Three empirical essays on absenteeism
Absenteeism is a widely observed phenomenon that has received a great deal of attention from academics who argue that it is an excellent proxy for individuals' attitude to work and commitment to their jobs. Unfortunately, very little of this work has been done by economists. The little economics that has been done has tended to view absenteeism as a measure of the supply of effort. Given the paucity of economic analysis on absenteeism, the psychology, sociology and management literature is reviewed to examine the extent to which their approach and that of an economist have common ground. Upon careful reading, it becomes evident these disciplines offer similar perspectives. Probably the most researched area of absenteeism is the relationship between absence and turnover. Although there is much contention as to what the relationship between these two phenomena should be, most researchers view this as a means to test the hypothesis of withdrawal. This thesis examines the problem somewhat differently and suggests that the approach of much of the empirical work is misguided. An alternative methodology to examine these phenomena is suggested and tested using a very large and detailed database. The results suggest there is a positive correlation between absence and turnover, although the relationship is more complex than described in the literature. One area where economists have made a great deal of theoretical progress is in the examination of why absence might vary across firms. The key insight is that production technology may affect the shadow cost of absence and if the costs of absence differ across firms, then there will be different levels of motivation to reduce it. It is argued that not only will the shadow cost of absence vary across firms, it will also vary over time and a theoretical model is developed to demonstrate this. There is a presumption in the literature that absenteeism is inversely related with the business cycle. However, the empirical work on the subject only models absence as a supply side phenomenon. This introduces a significant identification problem. At the very time when individuals are least likely to go absent, firms' demand for reliable labour will be at its lowest. The empirical work in the chapter models absence from both the supply and the demand side and the findings confirm that both play a significant role in determining absence. The finding that firms' demand for reliable labour may vary through the business cycle is novel and receives further investigation. The data is dissaggregated to determine the robustness of the relationship between demand side factors and the business cycle. At broad levels of disaggregation, the results remain quite strong, although there does appear to be a difference between unionised and nonunionised workers. At finer levels of dissaggregation the results are not as conclusive. This is attributed to the relatively small samples used to derive the individual absence series and the resulting increased volatility that emerges due increased variability from the use of small samples.