Absenteeism in the workplace: results from Danish sample survey data

Absentee data is analyzed using a worker environmental survey carried out in 1995. In a model which distinguishes between discretionary and involuntary absences individuals are assumed to maximize expected utility. This generates a probability distribution of days absent per year which is a mixture of two negative binomial distributions representing the behaviour of two unknown types of agent. This distribution is estimated separately for men and women. For each gender the number of days absent is quite small relative to what researchers have found in other countries. In the empirical model we identify different effects of the explanatory variables for workers with low rates of absenteeism and workers with high rates of absenteeism.