Search results

This paper uses data from the 1979 and 1997 National Longitudinal Surveys of Youth to estimate the changing returns to cognitive and non-cognitive skills with respect to college completion, and quantifies the extent to which gender differences in these skills are driving the college gender gap. The use of two distinct college graduation cohorts allows a dynamic analysis of the widening female advantage in college graduation. I decompose the increase in the college gender gap into three pertinent categories of measurable attributes: family background, cognitive skills, and non-cognitive skills (captured by school suspensions, behavioral problems, and legal infractions). A second decomposition is applied to the change in the gap between the two periods. The results show that roughly half of the observed college graduation gender gap in the NLSY97 is due to female advantages in observable characteristics, and roughly half is “unexplained”: due to gender differences in the coefficients. With respect to the change in the gap, approximately 29% of the difference in differences is the “explained” component, attributed to changes in the relative characteristics of men and women. In particular, declining non-cognitive skills in men are associated with about 14% of the increase in the gender gap.

We investigate how career disruptions in terms of job loss may impact morbidity for individuals diagnosed with type 2 diabetes (T2D). Combining unique, high-quality longitudinal data from the Swedish National Diabetes Register (NDR) with matched employer–employee data, we focus on individuals diagnosed with T2D, who are established on the labor market and who lose their job in a mass layoff. Using a conditional difference-in-differences evaluation approach, our results give limited support for job loss having an impact on health behavior, diabetes progression, and cardiovascular risk factors.

Purpose – This chapter compares the stability of the U.S. Dual Banking system's two bank groups, national and state banks, in light of the current financial crisis. The goal of the chapter is to answer three distinct questions: first, is there a difference in the (balance sheet) fragility between the two groups and, second, to what extent has the balance sheet fragility of both groups changed after the escalation of the financial crisis beginning in August 2007? Building on that, the third question asks to whether or not the respective regulatory agencies of both bank groups are responsible for these changes in balance sheet fragility in light of the financial crisis.

Methodology – To answer these questions the chapter uses U.S. Call Report data containing full quarterly balance sheets and P&Ls of all U.S. commercial banks over the period 2005–2008. Anecdotal evidence as well as univariate and multivariate difference-in-difference methodology focusing on the immediate pre-crisis period Q1/2005–Q3/2007 and the crisis period Q3/2007–Q4/2008 are applied.

Results – Highly significant and robust results show that, ceteris paribus, national banks reduced their potential balance sheet fragility after the escalation of the crisis in August 2007 by reducing lending and liquidity creation stronger than state banks. Anecdotal evidence supports the empirical findings. Although both FDIC and OCC did not anticipate the adverse effects of the crisis, the OCC publicly showed an earlier reaction to liquidity-related problems than the FDIC.

Originality – The chapter is the first of its kind to analyze bank fragility around the escalation of the financial crisis and the role of the regulatory agencies. The chapter holds especially interesting policy implications in the light of the current discussion about the future regulation of the banking markets.

This paper looks at the changes in the time allocation of welfare recipients in the United States following the 1996 welfare reform and other changes in their economic environment. Time use is a major determinant of well-being, and for policymakers to understand the broad influences that their policies can have on a population they ought to consider changes in all activities, not simply paid work. While an increase in market work of the welfare population has been well documented, little is known on the evolution of the balance of their time. Using the Current Population Survey to model the propensity to receive welfare, together with a multiple imputation procedure, I replicate previous difference-in-differences estimates that found an increase in child care and a decline in nonmarket work. However when additional data sources are used, I find that time spent providing child care does not increase. This is especially relevant as welfare recipients are overwhelmingly poor single mothers and the welfare reform increased time at work with ambiguous effects on time spent with children. I also find that time at work follows business cycles, with dramatic increases in work time throughout the strong economy of the late 1990s, accompanied by less time in leisure activities.

In this paper, we develop a simple model of the signaling value of the General Educational Development certificate (GED) credential. The model illustrates necessary assumptions for a difference-in-differences estimator that uses a change in the GED passing standard to yield unbiased estimates of the signaling value of the GED for marginal passers. We apply the model to the national 1997 passing standard increase, which affected GED test takers in Texas. We utilize unique data from the Texas Schools Micro Data Panel (TSMP) that contain demographic and GED test score information from the Texas Education Agency linked to pre- and post-test-taking Unemployment Insurance quarterly wage records from the Texas Workforce Commission. Comparing Texas dropouts who acquired a GED before the passing standard was raised in 1997 to dropouts with the same test scores who failed the GED exams after the passing standard hike, we find no evidence of a positive GED signaling effect on earnings. However, we find some evidence suggesting that our finding may be due to the low GED passing threshold that existed in Texas for an extended period.

This paper tests whether employers responded particularly negatively to African American job applicants during the deep U.S. recession that began in 2007. Theories of labor queuing and social closure posit that members of privileged groups will act to minimize labor market competition in times of economic turbulence, which could advantage Whites relative to African Americans. Although social closure should be weakest in the less desirable, low-wage job market, it may extend downward during recessions, pushing minority groups further down the labor queue and exacerbating racial inequalities in hiring. We consider two complementary data sources: (1) a field experiment with a randomized block design and (2) the nationally representative NLSY97 sample. Contrary to expectations, both analyses reveal a comparable recession-based decline in job prospects for White and African American male applicants, implying that hiring managers did not adapt new forms of social closure and demonstrating the durability of inequality even in times of structural change. Despite this proportionate drop, however, the recession left African Americans in an extremely disadvantaged position. Whites during the recession obtained favorable responses from employers at rates similar to African Americans prior to the recession. The combination of experimental methods and nationally representative longitudinal data yields strong evidence on how race and recession affect job prospects in the low-wage labor market.

In order to study whether college scholarships can be an effective tool in raising students’ performance in secondary school, we use one aspect of the Kalamazoo Promise that resembles a quasi-experiment. The surprise announcement of the scholarship created a large change in expected college tuition costs that varied across different groups of students based on past enrollment decisions. This variation is arguably exogenous to unobserved student characteristics. We estimate the effects of this change by a set of “difference-in-differences” regressions where we compare the change in student outcomes in secondary school across time for different student “length of enrollment” groups. We also control for student fixed effects. We find positive effects of the Kalamazoo Promise on Promise-eligible students large enough to be deemed important – about a 9 percent increase in the probability of earning any credits and one less suspension day per year. We also find large increases in GPA among African American students.

I examine changes in the incidence and consequences of job loss between 1981 and 2001 using data from the Displaced Workers Surveys (DWS) from 1984 to 2002. The overall rate of job loss has a strong counter-cyclical component, but the job-loss rate was higher than might have been expected during the mid-1990’s given the strong labor market during that period. While the job-loss rate of more-educated workers increased, less-educated workers continue to have the highest rates of job loss overall. Displaced workers have a substantially reduced probability of employment and an increased probability of part-time employment subsequent to job loss. The more educated have higher post-displacement employment rates and are more likely to be employed full-time. The probabilities of employment and full-time employment among those reemployed subsequent to job loss increased substantially in the late 1990s, suggesting that the strong labor market eased the transition of displaced workers. Reemployment rates dropped sharply in the recession of 2001. Those re-employed, even full-time and regardless of education level, suffer significant earnings declines relative to what they earned before they were displaced. Additionally, foregone earnings growth (the growth in earnings that would have occurred had the workers not been displaced), is an important part of the cost of job loss for re-employed full-time job losers. There is no evidence of a decline during the tight labor market of the 1990s in the earnings loss of displaced workers who were reemployed full-time. In fact, earnings losses of displaced workers have been increasing since the mid 1990s.

This chapter uses the nonlinear difference-in-difference (NL-DID) methodology developed by Athey and Imbens (2006) to estimate the effects of a treatment program on the entire distribution of an outcome variable. The NL-DID estimates the entire counterfactual distribution of an outcome variable that would have occurred in the absence of treatment. This chapter extends the Monte Carlo results in Athey and Imbens's (2006) to assess the efficacy of the NL-DID estimators in finite samples. Furthermore, the NL-DID methodology recovers the entire outcome distribution in the absence of treatment. Further, we consider the empirical size and power of tests statistics for equality of mean, medians, and complete distributions as suggested by Abadie (2002). The results show that the NL-DID estimator can effectively be used to recover the average treatment effect, as well as the entire distribution of the treatment effects when there is no selection during the treatment period in finite samples.

This chapter examines the effects that airports have had on economic development in cities from 1950 to 2010. It uses a novel dataset consisting of previously unexploited data on the origins and history of the aviation system in the United States. Applying the method of synthetic controls to a set of medium and small airports, I examine both the overall impacts and the heterogeneity within the outcomes of various airports. Then, I use regression analysis to determine key factors differentiating successful airports from less successful ones, as it pertains particularly to population and employment growth. I find that, first, on average, cities have benefited from airports over this period. Airports, overall, provided a causal contribution of 0.2– 0.6% per year on population and employment growth over the time period. Second, I show that city-level factors contributing to airport success include: (1) closer proximity to a major research university, (2) a capital city location, and (3) climate factors, particularly higher January mean temperatures and/or hours of sunshine. City size is a consideration as well; cities in larger metropolitan areas, with larger shares of employment in nontradables in the 1950s, were also better positioned to reap the benefits that airports provided on city growth. Significant differences were not found across regions, airport governance structures, or other factors.