Do Power Outages Impact Mental Health? Evidence from Maryland

• In Event: Infrastructure, Instability, and Health: Local Determinants of Community Well-Being

Abstract

Power outages are becoming more frequent in the United States due to aging grid infrastructure and increasing climate stress, yet their consequences for mental health remain poorly understood. Most existing studies rely on self-reported psychological distress following large-scale disaster events, leaving a gap in evidence on how routine, recurrent outages affect clinically diagnosed mental health outcomes. This study provides the first empirical evidence linking power outage exposure to mental health hospitalizations using high-resolution outage data and patient-level clinical records.

We combine 15-minute-interval power outage records obtained from the State of Maryland with administrative inpatient claims from the Maryland Healthcare Cost and Utilization Project (HCUP) State Inpatient Database, covering 2018 to 2023. Maryland is a particularly relevant setting given its mental health service utilization rate of 42.2 per 1,000 population—nearly twice the national average—and its aging electricity infrastructure. We aggregate data at the ZIP code–quarter level and define severe outage exposure as the count of episodes exceeding the 75th percentile of outage intensity and lasting at least one hour. We estimate Poisson fixed-effects models with ZIP code and quarter fixed effects, controlling for weather, socioeconomic characteristics, and healthcare access. To address potential endogeneity, we employ instrumental variable (IV) models using three grid infrastructure measures—transformer density, substation density, and transmission line density—as instruments. Because these instruments are time-invariant, the IV specification relies on cross-sectional identifying variation, controlling for time trends through year-quarter fixed effects and for ZIP-level confounders through a rich set of covariates.

Higher outage exposure is consistently associated with increased mental health hospitalization rates. The baseline Poisson model yields an incidence rate ratio (IRR) of approximately 1.017, indicating a 1.7% increase per additional severe outage episode. This association is robust to alternative severity cutoffs (50th and 90th percentiles), with effects concentrated in moderate-duration disruptions (1–3 hours) rather than prolonged outages. IV estimates indicate substantially larger causal effects (IRR ≈ 1.35–1.52), suggesting that conventional models understate the true impact. Heterogeneity analyses show that effects are statistically significant in large metropolitan areas (IRR = 1.016, p < 0.05), in lower-poverty ZIP codes, and in areas with higher Medicaid coverage.

These findings carry implications beyond Maryland. Globally, mental disorders affected an estimated 970 million individuals in 2019 and remained among the top ten leading causes of disease burden. As aging infrastructure and increasingly frequent climate-related extremes place energy systems worldwide at growing risk of outages, the mental health consequences documented here are likely to extend well beyond any single region. Climate change, the electrification of transportation and heating, and growing electricity demand from data centers will further strain grid infrastructure. Our results demonstrate that grid unreliability directly affects population health—an overlooked dimension of the societal impacts of energy systems—and underscore the need to integrate health considerations into reliability planning, prioritize energy access, energy equity, and energy justice for vulnerable communities, and strengthen continuity of mental health care during disruptions.

Author

• Yueming (Lucy) Lucy Qiu, University of Maryland, College Park
  Presenting Author