A comparison of international income inequality: an ordered probit model analysis

Abstract

In this study, the ordered probit model with marginal effect was applied to examine the key elements that influence changes in the Gini levels. The empirical results indicated that export incentives and poverty have a positive impact on the Gini levels. Moreover, the marginal effect revealed that export incentives increase income inequality for countries with higher inequality and reduce inequality for those countries with a relatively even income distribution. Likewise, the study found that the Asian Pacific Economic Cooperation (APEC) and the EU countries move from inequality to equality more efficiently than countries in the Latin American Integration Association (LAIA).

Keywords:

• income inequality
• Gini
• ordered probit model

JEL Classification:

• E64

Notes

¹ This article, based on the occurrence of the Asian financial crisis, splits the sample into two sub-samples. It spread through Russia and countries in Latin America causing the Russian financial crisis and the Latin American financial crisis in 1998. These events not only influenced member countries in APEC, but also hold lessons for those countries in EU and LAIA.

² Income inequality increases socio-political instability, which in turn impedes foreign investments. Therefore, income inequality and investment are conversely related.

³ The familiar interpretation of the Gini coefficient comes from the Lorenz curve and was first developed in the field of economics. It graphs cumulated income shares versus cumulated population shares when the population is ordered from low to high, as per capital incomes (Levy and Murnane, 1992; Partridge, 1997; Rodgers and Bertram, 1999; Mah, 2003; Haidich and Ioannidis, 2004).

⁴ The Ordinary Least Square (OLS) considers the difference in the dependent variable between 1 and 2 as equivalent to the difference between ranking 2 and 3.

⁵ We categorized a variety of the threshold value of the Gini coefficient and selected suitable values based on their significant levels after running the ordered probit model.

⁶ This database is available online at www.wider.unu.edu. Also, the values we applied were calculated from the formulas developed by Tony Shorrocks and Guang Hua Wan, as listed on the database of the World Institute for Development Economic Research (WIDER).

⁷ For developing countries, trade barriers’ reduction would lead to faster economic growth.

⁸ He used (Export + Import/GDP)×100% as the openness ratio.

⁹ They are Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, Netherlands, Portugal, Spain, Sweden, UK, Cyprus, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Malta, Poland, Slovakia and Slovenia.

¹⁰ They are Argentina, Bolivia, Brazil, Chile, Colombia, Cuba, Ecuador, Mexico, Paraguay, Peru, Uruguay and Venezuela.

¹¹ www.worldbank.com.

¹² They are Egypt, Jordan, Morocco, Tunisia and India.

¹³ In our sample, there are 14 countries in this group. They are Bolivia, Brazil, Chile, China, Colombia, Ecuador, Hungary, Lithuania, Mexico, Peru, Philippines, Poland, Thailand and Vietnam.

¹⁴ The greater openness to trade might widen initial inter-country differences in skill levels causing a divergence in the levels of per capita income. Thus, the income inequality gap expands.