Showing papers in "Journal of Human Resources in 2005"
TL;DR: This paper examined the validity of these instrumental variables and found that none of the candidate instruments is a useful source of identification in currently available data sets, and investigated the role of exclusion restrictions versus nonlinearity as the source for identification in bivariate probit models.
Abstract: Several previous studies have relied on religious affiliation and the proximity to Catholic schools as exogenous sources of variation for identifying the effect of Catholic schooling on a wide variety of outcomes. Using three separate approaches, we examine the validity of these instrumental variables. We find that none of the candidate instruments is a useful source of identification in currently available data sets. We also investigate the role of exclusion restrictions versus nonlinearity as the source of identification in bivariate probit models. The analysis may be useful as a template for the assessment of instrumental variables strategies in other applications.
421 citations
TL;DR: The authors analyzed the effects of Oklahoma's universal pre-kindergarten (pre-K) program for four-year-olds on children in Tulsa Public Schools (TPS) and found that the Tulsa pre-K program increases cognitive/knowledge scores by approximately 0.39 standard deviation, motor skills scores by roughly 0.24 standard deviation and language scores by about 0.38 standard deviation.
Abstract: Since the mid-1990s, three states, including Oklahoma, have established a universal pre-kindergarten (pre-K) program. We analyze the effects of Oklahoma’s universal pre-kindergarten (pre-K) program for four-year-olds on children in Tulsa Public Schools (TPS). The main difficulty with testing the causal impact of a voluntary pre-K program is that certain parents are more likely to select pre-K, and these parents might have other unobservable characteristics that influence the test outcomes of their children. Because TPS administered an identical test in September 2001 to children just beginning pre-K and children just beginning kindergarten, we can compare test outcomes of “old” pre-kindergarten students to test outcomes of “young” kindergarten students who attended pre-K the previous year. We find that the Tulsa pre-K program increases cognitive/knowledge scores by approximately 0.39 standard deviation, motor skills scores by approximately 0.24 standard deviation, and language scores by approximately 0.38 standard deviation. Impacts tend to be largest for Hispanics, followed by blacks, with little impact for whites. Children who qualify for a free lunch have larger impacts than other children.
298 citations
TL;DR: The authors used a natural experiment associated with human reproduction to identify the causal effect of teen childbearing on the socioeconomic attainment of teen mothers and found that the adverse consequences of early childbearing are short-lived.
Abstract: We exploit a "natural experiment" associated with human reproduction to identify the causal effect of teen childbearing on the socioeconomic attainment of teen mothers. We exploit the fact that some women who become pregnant experience a miscarriage and do not have a live birth. Using miscarriages an instrumental variable, we estimate the effect of teen mothers not delaying their childbearing on their subsequent attainment. We find that many of the negative consequences of teenage childbearing are much smaller than those found in previous studies. For most outcomes, the adverse consequences of early childbearing are short-lived. Finally, for annual hours of work and earnings, we find that a teen mother would have lower levels of each at older ages if they had delayed their childbearing.
276 citations
TL;DR: Using individual-level data on males from the 1988–91 National Health Interview Survey Multiple Cause of Death Files, the impact of relative deprivation within a reference group on health is examined and high relative deprivation in the sense of Yitzhaki is associated with a higher probability of death.
Abstract: While a large body of evidence relates low absolute income to premature mortality, a recent and growing literature argues that relative income influences health as well. Low relative income, or being deprived relative to one’s reference group, may cause stress and depression. These conditions are linked to mortality both directly (via heart disease, high blood pressure, and suicide) and indirectly (via increased smoking, poor eating habits, and alcohol abuse). Evidence from biology supports the notion that relative status may influence health outcomes. In this paper, we use restricted-use micro-level data from the National Health Interview Survey (NHIS) Multiple Cause of Death Files (MCOD) from 1988 to 1991 to examine whether relative deprivation increases the probability of dying. We define reference groups using a combination of characteristics including state, race, education, and age, and measure relative deprivation with Yitzhaki’s index. Our use of individual-level data allows us to for control for characteristics that are specific to reference group. Results indicate that high relative deprivation increases the probability of dying in all age groups and for those death categories with a high behavioral component. Those with high relative deprivation are more likely to self-report poor health, have high blood pressure or disabilities, and have a host of poor health habits including smoking, not wearing safety belts, high body mass index and not exercising. For nearly all health measures, our results suggest that much of the observed statistical relationship between absolute level of income and health found in previous work is actually measuring the impacts of relative deprivation on health.
269 citations
TL;DR: In this paper, the effects of income on health and mortality were studied using only the part of income variation due to a truly exogenous factor: monetary lottery prizes of individuals, and the results showed that higher income causally generates good health and that this effect is of a similar magnitude as when traditional estimation techniques are used.
Abstract: A vast literature has established a strong positive relation between income and health status and a negative relation with mortality. This paper studies the effects of income on health and mortality, using only the part of income variation due to a truly exogenous factor: monetary lottery prizes of individuals. The findings are that higher income causally generates good health and that this effect is of a similar magnitude as when traditional estimation techniques are used. A 10-percent income increase improves health by about 4-5 percent of a standard deviation.
265 citations
TL;DR: The authors investigated the relationship between improvements in per capita expenditure and child labor with a panel data set that spans this episode of growth in Vietnam and found that 80 percent of the decline in child labor that occurs in households whose expenditures improve enough to move out of poverty.
Abstract: Between 1993 and 1997, child labor in Vietnam declined by nearly 30 percent while the country’s GDP grew by nearly 9 percent per year on average. Using a simple, nonparametric decomposition, I investigate the relationship between improvements in per capita expenditure and child labor with a panel data set that spans this episode of growth in Vietnam. Improvements in per capita expenditure can explain 80 percent of the decline in child labor that occurs in households whose expenditures improve enough to move out of poverty. This finding suggests a previously undocumented role for economic growth in the amelioration of child labor.
244 citations
TL;DR: In this paper, the authors investigated whether the increased usage of nontenure-track faculty adversely affects undergraduate students' graduation rates using institutional level panel data from the College Board and other sources.
Abstract: During the last two decades, there has been a significant growth in the share of faculty members at American colleges and universities that are employed in part-time or full-time nontenure-track positions Our study is the first to address whether the increased usage of such faculty adversely affects undergraduate students’ graduation rates Using institutional level panel data from the College Board and other sources, our econometric analyses suggest that the increased usage of these faculty types does adversely affect graduation rates at four-year colleges, with the largest impact on students being felt at the public master’s level institutions
238 citations
TL;DR: This paper showed that families with children increased expenditures on women's clothing (relative to men's) after implementation of a policy that shifted a child subsidy payment from the father to the mother, which was interpreted as evidence that families do not pool their income but allocate consumption based on income source.
Abstract: This paper replicates results of an article showing that families with children increased expenditures on women’s clothing (relative to men’s) after implementation of a policy that shifted a child subsidy “payment” from the father to the mother. These results were interpreted as evidence that families do not pool their income but allocate consumption based on income source. However, the current paper also finds an increase in relative spending on women’s clothing among childless couples, a sample the policy change did not impact. Alternative explanations are explored for observing these patterns, but none can rule out either bargaining or income pooling.
196 citations
TL;DR: The authors investigated possible reasons for this effect, using data from the National Longitudinal Survey of Youth 1979-1992 and found that education appears to have a double effect on the earnings of women.
Abstract: Many studies have found that the impact of schooling on earnings is greater for females than for males, despite the fact that females tend to earn less, both absolutely and controlling for personal characteristics. This study investigates possible reasons for this effect, using data from the National Longitudinal Survey of Youth 1979–. One explanation is that education appears to have a double effect on the earnings of women. It increases their skills and productivity, as it does with men, and in addition it appears to reduce the gap in male and female earnings attributable to factors such as discrimination, tastes, and circumstances. The latter appear to account for about half of the differential in the returns to schooling.
189 citations
TL;DR: In this paper, the authors presented maximum simulated likelihood estimates of a system of limited dependent variables governing smoking and drinking patterns and income, and found that moderate drinking is associated with 10 percent higher income, while heavy drinking associated with 12 percent more income than drinking abstention.
Abstract: In an effort to increase understanding of the “alcohol/income puzzle”—the finding that drinking appears to lead to higher income—this paper presents maximum simulated likelihood estimates of a system of limited dependent variables governing smoking and drinking patterns and income. With all else held constant, moderate drinking is associated with 10 percent higher income, and heavy drinking associated with 12 percent higher income, than drinking abstention. Smoking is associated with larger effects on income than drinking: Single equation estimates suggest smokers earn 8 percent less than nonsmokers, and the smoking penalty rises to 24 percent after correcting for endogeneity.
186 citations
TL;DR: This article found that the effect of parental income on sons' family income and wages at age thirty has declined over this period, largely due to the increase in government investment in children, especially in their educational attainment.
Abstract: We use data from the PSID to assess whether the effect of parental income on son’s economic status has changed for cohorts born between 1949 and 1962. We find that the effect of parental income on sons’ family income and wages at age thirty has declined over this period. This is largely because the effect of parental income on son’s years of schooling has declined. The decline in the effect of parental income is not part of an overall decline in the effect of family background. We provide suggestive evidence that the decline is due to the increase in government investment in children, especially in their educational attainment.
TL;DR: This paper examined the link between income and living arrangements and found no association between pension income and elderly independence, but they did find that potential beneficiaries alter their household structure when they become pension-eligible.
Abstract: This study examines the link between income and living arrangements. Our identification comes from exploiting a discontinuity in age in the benefit formula for the social pension in South Africa. In contrast to previous literature, we find no association between pension income and elderly independence. We do find that potential beneficiaries alter their household structure when they become pension-eligible. Prime working-age women depart, and the presence of children younger than five and young women of childbearing age increase. These shifts in coresidence patterns are consistent with a setting where prime-age women have comparative advantage in work away from extended family relative to younger women. The additional income from old-age support may induce a change in living arrangements to exploit this advantage.
TL;DR: It is found that smoking during pregnancy doubles the chance an infant is born with a low birth weight, similar to single-equation estimates where maternal smoking is treated as exogenous.
Abstract: Using a statistical matching procedure to choose control groups, we find that four states that adopted large cigarette tax hikes had corresponding decreases in smoking participation of pregnant women. Using the tax hike as an instrument for smoking in birth-weight equations and pooling data across experiments, we find that smoking during pregnancy doubles the chance an infant is born with a low birth weight. Our estimates are similar to singleequation estimates where maternal smoking is treated as exogenous.
TL;DR: The authors found that the absence of age-related incentives in defined contribution plans leads workers to retire almost two years later on average compared to workers with defined benefit plans, and the evolution of pension structure can help explain recent increases in the typical retirement age, after decades of decline.
Abstract: Defined benefit pension plans have become considerably less common since the early 1980s, while defined contribution plans have spread. Previous research showed that defined benefit plans, with sharp incentives encouraging retirement after a certain point, contributed to the striking decline in American retirement ages. In this paper we find that the absence of agerelated incentives in defined contribution plans leads workers to retire almost two years later on average, compared to workers with defined benefit plans. Thus, the evolution of pension structure can help explain recent increases in the typical retirement age, after decades of decline.
TL;DR: In this article, the authors estimate the effects of HOPE on course-taking, treating nonresidents as a control group, and find that HOPE decreased full-load enrollments and increased course withdrawals among resident freshmen.
Abstract: A common justification for state-sponsored merit scholarships like Georgia’s HOPE program is to promote academic achievement. However, grade-based retention rules encourage other behavioral responses. Using longitudinal records of enrolled undergraduates at the University of Georgia between 1989 and 1997, we estimate the effects of HOPE on course-taking, treating nonresidents as a control group. First, we find that HOPE decreased full-load enrollments and increased course withdrawals among resident freshmen. Second, the scholarship’s influence on course-taking behavior is concentrated on students whose predicted freshmen GPAs place them on or below the scholarship-retention margin. Third, HOPE substantially increased summer school credits.
TL;DR: The authors explore the extent to which the large race gap in wealth can be explained with properly constructed income and demographic variables, and find that differences in savings behavior and/or rates of return play an important role.
Abstract: We explore the extent to which the large race gap in wealth can be explained with properly constructed income and demographic variables. In some instances we explain the entire wealth gap with income and demographics, provided that we estimate the wealth model on a sample of whites. However, we typically explain a much smaller fraction when we estimate the wealth model on a black sample. Using sibling fixed-effects models to control for intergenerational transfers and the effects of adverse history, we find that these factors are not likely to account for the lower explanatory power of the black wealth models. Our analysis of growth models of wealth suggests that differences in savings behavior and/or rates of return play an important role.
TL;DR: In this article, the authors examined the effects of minimum wages on the distribution of family income relative to needs in states and years with and without minimum wage increases, and found that their net effect is to increase the proportions of families with incomes below or near the poverty line.
Abstract: The primary goal of a national minimum wage floor is to raise the incomes of poor families with members in the work force. We present evidence on the effects of minimum wages on family incomes from March CPS surveys. Using non-parametric estimates of the distributions of family income relative to needs in states and years with and without minimum wage increases, we examine the effects of minimum wages on this distribution, and on the distribution of the changes in income that families experience. Although minimum wages do increase the incomes of some poor families, the evidence indicates that their net effect is, if anything, to increase the proportions of families with incomes below or near the poverty line. Thus, it would appear that reductions in the proportions of families that are poor or near-poor should not be counted among the potential benefits of minimum wages.
TL;DR: This article used a new methodology to assess the effects of court-ordered desegregation plans on segregation and white enrollment and found that white flight was particularly severe in districts with more public school districts in the same metropolitan area.
Abstract: This paper uses a new methodology to assess the effects of court-ordered desegregation plans on segregation and white enrollment. I then assess what characteristics of districts are predictive of having more or less white flight when desegregation plans are implemented. I exploit the wide variation in the timing of implementation of desegregation plans to identify their effects. I find strong evidence that segregation fell when districts implemented desegregation plans; plans were also associated with significant white enrollment losses that offset about one-third of the within-district reductions in segregation. White flight was particularly severe in districts with more public school districts in the same metropolitan area.
TL;DR: This article found that women who have experienced marital dissolution have considerably higher levels of personal income and annual wages than women who remain married, and that the higher wages of ever-divorced women mostly reflect increased labor supply intensity (hours and weeks of work) of woman who experienced divorce.
Abstract: The rise in the divorce rate over the past 40 years is one of the fundamental changes in American society. A seemingly ever-increasing number of women and children spend some fraction of their life in single female-headed households, leading many to be concerned about the economic circumstances of these women their and children. Estimating the cause-to-effect relationship between marital dissolution and female economic status is complicated because the same factors that increase marital instability may also affect the economic status and labor market behavior of women. We propose an instrumental variables solution to this problem based on the sex of the firstborn child. This strategy exploits the fact that the sex of the firstborn child is random and the fact that marriages are less likely to survive following the birth of girls as opposed to boys. Our IV estimates cast doubt on the contention that marital instability causes large declines in woman’s economic status. Once the negative selection into divorce is accounted for, we find that women who have experienced marital dissolution have considerably higher levels of personal income and annual wages than women who remain married. At the same time we find little evidence of differential poverty rates and equivalized household incomes among ever-divorced women and never-divorced women. We further show that the higher wages of ever-divorced women mostly reflect increased labor supply intensity (hours and weeks of work) of woman who experienced marital dissolution.
TL;DR: In this article, the causal effect of Social Security on living arrangements with an instrumental-variable approach that relies on the large shifts in benefits for cohorts born from 1910-21, the socalled Social Security notch.
Abstract: Previous studies of the effect of Social Security on elderly living arrangements generally have relied on data from the distant past or differences in benefits across families or cohorts that potentially were correlated with other determinants of living arrangements. Using data from the 1980–99 Current Population Surveys, we attempt to isolate the causal effect of Social Security on living arrangements with an instrumental-variable approach that relies on the large shifts in benefits for cohorts born from 1910–21, the socalled Social Security notch. Over all elderly households, the estimated elasticity of living with others with respect to Social Security income is -0.4, with elasticities of -1.3 and -1.4 for the widowed and divorced, respectively; most of the effects on living arrangements appear to be concentrated among the lesser educated as well. Our estimated elasticities are substantially larger than those from previous studies and suggest that reductions in current benefits would alter living arrangements significantly.
TL;DR: This paper investigated the relationship between welfare reform and health insurance, health care utilization, and self-reported measures of health status for women aged 20-45, using nationally representative data from the Behavioral Risk Factor Surveillance System.
Abstract: We investigate the relationship between welfare reform and health insurance, health care utilization,
and self-reported measures of health status for women aged 20–45, using nationally representative data
from the Behavioral Risk Factor Surveillance System.
TL;DR: In this paper, a method for estimating the mean impact of an assigned social program when it is not feasible to do a pre-intervention baseline survey but it is feasible to track ex-participants is proposed.
Abstract: We propose a method for estimating the mean impact of an assigned social program when it is not feasible to do a pre-intervention baseline survey but it is feasible to track ex-participants. In our triple-difference estimator; measured outcome changes are compared between continuing participants and matched ex-participants, after netting out the outcome changes for a matched comparison group who never participated. With sufficient followup observations one can test the joint conditions required for correctly identifying the gains to current participants. We apply the method to a workfare program in Argentina. Significant impacts on participants' current incomes
TL;DR: Although the estimates are less accurate than previously thought, most specifications by Angrist and Krueger (1991) are informative for returns-to-schooling and concern about the reliability of the model with 178 instruments is unfounded despite the low first-stage F-statistic.
Abstract: We evaluate Angrist and Krueger (1991) and Bound, Jaeger, and Baker (1995) by constructing reliable confidence regions around the 2SLS and LIML estimators for returns-to-schooling regardless of the quality of the instruments. The results indicate that the returns-to-schooling were between 8 and 25 percent in 1970 and between 4 and 14 percent in 1980. Although the estimates are less accurate than previously thought, most specifications by Angrist and Krueger (1991) are informative for returns-to-schooling. In particular, concern about the reliability of the model with 178 instruments is unfounded despite the low first-stage F-statistic. Finally, we briefly discuss bias-adjustment of estimators and pretesting procedures as solutions to the weak-instrument problem.
TL;DR: The authors analyzed how much of the gender earnings gap among physicians is due to womens greater family responsibilities, finding that women physicians earn 11 percent less for being married plus 14 percent more for having one child and 22 percent fewer for having more than one child.
Abstract: This paper analyzes how much of the gender earnings gap among physicians is due to womens greater family responsibilities. Women physicians earn 11 percent less for being married plus 14 percent less for having one child and 22 percent less for having more than one child. Before marrying/having children, women physicians who later became wives or mothers had higher earnings than those who remained single and childless, but sharply reduced their hours of work after marrying/having children. The results suggest that these earnings gaps do not reflect adverse selection but rather individual choices given time constraints imposed by family responsibilities.
TL;DR: In this article, a high-stakes television game show was used to determine whether wage differentials arise because of discrimination, and distinguishing between different theories of discrimination is extremely difficult and distinguishing different theories is harder still, and the results suggest no evidence of discriminatory voting patterns by males against females or by whites against blacks.
Abstract: Empirically determining whether wage differentials arise because of discrimination is extremely difficult, and distinguishing between different theories of discrimination is harder still This paper exploits a number of unique features of a high-stakes television game show to determine which contestants discriminate and why In the show, contestants take turns answering a series of trivia questions, and, at the end of each round of questions, one contestant is voted off by the other players in the round Our results suggest no evidence of discriminatory voting patterns by males against females or by whites against blacks However, somewhat surprisingly, we find that in the early rounds of the game women appear to discriminate against men We test three competing theories for the voting behavior of women: preference-based discrimination, statistical discrimination and strategic discrimination In doing so, we highlight the types of experimental designs that could be used to distinguish between these theories Only preference-based discrimination is consistent with the voting patterns
TL;DR: In this paper, the authors test for welfare-induced migration by comparing AFDC participation in border counties to interior counties in the same state, and they find that having a neighbor with benefits that are $100 lower increases AFDC expenditures in the border counties by 4.0-6.8 percent relative to the interior counties.
Abstract: I test for welfare-induced migration by comparing AFDC participation in border counties to interior counties in the same state. If migration costs are lower for border county residents, border counties on the high-benefit side of a state border should have higher welfare participation relative to the state's interior counties. Border counties on the low-benefit side should have lower welfare participation relative to the state's interior counties. The results obtained using county-level data from 1970-90 indicate that having a neighbor with benefits that are $100 lower increases AFDC expenditures in border counties by 4.0-6.8 percent relative to interior counties.
TL;DR: Estimating Long-Term Consequences of Teenage Childbearing - An Examination of the Siblings Approach as mentioned in this paper, presented in the Journal of Geriatrics and Obstetrics.
Abstract: Estimating Long-Term Consequences of Teenage Childbearing - An Examination of the Siblings Approach
TL;DR: The authors found a decline in the importance of ability in explaining educational performance, in part because low ability children with high economic status experienced the largest increases in educational attainment, while cognitive ability was not correlated with educational achievement.
Abstract: Most countries seek to reduce inequality by encouraging educational attain- ment, particularly by striving for better outcomes for able individuals from poor backgrounds. We analyse whether this has been a feature of Britain's substantial expansion of education during the past several decades. We use two unique longitudinal studies to test whether these improvements have been associated with changes in the role of cognitive ability and parental background in determining educational achievement. We find a decline in the importance of ability in explaining educational performance, in part because low ability children with high economic status experienced the largest increases in educational attainment.
TL;DR: It is found that moving from a junior high school system to a middle school system decreases on-time high school completion by approximately 1–3 percent.
Abstract: While nearly half of all school districts have adopted middle schools, there is little quantitative evidence of the efficacy of this educational structure. We estimate the impact of moving from a junior high school system, where students stay in elementary school longer, to a middle school system for on-time high school completion. This is a particularly good outcome measure because middle school advocates argued that this new system would be especially helpful for lower achieving students. In contrast to the stated objective, we find that moving to a middle school system decreases on-time high school completion by approximately 1–3 percent.
TL;DR: This paper used financial aid records of individual students at 28 highly selective private colleges and universities to calculate both the price the low-income students at these schools actually pay for a year's education, net of financial aid grants, and how the schools differentiate net price in recognition of their students' different family incomes.
Abstract: Working from the financial aid records of individual students at 28 highly selective private colleges and universities, we were able to calculate both the price the low-income students at these schools actually pay for a year’s education, net of financial aid grants, and how the schools differentiate net price in recognition of their students’ different family incomes - their pricing policies.