By Girish Kumar
Three books explain how — and what it would take to remember
In 1961, South Korea was poorer than Ghana. Its civil servants were underpaid, undertrained, and thoroughly corrupt. Its institutions were a shambles inherited from Japanese colonial administration — designed, like India’s, to control a subject population rather than serve a sovereign one.
What happened next is the part India’s policy class prefers not to dwell on. South Korea did not wait for better politicians or more foreign aid. It rebuilt its administrative state from the inside — new training academies, outcome-based evaluation, ruthless attention to what institutions were actually producing rather than what they were reporting. Within two generations, it had one of the most capable states in the world. Today its students outperform India’s on every global measure. Its citizens live, on average, a decade longer.
India, in the same period, wrote better laws.
This is not a counsel of despair. India’s democracy, its scale, its pluralism — these are genuine achievements that no technocratic shortcut could have produced. But they do not explain away a specific and stubborn failure: a state that spends roughly 14 percent of its GDP on education while producing learning outcomes that rank among the worst in Asia; that has built one of the world’s largest food security programs, covering 800 million people, while leaving half its children under five stunted; that has sent a spacecraft to the moon’s south pole but cannot reliably deliver a pension to a widow in rural Bihar without her losing a day’s wages to navigate the queue.
The problem is not money. It is not talent. India has both in remarkable abundance. The problem is that its institutions were built for a different century and a different kind of state — one designed to extract revenue, maintain order, and execute directives from above, not to deliver services, expand freedoms, or respond to citizens as agents with rights.
Three books — written across three decades, from three very different vantage points — describe, more precisely than anything else in the literature, exactly what India chose not to do. And what it still could.
The first is Amartya Sen’s Development as Freedom, published in 1999, which quietly detonated the old development consensus. Growth, Sen argued, is not the destination; it is merely one of the vehicles. Measure a society by its GDP per capita and you learn almost nothing about whether its citizens can read, vote, breathe clean air, escape poverty, or live without fear. The true measure is what people are actually able to do and to be.
His concept of capability — the real freedoms a person can exercise — shifts the administrator’s gaze from inputs to outcomes, from schemes to people, from disbursements to possibility. It demands that the state ask not “How much have we spent on the mid-day meal scheme?” but “Can the children who eat it actually read by class three?” It insists that a woman who is legally free to own property but practically unable to exercise that right because no court will hear her case is not, in any meaningful sense, free.
This reframing is not merely philosophical. It exposes the deep gap between India’s celebrated legislative architecture — the Right to Education, the National Rural Employment Guarantee, the National Food Security Act — and the lived reality of those the laws were meant to serve.
READ: Girish Kumar | The Full Plate: A powerful argument for grassroots change in India’s classrooms
Rights on paper that cannot be exercised are not rights; they are a kind of administrative theater. Sen gives policymakers a compass. He tells them what development is for. But a compass cannot build a road — and India’s roads have been notoriously hard to build.
That is where Abhijit Banerjee and Esther Duflo enter, with Poor Economics, issuing a challenge to the Indian planning tradition that has never quite been answered. Their book is a rebellion against grand theories. For decades, India’s development policy was driven by ideological certainty: first, that state-led planning would deliver growth; later, that liberalization and markets would do what the state had failed to do.
Both theories had confident adherents. Both produced genuine gains. And both, ultimately, left hundreds of millions of people behind — not because the theories were entirely wrong, but because they were applied as revealed doctrine rather than tested hypotheses.
Banerjee and Duflo show, through dozens of randomized controlled trials across South Asia and Africa, that the poor do not live inside ideological models. They live inside constraints: unreliable information, unpredictable incomes, weak institutions, and the daily calculus of impossible tradeoffs.
The question is never “markets or state?” The question is: within this specific context, with these specific people, what actually changes the outcome? The answers are frequently counterintuitive and always contextual.
In some settings, simply changing the timing of a conditional cash transfer increases school enrollment more than doubling the transfer amount. In others, providing textbooks — for decades the default solution to poor learning outcomes — has no measurable effect at all, because the children who receive them cannot yet read well enough to use them.
The lesson is not that textbooks are useless; it is that assuming they work, without checking, is a form of institutional negligence.
For India, where the evaluation of flagship programs has historically been either absent or captured by the agencies running those programs, this is not a methodological refinement. It is a cultural challenge. It asks the Indian state to treat its own certainties as hypotheses — and that is deeply uncomfortable for a bureaucratic tradition that runs on precedent and hierarchy.
Freedom, Banerjee and Duflo show, expands through systems that work, and systems work when they are built on evidence rather than ideology. But evidence, generated in pilots and labs, does not automatically become policy. For that, you need institutions capable of receiving evidence, acting on it, and sustaining the change across election cycles and transfer postings. For that, you need a theory of the state itself.
Satish Jha’s The Full Plate provides one. It enters where philosophy and empiricism meet the hard ground of governance, and it is here that the trilogy’s most practical and underappreciated argument lives. Jha does not write as a theorist observing India’s state from outside. He writes as someone who has watched, up close, how good intentions become bad systems: how a nutrition program becomes a procurement racket, how a school enrollment drive produces classrooms full of children who will never learn to read, how a technology deployment that works in the pilot collapses in the rollout because no one trained the functionaries, secured the infrastructure, or thought about what happens when the server goes down in a district with no internet.
His diagnosis is structural. India’s public institutions, he argues, suffer from three interlocking failures.
First, they are designed for compliance rather than capability — the incentive structure for a district official is overwhelmingly oriented toward avoiding mistakes, completing paperwork, and satisfying audit requirements, not toward whether children in her district are learning or whether the elderly are receiving their pensions.
Second, they treat citizens as administrative problems rather than as the point. The widow navigating the pension queue is not, from the system’s perspective, a person with rights and dignity; she is a file to be processed. This orientation — inherited from a colonial state designed to administer a subject population — is not merely unkind. It is operationally catastrophic. Systems that treat their users as obstacles generate friction, corruption, and exit; the citizens who can find alternatives do, and those who cannot are left with what remains.
Third, India’s institutions are fragmented in ways that make delivery structurally impossible. The child who arrives at school hungry cannot learn. The mother who cannot afford to be away from work cannot take her child to the vaccination camp. The farmer who cannot get clear title to his land cannot get a bank loan to buy better seeds. These problems are connected; the systems meant to address them aAre siloed. No single department is responsible for the whole person.
Jha’s framework — Capability, Dignity, Delivery — is both a diagnosis and a design principle. It asks institutions to measure themselves not by what they spend or report, but by what citizens are able to do after encountering them. It asks them to build for the user’s journey, not the bureaucracy’s convenience. And it asks them to use technology not as a substitute for institutional reform but as an enabler of it — a distinction that India’s enthusiasm for digital governance has too often ignored.
Together, these three books force a thesis into the open: India’s development failures are not primarily failures of resources or policy design. They are failures of institutional character. The state does not know what it is for, it does not know what works, and it is not organized to deliver even the things it does know work.
That indictment will meet an objection, and it should be heard. India has, in fact, built programs that work. The National Rural Employment Guarantee Scheme, for all its leakages, has transferred significant purchasing power to rural households. Jan Dhan brought hundreds of millions of unbanked citizens into the formal financial system. Direct Benefit Transfer has demonstrably reduced leakage in LPG subsidies. The state is not simply dysfunctional; it is capable of genuine achievement when it organizes around a clear, measurable goal with political will behind it.
The objection is fair — but it actually strengthens the trilogy’s argument rather than weakening it. Each of those successes shares a common feature: it was designed around a specific, measurable outcome, with feedback mechanisms, technology infrastructure, and political accountability aligned behind that outcome.
They succeeded, in other words, precisely because they embodied — even if accidentally — the principles Sen, Banerjee–Duflo, and Jha describe: clarity about what freedom you are trying to expand, evidence about what works, and institutions designed to deliver it.
The problem is that India has not been able to systematize this. Every successful program has been, to some degree, a bespoke achievement — dependent on a particular minister’s urgency, a particular collector’s competence, a particular moment of pol a itical will.
When the minister moves on or the collector is transferred, the system reverts. What India lacks is not the occasional capacity for good governance. It is the institutional culture of good governance — encoded in training, in incentives, in accountability structures, and in the norms that govern how administrators think about their work.
Consider what it would mean to build that culture deliberately. A young woman arrives at the Lal Bahadur Shastri National Academy of Administration after clearing one of the world’s most competitive examinations. She is brilliant, motivated, and almost entirely unprepared for what the job actually requires. She will spend her career making decisions that affect millions of people, in conditions of radical uncertainty, with inadequate data, inside institutions that were not designed to help her succeed.
She will be transferred every two to three years — a system meant to prevent corruption that also, reliably, prevents the accumulation of contextual knowledge and institutional memory. She will be evaluated on her ability to avoid scandal rather than her ability to produce outcomes.
Now imagine she encounters this trilogy before she takes her first posting. Sen teaches her to see the citizens she will serve as agents, not beneficiaries — people with capabilities to be expanded, not problems to be managed. The real question her work must answer is not “Was the scheme implemented?” but “What can the people in my district now do that they could not do before?”
Banerjee and Duflo teach her to be suspicious of her own certainties — to ask for the evidence, to design small experiments when it is absent, and to measure outcomes rather than outputs. Jha teaches her how to think about the institutions she will inherit and, eventually, lead: where the fragmentation is, why the incentives are misaligned, how technology can help and where it will fail, what it looks like to build a system that treats the citizen at the counter as the purpose of the enterprise rather than its inconvenience.
This is not a reading list. It is a formation — the construction of a professional identity organized around a different set of questions than the ones India’s administrative tradition has historically asked.
The reason it hasn’t happened is not intellectual. It is political. Teaching administrators to ask “What can citizens now do?” rather than “Was the file processed?” is threatening to systems organized around file processing.
Teaching them to treat their assumptions as hypotheses is threatening to hierarchies organized around the authority of certainty. Teaching them to design institutions that outlast individuals is threatening to a political culture that prefers dependence on individuals to the slower, less dramatic work of building systems.
The change India needs is not, ultimately, a pedagogical one. It is a change in what the state rewards. When the officer who improves learning outcomes in her district is promoted ahead of the one who processes more files, when the program that is rigorously evaluated is funded ahead of the one that is merely loudly announced, when the institution that treats the citizen with dignity is celebrated rather than the one that controls the citizen — then the ideas in these three books will have found their way into practice.
Until then, they are a diagnosis waiting for the political courage to become a prescription. India has the ambition. It has the talent. It has, in these three books, the intellectual tools. What it needs now is leaders willing to insist that the state be held to the standard these books describe.
South Korea figured that out in 1961. The question is whether India will wait another sixty years to try.