Public Transportation Networks and Urban Heat

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

Abstract

Traffic congestion is a plausible contributor to localized urban heat, a costly externality for the many who live and work near major roads. Yet its short-run causal effect on local surface temperature has not been credibly identified: existing estimates are associational or rely on broad natural experiments. We estimate it for California freeways using unplanned accidents as an instrument for congestion. The central threat is reverse causality: because congestion raises accident risk, accidents are in general endogenous to the congestion they are meant to instrument. We document this bias with an event study and remove it by restricting the instrument to accidents that strike while the segment is free-flowing, before congestion has formed. We measure land surface temperature from NASA ECOSTRESS at 70-meter resolution and exploit its varying overpass time to match each reading to same-hour congestion. Our central estimate is the reduced form: a free-flow-onset accident raises near-road surface temperature by about 0.08°C over the following two hours (p = 0.06), against flat pre-trends. The increase is about two-fifths of the 0.21°C by which the near-road ring is on average warmer than the 500–1,000-meter outer ring. Scaled by the accident’s effect on congestion, the reduced form implies roughly 0.4 to 0.8°C per 10 mph of speed reduction. The response is largest in the near-road ring and declines significantly with distance, tracing how a localized transport externality fades into the surrounding area.

Author

• Sierra Arnold, Xavier University
  Presenting Author