Optimal redistribution when stigma matters
(2003)
Files
Jacquet_L.pdf
Restricted Access - Adobe PDF
- 6.72 MB
Details
- Authors
- Supervisors
- Van Der Linden, Bruno
- Abstract
- A negative income tax (NIT) is a form of refundable tax credit and simply consists in paying benefits to those below an income threshold, and taxing the others. If instead of netting the transfers, a uniform, non-means-tested transfer is given to all, while practically everyone is taxed, you move from a NIT to a genuine basic income (BI). Optimal income tax theory, initiated by Mirrlees (1971), formally makes the case for a minimum level of income, which is intended for the least well-off members of society. However, it does not distinguish between providing this minimum income through a BI or through a NIT. In Chapter 2, these two instruments are compared and it is shown that BI has a number of advantages over NIT. It limits the problem of the inactivity trap; it leads to fairer intra-household distribution; it lowers administrative costs; and it reduces stigmatization. I also show that the size of the minimum income guarantee shrinks as the arsenal of fiscal instruments is expanded to include tagging (of specific categories of recipients), workfare, or in-kind transfers. However, introducing a concern with stigmatization at least partly offsets the advantages of such instruments. Chapter 3 questions the standard result according to which costless tagging is optimal from a utilitarian standpoint. Assuming that the intensity of stigmatization associated with tagging welfare recipients is exogenous and heterogeneous, it is shown that it should be recommended only under certain assumptions. The sensitivity of the utility functions to stigmatization as well as the distribution of skills play a decisive role. In Chapter 4, stigmatization is endogenously influenced by the extent of welfare fraud. Costly monitoring enables the decrease of the number of undeserving welfare recipients and thus, the reduction of stigmatization. Here, tagged and untagged people are modeled slightly differently than they were originally modeled by Akerlof (1978). Tagged people do not work while all the untagged people do work. It is assumed that the impact of the handicap on the level of utility is distributed on an infinite support. Then, I show that a tagging system is always optimal, whatever its (finite) costs. This is also true when no monitoring at all is optimal.
- Affiliations
Citations
Jacquet, L. (2003). Optimal redistribution when stigma matters. https://hdl.handle.net/2078.5/43399