A service of the

Download article as PDF

This article is part of A New Economic Paradigm

After decades of liberalising international trade and a strong increase in the openness of most economies, international trade relations have lately come under increased scrutiny. In Europe, protests all but stopped the Canada-EU Comprehensive Economic and Trade Agreement (CETA) and might derail the planned Transatlantic Trade and Investment Partnership (TTIP) with the United States as well – should negotiations ever be revived. In the US, the negative consequences of increased trade with China, and of globalisation in general, have been associated with a turn towards populism and ultimately the election of Donald Trump as president in November 2016.1

Yet, it is not just the public that seems to be increasingly dissatisfied with the outcomes of trade liberalisation; economists have recently toned down their traditional cheering for trade liberalisation and have started to acknowledge that the expansion of international trade might have had some negative side effects. International organisations such as the OECD and the WTO have underlined the need for complementary policies to make sure that “trade benefits all”.

This article gives an overview of the shifting views and policies on trade liberalisation. It will start by outlining the “old view” on trade, dominant since at least the 1980s but visible already in policymakers’ approach to trade liberalisation since World War II. It will then analyse how the empirical outcomes of several decades of trade liberalisation are (and should be) interpreted, pointing out successes but also highlighting areas in which increased trade has disappointed expectations. Finally, it will describe how these changes in the perception of outcomes of past trade liberalisation are now informing policy recommendations on trade.

The old paradigm in global trade relations

Since World War II, the world has seen an almost continuous process of trade liberalisation. Eight successful trade rounds were conducted under the framework of the General Agreement on Trade and Tariffs (GATT) from 1947 until the 1980s, each one further liberalising world trade, culminating in the creation of the World Trade Organization (WTO) in 1995. Average tariff rates for major economies fell continuously from 15% in 1947 to below five per cent in 1995 (see Figure 1), while the number of countries covered by GATT agreements grew from 23 to 123 in the Uruguay round (which ran from 1986 to 1994). Whereas the first GATT rounds focused mostly on the reduction of tariffs, the focus shifted towards non-tariff barriers to trade beginning in the Tokyo round in 1973. The ratio of global trade (exports plus imports) to GDP rose from slightly more than ten per cent in 1950 to almost 20% in 1960 and about 30% in 1990 (see Figure 2).Trade liberalisation received an additional boost during the 1990s, both from historical events (the fall of communism in Europe and the opening up of China) and from changes in the intellectual discourse among policymakers and economists. In the famous Washington Consensus outlined by John Williamson in 1989, trade liberalisation was one of the ten points recommended to developing countries.2

Figure 1
Average tariffs worldwide, 1945-2010
Average tariffs worldwide, 1945-2010

Source: Author, based on R.C. Feenstra, A.M. Taylor: International Economics, Worth Publishers, New York, 3rd edition, 2014, p. 14.

During the first decades after World War II, trade liberalisation was mainly conducted through multilateral trade rounds. While free trade agreements (FTAs) and customs unions had been possible under GATT rules for a long time, their use remained relatively limited until the late 1980s (see Figure 3). Only when it became obvious during the Uruguay round that multilateral trade negotiations had become more difficult to conclude did the US supplement its trade strategy by also conducting bilateral and regional trade negotiations with the aim of signing FTAs. At the same time, European countries pushed their integration project and moved from a customs union to a single market.

As tariffs outside the agricultural sector had already been strongly reduced, FTAs turned to addressing non-tariff barriers, especially technical barriers to trade (regulations and technical standards which act as an impediment to international trade) and sanitary and phytosanitary measures. As shown in Figure 3, from the 1990s onwards, newly signed FTAs increasingly were “deep” agreements which either went further than WTO provisions in areas that were in principle covered by WTO rules (such as customs regulations or technical barriers to trade) or included rules for areas which were not covered by general WTO rules (such as competition policy or investment).

A final element of global trade liberalisation since the 1990s has been the acceptance of large, formerly communist countries as new WTO members (e.g. China in 2001, Vietnam in 2007 and Russia in 2012), which allowed these countries the same level of access to the markets of advanced economies that was granted to other WTO members.

Figure 2
World merchandise trade, 1950-2016
World merchandise trade, 1950-2016

Sources: Author, based on data from World Bank: World Development Indicators Online Database, available at http://databank.worldbank.org/data/reports.aspx?source=world-development-indicators; and WTO: World Trade Report 2013, Geneva, 2013, p. 47.

Theoretically, the case for broad and rather indiscriminate liberalisation of trade had been supported originally by trade models in the tradition of Ricardo, Ohlin and Heckscher, which show that countries gain from free trade. Yet, theoretical economics never claimed that trade liberalisation would do no harm: in the Heckscher-Ohlin model, the scarce factor of production in each country loses when trade is liberalised. Thus, if a country with a relatively large endowment of capital per capita (such as an OECD member) liberalises trade vis-à-vis more labour-abundant countries (such as India or China), it can be expected that wages in the former will fall while the return to capital will increase.3 As Rodrik points out, the losses are of absolute nature: wage earners in labour-scarce countries lose in real terms, regardless of their consumption patterns. Hence, trade will have significant distributional effects.4

However, the problem of potential losers has generally not been seen as one of primary importance, as many of the estimates of the impact of globalisation on inequality from the 1990s led to the conclusion that, empirically, the impact of trade liberalisation on inequality would be very minor.5 Moreover, it was assumed that whatever small losses there were would somehow be compensated. This assumption relies on policy conclusions on the welfare criteria defined by Kaldor and Hicks,6 according to which any policy measure can be interpreted as advisable as long as welfare gains are in principle large enough that losers could be hypothetically compensated for their losses.

In practice, providing assistance to losers of trade has been at best an afterthought. While both the US and the EU have instruments in place to provide assistance to workers and industries hit by trade liberalisation, the financial means have remained very limited. In the US, the Trade Adjustment Assistance amounted to a mere $861 million in 2016, about 0.004% of GDP.7 The European Globalisation Adjustment Fund of the European Commission is even more limited in volume, with a maximum annual expenditure of €150 million.

Under the headings of “New Trade Theory” (from the 1980s onwards) and “New New Trade Theory” (from around 2000 onwards), additional arguments supporting trade were made. In these models, free trade leads to more product variety for consumers (and hence more welfare) and to lower prices (due to economies of scale).8 Contributions from the “New New Trade Theory” added that trade liberalisation would improve productivity (and hence increase incomes), as more productive firms would be allowed to grow and less productive firms would be pushed out of the market. According to these models, free trade would be even more beneficial than the traditional trade models claimed.

Figure 3
Number and depth of trade agreements, 1951-2015
Number and depth of trade agreements, 1951-2015

Source: Author, based on data from C. Hofmann, A. Osnago, M. Ruta: Horizontal Depth: A New Database on the Content of Preferential Trade Agreements, World Bank Policy Research Working Paper No. 7981, 2017.

It is safe to assume, however, that policymakers were much less informed by the more elaborate theoretical discussions than by the way economists communicated these ideas to a broader public. The problem with this communication, to use Rodrik’s words, is that

it has long been an unspoken rule of public engagement for economists that they should champion trade and not dwell too much on the fine print.…[E]conomists can be counted on to parrot the wonders of comparative advantage and free trade whenever trade agreements come up. They have consistently minimized distributional concerns [and] they have overstated the magnitude of aggregate gains from trade deals.9

A good example of this one-sided communication is the parable of “Isoland” in the defence of free trade in the best-selling textbook by Mankiw and Taylor,10 aimed at first-year students of economics and business administration. In this story, Mankiw and Taylor compare free trade to the invention of a new technology to more efficiently produce consumer goods (in their example, olive oil), and attempts to restrict trade are compared to attempts to limit technological progress by law. The message is straightforward: anyone who wants to restrict free trade reduces welfare as clearly as someone who wants to prohibit technological progress.

Outcomes of the old paradigm in trade relations

Empirically evaluating the welfare effects of past trade liberalisation is problematic because trade liberalisation never happens in a vacuum; often cross-border investment flows have been deregulated at the same time, domestic labour market institutions have changed and technology has progressed.

Nevertheless, there are some widespread notions about the effects trade liberalisation has delivered. Undisputably, trade liberalisation has increased consumer choices around the world and has brought down consumer prices for many durable and non-durable consumer goods, especially in advanced economies. Trade liberalisation is also a major component of the success enjoyed by Asian emerging markets, including China, in raising their income levels closer to those of advanced economies and of radically reducing global poverty. From 1993 to 2015, the number of people living in poverty globally declined from 1.939 billion to 836 million, with the decline in China alone accounting for a drop of more than 600 million people.

It needs to be kept in mind, however, that the Chinese experience has not been the result of all-out trade liberalisation in China itself. Instead, it seems to have been central that China was able to export into the world market. While China has opened up significantly for (import) trade since the 1980s, the country remains rather protectionist in a number of sectors. Similar arguments hold for other Asian countries.

While these benefits are rather undisputed, there have been a number of aspects in which trade liberalisation has produced fewer benefits than expected. First, empirically observed benefits from trade liberalisation often turned out to be much smaller than what had previously been predicted. For example, just prior to the implementation of the EU Single Market, it was estimated that it would increase GDP by between 4.3% and 6.4%.11 According to more recent work, the positive impact of the creation of the single market in the EU and the liberalisation of network industries together amounted to only 1.8% of GDP by 2006.12

On the other side of the Atlantic, the US Congressional Budget Office estimated that NAFTA would increase US incomes by about 1.5%, US output by 0.25%,13 and Mexican output by 6-12%.14 Current ex post estimates show the positive welfare effect of NAFTA (which should be roughly similar to income effects) for the US to be 0.08% and for Mexico to reach 1.31%.15

Second, the costs of adjustment, especially after severe trade shocks, seem to be much larger than previously thought. After the rapid integration of China into the world economy and the surge of US imports from China, regions that produced goods which were exposed to strong import competition recorded strong and persistent increases in unemployment, lower participation rates as well as strong and persistent declines in wages in the non-manufacturing sector.16 These impacts seem to have been much more severe and longer-lasting than what economists had previously predicted. For Germany, it has also been shown that increased exposure to trade with China (and, in the German case, with Central and Eastern Europe) can have long-lasting negative effects.17

The impact of trade on growing inequality within developed countries has also been reappraised. While in general the academic literature had long claimed that the impact of trade liberalisation on inequality was small or even statistically insignificant,18 more recent research questions this.19 By adding more potential channels through which globalisation could affect domestic wages, newer models yield the conclusion that it is “likely that the rapid growth of trade since the early 1990s has had significant distributional effects”.20

Another issue of growing public discontent has been the feeling that national policy space has been reduced by trade agreements, especially by “deep” trade agreements. The debate about Brexit is a case in point here: at the heart of the EU treaties is a trade arrangement which has gone much further than just removing tariffs and quotas; it also pushes to harmonise rules and regulations to create a friction-free single market of goods, services, labour and capital. The downside of this regulatory convergence has been that national parliaments are no longer able to legislate on certain issues which have already been decided at the EU level. In the British referendum about leaving the EU, concerns about “taking back control” played an important role.

Similar arguments were brought forward in the debate about CETA and TTIP. In both cases, resistance in Europe centred around regulatory cooperation and investor-state dispute settlements, which were seen as undermining democracy and the power of the state to regulate.

Moreover, despite the success in catching-up observed in some (East) Asian countries, the majority of developing countries do not appear to have benefited from trade liberalisation. While both Latin America and Africa have seen large increases in their exposure to trade, these countries have by and large achieved unimpressive rates of GDP growth over the past several decades. Also, levels of poverty have remained stubbornly high, and the impact of trade liberalisation on poverty is ambiguous.21 A number of economists, particularly Chang, have claimed that this failure to catch up has been closely related to the downside of the current global trading system, which does not allow less developed countries to implement industrial policies necessary for successful industrialisation.22

Evidence presented lately suggests that even prior to World War II, trade globalisation might have contributed to a divergence rather than a convergence of living standards between countries. Pascali examines trade and GDP data during the first wave of globalisation (1870-1913) and shows that only the countries at the top 25th percentile of the global income distribution seem to have benefited from the increased trade globalisation during that time, while GDP growth in the other countries actually appears to have been depressed.23 Including measures of institutional qualities, Pascali shows that countries with high institutional quality have benefited from trade, whereas others have lost. While this is a relatively new conclusion, if proven robust, it would seriously call into question the previous consensus that all countries benefit from trade liberalisation.

Potential for a new paradigm

Overall, there seems to be an emerging consensus that trade liberalisation has not delivered what it promised. The OECD has published a number of papers since 2010 looking at the conditions under which trade might have detrimental effects on employment and wages. Most recently, high-profile policy papers by the OECD, the IMF and the WTO acknowledge that, in certain instances, there may have even been some negative effects of trade liberalisation.24 The organisations now also underline the need to supplement trade liberalisation with a set of additional policy measures.

Figure 4
Degree of current consensus for new policy recommendations on trade relations
Degree of current consensus for new policy recommendations on trade relations

Source: Author’s elaboration.

What exactly needs to be done, however, is still disputed. A number of policy proposals have been discussed recently that could muster varying degrees of support among economists and international institutions. Figure 4 tries to sort the main proposals by the degree of support they have received so far.

In detail, the main proposals are:

  • Increased spending on labour market policies, social safety nets and redistribution: More active labour market policies are needed, including early activation strategies, training programmes to address skill gaps and specific job search assistance, and – if well-designed – wage subsidies and supplements to help workers build experience in other sectors; a more generous social welfare system and a larger government sector can provide security for all kinds of losers in the economy.25 High inheritance taxes would help to redistribute from winners to losers.26
  • Macroeconomic stabilisation policies: Displaced workers can find a new job much more easily during good economic times, and therefore, macroeconomic stabilisation policies should be employed to guarantee robust and stable economic growth to make trade adjustment less painful.27
  • Regional policies: As regions hit by a trade shock tend to remain depressed and suffer from high unemployment for a long time, regional policies are necessary to help new businesses form in these areas.28
  • Shallower trade agreements: While not a mainstream position, some economists – such as Rodrik – argue that policymakers should not excessively focus on further trade liberalisation – and especially not on negotiating “deep” trade agreements – as efficiency gains from such liberalisation can be expected to be minor while distributional effects can be relatively large.29
  • Increasing cross-border labour mobility: Rodrik and Milanović argue that the potential benefits from allowing (some) more workers to migrate from poor to rich countries could produce huge benefits while only creating relatively small costs (in terms of negative distributional effects).30
  • Allowing developing countries more space for industrial policy: Chang proposes to allow developing countries a more gradual process of trade liberalisation and more leeway for industrial policy, be it through direct government support and subsidies or selective protectionism.31


Nonetheless, many unanswered questions remain. On the conceptual side, it is not clear how much of the latest disappointment with the outcome of trade liberalisation is due to singular events or factors such as the integration of China into the world economy or the macroeconomic environment after the global financial and economic crisis of 2008-09. It is also unclear why the US labour market, long seen as one of the most flexible among OECD countries, has been so bad at adjusting to trade shocks. Moreover, very little is known about the role and impact of cross-border value chains or growing trade in services on incomes and inequality.

With the underlying effects and mechanisms disputed, there is also limited consensus about the best remedies. While hardly anyone would object to more training and active labour market policies, many economists are reluctant to support calls for a slowing down of further trade liberalisation. Hence, there is a lot of research to be done in the area of international trade, and this subfield of economics promises to become even more exciting over the coming years.

* I thank the participants of a workshop on a “New Economic Paradigm” in Edinburgh on 20-21 October 2017, and especially Dorothée Rouzet, for valuable comments.

  • 1 D.H. Autor, D. Dorn, G.H. Hanson, K. Majlesi: Importing political polarization? The electoral consequences of rising trade exposure, CEPR Discussion Paper No. 11511, 2016; D. Rodrik: Populism and the economics of globalization, NBER Working Paper No. 23559, 2017.
  • 2 One has to concede, however, that Williamson in his original contribution is much less radical than those who used the consensus points later for policy advice. See J. Williamson: What Washington means by policy reform, in: J. Williamson (ed.): Latin American adjustment: How much has happened?, Washington DC 1990, Institute for International Economics.
  • 3 W.F. Stolper, P.A. Samuelson: Protection and real wages, in: The Review of Economic Studies, Vol. 9, No. 1, 1941, pp. 58-73.
  • 4 D. Rodrik, op. cit., p. 5.
  • 5 See, for example, the survey by N. Chusseau, M. Dumont: Growing Income Inequalities in Advanced Countries, in: J. Hellier, N. Chusseau (eds.): Growing Income Inequalities: Economic Analyses, London 2013, Palgrave Macmillan UK, pp. 13-47; or by P. Brenton: Rising Trade and Falling Wages: A Review of the Theory and the Empirics, in: P. Brenton, J. Pelkmans (eds.): Global Trade and European Workers, London 1999, Palgrave Macmillan UK, pp. 18-38; or as examples for this claim, P. Krugman, R.N. Cooper, T.N. Srinivasan: Growing World Trade: Causes and Consequences, in: Brookings Papers on Economic Activity, Vol. 26, No. 1, 1995, p. 327, or J.E. Haskel, M.J. Slaughter: Have Falling Tariffs and Transportation Costs Raised US Wage Inequality?, in: Review of International Economics, Vol. 11, No. 4, 2003, pp. 630-650.
  • 6 See for example the recent elaboration in P. Antràs, A. de Gortari, O. Itskhoki: Globalization, inequality and welfare, in: Journal of International Economics, Vol. 108, 2017, pp. 387-412.
  • 7 B. Collins: Trade Adjustment Assistance for Workers and the TAA Reauthorization Act of 2015, CRS report for Congress No. R44153, 2016.
  • 8 See for example M.J. Melitz: The Impact of Trade on Intra-Industry Reallocations and Aggregate Industry Productivity, in: Econometrica, Vol. 71, No. 6, 2003, pp. 1695-1725.
  • 9 D. Rodrik: Straight Talk on Trade, Project Syndicate, 15 November 2016, available at https://www.project-syndicate.org/commentary/trump-win-economists-responsible-by-dani-rodrik-2016-11.
  • 10 N.G. Mankiw, M.P. Taylor: Economics, Andover 2014, Cengage, pp. 423ff.
  • 11 M. Emerson: The economics of 1992. An assessment of the potential economic effects of completing the internal market of the European Community, in: European Economy, No. 35, March 1988.
  • 12 F. Ilzkovitz, A. Dierx, V. Kovacs, N. Sousa: Steps towards a deeper economic integration: the Internal Market in the 21st century. A contribution to the Single Market Review, in: European Economy, No. 271, January 2017.
  • 13 The difference in the increase in US income and US output is due to the fact that US income includes profit income from Mexico to US citizens; these have increased especially strongly.
  • 14 Congressional Budget Office: A Budgetary and Economic Analysis of the North American Free Trade Agreement, 1993.
  • 15 L. Caliendo, F. Parro: Estimates of the Trade and Welfare Effects of NAFTA, in: The Review of Economic Studies, Vol. 82, No. 1, 2014, pp. 1-44.
  • 16 D.H. Autor, D. Dorn, G.H. Hanson: The China Syndrome: Local Labor Market Effects of Import Competition in the United States, in: American Economic Review, Vol. 103, No. 6, 2013, pp. 2121-2168.
  • 17 W. Dauth, S. Findeisen, J. Suedekum: The Rise of the East and the Far East: German Labor Markets and Trade Integration, in: Journal of the European Economic Association, Vol. 12, No. 6, 2014, pp. 1643-1675; W. Dauth, J. Suedekum: Globalization and local profiles of economic growth and industrial change, in: Journal of Economic Geography, Vol. 16, No. 5, 2016, pp. 1007-1034.
  • 18 See e.g. P. Krugman, R.N. Cooper, T.N. Srinivasan, op. cit.
  • 19 N. Chusseau, J. Hellier: Globalisation and Inequality: Where do we stand?, in: Journal of Income Distribution, Vol. 21, No. 3-4, 2012, pp. 7-34.
  • 20 P.R. Krugman: Trade and Wages, Reconsidered, in: Brookings Papers on Economic Activity, Spring 2008, pp. 103-137.
  • 21 L.A. Winters, A. Martuscelli: Trade Liberalization and Poverty: What Have We Learned in a Decade?, in: Annual Review of Resource Economics, Vol. 6, No. 1, 2014, pp. 493-512.
  • 22 H.-J. Chang: Kicking away the ladder: Development strategy in historical perspective, London 2003, Anthem Press.
  • 23 L. Pascali: The Wind of Change: Maritime Technology, Trade, and Economic Development, in: American Economic Review, Vol. 107, No. 9, 2017, pp. 2821-2854.
  • 24 International Monetary Fund, World Bank, World Trade Organization: Making Trade an Engine of Growth for All: The Case for Trade and for Policies to Facilitate Adjustment, 2017, available at https://www.wto.org/english/news_e/news17_e/wto_imf_report_07042017.pdf; OECD: Making Trade Work for All, 2017, available at https://www.oecd.org/trade/making-trade-work-for-all.pdf.
  • 25 Ibid.; D. Rodrik: Populism and..., op. cit.; B. Milanović: Global inequality: A new approach for the age of globalization, Cambridge, MA 2016, The Belknap Press of Harvard University Press.
  • 26 B. Milanović, op. cit.
  • 27 International Monetary Fund, World Bank, World Trade Organization, op. cit., p. 34.
  • 28 International Monetary Fund, World Bank, World Trade Organization, op. cit.; J. Südekum: Die Globalisierungsverlierer kompensieren –aber wie?, in: Wirtschaftsdienst, Vol. 97, No. 8, 2017, pp. 566-570.
  • 29 D. Rodrik: Straight Talk..., op. cit.
  • 30 Ibid.; B. Milanović, op. cit.
  • 31 H.-J. Chang, op. cit.

DOI: 10.1007/s10272-018-0733-8