Bridging Disciplinary Boundaries (January 11 - 14, 2007)

Purpose: This study investigates what factors affect the level and change rates of job satisfaction of the elderly during six years prior to retirement. Most of previous research regarded retirement as a discrete event rather than as a gradual process. Thus, little is known about which factors are critical for predicting the absolute level of job satisfaction of the old workers and how workers adjust their attitudes on jobs as approach to retirement.

Methods: The Data used for this study are the Korean Labor and Income Panel Study, which is a longitudinal data from 1998 to 2003. The sample includes individuals who were interviewed for all six years, aged 50 or more, and heads of household or their spouses. In order to consider variances within-individual and between-individuals, the Hierarchical Linear model with two levels, specifically the individual growth model, is used for this study. The level of job satisfaction is used as a dependent variable for the first and second level models. Independent variables in the first model include time-varying variables such as age, earnings, employment status (full-time or not) and job types. The intercept and the slope of age are assumed to be random. The second level models include time-invariant variables as independent variables, such as education levels, gender, and head of households.

Results: The results show that the absolute level of job satisfaction is influenced by the labor market factors, such as type of jobs and employment status. In addition, the level of education and earnings are significant predictors for the level of job satisfaction. Meanwhile, demographic factors, such as age, gender, and family relationships, have no impact on the level of job satisfaction. Regarding the change rates of job satisfaction, it turns out to be almost flat throughout six years, indicating no change in job satisfaction across all independent variables. The slopes differ only by the employment status. The job satisfaction levels of full-time workers are higher than those of part-time workers. While job satisfaction levels of full-time workers remain constant, those of part-time workers gradually increase as retirement comes close to.

Implications: These results have the following policy implications. First, it is necessary to activate labor participation of the elderly by the demand side policies rather than the supply side. The fact that there is no change in job satisfaction implies that the elderly retire involuntarily even though they are eager to work more. Thus, it would be useful to create a lot of jobs from demand side to make the elderly work. Second, it is necessary to enact social welfare policies for a gradual retirement to promote working activities of the elderly. However, this study is limited in not scrutinizing how a worker's job satisfaction is empirically connected to the retirement decision.