The Headmaster, the Chalk, and the Limits of Digital Public Finance
How long should it take a headmaster to buy a box of chalk?
In one Indian state, the answer runs something like this: a spending limit is set by the state government; the district enters it into the national financial management platform; the district informs the sub-district office—often over WhatsApp; that office informs the headmaster; the headmaster must then find a tax-registered vendor, an electronic invoice is raised, a form is filled, and the vendor is finally paid directly by the system. No cash ever touches the school. It is, in one sense, a marvel of fiscal engineering. It is also a lot of plumbing for a box of chalk.
I discussed this and more in the first episode of GovTech Talks, a new conversation series by MicroSave Consulting (MSC) on how technology is reshaping the way governments work. Our subject was public financial management, and the question at its heart is one every finance ministry now faces: what can digital reforms actually deliver, and what can they not do on their own?
The achievement is real. India's reform arc over fifteen years is among the most ambitious anywhere. What began in 2009 as an online fund-tracking platform grew into a national public financial management system—the backbone of a direct benefit transfer architecture that now moves money straight into the bank accounts of hundreds of millions of citizens.
Reforms from 2021 consolidated a bewildering sprawl of programme bank accounts—in a single state, from roughly 1.26 million scheme accounts to about 3,100. And the latest step, a just-in-time release system built on the central bank's payments infrastructure, is shifting the whole architecture from push to pull: money is released the moment a real transaction occurs, rather than parked as advances that sit idle in commercial banks while the government borrows at interest.
The Finance Ministry has cited interest cost savings of around $2.6 billion from these debt management changes. The underlying inefficiency here is an old and familiar one: governments borrowing at market rates to cover apparent shortages while their own money sits idle, unremunerated, in thousands of scattered bank accounts. What is new is the ambition to fix it with modern payment rails—precisely the digital-first redesign of treasury functions that the IMF's Digital Solutions Guidelines for Public Financial Management (2023) now recommends to all countries, and part of why India features so prominently in the Fund's work on transforming public finance through GovTech and in the World Bank's GovTech Maturity Index.
But digital systems don't sit in a vacuum. They interact with—and are deeply shaped by—administrative capacity, infrastructure, political incentives, and the lived realities of the people who navigate them. Three lessons from our research at the Foundation for Responsive Governance (ResGov) will, I suspect, sound familiar well beyond India.
First, street-level bureaucracy has not been eliminated; it has been transformed. Michael Lipsky's classic account of street-level bureaucrats argued that frontline workers' discretion shapes what policy actually is. Digitisation was supposed to change that. Instead, it has created a new frontline: the data entry operators who staff village council offices and are, quite literally, the human face of the digital state. In a study across five states of India, we found they receive almost no training—operators told us roughly 15 per cent of what they know came from formal instruction and 85 per cent from the job, often via informal messaging groups and screen-sharing apps, with all the data-security questions that raises. They answer to multiple bosses, and discretion persists in new forms: borrowed supervisor logins, judgement calls on who qualifies for a benefit. A sweeping digital transformation is resting on a strikingly weak human foundation of contractual, low-paid, largely invisible workers. The global evidence points the same way: as the World Bank's World Development Report 2016 put it, technology can genuinely improve how states deliver—but the dividends materialise only where the "analog complements" of skills, institutions, and accountability are in place.
Second, uniform design meets uneven context. We compared how the same centralised reform played out in two districts of the same state—one relatively better off, the other among the poorest in the country. In the poorer district, weaker internet connectivity, fewer registered vendors, and thinner administrative capacity produced far more friction on the ground. This is the "one-size-fits-all" trap that scholars of institutional reform have warned about for years—what the Building State Capability literature calls the gap between the form of a reform and its function. Buying chalk cannot carry the same process burden as building a school; reform design has to build in flexibility and calibrated discretion, with adequate controls, rather than assume the preconditions exist everywhere.
Third, generating data is not the same as using it. Digitisation has produced extraordinary visibility—an official sitting in the capital can now see the bank balance of a village childcare centre. But most of what gets used is aggregate: savings achieved, transactions processed, utilisation rates. There is limited institutional incentive to ask the uncomfortable questions the data could answer. Which communities and regions are benefiting less? What happens to a school that receives its annual funds with four months of the year left? Who is being quietly excluded—like the women who change their surname on marriage and find their identity records no longer match, blocking the maternity benefits they are entitled to? That is not a technology failure; it is a failure to design for the person the system is meant to serve. As the UN's report on digital welfare states reminds us, those who bear the cost of system failures are precisely those least able to opt out.
None of this is an argument against digital reform—the progress is genuine and worth defending. It is an argument, as researchers at ODI's Digital Public Finance Hub and the World Bank's Reimagining Public Finance discussions have also made, that digitalisation is not a substitute for governance. Wherever a country sits on the digital maturity curve, the next phase of reform looks similar: invest in human and institutional infrastructure with the same seriousness as digital infrastructure; design for contextual variation rather than assume it away; and build the incentives to actually use the data now being generated.
India built the pipes faster than almost anyone. Whether money, information, and authority flow through them to a headmaster who just needs chalk—that is the question public financial management exists to answer, in every country. Looking beyond the dashboard, is like looking beyond budget day, is not optional; it is essential.
This post draws on my conversation in Episode 1 of MSC's GovTech Talks series.
