There is a substantial literature that demonstrates a correlation between individual level of educational and health. There are four potential explanations for this correlation. Firstly, the level of education might be affected by the health of the individuals. Secondly, education might contribute to social stratification of individuals. Thirdly, education might provide a limited number of individuals with abilities which then are spread through social networks. Fourthly, education might provide the individual with abilities that will enhance health. From a policy perspective, the last potential mechanism is most important. The ideal research design to investigate this mechanism would be to carry out controlled trials. Obviously that is not feasible. A number of arguments indicate that natural experiments is the second best design, superior to cohort studies. The health effects of expanding schooling by means of changes of compulsory schooling laws have been extensively investigated. The results, however, are conflicting. In order to clarify the issue, a meta-analysis of the health effects European compulsory school law changes has been carried out. The effects of 26 analyses on mortality, 9 analyses on self-rated health and 7 analyses of the rate of obesity will be presented. The question of causal effects and potential mechanisms will be discussed.

Sven Bremberg, MD PhD, is a Senior consultant in Child and Adolescent Health at the Public Health Agency of Sweden and Associate professor, Department of Public Health, Karolinska Institute, Stockholm. He has been a contributor to the Swedish national strategies to tackle social inequalities in health.
Sofia Ljungdahl, Med dr, is a Public health planning officer at the Public Health Agency of Sweden. Dr Ljungdahl’s recent work addressed scientific evidence and recommendations in public health.