Key points are not available for this paper at this time.
In this pnper, two models ojindividunl labor supply are discussed. The first one is the hy now classical Hnusman-type model with convez piecewise linear budxet cnnstraints, in which both random prejerences and optimizatian errors are incorpornted by means ajnormally distrihuted random vnrinb(cs. Estimnted coefjScients are plnresible but the modef has the shortroming Niat unemployment jor males is nat cnptured nnd that the simufated hours dl.stributian misses the spikes in the sample distribution of warking hours. Therejore, an alternatis.e model is inlroduced which explicitly takes into account demand side restrictions on warking hnurs. The diJrerence with the standnrd model is the replacement of the optimization error by the assrrmption that each individunl can choose jrom afinite set oj wcrge hotus packages and either picks the job ofjer yielding highest utility or decides nor to work. It turns out that this model cnptrnes the sample distribution of working hours very well, jor mnles as K~ell a.c females. Wage and income elasticities according to the two models are similar nnd in line with other recent findings in TheArrhur vnn Socst is a projessor of ecnnomics at Tilburg Universiry in Thr Nenc~rland.r.Isolde Woirriez was a projessnr ojeconomics at Tilburg when the work nn rhis pnper hegnn..She is currendy af l,eyden University in The Nethtrlands.Tht aurhors nre grarrju!ro Peter Knrneman, the participants of the conjerence, and two anonymous rejerees for helpj~d rnmments.They rhank the OrgnniZntinn far Strategic Labar Market Resenrrh for kindly ntakinR Ihe dam nvailnble jor u.ce in this research.Financinl support hy rhe Netherlands OrRnnizntinn jur Scirnlific Re.cenrch is gratejrdly ncknmctedr;ed hy the secnnd nulhor.
Soest et al. (Mon,) studied this question.