Research
Working Papers
“The Effect of Merit-Based Free Community College” Link to Paper
with Kelly Hallberg, Elijah Ruiz, and Marvin Slaughter
Abstract
Free community college is often promoted as a way to expand access and reduce student debt, but may have unintended consequences if it reduces bachelor’s degree completion for students diverted from better resourced four-year universities. By examining a merit-based free community college program in Chicago called the Star Scholarship, we identify the impact of free community college on a distinct set of students: those likely to be deciding where to enroll rather than whether to enroll in college. Using a regression discontinuity design around the 3.0 GPA cutoff, we find that eligibility for merit-based free community college does not increase overall college enrollment, but does significantly shift students from starting at 4-year universities to first enrolling at community colleges. Notably, this diversion does not reduce the probability of eventually earning a bachelor’s degree within six years of graduating high school and eligible students are 2.1 percentage points more likely to earn an associate degree within three years. There is no evidence of a large decrease in the quality of the first college a student enrolls in nor do we see a decline in STEM degree completion for eligible students. Take-up is highest among students likely to be from immigrant families, highlighting unmet financial need among this group. These findings suggest that for the average student near the merit threshold, free community college can increase degree attainment without causing students to substitute two-year degrees for four-year degrees.
“Separation of Church and State Curricula? Examining Public and Religious Private School Textbooks” (Under Revision at The Review of Economic Studies) Link to Paper
with Anjali Adukia
Abstract
Curricular materials not only impart knowledge but also instill values and shape collective memory. Growth in U.S. school choice programs has increased public funds directed to religious schools, but little is known about what they teach. We examine textbooks from religious educational settings and from public schools in Texas and California, applying and improving upon computer vision and natural language processing tools to measure topics, values, representation, and portrayal over time. Political polarization suggests a narrative of divergence, but our analysis reveals meaningful parallels between the public school collections overall, while religious textbooks differ notably, featuring less female representation, characters with lighter skin, more White individuals, and differential portrayal of topics such as evolution and religion. Important similarities, however, also emerge: for example, each collection portrays females in contexts that are more positive but less active and powerful than males, and depicts the U.S. founding era and slavery in similar contexts.
“Unpacking the Long-Term Impact of Holistic Supports for Community College Students” Link to Paper
with Kelly Hallberg, Elijah Ruiz, and Marvin Slaughter
Abstract
This paper presents longer-term findings from a randomized controlled trial of One Million Degrees (OMD), a comprehensive support program for community college students in the Chicago metro area that provides financial, academic, personal, and professional assistance. Results from an initial evaluation found that an offer of a spot in the OMD program led to increased college enrollment, persistence, and associate degree attainment three years after randomization. With eight years of follow-up, we find that these effects persist, indicating the program causes applicants to enroll in and complete more degrees rather than solely accelerating completion. The impacts are concentrated among students who applied while still in high school compared to continuing community college students. For high school applicants, participation in OMD significantly improved labor market outcomes: in every year after randomization, they were more likely to be enrolled in school full time or employed full time, and by year seven they earned higher wages and held more stable jobs. For students already enrolled in community college, effects on labor market outcomes are positive but not statistically significant. For both high school and community college applicants, we observe positive, albeit insignificant, effects on bachelor’s degree attainment eight years after randomization. Taken together, these findings suggest that OMD improves long-term employment outcomes with effects operating through both increased degree attainment as well as the broader benefits of mentoring and advising. Compared to other holistic support models, we find smaller (though less precise) effects for students already in college, but larger gains in long-run attainment and earnings for students applying directly from high school. This highlights the importance of extending holistic supports to students at the critical decision stage of initial college entry, rather than limiting the offer of supports to those who have already enrolled.
“Supporting the Early Academic Momentum of Community College Students: Examining the Impact of Incorporating GPA into Course Placement” Link to Paper
with Kelly Hallberg, Adam Leader-Smith, Elijah Ruiz, Marvin Slaughter, and Courtney Washington
Abstract
Placing community college students into their first math or English course is a critical decision with significant implications. Students placed into coursework that is too challenging may struggle to persist, while those placed into unnecessary developmental coursework may experience reduced academic momentum, delayed graduation, and increased costs. Recent research has found that using multiple measures for placement, like high school GPA alongside standardized test scores, can improve placement accuracy and student’s academic outcomes. This study assesses the impact of a new multiple measure placement policy at City Colleges of Chicago, which “boosts” students with a high school GPA of 3.0 or above into higher math and English course levels, reducing their need for developmental education. In practice, the math and English placement boost applied to fewer than 5 percent and 8 percent of students, respectively, with even fewer students taking advantage of the boosted placement. Despite the small number of affected students, using a difference in regression discontinuity design we find that access to a placement boost in math or English courses decreased the number of developmental courses taken by students without affecting overall academic performance or persistence. However, it also led to delayed course-taking for these required math and English courses. Qualitative findings from interviews with students revealed a lack of awareness about the policy, suggesting a need for better communication and easier transcript sharing between Chicago Public Schools and City Colleges to maximize policy benefits.
Peer-Reviewed Publications
“What We Teach About Race and Gender: Representation in Images and
Text of Children’s Books”
Quarterly Journal of
Economics (November 2023)
with Anjali Adukia, Alex Eble, H. Birali Runesha, and Teodora Szasz
Abstract
Books shape how children learn about society and social norms, in part through the representation of different characters. To better understand the messages children encounter in books, we introduce new machine-led methods for systematically converting images into data. We apply these image tools, along with established text analysis methods, to measure the representation of race, gender, and age in children’s books commonly found in US schools and homes over the last century. We find that books selected to highlight people of color, or females of all races, consistently depict characters with darker skin tones than characters in “mainstream” books, which depict lighter-skinned characters even after conditioning on perceived race. Children are depicted with lighter skin than adults, despite no biological foundation for such a difference. Females are more represented in images than in text, suggesting greater symbolic inclusion in pictures than substantive inclusion in stories. Relative to the US Census, Black and Latinx people are underrepresented; whereas males, particularly White males, are persistently overrepresented. Our data provide a view into the “black box” of education through children’s books in US schools and homes, highlighting what has changed and what has endured over time.
[Press: Time Magazine, Wall Street Journal, School Library Journal, Code Together, Inequalitalks, FutureEd, The 74, Named one of the ten most significant studies of 2021 by George Lucas Foundation’s Edutopia]
“Tales and Tropes: Gender Roles from Word Embeddings in a Century of
Children’s Books”
Proceedings of
the 28th International Conference on Computational Linguistics
(October 2022)
Abstract
The manner in which gender is portrayed in materials used to teach children conveys messages about people’s roles in society. In this paper, we measure the gendered depiction of central domains of social life in 100 years of highly influential children’s books. We make two main contributions: (1) we find that the portrayal of gender in these books reproduces traditional gender norms in society, and (2) we publish StoryWords 1.0, the first word embeddings trained on such a large body of children’s literature. We find that, relative to males, females are more likely to be represented in relation to their appearance than in relation to their competence; second, they are more likely to be represented in relation to their role in the family than their role in business. Finally, we find that non-binary or gender-fluid individuals are rarely mentioned. Our analysis advances understanding of the different messages contained in content commonly used to teach children, with immediate applications for practice, policy, and research.
“Measuring Representation of Race, Gender, and Age in Children’s
Books: Face Detection and Feature Classification in Illustrated Images”
Proceedings
of the IEEE/CVF Winter Conference on Applications of Computer Vision
(January 2022)
Abstract
Images in children’s books convey messages about society and the roles that people play in it. Understanding these messages requires systematic measurement of who is represented. Computer vision face detection tools can provide such measurements; however, state-of-the-art face detection models were trained with photographs, and 80% of images in children’s books are illustrated; thus existing methods both misclassify and miss classifying many faces. In this paper, we introduce a new approach to analyze images using AI tools, resulting in data that can assess representation of race, gender, and age in both illustrations and photographs in children’s books.
Other Publications
“Redesigning Developmental Education Placement Policies: A Case Study of City Colleges of Chicago’s Approach” The Inclusive Economy Lab (May 2023)
“What Award Winning Books Teach Children About Race and Gender” The Brown Center Chalkboard (June 2023)
with Anjali Adukia, Alex Eble
Selected Works In Progress
“Natural Disasters, Labor Markets, and Economic Mobility of Young Adults”
with Mythili Vinnakota
“Community College Deserts and Labor Market Outcomes”
with Riley Acton, Kalena E. Cortes, Lois Miller, and Camila Morales
“(How) Do We Teach Emotions?”
with Anjali Adukia, Matthew Bonci, Paula Dastres, Jake Nicoll, and Teodora Szasz
“Skin Color, Marriage Markets, and Women’s Empowerment in Bangladesh”
with Anjali Adukia, Nina Buchmann, Erica Field, and Rachel Glennerster
“Influencing Identities: The Impact of Content Creators on Character Representation and Portrayal”
with Anjali Adukia and Jake Nicoll