This paper investigates the role of labor market attachment, on-the-job human capital accumulation, and occupational sorting in the narrowing gender wage gap over the past three decades. Many of the observe changes, such as labor market attachment, experience and occupation choice are not exogenous. We formulation of a dynamic general equilibrium model of earnings determination and estimate it using the PSID. In the model, statistical discrimination, labor market attachment, experience and occupational sorting are endogenous. In our model, workers have heterogeneous preferences regarding the degree of the labor market attachment and labor supply. Preferences are known privately to workers. It is costly for firms to replace workers; these costs may also vary across jobs (for example, some jobs may require greater investment in skill acquisition). The returns to labor market experience also differs across jobs. Thus, different jobs may offer different entry-level wages, hours and wage growth prospect. Employers observe age, experience, education, and other skills. Thus, weak labor market attachment may be more costly at different stages of a worker's career, and may vary with workers' skills. Based on observable characteristic, employers form beliefs regarding the degree of labor market attachment when they offer wage contracts. Wage contracts consist of job assignment, hours and earning. Thus, in equilibrium, statistical discrimination, based on degree of labor market attachment, may arise. We assume contracts cannot bind workers and employers to long-term employment. If firms believe that women are less likely to commit to long employment periods, they offer them lower wages. As a result, women work less hours and sort into occupations which have lower returns to labor market experience. The model allows us to estimate the gender wag gap accounting for the endogeneity of discrimination, labor market experience and occupational sorting. We then explore the effect of occupation specific aggregate productivity shocks, changes in cost of participation in the labor marker (may result from improved home production technology), and demographic changes on the decline in the gender wage gap. We developed a new multi-step semiparametric estimation strategy that allows us to estimate the model without solving it. This is particularly important in this class model which may exhibit multiple equilibria. The idea is to estimate the principal problem in the first step, along with other inputs from the agents Euler equation of consumption. In the second step, these estimates are used to nonparametrically estimate the firms' beliefs, the agents choice-specific probabilities and their derivatives. In the Final step, these estimates are combined with a tractable alternative representation of the agents choice-specific valuation to form moment conditions for the estimation of the structural parameters of the agents utility function. The estimates of the structural parameters are Root-N consistent (where N is the sample size) and asymptotically normal although the second step is estimated nonparametrically.
Download Info
To download:
If you experience problems downloading a file, check if you have the
proper application to
view it first. Information about this may be contained
in the File-Format links below. In case of further problems read
the IDEAS help
page. Note that these files are not on the IDEAS
site. Please be patient as the files may be large.
Publisher Info
Paper provided by Carnegie Mellon University, Tepper School of Business in its series GSIA Working Papers with number
2006-E40.
Length: Date of creation: Date of revision: Handle: RePEc:cmu:gsiawp:1143615504
Contact details of provider: Postal: Tepper School of Business, Carnegie Mellon University, 5000 Forbes Avenue, Pittsburgh, PA 15213-3890 Web page: http://www.tepper.cmu.edu/