Date of Award

8-2012

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Legacy Department

Economics

Advisor

Tamura, Robert F

Committee Member

Mroz , Thomas A

Committee Member

Simon , Curtis J

Committee Member

Jerzmanowski , Michal M

Abstract

The first chapter examines the effect of the number of years children spend living with a single-parent family instead of a two-parent family on children's completed schooling, based on a sample of children from the PSID. To deal with the endogeneity of mothers' family structure decisions, I exploit the variation across states and over time in unilateral divorce laws, unmarried fertility ratios, welfare rules, earned income tax credit rates, and labor market conditions that generate plausibly exogenous changes in mothers' family structure choices. I construct a set of extensive measures for these contextual variables and use them as instruments to estimate a child's human capital production function. Instrumental variable estimation indicates that one additional year spent in a single-parent family during childhood (ages 0-15) can cause a loss of 0.145 years in schooling. This result implies that the differences in family structure experiences over the early life course between white and nonwhite children can explain roughly 76% of the gap in educational attainment between the two groups. On the other hand, ordinary least-squares estimation only suggests 13%.
Children born to unmarried parents may receive lower human capital investments in youth, and therefore may be less likely to finish high school or to attend college. The second chapter explores these effects empirically using state level data over the period 1940-2000. We find that a steady-state increase in unmarried fertility ratio of 100 per 1,000 child births could lead to a 4.6 percent drop in high school graduation rate and a steady-state 4.2 percent decline in secondary school enrollment in the long-run. This result is important since Heckman and Lafontain (2010) found that since the late 1960s the high school graduation rate has fallen by 4-5 percentage points, despite the growing wage differentials between high school graduates and dropouts. Our analysis implies that the rise in unmarried fertility predicts a ceteris paribus drop in high school graduation rate of about 6.6% in the same time period, thus provides an important explanation for the dropout problem in recent decades.. Moreover, our results indicate a very weak link between abortion and child education, in contrast to the strong effect of abortion on crime documented in the literature.

Included in

Economics Commons

Share

COinS