Shade for People He Would Never Meet — What a developed India must never spend: its earth, its public roads and its institutions
Blossomed, Not Assembled · Volume VI of VI · Parts Twenty-Four to Twenty-Seven
The earlier volumes argued that India cannot copy its way to development, that its inheritance must be sifted rather than worshipped, that Kauṭilya built his state for officials who would be tempted, that a Tamil village wrote and then rewrote its own election rules in 921, and that the right to drink from a public tank at Mahad had to be walked to and defended long after it was written.
← Previous: V · Who May Drink From the Tank? · Series hub
A richer India could still be a poorer one. It could drain Punjab's water, let two apps crowd its payment rails and let its institutions quietly die. The last volume of Blossomed, Not Assembled names what a developed India must never spend.
In Brief: For readers who want Viksit Bharat to last beyond 2047. The volume names three things India cannot buy back once spent: the earth's power to grow back, the citizen's freedom to leave whoever holds her money and data, and institutions that the next generation can correct. It tests each on real cases (UPI, Punjab's aquifers, the data law, GDP, household debt), lists the eight ways good institutions die, and asks how a five-year electoral clock can pay for trees that take fifty years. It closes the six-volume series.
A developed India must never spend three things it cannot buy back: the earth's power to grow back, the citizen's freedom to leave whoever holds her money and data, and the institutions that let the next generation correct what this one built. The three share one feature. Spending any of them takes choices away from people not yet grown, and development should leave the next generation more choices, not fewer. This closing volume tests each on cases: India's payment rails, its data law, the aquifers under Punjab, the way it measures growth, and the ways good institutions have died on this soil. The old texts supply limits, not numbers. Most modern instruments for keeping those limits exist but are weakly used. Whether a five-year electoral clock can pay for a fifty-year tree stays open.
Opening — The Shade Aśoka Planted for People He Would Never Meet
In the ruins of Firoz Shah Kotla in Delhi stands a polished sandstone pillar more than two thousand years old. It was not made for Delhi. It was inscribed for the emperor Aśoka, who ruled most of the subcontinent in the third century BCE. It first stood at Topra, in present-day Haryana.² It carries seven of his edicts. The seventh appears on no other pillar.²
Twenty-seven years after he was crowned, Aśoka had these lines cut into the stone, here in Ven. S. Dhammika's translation:
"Along roads I have had banyan trees planted so that they can give shade to animals and men, and I have had mango groves planted. At intervals of eight krosas, I have had wells dug, rest-houses built, and in various places, I have had watering-places made for the use of animals and men."¹
A banyan takes decades to spread its roof. The king who ordered those saplings would be dead long before travellers rested in their full shade. He planted for strangers, for animals, and for people not yet born.
He also refused to make much of it. The next lines read: "But these are but minor achievements. Such things to make the people happy have been done by former kings."¹ What he wanted, he says, was for people to practise the dhamma, his word for a life of restraint, care and fair dealing.¹ The trees were the easy part.
Then comes the stop. Aśoka died in the 230s BCE. Over the next half-century his empire shrank, and in about 185 BCE its last king, Bṛhadratha, was killed by his own general, Puṣyamitra Śuṅga.³
The script of the edicts, an early form of Brāhmī, slowly passed out of use. In the fourteenth century the Sultan Firoz Shah Tughlaq, who ruled from 1351 to 1388, had this very pillar carried to the Yamuna, floated down to Delhi and raised on his new palace.² His court historian, Shams-i Siraj Afif, records that no one could read what was written on it.⁴ The edict about shade had become a puzzle. It stayed one until 1837. In that year James Prinsep, an official of the Calcutta mint, read and translated it. It was the first record of Aśoka read in modern times.²
The trees and the words had different fates. A banyan can seed itself and grow with no one to keep it, though nobody can show that any banyan on an Indian road today descends from Aśoka's. The edict needed a reader, and waited about two thousand years for one. Nothing in it was false, and almost nothing of the order it described was kept. A ruler can plant for the future. He cannot make the future keep what he planted, unless someone after him can read it, use it and change it.
The earlier volumes of this series asked how India should build. This last one asks the question that wealth makes harder to hear. What must a developed India never spend, however rich it becomes?
My answer takes the citizen's freedom to leave first, then the earth's power to grow back, then the institutions that let people who did not build a thing correct it, and the ways they die. The last part asks how a country that votes every five years can pay for any of them.
Part Twenty-Four — Who Will Taste the Honey When India's Wealth Is Data?
A ward committee with no budget is a meeting
By 2047 far more Indians will live in cities, among strangers. The society that once carried much of the country's welfare worked through the family, the caste council, the temple and the village. Those bodies work badly among strangers. They have to be rebuilt in forms that strangers can join: resident welfare associations, cooperatives, ward committees, professional bodies, schools, and temples, mosques, gurdwaras and churches acting as neighbourhood institutions.
The Constitution already has a place for one of these. Article 243S, added by the 74th Amendment in 1992, requires Wards Committees in every municipality with a population of three lakh or more. It leaves their membership, and how their seats are filled, to each state's legislature.⁵ A committee is only as strong as the money and power of the city above it, and Indian cities have little of either.
In November 2024 the Comptroller and Auditor General published a compendium of its audits of 393 city and town bodies in 18 states. Only 32 per cent of their revenue came from their own sources. The rest came as grants, and their spending ran 42 per cent ahead of their own resources.⁶
The amendment lists 18 functions for cities. Only four had been handed over with full autonomy, among them burial grounds and slaughterhouses.⁶ A ward committee inside such a city may have a chair and a register. It rarely has a budget.
Uttaramērūr, the Tamil village of an earlier volume, did better on one point. In 919 and 921 CE it gave its committees, the tank committee among them, a duty, a term and a check. A member whose accounts, or whose relatives' accounts, had not been rendered could not stand again.⁷
None of this means that wards should run everything. Subsidiarity, the principle that a job belongs at the lowest level that can do it well, was never a rule that everything should be local. Kauṭilya's state was itself central. The principle is about which level does which job. A ward can keep a lake, a park or a drain. It cannot run a payment system, an identity database or an electricity grid. Those jobs need scale, and scale brings a danger the village never faced.
The honey has moved
Kauṭilya, author of the Arthaśāstra, the old Indian treatise on statecraft, did not expect officials to be honest. In R. Shamasastry's translation he writes: "Just as it is impossible not to taste the honey or the poison that finds itself at the tip of the tongue", so an official who handles the king's revenue will taste at least a little of it. And "just as fish moving under water cannot possibly be found out either as drinking or not drinking water", an official cannot easily be caught taking money.⁸
He did not preach at the official. He designed for the temptation, listing in the chapter before forty ways revenue can be stolen, so that each can be watched.⁸
James Madison reached the same conclusion in 1788: "If men were angels, no government would be necessary." Since they are not, he wrote, "you must first enable the government to control the governed", and then "oblige it to control itself."⁹ Neither man trusted good character to do the work of good design.
In 2047 the most valuable revenue in India may not be the king's tax. It may be the record of what a billion and a half people buy, borrow, search, watch and owe, and the machine models trained on that record. I think the honey now sits on the tongue of whoever holds the data and the models. Kauṭilya's rule applies to them unchanged. Whoever holds it will taste it, whether a company, a ministry or a party. The useful question is not whether to trust them. It is whether the design lets anyone see them drink.
Is UPI a success or not?
It is both, and the two answers teach different lessons. UPI, the Unified Payments Interface, is India's shared system for instant payments between bank accounts. It has three layers. The National Payments Corporation of India (NPCI), a not-for-profit company owned by banks and set up at the initiative of the Reserve Bank of India and the Indian Banks' Association, runs the rails and writes the rules.¹⁰ Banks hold the accounts. Apps give the user a screen to pay from.
Any licensed bank or app can plug into the same rails on published terms. In September 2026 those rails carried about 24 billion transactions in one month.¹¹ As a public road, it works.
The crowding is at the app layer. In the same month PhonePe carried 45.6 per cent of UPI transactions and Google Pay 32.2 per cent, together 77.8 per cent. The next app, Paytm, carried about 8 per cent.¹¹
A high share is a warning, not proof of power. A user can keep two apps, a shop can accept both, and the money stays in her bank. Share becomes power through a loop. More users bring more shops, and more payments tell the app more about what people buy. That knowledge can sell loans and insurance to the same users, which pays for still more users. A newcomer who must win customers one at a time falls further behind, and the leader gains the power to set terms for shops and partners.
NPCI saw the risk early. In 2020 it said that no third-party app should carry more than 30 per cent of UPI traffic, measured over three months. In 2022 it moved the deadline to the end of 2024. At the end of 2024 it moved it again, to 31 December 2026.¹² As I write, in October 2026, that is still the deadline. Both large apps still carry far more than 30 per cent of a month's payments.¹¹
A cap is a blunt tool. It punishes an app for being chosen and does nothing about the loop of data. Rules that let a user carry her history to a rival, and that bar an app from favouring its own lending arm, reach the cause. The danger is not a fee today. It is the power to set terms tomorrow, when most of a country's daily payments pass through two companies' screens.
Four questions for any public road
Calling a road "public" settles very little. I would put four questions to any large shared layer, whether payments, identity, health records or a future public AI model:
Can any firm that meets published terms enter on the same terms as everyone else?
Can a user move to another provider with her history and carry on as before, at little cost?
Does any single operator carry so much of the traffic that it could set terms for the rest?
Who audits the rules, and who can change them?
A public road is governed in the public interest. An open road is one that nobody can close to newcomers or capture. A road can be either without the other, and the aim is both.
On open entry, UPI does well. On leaving, it does half the job. A user can switch apps and keep her bank account, because the money never sat inside the app. What the app has learned about her does not move with her.
A file she can download would not fix that. Leaving is real only when the systems work together, so that she can move without losing what the service did for her. On dominance, UPI fails today. On who writes the rules, the answer is NPCI, under the Reserve Bank's oversight. A citizen can read the rules. She has no part in writing them, and the one rule meant to keep the road open has been postponed twice.
What the data law does, and what it leaves open
India's Digital Personal Data Protection Act of 2023 is one guardrail on the new honey. It gives a person the right to a summary of what a company holds about her and with whom it was shared. She may have that data corrected or erased. She is owed a grievance process, and she may name someone to act for her if she dies or cannot act.¹³
The rules under the Act were notified on 14 November 2025, with eighteen months for most duties to come into force.¹⁴ The heaviest penalty, up to ₹250 crore, is for failing to keep data secure.¹⁴
Its limits should be stated as plainly. The Act gives no right to data portability, so a user cannot carry her history from one provider to another under it.¹³ Section 17 lets the central government exempt any instrumentality of the State it chooses to notify, on grounds such as security and public order.¹³ And section 44(3) rewrote a clause of the Right to Information Act, so that "information that is related to personal information" may be refused, without the earlier test of public interest.¹⁵
The Data Protection Board, which hears complaints, raises another question. Who appoints a regulator matters less than whether it can act against whoever appointed it. The Act calls the Board "an independent body". Yet the central government appoints its members for two-year terms, may reappoint them, may remove them after a hearing, and must approve its staff.¹³ A short term in the gift of a government whose own departments the Board may have to judge makes "independent" hard to rely on.
The law guards the citizen's data from companies. It lets the state exempt itself, and it narrows what citizens may ask of the state. It watches one tongue more closely than the other.
Will AI make Kauṭilya obsolete?
No. It moves the tongue. A model that decides who gets a loan, which complaint is heard first or which welfare claim is flagged as suspect does what Kauṭilya's revenue clerk did. It stands between a citizen and what is due to her. Three of the old principles carry over without strain:
Design for temptation. Assume the owner of a model will use it in their own interest, and build the audit before the trust.
Bounded discretion. Kauṭilya did not remove the official's judgment. He made it keep records and accept review.
Irreversibility. A decision that cannot be undone should not rest with a system no one can question.
A citizen does not need the model's inner workings explained. She needs the decision about her explained, which is a different thing. That means notice that a machine was involved, the reasons in her case, a way to contest them, and a named person who answers for the result. A model that gives none of these is Kauṭilya's fish in deeper water. A thick society in 2047 will need local bodies with real money and power, and central layers built as commons. Each will need to watch the other.
Part Twenty-Five — Can India Get Rich Without Draining Its Earth?
A hymn with a rule inside it
The Hymn to the Earth, in the twelfth book of the Atharvaveda, holds a line quoted so often that it has become decoration. It reads mātā bhūmiḥ putro ahaṃ pṛthivyāḥ, which Ralph Griffith translated as "I am the son of Earth, Earth is my Mother" (12.1.12).¹⁶ Quoted alone, it is a sentiment. Read with a verse from later in the same hymn, it becomes a rule:
yat te bhūme vikhanāmi kṣipraṃ tad api rohatu / mā te marma vimṛgvari mā te hṛdayam arpipam "Let what I dig from thee, O Earth, rapidly spring and grow again. O Purifier, let me not pierce through thy vitals or thy heart." Atharvaveda 12.1.35, trans. R. T. H. Griffith (1895–96)¹⁶
The verse does not say "do not dig". The hymn opens by asking the Earth to make ample room for its people, and its speaker ploughs, builds and digs.¹⁶ He is a taker, and he knows it. What he asks is that his taking grow back, and that he not strike the marma. Marma is the word Indian medicine used for the vital points of the body, the places where a wound does lasting harm.
The verse is not a theory of sustainable development. Its poet knew nothing of aquifers or carbon, and I am applying the verse, not reporting what it meant in its own time. Applied, it gives a rule in three steps. Take no faster than the thing grows back. Avoid damage that cannot heal. Where such damage truly cannot be avoided, decide it in public, after weighing the alternatives, and name who will pay. Money paid afterwards is not the same as restoring what was lost.
The second step matters most, because it names something that talk of "sustainability" often blurs. That something is irreversibility, damage that cannot be undone within any span a society can plan for. A forest cut and replanted can grow back in decades. An aquifer drained over decades does not refill within a lifetime. A species lost is lost. The weight of an act of taking depends not only on how much is taken but on whether it can return.
A pierced heart in Punjab
Punjab is the clearest case. Its farmers have had free electricity for farm pumps since 1997.¹⁸ Free power makes pumping almost costless, and the paddy grown on that water drinks deeply. The Central Ground Water Board measures a state's "stage of extraction", the water drawn in a year as a share of what it estimates can be drawn each year without running the aquifer down. In its assessment for 2025, Punjab's stage was about 156 per cent. The national figure was about 61 per cent.¹⁷ In 111 of Punjab's 153 blocks, more is drawn than can be replaced.¹⁷
The state's figure was 164 per cent in 2023.¹⁷ A state official credited canal water for rising water levels in some blocks, but the figures alone do not say why the share fell.¹⁷ The heart is still being pierced, more slowly.
Why does a farmer keep growing paddy over a falling water table? Because the system pays him to. Paddy and wheat are the crops the government reliably buys at its announced minimum support price, and the credit and post-harvest network of the state is built around them. Maize or pulses go to private traders who have no duty to pay that price.⁴² A farmer cannot be asked to leave paddy until another crop has an assured buyer, a market and storage. This is lock-in, the earth's version of the user who cannot leave her app.
Isn't environmentalism a rich country's luxury?
Not when the soil and the water are your own. India needs much more electricity, housing, roads and industry than it has. In 2024 its emissions of carbon dioxide from fossil fuels and industry were about 2.2 tonnes per person.¹⁹ The world average has stayed just below five tonnes per person for over a decade. The average American emitted 14.2 tonnes.¹⁹ The poor cannot be asked to protect the earth by staying poor. The verse does not ask that of them. It asks that the wound heal. It does not forbid taking.
Should India stop building coal plants?
The verse cannot answer that. It supplies the limit, not the number. By June 2025, half of India's installed power capacity came from non-fossil sources, five years ahead of the target it set at the Glasgow summit in 2021.²⁰ Capacity is not generation, though. In 2025 fossil fuels still produced 73 per cent of India's electricity.²¹
Clean power is not free of harm either. Solar parks need land, and by the International Energy Agency's count a typical electric car needs six times the minerals of a conventional one.⁴³ Who bears those costs, the coal worker who loses a job or the village beside a new solar park, belongs in the same public decision. Whether a given plant should be built is a question for engineers, economists and voters. What the verse adds is a discipline for every such decision:
Take less.
Waste less.
Make each unit of water, energy and land produce more.
Substitute where a cleaner means exists.
Restore what can heal.
Never cross a threshold that cannot be crossed back without treating it as the gravest kind of decision.
Not every road, mine or power plant India needs can be built without irreversible damage somewhere. For those the third step of the rule applies, and the unborn are among those who pay.
What does an ancient hymn know about aquifers?
Nothing. Modern science can measure an aquifer and count carbon, so what was once a prayer can now be kept as a ledger. The hymn's part is to say what the ledger is for.
What is GDP missing?
GDP measures income. It does not measure what the income cost. In 1989 the economist Robert Repetto and his colleagues at the World Resources Institute put it bluntly: "A country could exhaust its minerals, cut down its forests, erode its soil, and hunt its wildlife and fisheries to extinction, but measured income would not be affected as these assets disappeared."²² GDP records the value of what is extracted and does not deduct the loss of the asset it came from. It was never built to measure wealth. The fix is not to abolish it but to stop treating it as the whole balance sheet.
The World Bank's Changing Wealth of Nations 2024 counts a country's wealth as its produced capital, natural capital, human capital and net foreign assets. Between 1995 and 2020, in more than a quarter of countries, GDP per person rose while real wealth per person fell.²³ Those countries grew richer by the usual measure and poorer by the one a careful family would use.
The tools to keep a fuller account already exist. The UN Statistical Commission adopted the System of Environmental-Economic Accounting, known as SEEA, as an international statistical standard in 2012. In March 2021 it adopted a second standard, for ecosystems.²⁴ India's statistics ministry, MoSPI, has published EnviStats India: Environment Accounts on this framework every year since 2018. The accounts cover forests, water, minerals, energy, waste and the services ecosystems provide.²⁶
Its forest accounts for 2025 valued the climate services of India's forests, such as holding carbon, at about 2.63 per cent of GDP.²⁵ That is a value set beside GDP, not added to it, and GDP shows none of its loss. In October 2026 the ministry set out a strategy for 2026 to 2030 to make these accounts longer, more timely and better connected. It has not adopted a "green GDP".²⁶
I would not invent a new index. A ledger with more columns is enough, and it is the kind a family already keeps in its head. Beside income, a natural column would show the water table, the forest and the soil.
A human column would show whether children can read. In 2024 fewer than half of Class 5 children in rural government schools could read a text meant for Class 2.²⁷ A social column would show the care families give their old and their sick, which no national account records. A year in which income rose and all three columns fell would then be reported as what it was.
Desire, served or farmed
In the old Indian list of human aims, kāma, desire and enjoyment, is a legitimate aim. The tradition held it within dharma, the order of right conduct, and never treated it as shameful. An economy can serve desire, or it can farm it. Farming desire means using what a seller knows about a buyer to create wants, financing them on easy credit, and counting the sales as growth.
The Reserve Bank's Financial Stability Report of December 2025 put household debt at 41.3 per cent of GDP in March 2025, above its five-year average of 38.3 per cent. Retail loans other than for homes, such as personal loans, credit cards and loans for vehicles and consumer goods, made up 55.3 per cent of household borrowing. The Bank describes them as mostly for consumption.²⁸ Not all are. A vehicle loan can buy a delivery van.
The June 2026 report, as summarised by the journalist Vivek Kaul, put that share at 58.4 per cent by March 2026. A credit bureau found that 65 per cent of the customers of digital lenders were under thirty-five.²⁹ That is a pattern worth watching, not proof that the young are being exploited.
Neither credit nor desire is the problem. The danger is a mechanism. The same data that reads a buyer can time a credit offer to the moment of wanting, before the wanting has passed. A loan that buys a skill carries a young person across a river. A debt run up for a craving the app was built to create strikes a marma early, and the wound lasts longer than the pleasure.
The marma beyond the soil
The ledger has more columns than ecology. This extension is my analogy, not the hymn's, which speaks only of the earth.
A language that its children are taught to be ashamed of can die within two generations, and nothing brings it back. Children's attention, captured young by devices designed to farm it, may not be won back at the same age later. A debt, financial or ecological, that outlasts every living payer is a wound inflicted on people who cannot consent. These losses are unlike each other and unlike a drained aquifer. What they share is that some damage removes a capacity that cannot easily be restored.
A tenant, not an owner
The Īśa Upaniṣad, one of the shortest of the old philosophical texts, opens with a verse about ownership. Max Müller translated its first line as "All this, whatsoever moves on earth, is to be hidden in the Lord (the Self)."³⁰ It continues: tena tyaktena bhuñjīthāḥ mā gṛdhaḥ kasyasvid dhanam, which I render as "enjoy it by letting go, and do not covet anyone's wealth."
Many readers take the verse as a call to renunciation. I read it, as one reasonable reading among others, as tenancy. The earth is held in trust, and its users are tenants, not owners. A good tenant enjoys the house fully, and does not burn the furniture.
Indian tradition spoke of three debts, ṛṇa, that every person is born owing. One of them was owed to the powers that sustain life, and it was paid by returning at least what one took. A country that becomes rich by exhausting its own soil and water has not become viksit, the word for a flower that has opened. It has spent its inheritance and called the spending income.
Part Twenty-Six — Eight Quiet Ways a Good Institution Dies
Why do good institutions die?
Some deaths are loud. Most are slow, of causes that never look like death on the day. On the evidence of these volumes I count eight ways, each with an Indian example already met in the series. They are not separate boxes. They feed one another. A blind institution is easier to capture, and a captured one soon learns to mimic.
An institution can die of personalisation. The founder will not leave, or leaves nothing that works without him. Aśoka appointed officers of dhamma to carry his teaching through the empire.¹ No source I have read shows them working after the dynasty fell. That is an absence of evidence, not proof that they vanished. The early sign is a body in which every decision waits for one person.
It can be captured. Kin, a caste or a party takes the committee and runs it for itself, and often wraps itself in the institution's sanctity, so that questioning it looks like an attack on the faith. By the early twentieth century many Sikh gurdwaras were controlled by hereditary custodians called mahants, with the colonial government's support. At Nankana Sahib in February 1921 a custodian's men killed about 200 reformers, by the Encyclopaedia of Sikhism's count. A law of 1925 handed the historic shrines to an elected Sikh body.³¹ The early sign is the same few families in every office.
It can learn obedience from its grants. An institution that lives on a renewed grant learns to please its funder rather than serve its purpose. An earlier volume argued, as my inference rather than a finding, that India's public universities, funded year by year, are kept short-sighted this way. The early sign is a report written for the funder, not the public.
It can silt up. When no one is paid, honoured or held responsible for maintenance, the tank fills slowly, and nobody notices until the rain fails. Uttaramērūr kept a standing committee for its tank, and a tank needs one, because silt never stops.⁷ A city drain, a court's backlog and a hospital's equipment silt the same way. The early sign is a repair budget that is cut first every year.
It can keep a franchise too narrow, the circle of people allowed to choose and to use. Those shut out have no reason to defend it, and many reasons to welcome its fall. Uttaramērūr's assembly was open only to landholding Brahmins with their own house, between thirty-five and seventy, who knew the Vedic texts.⁷ At Mahad, people were kept from the town's public tank for years after a resolution had opened it to them.³⁶ The early sign is that those most affected have no seat.
It can lose its practice. The record survives, but no second generation carries the skill of using it. Brāhmī did not vanish from the stone. It vanished from the hands that could read it, and by Firoz Shah's time no one at his court could.⁴ The early sign is one person who alone knows how the thing works.
It can survive as mimicry, a copy of the forms after the function has gone. The committee meets, the register is kept, the report is filed, and nothing is produced.
The economists Lant Pritchett, Michael Woolcock and Matt Andrews named this in 2013, after a harmless snake that wears a poisonous one's colours.³² The rural school met earlier in this series is its Indian specimen. It has the building and the payroll, and fewer than half its Class 5 children read a Class 2 text.²⁷ Nothing looks dead, which makes this the hardest death to see. The early sign is a form that is always reported and a result that never is.
It can go blind. An institution stops finding out what it is doing. In November 2019 the statistics ministry withheld the 2017–18 consumer expenditure survey, citing "data quality issues". A leaked version had shown average spending per person falling, for the first time in about four decades.³³ A new survey, for 2022–23, came out in 2024, but its methods changed, and the ministry itself warns that this must be kept in mind when comparing it with earlier rounds.³³ No new national survey of household spending was published between the one for 2011–12 and the one for 2022–23. The early sign is a number that stops appearing.
Some of the eight are failures of first design, some are failures of the people who inherit an institution, and most are both. Uttaramērūr's narrow franchise was written into its rules from the start. The same rules foresaw two other deaths, relatives and accounts. No design can foresee all eight, so the remedy has to be instruments that keep working after the designers have gone, and that their successors can change.
Part Twenty-Seven — How Can a Five-Year Government Plant a Fifty-Year Tree?
Why plant a tree you will never sit under?
The Bhagavad Gītā gives one form of the discipline that long work needs:
karmaṇy evādhikāras te mā phaleṣu kadācana / mā karmaphalahetur bhūr mā te saṅgo 'stv akarmaṇi "Your right is to the work alone, never to its fruits. Do not let the fruit be your motive, and do not cling to doing nothing." Bhagavad Gītā 2.47, my rendering³⁴
Kṛṣṇa says this to Arjuna on a battlefield, about a warrior's duty. He is speaking to a person, not a nation. I take one thing from it here.
Institutions that last need people willing to act when most of the return will belong to others. One who needs to see the result will build only what can be finished within a term of office. One freed from the fruit can build what takes a century. Freedom from the fruit is not freedom from consequences, though. A builder must not need the credit, but she must answer for whether the tree survives.
The second volume of this series set out six aims of life. Two of them are my additions to the classical four: jñāna, knowledge, and saśaktīkaraṇa, empowerment, by which I mean more effective choice over the conditions of one's own life, and power handed on to the next generation. Aśoka's banyans are the picture of the second.
Two clocks, the file and the tree
A republic runs on two clocks at once. The file's clock is really several short ones: the election, the annual budget, the transfer of officers, a company's quarterly results and a poor household's need to get through this season. The tree grows on the rhythm of generations. India's centenary is twenty-one years away. Many of those shaping it will not see it, and most of those who see it will not know who shaped it.
Hoping for more patient politicians will not do it. The file has to be made to pay for the tree, and that takes instruments that outlast the people who set them up:
An endowment whose capital no minister can spend. Uṣavadāta deposited 3,000 kārṣāpaṇas, silver coins, with two weavers' guilds near Nāsik. Only the interest was to be spent, on the monks of a cave. The deed is usually placed in the first or second century CE, but the era of its date is disputed.³⁵
A Finance Commission, which the Constitution requires the President to set up at least every fifth year, whoever is in power. It recommends how taxes are divided between the Union and the states and among the states, and what grants they receive.³⁷
An amendment rule that makes change possible but not easy. Article 368 asks for a majority of each House's whole membership and two-thirds of those present and voting. Some changes also need the consent of half the states.³⁸
A statistical office whose numbers are published whether or not they flatter the government.
A law that protects a forest, a river or an aquifer for longer than any government's term.
A market-share rule enforced on the date it names, and a right to move one's history from one provider to another.
These are the modern forms of Aśoka's trees. They make the Gītā's discipline payable, and they carry a danger of their own. Durability without accountability becomes entrenchment. Accountability without durability becomes short-termism. A body built to outlast elections needs a clear mandate, independence in its daily work, public reports and a review at fixed intervals, or the fifty-year tree becomes a fifty-year excuse.
Can the founder leave?
The list of deaths gives a test for anyone who builds an institution, whether a minister, a mayor, a temple trust or a start-up founder. A weak institution says, "Everything depends on me." A good one says, "I can leave." A great one says, "My successor can change what I built," and it outlives its founder without becoming the property of its successors. Succession without the power to correct produces a bureaucracy. Correction without succession produces a revolution every generation. A developed country needs both.
Can India really teach the world how to develop?
The public talk that set off this series ended on a hope that India will one day tell the world how to develop.³⁹ I am not in a position to say what India will teach in 2047, and I doubt anyone is. I think the sober version of the hope is narrower, and stronger for being narrow. A country can teach only what it has tested and corrected in its own hands.
India can teach the payments road, with the warning about its two crowded on-ramps. It can teach the high-yielding wheat of the 1960s, a borrowing adapted to its own soil, with Punjab's aquifers as the warning attached. It can show, on a temple wall at Uttaramērūr, an election rule written, tried and rewritten within two years.⁷
It cannot yet teach how to make every child read, or how to farm without draining the water under the farm. Other countries do not learn from a country's claims. They learn from its records, including the record of what failed and how it was corrected.
Ending — The Stone Outlived Everything It Recorded
Uṣavadāta meant his endowment at Nāsik to last. The weavers' guilds that held his coins are long gone, and for many centuries no monk in that cave has drawn robe-money from them. The cave wall still gives the sums, the guilds, the rate of one per hundred a month, and the fact that the deed was registered in the town's assembly.³⁵
The committees of Uttaramērūr stopped meeting long ago. The temple wall still gives the age limits, the land a candidate had to own and the rule about unrendered accounts. It still describes the palm-leaf tickets, and the pot from which a young boy drew the names.⁷ In both places the record of how the thing was built has outlived the thing. The stone keeps the information, not the institution. It is not enough to build the thing again. It is enough to begin asking how it was built, whom it served and whom it shut out, and then to build better.
The franchise has changed since the stone was cut. Article 15 of the Constitution now forbids any bar on grounds of religion, race, caste, sex or place of birth to "the use of wells, tanks, bathing ghats, roads and places of public resort" kept with public money or dedicated to public use.⁴⁰
At Mahad, a tank opened by resolution in 1924 was drunk from in 1927, and the Bombay High Court dismissed the caste Hindus' last appeal over it only in March 1937.³⁶ In a national survey of 2011–12, 27 per cent of respondents still told surveyors that they practised untouchability in some form.⁴¹ The old walls can be read again because scholars from Prinsep onward relearned their scripts. The franchise that could use them now includes every citizen of the republic, by law if not yet in use.
A banyan needs no successor. Aśoka's trees could grow on without him. An institution cannot. It needs people who understand it, keep it and are free to change it, and a franchise wide enough that all of them may. The earth that grows back, the road a citizen can leave and the institution the young can correct have that one thing in common. Each leaves the next generation more choices than it would otherwise have had.
Words to keep
Arthaśāstra: the old Indian treatise on statecraft and economy, attributed to Kauṭilya.
Brāhmī: the script of Aśoka's edicts. It fell out of use and was deciphered by James Prinsep in 1837.
dhamma: Aśoka's word (Sanskrit dharma) for a life of restraint, care and fair dealing.
dharma: the order of right conduct, the duty that holds a person and a society together.
jñāna: knowledge. The first aim the author adds to the classical four.
kāma: desire and enjoyment, a legitimate human aim when held within dharma.
kārṣāpaṇa: a silver coin of ancient India.
mahant: a hereditary custodian of a shrine. Sikh reformers took the gurdwaras back from them between 1920 and 1925.
marma: a vital point of the body, where a wound does lasting harm.
ṛṇa: debt. Tradition spoke of three debts every person is born owing.
sabhā: a village assembly, such as the one at Uttaramērūr.
saśaktīkaraṇa: empowerment, in the author's sense more effective choice over the conditions of one's own life, and power handed on. The second aim he adds to the classical four.
viksit: opened, as a flower opens. From Sanskrit vikasita, the word in "Viksit Bharat".
Sources
Aśoka, Pillar Edict VII and Rock Edict V, in Ven. S. Dhammika, The Edicts of King Asoka: An English Rendering (Buddhist Publication Society, Wheel 386/387, 1993). https://www.accesstoinsight.org/lib/authors/dhammika/wheel386.html
E. Hultzsch, Inscriptions of Asoka, Corpus Inscriptionum Indicarum vol. 1, new edition (Oxford: Clarendon Press, 1925), Introduction, pp. xv–xvi (the Delhi-Topra pillar, its unique seventh edict, its removal by Firoz Shah, Prinsep's reading of 1837). https://archive.org/details/InscriptionsOfAsoka.NewEditionByE.Hultzsch
"Shunga dynasty", Encyclopaedia Britannica (Puṣyamitra assassinated Bṛhadratha, the last Mauryan ruler, c. 185 BCE), https://www.britannica.com/topic/Shunga-dynasty, and "Ashoka", Encyclopaedia Britannica, https://www.britannica.com/biography/Ashoka. See also Romila Thapar, Aśoka and the Decline of the Mauryas (Oxford University Press, 1961).
Shams-i Siraj Afif, Tārīkh-i Fīrūz Shāhī, in H. M. Elliot and J. Dawson, The History of India as Told by Its Own Historians, vol. 3 (1871), pp. 352–353, as retold in "Journey of a pillar", The Tribune (Spectrum), 27 November 2011. https://www.tribuneindia.com/2011/20111127/spectrum/art.htm
Constitution of India, Article 243S (inserted by the Constitution (Seventy-fourth Amendment) Act, 1992). https://www.constitutionofindia.net/articles/243s-constitution-and-composition-of-wards-committees-etc/
Risha Chitlangia, "Over 30 yrs after 74th Amendment, CAG flags 'weak compliance' with law empowering urban local bodies", ThePrint, 14 November 2024, reporting the CAG compendium of performance audits on the 74th Amendment. https://theprint.in/india/governance/over-30-yrs-after-74th-amendment-cag-flags-weak-compliance-with-law-empowering-urban-local-bodies/2357067
Uttaramērūr inscriptions of Parāntaka I, regnal years 12 and 14 (c. 919 and 921 CE), Vaikuṇṭha Perumāḷ temple: V. Venkayya, "Two Inscriptions at Uttaramallūr", Archaeological Survey of India, Annual Report 1904–05 (1908), pp. 131–145, https://archive.org/details/in.ernet.dli.2015.35485; dating after the DHARMA project edition (Emmanuel Francis, 2024), https://erc-dharma.github.io/tfa-uttiramerur-epigraphy/texts/htmloutput/DHARMA_INSUttiramerur00002.html; K. A. Nilakanta Sastri, Studies in Cōḷa History and Administration (University of Madras, 1932).
Kauṭilya, Arthaśāstra 2.9.32–33 (honey on the tongue; fish and water) and 2.8 (forty ways of embezzlement), trans. R. Shamasastry (1915), "Examination of the Conduct of Government Servants". https://www.wisdomlib.org/hinduism/book/kautilya-arthashastra/d/doc366055.html. Verse numbers follow R. P. Kangle's edition (Bombay, 1960–65).
James Madison, The Federalist No. 51 (1788). https://avalon.law.yale.edu/18th_century/fed51.asp
National Payments Corporation of India, "About us". https://www.npci.org.in/who-we-are/about-us
"PhonePe & Google Pay's combined UPI app share fell to 77.8% in Sept 2026", MediaNama, October 2026, from NPCI's app-wise data (published at https://www.npci.org.in/what-we-do/upi/upi-ecosystem-statistics). https://www.medianama.com/2026/10/223-phonepe-google-pay-upi-share-sept-2026/
"NPCI extends UPI market share cap deadline to 2026", MediaNama, 2 January 2025 (NPCI circular UPI-OC-No-210-FY-24-25). https://www.medianama.com/2025/01/223-npci-extends-upi-market-share-cap-deadline-to-2026-google-phonepe-relief/
The Digital Personal Data Protection Act, 2023 (No. 22 of 2023), sections 11–14, 17, 19–21, 24 and 28. https://www.meity.gov.in/static/uploads/2024/06/2bf1f0e9f04e6fb4f8fef35e82c42aa5.pdf
Press Information Bureau, backgrounder on the Digital Personal Data Protection Rules, 2025 (notified 14 November 2025). https://static.pib.gov.in/WriteReadData/specificdocs/documents/2025/nov/doc20251117695301.pdf
Shivam Jadaun, "Section 44(3) of the DPDP Act: Cloaking India's right to know in secrecy", Bar and Bench, 30 December 2025. https://www.barandbench.com/columns/section-443-of-the-dpdp-act-cloaking-indias-right-to-know-in-secrecy
Atharvaveda (Śaunaka) 12.1.1, 12.1.12 and 12.1.35, trans. Ralph T. H. Griffith, The Hymns of the Atharvaveda (1895–96). https://sacred-texts.com/hin/av/av12001.htm. Compare W. D. Whitney, Atharva-Veda Saṃhitā (Harvard Oriental Series, 1905).
Central Ground Water Board, Dynamic Ground Water Resources of India 2025, as reported in "Groundwater extraction drops in Punjab as canal irrigation expands: Report", The Tribune, 22 June 2026, https://www.tribuneindia.com/news/punjab/groundwater-extraction-drops-in-punjab-as-canal-irrigation-expands-report/, and the Minister of State for Jal Shakti's written reply to the Rajya Sabha, reported in "Nearly 11 per cent groundwater assessment units over-exploited: Govt", ETV Bharat, 21 July 2026, https://www.etvbharat.com/en/bharat/nearly-11-per-cent-groundwater-assessment-units-over-exploited-govt-enn26072007475
"Punjab spent Rs 1.34 lakh crore on free farm power since rollout", The Tribune, 22 April 2025 (free power from 1997). https://www.tribuneindia.com/news/punjab/punjab-spent-rs-1-34-lakh-crore-on-free-farm-power-since-rollout
Global Carbon Budget (2025), via Our World in Data (CO₂ from fossil fuels and industry): India country profile (2.2 tonnes per person in 2024), https://ourworldindata.org/co2/country/india; United States profile (14.2 tonnes), https://ourworldindata.org/co2/country/united-states; per-capita chart (world "just below five tonnes per person for over a decade"), https://ourworldindata.org/grapher/co-emissions-per-capita
"India achieves 50% non-fossil power capacity five years ahead of target", Mongabay India, July 2025 (capacity as of 30 June 2025). https://india.mongabay.com/short-article/2025/07/india-achieves-50-non-fossil-power-capacity-five-years-ahead-of-target/
Ember, India country profile (fossil fuels 73% of generation in 2025). https://ember-energy.org/countries-and-regions/india
Robert Repetto et al., Wasting Assets: Natural Resources in the National Income Accounts (World Resources Institute, 1989), quoted in FAO, Forestry and environmental accounting. https://www.fao.org/4/t9500e/t9500e0c.htm
World Bank, The Changing Wealth of Nations 2024. https://www.worldbank.org/en/publication/the-changing-wealth-of-nations
United Nations, System of Environmental-Economic Accounting: "SEEA Central Framework" (adopted by the UN Statistical Commission at its 43rd session, 2012), https://seea.un.org/content/seea-central-framework, and "Ecosystem Accounting" (adopted at the 52nd session, March 2021), https://seea.un.org/ecosystem-accounting
MoSPI, EnviStats India 2025: Environmental Accounting on Forest, as summarised by Drishti IAS, 29 September 2025. https://www.drishtiias.com/daily-updates/daily-news-analysis/environmental-accounting-on-forest-2025
"MoSPI sets 2026–30 roadmap to improve how India tracks natural resources", The Policy Edge, 7 October 2026. https://www.policyedge.in/p/mospi-sets-202630-roadmap-to-improve-how-india-tracks-natural-resources
ASER Centre, Annual Status of Education Report (Rural) 2024 (government schools: 44.8% of Class 5 children can read a Class 2 text), https://asercentre.org/aser-2024/, as reported in "ASER 2024: Reading & arithmetic skills improve…", Millennium Post. https://www.millenniumpost.in/big-stories/aser-2024-reading-arithmetic-skills-improve-among-class-3-and-5-students-led-by-govt-schools-596643
Reserve Bank of India, Financial Stability Report, December 2025, as reported in "India's household debt rises to 41.3 per cent of GDP: RBI FSR", Business Standard, 31 December 2025. https://www.business-standard.com/economy/news/indias-household-debt-rises-to-41-3-per-cent-of-gdp-rbi-fsr-125123101081_1.html
Vivek Kaul, "The Great Borrowing Boom: Why Indian Households are Taking on More Debt", The India Forum, 15 July 2026, drawing on the RBI Financial Stability Report, June 2026 (https://www.rbi.org.in/Scripts/PublicationReportDetails.aspx?UrlPage=&ID=1319) and CRIF High Mark (September 2025). https://www.theindiaforum.in/economy/great-borrowing-boom
Īśa Upaniṣad 1, trans. F. Max Müller, The Upanishads, Part 1, Sacred Books of the East vol. 1 (1879). https://hinduwebsite.com/sacredscripts/hinduism/upanishads/vagasaneyi.asp
The Encyclopaedia of Sikhism, ed. Harbans Singh (Punjabi University, Patiala), "Shiromani Gurdwara Parbandhak Committee". https://eos.learnpunjabi.org/SHIROMANI%20GURDWARA%20PARBANDHAK%20COMMITTEE.html
Lant Pritchett, Michael Woolcock and Matt Andrews, "Looking Like a State: Techniques of Persistent Failure in State Capability for Implementation", Journal of Development Studies 49:1 (2013). https://doi.org/10.1080/00220388.2012.709614
"Govt says consumer spending survey withheld over 'data quality' issues", The Week, 15 November 2019, https://www.theweek.in/news/biz-tech/2019/11/15/govt-says-it-withheld-consumer-spending-survey-over-data-quality-issues.html, and MoSPI, Household Consumption Expenditure Survey 2022–23: Fact Sheet (2024). https://mospi.gov.in/sites/default/files/publication_reports/Factsheet_HCES_2022-23.pdf
Bhagavad Gītā 2.47, Sanskrit text at IIT Kanpur Gita Supersite. https://heritage.iitk.ac.in/SBG/2.47
Inscription of Uṣavadāta, Nāsik Cave 10, year 42, in É. Senart, "The Inscriptions in the Caves at Nasik", Epigraphia Indica vol. 8 (1905–06), Nasik no. 12. https://whatisindia.com/inscriptions/epigraphica_indica/vol8_1905-1906/senart24.html. The era of the date, and so its century, is disputed.
Supreme Court of India, Dr B. R. Ambedkar gallery, "Mahad Satyagraha", https://main.sci.gov.in/AMB/mahadsatyagraha.php, and Bombay High Court, Narhari Damodar Vaidya v. Bhimrao Ramji Ambedkar, Second Appeal No. 462 of 1933, judgment of 17 March 1937, AIR 1938 Bom 146, https://www.courtkutchehry.com/Judgement/Search/t/455024-narhari-damodar-vaidya-appellant-hash. See also Anand Teltumbde, Mahad: The Making of the First Dalit Revolt (Aakar, 2016).
Constitution of India, Article 280(1) and (3). https://www.constitutionofindia.net/articles/article-280-finance-commission/
Constitution of India, Article 368(2). https://www.constitutionofindia.net/articles/article-368-power-of-parliament-to-amend-the-constitution-and-procedure-therefor/
Report of a public talk on "Ancient Indian Wisdom as a Guide for Viksit Bharat", The Morning Standard, 26 September 2026.
Constitution of India, Article 15(2). https://www.constitutionofindia.net/articles/article-15-prohibition-of-discrimination-on-grounds-of-religion-race-caste-sex-or-place-of-birth/
India Human Development Survey II (2011–12), NCAER and University of Maryland, reported in "Biggest caste survey: One in four Indians admit to practising untouchability", NCAER. https://www.ncaer.org/news/biggest-caste-survey-one-in-four-indians-admit-to-practising-untouchability
Saroj, "Punjab and Haryana are unable to get out of paddy-wheat cycle; why is that?", Down To Earth, 28 June 2023, drawing on a 2021 ICAR policy paper and Ranjini Basu (2022). https://www.downtoearth.org.in/news/agriculture/punjab-and-haryana-are-unable-to-get-out-of-paddy-wheat-cycle-why-is-that--90289
International Energy Agency, "Clean energy demand for critical minerals set to soar as the world pursues net zero goals", press release for The Role of Critical Minerals in Clean Energy Transitions (May 2021). https://iea.org/news/clean-energy-demand-for-critical-minerals-set-to-soar-as-the-world-pursues-net-zero-goals
How this was written
Read: the earlier volumes, Dhammika's translation of the edicts, Hultzsch's introduction, Griffith's Atharvaveda, Shamasastry's Arthaśāstra, the Data Protection Act, the 1937 Mahad judgment, the MoSPI fact sheet, the SEEA pages, Our World in Data, Ember and the World Bank.
Could not check directly, so taken through the reports named: the Reserve Bank's June 2026 report (Kaul), the CGWB 2025 report (a press report and a parliamentary reply), NPCI's app-wise table (MediaNama), MoSPI's forest accounts (Drishti), the CAG compendium (ThePrint), Afif's chronicle (Hultzsch and The Tribune), the Uttaramērūr translation pages (Venkayya as read for Volume IV) and the Morning Standard report. Wikipedia citations for the pillar, the Mauryan end, Uttaramērūr, the gurdwaras, SEEA and Mahad were replaced. The Mauryan end now rests on Britannica, a reference work, and Thapar, which I could not open.
The author's own construction: the four road questions, the three-step reading of the hymn's rule, the ledger with more columns, the reading of the Īśa verse as tenancy, the extension of marma to language, attention and debt, the eight ways of dying as a list and their early signs, the founder's test, and the claim that the three things not to spend share one feature.
What would change the conclusion: evidence that UPI's two large apps cannot in practice set terms for the rest, or that GDP growth in India has not come with any loss of natural wealth, would weaken the first two claims. A record of Aśoka's dhamma officers working after the Mauryas would remove the example for personalisation.
No imagined persons appear in this volume.
Comments