A public pension system is one of the most basic components of a social safety net for the elderly. Following developed countries, other developing nations have adopted public pension programs in recent decades. The importance of a pension system is more pronounced in South Korea because it is one of the fastest aging countries. South Korea’s old-age dependency ratio, the ratio of elderly persons divided by those of working age, is expected to increase to 29.8% by 2030 (South Korea, Office of Statistics), up from 14.9% in 2009 (World Bank).

The Korean National Pension Plan (NPP) was introduced in 1988 on a pay-as-you-go basis. The NPP experienced three major coverage expansions. In 1988 when the NPP was implemented, it covered workers in establishments with 10 or more workers. The compulsory coverage was expanded to firms with 5 or more employees in 1992. In 1999, the NPP coverage was extended to self-employed individuals and non-income earners and it achieved the status as a universal pension plan targeting the entire general public. As anticipated, participation rates in the NPP increased sharply and substantially with each round of legislation: from 4,433,000 in 1988 to 16,230,000 in 2000.

We construct three birth cohorts of men (born between 1936-1939, 1940-1943, and 1944-1947) who were subject to the different NPP eligibility criteria. We use successive waves of the 1990-2010 Korean Family Income and Expenditure Surveys (FIES) to investigate the impacts of expansions in mandatory participation in the public pension plan on the economic outcomes of the elderly. Each wave of the FIES, administered by the Office of Statistics in South Korea, includes detailed data on approximately 10,000 households. 

We use the birth cohorts as instrumental variables to estimate the impacts of expansions in the pension plan on the income, expenditures and labor market participation of the elderly. Our LATE estimates are unbiased providing the IVs are correlated with NPP participation as is clearly demonstrated by the sharp increases in NPP uptake with each expansion in mandatory participation.  Additionally, because the differences in age between the cohorts are not large, we argue that the IVs are not correlated with unobserved individual characteristics that affect NPP participation. 

While prior research on the South Korean NPP has focused on the solvency of the system, this study evaluates its impacts on beneficiaries. The study has important implications for labor market activation and anti-poverty policies for the elderly, a demographic group that is increasing rapidly in South Korea and globally.