countries of whatever trade liberalisation measures they agree to within the WTO. The two cannot be seen isolation. This provides the basis of one of the important recommendations that we make below.
Procedural fairness is particularly important when there is disagreement about exactly what constitutes a fair agreement. Its main components are transparency and the way in which negotiations are conducted. Historical GATT negotiations have been unduly secretive, particularly the 'Green Room' process used in the Uruguay Round, in which developed countries would hold substantive negotiating sessions to which only a handful of key developing countries were invited. Developing countries begin at an informational disadvantage compared to other members — this should be diminished to the extent possible by the open provision of information about ongoing negotiations.
In the Uruguay Round, the scope of the WTO's operations grew alarmingly. The WTO provides a convenient negotiating mechanism and ready-made dispute settlement architecture, but there are important reasons to be cautious about the expansion of the WTO's scope:
A 'principle of conservatism' needs to be introduced to the WTO to oppose this expansion in scope, consisting of three elements:
[M]odern trade agreements have been extended into areas which intrude into national sovereignty with no justification based on the need for collective action and without clearly identified and fairly distributed global benefits. The presumption of consumer sovereignty is based on the premise that society should only interfere with individual choices when those choices have consequences for others, when there is a need for collective action, and the same is true in trade. —p86
One of the most important challenges for a development round is to develop a system that treats less developed economies differently to more developed economies, whilst remaining predictable, rules-based and fair. In the WTO context, this is known as “special and differential treatment, or SDT. Considerable SDT has been a feature of the WTO system for decades. The WTO recognises that trade policies which are likely to maximise welfare in developed countries may not have the same effect in the very different economic environment of developing countries. Common existing SDT provisions include:
SDT is controversial. Whilst developing countries argue that SDT is necessary to permit them to pursue appropriate trade policies, opponents argue that:
SDT has always been an important demand of the developing countries in Doha, with the G33 its main proponent. After the Cancún ministerial walk-out, in 2004 the EU Trade Commissioner wrote to trade ministers he proposed that LDCs11) and “other weak or vulnerable developing countries” should not have to undertake further tariff reductions during Doha, whilst still benefiting from other members' reductions. These countries would have the “round for free” (RFF).
The danger of the RFF approach is that it would reduce the participation of the RFF countries in the round — indeed, its intention was surely to overcome opposition to developed country proposals by the RFF countries by reducing their incentive to get involved. The result might be that Doha came to mimic early GATT rounds in which the “GATT operated as a club for the advancement of rich-country interests.”12) Whilst it would enable poorer economies to benefit from new tariff reductions, these reductions are more likely to be on goods of interest to exporters in other developed countries, of limited relevance and benefit to RFF economies. An additional problem with the approach is that it allows RFF countries to retain a veto over a round in which they are contributing little — this could hold up negotiations, further encouraging the proliferation of bilateral agreements outside the WTO.
More importantly, the RFF approach neglects the opportunity for South-South liberalisation, which could bring developing countries real gains. South-South tariffs are currently a significant problem: Latin American manufacturers face tariffs seven times higher in other LA countries than in industrialised countries; East Asian exporters face tariffs 60 per cent higher in other EA economies than in rich nations. The World Bank estimates there could be welfare gains of $30 billion in each of agriculture and industrial goods from total South-South liberalisation.
In order to capture some of these gains, the authors set out a Doha market access proposal (MAP). This, conceptually, is very simple. Countries are ranked according to economic size (raw GDP) and wealth (per capita GDP): all members fully liberalise13) with respect to economies that are both smaller and poorer than themselves. For instance, Egypt would gain full market access to the US (which is both larger and richer) but would have to give full market access to Uganda (both smaller and poorer), whilst retaining MFN tariffs in both directions with Antigua (which is smaller but richer).
The principle is of a rules-based system based on objective criteria that is nevertheless progressive. Economic size (GDP) is used as a measure of economic scale, recognising that large economies (like China) benefit from economies of scale relative to smaller countries. Wealth (GDP per capita) is used as a measure of advancement (in technological and organisational terms), recognising that developing countries may find it difficult to compete in the short-to-medium term with more advanced competitors. All countries retain the ability to use trade policy against countries that are either larger or richer, but open up to those both smaller and poorer. The scheme has many advantages:
Although not discussed explicitly, details such as adjustment periods, provisions for specific sectors and many other complexities would have to be negotiated as part of this overarching policy.
There is now wide recognition that WTO “behind the border” rules may have gone too far. Many policies proscribed by the Subsidies Agreement and TRIMs were used, arguably to good development effect, by the East Asian economies during their rapid industrialisation. The TRIPS agreement contains much that will make it more difficult and expensive for developing countries to adopt new technologies from more advanced competitors, locking in the technology gap.
Developing countries should demand exemptions from such multilateral rules, including in proposed Doha agreements on competition and investment. These agreements would restrict developing country governments from pursuing advantageous policies. Further agreements to establish binding good practice in behind the border areas should be undertaken on a (voluntary) plurilateral basis, not as part of the compulsory “single undertaking” of the WTO.
Compulsory commitments should not be made in areas where the purported gains are controversial and where implementation and opportunity costs are high. And where the gains are certain and the costs are low for developing countries (such as allegedly in efforts to liberalise some aspects of FDI regulations), there is no need for compulsion, because developing countries can be persuaded of the benefits of taking unilateral action. —p104
A blanket proscription against government subsidies to technology (industrial policies) is likely to have an adverse effect on developing countries and, indeed, it is likely in practice to be unfair: the United States conducts its industrial policy largely through the military, which supports a wide variety of technological developments that eventually have important civilian applications. And it is hard to conceive of a trade agreement that would prohibit the development of such technologies through defence programmes. (Even the EU has complained about America's use of defence expenditures as a hidden subsidy for its aerospace industry.) —p105
The problem, of course, is that political globalisation has not kept pace with economic globalisation: issues of international trade agreements are seldom looked at through… the same kind of lens through which we look at domestic legislation In national economic debates, we do not demand that the poor give up and amount commensurate with what they get. Rather, we talk about social justice and equity. —p107
The current focus of Doha, on areas of dubious benefit to developing countries and neglecting important developmental areas, is inappropriate for a development round, and ignores the historical context in which developed countries have accrued far greater benefits from the multilateral system than the poor. There is an important asymmetry of power in the negotiations: the impact of concessions in developed countries will be small and easily manageable, whereas the impact of opening up markets in developing countries is large and will place significant strain on their governments.
The authors' main proposals:
In other words, reciprocity should not be the basis of negotiations.
The authors present further, less central, proposals:
The chapter contains a detailed table of further proposals (table 7.1, p112–4), which itself summarises the analysis presented in the appendices. Some highlights (particularly those not discussed elsewhere):
The basic rationale for trade liberalisation is to increase economies of scale and improve efficiency in the allocation of resources. Efficiency gains stem from the discrepancy in factor payments across countries — where factor payments vary, an opportunity exists to improve global productivity through liberalisation. It is worth noting then, that, assuming factor payments equal marginal products, the largest discrepancies exist in unskilled labour, followed by skilled labour, and lastly capital. In other words, an agreement that liberalises the movement of unskilled labour will do the most to improve global efficiency. In addition to the efficiency gains, the movement of unskilled workers would create large development benefits.
Notwithstanding this, liberalisation of unskilled labour has lagged far behind the liberalisation of goods and services generally. It fits into the WTO framework as under the GATS as “mode 4”, the “movement of natural persons”. So far, mode 4 commitments have lagged behind the other three modes, and have focused on the most skilled workers.
In order to maximise the development benefit of migrant labour, governments (and potentially the WTO) can get involved in improving systems of remittances to make the remittance of wages cheaper, safer and more convenient (at present a large portion of remittances flow through informal channels).
There are further GATS sectors that are of particular interest to developing countries in modes 1 and 2 (the expansion of offshore outsourcing has already lead to rapid growth in some of these sectors):
There are also non-market access reforms that could provide development benefits. Tourism is particularly important to many developing countries. Although the sector is usually liberal in developing economies, they nonetheless suffer from “rampant” anticompetitive practices in the industry, primarily in the North, which “minimise spillover and multiplier effects”.16) A multilateral anti-trust framework in tourism and maritime transport could create development benefits and support further liberalisation.
A similar situation exists in mode 3 liberalisation — many developing countries have unilaterally liberalised FDI rules, but a WTO agreement limiting the negative effects of competition between governments to provide the most attractive financial incentives to foreign investors could increase the development benefits of this liberalisation.
A final opportunity for improvement exists to provide greater scope for governments to adopt appropriate social policy. Developing countries have experienced negative consequences in the liberalisation of sectors such as the media and utility provision — governments need the space to protect vulnerable groups, prevent price rises caused by the market power of foreign firms and guarantee local content in appropriate industries (particularly the media).
Chapter 3 discussed the high levels of agricultural protection in the OECD and the negative effects on developing country production. However, the effects are ambiguous: by lowering commodity prices the subsidies hurt producers (often poor farmers) whilst benefiting net consumers (often better off urban classes). But the balance varies from one product to another — so reductions should be targeted to prioritise the most development friendly, and proceed more gradually (with substantial adjustment assistance) in areas that will have large impact on poor, vulnerable net-importing countries. Liberalisation should proceed thus: - Begin by reducing border protection (tariffs and export subsidies), with reduction most rapid on goods produced primarily in developing countries and those consumed primarily in developed countries (eg sugar, tropical products, cotton). - Production subsidies on price-sensitive necessities consumed widely in developing countries should be reduced gradually, with part of the savings in subsidy budgets committed to developing countries to cover adjustment costs (North Africa, Sub-Saharan Africa and Latin America (except Brazil, Argentina, Mexico) rely on imports of grains and oilseeds subsidised in the OECD). - Domestic support should switch to payment systems that do not encourage production (such as land-based payments). It's important to recognise that most “non-trade distorting” subsidies actually encourage export. Where domestic demand is price inelastic, subsidies that encourage production will also encourage export (the WTO presently makes too much of the difference). Additionally, where farmers face credit constraints, even income support can provide additional finance that farmers can use to expand production, which they could not have sourced elsewhere.
Developing country interests have been neglected in previous tariff liberalisation, and consequently serious problems remain, particularly
For example, in 2001, clothes and shoes accounted for only 6.5 per cent of US imports in value terms but they brought nearly half of the $20 billion of US tariff revenue.
In analysing the impact of escalation, it is important to calculate the effective tariff rate on value adding; analysis based only on nominal rates will not reveal the true development impact. Effective tariff rates on value added in food processing, for instance, is very high.
Additionally, the liberalisation of South-South tariffs could bring large welfare gains, as discussed in Chapter 6.
Legal commitments to liberalise do not eliminate protectionist sentiment or the “politics of special interests”.17) Accordingly, as the North has reduced tariffs through WTO liberalisation, its use of non-tariff barriers as a replacement has escalated. Although non-tariff barriers are primarily used by developed countries, their use by larger developing countries is also rising. They take four forms:
The general problem is that non-tariff barriers can be applied too easily by developed countries:
America's misuse of dumping duties illustrates both problems. The country accused of dumping must respond in a short period of time to a demand for extensive information in English. When unable to comply, the US government acts on the “best available information” (BIA) — ie that provided by the injured American firm. Large tariffs are applied, which are often later reduced, but can nevertheless be effective in the long term in two ways:
The calculation of “cost” used to determine whether a firm is dumping is also inappropriate. Many countries use average costs, whereas economists agree that the relevant measure is marginal cost — a firm with high fixed costs, selling below marginal costs during a downturn should not be considered to be dumping. Secondly, the use of a “similar” comparator country in order to approximate production costs can be misused — for instance, “the US used Canada as the country most similar to Poland.”18)
Safeguards are overused by the United States and probably underused by developing countries.
If the richest country in the world, the United States, which a strong safety net, relatively high employment level, etc, has to resort to safety measures to protect itself against a surge of imports, how much more justified are developing countries in imposing such measures. Indeed, it is hard to conceive of many important liberalisation measures against which safeguard protections could not justifiably be invoked by developing countries.
Clearer standards are required at the international level to determine when safeguards can be applied, and these should include SDT for developing countries. For instance, a member should have to show that the loss must be of at least 1 per cent of all employment and that the country's social safety net will struggle to absorb it, with lower thresholds for developing countries.
Reform requires three components:
The determination of what constitutes a subsidy also requires clarification to rectify current problems:
Development-friendly behind-the-border measures that should be included in a Doha agreement:
The fact that the US and EU put [the Singapore Issues] on the agenda and continued to push them for so long within the so-called Development Round is of concern: were they merely bargaining chips? Was there no real comprehension about what should be meant by a Development Round? —p141
The authors believe that the balance of intellectual property rights in the US and embodied within TRIPS favours rights holders more than is beneficial for economic efficiency and development. Intellectual property rights create a form of monopoly and are therefore always statically inefficient — the trade-off is between static inefficiency and the dynamic effects of greater incentive to innovate. They note that the basic research that drives new technology is conducted by the state at public expense and would be no more successful with stronger protection. State support is justified because knowledge is a public good.
There are many development dangers from IP law that is too strong:
Before NAFTA, Canada routinely granted compulsory licenses on pharmaceutical products for the purpose of reducing health costs through widely available generic drugs. Canada assigned royalties to the patent holders, usually of 4 per cent of the generic competitor's sales price. In sharp contrast, and despite the HIV/AIDS public health crisis, no African country has issued a compulsory license for any medicine. —p143
The authors believe that TRIPS causes such comprehensive harm that it should be rolled back. Because of the principle of conservatism discussed in chapter 5, IP should not be discussed within the WTO; governments should be represented by scientists in any discussion of IP, not trade ministers. Any new agreement should recognised that developing countries are disadvantaged in any legal dispute process.
If TRIPS is not to be eliminated, then it needs to be revised in various areas, and developing countries need further support to fully exploit its provisions. Particularly:
Discussion in Doha has centred more on improving access of developed country firms into developing markets than on making markets more competitive or improving access by developing country firms into developed markets. However, there could be value in a competition agreement that harmonised anti-trust law at the highest common denominator. This ought to feature:
According to the conservatism principle, investment should not be part of a WTO trade agreement. It is also unbalanced to negotiate an investment agreement without concomitant labour and environmental agreements. Reasons not to negotiate a pure investment agreement within the WTO:
If an agreement were to be negotiated, the most important development issue would to prevent a “race to the bottom”, so countries should be restricted in using:
as a means of attracting foreign investment.
Previous services negotiations had a considerable focus on financial services, which should be re-examined. Developing countries should be permitted to force banks to lend to underserved populations.
In at least some developing countries there are concerns that the purchase of local banks by foreign banks has reduced the flow of credit to small and medium-sized domestic enterprises, and thus impeded economic growth. (There is a long history of such concerns, evidenced in the US for instance by restrictions on interstate banking, intended to prevent New York and other money-centre banks from buying up other banks, thereby impeding regional, and especially rural, development.) —p152
Developing countries are concerned that new trade agreements will create additional barriers to the entry of their goods, particularly “blue tariffs” (impediments on labour standards) and “green tariffs” (environmental standards). Economic theory argues that weak standards do not generally improve a country's competitiveness, so such standards should not be included in a trade agreement. There are exceptions:
Whilst some argue that these would be better addressed through a channel other than a trade agreement, in reality there are few other options.
The US has recently attacked China for allegedly manipulating its exchange rate. The situation is so complex that in almost all cases it is impossible to definitively determine whether this is truly a deliberate government policy. The threshold for action should be very high, at the least that only high multilateral surpluses be admissible as evidence.
Article XII of the Marrakech Agreement provides for the possibility that non-members may join the WTO, but “for an organisation which prides itself on being 'rules-based', the accession process is remarkably vague.”21) In practice, developing countries as a bloc retain a veto on the accession package, and use this power to demand much greater concessions from acceding countries than have been made by existing members. As part of the Doha Round, in December 2002 the General Council adopted new guidelines for LDC accession: the process would be streamlined and simplified, and acceding LDCs would not be required to make commitments beyond existing LDC members. In practice these guidelines have been ignored. Empirical studies show that, over time, accession is becoming slower and more costly (in terms of commitments). Acceding members have been forced to accept the following:
> It seems strange that the WTO's developed country members should force acceding countries, particularly small and poor countries like Cambodia and Nepal, into such strong concessions. Grynberg and Joy (2000) suggest that the motivation lies in the developed countries' desire to create a precedent that can be applied to future negotiations. —p161
The WTO urgently needs to implement the December 2002 General Council decision to make accession more transparent, and to ensure that accession terms are comparable to commitments made by members at a similar level of economic development.
Whilst Doha has floundered, the US has been aggressively pursuing bilateral treaties. These are bad for development because
As discussed, there are large potential gains from the liberalisation of South-South trade. This could be achieved by permitting developing countries to establish preferential market access deals to one another, similar to the EU's Everything but Arms initiative. These would be more useful than similar schemes provided by, for instance, the US, because the political conditionality imposed by the US reduces certainty and therefore value, whereas a “South-South agreement would likely focus more exclusively on trade issues, and would not suffer from these abuses.”23)
The result [of secretive trade negotiations without parliamentary oversight] is agreements, like Chapter 11 of the NAFTA… or the TRIPS Agreement, which contain provisions that would probably never have been accepted by a democratic parliament with open discussion in a deliberative process. —p167
Reform in the way in which trade negotiations are conducted and the WTO is structured is urgent. Many features of the current structure of negotiations disadvantage developing countries. As noted in chapter 5, the fairness of an international regime depends on its procedures as well as its outcomes. The authors propose the following:
This chapter presents more supporting empirical information than previous chapters.
Understanding of adjustment costs is important because:
Liberalisation brings various different types of cost:
Adjustment costs disproportionately impact developing countries, because:
Studies quantifying adjustment costs in developed countries find that costs accrue mainly to workers, and that they are small compared with consumer benefits (eg Baldwin, Mutti and Richardson (1980) analyse a 50 per cent cut in US tariffs and find 90 per cent of costs are borne by laid-off workers (the remainder by capital stock) but that total costs are only 4 per cent of total gains — other studies give similar results). Evidence in developing countries is much weaker, but costs are likely to be much higher.
For example, liberalisation of the cashew market in Mozambique in the late 1990s led to the loss of 85 per cent of the workforce employed in the local processed-cashew industry. Recent evidence suggests that whole towns have shut down as a result of the factory closures. —p175
Costs in developing countries are exacerbated by:
The authors present a range of evidence suggesting that the benefits of non-reciprocal tariff preferences (both GSP and especially LDC schemes such as EBA and AGOA) are small, and more than offset by the likely benefits in MFN tariffs, especially by developing countries who don't have GSP schemes but do currently have higher tariffs than developed countries. One reason for this is restrictive rules of origin, particularly in the EU, which undermine the value of such benefits (the American scheme has much lower coverage but a higher rate of take-up — this probably reflects the costs of meeting the EU's RoO). However, the authors' analysis underestimates the value of EBA preferences as it uses Cotonou preferences as its counterfactual, not MFN tariffs (and Cotonou preferences expired in 2008). One reason that the GSP is ineffective is that it fails to compensate for a tariff regime that discriminates against products exported by developing countries — thus for the EU, Canada and Japan the (weighted) average duty paid under the GSP is still larger than the average duty paid under MFN.
In a sense the GSP only partially compensates for the discrimination by developed countries against the goods produced by developing countries. —p183
Adjustment costs vary not only by country but between different vulnerable groups within each country. For this reason (and because many ignore unemployment and weak capital and risk markets)26) most analysis currently undertaken is inadequate. A useful analysis must look at the sub-country level, take market imperfections into consideration, and analyse the effects on
Fiscal effects also vary a lot between countries. “Senegal pursued trade liberalisation in the mid-1980s, following which there were large revenue shortfalls. Lost tariff revenue combined with slow growth in trade volumes and weaknesses in economic management led to dire fiscal consequences.” The process had to be abandoned. Other countries have been fairly successful in replacing lost revenue. Institutions such as the IMF believe that switching from trade tax to VAT is welfare-enhancing, but recent theoretical research shows that in a country with an untaxable informal sector it is desirable to retain some trade taxes.
On the basis of proper per-country analysis, assistance must be provided to governments to improve safety nets for workers and to improve credit markets for firms (if necessary to counteract misguided IMF policy artificially inflating interest rates). Evidence suggests that even if firms' average cash flows improve, if credit markets are weak the adverse effects on losers can more than offset gains leading to overall losses. Technical assistance is currently inadequate and must be expanded and improved. Programmes must support countries in developing their own policies, unfettered by IMF conditionality, and new tax structures that undermine long-term growth based on an inadequate understanding of developing economies should not be pushed on governments. Assistance for LDCs must also help them to capture new (and existing) opportunities for export, by addressing supply constraints (often linked to capital markets and infrastructure) and product standards through participation in standard-setting and assistance for national programmes to measure and improve conformance.
Two appendices contain empirical reviews of market access and Singapore issues.