Is a welfare system a welfare improving? On the one hand, this form of social insurance gives money to those who have come upon rough times, facing low income and unemployment. On the other hand, giving individuals money conditional on not having low income gives these same people an incentive not to work. The field of economics can help policy makers by measuring the benefits individuals receive from the program and also calculating the work disincentives. Most of the economics literature has focused on the cost of the welfare system and one of these papers is the Lemieux and Milligam (2004).
In this paper, the authors look at the the Social Assistance (S.A.) program in Canada. Prior to 1989, Quebec offered a monthly benefit of about $506 CDN to poor single males 30 years old and older but a benefit of only $185 CDN to those under the age of 30. The authors hypothesize that 29 year-olds should work more than 30 year-olds solely due to this difference in benefits. Using a regression discontinuity framework, the authors aim to test this empirically.
Using a variety of specifications and testing their results against other Canadian provinces, Lemieux and Milligan find that employment rates for 30 year olds in Quebec are 3-5% lower than their 29 year old counterparts. Macro-economic conditions were similar in Quebec to the rest of the nation and their finding seem relatively robust. Since the authors do not measure the benefit of the program, it is difficult to determine whether or not the Canadian S.A. program is worth the cost, but the authors’ evidence that welfare programs create employment disincentives is important.
Lemieux and Milligan (2004), “Incentive effects of social assistance: a regression discontinuity approach,” NBER WP 10541.