A few days ago, my mother heard that her cousin had died at ninety-two. It was the kind of death that draws people together rather than apart. The cousin had been, by every account, a woman whose open smile felt like an invitation to interact, whose humour made the people around her feel that life, for all its complications, was not a bad thing at all. Her final act was to donate her body to medical research.
My mother, who spent her working life in social work, was moved. She asked me a string of practical questions: did she need to sign a form; which institution would accept the donation; could she also pledge her corneas to an eye bank; and, quietly underneath all of it, did I, as her son, accept this decision. I said yes without hesitation. So did my sister.
What struck me afterwards was not the decision itself but its timing. My mother’s elder brother and his wife had made an identical pledge years earlier. She had known about it, presumably discussed it, and had done nothing. It took a cousin’s death, someone whose warmth and humour she admired, rather than a sibling she simply loved, to finally move her into action. I am aware that one family’s turn of heart proves nothing about a national system. What it can do is sharpen the question worth asking of one, which is what the rest of this essay tries to do.
My mother’s decision carries a specific irony, because she is not simply a religious woman bucking at her own tradition’s view on the matter. In the 1980s, she authored a report for Sankara Nethralaya that documented, at some length, the strength of religious objection in India to disposing of or altering the body outside practices prescribed by scripture. She had established exactly how widely held that resistance was. That she arrived, decades later, at the opposite choice for her own body is not evidence that the objection she once documented was thin or overstated. It is evidence of something more specific: that even a genuine, scripturally grounded reluctance, one she had personally confirmed was real, can be moved not by better information but by a particular kind of encounter.
What her cousin’s death gave her was something closer to permission modelled by someone she loved, at a moment she was paying attention, to set a known and real objection aside. That distinction, between an objection someone actually holds and one they’ve simply inherited, matters a great deal for how India’s donation numbers should actually be read.
I went looking for the practicalities on her behalf and found that several government medical colleges accept body donations, and a number of eye banks take corneas; the PROs at Madras Medical College and at Sankara Nethralaya were reasonably prompt in furnishing the necessary forms and information. What surprised me more was a comparison one of them mentioned in passing: Sri Lanka, a country with a tiny fraction of India’s population, retrieves more corneas annually than India does; more than one person running an eye bank told me as much, unprompted, as though it barely needed saying. My first instinct was to read this as a verdict on generosity: that Indians are, in some diffuse way, less willing to give of themselves. I don’t think that instinct survives scrutiny, and the reason it doesn’t tells us something about a much larger pattern in how public systems succeed and fail.
India’s last few decades present a genuine puzzle. In some domains the country has closed the gap between policy and citizen with real speed; telecom penetration reaching villages that never had a landline, rural road-building connecting places once effectively cut off, digital payments and Aadhaar enrolment and subsidised cooking-gas connections under the Ujjwala scheme reaching hundreds of millions of people within a few years. And yet outcomes that depend on a conversation rather than a connection remain resistant to improvement, even where funding is not obviously the binding constraint. Organ and cornea donation, tuberculosis treatment adherence, and immunisation follow-through in remote districts are the clearest examples.
The distinction I want to draw is between two categories of delivery that look similar from a distance but fail and succeed for almost opposite reasons. The first is transactional infrastructure: a road, a SIM card, a gas cylinder, a digital payment. These tend to succeed because the last mile is one more standardised unit in a repeatable process, closed by someone with a direct stake in closing it; an agent paid per activation, a contractor paid per kilometre, a government claiming credit for a countable achievement. The incentive sits at the point of delivery, and the transaction, once designed, repeats at scale without needing to be reinvented for each citizen.
The second category has a name in public administration: the American political scientist Michael Lipsky called it street-level bureaucracy in his 1980 study of the same title. The class of public service that depends on the discretion, judgement, and motivation of the frontline worker actually facing the citizen: a teacher, a police officer, a hospital counsellor. Later scholars extended the framework to argue that these workers don’t just implement policy, they effectively make it, through the thousands of small judgement calls a rulebook can never fully specify.
Cornea retrieval sits squarely in this category. It is not a standardised transaction but an emotionally fraught, time-bound, individually negotiated conversation with a grieving family, conducted by an underpaid or entirely absent hospital counsellor with no direct stake in whether the conversation succeeds. Nobody is paid per conversion. Compare this with Aadhaar enrolment, which had a clear administrative target and an incentivised agent for every completion, and the outcome gap stops looking mysterious.
Reports on Sri Lanka’s cornea programme suggest its success rests less on the generosity of individual citizens than on a deliberately built hospital retrieval system of trained counsellors approaching families within the narrow window during which corneas remain viable, backed by decades of consistent public messaging rooted in the Buddhist ethic of giving. It is closer to proof that a culturally resonant story, delivered by a properly resourced human being at the right moment, can accomplish what no single awareness campaign could.
The framework is tidy enough to be suspicious of, and it’s worth testing it against a case that resists it. Ujjwala, the Indian scheme providing subsidised cooking-gas connections to poor households, is usually cited as a transactional-infrastructure success with millions of connections issued quickly. But it is an open enough secret in the sector, one even government reviews don’t dispute, that a large share of beneficiary households, having received the connection, do not refill the cylinder and revert to biomass fuel. The infrastructure delivered; the behaviour it was meant to produce did not follow.
That failure is not really an infrastructure problem, it’s a street-level one, dressed as infrastructure. Nobody was tasked, paid, or embedded locally to have the sustained conversation about cost, habit, and safety that determines whether a household actually keeps using the connection. The lesson isn’t that the transactional/relational distinction is wrong, but that it describes the nature of the problem, not the sector. A programme can be designed as if it were a road when it is actually, underneath, a conversation and that mismatch is where these efforts quietly fail.
Brazil’s Family Health Strategy is among the better documented street-level successes: community health workers assigned a small, fixed roster of households, frequently drawn from the same neighbourhood, paid specifically to build a sustained relationship rather than to hit a visit quota. Multiple studies over the programme’s roughly three-decade run have linked it to meaningful declines in infant mortality, crediting the model’s insistence on continuity and local trust rather than one-off contact. Set against this, the United States’ foster care and child-protection system is a sobering counter-case from a wealthy country: caseworkers chronically overloaded, funding not obviously scarcer than in comparable systems, and outcomes still poor, because the job cannot structurally be done well at the caseloads typically assigned. The pattern is consistent: success needs a worker with a roster small and stable enough to build real trust, and some accountability to a community rather than only to a distant bureaucracy.
India’s eye bank counsellors have neither of those things. They meet a grieving family once, under time pressure, with no ongoing relationship and no local accountability pulling them toward a good outcome. That is a design problem, not a fixed national trait — and the fix looks like Brazil’s, not like Aadhaar’s. It means professionalising and properly resourcing the person who asks, not printing more pledge forms.
There’s a small irony worth returning to, because it’s where this essay started rather than a stray detail I’ve since moved past. My mother’s decision did not arrive through a form, a hospital counsellor, or an awareness campaign. It arrived through a cousin someone she loved and admired, delivering the idea in exactly the way Lipsky’s frontline worker is meant to but so rarely manages in practice: personally, with warmth, at a moment that mattered to her specifically. This is, in miniature, the same mechanism Brazil’s programme institutionalises at scale with proximity, continuity, and a messenger who is known rather than official.
The Indian state has not yet found a way to manufacture that deliberately. But Brazil and Sri Lanka have come close, by treating the conversation as real, paid, professional work rather than an afterthought bolted onto a hospital’s other business. The comfortable explanation for India’s low donation numbers, that a religious, tradition-bound population simply lacks the facts, or needs a better awareness campaign, does not survive contact with her case.
The generosity, on the evidence of my own family, was never straightforwardly withheld or freely given. It was something a trusted person could reach in, over decades of stated objection, and unlock. The mechanism for that reaching-in is what India’s institutions still lack — and what Brazil and Sri Lanka, in their different ways, have built.