Among economists and policymakers alike, there is now a sense of agreement that import competition from low-wage countries has caused a decline in the relative wage of unskilled workers in rich nations, probably best summarised by Krugman’s verdict that the impact of trade on wages “is big, and getting bigger” (see Krugman 2007 on this site).1
Much less discussed is the flipside of this argument, namely that trade with richer nations tends to decrease the relative wage of skilled workers in poorer nations, which reduces the incentives to invest in human capital. So, does trade with rich nations "de-skill" emerging economies? In a new study (Auer 2010), I develop an empirical methodology to answer this question.
Exploiting geographic and over-time variation
The basic problem for this empirical analysis is that even in a static world with fixed supply of skilled and unskilled labour, any measure of the factor content of trade would be correlated with domestic education levels.
I propose a two-step methodology to address this problem and establish the causal relationship between trade and domestic education decisions. In the first step, I isolate the human capital component of imports that is exclusively related to the skill supply abroad, i.e., I isolate the geographic component of the factor content of trade that can be explained by a country's geographic proximity to international supply of skilled or unskilled labour.
More specific, I run gravity-style regressions relating the skill content of imports to geographic information. Then, I isolate the geographical component following the approach developed by Frankel and Romer (1999), who construct measures of how much other nations are likely to export to a given nation conditional on the geographic bilateral distance. Whereas Frankel and Romer are interested in how much other nations are likely to trade with a given country in total, I am interested in how much skilled labour content other countries are likely to export to this nation. The variables constructed in this way are termed "geographic proximity to skilled labour” and "geographic proximity to unskilled labour”.
In the second part of the empirical analysis, I use the constructed measures of geographic proximity to skilled/unskilled labour to identify whether trade affects human capital decisions. The analysis takes care of two potential problems.
The first potential problem is that the geographic instrument for trade is collinear with country-specific determinants of development such as institutional quality (see, in particular the critiques by Dollar and Kray 2003 and Rodrik et al. 2004). This issue is tackled following the solution proposed by Dollar and Kray (2003) and Feyrer (2009). I do not rely on the cross-sectional information as in Frankel and Romer's original analysis, but rather use the time-series variation within each country.2
The second potential problem is that countries which are geographically open to skilled labour also tend to be geographically open to trade in general. I thus also include a country's general level of geographic openness to trade (again varying over time) in all specifications, thus controlling for general openness to trade in the Frankel and Romer sense.
Findings: School dropouts made in the OECD?
Table 1 below presents my main empirical findings. All panel estimations span five 5-year intervals from 1972 to 1992 in 41 non-OECD nations. All specifications include three sets of control variables:
- Country-fixed effects to capture time-invariant country-characteristics.
- A linear trend to capture the general upward movement in both openness and education.
- A time-varying measure of a country’s general openness to trade that controls for the general effect of being geographically open to trade.
In addition, the specification includes the main variable of interest, namely:
- Time-varying measures of a country’s geographic proximity to skilled labour (columns 1 to 3) or unskilled labour (in columns 4 and 5). Both measures of geographic proximity are standardised.
Overall, these specifications thus exploit variation in how some countries tend – compared to their relative openness to all goods – to be relatively close to the supply of skilled or unskilled labour and how this relative proximity varies throughout time.
In column (1), the dependent variable is the (logarithm of) average years of total education in the workforce over 15 years of age. I find that the coefficients of proximity to skilled labour is statistically highly significant and is also very large in economic terms. Holding country, time, and general openness to trade constant, a one-standard deviation higher geographic proximity to skilled labour is associated with a 12% lower average education length of the country’s workforce. This is economically very sizeable and also comparable to the effect of geographic openness itself (positive coefficient 0.191, also this variable is standardised).
Columns (2) and (3) document that different types of education are affected differently by imported skills. In specification (2), the dependent variable is the logarithm of average years of primary education in the workforce. In specification (3), the dependent variable is the logarithm of average years of higher (secondary plus tertiary) education in the workforce. While the effect of proximity to skilled labour is significant for both types of education, the coefficient is larger for higher education. A one-standard deviation higher proximity to skilled labour is associated with a 15% reduction of higher education, but only a 10% reduction in primary education.
Columns (4) and (5) evaluate the flip side of the results presented in the previous columns – does proximity to unskilled labour increase domestic education? The dependent variables are again the logarithm of primary (in column 3) and the logarithm of higher education (in column 4). Both primary education and higher education significantly increase if a country is closer to supply of unskilled labour. Again, I find that this relation is more pronounced for advanced education than for primary education.
Table 1. Regression results
Can globalisation cause divergence?
Table 1 documents that in a sample of 41 emerging economies, being close to the global supply of skilled labour decreases domestic human capital, whereas being close to unskilled labour increases it.
In Auer (2010), I next analyse the welfare and income consequences of trade-induced (dis-)accumulation of human capital in a model featuring within-country worker heterogeneity and across-country differences in the relative productivity of human capital. This model is based on heterogeneous workers who make educational decisions in the presence of complete markets. It features no externalities and consequently all countries must strictly gain from trade.
Yet globalisation might still cause poor countries to have lower income (reflecting the lower investment levels in human capital) and it might also imply divergence of welfare. When heterogeneous workers invest in schooling, high type agents earn a surplus from their investment. Trade shifts this surplus to rich countries that can use these skills more efficiently. The dynamic effects of trade liberalisation thus tend to occur mostly to initially rich countries, implying that it is the already well-off that gain the most from trade, i.e. while trade results in a uniformly better off world, it may also be associated with greater inequality across nations.
Auer, Raphael (2010), “Are Imports from Rich Nations Deskilling Emerging Economies? - Human Capital and the Dynamic Effects of Trade”, Working Paper 2010-18, Swiss National Bank.
Auer, Raphael and Andreas M Fischer (2010), "The effect of low-wage import competition on U.S. inflationary pressure", Journal of Monetary Economics, 57(4):491-503, May.
Auer, Raphael and Andreas M Fischer (2008), “The impact of low-income economies on US inflation”, VoxEU.org, 13 June.
Auer, Raphael, Andreas M Fischer, and Degen Kathrin (2010), "Globalization and Inflation in Europe”, MIMEO, Princeton University (new version available on request).
Bernard, AB, JB Jensen, PK Schott (2006), “Survival of the best fit: Exposure to low-wage countries and the (uneven) growth of US manufacturing plants”, Journal of International Economics, 68(1):219-237.
Dollar, David and Aart Kraay (2003), "Institutions, trade, and growth", Journal of Monetary Economics, Elsevier, 50(1):133-162, January.
Feyrer, James (2009), "Trade and Income -- Exploiting Time Series in Geography", Mimeo, Darthmouth University.
Frankel, A Frankel and David Romer (1999), "Does Trade Cause Growth?", American Economic Review, 89(3):379-399, June.
Krugman, Paul (2007), “Trade and wages, revisited”, VoxEU.org, 15 June.
Krugman, Paul (2008), “Trade and Wages, Reconsidered”, Presented at the 2008 Spring Meeting of the Brookings Panel on Economic Activity, February.
Rodríguez, Francisco and Dani Rodrik (1999), "Trade Policy and Economic Growth: A Sceptic's Guide to the Cross-National Evidence," CEPR Discussion Papers 2143, CEPR Discussion Papers.
Rodrik, Dani, Arvind Subramanian, and Francesco Trebbi (2004). “Institutions Rule: The Primacy of Institutions Over Geography and Integration in Economic Development”, Journal of Economic Growth, Springer, 9(2):131-165.
1 Also the academic literature has found evidence that the effect of such import competition is substantial. For example, Bernard et al. (2006) document that in the US, industries facing a high degree of import competition from China are characterised by industry exit and low domestic growth rates. Auer and Fischer (2008 and 2010) and Auer et al. (2010) document that sectors facing a high degree of low wage import-competition experience strong price declines.
2 Rodríguez and Rodrik (1999) argue that it is not certain whether the insights of this literature can be used for trade policy since they relate geographic characteristics to economic development, and trade policy does not affect geography.