Health, disability insurance and labour supply: evidence from a dynamic structural model

This work presents a dynamic structural model of saving, labour supply, disability insurance and private pension claiming decisions in order to provide a quantitative assessment of the labour market effects and the distributional consequences of disability insurance in the UK. I develop and calibrate the model using the English Longitudinal Study of Ageing, focusing on males living with a partner at the end of their working life. In the model, individuals face uncertainty on wage realization, health development and life expectancy. They make a discrete choice about whether to participate in the labour market and in case of participation they choose hours of work. Moreover, if their health level is below a calibrated threshold they can apply for contributory disability insurance. I perform several policy experiments altering the parameters of the disability benefit programme. The policy experiments reveal the presence of unused working capacity among disability insurance recipients: among benefit recipients with health close to the eligibility threshold 41% would be active in the absence of the programme. The benefit has important distributional consequences and reduces income inequality caused by health effects on productivity. The insurance value provided by the benefit is largely heterogeneous across health and wealth groups: individuals with health and wealth in the first quartile of the distribution at age 50 would need 44% of their assets to be compensated from benefit removal. Strengthening the health assessment to receive disability insurance might help target efficiency and increase labour market participation overall, but at the cost of an increase in the dispersion of the income distribution. A policy reducing the cost of re-entering the labour market by 30% (for those with health below the disability insurance threshold) is promising, as it increases labour supply by 3% without increasing inequality. Finally, marginal changes in the generosity of the benefit amount has a small effect on labour supply. Focusing instead on individuals younger than 55 the effects are larger: the elasticity of non-participation to the labour market with respect to benefit generosity is 0.34 and the elasticity of IB rate is 0.41. Having a comprehensive health measure and a continuous process for health allow to better evaluate heterogeneity by health level in the policy effects.