Job Loss and Work Among Older Workers

Abstract

In recent decades the older worker population has grown significantly, and its sociodemographic composition has become more diverse. Yet it is not clear what the implications of these demographic changes are for work in older age. Prior research on aging and work has predominately studied more advantaged workers. This dissertation improves our understanding of the labor market experiences of marginalized workers in older age. In Chapter 2, I evaluate whether Black-White disparities in reemployment likelihood increase, decrease, or remain constant among older workers relative to younger workers and what factors explain these differences. I draw on scholarship on racial disparities in health and evaluate three hypotheses: cumulative dis/advantage (inequality increases over time), aging-as- leveler (inequality decreases over time), and persistent inequality (inequality does not change over time). I test these hypotheses using hazard models and data from the Survey of Income and Program Participation. I find that Black-White inequalities in reemployment narrow by about 70 percent in older age, supporting the aging-as-leveler perspective. Relative to their younger counterparts, White older workers experience a greater decline in reemployment likelihood than Black older workers. Though Black older workers face the lowest chance of reemployment. Prior experience, education, and skills do not explain Black-White differences in reemployment across age, suggesting that hiring discrimination and other post-job-loss factors likely underlie these differences. In Chapter 3, I generate more comprehensive estimates of Black-White differences in the costs of job loss across age and test which theories account for these differences. Scholarship on job loss typically examines changes in earnings only for workers who reemploy and likely underestimates the financial implications of late-career job loss, particularly for minority older workers. In my sample I include individuals who do not reemploy to account for Black and older workers’ lower likelihood of reemployment. Using fixed effects models and data from the Survey of Income and Program Participation, I find that workers experience a large decline in earnings following job loss, ranging from 65 to 81 percent 1-3 years after job loss. Older and Black workers experience greater declines in earnings than their White and younger counterparts, but I do not find evidence of Black-White disparities changing across age. I find that job tenure partially accounts for age differences, pointing to human capital, job matching, and deferred compensation theories. However, age and race differences in earnings losses remain either partially or totally unexplained, suggesting that differences stem from unobserved factors that likely emerge after job loss, such as job search characteristics and age and racial discrimination. In Chapter 4, I examine why less-educated men are more likely to exit the labor force than their more-educated counterparts prior to retirement age. In recent decades there has been an increase in early retirement (labor force exit by around age 60) among men, particularly among those without a college degree. This gap in labor force attachment by educational attainment (what I refer to as the education gap) is of concern because working into older age is critical for creating a sufficient nest egg for retirement. Is the education gap in labor force participation due to choice or constraint? I use the Health and Retirement Study to estimate how much of the education gap is a function of less-educated men disproportionately being unable to work (health status) and disproportionately lacking access to higher-quality jobs with employee retirement benefits that encourage working longer (pensions). I find that differences in health status and in pension coverage account for 67 percent and 22 percent of the gap, respectively. Together these findings suggest that less-educated men likely exit the labor force due to constraint, prior to saving enough for retirement. The education gap in labor force participation could therefore be a precursor to retirement security inequalities in older age.