Abstract: Greek-letter student social groups, better known as fraternities and sororities, are a ubiquitous feature on many American higher education campuses. These organizations, especially fraternities, have a reputation for encouraging unruly and improper behaviour among both members and non-members. This paper investigates the effect of the degree of prevalence of these Greek organizations at a campus, as measured by the percentage of students who are members of fraternities and sororities, on the instances of liquor and drug law violations on campuses, as measured by the number of arrests for liquor and drug laws violations. Using a unique dataset, which combines data from three sources, we address any potential selection bias by including several controls associated with party culture and through the inclusion of institution-level fixed effects. We find that a larger percentage of students in fraternities (but not sororities) is associated with an increase in the number of arrests for drug law violations. A larger percentage of students in sororities (but not the percentage of students in fraternities) is associated with a larger number of arrests for liquor law violations. This result is highly significant and is robust across various specifications.
Abstract: In this study, we use institution-level data for the period 2004 to 2016 from the Integrated Postsecondary Education Data System to examine the excess revenues of private, four-year nonprofit institutions. We present data on the magnitude of excess revenues for these institutions over this period, examine how excess revenues are associated with different types of private institutions, and how within-institution excess revenues are affected by changes in time-varying factors, such as their size, selectivity, revenue structure, and expense distribution. We find that across most years in our sample, private, four-year nonprofits averaged double-digit excess returns. The results show that variations over time in excess revenues are related to a number of factors, including institution size, yield rates, net tuition revenue, and tuition discount rates.
Abstract: Colleges and universities offer classes that meet for different lengths of time and different numbers of days per week, such as classes that meet 2 days and those that meet 3 days. Traditionally triweekly classes that met for a shorter duration were more common than classes that met biweekly for a longer duration. Biweekly classes are becoming more popular with time. However, there is some concern that classes that meet more often are better suited for student learning than others. This paper, using data from a small liberal arts college, finds that after controlling for the starting time of the class meeting and course fixed effects as well as faculty and student fixed effects, student learning across 2 and 3 days classes is essentially the same.
Abstract: This econometrics pedagogy note points to online material that demonstrates the importance of using cluster standard errors (SEs) with data generated from complex surveys. Simulation is used to show that both classic ordinary least squares and robust SEs perform poorly in the presence of within-cluster correlated errors, while cluster SEs perform much better. We take advantage of Excel’s spreadsheet interface to produce clear and intuitive visuals of the data generation process and explain key results. Customizable Stata and R implementations, which help in further analysis by taking advantage of the unique different capabilities of Stata and R, are also provided. We conclude with suggestions for how to use these files in the classroom.
Abstract: Research in psychology has shown that early morning classes are not conducive to learning because of the peculiar sleep cycles of adolescents and young adults that cause them to be especially groggy in the morning. Our study examines the relationship between the times that classes are offered and the grades that students in these classes earn at a highly selective liberal arts college. Our main findings are that morning classes are harmful for student achievement. Grades are especially lower for classes that were scheduled at 8 am and 9 am. Moreover, while students of both genders are adversely affected by early morning courses, the effects are particularly pronounced for male students. This institution assigns students randomly to different sections of the same course, thus creating a quasi-natural experiment and enabling us to control for unobserved characteristics of students. In addition, we include student and faculty fixed effects.
Abstract: In this paper, we explore whether there is a relationship between average grades earned in a course and the national average salaries of graduates of the major associated with the course. Using student-level data from a selective private liberal arts college, we find an inverse relationship. The result suggests that students face a trade-off between grades earned in college versus higher expected earnings in the future. This relationship is stronger for students with lower math SAT scores but not for those with lower verbal SAT scores. Finally, the female advantage in course grades diminishes significantly in majors with higher salaries.
Abstract: Colleges want to increase retention and graduation rates, but they are also under pressure to control costs. Increasing class size is a common method to reduce per student costs. This paper examines the relationship between class size and student achievement. Using data from a selective liberal arts college, we show that grades of students decrease as class size increases. Moreover, relatively vulnerable students such as first-years or those with low SAT scores experience on average larger negative effects from increases in class sizes. The findings suggest that attempts to control costs may harm students, particularly those least likely to graduate.
Abstract: This paper focuses on econometrics pedagogy. It demonstrates the importance of including probability weights in regression analysis using data from surveys that do not use simple random samples (SRS). We use concrete, numerical examples and simulation to show how to effectively teach this difficult material to a student audience. We relax the assumption of simple random sampling and show how unequal probability of selection can lead to biased, inconsistent OLS slope estimates. We then explain and apply probability weighted least squares, showing how weighting the observations by the reciprocal of the probability of inclusion in the sample improves performance. The exposition is non-mathematical and relies heavily on intuitive, visual displays to make the content accessible to students. This paper will enable professors to incorporate unequal probability of selection into their courses and allow students to use best practice techniques in analyzing data from complex surveys. The primary delivery vehicle is Microsoft Excel®. Two user-defined array functions, SAMPLE and LINESTW, are included in a prepared Excel workbook. We replicate all results in Stata® and offer a do-file for easy analysis in Stata. Documented code in Excel and Stata allows users to see each step in the sampling and probability weighted least squares algorithms. All files and code are available at www.depauw.edu/learn/stata.
Abstract: It is natural to suppose that a prosecutor’s conviction rate—the ratio of convictions to cases prosecuted—is a sign of his competence. Prosecutors, however, choose which cases to prosecute. If they prosecute only the strongest cases, they will have high conviction rates. Any system that pays attention to conviction rates, as opposed to the number of convictions, is liable to abuse. As a prosecutor’s budget increases, he allocates it between prosecuting more cases and putting more effort into existing cases. Either can be socially desirable, depending on particular circumstances. We model the tradeoffs theoretically in two models, one of a benevolent social planner and one of a prosecutor who values not just the number of convictions but the conviction rate and unrelated personal goals. We apply the model to U.S. data drawn from county-level crime statistics and a survey of all state prosecutors by district. Conviction rates do have a small negative correlation with prosecutorial budgets, but conditioning on other variables in regression analysis, higher budgets are associated both with more prosecutions and higher conviction rates.