Comment on “Income and Wealth Inequality in Asia and the Pacific: Trends, Causes, and Policy Remedies”

Abstract

Zhuang (2023) covers a large amount of ground when it comes to issues, countries, and literature. The latter is indicated by the fact that the list of references includes more than one hundred titles.

My first comment is that Zhuang's paper actually does not do what the title originally indicates. On one side, the spatial coverage is broader than suggested by the title. It covers Australia and New Zealand, that usually are not considered parts of Asia. More importantly, the coverage is narrower as it does not treat income and wealth inequality in Asia as a unit. True the paper deals with the development of inequality in Asian countries and its possible causes. However, it does not attempt to address how inequality at the household level in Asia as an entity has developed.

For some years a literature studying how household income inequality from a global perspective has evolved. In a recent contribution, Milanovic (2022) reports that the Gini coefficient at the household level for income in Asia as a whole decreased from 59% in 2008 to 55% in 2013, a rather large decrease over a short period.

It can be claimed that the evolution of income inequality in Asia during the most recent years does not at most, have to do with how inequality within countries has evolved. Instead the main factor is how the average incomes in various countries have changed. For example average income in China and in India has increased more rapidly than in Asia's high-income countries like Japan. As a consequence the middle classes in China and India have grown. On this see, for example, Sicular et al. (2022) who define the “global middle class” as being neither poor nor rich if the people are living in the developed world. In 2018 China's global middle class constituted not less than 25% of China's population and the middle class in India was estimated to 6% of its population. The absolute size of the Chinese middle class was in 2018 nearly double the size of the global middle class in the USA and similar in size to that in Europe.

My second comment relates to if we should care about rising inequality. Zhuang touches on this issue in his concluding section and refers to the literature on the inequality of opportunity (IOp). A point of departure taken in this literature is that public policy better not try to counteract inequality that is due to effort, but it should focus on inequality due conditions individuals cannot affect (circumstances). Most of the empirical literature aiming to quantify IOp concerns high-income countries but by now there are some papers on China. For example, Yang et al. (2021) show that between 2002 and 2013 and especially between 2013 and 2018 IOp in China declined. In 2002 the large contributors to IOp were region and hukou type at birth. However, in 2018 the contributions of those circumstances had decreased, but that of parents' education had increased. This study also finds that IOp is higher for older than younger birth cohorts.

My third comment relates to the data situation. An important issue is how comparable the data is across countries and time. For several decades there have been ambitious efforts to harmonize survey data on income for different countries in the Luxembourg Income Survey (LIS). Starting with high-income countries, now more and more countries have been included in the databank. Today LIS has data from the following Asian countries: China, India, Israel, Japan, Laos, Palestine, South Korea, Taiwan, and Vietnam. Thus, the LIS data covers slightly more than 70% of Asia's total population. One would expect that progress beyond the present paper can be made by using LIS data more intensively.

My last comment relates to how research represents inequality and its changes. Today there is a large awareness that in many cases it is not advisable to summarize the degree of inequality through the numerical value of only one index. Perhaps more progress can be made in the future by considering more than one dimension for a particular household, see, for example, the study on the USA by Fischer et al. (2022).