An $11-million bond, 200,000 children, and a 2.5x learning gain certified by an independent evaluator: the architect of the Quality Education India DIB explains why the money should move only after the result — and why the tool should retire once the discipline sticks.
44:04
Abha Thorat-Shah's argument starts blunt: giving money away without measuring the outcome is wasted money, because in development the only return is a social one — and an unmeasured social return doesn't exist. Her fix was not to lecture non-profits but to re-plumb the finance. The Quality Education India Development Impact Bond (2018–2022) raised $11 million, handed organisations like Kaivalya Education Foundation, SARD and Educational Initiatives unrestricted working capital, put a data partner inside their classrooms, and made one thing sacrosanct: learning outcomes, independently evaluated. Investors earned 6–8% only after 200,000 children measurably learned 2.5 times more than peers; over-performing non-profits earned an 8% bonus of their own. India's crisis is learning, not enrolment, so the cheapest fix sits in the primary grades — roughly ₹1,000 a child, layered on top of government budgets rather than replacing them. And once a bond proves the point, she argues, you retire it and mainstream the discipline.
Every block is a chapter, coloured by what it's about. Click any of it to jump straight to that minute on YouTube.
Normal development finance pays upfront: whether or not the work is delivered, the money is gone. The impact bond inverts this — an investor fronts the working capital, an independent evaluator measures the children, and the outcome funder repays the investor with a 6–8% return (capped at 8%) only if the impact is validated. Risk shifts to the investor; the funder buys a result, not an activity.
Non-profits normally run with their hands tied — the funder dictates line items ('here's money for textbooks, here's money for computers'), so even a great leader can't redirect spend mid-program. The QEI DIB gave operators unrestricted working capital and made only the outcome sacrosanct, with a data person sitting alongside flagging which 30% of the classroom was struggling in maths. Abha notes that freedom is 'pretty rare' in development finance, though private-equity people consider it obvious.
Right-to-education put a school within a kilometre of every child and enrolment near 90%, so building schools is not where philanthropy adds value. The crisis is that children fall two years behind in the primary grades, then either drop out or get passed upward until class-10 science and algebra are incomprehensible. Catching the problem at the root is the 'biggest bang for your buck' — the data shows children who gain from learning stay in education.
Clean causality data didn't exist because nobody pays cash-strapped non-profits to build data systems. So BAT selected on promise, aptitude and leadership culture — 'like investing in a bunch of entrepreneurs' — choosing Kaivalya Education Foundation, SARD and Educational Initiatives after cross-analysing funder data, including the Dell Foundation's. Each partner ran a different theory of change: sit with the child, upgrade the teacher, transform the principal, or focus on the bottom-20 students.
Keeping a child in school costs less than or close to ₹1,000 because the government already spends millions on schools, teachers and infrastructure. The philanthropist's job is the value-add on learning outcomes riding on that base — in Abha's image, moving the system from a Maruti 800 to an Audi rather than building the car. The unlock she wants next is budgetary flexibility, so a principal can fund a computer lab instead of a mandated playground.
Asked why the bond ran only four years, Abha says repeating impact bonds after the point is proven is unnecessary: the partners now raise outcomes-style money without BAT, and Educate Girls scaled to many more girls on its own. The instrument migrates to the next stuck problem — skills, secondary education, foundational literacy and numeracy — while simple grant-making absorbs the data-and-evidence discipline. 'Impact bonds are really critical when something is not working.'
Investors do get paid — put in £100, get back roughly £108 — but Abha insists the return is the carriage, not the carrot: what hooks them is the sharp focus on results. When their money is at risk, funders sit in and ask 'what's gone wrong?' in a way she never sees in grant-making. With a tsunami-like private-sector move toward social return and ESG, measured instruments are also the guard against impact washing.
Every figure below was said on air — timestamps included, caveats kept.
The impact report came after the program. The data from the impact report sat on a shelf and didn't have an influence on the way the program was run.
— Abha Thorat-Shah ▶ 9:53
The word 'system' is the entire car. It's not the engine, it's not the steering wheel — it's the whole machine working together.
— Abha Thorat-Shah ▶ 24:22
We had to look for promise and aptitude over data and analysis — and the ones who showed us that were the ones who were part of it.
— Abha Thorat-Shah ▶ 20:21
When people's money is at risk, they behave really differently. They really sit in there and say, 'What's gone wrong?' I don't see that elsewhere.
— Abha Thorat-Shah ▶ 40:57
The cleanest plain-English walkthrough of DIB plumbing: who fronts cash, who validates, and why the return is capped at 8%.
Abha's origin logic: if the impact report can't change the program, change the finance so data steers money in real time.
Why primary grades are the highest-leverage cheque in Indian education — and the Educate Girls pilot that set the template.
How BAT screened non-profits like a VC — promise over data — and backed four levers at once: child, teacher, principal, bottom-20.
India is one of the only countries where women are leaving the workforce; the new Skills Impact Bond pays for retention, not enrolment.
Auto-generated captions, lightly cleaned. Click a timestamp to open that moment on YouTube.