Body
Last month we posted on new data on income inequality in South Korea that builds on the Piketty-Saez approach to the issue. The work on Korea for the World Top Incomes Database was done by Kim Nak Nyeon and Kim Jongil in a paper called “Top Incomes in Korea, 1933-2010: Evidence from Income Tax Statistics” (WTID Working Paper 2014/2). An advantage of this approach over others is that it does a better job at capturing top incomes than standard household income and expenditure surveys.
In a comment on that post, Charles Wolf asked one of the most important ethical issues of our time. To paraphrase: “So what if inequality is increasing? The key question is what is happening to incomes at the bottom of the income distribution. If they are going up too, then can’t we say that the rising tide is lifting all boats?”
The short answer is that while not as bad as the US, growth at the bottom of the Korean income distribution is sluggish. A recent paper from the Bureau of Labor Statistics takes the long-standing story of stagnant wages and total compensation in the US forward to post global financial crisis time period (2007-2014). Maybe it is not surprising that trends at the outset of the period were weak, but the problem is that the recovery for those at the bottom never comes. What is surprising, however, is the size of the swath of the workforce that constitutes "the bottom." You need to get into the 80th percentile (for wages) and the 90th percentile (for total compensation) before you see positive growth over the entire period.
Statistics Korea provides information that permits a roughly comparable exercise for South Korea with monthly incomes of non-farm households of two or more; the income distribution data is here. Again, caveat emptor: this data is from surveys, not tax returns, and thus the series are not directly comparable to Kim and Kim; nonetheless, they do break out “market” and “disposable” income.
Given our interest in the market sources of inequality, we simply tracked real market income growth in the 10th and 90th percentile (first figure) and then plot the ratio of the two (second figure). As can be seen, the incomes at the bottom ten percent (technically, the border between p10 and p20) peak right before the Asian financial crisis and basically are flat since. In the third figure we show that the ratio of top quintile incomes to the bottom quintile incomes goes from about 4 in 1990 to 6 in 2009 before falling back slightly. (Note: this figure includes total income in the top and bottom quintiles, not just the income figure at the quintile marks).
When we look at real disposable income growth, the picture is not quite as bad; in the final chart we plot p 10, p 20, p 50 and p 90. The stagnation at the very bottom after the Asian financial crisis remains long-lived, but incomes actually recover after the global financial crisis, perhaps pulled up—rather than down—by closer integration with China. Yet as you move toward higher income deciles, the pace of income growth also goes up, with the top 10% clearly doing the best.
The inequality tracked in the Kim and Kim dataset is clearly driven in the first instance by the rapid explosion of top incomes, the full extent of which is not captured in the Statistics Korea data. But income growth at the bottom of the income distribution is clearly sluggish. A useful new paper from the Korea Economic Institute by Yiagadeesen (Teddy) Samy that we cited last time (using still other data; Frederick Solt’s Standardized World Income Inequality Database) walks through the hypotheses, which include deindustrialization and increasingly dualistic labor markets. The relationship with China figures in these explanations, and is a mixed bag. On the one hand, there are clearly benefits of being proximate to the Chinese growth engine; on the other hand Korea is undoubtedly feeling pressure on wages at the bottom of the income distribution from growing integration.
