The award of disability pensions by municipalities and referral to flex jobs
Christophe Kolodziejczyk, Henrik Lindegaard Andersen, Kræn Blume Jensen og Signe Hald Andersen, August 2009
Summary
During the last two decades unemployment has continued to fall, and it reached a historic low in 2008. During the same period the number of individuals awarded a disability pension and/or a flex1 job has continued to rise, due to an increase in the number of applicants. However, this general increase conceals the fact that there are considerable differences at municipal level. Some municipalities award far more disability pensions and flex jobs than others.
This working paper investigates these differences and postulates a benchmark model controlling for a large number of factors at both individual and municipal level. The model is used to predict the inflow of applications for disability pensions and/or flex jobs at the municipal level and compare the predicted inflow to the actual inflow. The ratio between predicted and actual inflow is the benchmark indicator which is used to rank the municipalities.
Several factors influence a municipality’s propensity to award disability pension and flex jobs. These factors can be divided into three main groups. Firstly, there are personal factors such as demographics, education and the labour market, as well as health-related factors which take centre stage in this analysis. Health-related factors are the frequency of visits to the doctor, hospitalization and the use of prescribed medicine. Secondly, there are municipality-related factors such as the constraints under which the municipality is placed (such things as size and the conditions obtaining on the local labour market). Thirdly and lastly, factors relating to caseworker practice and other invisible institutional factors.
In the empirical section, individual and municipality related factors are included in the models. After controlling for these factors, unexplained differences may be largely attributed to differences in the last group of factors, i.e. caseworker practice and other invisible institutional factors. The empirical model is a logit model with fixed effects at the municipal level, and the marginal effects are calculated. Based on simulations the model calculates the predicted inflow of applications for disability pensions and flex jobs and a standard error. The analysis focuses on applications for disability pensions and/or flex jobs by people aged 18-64 during the period 2006-2007 and register based data are used as the empirical foundation.
The analysis shows that the older people are, the more likely they are to obtain disability pension or flex job. The same is true if their attachment to the labour market is weak and they are in poor health, while factors affecting the municipality and the constraints affecting it do not seem to make any difference. The model revealed that, even after controlling for the factors mentioned, large differences existed between municipalities. In fact, the number of disability pensions and/or flex jobs awarded could be reduced by 43% if all municipalities awarded them as sparingly as the 10 municipalities who used them least. The analysis also shows that those municipalities who award (relatively) the greatest number of disability pensions and flex jobs use the diagnosis “Mental disorders” more than other municipalities, have a shorter case processing time and more rarely reject applications.
- Flex job. Light work, possibly less than full time. The employer receives a subsidy from the municipality for each person so employed.
Danish Institute of Governmental Research | Købmagergade 22 | 1150 København K | E-mail: