Economics of Higher Education
Lightning Round SessionSaturday, Jan. 4, 2025 8:00 AM - 10:00 AM (PST)
- Chair: Caroline Hoxby, Stanford University
College Course Shutouts
Abstract
What happens when college students are not able to enroll in the courses they want? We use a natural experiment at Purdue University in which first-year students are conditionally randomly assigned to oversubscribed courses. Compared to students who are assigned a requested course, those who are shut out are 40% less likely to ever take the oversubscribed course and 30% less likely to ever take a course in the same subject. While a course shutout is equally likely to occur to female and male students who requested the course, shutouts are much more disruptive for female students. In the short run, shutouts decrease the credits female students earn as well as their GPA. In the long-run, shutouts increase the probability female students drop out of school in the first year, decrease the probability they choose majors in STEM fields (Science, Technology, Engineering, and Math), decrease cumulative GPA, and decrease the probability of graduating within four years. In contrast, shutouts have no effects on short-run credits earned, dropout, majoring in STEM, cumulative GPA, or four-year graduation for male students. Shutouts do have one large measurable long-run impact on male students---shutouts significantly increase the probability that men choose a major from the business school.Do Double Majors Face Less Risk? An Analysis of Human Capital Diversification
Abstract
We study how human capital diversification, in the form of double majoring, affects the response of earnings to labor market shocks. Double majors experience substantial protection against earnings shocks, of 56%. This finding holds across different model specifications and data sets. Furthermore, the protection double majors experience is more pronounced when the two majors are more distantly related, highlighting the importance of diverse skill sets. Additional analyses demonstrate that double majors are more likely to work in jobs that require a diverse set of skills and knowledge and are less likely to work in occupations that are closely related to their majors.Gender Bias in Emerging New Research Topics: The Impact of COVID-19 on Women in Science
Abstract
We investigate the impact of new research opportunities on the long-standing under-representation of women in medical and academic leadership by assessing the impact of the emergence of COVID-19 as a new research topic in the life sciences on women's authorship. After collecting publication data from 2019 and 2020 on biomedical publications, where the position of first and last author is most important for future career development, we use the major Medical Subject Heading (MeSH) terms to identify the main research area of each publication and measure the relation of each paper to COVID-19. Using a Difference-in-Difference approach, we find that although the general female authorship trend is upwards, papers in areas related to COVID-19 are less likely to have a woman as first or last author compared to research areas not related to COVID-19. Conversely, new publication opportunities in the COVID-19 research field increase the proportion of women in middle, less-relevant, author positions. Stay-at-home mandates, journal importance, and access to new funds do not fully explain the drop in women's outcomes. The decline in female first authorship is related to the increase of teams in which both lead authors have no prior experience in the COVID-related research field. In addition, pre-existing publishing teams show reduced bias in female key authorship with respect to new teams specifically formed for COVID-related research. This suggests that "opportunistic" teams, transitioning into research areas with emerging interests, possess greater flexibility in choosing the primary and final authors, potentially reducing uncertainties associated with engaging in productions divergent from their past scientific experiences by excluding women scientists from key authorship positions. Understanding the roots of gender inequality in scientific production is essential to develop and implement effective gender policies to achieve institutional goals of gender equal access to emergent scientific topics.Getting Across the Finish Line: How to Boost Completion Rates for Underrepresented Student Groups
Abstract
Despite increasing enrollments, there remains a persistent racial gap in college completion rates in the United States. This gap has long-run implications for inequalities in employment, income and wealth. Universities have focused on creating a diverse student body (peer effects), a diverse faculty (role model effects) and on providing resources to underrepresented students in an effort to boost completion rates. But studies testing completion rates for Black and Hispanic students have produced a variety of sometimes conflicting results. This paper contributes to the literature by testing the different effects together in a panel data set of 9,800 observations from 1,493 universities over 2011-2018. Ordinary Least Squares and Fixed Effects regressions test the determinants of Black and Hispanic completion rates both in absolute terms and relative to White completion rates. The results show that peer effects disappear when role model effects are included. This suggests that previous literature testing only peer effects could be suffering from omitted variable bias. By contrast, role model effects stay significant in improving completion rates even when accounting for peer effects. Faculty salaries and the percentage of full-time faculty are highly significant in reducing the racial gap for Black students but not for Hispanic students, suggesting universities need to be mindful the same approach may not work with all underrepresented groups. The results strongly suggest institutions relying solely on peer effects from recruiting a more diverse student body will not be able to remedy the completion gap. Institutions will require additional resources for faculty salaries and for the recruitment of full-time, diverse faculty members.Impact of College Major Skills on Lifetime Earnings and Sorting
Abstract
This project studies the implications of mismatch in the college labor market, where mismatch is defined as the gap between the skills firms demand and the specific skills different college majors graduate with. In our baseline regression, we estimate how earnings are affected based on the initial mismatch in the first job out of college before we estimate the effect of initial mismatch over lifetime earnings. In our first stab at this, we estimate mismatch penalties over the life-cycle using publicly available data and Lightcast data (formerly known as Burning Glass Data). We find early mismatch penalties associated with customer service, people management, and social skills. Over the life-cycle, we see the penalty for mismatch declines for cognitive, people management, project management, and customer service skills. We use a search model to understand the decline over the life cycle and the importance of each skill and the level of skill a student in each major graduates with. By combining Lightcast data with restrictive use U.S. Census Bureau data (NSCG (National Survey of College Graduates) and ACS (American Community Service)), we can track students throughout their early career stages. With this data and our model, we are able to contribute to the literature by adding which relevant college skills and major specific skills allow students to succeed in the labor market, at least in terms of earnings. This allows us to dissect the importance of 10 different general skills that all college majors acquire some of, as well as major specific skills.Life-cycle Effects of Income-Driven Repayment on Credit Outcomes, Future Student Loan Borrowing, and Labor Market Outcomes
Abstract
This paper studies the short- to medium-run impacts of income-driven repayment (IDR) on student loan borrowers' access to credit and labor market outcomes. We use a novel instrument, which leverages the randomly assigned student loan servicer, to instrument the likelihood a borrower is on income-driven repayment. Using a 2% sample of all individuals with credit in the US from the University of California Credit Panel (UC-CCP), we have estimated the impact of IDR on credit scores, loan balances, having a mortgage, having a small business loan, entering and staying in a licensed career field, and future student loan take-up for borrowers 1-5 years after the end of the first deferment spell.We find that IDR has positive effects on the status of the student loan. We find positive initial impacts on credit followed by negative or null effects in later years. We see increasing student loan balances and an increased likelihood of an individual returning to school and taking out a future student loan. We are the first to show that this may be due to moral hazard where borrowers with IDR access were more likely to take out an additional future student loan.
We additionally show IDR decreases in wealth-generating lines of credit such as mortgages and small business loans. Lastly, we are the first to examine how IDR changes the likelihood of an individual working in licensed professions. We see largely null initial effects on the likelihood an individual is licensed as a teacher or health professional; in the later half of the sample, these effects become negative--suggesting promises of forgiveness through PSLF alone do not increase the likelihood individuals stay in nursing or teaching.
Need-based Financial Aid and Major Choices
Abstract
This paper presents new evidence on the impact of a need-based financial aid program on first-year undergraduate students' major choice. Using rich administrative data, we exploit sharp discontinuities in grant eligibility in Canada to identify the causal effect of eligibility for low-income and middle-income students separately. Our findings show that grant eligibility leads to more grants and fewer loans, as well as a higher likelihood of majoring in the humanities rather than business among low-income students. Students from middle-income families are more likely to pursue health-related majors when they are eligible for more grants instead of loans. However, we do not observe significant differences in major-specific degree completion rates within six years.Signaling Value of Elite Colleges: Experimental Evidence from Bangladesh
Abstract
Recent literature documents higher returns in the labor market from attending elite colleges. These advantages may stem from the value added by these institutions. Alternatively, this can be due to signaling. The study elucidates these mechanisms through an audit-based randomized control trial. We sent around 4,000 fictitious resumes in response to online job advertisements, where we systematically varied the name of the college from which the candidate graduated in the otherwise similar resumes. We find evidence of a significant premium for attending elite colleges -- candidates from elite colleges were 34 percent more likely to receive a callback than similar candidates from non-elite colleges. Utilizing a heteroskedastic probit model that allows decomposition of the marginal effect of graduating from elite college into two distinct components, the "level" effect, and the" variance" effect, we find evidence in favor of taste-based discrimination rather than statistical distribution. We also discern disparities in callback rates for women, who are 30 percent less likely to be called back. Elite college attendance mitigates this effect, especially for resumes of better quality. Our research holds significant implications for policy considerations. We demonstrate that labor markets exhibit discrimination based on the specific colleges individuals attend, even when their abilities are comparable. This bias can perpetuate intergenerational inequality, as students from low- and middle-income families often face barriers in accessing elite private universities due to elevated tuition fees. Consequently, it becomes imperative to formulate policies, such as establishing independent certification markets, to rectify these asymmetries.The Diversity Impacts of Removing Standardized Testing Requirements in College Admissions
Abstract
Test-optional policies for college admissions are currently under debate. At the onset of the COVID-19 pandemic, many schools implemented test-optional policies and are evaluating whether to make these policies permanent. One dimension of interest for colleges is the impact that test-optional policies have on campus diversity. We utilize the Common Data Set (CDS), a higher education survey published annually. Our research focuses on the most recent currently available five years (2018-19 through 2022-23 school years). Our control group includes colleges that maintained their standardized testing requirement throughout the study timeframe and our treatment group includes colleges that went test-optional beginning in the 2020-2021 school year. Using a difference-in-differences framework, we find that shifting to a test-optional policy is associated with increases in female applicants of about 1 percentage point (95% CI: 0.002 to 0.017, p=0.02) by the second year of the new policy. The share of student enrolling by the second year of the policy who are Black also increased by about 1 percentage point (95% CI: 0.001 to 0.018, p = 0.03). The share of students enrolling who are White was unaffected but the share of students enrolling who are Asian decreased by about 1.5 percentage points (95% CI: -0.030 to -0.001, p = 0.04) following the change. These results underscore the complex dynamics involved in promoting diversity through admission policies.JEL Classifications
- I2 - Education and Research Institutions