Welfare Benefits and Labour Supply: Evidence from a Natural Experiment in Japan.

Work incentives of welfare programs that provide means-tested income transfer for low-income population has been a central issue in the context of ulitisation of labour force and/or social inclusion. Many existing studies show the evidence that welfare benefit levels have negative impact on labor supply of recipients. Accordingly, various welfare reforms have been proposed and implemented in developed countries. In Japan, Public Assistance (PA) use has been increasing since the early 1990s. Although the main cause of surging PA use is population aging, the proportion of working population to total recipients is also increasing. Therefore, an examination of the causal relationship between work incentive and the PA is needed. In this study we use municipality mergers implemented in Japan between 2000 and 2005 to identify the impact of welfare benefits on labour supply. In Japan, benefit levels of PA depend on recipients’ resident area as well as individual attributes. Each municipality is assigned to one of six area classifications with different benefit levels. An important feature of municipality mergers was exogenous increase of benefits for residents in case of the merger among municipalities with different benefit levels. To identify the effect, we conduct the difference in differences estimation, using panel data of Japanese municipalities. The results show that the increase in PA benefit levels caused by municipality merger during observation period raise the dependency rate of PA by 0.4-0.6‰ points and lower the employment rate of working population by 0.1-0.3% points on average. The magnitude of the estimated effects is not negligible compared to average change of dependency rate (+3.7‰ points) and employment rate (-0.6%points) during same period.