Search Results

DESIGN/METHODS: We use data from the National Longitudinal Survey of Youth (NLSY), a nationally representative sample of 14-21 year olds in 1978. Our analysis includes all female respondents from this longitudinal survey who gave birth since 1978. Using Ordinary Least Squares regression for children between the ages of 5 and 7 years of age, we model the child's score on the Behavioral Problems Index (BPI) as a function of mother's age, child characteristics, mother's social characteristics, and grandparent's social characteristics.

RESULTS: We report on several models. In the first, maternal age (age 15-19 at first birth vs age 20 or older) is used as the sole predictor of BPI. Status as a teen mother is highly predictive of a child's later behavioral problems. As control variables are added in subsequent models, the ability of maternal age to predict child behavior problems decreases dramatically. Rather, measures of poverty, mother's level of education, and grandparent's level of education best predict behavioral difficulties in the children.

CONCLUSIONS: A simple model suggests that teen motherhood predicts behavior problems. This relationship disappears when other social factors are considered. This work suggests that adolescent childbearing and parenting cannot be considered in isolation from the factors that promote its occurrence. Pediatricians working with young mothers must consider the larger social influences at work in the lives of adolescents.

DATA: In this study, we will analyze data from the National Longitudinal Survey of Youth 79 (NLSY79) and Children of the NLSY79. The NLSY79 is a panel survey of 12,686 men and women who were 14-21 years old in 1978 and is designed to gather detailed data about employment, education/training, income, fertility, and family characteristics. The data are nationally representative of people living in the United States in 1978. The Children of the NLSY79 is a supplemental survey of all children born to the 6,283 women in the original sample. The supplemental survey provides data on the cognitive development of the children born to these mothers.

METHODS: We use a lagged dependent variables approach to model the relationship between early childhood cognitive scores and unemployment insurance receipt. It is possible that parental characteristics associated with employment termination and layoffs are also correlated with child cognitive outcomes. To minimize this selection bias threat, our sample includes only those families who faced an unanticipated firm closure. Our baseline models control for as many measurable characteristics as possible that might differ between short-term and long-term UI participants and be related to child outcomes. However, endogeneity is still a substantial problem because those who leave UI may be systematically different from those who experience prolonged exposure to UI in unmeasured ways that are correlated with child outcomes. To further reduce the possibility of selection bias, we use a family fixed effects model that compares siblings' outcomes when UI was received to those when UI was not received.

POLICY IMPLICATIONS: This research can inform public policy in important ways. UI has often been criticized for creating disincentives to find work. However, one very good reason for states to provide UI to displaced workers is to minimize the negative effects of unemployment spells that might be associated with reduced income levels. Currently, very little is known about the effects of parental UI receipt on children's cognitive or behavioral outcomes. This study seeks to address this area of need in the research base, given the current focus and multiple expansions of UI eligibility across the states.