Date of Award
Campus Access Dissertation
Doctor of Philosophy (PhD)
Jeffrey A. Burr
Wang (2013) noted a paucity of retirement research that utilized longitudinal study designs. Following Tang and Burr's (2015) retirement research, this study analyzed retirement trajectories using a combination of latent class analysis and group-based trajectory modeling. This study identified the number and types of retirement trajectories using multiple indicators of labor force participation and retirement. Using the Health and Retirement Study data (1998-2004), this study found at baseline that four latent classes (not retired, retired, partially retired, homemakers) produced an optimal fit to the data. Thereafter, group-based trajectory modeling showed two divergent retirement trajectories over the remaining time periods. Chi-square tests were applied to the baseline classes to test for differences by gender or race. Although chi-square tests showed differences in retirement status by gender and race, once controls were added for age, health, marital status and education, the race and gender differences in retirement status were no longer statistically significant. Consistent with the research literature, the following covariates were associated with a greater likelihood of transitioning to retirement: increasing age, poor health, spousal influence, and less education. Further retirement research using the novel methodologies of latent class analysis and group-based trajectory modeling is warranted.
Haimowitz, Bruce R., "Trajectories of Retirement in Later Life: A Group-Based Modeling Approach" (2017). Graduate Doctoral Dissertations. 359.