Abstract
Short summary of the published abstract. A randomized field experiment with 1,607 frequent public transport users in Bogotá tests vouchers loaded onto travel cards for four months. Vouchers increase ridership, particularly on weekdays and at peak hours, and reduce the burden of public transport expenditures. The study evaluates voucher delivery as an alternative way to provide demand-side subsidies.
Empirical strategy
The experiment randomly offered monthly vouchers of COP 28,000 (A) or COP 21,000 (B), alongside a control group. Administrative travel-card records measure boardings before, during, and after the intervention. Person and week fixed effects absorb stable individual differences and common shocks. The study reports both intention-to-treat (ITT) estimates and effects among voucher activators (ATT), using assignment as an instrument for activation. Affordability is measured using SITP spending relative to income; this index excludes other transport modes.
Main equation
Equation (1), written with the two voucher arms explicitly, is:
\[ Y_{it}=\gamma_i+\lambda_t+ \sum_{g\in\{A,B\}}\left(\beta_g T^g_{it}+\tau_g P^g_{it}\right)+\epsilon_{it}. \]
\(Y_{it}\) is boardings, \(T^g_{it}\) identifies assignment to arm \(g\) during the intervention, and \(P^g_{it}\) identifies its post-intervention period. \(\beta_g\) is the ITT effect; standard errors are clustered by participant.
Selected results
Ridership
| Voucher group | ITT: all days | ATT: all days | ATT: weekdays | ATT: weekends |
|---|---|---|---|---|
| A: COP 28,000 | 7.22% | 8.73% | 9.58% | 4.13% |
| B: COP 21,000 | 6.50% | 8.04% | 8.98% | 4.15% |
Affordability
| Group | Before | During | After |
|---|---|---|---|
| Control | 19.1% | 20.5% | 23.6% |
| A: COP 28,000 | 20.5% | 16.3% | 24.3% |
| B: COP 21,000 | 21.2% | 18.0% | 26.2% |
- Usage: activators make about 8–9% more trips; gains concentrate on weekdays and peak periods.
- Affordability: the burden falls during voucher delivery, with estimated monthly savings of about USD 4.10.
- Persistence: post-treatment ridership differences are not statistically significant. The evidence concerns frequent users in this experiment rather than a citywide rollout.
Citation
@article{GuzmanGomezCardonaOchoa:2024,
title = {Increased ridership and improved affordability: Transforming public transport subsidies in developing contexts},
author = {Guzman, Luis A. and Gómez Cardona, Santiago and Ochoa, Jorge Luis},
doi = {10.1016/j.retrec.2024.101492},
journal = {Research in Transportation Economics},
year = {2024},
volume = {108},
pages = {101492}
}