Health care labor costs: Evidence from relative wages of janitors

This dissertation is the first empirical study in estimating the earning differential between hospital and non-hospital janitors. In estimating the earning differential, this study utilizes data from Current Population Survey (CPS) and the Bureau of Labor Statistics (BLS) (for injury and illness data) for 1985--1996 period. Understanding the cause of wage differentials of health care workers is an important prerequisite in developing more effective cost saving policies in the health care industry. Previous research on the earning differential in the health care industry mostly considers an occupation that has industry specific job responsibilities and employs service workers of non-health sector industries as a control group. A comparison between two different occupations with absolutely different job responsibilities turns out to be largely imprecise in nature. In contrast, the relatively homogeneous job responsibilities of the janitorial occupation across industries clearly allow a cross-industry wage comparison of janitors while minimizing wage distortion. Furthermore, it allows us to examine the labor cost in a relatively non-competitive industry.;Janitors have a highly elastic labor supply curve, and so, they do not have much intrinsic bargaining power to raise their wages above the competitive level. However, janitors may obtain rent by controlling labor supply through unionization. Alternatively, they may be compensated for the risk they endure in the work site. This study employs two different models: the rent sharing and the compensating wage differentials models. Initial findings suggest that unionized hospital janitors receive 2.88% less than their unionized non-hospital counterparts. The lower percentage of unionized janitors in the hospital sector may explain this discount. Interestingly, the non-union hospital janitors receive 8.87% higher wages compared to their non-hospital counterparts in other industries. The hospital premium received by its janitors over the baseline comparison group of non-union janitors in other industries is not indicative of rent sharing, and so, it reveals the need to investigate whether this premium represents a compensating differential.;This dissertation estimates earnings equations which correct for risk endogeneity, workers' heterogeneity associated with workplace risks, and union choice. Compensation for workplace risk helps explain the hospital premium for janitors. The findings demonstrate that extra compensation is paid to those hospital janitors who own unmeasured characteristics such as attention to detail to prevent the spread of disease that make them more adept in a high-risk worksite. Because these unmeasured characteristics are highly valued, it is difficult to conclude that the high pay to janitors raises the operating costs of hospitals. These characteristics may provide some safety to janitors by reducing the percentage of workdays lost to injury and illness, thereby, possibly, lowering costs to hospitals.