Occupational Strain, Health, and Compensating Wage Differentials
Occupational Strain, Health, and Compensating Wage Differentials
Monday, June 11, 2018: 3:50 PM
1055 - First Floor (Rollins School of Public Health)
Discussant: Kathryn L. Wagner
In 2014, 28 percent of Americans were completely or somewhat dissatisfied with their level of on-the-job stress (Gallup, 2014). A large medical literature shows that job stress is associated with higher rates of cardiovascular disease, workplace injury, musculoskeletal disorders, and suicide. A growing literature shows that job stress affects productivity measures such as absenteeism, tardiness, and intention to quit.
Despite of the known adverse consequences of job strain, the economics literature has found little to no evidence of compensating wage differentials for any type occupational strain for any sort except risk of death (French, 1998). However, if young workers receive compensating differentials for stressful job characteristics, then any observed correlation between job stress and health and productivity outcomes is in part driven by selection into occupation. Consider a particular type of job strain (e.g., a competitive, pressure-cooker environment). If workers have heterogeneity in preferences for job strain, then we would expect that workers who prefer pressure-cooker environments to select into those occupations. Failing to control for occupational selection will bias any estimates of compensating wage differentials. Additionally, the selection bias may vary over time. Workers’ non-monetary utility from job strain may change as they either habituate to a given or strain or begin to burn out.
In this paper, we evaluate the extent to which young adult workers are willing to tradeoff mentally stressful job characteristics for higher wages. We merge longitudinal data from the National Longitudinal Survey of Youth, 1997 cohort, to rich job characteristic data from the Occupational information Network (ONET). Using factor analysis, the ONET data allow us to characterize an individual's job on several stress dimensions. To address selection, we estimate a dynamic stochastic occupational choice model. Unlike traditional occupational choice models where there a discrete number of alternatives, we employ a continuous choice model developed by Yamaguchi (2012), where occupations are defined by a vector of job requirements and job strain. We allow each element of job strain to enter the wage expression and affect the nonmonetary utility of a particular occupation. The model allows workers to habituate to job strain, and incur switching costs (positive or negative) from choosing occupations that are distinctly different from the environments to which they are habituated. The model allows for heterogeneous preferences for occupational characteristics (job strain) and heterogeneity in the habituation process. We are in the process of estimating the model using the Kalman filter and will have results within the next six weeks. We will use available job satisfaction measures and reported intent to quit as to validate our results. Using the estimated parameters from the model, we calculate the compensating differentials for several types of job strain – What wage increase is needed to induce an individual of a particular type to be indifferent between more/less strenuous occupations?
Despite of the known adverse consequences of job strain, the economics literature has found little to no evidence of compensating wage differentials for any type occupational strain for any sort except risk of death (French, 1998). However, if young workers receive compensating differentials for stressful job characteristics, then any observed correlation between job stress and health and productivity outcomes is in part driven by selection into occupation. Consider a particular type of job strain (e.g., a competitive, pressure-cooker environment). If workers have heterogeneity in preferences for job strain, then we would expect that workers who prefer pressure-cooker environments to select into those occupations. Failing to control for occupational selection will bias any estimates of compensating wage differentials. Additionally, the selection bias may vary over time. Workers’ non-monetary utility from job strain may change as they either habituate to a given or strain or begin to burn out.
In this paper, we evaluate the extent to which young adult workers are willing to tradeoff mentally stressful job characteristics for higher wages. We merge longitudinal data from the National Longitudinal Survey of Youth, 1997 cohort, to rich job characteristic data from the Occupational information Network (ONET). Using factor analysis, the ONET data allow us to characterize an individual's job on several stress dimensions. To address selection, we estimate a dynamic stochastic occupational choice model. Unlike traditional occupational choice models where there a discrete number of alternatives, we employ a continuous choice model developed by Yamaguchi (2012), where occupations are defined by a vector of job requirements and job strain. We allow each element of job strain to enter the wage expression and affect the nonmonetary utility of a particular occupation. The model allows workers to habituate to job strain, and incur switching costs (positive or negative) from choosing occupations that are distinctly different from the environments to which they are habituated. The model allows for heterogeneous preferences for occupational characteristics (job strain) and heterogeneity in the habituation process. We are in the process of estimating the model using the Kalman filter and will have results within the next six weeks. We will use available job satisfaction measures and reported intent to quit as to validate our results. Using the estimated parameters from the model, we calculate the compensating differentials for several types of job strain – What wage increase is needed to induce an individual of a particular type to be indifferent between more/less strenuous occupations?