On this page
In 2006 the Government of Bihar started giving every girl who enrolled in Class 9 money to buy a bicycle. Other states copied it. But Bihar was also growing fast, so rising enrolment alone could not prove the bicycles worked. Economists Karthik Muralidharan and Nishith Prakash tested it in their paper Cycling to School, published in the American Economic Journal: Applied Economics (2017).
What was Bihar's bicycle scheme for girls?
The scheme, which the paper calls the Cycle program, paid Rs 2,000 per girl (about $45 in 2006) to buy a bicycle, according to the authors' final version of the paper.pdf). The money was handed out at public school ceremonies, and principals collected receipts. In 2007-08 the state spent Rs 310 million on bicycles for 160,000 girls.
The scheme also reached the girls it was meant for. A separate implementation study by Ghatak and colleagues, which the authors cite, found that only 3% of eligible households did not receive the money, and only 2% of those who received it did not buy a bicycle.
How did the researchers test whether the bicycles worked?
There was no lottery: every eligible girl got the money. So the authors built a comparison from survey data (same paper.pdf)):
Data: the District Level Health Survey (DLHS-3) of 2007-08, about 18 months after the scheme began. Their main estimate uses about 30,000 children aged 14 to 17 in Bihar and Jharkhand.
Treated vs untreated: girls aged 14-15 were deciding about Class 9 after the scheme began; girls aged 16-17 had already decided before it.
Two comparison groups: boys of the same ages, and neighbouring Jharkhand, part of Bihar until 2001, which had no such scheme.
This triple difference is designed to strip out changes that hit everyone, such as economic growth or new school spending. The authors note it cannot rule out every change that happened at the same time.
What did the study find?
Being in a cohort exposed to the scheme raised girls' age-appropriate secondary enrolment by 5.2 percentage points, from a base of 16.3%, which is a 32% increase (Muralidharan and Prakash 2017).
It closed 40% of the 13-point gender gap in age-appropriate secondary enrolment.
Distance mattered most. In villages 3 km or more from a secondary school, girls' enrolment rose by 8.75 percentage points (87%) and the gender gap fell by 54%; in villages closer than 3 km, the effect was close to zero (authors' final version.pdf)). The authors read this pattern as evidence that the main mechanism was the bicycle cutting the time and safety cost of the journey, though they cannot separate it from other parts of the scheme, such as the public ceremonies.
School records told a similar story. The number of girls appearing for the Class 10 board exam rose by 18%, and the number passing rose by 12%. At under $1 a month per girl, it raised enrolment more cheaply than a comparable cash-transfer programme in Pakistan, the authors estimate.
What does the study not show?
It is not a randomised trial. The design is careful, but it rests on assumptions, such as Jharkhand being a fair comparison.
It is one snapshot. The survey measures on-time enrolment in Class 9, not how long girls stayed.
Exam gains were smaller. Only about 37% of the extra enrolment showed up as extra board exam passes, and the authors say education quality needs attention too.
Empowerment was not measured. Effects on confidence, marriage age or work are discussed but not estimated.
A later randomised trial in Zambia, Wheels of Change by Fiala, Garcia-Hernandez, Narula and Prakash (IZA discussion paper, 2022), gave bicycles to school-going girls living more than 3 km from school. Nishith Prakash co-authored both studies, so it is not a fully independent replication. After one year, commuting time fell by 35%, late arrivals by 66% and absenteeism by 27%. Different country, same direction.
What can students and families take from it?
This part is our inference, not a finding of the study. The study covers Bihar in 2006-08, so its numbers may not carry over elsewhere; the broader point is that a long, tiring or unsafe journey can quietly decide whether a student keeps going. If you are in Class 9-12 or preparing for JEE or NEET, the journey is part of your study plan.
Time the commute. Note door-to-door minutes each way for a week.
Ask what it costs you. Tiredness, missed first periods and safety worries all count.
Look for a shorter or safer option, such as cycling with classmates.
Use the time you win back for one short, focused revision block.
If you are unsure what to do with that block, Fermi's AI tutor can work out what you already know and build a plan for the session, so the saved time has a clear first step.
For more on rural schooling, read what ASER 2023 found about 14-18 year olds, what UDISE+ 2025-26 data shows about India's schools and how Teaching at the Right Level fared in Indian trials.



