Beliefs are a critical component of economic decision-making. Since they are not directly observable, they are generally elicited through survey questions. Hence, it is important to know whether and how incentivization affects the elicited beliefs. Using an experiment, this paper tests the effect of incentivized versus unincentivized belief elicitation on students’ beliefs about returns to effort in an educational setting. Initially, incentivization significantly reduces the bias in beliefs, without impacting the accuracy of beliefs. It also reduces the frequency of round answers and increases time spent responding to the belief questions, both of which suggest increased effort. However, these initial effects on bias and response times diminish or become nonexistent in later weeks, suggesting short-lived effects.

This study examines the role of job-related and non-job-related factors in college major choice. Using a staggered intervention, we provide students information on various aspects of majors and assess the impact of different pieces of information on their stated choices. We show that major choices depend on a wide set of factors, especially for students who are initially unsure about their major choice. The non-job-related factors, such as a major’s course difficulty and gender composition, are particularly important to students. Male and female students value different major characteristics in different ways. Female students – particularly those with below-median high school GPA – avoid majors that are more difficult than they originally believed, while male students are averse to majors with more female faculty but prefer those with more female students. Our findings help us understand gender gaps in college major choice and have a number of implications for researchers and policymakers seeking to study major choice or influence those choices.

How might the visibility of an author's name and/or institutional affiliation allow bias to enter the evaluation of economics papers? We ask highly qualified journal editors to review abstracts of solo-authored papers which differ along the dimensions of gender and institution of the author. We exogenously vary whether editors observe the name and/or institution of the author. We identify positive name visibility effects for female economists and positive institution visibility effects for economists at the top institutions. Our results suggest that male economists at top institutions benefit the most from non-blind evaluations, followed by female economists (regardless of their institution).

While distance learning has become widespread, causal estimates regarding returns to effort in technology-assisted learning environments are scarce due to high attrition rates and endogeneity of effort. In this paper, I manipulate effort by randomly assigning students different numbers of lessons in a popular online language learning platform. Using administrative data from the platform and the instrumental variables strategy, I find that completing 9 Duolingo lessons, which corresponds to approximately 60 minutes of studying, leads to a 0.057–0.095 standard deviation increase in test scores. Comparisons to the literature and back-of-the-envelope calculations suggest that distance learning can be as effective as in-person learning for college students for an introductory language course.

This paper investigates the pricing of malls in an environment where shoppers choose between a car and public transportation in getting to a suburban mall. The mall implicitly engages in mixed bundling; it sells goods bundled with parking to shoppers who come by car, and only goods to shoppers who come by public transportation. There are external costs of discomfort in public transportation due to crowdedness. Thus, shoppers using public transportation crowd out each other. The mall internalizes these external costs, much like a policy maker. To do so, it raises the sales price of the good and sets a parking fee less than parking's marginal cost. Hence, parking is always a loss leader. Surprisingly, this pricing scheme is not necessarily distortionary.

Women remain underrepresented in STEM occupations, largely due to educational choices. This study investigates two interventions—performance feedback and advice—aimed at narrowing the gender gap in selecting math tasks over verbal tasks. In an online lab experiment, participants chose between completing a math task or a verbal task. Before making their choice, they received either performance feedback on previous tasks, advice on which task to choose, or both, with advisor gender varied. A significant gender gap was observed in math task choices in the group who did not receive the interventions. Performance feedback intervention eliminated this gap by increasing the proportion of females selecting the math task, while leaving choices of males unaffected. Being advised to choose math task increased math task selection only among females, while being advised to choose verbal task decreased math task selection only among males. The advisor’s gender generally had no significant impact on the effectiveness of advice. This study contributes to our understanding how performance feedback and advice affect educational decisions, highlighting the complexity of these interventions.

Teamwork is a critical skill in the workforce and is often the most despised  aspect of classroom instruction by students. Team contracts require students to decide and  document how they will communicate, divide work, and resolve potential conflicts before beginning their course-specific group work. We examine the effects of team contracts on students’ attitudes towards teamwork, individual teamwork skills, team dynamics, and their sense of classroom community using a large scale randomized experiment administered through the Economic Education Network for Experiments (EENE). Estimated effects are small and not distinguishable from zero.

This study examines how access to LLMs affects cooperative learning and student performance in undergraduate economics education. While generative AI tools are increasingly common in classrooms, little is known about how they shape peer interaction and collaborative learning environments. We conduct a randomized classroom experiment in an introductory economics course at the University of Chicago. Students work in small groups during weekly discussion sections and in each session, some groups are randomly assigned to receive access to a custom course-specific AI tutor. To balance fairness and experimental variation, each student receives AI access in exactly half of the discussion sections. The design generates within-student variation in AI exposure over time as well as across-student variation in AI exposure before midterm. Primary outcomes include quiz and exam performance. Secondary outcomes measure peer interaction, teamwork experiences, and attitudes toward AI.