Search Results

The Committee on National Statistics (CNSTAT), in consultation with the Institute of Medicine, convened a two-day workshop on October 14-15, 1999 to identify ways to advance the often conflicting goals of exploiting the research potential of micro-data while maintaining the appearance and reality of preserving confidentiality. Special attention was given to longitudinal data that have been linked to administrative records, because such data have greater research potential but are also more vulnerable to disclosures. The workshop brought together data producers from federal agencies and research organizations; data users, including academic researchers; and experts in statistical disclosure limitation techniques, confidentiality policies and administrative and legal procedures. Norman Bradburn, National Opinion Research Center, chaired the workshop.

The purpose of this workshop was to review the benefits and risks of providing public use research data files and to explore alternative procedures for restricting access to such data, especially longitudinal survey data that have been linked to administrative records. The workshop discussions addressed the impact of measures designed to reduce disclosure risk on each group involved--survey respondents, data producers, and data users. Workshop participants identified best current practices and suggested improvements to maximize the social return on investments in such databases, while fully complying with legal and ethical requirements.

The workshop discussion accomplished several goals:

• reviewed current practices and concerns of federal agencies and other data producing organizations;
• reviewed the types of research that are enhanced, or only made possible, using linked lon gitudinal data;
• provided an overview of administrative arrangements to preserve confidentiality;
• identified ways to foster data accessibility in secondary analysis; and
• assessed the utility of statistical methods for limiting disclosure risk.

The Committee prepared a report covering workshop proceedings, which will include presentation and discussion summaries, and suggestions for future research. This report will be published in September 2000.

Plans are underway for developing follow-on activities to this workshop that could include convening a panel that would oversee one or more workshops and would then write a report with recommendations. It seems likely that future work would shift the focus away from what agencies are currently doing--which, after the first workshop, is fairly clear--and toward looking in more detail at: (1) cutting edge statistical techniques for manipulating data in ways that preserve important statistical properties and allow for broader general data release; (2) new, less burdensome ways (e.g., internet, remote access etc.) of providing researchers with access to restricted data sets; (3) researcher reaction to these potential new directions; and (4) the issue of licensing coupled with graduated civil and criminal penalties for infringement. Future efforts would focus more carefully on longitudinal microdata.

This activity is sponsored by the National Institute on Aging, the Agency for Health Care Policy and Research, the Bureau of Labor Statistics, the National Library of Medicine, the Office of the Assistant Secretary for Planning and Evaluation of the U.S. Department of Health and Human Services, the Office of Planning and Evaluation of the U.S. Department of Health and Human Services, and the Social Security Administration

I examine three measures of past 30 day alcohol use in middle age: number of alcoholic drinks, heavy drinking (≥ 60 drinks), and weekly binge drinking. Data are drawn from the National Longitudinal Survey of Youth 1979 Cohort. I proxy economic conditions at school-leaving with the state unemployment rate, and exploit variation generated by volatility in the U.S. business cycle between 1976 and 1992 to identify alcohol use effects. Because the severe recession of the early 1980s lies within this time period, I compare cohorts that left before, during, and after this recession. To address the potential endogeneity of the time and location of school-leaving, I instrument the school-leaving state unemployment rate.

Design and Methods: Group-based trajectory modeling and National Longitudinal Survey of Youth 1979 data from 1990 to 2008 were used to quantify weight trajectories through midlife for 10,038 young adult men and women stratified by BMI category. Logistic regression was used to assess the association of trajectory membership with obesity-related conditions (hypertension, diabetes, arthritis) in middle age.

Results: Annual weight gain averaged 0.53 kg (1.17 lb) across the entire sample. However, there was considerable variation by and within BMI categories. More than 98% of men and 92% of women were on upward-sloping trajectories, generally moving into a higher BMI category by middle age. Those who experienced early and rapid weight gain during young adulthood were most likely to be on a steeper trajectory and had greater risks for obesity-related conditions.

Conclusion: This study points to the health and weight benefits of entering young adulthood with a normal BMI, but further reveals that this is no guarantee of maintaining a healthy weight through midlife. For those who are young adults today, weight maintenance is unlikely to occur without significant environmental or technical innovation.

A recent strand of literature in demography argues that young unmarried Americans value marriage so highly that it is perceived as a family status to be chosen after certain economic preconditions are fulfilled--after they have achieved the so-called "white picket fence dream" (a house, surplus income etc.). Motivated by these claims, in the second essay we use data from the NLSY79 to examine whether there is any direct relationship between the individual's housing and financial assets and his/her transition into marriage or cohabitation. For both men and women, analysis using a proportional hazard model indicates a positive association of asset ownership with transition into marriage, but not with transition into cohabitation. However, instrumental variables probit estimations, designed to account for the endogeneity of asset-accumulation, either remove the statistical significance of the association between asset ownership and family union transitions, or identify effects that are in the opposite direction to those derived from the time-to-event analysis, indicating dissuading effects of asset ownership on transition to marriage.

The existing theoretical literature on household decision-making makes no distinction between different institutional processes of household formation, namely, cohabitation and marriage. In the third essay, we develop a simple two-period model of family union that distinguishes cohabitation and marriage. The analytical results of the model suggest that compared to marital unions, cohabiting unions have higher risk of dissolution in the future, and involve less intra-household specialization. The model also indicates that improved labor market conditions for men provide stronger incentives for marriage than for cohabitation; and that cost of divorce affects married women's labor supply choice.

The second part of this dissertation evaluates the impact of being born under negative conditions in the labor market on human capital formation, and what parental behavior could be leading to those effects. I estimate the impact of the unemployment rates on children's assessment outcomes in cognitive and noncognitive skills. Counterintuitively, the results suggest that higher unemployment rates are linked to positive child development outcomes later in childhood. In my main specification, an increase of 1 percentage point in state unemployment causes an increase of 2.5% of a standard deviation in cognitive test scores after controlling for income at birth, hours worked at birth, and other variables.

The third chapter exploits panel data from the 1979 National Longitudinal Survey of Youth to investigate the efficacy of vocational education as a means of retraining adults. Using Altonji, Elder and Taber's (2005) method to correct for selection into training, I estimate a 10-year profile of returns to vocational training for those changing jobs. While fairly imprecise, the estimates I obtain suggest that the average spell of vocational training increases hourly wages by 6 percent for the first couple of years after training and around 8-10 percent thereafter, with the effects persisting for at least 10 years. I also find positive effects of training on employment, weeks worked, and income, but these results are both less precise for each year and less consistent across years. Under further assumptions regarding the cost of training and the pattern of returns after ten years, I also show that the estimated returns are sufficient to make vocational training a sound investment.

Published as: Manski, Charles F. and John V. Pepper.
"Monotone Instrumental Variables With An Application To The Returns To Schooling," Econometrica 68,4 (July 2000): 997-1010. Also: http://www.jstor.org/stable/2999533

Methods: Empirical, English-language articles examining the effects of economic change at either the aggregate or individual level on birthweight, length of gestation, neonatal mortality and the secondary sex ratio were identified using PubMed and ISI Web of Knowledge. Studies were organized by level of analysis and birth outcome and evaluated for internal and external validity.

Results: One individual-level study reported a strong association between individual shift to inadequate employment and decreased birthweight. Of seven aggregate-level studies on birthweight, five exhibited moderate to strong validity but reported inconsistent findings. Similarly, findings from five studies (four with moderate to strong validity) examining rates of neonatal mortality reported inconsistent findings. Three of four moderate to strong studies reported a reduced secondary sex ratio following economic contraction.

Conclusions: Associations between economic contraction and birthweight, neonatal mortality and the secondary sex ratio remain speculative. Consensus on methodology is needed to compare findings across studies. Further research on economic contraction and the secondary sex ratio, as well as individual-level birthweight and length of gestation, is warranted.

Methods: Data on health behaviors among 7,074 pregnancies in the National Longitudinal Survey of Youth 1979 were linked to monthly unemployment rates in maternal state of residence. The study examined associations between exposure to unexpected economic contraction (higher than expected state-level unemployment) during each trimester of pregnancy and maternal smoking, alcohol use, and gestational weight gain using generalized linear models.

Results: Economic contraction was not associated with maternal smoking or gestational weight gain. Associations between economic contraction and maternal alcohol use differed by maternal race/ethnicity and education. Among Black/non-Hispanic women, exposure to economic contraction during the first and second trimester of pregnancy were associated with a 42% (95% CI: 1.08, 1.85) and 33% (95% CI: 1.01, 1.74) increased risk of alcohol use, respectively.

Conclusions: Findings suggest that exposure to extreme economic contraction during pregnancy may be associated with increased use of alcohol with differences by maternal race/ethnicity and educational attainment. Economic contraction was not associated with other maternal pregnancy behaviors.

Methods. I examined the association between maternal exposure to state-level labor market perturbations during each trimester of gestation and fetal growth, as measured by birth weight for gestational age percentile. The study population included 6,715 gestations and births between 1982 and 2000 to women enrolled in the National Longitudinal Survey of Youth 1979 (NLSY79). I calculated birth weight for gestational age percentiles using national reference data and categorized births <10th percentile as small for gestational age (SGA).

I defined perturbations to labor market security as months in which the state unemployment rate was higher than its statistically expected value (i.e., unexpectedly high labor market insecurity) and months in which the state unemployment rate was lower than its statistically expected value (i.e., unexpectedly high security). I derived statistically expected values using ARIMA modeling me thods to account for autocorrelation. Gestations in the NLSY79 were classified as either exposed or unexposed to labor market insecurity or security in the first, second, and third trimester if one of these labor market perturbations occurred in the maternal state of residence during that trimester.

I used linear and logistic regression models to examine the association between labor market perturbations in each trimester and birth weight percentile and odds of SGA. I also examined whether any observed associations differed by maternal race/ethnicity, childhood socioeconomic status, educational attainment, marital status, employment status, or poverty status. Finally, I explored whether any observed associations were mediated by individual economic change (i.e., changes in maternal employment status or household income) or maternal pregnancy behaviors (i.e., smoking, first trimester utilization of prenatal care, or net gestational weight gain). If associations were mediated by one these factors, I calculated the proportion of the total association explained by that factor.

Results. Exposure to labor market insecurity in the first trimester was significantly associated with a decrease in birth weight for gestational age of 4.05 percentile points (95% CI = -6.87, -1.22) and higher odds of SGA (OR = 1.50, 95% CI = 1.21, 1.86). Exposure to labor market insecurity in the second and third trimesters was not significantly associated with either outcome. Exposure to labor market security was not associated with birth weight for gestational age percentile or SGA.

The association between exposure to labor market insecurity in the first trimester and birth weight percentile differed significantly by maternal childhood SES, educational attainment, and employment status but not by race/ethnicity, marital status, or poverty status. Exposure to labor market insecurity in the first trimester was associated with decreases in birth weight percentile of 5.52 (95% CI = -10. 0, -1.04) and 8.66 points (95% CI = -14.04, -3.29) among women with average or high childhood SES, respectively, while the association was not significant among women with low childhood SES. Exposure to labor market insecurity in the first trimester was associated with a decrease in birth weight percentile of 9.22 points (95% CI = -15.77, -2.88) among women with <12 years educational attainment, while the association was not significant among women with 12 years or >12 years educational attainment. Exposure to labor market insecurity was associated with a decrease in birth weight percentile of 7.10 points (95% CI = -12.33, -1.87) and 10.27 points (95% CI = (-18.82, -1.71) among women keeping house and out of the labor force, respectively, while the association was not significant among employed and unemployed women.

My exploration of mediation by individual economic change and maternal pregnancy behaviors found that approximately 11% of the association between exposure to labor market insecurity in the first trimester and birth weight percentile was explained by net maternal gestational weight gain. The association also differed significantly by maternal smoking, with the association only significant among smokers. No other individual economic change or maternal pregnancy behaviors mediated greater than one percent of the association.

Conclusions. Findings support my hypothesis that fetal growth responds to a contemporary ecological perturbation, i.e., unexpectedly high labor market insecurity. Exposure to this perturbation appears to have more impact on fetal growth if it occurs in the first trimester of gestation. The finding that associations between exposure to labor market insecurity and birth weight percentile were stronger among women with high childhood SES, <12 years education, and those keeping house or out of the labor force suggests that these women may be more vulnerable to economic perturbations. Although further research on mediation is needed , initial findings suggest that maternal gestational weight gain may represent one pathway through which economic perturbations affect fetal growth.

Methods: We used a multilevel design to examine associations between exposure to state-level unexpected economic contraction during each trimester of gestation and birth weight for gestational age percentile and small for gestational age (SGA), using the National Longitudinal Survey of Youth 1979. We examined differences in observed associations by maternal educational attainment, race/ethnicity, employment status, and poverty status.

Results: Exposure in the first trimester was associated with a 3.7 percentile point decrease in birth weight for gestational age (95% confidence interval [CI] = −6.8 to −0.6). This association appeared stronger for women “keeping house” or with <12 years education. Exposure in the first trimester was also associated with increased odds of SGA (odds ratio = 1.5 [95% CI = 1.1 to 2.1]) and term SGA (odds ratio = 1.6 [95% CI = 1.2 to 2.3]).

Conclusions: Unexpected economic contraction during early pregnancy may be associated with reduced fetal growth.

STUDY DESIGN: We estimated associations between GWG (continuous and within categories defined by the Institute of Medicine), maternal prepregnancy body mass index, and each outcome in 4496 births in the National Longitudinal Survey of Youth 1979, which was a prospective cohort.

RESULTS: GWG (kilograms) was associated with decreased risk of SGA and increased risk of LGA, cesarean delivery, postpartum weight retention, and child overweight independent of maternal demographic and pregnancy characteristics. Gain above the Institute of Medicine guidelines was associated with decreased risk of SGA and increased risk of all other outcomes.

CONCLUSION: Excessive gain may have long-term consequences for maternal and child body size, but the benefits of lower gain must be balanced against risk of SGA. Copyright © 2010 Mosby, Inc. All rights reserved.

Chapter one of the dissertation provides background on America's complicated health care system where eligibility for subsidized health insurance depends on employment, spouses' employment, employer characteristics, family income and family structure. The idiosyncratic nature of the health care system almost guarantees that unintended consequences will arise from the interactions between individual choices, health insurance availability and government intervention. This chapter addresses some of these unintended consequences, while chapters two and three provide case studies.

Chapter two examines whether health and access to health insurance impact the decisions to marry and divorce. I argue that individuals in ill health face difficulties on the marriage market. Furthermore, since sickness during marriage is an unanticipated shock that reduces the size of the marriage surplus, sickness should contribute to divorce. Finally, I investigate if access to subsidized health insurance through marriage affects the probability of divorce. Proportional hazard models estimated on panel data from the National Longitudinal Survey of Youth indicate that ill health increases the time until first marriage. Furthermore, illnesses that were present at the time of marriage have no impact on divorce but health shocks that occur during the marriage increase the propensity to divorce for nonblack men. When the source of health insurance is added to the analysis it appears that sick women remain married to keep health insurance from their spouse's employer and are more likely to divorce when doing so would enable them to qualify for Medicaid.

Chapter three investigates the relationship between the minimum wage, employer-provided health insurance and part-time labor. I argue that minimum wages prevent the tradeoff of wages for benefits and result in lower rates of subsidized employer-provided health insurance. Additionally, increasing the minimum wages creates an incentive to hire more part-time workers who can be excluded from the firm's health insurance plan. The essay uses the 1988-1993 Current Population Surveys and exploits cross-state and federal variation in minimum wages. I find that higher minimum wages reduce the receipt of employer-provided health insurance, particularly for low-income workers. Furthermore firms, governed by the nondiscrimination law, switch to part-time workers who can be denied health insurance.

Methods: I apply hybrid mixed effects models that disaggregate within- and between person variation to investigate incarceration’s relationship with home ownership and net worth, using National Longitudinal Study of Youth data from 1985 to 2008. I also incorporate a set of mediation models in order to test for indirect effects of incarceration on wealth through earnings, health, and family formation.

Results: My results show that incarceration limits wealth accumulation. Compared to never-incarcerated persons, ex-offenders are less likely to own their homes by an average of 5 percentage points, and their probability of home ownership decreases by an additional 28 percentage points after incarceration. Ex-offenders’ net worth also decreases by an average of $42,000 in the years after incarceration.

Conclusions: When combined with previous research on incarceration, my findings show that incarceration acts as an absorbing status, potentially leading to the accumulation of disadvantage. Although incarceration’s negative effects on wealth accumulation were partially mediated by its relationship with earnings and family formation, incarceration directly affected home ownership and net worth. In most cases, former inmates began with flatter wealth trajectories and experienced additional losses after incarceration.

The first chapter addresses a long-standing question concerning the sources of the positive wage differential between large and small US cities. I build a spatial equilibrium model that I use to measure the contribution of three channels. In the model, larger cities are characterized by a higher frequency of interactions in the labor market--hence better matches between workers and firms (matching)--a better composition of peers workers learn from (knowledge diffusion), and positive sorting of high-skilled workers through migration across cities (sorting). I find that the aggregate implications of policies that change the size and composition of cities are determined by how such policies influence matching, knowledge diffusion, and sorting. Concretely, I show that an expansion of housing supply in large, productive, cities reduces the extent of sorting and knowledge diffusion in those cities. As a consequence, the aggregate income gain from implementing such policy is considerably smaller than in a hypothetical scenario in which the productivity of cities was invariant to the policy.

The second chapter extends the traditional search and matching model of the labor market to account for long-run growth in the efficiency of the search technology. We provide necessary and sufficient conditions under which such long-run growth does not trigger a secular decline in unemployment (consistently with US data), but it contributes to labor productivity growth. Intuitively, a higher meeting probability in the labor market allows firm-worker pairs to be more selective with respect to the quality of the matches they create. Simple calculations show that this channel may be responsible for approximately one fourth of US labor productivity growth over the last 30 years.

This paper describes recent patterns in the subsequent fertility of women in the United States by comparing rates of second and higher-order births for women with teen and nonteen first births. We expand upon previous work by paying particular attention to issues related to the spacing of second births. We pose and answer three questions concerning the pace and tempo of second and higher-order births. First, is an early first birth associated with a quicker overall pace of second births? Second, is an early first birth associated with higher second birth rates during the first few months postpartum? And third, does an early first birth or a closely spaced second birth speed the pace of higher-order births? For black women, we find that a first birth before age eighteen speeds the overall pace of second births by about 25 percent, and that part of this overall difference is due to a doubling of the pace of second births in the first 15 months postpartum. For white women, we find no association between a first birth before age eighteen and the overall pace of second births, but a strong positive association between an adolescent first birth and the pace of second births in the 15 months postpartum, and a negative association at longer durations. For black and white women, an early age at first birth is associated with a faster pace of third births, an effect that appears to persist for fourth births. However, a closely spaced second birth is at least as strongly associated with a fast pace of third and later births as is an adolescent first birth. These results suggest substantial spacing effects on second and higher order births that, in several instances, outweigh effects of a teen first birth. We discuss implications of these findings for policies concerning teen childbearing and argue for special attention to pregnancy prevention in the six months immedia tely following a first birth.

Methods: Data are from 301 mothers in the National Longitudinal Survey of Youth 1979 who gave birth prior to age 19. We estimated path models to assess women's incomes, partner characteristics, and health behaviors at age 40 as mediators of the relationship between their educational attainment and self-rated general health at age 50.

Results: After accounting for observed background factors that select women into early childbearing and lower educational attainment, higher levels of education (high school diploma and GED attainment vs. no degree) were indirectly associated with higher self-rated health at age 50 via higher participant income at age 40.

Copyright of Journal of Health & Social Behavior is the property of American Sociological Association and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts)

Copyright of Law & Society Review is the property of Blackwell Publishing Limited and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts)

Three main findings emerge from the analysis. First, incarceration is a powerful predictor of mid-life mental and physical health functioning. Second, these effects are evident in models that account for the non-random nature of incarceration. That is, even when models are employed to correct for sample heterogeneity, the significant physical and mental health effects remain. Finally, the results indicate that incarceration contributes to racial inequalities in physical health. When incarceration is included in models estimating physical health functioning, racial differences in health become non-significant.

A variety of factors come together to explain these findings. First, incarceration fractures social bonds associated with health functioning. Additionally, incarceration lowers the likelihood of gainful wages and depresses wages. Furthermore, incarceration exposes individuals to high levels of infectious diseases and stress. Finally, most ex-inmates find themselves near the bottom of the social hierarchy, making it difficult for them to fully control their lives and participate in society upon release.

These factors, with two additional considerations, also explain how incarceration contributes to racial inequalities in physical health. First, relative to whites, blacks are much more likely to be incarcerated and, therefore, more likely to be exposed to the negative health effects of the penal system. Second, it appears more difficult for blacks than whites to reestablish social bonds to social institutions that protect or enhance physical health.

Many families are struggling to provide a supportive environment on limited resources. Some work full-time or part-time, some are unemployed, others are homeless, others are receiving Temporary Assistance for Needy Families (TANF) and other forms of public assistance. These families, working or not, face many obstacles in their struggle to become self-sufficient (Beaulieu, 1999). During the 1990s revisions to the United States public assistance programs began moving families from welfare to work. However, work does not always lift or keep a family out of poverty. If family, individual, and employment characteristics that distinguish the working poor from the working non-poor can be identified, better public policy and outreach programs can be designed. This panel discussion focused on characteristics of a sample of working poor and working near-poor and the availability of fringe benefits from their employers, ideas regarding transitional assistance and strategies for reaching the working poor population, and finally provides resources related to the Workforce Investment Act of 1998.

How do changes in young mothers' marital status relate to the likelihood of exiting from and reentering into poverty? Does marrying get them out of poverty as strongly as being unmarried puts them back into poverty? Using two-way transition models from event history analysis and a sample of young mothers who experienced poverty from the National Longitudinal Survey of Youth 1979 data, we will address these questions.

We expect to find that getting married in a year is associated with the increased likelihood of exiting poverty in the year and with decreased likelihood of going back into poverty. Being unmarried is hypothesized to be associated with an increased likelihood of reentering poverty and decreased likelihood of exiting from poverty when other conditions are kept equal. We also expect that the economic benefit of marrying for young mothers is smaller than the negative economic consequences from being unmarried.

The report opens with an examination of theoretical perspectives that hypothesise why parental income might affect children's outcomes. It discusses a range of methodological issues that confront researchers in this field. It documents the findings of a range of research on the effect of parental income on six broad areas of child outcomes: cognitive test scores; socio-emotional functioning, mental health and behavioural problems; physical health; teenage childbearing; educational attainment; and future economic status. It considers whether the source of parental income matters for child outcomes, whether the effect of parental income might vary according to the age of the child, and whether the effect of parental income depends on the child's gender or race. The report concludes with a discussion of policy insights that might be gleaned from the research literature in this field.

Parental income is positively associated with all outcomes covered in the review. When family background variables are controlled, however, the estimated size of the effect of parental income reduces, and the residual effects are generally small to modest on most outcomes. The size of the effect of income differs across different outcomes: it appears to have its largest effect on cognitive test scores and educational attainment. For some outcomes, such as health, there is too little research to draw strong conclusions about the effect of income. The effect of income is larger when incomes are measured over a longer period - that is to say, extended durations on low income have stronger adverse effects on children than short periods on low income. There is some evidence that the effect of income is larger for low-income than for high-income children.

No general conclusions can be drawn about whether parental income is more important at different stages of childhood; however, there is some evidence to suggest that income is more important in early childhood for schooling outcomes. There is little evidence to suggest that income has differential effects on children of different gender or race. Welfare income is found to be negatively associated with a range of children's outcomes; however, this seems to be due not to welfare receipt per se but to parental characteristics that make some parents more prone to be on welfare than others. Finally, it is noted that most of the research has been done in the US and there is insufficient evidence to draw any conclusions about whether the effect of parental income varies across countries.

Although parental income generally has only a small to modest effect on any particular outcome, it contributes to many aspects of children's well-being. This means that income gains have the potential to make a significant cumulative difference to the lives of children.

Specifically, the authors find that mothers' own characteristics, measured when she herself was an adolescent, can predict her future income and education, and the latter, in turn, predict her daughter's adolescent characteristics, which presumably predict the daughter's future income and education. These findings are important for research and policy on several levels. In short, the authors argue that the importance of socioeconomic status will be overstated if researchers omit a mother's own adolescent characteristics in their measurement models.

The author finds that about half of earnings inequality in the U.S. can be explained by family and community influences during childhood. He also finds that these influences have become more important in recent decades. Copyright: Federal Reserve Bank of Chicago: 2004

Methods. We used the National Longitudinal Survey of Youth (NLSY), a nationally representative longitudinal data set. We used harmful and binge drinking at ages 17 to 20 to predict harmful and binge drinking at ages 30 to 31, stratifying for gender and controlling for confounders.

Results. Data were available on harmful drinking at both the adolescent and the early adult age period for 3790 individuals and on binge drinking for 2387 individuals. Harmful drinking during adolescence was significantly associated with harmful drinking at ages 30 to 31 for men only. Among male adolescents, 14% of harmful drinkers continued harmful drinking at ages 30 to 31, compared with 4% of nonharmful drinkers who became harmful drinkers. In Poisson regression models, binge drinking during adolescence was associated with binge drinking at ages 30 to 31 for both men and women, generating relative risks of 2.3 (95% confidence interval: 1.8--3.0) and 3.0 (95% confidence interval: 2.4--4.8), respectively. Half of binge-drinking male adolescents and one third of binge-drinking female adolescents engaged in binge drinking into early adulthood, compared with 19% for non--binge-drinking male adolescents and 8% of non--binge-drinking female adolescents.

Conclusions. Problem drinking during adolescence is associated with problem drinking in early adulthood. Efforts to prevent and treat adolescent problem drinking could have an impact on the progression of alcohol-related disease. [ABSTRACT FROM AUTHOR]

In the second phase of this research, I link parent data to their young adult children to assess the impact parental indebtedness has on their children’s social mobility. Because consumer debt can be used as a mechanism to both invest in children and to drain fiscal resources that could be invested in children, I initially examine the impact of parental indebtedness on financial investment in their young adult children. I find that young adult children whose parents hold significant amounts of consumer debt in later life (groups 2, 6, and 7) are much more likely to receive financial support from their parents. These parents are more likely to support their young adult children by paying their children’s bills, financing their education, and letting them live at home than are children whose parents consumer carry little debt when their children become young adults. My findings suggesting that some parents likely incur consumer debt as a means of investing in their young adult children. I next examine the impact parental indebtedness has on their young adult children’s financial behaviors. I find that young adults whose parents have consistently low debt (group 1) are unlikely to become consumer debtors themselves during young adulthood and that young adults whose parents have consistently high consumer debt (group 7) are very likely to hold sizeable consumer debt balances early in life. I also find, however, that young adults with parents who, in their own young adulthood, followed a trajectory compatible with the life course hypothesis (groups 2, 3, 4, 5, and 6) have indistinguishable debt patterns. Because important distinctions of consumer behavior did not emerge until later ages among their parents’ cohort, these findings still support my hypothesis that carrying debt is a somewhat socialized behavior. My dissertation finally examines how parental indebtedness impacts educational outcomes. Preliminary findings suggest that young adults whose parents hold significant amounts of debt in later life (groups 2, 6, and 7) are more likely to attend college than children of parents who have consistently low consumer debt (group 1), but that only children of parents who repay their own debts mid-life and re-emerge as consumer debtors (group 6) matriculate faster than the children of low debtors (group 1) when enrolled.

I analyze data from the 1979 cohort of the National Longitudinal Survey of Youth to examine the patterns and payoffs to non-searching across the work career. I identify three main types of non-searching—entry-level, reentry-level, and elite—among employed workers both early in their work careers (age 17-25) and in the early prime of their careers (age 33-41). Entry-level non-searching occurs early in the career among people who are often students that are accessing their first few jobs. Reentry-level non-searching occurs later in the career primarily among women transitioning back into the workforce after taking time off to take care of family responsibilities. Elite non-searching occurs mainly among men in the prime of their careers who have the most work experience. Non-searchers vary substantially in their personal and job characteristics, with the bulk of the benefits to non-searching going to elite non-searchers. These elite non-searchers are recruited from some of the best jobs around and are more likely than active searchers to receive high wage managerial jobs with a great deal of authority and prestige. These findings re-assert the importance of social capital in determining labor market outcomes, suggesting that future research needs to account for non-searchers in order to assess the full extent of informal job matching and embeddedness in the labor market.

The second essay models how screening workers through social networks impacts labor mobility in markets with adverse selection. When there is asymmetric information in labor markets, worker mobility is constrained by adverse election in the market for experienced workers. However, if workers can acquire references through their social networks then they can move more easily between jobs. In this chapter I develop a simple labor market model in which workers can learn the productivity of other workers through social interaction. I show that networks increase wages and mobility of high-productivity experienced workers; however, networks discourage workers from accepting jobs outside their job-contact network, because of adverse selection.

The third essay in this dissertation examines the importance of social networks in labor markets when work is produced jointly. Most employers cite engage in work together. In this essay, I explain why it might be rational for firms to hire through social networks even when worker skill is observed perfectly, if these workers are better able to do joint work with the firm

Chapter 1 considers how locus of control--the degree to which one believes one's actions influence outcomes--affects unemployed job search. Assuming that locus of control is a determinant of beliefs about the efficacy of search effort, the model predicts that "internal" individuals (who believe their actions determine outcomes) search more intensively and set higher reservation wages than their "external" counterparts (who believe their actions have little effect on outcomes). Using the 1979 National Longitudinal Survey of Youth, I find that, consistent with these predictions, "internal" job seekers search more intensively and set higher reservation wages than do their more "external" peers, but are no better at converting search effort into job offers and earn no more than their peers upon finding employment conditional on reservation wages. The findings also indicate that very "internal" individuals hold out for excessively high wages while very "external" individuals search too little. As a result, both groups spend more time unemployed than individuals with average loci of control.

Chapter 2 tests the hypothesis that locus of control affects search behavior by influencing beliefs about the efficacy of search effort in a laboratory experiment in which subjects exert effort to generate offers. There are two experimental treatments: a limited information treatment in which subjects exert effort without knowledge of how their effort influences the generation of offers and a f ull information treatment in which subjects are informed of this relationship. I find that in the limited information treatment more "internal" subjects exert more effort and hold out for higher offers than more "external" subjects, but there is no such relationship in the full information treatment when uncertainty about the connection between effort and outcomes does not exist. In both treatments, however, I find that "externality" is positively related to the probability that an individual "quits" searching, suggesting that locus of control may play a key role in explaining "discouragement" among searchers

Chapter 3 examines how learning disabled youth fare in high school relative to observationally equivalent peers in terms of cognitive and noncognitive skills. Learning disabled youth in my sample are six percentage points more likely to graduate from high school than peers with the same measured cognitive ability. This difference cannot be explained by differences in noncognitive skills, families, or school resources. Instead, I find that learning disabled students graduate from high school at higher rates because of high school graduation policies making it easier for learning disabled youth to obtain a high school diploma. The effects of these graduation policies are even more remarkable given that I find evidence after high school that learning disabled youth have less unmeasured human capital. [NOTE: This chapter is based on the Children of the NLSY79 and the NLSY79 Young Adult.]

The current study examines the impact of maternal cohabitation and remarriage on children's father/father-figure relationships, using longitudinal data from the National Longitudinal Survey of Youth, 1979 Child cohort. Data will be taken from the 1998 survey year. Children of never-married, divorced, and separated mothers are included. Dependent variables differ by child's age, with greater use of child-report data for older children. The relationships between children and fathers under age ten will be assessed by mother report of child-father closeness and child involvement with father-figures. For children ages ten to fourteen, measures of father involvement, time with father, father closeness, stepfather involvement, time with stepfather, and stepfather closeness will be constructed directly from child reports. For children ages fifteen and up, dependent variables will be child reports of frequency of visitation and phone conversations with the nonresident father. The payment of child support will be examined for children of all ages using maternal reports. It is hypothesized that maternal cohabitation is associated with child relationships with the biological father both directly, and through a spurious selection effect. There is some evidence that nonresident fathers feel displaced when another potential father-figure enters the child's life. This effect is not expected, however, to be as strong for maternal cohabitation as it will be for maternal remarriage, as the child's relationship with the new father-figure is more tenuous in cohabitations than remarriages. Selection effects are also expected to be very different between maternal cohabitation and maternal remarriage. Cohabiting mothers are expected to be of a lower socioeconomic status on average than remarried mothers, and previous evidence indicates that maternal education is positively associated with father-child contact in disrupted families. Mother SES is also likely to be correlated with father SES, and father SES is also positively associated with father-child contact and payment of child support. In part because of the socioeconomic effect, it is also expected that children of never-married parents will be more heavily represented in cohabiting than in remarried households, and fathers are less likely to see and contribute to children if the children were not born in marriage....The study uses ordinary least-squares and logistic regression to examine these hypotheses, with special attention paid to separating spurious effects based on selection from direct effects caused by maternal marital/cohabitation status.

Methodology/approach: The chapter consists of descriptive analysis of data from the National Longitudinal Survey of Youth’s 1979 cohort. Respondents were followed from 1979 to 2006.

Findings: The economic consequences of single motherhood are persistent. Women who have once been divorced or never-married mothers remain poorer through middle age, no matter how their family structure subsequently changes.

Social implications: A critical feature of the modern economic and demographic landscape is the intersection of individual and family characteristics. Many divorced and, especially, never-married mothers experience profound disadvantage even before they become mothers.

These changes have created tremendous uncertainty in children's lives and have led to considerable speculation among policy makers and the public more generally about the consequences of father absence. Some analysts argue that growing up with a single mother is the primary cause of many of the country's most serious social problems, including poverty, high school dropout, teen pregnancy, and delinquency (Popenoe, 1988, 1996; Whitehead, 1993; Blankenhorn, 1995). Others argue that poverty and economic insecurity are the real culprits, causing both father absence and adolescent behavioral problems (Skolnick, 1991; Stacy, 1993). Still others claim that the problems associated with family disruption are rooted in marital discord that begins long before the parents separate or divorce.

To bring empirical evidence to bear on this debate, my colleagues and I have been analyzing several large, nationally representative surveys that contain information on children's family structure growing up as well as their educational attainment and social adjustment in young adulthood. In this chapter, I summarize the major findings from this work as it pertains to the following questions:

Are children raised apart from their biological fathers worse off than children raised by both parents?
How large are the differences and which groups of children are most affected?
What factors account for the lower achievement of children in one-parent families? What factors are associated with resilience?
And finally, Do differences in children's wellbeing predate family disruption or are they are a consequence of father absence?

Our investigation has been going on for over 10 years now and covers more than 10 data sets. The most important of these are the Panel Study of Income Dynamics (PSID), the National Longitudinal Survey of Youth (NLSY), the High School and Beyond Study (HSB), and the National Survey of Families and Households (NSFH). All of these surveys are large enough to allow us to distinguish among different types of single parent families, including families headed by never-married mothers as well as families headed by divorced or separated mothers and remarried mothers. These surveys also allow us to compare differences between boys and girls raised in one- and two-parent families as well as differences between children of different racial and ethnic backgrounds and different social classes.

To summarize briefly, we find that children who grow up apart from their biological fathers do less well, on average, than children who grow up with both natural parents. They are less likely to finish high school and attend college, less likely to find and keep a steady job, and more likely to become teen mothers. The differences are not huge. Indeed, most children who grow up with a single parent do quite well. Nor are they large enough to support the claim that father absence is the major cause of our country's most serious social problems. However, the differences between children in one- and two-parent families are not so small as to be inconsequential, and there is fairly good evidence that father absence per se is responsible for at least some of them.

NLSY 79 data is used to establish that about half of unmarried parents who are cohabiting at birth are still living together after six years.