Resource title

Discrete Choice Labor Supply : Conditional Logit vs. Random Coefficient Models

Resource image

image for OpenScout resource :: Discrete Choice Labor Supply : Conditional Logit vs. Random Coefficient Models

Resource description

Estimating labor supply functions using a discrete rather than a continuous specification has become increasingly popular in recent years. On basis of the German Socioeconomic Panel (GSOEP) I test which specification of discrete choice is the appropriate model for estimating labor supply: the standard conditional logit model or the random coefficient model. To the extent that effect heterogeneity is present in empirical models of labor supply functions, the application of a random coefficient model is necessary to avoid biased estimates. However, because of the complex structure, random coefficient models defy calculating confidence intervals of marginal effects or elasticities. Therefore, if heterogeneity is nonexistent or does not lead to a significant bias in the derived labor supply elasticities, standard discrete choice models provide the more favorable choice. Due to their simple structure, conditional logit models are far less computational intensive providing standard tools to calculate confidence intervals of elasticities. My findings suggest that effect heterogeneity is present when estimating a discrete choice model of labor supply drawing on data of the GSOEP. However, the labor supply elastisities derived form the specifications with and without random effects do not differ significantly. That leads to the conclusion that the standard discrete choice model, attractive for its simple structure, provides an adequate model choice for the analysis of labor supply functions based on the GSOEP.

Resource author

Peter Haan

Resource publisher

Resource publish date

Resource language

eng

Resource content type

text/html

Resource resource URL

http://hdl.handle.net/10419/18091

Resource license

Adapt according to the presented license agreement and reference the original author.