Health status and wage differences: measuring productivity penalty and discrimination patterns

Abstract

In this study by utilizing the 2008 Athens Area Study (AAS) data set, I examine to what extent inequalities in the Greek labour market between health-impaired and healthy employees may be attributed to differences in productivity and/or discrimination. A group of health-impaired male and female employees who self-reported that their productivity is not affected by their impairment is used to measure the effects of discrimination separately from the effects of poor health on wages. A penalty for heath-impaired employees having work limitations exists. Evidence of wage discrimination against health-impaired employees having no work limitations is also observed. Both findings are statistically significant. Currently, health-impaired people do not appear to face a level playing field in the Greek labour market.

Acknowledgements

I am grateful for the helpful comments from the editor of the journal Applied Economics Letters, Prof. Mark Taylor and from two anonymous referees on an earlier draft of this article.

Notes

¹Definitions of detailed variables and descriptive statistics are available with the author and will be given on request.

²The explanatory variables that I use to explain and identify health status conditional on work limitations are detailed health measurements: functional disabilities, limitations in daily activities and other instrumental activities, the presence of diabetes, heart disease, arthritis, psychological problems and cancer, as well as the presence of disability/health benefits.

³The controls for endogeneity of self-assessed health responses are statistically insignificant in all regressions. I can suppose then that the estimated wage discrimination would be its true level. Sensitivity analysis is available with the author and will be given on request. Note that the regression outcomes show that employees' choice to report a health impairment increases with the existence of limitations in functional, daily and instrumental activities, the existence of objective illnesses such as diabetes, heart attack, arthritis, psychological problems and cancer, as well as having disability/health benefits. The magnitudes of these effects are stronger for the health-impaired employees with work limitations than for the health-impaired employees with no work limitations.

⁴With respect to the other exogenous variables of interest, the results are as expected.