Effects of new transit lines on commuting: Evidence from restricted-use Census Bureau microdata

Abstract

Urban planners face challenges in justifying the behavioral impacts of public transit investments with causal evidence. In this study, we take a cost-effective approach to data collection by utilizing restricted-use Census Bureau microdata that provide rich individual characteristics and fine-geographical-resolution block information, and then construct a natural experiment. We evaluate the impacts of new rail transit lines on commuting behavior in Charlotte (NC), Dallas (TX), and Los Angeles (CA). We find that new transit lines lead to 10%, 6%, and insignificant increases in public transit ridership for Charlotte, Dallas, and Los Angeles, respectively, signaling potentially diminishing marginal returns of public transit investments. Also, we find no significant effects on commuting time across the three cities. This research demonstrates the potential of using Census Bureau microdata to assess the causal impact of urban infrastructure.