VoxEU Column Competition Policy International trade

Market access in global and regional trade

Do trade costs still matter in a modern era characterised by a fall in transaction costs? This column argues that there is a dearth of good analysis in the debate around market access difficulties. Complaining about restrictions in accessing foreign markets is political leaders’ current favourite hobby yet. In light of stalled WTO negotiations, shouldn’t rigour, not rhetoric, lead this debate?

Considering the current context, with WTO negotiations seemingly stalled and rising protectionist pressures since the beginning of the crisis, a rigorous measure of market access difficulties, as encountered by exporters, will encourage a more informative policy debate.

A good illustration is the case of least developed countries (LDCs), on which current WTO talks are largely focused. Despite complex and wide-ranging preferential access granted by rich countries to LDCs' exporters – essentially dropping all tariffs and quotas on manufacturing goods – there are claims that northern market access remains limited. Those claims are seemingly supported by the low level of their market shares in rich countries. The share of LDCs in total imports of the most developed countries offers a simple, although very crude, measure of their market access. This import share is rarely above 1%. Indeed, the import share of LDCs in the EU grew from about 0.4% in 1990, to 0.5% in 2000 and to 0.55% in 20061.

Measuring market access difficulties

However instructive, market shares cannot be a sufficient measure from which to draw accurate conclusions on the level of market access experienced by southern exporters on northern markets. The main limitation is that we do not know a priori what to compare those numbers to. Any assessment of market access based on trade flows needs to specify a benchmark of trade patterns, against which actual international exchanges of goods can be compared.

A theoretical benchmark

In a recent paper (De Sousa, Mayer and Zignago 2012), we use a micro-founded heterogeneous consumers model to estimate the impact of national borders on global and regional trade flows. Difficulties in market access are measured as a negative deviation from a theoretical benchmark. This benchmark is given by the ‘border effects method’, which compares imports from foreign countries to ‘imports’ from domestic producers. The benchmark is thus based on a hypothetical situation of best possible market access as faced by domestic producers. Market access difficulties are thus revealed by distortions in trade flows, after having controlled for supply and demand capacity and bilateral frictions such as tariffs.

This approach – using a gravity-based benchmark to reveal market access – is not the only one. Two alternatives exist; one measures protection directly through the collection of formal trade barriers; the other one uses deviations form the law of one price2.

We apply the border effects method to provide new results on market access difficulties, distinguishing in particular between northern and southern exporters, over the period 1980-2006, for a large set of industries. This is made possible by the construction and use of large interconnected data sets on production, trade, and various bilateral frictions such as tariffs3.

Falling market access difficulties

Our estimates imply that, on average during the period 1980-2006, each country traded around 391 times more within its national borders than with another country, all other things remeaining equal. Using standard estimates of trade elasticities from the literature, one can calculate tariff-equivalents of those border effects. For the north-north border effect over that period, it is as high as 82% while the figure is 118% for imports of rich countries coming from developing ones. Although north-north trade is far from free, expressed in tariff equivalent, south-north trade is therefore about 35 percentage points harder. Furthermore, among developing countries, it appears that the more restricted access in Northern markets is encountered by lower income exporters: those face a tariff equivalent of the border effect of 128%, while the figure for upper-middle income exporters is 111%.

As shown in Figure 1 the difficulties to penetrate foreign markets have experienced a noticeable fall since 1980 in both Southern and Northern markets. The worldwide estimated border effect has decreased from a factor of 764 in 1980 to 131 in 2006 (cf. the left panel).

Figure 1.

Note: Annual estimates of the national border dummy coefficient with their confidence interval. Source: De Sousa, Mayer, and Zignago 2012.

The right panel of Figure 1 shows that the southern border effect is much larger than the Northern border effect. However, both border effects have strongly decreased since the 1980s, mirroring the evolution of the average world border effect.

Explaining border effects

The first obvious explanation for the existence of such border effects is trade policy. The dataset at hand enables us to see if tariffs contribute to explain this overwhelming tendency to consume domestic goods. While tariffs still have a large influence on trade patterns in our regressions, they do not seem to explain a dominant part of the border effect. Controlling for tariffs, the tariff equivalent of the worldwide border effect is 164%, against 186% without tariffs as a control. The difference in the relative access of southern and northern exporters to rich markets is also preserved. This observed difference could be attributable to differences in infrastructure and trade facilitation such as cumbersome documentation requirements, restrictive administrative regulations, and other unwieldy border procedures, all of which impose higher costs on exports for developing countries.

Maps 1 and 2 show the level and the evolution of the market access of 69 developing countries into Northern markets. Unsurprisingly, east Asian exporters, China in particular, are among those for which changes in access to northern markets are more favorable over the period (Map 1). This result is in line with the finding that Japan has favoured access to its market by increasing economic relations with the developing Asian countries. Neighbours of western European countries, such as Bulgaria and Romania, also improve largely their access to rich countries after 1989. Latin America's largest economies are also facing less difficulties than before in reaching northern markets. On the contrary, African countries are in general below the median levels, except for Nigeria which substantially reduced the border effect it faces. These changes also translate into different levels of access to northern markets in the end of the period, which are represented in Map 2.

The impact of regional trade agreements

Lastly, our theory-based measure offers a renewal of the assessment of the impact of regional trading arrangements (RTAs). We investigate four main RTAs: EU, NAFTA, MERCOSUR, and ASEAN. Figure 2 shows the evolution of ‘border effects coefficients’ both for the world and for each of the considered RTAs.

Figure 2.

The impact of each RTA is expected to be different. The EU is undoubtedly the largest experiment of regional integration in the recent period, characterized by a long-term commitment of member countries to achieve wide-range integration. The left panel of Figure 2 shows the EU as an evolving (enlarging) entity, with several integration steps shown on the EU line. Each enlargement then naturally witnesses a jump in the border effect, the new members being initially less integrated than existing ones. The right panel shows the evolution for a constant definition of the EU, the EU at 15. The fall is then much smoother as expected. MERCOSUR is a customs union signed in 1991 between Argentina, Brazil, Paraguay and Uruguay (during the years of our sample) but implemented in 1995, with member countries substantially liberalizing their internal trade during the transition period. Moreover, the common external tariff covered 85% of tariff lines in 1995. A schedule for convergence towards a complete common external tariff and free trade was then agreed upon but significantly disturbed by the macroeconomic problems in Brazil and Argentina at the end of the nineties. NAFTA is a free trade agreement that entered into force between the USA, Canada and Mexico in January 1994. Tariff reductions among member countries were scheduled on a 10/15 years agenda. An interesting aspect is its North-South nature. ASEAN is officially a free trade agreement between Indonesia, Malaysia, Singapore, Thailand and the Philippines since 1977, but intra-bloc trade liberalization was really implemented on a large scale starting with the ASEAN free trade agreement in 1992 (Soloaga and Winters 2001).


As shown in Figure 2, the EU, NAFTA, ASEAN and MERCOSUR agreements all tend to reduce the estimated degree of market fragmentation within those zones, with the expected ranking between their respective trade impact. A striking characteristic is the apparent convergence in the absolute level of integration of the EU, NAFTA and ASEAN until the end of the 1990s. The EU begins as a far more integrated areas than the other two zones, but these zones gradually catch up, becoming close to the level of EU integration at the end of the 1990s. For the most recent period, after 1999, there seems to be a clear ranking of integration with EU countries being the most integrated zone followed by NAFTA, ASEAN and then MERCOSUR, for which border effect coefficients fall markedly since the period 1993-1995. This result is especially interesting since 1995 is the date where most internal trade liberalisation should have been completed.


Anderson, J E, and E van Wincoop (2004) "Trade Costs", Journal of Economic Literature, 42(3), 691-751.

De Sousa J, T Mayer and S Zignago (2012), "Market Access in Global and Regional Trade", CEPR Discussion Paper, 9085.

Evenett, S (2012), "Débâcle: The 11th GTA Report on Protectionism", VoxEU.org, 14 June.

Soloaga I, and A Winters (2001), "Regionalism in the Nineties: What Effect on Trade?", North American Journal of Economics and Finance, 12, 1-29.

1 The EU market is understood here as the first 15 EU members. The 50 LDCs are retained according to the UNCTAD's list (as of 2006).

2 cf. Anderson and Van Wincoop 2004 for a survey of those types of evidence, and the major data issues involved

3 This data set is made publicly available at http://www.cepii.fr/anglaisgraph/bdd/TradeProd.asp.

4,514 Reads