Evaluating Alternative Representations of the Choice Sets in Models of Labor Supply

Abstract

During the last two decades, the discrete choice modeling of labor supply decisions has become increasingly popular. Within the literature adopting this approach, however, there are two potentially important issues that so far have not been given the attention they might deserve. A first issue concerns the procedure by which the discrete alternatives are selected to enter the choice set. Most authors choose (not probabilistically) a set of fixed points identical for every individual. Some authors adopt instead a sampling procedure and also assume that the choice set may differ across households. A second issue concerns the availability of the alternatives. Most authors assume all the values of hours of work within some range are equally available. At the other extreme, some authors assume only two or three alternatives (for example, nonparticipation, part-time and full-time) are available for everyone. Some studies account instead for the fact that not all the hour opportunities are equally available to everyone specifying a probability density function of opportunities for each individual. In this paper we explore by simulation the implications of (i) the procedure used to build the choice set (fixed alternatives versus sampled alternatives); (ii) accounting or not accounting for a different availability of alternatives. The results of the evaluation performed in this paper show that the way the choice set is represented has little impact on the fitting of observed values, but a more significant and important impact on the out-of-sample prediction performance. Thus, the treatment of the choice sets might have a crucial effect on the result of policy evaluations.

Structural Labor Supply Models and Wage Exogeneity

“…The first decision in the estimation relates to the construction of the choice set (see Aaberge et al, 2009, for a detailed discussion of this issue). Most authors simply pick a set of representative levels of hours of work and assume (small) identical choice sets for the whole population.…”

“…As working hours are typically concentrated in only few hours categories, most authors include in their models fixed costs of working, hours restrictions, or both. Aaberge et al (1995Aaberge et al ( , 2009) Blundell et al ( , 2000 Quadratic…”

“…Fixed costs and hours restrictions can also be interpreted as measures for the availability of the respective choice alternatives (Aaberge et al, 2009). Less than half of the models explicitly allowed for stigma effects and non-take-up of welfare benefits.…”

Labour supply and taxation with restricted choices

“…The φ and γ parameters refer to the exponents on hours (non-market work) and on consumption as described in the utility specification (1) of section 3. The next panel refers to the parameters that influence the marginal utility of hours through the specification of ln(a) in equation (2). Following these are the parameters of fixed costs (3).…”

“…There is a long history of incorporating such restrictions into labour supply models, including, Aaberge (1999Aaberge ( , 2009), Altonji and Paxson (1992), Bloemen (2000Bloemen ( , 2008, Dickens and Lundberg (1993), Van Soest et al (1990), Ham and Reilly (2002) and Dagsvik and Strom (2006); see also the recent discussion in Chetty (2012). The ideas we develop here place these models of hours retrictions in a constrained rational choice setting in which the set of alternative choices on offer is restricted.…”