Analyzing disparities in app-hailed travel during extreme heat in New York City

• Ridership increased by six to nine percent compared to normal-weather days. • Fixed effects were used to analyze variation across low- and high-income areas. • Afternoon peak higher by 1.7 to 3.2 rides per 1,000 people in median income areas. • Ridership increased by three percent for every $10,000 increase in income. • Baseline per cap trips were lower in low-income areas compared to high-income areas. To understand extreme weather effects on travel behavior, we examine changes in New York City app-hailed travel during extreme-heat days and estimate whether such changes vary across low- and high-income neighborhoods. We use trip records for July 2019, when several heat-related messages were issued by the National Weather Services, and find that daily ridership was six to nine percent higher on days when heat messages were issued than on matching weekdays in the same month. Our fixed effects regression models estimate that for median income neighborhoods afternoon peak-hour ridership was 1.7 to 3.2 rides higher per 1,000 people during the heat events than on matching weekdays in the same month. This effect increased by an additional 0.5 to 0.6 rides per 1,000 people for every $10,000 increase in average neighborhood per capita income, suggesting that higher income travelers increased trip frequency at a higher rate than lower-income travelers.