Absenteeism Predictors: Least Squares, Rank Regression, and Model Selection Results

This paper examines the determinants of absenteeism using ordinary least squares rank-based regressions and a model selection procedure. The results show that personal attributes are the most important determinants of long-term absences. For total working days lost, the penalty factors are the most significant predictors. The results also show that absenteeism tends to be lower among firms with more part-time workers. Unionization, on the other hand, increases the total days lost due to absenteeism.