Showing papers by "Joshua D. Angrist published in 1996"

Identification of Causal Effects Using Instrumental Variables

Abstract: We outline a framework for causal inference in settings where assignment to a binary treatment is ignorable, but compliance with the assignment is not perfect so that the receipt of treatment is nonignorable. To address the problems associated with comparing subjects by the ignorable assignment—an “intention-to-treat analysis”—we make use of instrumental variables, which have long been used by economists in the context of regression models with constant treatment effects. We show that the instrumental variables (IV) estimand can be embedded within the Rubin Causal Model (RCM) and that under some simple and easily interpretable assumptions, the IV estimand is the average causal effect for a subgroup of units, the compliers. Without these assumptions, the IV estimand is simply the ratio of intention-to-treat causal estimands with no interpretation as an average causal effect. The advantages of embedding the IV approach in the RCM are that it clarifies the nature of critical assumptions needed for a...

Children and Their Parents' Labor Supply: Evidence from Exogenous Variation in Family Size

Abstract: Although theoretical models of labor supply and the family are well developed, there are few credible estimates of key empirical relationships in the work-family nexus. This study uses a new instrumental variable, the sex composition of the first two births in families with at least two children, to estimate the effect of additional children on parents' labor supply. Instrumental variables estimates using the sex mix are substantial but smaller than the corresponding ordinary least squares (OLS) estimates. Moreover, unlike the OLS estimates, the female labor supply effects estimated using sex-mix instruments appear to be absent among more educated women and women with high-wage husbands. We also find that married women who have a third child reduce their labor supply by as much as women in the full sample, while there is no relationship between wives' child-bearing and husbands' labor supply. Finally results to estimates produced using twins to generate instruments. Estimates using twins instruments are very close to the estimates generated by sex-mix instruments, once the estimators are corrected for differences in the ages of children whose birth was caused by the instruments. The estimates imply that the labor supply consequences of child-bearing disappear by the time the child is about 13 years old.

Children and their parents labor supply: evidence from exogenous variation in family size.

Abstract: Although theoretical models of labor supply and the family are well developed there are few credible estimates of key empirical relationships in the work-family nexus. This study uses a new instrumental variable the sex composition of the first two births in families with at least two children to estimate the effect of additional children on parents labor supply [in the United States]. Instrumental variables estimates using the sex mix are substantial but smaller than the corresponding ordinary least squares (OLS) estimates. Moreover unlike the OLS estimates the female labor supply effects estimated using sex-mix instruments appear to be absent among more educated women and women with high-wage husbands. We also find that married women who have a third child reduce their labor supply by as much as women in the full sample while there is no relationship between wives child-bearing and husbands labor supply. (EXCERPT)

Identification of Causal Effects Using Instrumental Variables: Comment

Abstract: We outline a framework for causal inference in settings where assignment to a binary treatment is ignorable, but compliance with the assignment is not perfect so that the receipt of treatment is nonignorable. To address the problems associated with comparing subjects by the ignorable assignment-an intention-to-treat analysis-we make use of instrumental variables, which have long been used by economists in the context of regression models with constant treatment effects. We show that the instrumental variables (IV) estimand can be embedded within the Rubin Causal Model (RCM) and that under some simple and easily interpretable assumptions, the IV estimand is the average causal effect for a subgroup of units, the compliers. Without these assumptions, the IV estimand is simply the ratio of intention-to-treat causal estimands with no interpretation as an average causal effect. The advantages of embedding the IV approach in the RCM are that it clarifies the nature of critical assumptions needed for a causal interpretation, and moreover allows us to consider sensitivity of the results to deviations from key assumptions in a straightforward manner. We apply our analysis to estimate the effect of veteran status in the Vietnam era on mortality, using the lottery number that assigned priority for the draft as an instrument, and we use our results to investigate the sensitivity of the conclusions to critical assumptions.

Schooling and Labor Market Consequences of the 1970 State Abortion Reforms

Abstract: This study uses the 1970 state abortion reforms to estimate the effect of teen and out-of-wedlock childbearing on the schooling and labor market outcomes of mothers observed in 1980 and 1990 Census microdata. Reduced-form estimates suggest that state abortion reforms had a negative impact on teen marriage, teen fertility, and teen out- of-wedlock childbearing. The teen marriage effects are largest and most precisely estimated for white women while the teen fertility and out-of-wedlock childbearing effects are largest and most precisely estimated for black women. The relatively modest fertility and marriage consequences of abortion reform for white women do not appear to have changed schooling or labor market outcomes. In contrast, black women who were exposed to abortion reforms experienced large reductions in teen fertility and teen out-of-wedlock fertility that appear to have led to increased schooling and employment rates. Instrumental variables estimates of the effects of teen and out-of- wedlock childbearing on the schooling and employment status of black women, using measures of exposure to abortion reform as instruments, are marginally significant and larger than the corresponding OLS estimates.

Short-run demand for Palestinian labor

Abstract: Palestinian residents of the West Bank and Gaza Strip who work in Israel generally earn more than Palestinians employed locally, but this wage premium is highly volatile. Beginning with the 1987 Palestinian uprising, changes in wage differentials by work location parallel Palestinian absences from work in Israel. This article interprets changing location differentials in response to exogenous shocks as movements along an Israeli demand curve for migrant workers. Estimates of a model of the West Bank and Gaza Strip labor market are used to evaluate the effect of policies governing Palestinian access to the Israeli labor market.

Schooling and Labor Market Consequences of the 1970 State Abortion Reforms

Abstract: This study uses the 1970 state abortion reforms to estimate the effect of teen and out-of-wedlock childbearing on the schooling and labor market outcomes of mothers observed in 1980 and 1990 Census microdata. Reduced-form estimates suggest that state abortion reforms had a negative impact on teen marriage, teen fertility, and teen out- of-wedlock childbearing. The teen marriage effects are largest and most precisely estimated for white women while the teen fertility and out-of-wedlock childbearing effects are largest and most precisely estimated for black women. The relatively modest fertility and marriage consequences of abortion reform for white women do not appear to have changed schooling or labor market outcomes. In contrast, black women who were exposed to abortion reforms experienced large reductions in teen fertility and teen out-of-wedlock fertility that appear to have led to increased schooling and employment rates. Instrumental variables estimates of the effects of teen and out-of- wedlock childbearing on the schooling and employment status of black women, using measures of exposure to abortion reform as instruments, are marginally significant and larger than the corresponding OLS estimates.

The Effect of Teen Childbearing and Single Parenthood on Childhood Disabilities and Progress in School

Abstract: be responsible for poor health and low levels of schooling among the children of young mothers. This paper uses special disability and grade repetition questions from the school enrollment supplement to the 1992 Current Population Survey to estimate the effect of maternal age and single parenthood on children's disability status and school progress. Our results suggest that there is little association between maternal age at birth and children's disabilities. But the children of teen mothers are much more likely to repeat one or more grades than other children, and within-household estimates of this relationship are even larger than OLS estimates. The grade repetition findings from the CPS are replicated using a smaller sample from the National Longitudinal Survey of Youth. Another finding of interest is that having a father in the household is associated with lower disability prevalence and fewer grade repetitions. But many of the effects of single parenthood on disability, as well as the effect on grade repetition, appear to be explained by higher incomes in two-parent families.

The Effect of Teen Childbearing and Single Parenthood on Childhood Disabilities and Progress in School

Abstract: be responsible for poor health and low levels of schooling among the children of young mothers This paper uses special disability and grade repetition questions from the school enrollment supplement to the 1992 Current Population Survey to estimate the effect of maternal age and single parenthood on children's disability status and school progress Our results suggest that there is little association between maternal age at birth and children's disabilities But the children of teen mothers are much more likely to repeat one or more grades than other children, and within-household estimates of this relationship are even larger than OLS estimates The grade repetition findings from the CPS are replicated using a smaller sample from the National Longitudinal Survey of Youth Another finding of interest is that having a father in the household is associated with lower disability prevalence and fewer grade repetitions But many of the effects of single parenthood on disability, as well as the effect on grade repetition, appear to be explained by higher incomes in two-parent families

Conditional Independance in Sample Selection Models

Abstract: In this note, I describe the conditional independence properties that make it possible to use the selection propensity score to control selection bias in a general sample selection model. The resulting characterization does not rely on a latent index selection mechanism and imposes no structure other than an assumption of independance between the regression error term and the regressors in random samples. This approach leads to a simple rule that can be used to determine if conditioning on the selection propensity score is sufficient to control selection bias.

Children and Their Parents' Labor Supply: Evidence from Exogenous Variation in Family Size

Abstract: Although theoretical models of labor supply and the family are well developed, there are few credible estimates of key empirical relationships in the work-family nexus. This study uses a new instrumental variable, the sex composition of the first two births in families with at least two children, to estimate the effect of additional children on parents' labor supply. Instrumental variables estimates using the sex mix are substantial but smaller than the corresponding ordinary least squares (OLS) estimates. Moreover, unlike the OLS estimates, the female labor supply effects estimated using sex-mix instruments appear to be absent among more educated women and women with high-wage husbands. We also find that married women who have a third child reduce their labor supply by as much as women in the full sample, while there is no relationship between wives' child-bearing and husbands' labor supply. Finally results to estimates produced using twins to generate instruments. Estimates using twins instruments are very close to the estimates generated by sex-mix instruments, once the estimators are corrected for differences in the ages of children whose birth was caused by the instruments. The estimates imply that the labor supply consequences of child-bearing disappear by the time the child is about 13 years old.