The numbers Brief Episode map Timeline Takeaways Charts Quotes Clips Glossary Go next Transcript

Episode 63 · Impact · 46 min

The Rube Goldberg device that feeds India

Before anyone kicks the green revolution, Mark Kahn wants credit given: one-sixth of humanity went from a ship-to-mouth existence in the 1960s to food secure within a decade. His argument is that the contraption which did it (input subsidies, MSP, the FCI, a ration system designed before credit cards existed) has outlasted its assumptions, and the honest replacement is to stop delivering welfare through prices and cut the farmer a cheque. PM-Kisan's ₹6,000 a year, he says, is an arbitrary number.

Supply chain & agri Regulation & policy India macro

Mark Kahn

Co-founder, Omnivore · with Vishal Krishna

45:44

Said in this episode

▶ 1:01

195M

India's cultivated area, as stated on air

The host's opening figure, with the unit garbled in the auto-captions. Of it, 63% is rain-fed and 37% irrigated.

▶ 21:55

₹6,000/yr

PM-Kisan's direct transfer per farmer

Kahn treats the figure as arbitrary: fold in the other programmes and it could be ₹60,000, or a ₹1.5 lakh floor under farm incomes.

▶ 19:45

₹70,000 cr

India's vegetable-oil import bill

The host's number for essentially the only thing Indian agriculture buys abroad; Kahn calls the country otherwise an export-surplus nation in agriculture.

▶ 17:41

₹6,000 cr

Punjab's annual mandi taxation

Kahn's figure for what the state collects from its mandi system each year, largely unaccounted for, and part of what the farm-law protests were defending.

▶ 38:15

20–25%

Yield and profit gain from better agronomy alone

Kahn's estimate of India's low-hanging fruit, against a US frontier where further gains require pushing the technology itself.

▶ 1:16

33+

Companies Omnivore has backed since 2010

Across three funds, per the host's introduction; Kahn says agritech was too small and too peripheral to fund until about five years ago.

The brief

The argument in sixty seconds

Kahn's case runs in two movements. The first is a defence: before anyone kicks the green revolution, notice that one-sixth of humanity went from ship-to-mouth in the 1960s to food secure within a decade even as the population exploded, and that the machinery which did it (high-yielding varieties, fertiliser subsidies, the FCI, minimum support prices, a ration system built before credit cards or digital identity existed) was, for all its absurdity, a Rube Goldberg device that worked. The second is the indictment: those programmes have outlasted the boundary conditions they were designed under, and each now carries a bureaucracy and a constituency that eats off it, which is why the fertiliser subsidy survives every obvious argument for direct transfer, and why part of what Delhi's farmer protests defended was the roughly ₹6,000 crore a year Punjab collects, largely unaccounted, from its mandi system. His prescription is blunt: stop using price to deliver farmer welfare, because that produces Soviet-style decisions like growing subsidised, groundwater-hungry rice a thousand kilometres from Delhi. Cut the farmer a cheque instead. PM-Kisan's ₹6,000 could be ₹60,000, or a ₹1.5 lakh floor under farm incomes, if the other programmes were folded in. On land he is equally contrarian: consolidation matters less than an information layer separating who owns from who tills, since Pakistan's true zamindari estates do not beat India on yields and China's smaller plots do. And on his own trade, three funds and 33-plus companies since 2010, he insists India must build the informational layer before the robots, because you cannot sell a smallholder half a million dollars of kit.

Worth your time if you are

Agritech founders selling to smallholder farmers

Policy readers arguing about MSP, DBT and the mandis

Investors sizing India's rural economy

City families still holding ancestral farmland

Episode map

Where the conversation travels

Every block is a chapter, coloured by what it's about. Click any of it to jump straight to that minute on YouTube.

Supply chain & agriRegulation & policyIndia macroVenture capitalImpact & outcomesFounder journey

01Cold open: Houston, ITC and 195 million acres 0:00 Vishal opens with his own family selling its land fifty years ago and the scale of the sector (roughly 70% of Indians rural, a cultivated area of 195 million split 63% rain-fed to 37% irrigated) before Kahn traces his route from a Houston neighbourhood full of NRIs to graduate-school projects for ITC's agribusiness division and a move to Mumbai in late 2007. 02The third leg of the stool 3:31 Agriculture struck Kahn as the one sector never fully liberalised, and his view of India's future has three legs: value-added engineering, software and services, and an agriculture with the global ambition of a Brazil or a Turkey. A democracy with half its people on farms cannot follow China's path without bringing the countryside along. 03Land reform without the zamindars 6:50 People overstate consolidation: Pakistan never did land reform and still has 10,000- and 20,000-hectare estates without beating India on yields, while China's smaller holdings do, so the real work is a digital layer separating owner from tiller, directing subsidies to the cultivator and preserving the option value that stops families becoming the Gurgaon farmer who sold in 1990. 04Congratulate yourselves before you tear it down 11:28 Kahn insists on the achievement first: one-sixth of humanity going from ship-to-mouth in the 1960s to food secure within a decade, with edible oil the only meaningful import. He defends the Rube Goldberg device that delivered calories, if not nutrition, noting the PDS predates credit cards and digital identity. 05Programmes that outlast their assumptions 15:37 Large government programmes acquire bureaucracies and stakeholders who eat off them, which is why the fertiliser subsidy resists direct transfer despite every argument for it, why Punjab's ₹6,000 crore of mandi taxation shaped the farmer protests, and why reform will move like a waltz rather than a straight line. 06Five acres and cousins who want engineering 18:28 Vishal turns personal, with cousins in Karnataka on five or ten acres who would rather sell up and study engineering, and adds the counterweight number: about ₹70,000 crore of vegetable-oil imports against an otherwise export-surplus agricultural economy. 07Decouple welfare from price 20:00 India's method has been to subsidise inputs and inflate outputs, which produces Soviet-style choices: cheap fertiliser overused, supported rice grown near Delhi on groundwater that cannot spare it, and the field burning that follows. Kahn's fix is to cut the farmer a cheque and stop steering behaviour through prices. 08The poverty trap and the better intermediary 22:42 The typical smallholder owes a moneylender who also sells them poor inputs and buys back their crop, and agritech platforms attack that loop hat by hat: market the produce above the local trader's price, supply honest inputs, then move the farmer towards institutional finance, displacing intermediaries and large agribusinesses at once. 09Tomatoes dumped along the Kolar highway 25:00 Forty years of tomatoes rotting by the roadside meet the limits of policy instruments (warehouse receipts do not work for a perishable) while horticulture is quietly a success story, with avocados in Arunachal and dragon fruit everywhere, and the cure for periodic gluts is disintermediation plus decentralised processing. 10Too small, too peripheral, too weird 28:19 Until about five years ago there was simply no money in Indian agritech, and Kahn recalls one of the country's most senior agribusiness figures rubbishing his entire enterprise from a panel. It is vindication he now allows himself to enjoy. 11Universal basic farmer income 29:24 Given nine or ten kinds of subsidy, Kahn picks DBT over price support and cash over even a DBT-linked fertiliser subsidy: make it ₹1.5 lakh, put a floor under farm incomes, and let farmers choose bio-fertiliser, organic or permaculture instead of subsidising urea into the groundwater and paddy into Delhi's air. 12The rural vote does not swing one way 31:31 Kahn, who calls himself an immigrant rather than an expat, argues rural votes have toppled governments in both directions, and while he did not love seeing basic reforms rolled back over politics, it was the will of the voter, so the next reformer should talk to farmers and acknowledge a century of terms of trade running against agriculture. 13UPI, logistics, and the agri stack 33:52 After the payments stack and a logistics stack, the agri stack is already a thing, and Kahn's answer to whether logistics or the farmer side should be fixed first is that the whole ecosystem has to move together, which leads back to 2010, when the word agritech barely existed in India. 14The storm they wanted to get ahead of 35:51 Years inside a domestic agribusiness showed Kahn transformation happening, but far too slowly, while the digital change of the early Flipkart era refused to filter into agribusiness at all, so Omnivore was built, with Godrej Agrovet as anchor, to catalyse an ecosystem that did not yet exist. 15Why India gets platforms and America gets robots 37:28 American farmers are effectively professional agronomists ageing into their 60s and dependent on migrant labour, so the US innovates autonomous tractors and robotic pickers; India's upside is 20-25% from better agronomy alone, which means an informational and organisational layer first and use-case deep tech, like the farm robots of a portfolio company, riding on top of it. 16Walk alone, and read India After Gandhi 41:16 Kahn's guide to happiness is to stop measuring yourself against everyone else (Tagore's walk-alone line is tattooed on his arm) and his reading list runs from Michael Pollan's The Omnivore's Dilemma, which gave the fund its name, to Sarah Mock on farmers as landowners and Ramachandra Guha on what 1947's jailed freedom fighters got right once they were handed ministries.

Timeline

The arc, briefly

India's farm machine, and the fund Kahn built alongside it, dated from the conversation.

1960sShip to mouth, then food secure. One-sixth of humanity goes from a ship-to-mouth existence to food security within a decade, even as the population explodes, on high-yielding varieties, fertiliser subsidies, the FCI and minimum support prices. Kahn calls the machinery a Rube Goldberg device that was good at calories, less so at nutrition.

1985-90DLF comes together and Gurgaon farmers sell. Kahn's picture of the most miserable man in India is the farmer who sold for a fraction of what the land became, which is why families hold the option value rather than consolidate.

2006Vishal asks in print whether retailers can fix Kolar. He runs an article in Business World on whether organised retail could intervene with price, quality and buy-back. Forty years of driving that highway and he still sees tomatoes dumped from Kolar to Mulbagal.

2007-08Kahn settles in Mumbai. Graduate-school projects for ITC's agribusiness division pull him to India in late 2007, then years inside a domestic agribusiness where he watches the sector transform far too slowly while Flipkart-era digital change refuses to filter in.

2010Omnivore is founded. The word agritech does not yet exist in India. Godrej Agrovet comes in as the fund's first anchor, on a bet that a digital storm was arriving and there was no ecosystem to catch it. One of the country's most senior agribusiness figures used a panel to rubbish the whole idea.

2017Still too small to fund. Until about five years before this conversation, Kahn says, Indian agritech was small, peripheral and weird enough that there was simply no money in it.

2022The device has outlasted its assumptions. The PDS was designed before credit cards or digital identity, so ration shops were the only mechanism available. Kahn's point is that the country no longer needs to deliver welfare that way, and that every surviving programme has a bureaucracy eating off it. Punjab collects roughly ₹6,000 crore a year from its mandi system, largely unaccounted, and that was part of what the Delhi protests defended.

TodayThree funds, 33 companies, and an argument for a cheque. Kahn wants welfare decoupled from price: PM-Kisan pays ₹6,000 a year, a number he treats as arbitrary against ₹60,000 or a ₹1.5 lakh floor if the other programmes were folded in. Platforms come before robots, because you cannot sell a smallholder half a million dollars of kit.

Takeaways

Ideas to carry out of this hour

India's farm machine worked, then outlived its assumptions

Kahn refuses to start from failure. One-sixth of humanity went from a ship-to-mouth existence in the 1960s to food security within a decade even as the population exploded, delivered by high-yielding varieties, fertiliser subsidies, the FCI and minimum support prices. A Rube Goldberg device, in his phrase, that was pretty good at making sure people had calories, if not nutrition. The PDS was designed before credit cards or digital identity, so physical ration shops were the only mechanism available. His complaint is not that any of it was wrong, but that like most large government programmes it has outlasted the boundary conditions it was created under.

▶ 12:29 ▶ 14:06 ▶ 14:51

Every obsolete subsidy has a constituency that eats off it

There is not a single reason in the world, Kahn argues, not to run the fertiliser subsidy as a direct benefit transfer, except that everyone in the system loots it. The same holds for the PDS and the FCI's wheat and rice purchases: successful programmes acquire massive bureaucracies and stakeholders who will not let them go. He reads part of the Delhi farmer protests the same way: alongside genuinely legitimate grievances sat the roughly ₹6,000 crore a year Punjab collects from its mandi system, money he says goes largely unaccounted.

▶ 16:08 ▶ 16:24 ▶ 17:41

Deliver welfare as welfare, not as a price

India's method of helping farmers has been to subsidise inputs and inflate outputs, passing the surplus along indirectly. Kahn's objection is that using price as the delivery mechanism produces Soviet-style decisions: fertiliser is cheap, so more goes on than the crop needs; rice is supported, so it gets grown a thousand kilometres from Delhi on groundwater the region cannot spare, with the field burning that follows. Decouple the two, cut the farmer a cheque and let them pick bio-fertiliser, organic or permaculture, and the externalities stop being subsidised.

▶ 20:50 ▶ 21:21 ▶ 21:38

PM-Kisan's ₹6,000 is an arbitrary number

Kahn credits the union government for creating a direct farmer transfer at all, following experiments that began in Andhra Pradesh and Odisha. But he treats the amount as a policy choice rather than a constraint: nothing says it must be ₹6,000 a year rather than ₹60,000, or a lakh, if the other programmes were folded into it. He is willing to go further still, to a ₹1.5 lakh floor and a universal basic farmer income, provided it replaces the price-based machinery instead of sitting on top of it.

▶ 21:55 ▶ 22:11 ▶ 30:27

Land reform is an information problem first

The instinct that Indian farms are simply too small is only half right. Pakistan never did land reform and still has true zamindari estates of 10,000 and 20,000 hectares, yet India generally beats it on yields; China's holdings are smaller than India's and beat India's. What actually blocks consolidation is fear that an informal arrangement hardens into lost title once someone cuts a deal with the village patwari, plus the option value of land, the reason nobody wants to be the Gurgaon farmer who sold in 1990. His answer is a layer of digital ownership recording who owns, who cultivates and on what terms, which enables viable units without anyone surrendering the upside.

▶ 7:39 ▶ 8:52 ▶ 10:06

The farmer's poverty trap is one counterparty wearing three hats

The typical smallholder is in debt to a moneylender who also supplies them less-than-great inputs and also buys back their crop. It is a loop, in Kahn's telling, straight out of the development-economics literature on poverty traps. Agritech platforms attack it hat by hat: market the produce for more than the local trader or market yard pays, then supply better inputs at better prices rather than the highest-margin ones, then move the farmer towards low-cost institutional finance. The prize is the super-normal profit currently taken by intermediaries, which is why he expects these businesses to disrupt the arhtiyas and the large agribusinesses at the same time.

▶ 23:12 ▶ 23:41 ▶ 24:28

India needs a platform layer before it needs robots

American farmers are effectively professional agronomists, so the low-hanging fruit there is gone and gains come only from pushing the technology frontier. With the workforce ageing into its 50s and 60s and leaning on migrant labour, that frontier looks like autonomous tractors and robotic pickers. India's inefficiency is its advantage: a little better agronomy lifts yields and profits 20-25%, which is platform work, not frontier work. Kahn backs a farm-robot company built for Indian conditions, but expects it to scale by riding on agritech platforms, because you cannot walk up to a smallholder and sell half a million dollars of equipment.

▶ 37:43 ▶ 38:15 ▶ 40:46

The numbers, drawn

What the episode measures

Every figure below was said on air, with timestamps included and caveats kept.

How India's cropland gets its water

% of cultivated area

Rain-fed63%

Irrigated37%

From the host's opening framing: a cultivated area of 195 million (the caption garbles the unit), 63% of it rain-fed against 37% irrigated.▶ 1:01

What a farmer's cheque could be

₹ per farmer per year

As argued on air: PM-Kisan pays ₹6,000 a year, a number Kahn calls arbitrary. Scrap the other programmes and it could be ₹60,000, or a ₹1.5 lakh floor under farm incomes.▶ 21:55

One counterparty, three hats

the smallholder's loop, and how a platform prises it open

Kahn's reading of the poverty trap, which he says comes straight out of the development-economics literature. The order matters: sell the crop above what the local trader or market yard pays, then supply inputs on price rather than margin, then move the farmer to institutional credit.▶ 23:12

Why one country gets robots and the other gets platforms

where the next yield gain has to come from

Kahn backs a farm-robot company built for Indian fields, but expects it to reach farmers on top of agritech platforms rather than through a direct sale. The American frontier is where it is because the easy agronomy gains there were taken decades ago.▶ 38:15

Worth keeping

Lines that stay

Before we kick the green revolution's ass, let's take a step back and congratulate ourselves a little bit. India went from a ship-to-mouth existence in the 1960s to, within a decade, becoming food secure — even as the population exploded.

▶ 12:29

These things that were very successful have massive bureaucracies. They're not required anymore, but everyone wants them to continue — because everyone eats off of them.

▶ 16:08

What we have done to deliver farmer welfare is subsidise inputs and inflate outputs. There's another way to do it: cut him a goddamn cheque. Just decouple it.

▶ 21:38

I don't think there's any person more miserable than a farmer from Gurgaon who sold their land in 1990.

▶ 10:06

You need that platform layer to get built for deep tech to be viable in an Indian context. Unlike America, you can't go to a smallholder farmer and sell them half a million dollars worth of goods.

▶ 40:46

Clips that travel

Short on time? Start here

City families still holding ancestral farmland

Land reform without the zamindars

Pakistan's 10,000-hectare estates, China's smaller plots, the fear of the village patwari, and the option value that keeps land unsold.

7:04 → 11:28 · 4 min ▶ Watch clip

Anyone who reads Indian farm policy as pure failure

Congratulate yourselves before you tear it down

One-sixth of humanity going food secure in a decade, and the Rube Goldberg device that delivered calories if not nutrition.

12:10 → 15:37 · 3 min ▶ Watch clip

Policy readers arguing about MSP, DBT and the mandis

Cut the farmer a cheque

The thesis at its sharpest: why using price to deliver welfare ends in subsidised rice drinking Delhi's groundwater.

20:00 → 22:42 · 3 min ▶ Watch clip

Economists sizing up cash-transfer schemes

Universal basic farmer income

Kahn's willingness to put a ₹1.5 lakh floor under farm incomes, if it buys an end to input subsidies and their externalities.

29:24 → 31:31 · 2 min ▶ Watch clip

Agritech founders selling to smallholder farmers

Why India gets platforms and America gets robots

Ageing US farmers and migrant labour against India's 20-25% agronomy upside, and why farm robots need a platform to ride on.

37:28 → 41:16 · 4 min ▶ Watch clip

Glossary

The jargon, unpacked

MSP (Minimum Support Price)

The floor price at which government agencies buy notified crops, and part of the price-based welfare machinery Kahn wants replaced by direct cash.

FCI

The Food Corporation of India, the state buyer and warehouser of wheat and rice that gives the minimum support price its teeth.

PDS

The Public Distribution System of ration shops. It was designed, as Kahn points out, in an era before credit cards or digital identity made any alternative possible.

DBT

Direct Benefit Transfer: paying a subsidy as money into the beneficiary's bank account instead of distorting the price of the thing being subsidised.

PM-Kisan

The union government's income-support scheme paying enrolled farmers ₹6,000 a year directly. The model Kahn wants scaled up until it replaces everything else.

Mandi (APMC yard)

The regulated wholesale market where farmers must sell notified produce; Punjab's system raises around ₹6,000 crore a year for the state, which Kahn says is why it gets defended.

Arhtiya

The commission agent in the mandi who buys the farmer's crop and often lends against it, and one of the intermediaries agritech platforms try to displace.

Agri stack

Public digital infrastructure for agriculture (farmer, land and crop records) following India's payments and logistics stacks; Kahn says it is already under way.

Connections

If this resonated, go here next

EP 159 · Deep tech

Full transcript

The whole conversation, searchable

Auto-generated captions, lightly cleaned. Click a timestamp to open that moment on YouTube.