Wage Discrimination by Sex and its Correlates Argentina and Mexico

Henry A. Selby
Department of Anthropology,
University of Texas
Austin, TX 78712
Fax: 512-471-6535
E-mail: selbyhenry@mail.utexas.edu

An examination of the income distributions for four cities in Mexico and Argentina reveal that there is systematic discrimination against women even given the disparity in hours worked. The Blinder Oaxaca algorithm was used to determine how much of the di scrimination was accounted for by undercompensation to the usual human capital variables, and around 50% was found. The other half of the variance is residual, or to be accounted for by "pure discrimination". The human capital variables were education and job experience (as estimated by age).

Wage determination equations were also derived using education, marital status, experience, experience-squared, industry type, job type and city as variables. Education was found to behave as it should, with higher educational attainments being associated with higher compensation. Marital status was more interesting: it was positive for the men and negative for the women in both countries, suggesting that recent changes in household structure that have favored the increase of mother-only households have n ot entailed the same financial disasters that were predicted. And examination of work experience and the squared value shows that wages peak at around 30 years of experience (somewhat earlier for Mexico) declining after that. Pensions in Argentina are sug gested as the differential here, since people who are pensioned often continue working because the pensions are so inadequate.
Work is continuing not just on income distributions, but on the factors that facilitate or impeded the insertion of workers into the work force. For these purposes a household sample has been constructed, but only preliminary analyses are avaialble.