This website uses cookies.

By clicking the "Accept" button or continuing to browse our site, you agree to first-party and session-only cookies being stored on your device to enhance site navigation and analyze site performance and traffic. For more information on our use of cookies, please see our Privacy Policy.

  1. Home
  2. ASSA Annual Meeting
  3. 2025
  4. Program
  5. Economics of Education
« Back to Results

Economics of Education

Lightning Round Session

Saturday, Jan. 4, 2025 8:00 AM - 10:00 AM (PST)

Parc 55, Hearst
Hosted By: American Economic Association
  • Chair: Christopher Walters, University of California-Berkeley

Buy In, Opt Out? The Effect of the School Choice Lottery on Parental Educational Spending and Investment Decision: Evidence from a Quasi-Experiment in China

Guodong Chen
,
New York University-Shanghai
Yu Zhou
,
New York University-Shanghai
Cheng Jiang
,
Peking University
Jianye Yan
,
China Agricultural University

Abstract

Over-subscription in the school choices is common in developing countries. As
a result, the school choice lottery has become a popular tool to solve the oversubscription
issues. Using a unique user level data from a large mobile platform, we examine the impact of the school choice lottery on the parental educational spending and other consumption and investment decision at China. Our results show that parents respond to the reform significantly after the announcement of the policy. In addition, we find that there exists a large heterogeneity across the gender and migrant history as well as their consumption level. Furthermore, we explore the potential source to finance the increase in educational spending and find that users with low to medium wealth will crowd out the general merchandise spending, but the users with high wealth level will cash out their mutual fund investment. Our results imply that there exists a large magnitude of undocumented noncompliance, which could potentially bias the intended policy effect.

When The Student Becomes the Master: A Field Experiment on Learning by Teaching

Rohen Shah
,
University of Chicago

Abstract

Learning by Doing (Arrow, 1962) is a longstanding theory on skill development, but it is unclear what the "doing" might be for students learning an academic subject. In this paper, I propose that for students, the "doing" of learning is actually teaching. This study presents evidence from a large-scale field experiment that randomly assigned students to experimental conditions in which they are either (1) assigned to make ``explanation'' videos, (2) assigned additional practice problems, or (3) placed in a pure control condition. The explanation treatment improved short-run scores by 0.17 standard deviations and long-run grades by 0.07 standard deviations relative to the practice-problem group. Notably, while both treatment groups improved relative to control, only the explanation treatment improved performance on novel problems, suggesting that explaining concepts enhances one's ability to understand deeply and generalize concepts.

Inspirational or Demotivational: Experimental Evidence on the Impacts of Role Models

Aurelia Aochun Di
,
University of Washington

Abstract

High-achieving role models can raise aspirations and improve performance, but may negatively affect individuals who fail to meet their goals. In this paper, I study what role model characteristics influence their effectiveness in improving academic performance. Using a randomized controlled trial with students across five middle schools in China, I compare the impacts of role models with different success levels. Two months later, students exposed to higher-achieving role models improved test scores by 0.07-0.18 standard deviations, whereas those exposed to moderately achieving role models experienced an average 28.8% and 26.6% reduction in the likelihood of feeling depressed and stressed, respectively. Higher-achieving role models improve low-performing girls’ academic outcomes but negatively affect their mental health, as these girls invested more effort but still found their improved performance falling short of their elevated aspirations. This paper highlights the negative impacts of role models on mental health as a trade-off for enhancing performance in an underperforming subgroup, emphasizing the need to consider mental health when implementing role model interventions.

Not Too Early, Not Too Late: Encouraging Engagement in Education

Tracy Xiao Liu
,
Tsinghua University
Ulrike Malmendier
,
University of California-Berkeley
Stephanie Wang
,
University of Pittsburgh
Shuhuai Zhang
,
Tsinghua University

Abstract

A common reason why individuals fail to reach educational and other longterm goals is that it is hard to maintain the required effort over time. One example is the decline in attendance and student engagement over the course of a semester. While educators often attempt to address these issues with information and motivational material at the beginning of a class, or with remedial sessions at the end of the class, we argue that the critical time in longterm sustained efforts falls instead in the middle. To study the effect of timing, we conduct a field experiment that assigns a task aimed at engaging stude nts with the class material at different times throughout the semester. We show that assigning tasks to low-performing students in the middle of the term, compared to early or late in the semester, improves their performance along several dimensions: attendance, homework grades, and exam grades. Consistent with our theoretical framework, our findings suggest that accumulating frictions over the course of a semester and the corresponding decline in effort can be addressed by an intermediate timing of costly "engagement'’ tasks.

Overconfidence and Inaccurate Self-Assessment: Evidence from Student Grade Estimations

Huan Cai
,
Cornell College
Jie Zheng
,
Shandong University

Abstract

We conduct a series of surveys in the 12th grade of a senior-high school in China for a cohort of around 1,800 students who took the annual National College Entrance Exam (also known as Gaokao) in July of the years 2022 and 2023. Students were asked to fill in multiple surveys for five sets of mock exams and Gaokao: one before each exam, and the other after the exam but before the exam results are revealed.

Using over 1,300 students’ self-reported estimations of their exam performance on six different subjects, we find stronger effects of overplacement (of their ranks) than overestimation (of exam scores) for the same task, following the definitions of overconfidence in Moore and Healy (2008). We also find that overestimation is most severe when estimating Gaokao scores in the subject of mathematics.

In estimates of the current task (of mock exams), we also confirm the Dunning-Kruger effects that students who are less competent in a subject are more likely to overplace their performances. Such Dunning-Kruger effects are stronger in the estimates of a future task (of Gaokao exams).

To further understand these new findings, we decompose the traditional overconfidence measures into two separate perspectives: biased estimates (measured by the signs of the estimation errors), and inaccurate self-assessments (measured by the absolute value of the estimation errors). We show that, both perspectives contribute to the Dunning-Kruger effects: students who are less competent in a subject are more likely to overplace their performances and have less accurate self-assessments. However, biased estimates seem to play a more important role, and taking the survey after the exam only made it worse. But the exam does help reduce the gap in self-assessment accuracy between students with different competency.

The Effect of Losing Pell Grant Eligibility on Student Outcomes

Shinyoung Kim
,
Iowa State University

Abstract

This paper examines the effects of Pell Grant eligibility on student outcomes using a regression discontinuity (RD) design and bounds treatment effects to account for selection effects arising from the loss of Pell Grant eligibility- a question unexplored in previous research. While initial eligibility is determined by financial need, students must achieve Satisfactory Academic Progress (SAP) to retain the grant. The findings show that students who initially received more grant aid are 26 percentage points less likely to persist in the year they lose Pell and 8 percentage points less likely to graduate within 4 years compared to those with less aid. Moreover, those who did not persist had significantly lower SAT math scores, suggesting that non-persistence is not random; thus, the RD estimand is confounded by selection effects. Finally, I estimate bounds on the treatment effects, accounting for selection effects. The bounding results suggest that the previously estimated effects of additional grants on student outcomes may be underestimated. For example, while naive RD estimates find no effect on 4-year graduation rates, accounting for selection effects reveals that students receiving additional grant aid are 7 percentage points more likely to graduate within 4 years compared to those with less aid. One may argue that the SAP policy enhances aid efficiency by pushing out students whose costs of finishing college exceed the benefits; however, evidence suggests that it also dismisses students who may be capable of completing and potentially benefiting from college.

The effects of STEM versus humanities in high school: evidence from admissions cutoffs, administrative data, and large-scale surveys

Robert Ainsworth
,
University of Florida
Rajeev Dehejia
,
New York University
Andrei Munteanu
,
University of Quebec-Montreal
Cristian Pop-Eleches
,
Columbia University
Miguel Urquiola
,
Columbia University

Abstract

We estimate the impacts of being assigned to a STEM- versus humanities-focused high school curriculum. We use a regression discontinuity design based on cutoffs in Romania’s high school assignment system, together with administrative data (to measure high school enrollment and performance) and large-scale surveys (to measure college enrollment, career plans, high school and college satisfaction, wellbeing, time use, and beliefs and preferences). STEM assignment makes students 67 pp more likely to graduate from high school STEM, 24 pp more likely to enroll in a college STEM program, and 23 pp more likely to plan to pursue a STEM career. Further, it leads students to have greater confidence in their STEM abilities and to enjoy STEM more. During high school, it makes students spend more time on homework and less on social media and reading, have more male and fewer female friends, and report higher wellbeing. Importantly, the STEM curriculum is risky: STEM assignment causes students to score worse on a high-stakes high school exit exam, leading to a lower probability of attending any college and a decline in high school satisfaction by the first year after high school. The risk is especially pronounced for students with low baseline achievement, for whom the reduction in college enrollment is 16 pp. Also, STEM assignment makes females report lower college satisfaction, primarily due to negative experiences with professors and peers—though it does not influence regret over high school or college application choices. Finally, among males, STEM assignment leads to more traditionalist expectations and greater political conservatism.

To Inspire and To Inform: The Role of Role Models

Marcos Agurto
,
University of Piura
Muchin Isabel Ayen Bazan Ruiz
,
Georgia Institute of Technology
Siddharth Hari
,
World Bank
Sudipta Sarangi
,
Virginia Tech

Abstract

Bridging the gender gap in STEM fields has emerged as a concern for policymakers both in developed and developing countries. This paper examines the impact of light touch intervention where female engineering students act as role-models for high school students in Peru. We find that a brief 20-minute interaction with the role-models led to sharp increases in preferences towards engineering, with the effects being concentrated on female students with high math aptitude. We find that these results are driven by increased self-confidence as a result of exposure to role models. Set in the context of a developing country, our results show that low-cost interventions can be helpful in reducing the STEM gender gap but cannot address broader deep seated gender stereotypes.

What Is Essential Is Visible to the Eye: Saliency in Primary School Ranking and Its Effect on Academic Achievements

Francois-xavier Ladant
,
Northwestern University and Harvard University
Julien Hédou
,
Stanford University
Paolo Sestito
,
Bank of Italy
Falco Bargagli-Stoffi
,
Harvard University

Abstract

We propose a new strategy to identify the impact of class rank, exploiting a “visible” primary school rank from teachers’ exam grades, and an “invisible” rank from unreported standardized test scores. Leveraging a unique panel dataset on Italian students, we show that the visible rank has a substantial impact on students’ perceptions, which affects subsequent academic performance. However, the effect of being surrounded by higher-SES or higher-achieving peers remains positive even accounting for the decrease in rank. Higher-ranked students self-select into high schools with higher average student achievements. Finally, exploiting an extensive survey, we identify psychological mechanisms channeling the rank effect.
JEL Classifications
  • I2 - Education and Research Institutions