A causal evaluation of Bogota's cable car illustrates the transformative potential of mobile phone data for policy analysis

TL;DR

This study demonstrates how high-resolution mobile phone GPS data can be used for causal evaluation of transport policies, revealing impacts on mobility and socioeconomic interactions in Bogota's cable car project.

Contribution

It introduces a novel method using mobile phone data for causal impact evaluation of urban transport policies, especially in low-data environments.

Findings

Cable car increased mobility by 6.5 trips per person per month

Most trips were within local neighborhoods and to the city center

Limited evidence of increased socioeconomic mixing

Abstract

Transport infrastructure is vital to the functioning of cities. However, assessing the impact of transport policies on urban mobility and behaviour is often costly and time-consuming, particularly in low-data environments. We demonstrate how GPS location data derived from smartphones, available at high spatial granularity and in near real time, can be used to conduct causal impact evaluation, capturing broad mobility and interaction patterns beyond the scope of traditional sources such as surveys or administrative data. We illustrate this approach by assessing the impact of a 2018 cable car system connecting a peripheral low-income neighbourhood in Bogota to the bus rapid transit (BRT) system. Using a difference-in-differences event study design, we compare people living near the new cable car line to people living in similar areas near planned stations of a future line. We find that the cable car increased mobility by approximately 6.5 trips per person per month, with most trips within the local neighbourhood and to the city centre. However, we find limited evidence of increased encounters between the low income cable car residents and other socioeconomic groups, suggesting that while the cable car improved access to urban amenities and quality of life, its impact on everyday socioeconomic mixing was more modest. Our study highlights the potential of mobile phone data to capture previously hard-to-measure outcomes of transport policies, such as socioeconomic mixing.