Disentangling absence patterns

Can a compulsory dialogue meeting (DM) between the sick-listed worker, the employer, the GP and the social security administration promote return to work? This is studied by Ragnhild Schreiner, Knut Røed and Simen Markussen in the paper «Can compulsory dialogue meetings nudge sick-listed workers back to work?» The study exploit the substantial variation between Norwegian counties in the extent such meetings actually are held, to study whether return-to-work probabilities actually change more in counties with high DM-propensities than elsewhere at the durations meetings normally are held. The findings are that these meetings have a substantial impact on return-to-work propensities and that they are clearly cost-efficient. Furthermore, separating between the effects of notification (receiving the letter saying that a meeting will be held e.g. 3 weeks from now) and actual participation we find that both are approx. equally important. In other words, a substantial part of the absence reduction due to DM occurs before the meeting actually is held, as a response on receiving the notification letter. The paper is published in the Economic Journal. Is sickness absence higher than necessary as a consequence of the market organization of GPs? In Norway, patients are free to choose their own GP. These GPs get part of their payment in terms of a capitation fee for each of these patients. And, the same GPs are set to provide sickness certificates to qualify for paid sick leave. This may provide GPs with an incentive to be lenient towards their patients in order to attract more customers. Whether this is the case, and its quantitative relevance for total sick leave, is studied by Knut Røed and Simen Markussen in the paper «The marked for paid sick leave». First they consider situations where patients choose a new GP, and estimate a statistical model in order to test whether GP-leniency affects patients? choice. Their results indicate that patients do prefer lenient GPs. This is particularly the case for patients with a sick-leave episode recently. Secondly they study whether changes in the degree of competition (with other GPs in a local market) affect GP leniency. They find that GPs do become more lenient when they face fiercer competition from the other GPs nearby. In total they find that the market organization of the GP-system leads to a modest increase in total sick leave. The paper is published in Journal of Health Economics Medicalized youths? A distinct trend in Norwegian social insurance has been the increasing dominance of health related benefits, relative to unemployment benefits or welfare. In the paper Unemployed or Disabled? Disability Screening and Labor Market Outcomes of Youths, Ragnhild Schreiner study whether this trend has impacts on subsequent labor market participation and benefit take-up for young adults. Her identification strategy is built on exploiting practice style variation across social security offices over time. Her results indicate that being granted a health related benefit, instead of unemployment benefits or welfare, has substantial negative impacts on future labor market participation. The paper is part of Schreiners phd-dissertation. Intelligence and mortality: A number of studies have shown that higher intelligence, measured early in life, is associated with lower mortality ? but why is this so? Does high IQ protect against accidents or harmful habits? Is high IQ a biological signal of a well-functioning and healthy body? Or can it be something else causing the correlation? This is studied in a paper by Bratsberg and Røgeberg called Childhood socioeconomic status does not explain the IQ-mortality gradient. A prominent hypothesis is that the IQ-mortality correlation really is caused by a relationship between social class and mortality. In order to examine this, the researchers used Norwegian administrative data and compared mortality of brothers ? for which the socio-economic status is the same. The vast majority of Norwegian males born between 1962 and 1992 conducted an intelligence test in the armed forces. Since these are population data, the researchers can employ a research design where the study the IQ-mortality correlation among brothers, hence completely excluding the potential channel from socio-economic background. Their results indicate that the IQ-mortality correlation is just as strong within families as it is between families. The paper is published in Intelligence.

This project will contribute to the scientific literature on sickness absence. It is an extension, as well as a continuation, of ongoing research at the Frisch Centre using administrative data for Norway. The project will also finance a Phd-student. It ha s four topics: 1) Health and absenteeism: Recent research has found the use of graded sick leave to reduce sick leave duration as well as strengthen subsequent labour force attachment. We will extend this literature by bringing forward more specific know ledge by letting the effect of sick leave occurrence (yes/no) and type (duration and grading) be made conditional on health. If successful, this may improve official guidelines for sick-leave prescription. 2) Disability across generations: Disability is s trongly correlated across generations. This possibly puts a bound on the effect of current policies, as disability seem partly predetermined by social or biological inheritance. We will extend the large literature on intergenerational persistence in earni ngs and education by (i) carefully try to disentangle and quantify the causal mechanisms creating such correlation across generations, and (ii) evaluate policies which may break or weaken the intergenerational link. 3) The role of employers and workplace s: Arguably, the workplace is the most important arena for determining sickness absence. Yet, the mechanisms behind the variation in absenteeism between firms are not well understood. Using a rich and extensive employer-employee dataset we will seek to un derstand why some firms have much higher sick leave rates than others. In particular we will investigate the role of selection/matching, workplace characteristics and leadership. 4) The gender gap in absenteeism: There is a substantial and partly unexplai ned gender gap in absenteeism. We will test whether this gap can be explained by caretaking for children, statistical discrimination or biological sex-differences in health or risk preferences.