Worker absence and shirking: evidence from matched teacher-school data

We utilise a unique matched teacher-school data set of absenteeism records to quantify shirking behaviour in primary and secondary schools. Shirking behaviour is shown to vary systematically across schools, and hence schools are characterised as either healthy (low absenteeism) or sick (high absenteeism). Using count data techniques, and allowing for the problems of unobserved heterogeneity and partial observability in our data, we find that teachers in sick schools have higher absence rates. Our estimates suggest that shirking behaviour can account for 24 percent to 38 percent of recorded absenteeism. Furthermore, a teacher who moves from a healthy school to a sick school is likely to face an increased risk of absenteeism of up to 70 percent. As the factors a¤ecting involuntary absenteeism are unlikely to change in the short run, we argue that this increased incidence in absenteeism re?ects the impact of the change in school environment on shirking behaviour.