Perceived Job Security, Health Status and Sick-Leave: A Study on Moral Hazard in the Swedish Social Sickness Insurance System

Abstract

BACKGROUND In the Swedish sickness insurance system, short term sick-leave (STSL) is self-reported, while the long term sick leave (LTSL) is associated with medical certification. The aim of this study is to investigate the difference in sick listing between STSL and LTSL in connection with perceived job security and self-reported health status among employees. Given health status, higher job security may increase STSL due to moral hazard among employees. DATA Data from the cross-sectional postal Public Health Survey 2002 to a representative sample of the Stockholm County population aged 18-84 years (response rate 62.9%) were used. We include currently employed persons at age 25-64 years in the analyses (n=15 565). Those individuals who answered that they had been on sick-leave 1-30 days during the last 12 months with a sick listing frequency of 2-4 times were coded as STSL. Those who reported sick-leave more than 30 days during the same period were coded as LTSL. Perceived job security have been reflected on the answers to questions on individuals´ unemployment history (if had been unemployed in the last two years), type of job contract (if currently permanent job) and ethnic background (if born in Sweden, in other OECD countries or in non-OECD countries). Health status was measured by the five dimensional generic health-related quality of life measure the EQ-5D. Health risk factors (age, gender, living style, occupational status, working environment, and income) were either categorical or continuous. METHOD Logistic regression analyses have been employed using sick-leave as the dependent variable and perceived job security, health status and health risk factors as explanatory variables. Both STSL and LTSL models for both male and female and altogether have been analyzed. A full model with all variables included and a recommended reduced model have been established for each type of analyses. RESULTS Generally, health status is better for individuals with no unemployment history, with permanent job contract and among born Swedes. Given health status and other risk factors, individuals with unemployment history, no permanent job contract and born in other OECD and non-OECD countries are more frequently on STSL. On the contrary, only individuals without permanent job contract are more frequently on LTSL. For females, all three variables that indicate perceived job security have significant influence on STSL. However significant effect of employees born in non-OECD countries is not found compared with born Swedes, but found among employees born in other OECD countries. LTSL on the other hand, is significantly influenced by only type of job contract. Males employees with unemployment history and born outside Sweden are more frequently on STSL, while LTSL is higher only among employees born outside Sweden. Health status has significant effect on both STSL and LTSL, however the effect is stronger on the LTSL. CONCLUSIONS Moral hazard has been observed as higher perceived job security, for given health status, increases STSL more strongly than it does to LTSL. Moral hazard is to some extent found both among female and male.