Trade, Globalization and Sustainability Impact Assessment

A Critical Look at Methods and Outcomes

Paul Ekins, Tancrede Voituriez, Paul Ekins, Tancrede Voituriez

• Publisher: Routledge
• Year: 2012
• ISBN:
  9781136551116
• Edition: 1
• Topic: Jura
• Subtopic: Internationales Handelsrecht

Trade liberalization, as promoted by the World Trade Organization (WTO), has become one of the dominant drivers and most controversial aspects of globalization. Trade sustainability impact assessments (SIAs) were introduced as a means of generating better understanding especially of the social and environmental impacts of trade liberalisation, and of making those impacts more consistent with sustainable development. This book takes a hard look at the experience of Trade SIAs to date, and the extent to which they have achieved their objectives and improved the outcomes of trade negotiations. It proposes several ways in which Trade SIAs could be made more effective, and illustrates these in respect of controversial sectors in which trade liberalisation has been implemented or proposed, including commodities, services and investment. Finally the book makes proposals beyond SIA through which some of the conflicts between trade liberalization and sustainable development could be more effectively addressed.Written by top researchers and experts on trade SIAs, this book is vital for researchers, academics, post-graduate students and policy makers working on any aspect of impact assessment, international trade or globalisation more generally. In addition, the book will provide a particularly useful background for those considering how the environment and trade interrelate at both global and regional levels, with some particular insights on climate change and trade policies.

Part 1

The Context: Trade, SIAs and
Development

1

Trade-induced Changes in
Economic Inequality: Assessment
Issues and Policy Implications for
Developing Countries

Sylvain Chabe-Ferret, Julien Gourdon,
Mohamed Ali Marouani and Tancrède Voituriez

Introduction

The Doha Development Round stresses the need to rebalance the expected gains from freer trade toward developing countries. The development dimension given to the round echoes a double-meaning acceptation of fairness in the trade liberalization process. First, World Trade Organization (WTO) member countries are now committed to design trade liberalization agreements such as to correct past unfairness, according to which developing countries were prevented from developing as much as they ought to through effective trade openness among their partner countries. Secondly, trade liberalization should be fair according to the consensual meaning given to ‘development’ by international aid agencies and UN bodies for about the last ten years, which equates development with poverty reduction. A fair trade liberalization round should hence equally reduced poverty.

Trade economists have spared no effort to check that under the various scenarios on the table before the WTO Hong-Kong Ministerial in December 2005, expected gains from freer trade provided by trade models actually matched the expectation of fairness placed on the round in progress (Anderson and Martin, 2006; Hertel and Winters, 2006; Polaski, 2006). Computable general equilibrium (CGE) models highlighted the cost of protection and of distorting supports for the very countries that did resort to such policies, making the most protectionist countries the most beneficiaries from trade liberalization. Countries or regions such as the EU, Japan and the US hence came first out of the hat, which was not the kind of fairness or rebalancing effect one would have expected from a genuine development round. Estimates of possible changes in poverty headcounts induced by trade liberalization gave such tiny figures on their side – particularly when compared with the first assessments made about five years ago – that taunting comments began to appear, mocking the Doha pro-development posture, when ‘much ado about nothing’ would have conferred indeed a more appropriate title on it (Ackerman, 2005; CEPR, 2005). Ironically, because of the wide-scale implication of trade modellers in the advocacy of trade liberalization since the onset of the development round, the uneasiness and awkwardness of the Doha Round in fulfilling its development mandate seem to pervade the modellers’ work, suggesting at least possible fallacies in the message conveyed by their estimates to trade negotiators, non-governmental organizations (NGOs) and the media (Panagariya, 2004).

What indeed came out of nearly a decade of debate on the trade-growth-poverty-inequality nexus? Ex-post evaluations based on cross-country studies fuelled harsh methodological controversy; they displayed weak evidence of a positive trade-and-growth linkage (Rodriguez and Rodrik, 1999), while the estimated impact of trade opening on income inequality turns out to be small because trade liberalization seems to favour the relative demand for skilled labour (Anderson, 2005). On the ex-ante modelling side, estimates of aggregate welfare changes from trade liberalization produced by static CGE models nourished blunt criticism for ignoring most – not to say all – market failures crippling developing economies (Stiglitz and Charlton, 2005).

Another sharp criticism of standard CGEs is their reliance on the representative agent hypothesis that impedes them from analysing the impact of trade liberalization on income distribution. Bourguignon et al (1991) and Cogneau et al (2003) proposed new methodologies, taking into account households’ behaviour, to analyse the impact of policies on households. By linking CGEs with microsimulation models, such methodologies permit the analysis of the impact of trade liberalization on income distribution and poverty.

As a result, trade liberalization could be poverty-alleviating in the long run and on average, while it is acknowledged that trade liberalization brings about distributional changes that may make the poor even worse off in the short term in particular countries, and, notably, in the poorest ones (Winters et al, 2004). Then there is a third dimension of fairness in the trade liberalization process: according to whether a country faces short-term or long-term gains, its political capability to rally the support of its population for joining the liberalization project – and hence benefit in due time from effective gains from freer trade – will differ dramatically. A fair liberalization round should place them on an equal footing and hence take into account not only the distribution of gains and losses among countries and among households, but as well its distribution over time. This raises the issue of dynamic modelling and its weaknesses, especially in the way expectations are treated.

The starting point of this paper is given by country situations where trade liberalization is expected to be poverty- and inequality-alleviating in the long run while inducing a short run increase in poverty or in inequality. We hence focus on a trade-induced social injustice case, which is a different animal from the ‘loud losers’, lobby-based explanation of government reluctance to move down the liberalization road we are used to finding in political economy analyses. In this latter case, short-term gains do exist but they are politically risky to tap, which is not the case we embrace here. The question we ask is what are the distributive aspects of trade that are worth documenting to help governments better integrate trade policies within a global policy framework so as to enhance growth and reduce poverty and inequality. The method followed is a literature review, organized according to three different interpretations of fairness implied by the ‘development’ objective of the world trade liberalization agenda. Because a ‘pro-development’ trade liberalization agenda should first correct past unfairness in the trade regime, which raises the broad issue of country-level trade liberalization's ex-post impact assessment, we start in the next section with the review of the main findings of country-level ex-post assessments. But a ‘pro-development’ trade liberalization agenda should also reduce poverty at household level, a point we address in the third section. Last, because development is basically a dynamic process, the distributive-dynamic effects of trade liberalization are also considered. Across all these three definitions of fairness, the development objective of the Doha Round proves to be an objective that trade liberalization cannot systematically achieve. A synthesis of our ten main results concludes the paper.

Fairness interpretation one: Country-level ex-post assessment

The ‘ex-post’ empirical evaluation of trade liberalization's impact on inequality over the last decade provides interesting but no clear-cut results. Two main approaches have been followed, assessing:

1 wage inequality in the manufacturing industry between unskilled and skilled labour, using time series analysis;

2 aggregate inequality, measured by the Gini coefficient on various sources of revenue (land, capital, wages) on a cross-country basis.

These two approaches build upon the Heckscher-Ohlin (HO) model to test predictions of changes in income inequality among developing countries. Assuming that unskilled labour is the relative abundant factor in developing countries, trade liberalization should increase its relative returns when compared to capital and skilled labour, and hence reduce inequality. The results of studies on wage inequality reject HO predictions for developing countries in Latin America in the process of trade reform. Results in Asia are more heterogeneous. Concerning aggregate inequality, the first studies on global inequality that basically test the impact of openness in developing countries do not exhibit robust results either, producing insignificant effects, or rejecting the prediction, except in Calderon and Chong (2001). With both approaches, initial tests did not conform to the theory: namely the wage premium for skilled workers and overall inequality often increase in developing countries when trade is liberalized.¹

Faced with this puzzle, authors have improved their theoretical approach and empirical assessment methods.² Several routes are liable to explain the increase of the skill premium and the widening of global inequality. All deal with heterogeneity among developing countries, be it heterogeneity in human capital endowment, heterogeneity in natural resources endowment, heterogeneity in outsourcing and foreign direct investment (FDI), or heterogeneity in technology. For each of them, outcomes and salient results are listed below. Unaddressed issues complete our review.

Heterogeneity in human capital endowment

We briefly review explanatory arguments as well as some test results that such arguments might have led to.

Argument one: One should consider heterogeneity in developing countries’ human capital endowment, on the ground that some developing countries may not actually display a comparative advantage in unskilled-labour-intensive goods.

Wage studies: To explain the difference of liberalization's impact on wage inequality between Latin American and Asian countries, a possible candidate seems indeed the timing of trade policy reform. At the time when Latin American countries started to liberalize, they were no longer unskilled-labour abundant, contrary to East Asian countries that liberalized at a time when they were unskilled-labour abundant (Wood, 1997). Several studies (Harrison and Hanson, 1999) on wage inequality in Latin America provide evidence that unskilled-labour-intensive sectors were protected with the highest tariffs prior to trade reform. Such industries have experienced the largest tariff reductions in the process of trade reform. Hence ‘the increase in the skill premium’ matches trade theory predictions: provided that trade liberalization focused on unskilled-labour-intensive sectors, the economy-wide return to unskilled labour predictably shrank.

Gini studies: When testing the impact of trade openness accounting for human capital endowment, Spilimbergo et al (1999) and Fischer (2001) show that developing countries that were relatively less endowed in human capital experienced a lower increase in inequality after trade liberalization. Gourdon et al (2008) do not reproduce these results when taking into account heterogeneity in data sources and using different indexes of trade liberalization.³ Nonlinearities in the relationship between human capital and inequality during trade liberalization seem to prevail.

To summarize, studies accounting for heterogeneity in human capital endowment do not refute the fact that trade liberalization benefits the relatively abundant factor in developing countries. They basically argue instead that all developing countries do not display a comparative advantage in unskilled labour, contrary to a widespread assumption.

Argument two: Different types of unskilled labour coexist in developing countries (basically educated and uneducated), which requires detailed assessment of trade impacts along them.

Wage studies: Wood (1994) argues that North-South manufacturing trade not only raises the wage of workers with basic education level relative to that of uneducated workers, but that it also raises the wage of skilled workers with basic education relative to uneducated skilled workers. This is mainly due to the impossibility for uneducated workers to be hired in export-oriented manufacturing activities.

Gini studies: Milanovic (2005) shows that trade liberalization increases income inequality in low-income countries but decreases inequality among middle-income economies. Milanovic interprets this result as a trade-off between liberalization and education: openness in developing countries might increase inequality by helping those with basic education, and leaving even further behind those with no education. The lowest income deciles begin to benefit from increased labour demand only when the poor become reasonably skilled. Gourdon et al (2008) differs from Spilimbergo et al (1999), by showing that (relative) abundant endowment in uneducated labour increases inequality when a country opens to trade whereas (relative) abundant endowment in basically educated labour significantly reduces it.

To summarize, taking into account heterogeneity in human capital endowment across developing countries explains that increased openness will only lead to an increase in basically educated labour demand, and in turn in its remuneration, while the demand for uneducated labour will fall, magnifying the skill premium effect.

Heterogeneity in natural resources endowment

In the literature on inequality, natural resource endowment is viewed as a possible factor that inverts the basic HO prediction.

Wage studies: Abundant endowment in natural resources may lead to wage inequality in manufacturing since processing industries of primary goods are more skill- and capital-intensive than low-skill manufactures. Bourguignon and Morrisson (1990) corroborate this intuition on a set of countries from Asia, Latin America and Africa.

Gini studies: Theory suggests that openness should lead to an increase in natural resource returns in countries where this factor is relatively abundant.

Leamer et al (1999) show that an increase in exports of primary commodities is positively correlated with income inequality, but they do not control for a country's relative abundance in natural resources. Large export volumes in primary commodities may indeed reflect high endowment in unequally distributed natural resources and cause inequality upsurge independently of trade openness. Spilimbergo et al (1999) and Fischer (2001) control for relative abundance in natural resource endowments. Their results indicate that while natural resources significantly increase inequality, trade liberalization in a land-abundant country has no clear effect. The fact is that the distribution of natural resources is as important as their relative abundance. For instance, in a country like Brazil where land is unequally distributed, openness might lead to an increase in inequality. Suc...