Experimental Evidence on Subsidized Transportation for Girls in Remote Contexts

How to cite

Amaral, Sofia; Vyborny, Kate; Gaddis, Isis; Srinivasan, Shyam. 2025. "Experimental Evidence on Subsidized Transportation for Girls in Remote Contexts." AEA RCT Registry. AEARCTR-0014355. https://doi.org/10.1257/rct.14355-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Sofia Amaral* , Kate Vyborny , Isis Gaddis , Shyam Srinivasan

* Primary investigator

Registry ID
AEARCTR-0014355
DOI
https://doi.org/10.1257/rct.14355-1.0
First registered
Last updated
Status
On going
Trial dates
2024-07-15 – 2026-09-30
Countries
Pakistan (Khyber-Pakhtunkhwa)
JEL codes
J16, O18, R41
Registry record
http://www.socialscienceregistry.org/trials/14355

Abstract

This study evaluates a large-scale intervention that subsidizes safe school transport for girls in rural Khyber Pakhtunkhwa (KP), Pakistan, a region where steep terrain, long distances to school, and entrenched gender norms result in one of the widest gender gaps in education in the country. Using a cluster-randomized controlled trial across schools in 7 districts of KP - of which 168 school clusters were selected for phase 1 of the study, and an additional 203 school clusters to a phase 2-, the evaluation tests the impact of subsidized, girls-only pick and drop transport services on school enrollment, attendance, and learning outcomes among middle-school–age girls. The intervention targets both in-school and out-of-school girls who reside roughly more than 1.6 km from the nearest government middle school. Transport is organized through local parent-teacher councils (PTCs), with private drivers vetted for safety. To understand mechanisms we cross-randomize among treated cluster a "safety salience" treatment designed to highlight risks and mitigation strategies to drivers. The research aims to disentangle three potential constraints to girls’ schooling—affordability, accessibility, and safety—and to explore how social norms shape take-up and impact. We also assess spillovers across siblings. This study contributes to a growing literature on gender, mobility, and education by testing a scalable, government-led transport model in a high-need, low-accessibility setting. Findings will inform program expansion and broader education and gender policy in Pakistan and similar contexts.

Intervention

The main intervention to be evaluated consists of the provision of subsidized transport to girls. The subsidized transportation will consist of the recruitment of privately run vans with drivers vetted by the Government and committed to ensuring safe transportation for girls. The level of subsidy, as a fraction of the unit cost of providing the transport, will be determined after the baseline survey, based on the respondents’ (parents of girls eligible to attend middle school) reported willingness to pay for transport. Girls will travel in vans designated only for girls’ school transport on a subscription basis. The intervention will therefore address barriers to girls’ education in three dimensions: i) affordability by alleviating the economic burden of traveling to school for girls from remote areas, ii) accessibility, by ensuring girls have a reliable mode of transportation available, and ii) safety by ensuring only vetted drivers and other female passengers are able to travel in these vans. The intervention will be delivered and managed locally by parent-teacher councils (PTCs). To better understand the role of safety, we cross-randomize a safety salience nudge to drivers. Drivers will received regular calls to make salient the role of importance girls' safety and of adequate conduct of service providers.

Experimental design

To evaluate causal effects of the intervention we design a cluster-randomized control trial across school clusters. In school clusters randomly assigned to the treatment group, eligible girls are offered subsidized transport in designated vans. In school clusters randomly assigned to the control group, eligible girls will not be offered subsidized transport. The entire catchment area of one middle school is the cluster. In phase 1 of the study (academic year of 2024-2025) the study sample consists of 168 school clusters of which about 86 clusters will be randomly allocated to the treatment group. Within each treated school cluster, all girls eligible to attend middle school, who live at least 1.6 km by road from the middle school, will be offered subsidized transport. This includes girls who are currently enrolled in the middle school, girls who were enrolled in grade 5 of primary school during the previous academic year and are eligible to transition to middle school for the 2024-25 academic year, and girls who have dropped out but could potentially be enrolled back into the education system at the middle school level, if their mobility barriers are addressed. It is worth noticing that in this context there aren’t mixed-gender schools and thus, the project is only focusing on studying the effects of the program on girls. In phase 2 of the trial, the pilot will be extended to additional 3 districts. In phase 2 (academic year of 2025-2026) 203 school clusters were sampled to take part in the study of which 102 are part of the treatment group. We further create experimental variation in a safety salience treatment designed to make safety concerns salient to drivers (cross-randomized at the level of treated school clusters).

Design details

Randomization method
Randomization done in office by a computer using Stata.
Randomization unit
School cluster.
Clusters
371 school clusters.
Observations
The final sample will be finalized depending on budget restrictions and negotiations with the implementing partner.
Arms
Treated: 188 -- 86 school clusters in phase 1 & 102 school clusters in phase 2 Control: 183 -- 82 school clusters in phase 1 & 101 school clusters in phase 2
Primary outcomes
1- School Enrollment (binary); 2- Grade Completion (binary); 3- Attendance Rate (Continuous).
Secondary outcomes
1- Learning; 2- Commute experiences (we will measure daily commute time, mode of transport, safety and overall satisfaction with commute); 3- Parents’ perceptions of safety and commute quality; 4- School expenditures (including transport expenditures); 5- Gender attitudes, aspirations and social norms; 6- Mobility.
Partners
Name: Government of Khyber Pakhtunkhwa Partner_type: government

Cited by

No citation data available for this registration yet.

Keywords