PUBLICATIONS
Ozier, A., Charron, D., Chung, S., …& Olopade, C. O. (2018). Building a consumer market for ethanol-methanol cooking fuel in Lagos, Nigeria. Energy for Sustainable Development, 46, 65-70.
Wright, K. O., Wright, E., Ottun, T. A., Oyebode, M. O., Sarma, V., & Chung, S. (2018). Economic recession and family planning uptake: Review of a Nigerian health institution. Tropical Journal of Obstetrics and Gynaecology, 35(2), 147-152.
WORKING PAPERS
Chung, S., Persico, C., & Liu, J. (2025) The Effects of Daily Air Pollution on Students and Teachers.” NBER Working Paper No. 33549.
Abstract: Recent empirical research shows that air pollution harms student test scores and attendance and increases office discipline referrals. However, the mechanism by which air pollution operates within schools to negatively affect student and teacher outcomes remains largely opaque. The existing literature has primarily focused on the effects of prolonged exposure to pollution on end-of-year test scores or total absence counts. We examine how ambient air pollution influences student-by-day and teacher-by-day outcomes, including absences and office discipline referrals, using daily administrative data from a large urban school district in California between 2003 and 2020. Using wind direction as an instrument for daily pollution exposure, we find that a 1 μg/m3 increase in daily PM2.5 causes a 0.56% increase in full-day student absences and a 2.76% increase in office referrals in a three-day window. In addition, over three days, a 1 μg/m3 increase in daily causes a 1.24% increase in teacher absences due to illness. Importantly, the effects are driven by low-income, Black, Hispanic, and younger students. Our research indicates that decreasing air pollution in urban areas could notably enhance both student and teacher attendance, minimize disruptive behavior in educational settings, and consequently, contribute to an improvement in students’ academic performance.
WORKS IN PROGRESS
Chung, S. The effects of air pollution on arrest: Evidence from violent crime, minor crime, and property crime (draft completed)
This study examines whether short-term exposure to elevated levels of ambient air pollution impacts police arrests, using daily individual-level arrest data collected by the New York Police Department from 2006 to 2019. Estimating the causal effects of social and environmental factors on crime is crucial for policymakers aiming to prevent violence and reduce the socioeconomic costs associated with crime. Air pollution is an understudied factor that may increase criminal opportunity and warrants exploration as an extension of crime prevention through environmental design. While most prior studies focus on criminal behavior, my research shifts the focus to arrests—a critical measure of police performance—due to offenses such as petty crimes, low-level drug dealing, public intoxication, and loitering, in addition to violent and property crimes.
Chung, S. & Ki, N. Who gets to breathe? Subgroup inequality and the limits of institutional protection in homeless health (draft completed)
This study investigates how environmental hazards generate both between-group and within-group health disparities among homeless populations and whether institutional performance mitigates these inequalities. Using panel data from 27 Continuum of Care (CoC) districts in New York State (2017–2019), we develop three disparity indices to capture inter- and intra-group variation in pollution-related hospitalizations across gender, age, and race. Applying Poisson fixed-effects models with HUD’s System Performance Measures (SPMs), we find that air pollution significantly increases hospitalization disparities across multiple subgroups, while higher-performing CoCs—characterized by stronger administrative capacity and cross-sector alignment—substantially reduce both between-group and within-group inequalities. These results suggest that CoCs function as institutional buffers, translating coordination and data capacity into tangible equity outcomes. By conceptualizing CoC governance as an equity infrastructure, this research advances public management theory on how administrative systems condition distributive outcomes under environmental stress, reframing equity in homelessness policy from equity by intent to equity by institutional function.
Xu, X. & Chung, S. The Effects of Policy Narratives – A Survey Experiment in Sustainability Policy (draft completed)
The Narrative Policy Framework (NPF) explores the role policy narratives play in policy process. While current literature examines how narratives can influence public’s support for policies, there is a lack of research on the effects of narratives on public’s overall perceptions of the government and service delivery. In this paper, we use a survey experiment with 800 U.S. respondents to examine how elements of the NPF, policy setting, characters, and moral, influence public perceptions and evaluations of local government in the context of sustainability policy. Specifically, we examine large bulky item recycling. Analysis suggests that prior agency performance, as the setting of changing the policy, has the strongest effects on public’s perceptions and evaluations of government performance. Moreover, who receives the burden and the moral of the policy also influence public’s ratings. As for public’s willingness to coproduce, only prior agency performance has positive effects.
Chung, S., Shen, R., & Curley, C. Scaling up or spreading thin? The role of water utility numbers in ensuring safe drinking water (in progress)
This study aims to investigate how water utility size influences hospitalization rates across multiple demographic groups in New York State. Utilizing a longitudinal panel dataset of hospital admissions from the Healthcare Cost and Utilization Project’s State Inpatient Databases, linked to ZIP code–level water system characteristics, we estimate the extent to which smaller utilities’ higher violation rates translate into increased hospital utilization for water-related illness. A central component of this research involves examining whether the frequency and severity of these violations mediate the relationship between water utility size and health outcomes. By employing the Poisson Pseudo-Maximum Likelihood with High-Dimensional Fixed Effects (PPMLHDFE) modeling approach, the analysis will control for time-invariant local factors and potential over-dispersion in hospitalization data.
Gao, X., Curley, C., & Chung, S. Understanding Energy Poverty: Exploring patterns and profiles of energy behaviors and consequences to refine energy poverty measures (preliminary findings gathered)
In collaboration with colleagues, we are investigating patterns of energy behaviors to refine measures of energy poverty. By analyzing household energy consumption and behaviors, we aim to inform policies that address energy insecurity and promote equitable access to energy resources.
Lee, J., Ragunathan, R., Evans, S., Law, A., …& Santos, H. Neighborhood opportunity and child externalizing behaviors: An ECHO consortium study (under review)