The Unintended Consequences of High-Intensity Streetlights on Student Outcomes: Evidence from a Crime Deterrence Experiment

• In Event: Neighborhood Contexts and Educational Opportunity: Exploring Local Determinants of Student Success

Abstract

In 2016, New York City launched a randomized controlled trial (RCT) that installed high-intensity streetlights across 80 high-crime public housing developments. While prior research has documented significant crime reductions from this intervention (Chalfin et al., 2022; Mitre et al., 2022), whether there were unintended consequences on residents - particularly children - remain unknown. Critics warn that constant, glare‑level lighting can disrupt sleep and circadian rhythms, undermining daytime cognition and increasing obesity risk (American Medical Association, 2016). Further, residents - specially those living on lower floors-  report sleeplessness, heightened stress, and a sense that the lights serve as an intrusive policing tool, generating feelings of surveillance rather than safety (Bittle, J. & Kraven, J., 2018, Loeffler et al., 2022).

This paper examines how exposure to extremely bright lights (~600K lumens per tower, while reportedly the Yankee Stadium’s lights are of 150K lumens) affects the academic, behavioral, and mobility outcomes of K-12 students living in treated developments. Leveraging student-level longitudinal administrative data from the NYC Department of Education (2012–2019), we link ~355,000 student-year observations to detailed residential histories and the experimental assignment of lights. The experiment randomized both treatment assignment and dosage across 40 matched development pairs. We begin by assessing covariate balance at the development level and estimating intent-to-treat effects dosage-response relationships using simple differences in means for the post-treatment period. We then exploit this design to estimate difference-in-differences and event study models with development-pair and student fixed effects. These strategies address imbalances in baseline student characteristics, enhancing the internal validity of our findings. We also examine heterogeneity by building height, race, and grade level.

Preliminary results suggest that treatment assignment -receiving high-intensity lights- leads to modest improvements in standardized math scores that grow over time, as well as short-term reductions in grade retention. We find no significant effects on other academic or residential mobility outcomes. However, when we examine treatment intensity, defined as living in a building within a treated development that was assigned more than six additional lights, we observe that while higher light exposure reduces nighttime index crime, it decreases students’ academic achievement and increases the probability of exiting public housing, and BMI. These results are driven by students that live in low-rise buildings, who are more directly exposed to the light towers.

These findings suggest that the negative effects of high light exposure within a student’s own building may offset potential academic benefits associated with reduced crime. This raises policy-relevant questions about how to weigh public safety objectives against educational impacts in the context of environmental design interventions in disadvantaged communities.

Author

• Mayra Samantha Cervantes, Syracuse University
  Presenting Author