How the Green Revolution Built the Rent Stack, 1940-1970

The Package, As Shipped

The object historians call the Green Revolution was not a single breakthrough. It was a bundle. Four components moved together: semi-dwarf hybrid or pure-line cereal varieties bred for response to heavy fertilisation; synthetic nitrogen and phosphate at rates unprecedented in tropical smallholder systems; controlled irrigation through tube wells and canal expansion; and a chemical pest-and-weed-control programme of organophosphates, organochlorines, and later herbicides. Each component made the others load-bearing. Semi-dwarf wheat without fertiliser yielded less than the tall landraces it replaced; heavy fertiliser without irrigation burned the crop; irrigated monoculture without pesticide collapsed under pest pressure (Perkins, Geopolitics and the Green Revolution 1997; Cullather, The Hungry World 2010). The yield arithmetic and the input arithmetic were the same arithmetic.

The institutional timeline is short and specific. The Rockefeller Foundation established the Mexican Agricultural Program (MAP) in 1943 at the invitation of the Mexican government, staffed initially by plant pathologist Elvin Stakman, soil scientist Richard Bradfield, and horticulturalist Paul Mangelsdorf, with J. George Harrar directing field operations from Chapingo (Rockefeller Foundation Annual Report 1943; Fitzgerald, Exporting American Agriculture 1986). Norman Borlaug joined the programme in 1944 and ran the wheat-breeding work that produced the Pitic 62, Penjamo 62, and later Sonora 63 and Sonora 64 semi-dwarf varieties. The MAP became the organisational precursor of CIMMYT (Centro Internacional de Mejoramiento de Maíz y Trigo), formally constituted in 1966 and joined to CGIAR at the latter's founding in 1971 (CIMMYT founding charter 1966; CGIAR constitutive documents May 1971).

The Ford Foundation entered the architecture in 1951 with a grant to India, scaled dramatically in the 1960s through the Intensive Agricultural District Programme (IADP). IADP concentrated package-based inputs in seven initial districts and later fifteen, under the directorship of Sherman Johnson and with Ford field advisers embedded in Indian state agricultural services (Ford Foundation Annual Reports 1960-1968; Brown, Seeds of Change 1970). The International Rice Research Institute (IRRI), jointly established by Rockefeller and Ford in 1960 in Los Banos, Philippines, did for rice what Mexico had done for wheat: IR8, released in 1966, became the single-variety engine for rice-yield growth across Asia (Chandler, An Adventure in Applied Science 1982; Dalrymple, Development and Spread of High-Yielding Rice Varieties 1986). USAID underwrote the diffusion through PL 480 food-aid conditionality and technical-assistance agreements that tied recipient governments to the package as a matter of foreign-policy alignment (Cullather 2010).

In May 1971 the World Bank convened the founding charter of the Consultative Group on International Agricultural Research. CGIAR inherited CIMMYT, IRRI, IITA (Nigeria, 1967), and CIP (Peru, 1971), with founder funders including the World Bank, the Rockefeller Foundation, the Ford Foundation, the FAO, the UNDP, USAID, and Canadian CIDA (Baum, Partners Against Hunger 1986). By the late 1970s the federation had expanded to thirteen centres. The institutional design was not neutral. The research questions CGIAR was set up to fund were questions that presupposed the package. Alternatives, landrace improvement without synthetic inputs, nitrogen-fixing intercrop systems, pastoralist seed networks, received funding of a different order of magnitude.

The Yield Gains Were Real

A structural critique is only credible when it starts with what actually happened at harvest. What happened, across four adopting regions in the first two decades of the package, was that cereal yields rose faster than population growth for the first time in the modern history of those regions. Indian wheat production rose from approximately 11 million tonnes in 1960 to approximately 55 million tonnes by 1990; rice production rose from approximately 35 million tonnes to approximately 110 million tonnes over the same period (FAOSTAT production data 1960-1990, cross-checked with Indian Agricultural Statistics at a Glance, Ministry of Agriculture, multiple editions). Mexican wheat yields approximately doubled between 1948 and 1970 under MAP varieties (Rockefeller Foundation Annual Report 1970; Perkins 1997). Pakistani wheat yields nearly doubled across 1965-1970 as the IADP-descended package reached Punjab operators. Filipino rice yields rose on the back of IR8 diffusion from approximately 1.2 tonnes per hectare in 1965 to over 2.5 tonnes per hectare in the best-adopting districts by the late 1970s (Dalrymple 1986; IRRI Annual Reports 1966-1980).

The Malthusian prediction that hundreds of millions of people would die of starvation during the 1970s and 1980s, set out most influentially in Paul Ehrlich's The Population Bomb (1968), did not materialise. Borlaug received the Nobel Peace Prize in Oslo in December 1970 for the breeding programme that anchored the Mexican and Indian wheat gains; his lecture that day explicitly framed the yield breakthrough as a buying time against population growth, not a solution (Borlaug, Nobel Peace Prize Lecture, December 11, 1970). The famine prevention deserves to be named as a genuine accomplishment. A naturalist counting surviving bodies cannot call this arithmetic small.

The qualifier, which the counter-frame section develops, is that aggregate tonnage says nothing about who harvested it, who kept the margin, and what the package cost the operator who planted it. A region whose total cereal output triples while two-thirds of its smallholders are pushed off their land has experienced an aggregate gain and a distributional capture at the same time. The arithmetic of the next two Moves is what happens to that gain on the balance sheet of the operator who made it happen.

What the Package Did to the Operator's Ledger

The package's second arithmetic ran on a different clock. Hybrid seed does not breed true. The F1 vigour that gave the semi-dwarf wheats and the IR-series rices their yield response degrades in the F2 generation, so the operator who saves seed loses yield and has to repurchase. Matched chemistry is sold on the same cycle: the herbicide programme designed around the variety, the insecticide programme designed around the monoculture, the foliar feed designed around the heavy N rate. The 1943-1971 package did not invent the repurchase cycle. It scaled it. The fertiliser demand curve was even starker. A smallholder applying 10-20 kilograms of nitrogen per hectare on a traditional landrace does not need much of an input-manufacturing industry. A Punjab operator applying 150-250 kilograms of nitrogen per hectare on IR8 or Sonora 64 needs the international fertiliser trade at the door every season (FAO fertiliser consumption series 1965-1990; International Fertiliser Association production data).

The incumbent input manufacturers of the period were positioned to meet that demand. BASF, Bayer, Hoechst, and ICI dominated European agrochemistry; Dow Chemical, Monsanto, DuPont, and Standard Oil subsidiaries dominated the American side; Japanese Sumitomo and Nippon Soda expanded through Asian distribution (Hayami and Ruttan, Agricultural Development 1985; Agrow Reports industry histories). The commercial alignment was not conspiratorial. It was structural. The varieties CIMMYT and IRRI released were varieties that responded to the inputs these firms already produced. Research programmes at the centres interacted continuously with chemistry-industry representatives because the package required chemistry-industry inputs to function at yield. The funding flows match. The Rockefeller Foundation's agricultural programme commitments from 1943 to 1970 ran into hundreds of millions of dollars of period-specific value, with Ford, USAID, and World Bank co-funding scaling the architecture by an order of magnitude by the early 1970s (Rockefeller Foundation financial reports; Ford Foundation Annual Reports; US AID mission archives; World Bank agricultural lending series 1965-1975).

The operator-level consequences compounded on a twenty-to-forty-year horizon. Debt-treadmill dynamics are documented across adopting regions: Punjab and Haryana operators who prospered in the first decade of the package accumulated input-credit obligations and irrigation-capital obligations that transformed them from principals of their own production into agents of input suppliers, commission agents, and nationalised banks (Shiva, The Violence of the Green Revolution 1991; Singh, Punjab Agriculture, multiple editions; National Sample Survey Organisation household debt data). Groundwater tables across the Indo-Gangetic Plain fell approximately 10-30 metres across the 1960-2010 period in the highest-extraction districts (Tiwari et al., GRACE satellite analysis, Geophysical Research Letters 2009; Central Ground Water Board of India Annual Reports). Eutrophication of surface water under heavy nitrogen applications, pesticide-resistant pest complexes (brown planthopper in Asian rice, bollworm in Indian cotton), and a long pesticide-health-cost tail are documented in the FAO International Code of Conduct on Pesticide Management review series and in multiple epidemiological studies out of the Punjab Agricultural University (Thapar and Jaswal 2018 synthesis).

Landrace diversity collapsed on a timescale that matters for long-run resilience. The FAO Commission on Genetic Resources for Food and Agriculture, in its State of the World's Plant Genetic Resources reports (1996, 2010), documents diversity losses across the major cereals of adopting regions in the range of 80-95 percent of pre-1950 varietal baseline. Nikolai Vavilov's 1920s-1930s Soviet collections had catalogued thousands of distinct wheat, rice, and maize landraces across the centres of origin; by the 1990s, the rest survived only in gene-bank storage (Fowler and Mooney, Shattering: Food, Politics, and the Loss of Genetic Diversity 1990; Mooney, The Law of the Seed 1983). The genetic floor under the package's yield gains narrowed at roughly the same speed the yield gains compounded.

These consequences are not character failures of the operators who adopted the package. A Punjab smallholder in 1967 who planted Mexican wheat got a yield response that paid for the nitrogen. The smallholder's son in 1987 inherited the nitrogen demand curve and the tube-well depreciation schedule without the 1967 marginal-return-to-inputs. The package was, by then, the default. An operator looking at that ledger sees that 35-50 percent of variable cost now leaves before a crop is harvested, which is the arithmetic the input sovereignty analysis and the seed sovereignty analysis decompose for the present.

The Package Did Prevent Famine. It Also Captured the Ledger.

Three counter-frames have to be carried side by side for the argument to survive contact with the evidence. The first is the Ehrlich-rebuttal frame. The Population Bomb (Ehrlich 1968) predicted imminent mass starvation across South Asia and Latin America; the prediction failed. Hundreds of millions of people who would have been expected to die under the 1968 projection lived, ate, and raised children. A writer who skips past that fact in order to get to the structural critique has not taken the evidence seriously. The package prevented a famine that was on a demographic trajectory to occur. Any subsequent claim has to enter the room through that door.

The second counter-frame is the distributional critique, documented by three lineages of scholarship. Frances Moore Lappe and Joseph Collins's Food First: Beyond the Myth of Scarcity (1977), arguably the founding text of the food-sovereignty critical tradition, argued that the Green Revolution's aggregate yield gains concentrated with larger operators who could access credit, irrigation capital, and input-supply chains, while smallholder operators who could not finance the package either exited farming or remained at traditional yield levels while facing the prices generated by package-adopting neighbours (Lappe and Collins 1977; Lappe, Collins, and Rosset, World Hunger: Twelve Myths, 1998 revised edition). Vandana Shiva's The Violence of the Green Revolution (1991), grounded in Punjab fieldwork, documented the social and ecological consequences of the package at regional scale: women-farmer displacement, smallholder capitalisation-gap, groundwater depletion, pesticide-health burden, and a political radicalisation in the 1980s Punjab insurgency that Shiva argued traced back to the economic squeeze of the input-debt cycle. Raj Patel's Stuffed and Starved (2007, 2012 revised) extended the critique into the global food system and developed the argument that the same package that prevented famine in Asia underwrote the consolidation of the trading and processing architecture that governs world food supply today.

The third counter-frame is the measured rent-stack-capture analysis, which is the one this page extends. The claim is not that the Rockefeller Foundation, the Ford Foundation, or the scientists at CIMMYT and IRRI intended the 2026 Bayer Crop Science balance sheet. They did not. The claim is that the institutional architecture they built selected for a package that had incumbent-input-manufacturer alignment baked into its mechanism, and that over sixty years that alignment compounded into the present rent stack. The operator-level debt-treadmill, the genetic-diversity floor, and the ecological-externality profile are consequences of design choices in 1943, 1960, and 1971 that did not contemplate sixty years of compounding. Structural outcomes do not require structural intent. Institutional incentives are sufficient.

The three frames are not alternatives. The package prevented famine. The distributional capture was real. The rent-stack descent was the institutional design operating as it was designed to operate.

Why the Bundling Was Not Separable

The package's four components were technologically novel in combination, and that combination was commercially aligned with incumbent input manufacturers. These two facts are not independent. They are the same fact seen from two angles. A breeding programme designed around fertiliser response produces varieties that require the fertiliser industry. A breeding programme designed around drought tolerance in low-input systems would have produced varieties that required different industry alliances or no industry alliance at all. The Rockefeller Foundation's Mexican Agricultural Program from 1943 forward, the International Rice Research Institute from 1960 forward, and CGIAR from 1971 forward worked on the first kind of programme. They worked on that programme because that was what the funders were set up to fund (Cullather 2010; Perkins 1997; Fitzgerald 1986).

The contemporaneous alternative was not hypothetical. In the 1920s and 1930s Nikolai Vavilov's Institute of Plant Industry in Leningrad had built the world's largest collection of crop landraces, organised around centres of origin and adapted to low-input traditional cropping systems (Pringle, The Murder of Nikolai Vavilov 2008; Nabhan, Where Our Food Comes From 2008). Vavilov was arrested by the Soviet regime in 1940 and died in Saratov Prison in January 1943, the same year the Mexican Agricultural Program was founded. His collection survived the siege of Leningrad because his staff refused to eat the seeds. An international agricultural research architecture built on Vavilov-style landrace improvement in situ with farmer-breeding partnerships would have produced different varieties, different input-demand curves, and different industry alliances. That architecture did not get built. The architecture that did get built was the one that fit within the funding constraints of American private-foundation philanthropy under early-Cold-War geopolitics (Cullather 2010; Fitzgerald 1986; Perkins 1997).

The present-day rent stack is the institutional descendant of those choices. Bayer Crop Science, Corteva, Syngenta Group, and BASF together control approximately 60 percent of the global commercial seed market; in the United States, 70 to 80 percent of maize and soy acreage is planted to patented traits (USDA Economic Research Service 2024). Nutrien, Yara, Mosaic, and CF Industries dominate global fertiliser supply; OCP Morocco holds approximately 70 percent of world phosphate rock reserves (US Geological Survey Mineral Commodity Summaries 2024). John Deere holds approximately 53 percent of US large-tractor share (Association of Equipment Manufacturers 2024). ADM, Bunge, Cargill, and Louis Dreyfus handle approximately 70-90 percent of global grain trade depending on crop (IATP 2022; Murphy and Burch 2021). Each of these concentrations has a lineage that traces back to the demand curve the Green Revolution package created, the monoculture uniformity the package required, and the institutional research agenda CGIAR stabilised. The concentration is not a corruption of the package. It is the package's mature state.

The counter-response documented at Navdanya in twenty-two Indian states, the MASIPAG farmer-breeding network in the Philippines, and the seed-sovereignty movements that trace through La Via Campesina and the 1996 World Food Summit, are not nostalgia for a pre-package agriculture. They are operator-level reconstructions of the architecture the 1943-1971 choices excluded. Open-pollinated varieties that reproduce. Mixed-variety fields that carry their own genetic insurance. Locally adapted breeding in partnership with the farmers who grow the crop. The arithmetic those reconstructions run is not traditional. It is the arithmetic of a balance sheet that does not invoice the operator for seeds that used to reproduce and soil biology that used to work for free. Biochar, mycorrhizal networks, on-farm composting, and nitrogen-fixing rotations deliver, at the mechanism level, what the Green Revolution package chose not to develop because the funding architecture was not set up to fund it.

The structural balance-sheet argument is the companion diagnosis for the present tense. What this page adds is the history. The 2026 rent stack did not emerge from nowhere. It emerged from a package that was bundled, funded, and institutionalised across three decades in the middle of the last century, by foundations and agencies that were solving the problem they were set up to solve, using the instruments their funding structure permitted. The famine that was prevented deserves its place in the record. The balance sheet that was captured deserves its place in the same record. The Green Revolution solved a yield problem by creating a balance-sheet problem. The balance sheet was the point.