Labor Force Participation of Married Women When Wages and Employment Are Risky: Theory and Estimation

144 p.Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 1986.Over the past 25 years, literature on the growth of married women's labor force participation, has concluded virtually without exception that the principal source of the growth has been the concurrent growth of women's real wages. In fact, while the rapid real wage rate growth of the 1950's and 60's virtually disappeared during the 1970's, participation rates grew steadily throughout the period. Thus, the principle source of participation rate growth for married women, as identified by the literature of the last three decades, essentially vanished during the 1970's, while participation rates continued their rapid increase.This thesis asks whether the economic riskiness fostered by the economic instability of the 1970's may have contributed to the growth of participation rates by married women during this period.A two-period theoretical model is developed in which employment, real wages and real incomes are risky in the second period. Comparative static analysis shows that decreases in the expected probability of employment, and in the expected mean real wage and the expected mean level of real income encourage current participation by married women. Moreover, a mean-constant increase in the variance of the woman's expected real wage distribution (this is interpreted as an increase in the overall level of real wage risk) also tends to pull her into the labor force. The effect of a mean-constant increase in the variance of her husband's expected real income distribution is ambiguous, but seems likely to discourage her participation.Empirical tests are carried out through the use of highly aggregated annual time series data covering the years 1956-1982. Engle's ARCH model is used to estimate expected variances of the real wage and income distributions. The estimates are consistent with previous studies, and confirm the importance of real wage growth. Expectations variables almost always have the expected signs, and are often statistically significant. Within-sample predictions of the participation rate are enhanced by the inclusion of the expectation terms.U of I OnlyRestricted to the U of I community idenfinitely during batch ingest of legacy ETD