Extreme weather and Student Aggression: Large-Scale Causal Evidence on Climate-Driven School Discipline Disparities

• In Event: The effects of environmental risks on children

Abstract

Numerous studies show a strong positive relationship between temperature and violence, suggesting extreme heat could be associated with increased aggression among students at school. Yet very few studies have examined the relationship between extreme weather events, specifically extreme heat and cold, and school disciplines, which are strongly correlated with worse academic outcomes. In this study, we examine how extreme weather events affect office discipline referrals (ODRs) and punitive school disciplinary incidents (e.g., suspensions/expulsions) by linking day-level weather data with day-level, school disciplinary infraction data from the three large, western states (Washington, Oregon, and California) between 2013 and 2023.

Our day-level, school discipline data comes from the over 2,700 schools that use the School-Wide Information System (SWIS) data system which includes administrative data from all schools implementing Positive Behavioral Interventions and Supports (PBIS), a multi-tiered behavioral management system in the three states. The dataset captures minor behavioral incidents that typically go unrecorded in official administrative databases, allowing us to peer into a variety of school discipline outcomes and potential mechanisms linking heat and aggression in schools. Our sample includes > 1.5 million school-day observations. We obtain the daily area average of maximum temperature from the U.S. National Oceanic and Atmospheric Administration’s (NOAA) National Centers for Environmental Information using the nClimGrid-Daily product. Our data are at the daily, census-tract level. These data represent one of the largest datasets ever assembled on day-level school disciplines with corresponding temperature variation. To evaluate the effects of extreme weather on ODRs, we use a natural experiment, leveraging the fact that extreme weather is plausibly exogenous to other factors that affect ODRs since the location and timing of these weather events are based on atmospheric conditions.

Our preliminary results show an additional day when and where daily maximum temperature exceeds 90oF in a given census tract is associated with an increase of 0.45 total ODRs, which represents 4.6% of mean daily ODRs in a given tract. However, an additional very cold day when daily maximum temperature is < 40oF is associated with a reduction in ODRs of 0.79. The latter result is likely related to lower overall attendance on snow days. We further find differential effects based on type of disciplinary incident. Punitive disciplinary incidents, which are disciplinary infractions severe enough to warrant a punishment (e.g., a suspension), also increase by 0.019 to 0.039 on days when and where temperature exceeds 80oF. This represents 2.8 to 5.7% of mean punitive disciplinary incidents in a given tract. We conduct several robustness checks, including tests for selection bias, comparing results with different sets of controls, and the inclusion of cross-sectional air conditioner data. Our results show that as climate change is expected to increase the frequency of extreme weather events over time, finding ways to mitigate its impacts on the U.S. educational system will be critical in ensuring ODRs and their associated academic outcomes do not worsen.

Author

• Nicole S Ngo, University of Oregon
  Presenting Author