The Impact of Later-Year Employment and Income Security on Mortality Among Older Adults

Background and Purpose: The trend of "late retirement," which involves older adults delaying their retirement, has become increasingly prevalent across the OECD countries. This phenomenon can be attributed to various factors, such as improvements in health and education and pension reforms that have raised the retirement age. Despite structural challenges such as an underdeveloped old age income security system and a segmented labor market, older adults in South Korea have also exhibited a tendency towards late retirement. The purpose of this study is to examine the characteristics of work of older adults in South Korea, and to investigate the impact of later- year employment and the old age income security on mortality in this population.

Methods: Using data from the Korean Labor & Income Panel Study (KLIPS), this study examined the impact of later-year employment and old-age income security (public pension) on mortality rates. We restricted our sample to respondents aged 60 and above who participated from wave 6 (2003) to wave 21 (2018) and had complete information regarding labor market participation and mortality throughout the entire study period. The final sample included 7,692 individuals (53,664 observations). To address the issue of right censoring, where multiple samples are observed without experiencing the event of death during the study period, a discrete-time event history model using a logistic regression was employed. This model is particularly useful when the timing of events is measured in discrete time intervals, allowing for the incorporation of time-varying covariates that account for changes in the risk of the event over time. Using discrete-time event history model, we examined the extent to which the employment status and public pension receipt predict time to event (i.e. mortality), controlling for various factors including age, sex, educational attainment, income quantiles, self-reported health, household type, marital status, and the logarithm of public transfer income.

Results: Older adults who were wage and non-wage workers had a lower mortality risk compared to those who were economically inactive or retired (37% to 50%, respectively). Moreover, the finding indicated that receiving public pensions was significantly associated with a lower mortality risk (25%) among older adults; household income quantiles did not have a significant effect on mortality. To ensure the validity of these results, robust analyses were performed, including adjustments for reference groups and control of a two-year lagged self-report health, which still yielded consistent findings.

Conclusions and Implications: The study findings showed a clear influence of public pension receipt on reduction in the likelihood of mortality among the older adults. Most of those among the economically active population who were not eligible for public pensions were workers at small firms, non-regular, or self-employed workers. We had a longer follow-up period than in most previous studies; therefore, our study had the strength to find more definitive evidence on a significant relationship between the public pension and mortality in older adults. This finding demonstrates the need for support of low-income older adults through improving the coverage and level of public pensions.