The Impact of Trade Liberalization on Income Inequality: Does the Direction of Trade Matter?

Abstract

Recent trends in inequality have raised concerns among researchers and policymakers globally. The role of globalization, one of the leading forces driving this trend, continues to be intensely debated in academic and policy circles. Invoking standard trade theory, this paper analyses whether and the extent to which trade liberalization has contributed to the recent trends in inequality. The approach and findings of the paper are novel: previous studies of trade liberalization’s impact on inequality do not explicitly control the direction of trade. The empirical results show that trade liberalization is associated with decreasing income inequality overall, but contingent on the direction of trade, it has opposing effects: North–North and South–South trade are inequality-reducing while North–South trade is inequality-increasing. Simply put, liberalizing trade between countries of similar developmental levels does not raise inequality. This paper affirms, using recent data, that trade with developing countries raises inequality in developed countries. Additionally, it finds that North–South trade (particularly imports from high-income to low-income countries) may also raise inequality in developing countries, contrary to Heckscher–Ohlin–Stolper–Samuelson model predictions. Skill-biased technical change, a consequence of trade liberalization between North and South, is the main mechanism driving inequality increases in developing countries.

KEYWORDS:

• Income inequality
• trade liberalization
• direction of trade
• tariffs

JEL Codes:

• F1
• F6
• O15

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Only the preliminary data analysis uses the full dataset (1980 – 2018). The subsequent regression analyses will use a subset of the data, 1990 – 2018, due to severely limited (missing) observations in the period 1980 – 1989.

2 The dataset used in this section covers the period 1990 – 2018. Data for 1980 -1989 are not usable for regressions due to a lot of missing values for many countries

3 Hausman test of fixed versus random effects: as shown by the small Hausman test chi-sq statistics and large p-values in Table 3, we cannot reject the null hypothesis that there is no systematic difference in coefficients between the fixed effects and random effects estimation. Since the fixed effects coefficients are consistent under both the null and alternative hypothesis, we are indifferent between the FE and RE models and choose to report the FE coefficients. The reason to prefer RE estimation is because RE coefficients have smaller standard errors, but since we are using Driscoll-Kraay standard errors, which are robust and efficient, RE estimation is unnecessary.

4 The derivative of the estimated equation (gini index) with respect to the tariff rate depends on the direction of trade variables, i.e. the interaction terms. When this marginal effect is evaluated at the means of the direction of trade variables (Merc_exp_hinc, Merc_imp_hinc, Merc_exp_lminc and Merc_imp_lminc) the effect of tariff on inequality is positive (i.e. inequality-increasing), although the estimated coefficient of Tariff rate*Merc_imp_hinc and Tariff rate*Merc_imp_lminc are negative.

Additional information

Notes on contributors

Cephas Naanwaab

Dr. Cephas Naanwaab is an Associate Professor in the Department of Economics at North Carolina A&T State University, Greensboro, North Carolina, USA. His research interests include international trade and economic development, economic impact studies, and economic inequality. He has published more than ten peer-reviewed journal articles and presented his research at more than fifteen professional conferences. He has recently been published in the International Economic Journal and the Global Economy Journal.