Income Insecurity of the Elderly, 1998-2013

Abstract

A surge of interest in economic situation of the elderly Koreans has led many to question the viability and effectiveness of public pensions. The broadest encompassing form of social safety nets in Korea, public pensions are on the one hand criticized for its weak provisions, and on the other for its lack of sustainability. The two sides are seemingly at odds against each other, with one calling for its expansion and the other its reduction. Yet, no constructive agreement can be reached without understanding the way in and the extent to which public pensions affect financial circumstances of the elderly. This paper attempts to fill the gap surrounding this debate by assessing the role of public pensions in mitigating the economic insecurity of the elderly Koreans from a longitudinal perspective – that is, by focusing on long term changes in the extent of income insecurity during old age. I first attempt to better document the changes of income insecurity of elderly Koreans using half the squared coefficient of variation (GE2) as a measure for overall income insecurity and decomposing the measure into income inequality and income volatility. I document the situation of different cohort groups over the last 16 years

for a rigorous assessment, I compare the situation of the older old and the younger old to see whether the situation is worsening for the emerging elderly. I then examine the effect of public pensions on overall income insecurity as well as the two components. I compare its effect with that of other subsidiary income sources, mainly private transfer income, which has been for long the predominant form of income source for elderly Koreans. The deployed dataset is Korean Labor & Income Panel Study (hereafter KLIPS) from 1998 to 2013 (wave 2 – 17). I provide trends of GE2, as well as its two components inequality and volatility, following a method developed by Nichols (2008). The trends reflect changes arising from differences in the three factors: age, cohort, period. I take this into consideration by using a standard cohort table for GE2, income inequality and income volatility. This allows me to draw a comprehensive and careful interpretations of the trends. Main findings are as follows. First, income insecurity first tends to rise then returns back to a level similar to or slightly higher than where it began. Of the two components of income insecurity, inequality was found to increase over time while volatility declined. Hence, increase in income insecurity was arguably driven by inequality, while the drop in latter periods was driven by growing stability of incomes. Second, in regards to the economic situation during old age, I found income insecurity to be much higher during old age. However, younger elderly displayed greater level of income security than their predecessors in both of the dimensions considered: inequality and volatility, suggesting that the situation during old age has been, in fact, improving. Lastly, public pensions were found to have gained in significance, reducing both inequality and fluctuations for the emerging old at a much higher proportion than their predecessors. Its impact was more forceful in reducing fluctuations of income, and less so for inequality, however. In contrast, effects of private transfers were eroding, suggestive of a shift of risk from household to society. This paper is not without limitations. Especially worth consideration is the measurement of economic insecurity employed in this paper

income insecurity and volatility measures fail to capture the all the complexities surrounding economic insecurity, for example the on-going insecurities or future risks that elderly households face. However, it is a good indicator of the observed insecurities reflected in income changes, and is significant in that it provides a comprehensive, unified framework and one that is practical and feasible to use with given data. Especially noteworthy finding, in this paper, is the growing presence of public pensions in determining situations of the elderly. My method also revealed its differential effects on inequality and volatility, an aspect that should be taken into account when designing related policies.