Longitudinal Studies on Sickness Absence in Sweden

Abstract: The research object of this thesis is sickness absence taking behavior in Sweden. In four empirical studies, using longitudinal data, the thesis investigates sickness absence taking behavior as a result of the impact of specific structural factors.The first empirical study (Chapter II) examines the relationship between the annual national sickness absence and unemployment rates for the period 1935-1990. The effects of changes of the replacement level are also studied. The analyses show that the widespread notion of an inverse relationship between the unemployment rate and the sickness absence rate cannot be verified by a closer analysis of available data.The remaining three empirical studies use micro-data to study sickness absence behavior. Chapter III closely examines the effect of the reduction in sickness cash benefits of March 1 1991 on the short-term absence rate. The theoretical perspective is derived from rational choice theory. Intensity regressions on duration data show that the objectives that were linked to the reduction of benefits-lower sickness absence rates and an equal distribution of the burdens brought about by the reduction-are incompatible. The weaker groups that are likely to have the greatest need for sickness absence reduce their absence taking more than stronger groups in the labor market.Chapter IV, co-authored with Joakim Palme, addresses the question of how conditions in childhood affect absence taking in adulthood. Analyses of data of a Stockholm cohort reveal how conditions in childhood and early adolescence structure the absence taking behavior of individuals. The chapter shows the endurance of these effects, a finding that is most clearly manifested in what has been labeled "the social imprint effect".The fourth of the empirical studies (Chapter V) treats the issue of gender differences in short-term absence rates. The study focuses on the impact of the gender composition of work places, but hierarchical positioning and integration among workers are also investigated. The results indicate that numerical representation conditions the short-term absence rate of women in the sense that women at workplaces where they constitute a small minority have a lower short-term absence rate than other women. For men, the hierarchical position in which they work is a more important determinant for the short-term absence rate.The results provide new insights for the study of sickness absence from a sociological perspective by specifying the mechanisms through which the social structure, in terms of institutional constraints, incentives, social stratification, and organizational traits of job sites, influences the behavior of individuals.