Daniel Sumner on the Political Economy of Agriculture
What this covers
Daniel Sumner, an agricultural economist at UC Davis, joins Russ Roberts to trace the political and economic anatomy of U.S. farm subsidies from the New Deal to the present. Rather than functioning as poverty relief—their original justification—Sumner argues that these programs have become primarily wealth transfers to large landowners and agribusiness operators. The conversation maps how subsidies actually work: by raising crop prices, they necessarily benefit larger producers most, capitalize into land values, and get absorbed by landowners in rental rates. Sumner walks through the mechanics of several policy eras—from New Deal parity schemes and legislated price floors through deficiency payments to modern crop insurance subsidies running $10–15 billion annually, where taxpayers effectively function as reinsurance providers for private insurers.
The substance turns on two interlocking insights. First, following Mançur Olson and Gary Becker, political power flows to organized coalitions with concentrated stakes. Industries that unite growers, processors, and handlers—cotton, sugar, dairy, corn—win subsidies; fragmented ones like cattle and pork do not. Subsidized farms show no measurable economic advantage over unsubsidized competitors like California vegetable growers or almond producers. Second, the strongest surviving rationale is simply inertia: subsidies persist because they have always existed, creating political unease about removal even among Sumner's economists. The conversation also addresses spillovers: how U.S. cotton subsidies depressed world prices enough to trigger a WTO ruling against America and a settlement with Brazil, and how the shift toward revenue insurance distorts planting decisions at the margin. Throughout, Sumner notes where received wisdom breaks down—on farm consolidation trends, on whether corn subsidies drive obesity, and on whether subsidies explain American agricultural success.
Sumner argues that U.S. agricultural subsidies, originally framed as relief for poor farmers, predominantly enrich large landowners and have no compelling economic rationale—surviving instead on political organization and the inertia of having always existed.
- Because subsidies raise output prices, benefits scale with farm size and capitalize into land values, so the largest farms and landowners gain most
- Industries that organize into single-voice coalitions (sugar, cotton, dairy) win subsidies while fragmented ones (cattle, pork) do not
- The strongest surviving justification is simply that subsidies have always existed, creating uncertainty about removing them
Farm subsidies primarily benefit large landowners and wealthy farm businesses, not poor rural people.
- The dramatic losers from farm programs are taxpayers and, to some extent, consumers—though the consumer effect is mixed since heavy corn-eaters (via cheaper hamburger and pork) can benefit when programs stimulate extra corn production.
“the dramatic losers of this program are you and me, who pay more--and taxpayers, generally.”
- Because farm programs raised crop prices to raise farming profitability, the more you produced the more subsidy you received—making it inherently an industrial policy where the biggest farms were the biggest beneficiaries by design, never functioning as poverty relief despite the original justification of helping poor rural people.
“if you produced a lot, you got a lot of subsidy out of the deal. And in that sense it was an industrial policy. So, the bigger the farm, the more benefit. And that was on purpose.”
- Because land is the inelastic resource in farming while labor and crop-switching are flexible, subsidies that raise the whole configuration of crop prices get capitalized into higher land rental rates and values, so the true long-run beneficiaries are landowners (who may live anywhere) rather than tenant farmers, whose higher revenue is offset by higher land rental costs.
“The farmer who is out there farming the land, say, renting everything he farms, gets essentially nothing out of this. He has higher total revenue and higher land rental costs. It's a wash for him.”
- Farm program subsidies now run a steady-state $15-20 billion per year covering about $300 billion of agriculture, representing roughly 6-10% of covered farm revenue (higher for cotton, corn, and wheat; lower for soybeans), transferred to businesses with average equities in the millions—some with $100 million or even a billion dollars of equity.
“We're now at a steady state between $15 and $20 billion dollars. That covers about $300 billion worth of agriculture.”
Price supports cause predictable overproduction cycles requiring government stockpiling and land-idling to manage.
- Setting legislated minimum prices above market-clearing levels predictably caused both excess production (farmers grow more) and reduced demand (buyers buy less), forcing the government to become the buyer of last resort, acquire huge unsustainable stockpiles, and impose land-idling requirements to hold back production—a self-reinforcing cycle.
“the buyer of last resort was the government. So the government acquired a lot of corn”
- Over the last 30-40 years, farm policy shifted from government stockpiling to 'deficiency payments,' where the government set a target price for farmers and made up the difference between that and the market-clearing price via direct checks—allowing low consumer prices while guaranteeing high farmer prices and still stimulating production, functioning more like a traditional subsidy.
“the government said, 'I will set the market price for corn, but I will enforce that by making up the difference between the government-set price and whatever the market ends up being.'”
- Highly subsidized revenue insurance still stimulates additional production at the margin, especially where insurance is good for one crop but not another (e.g., wheat versus barley) or in outlying areas poorly suited to a crop—encouraging farmers to plant wheat in the Dakotas on land that might otherwise have stayed in pasture.
“I provide wheat in the Dakotas and maybe I might not have grown wheat: maybe that land would have stayed in pasture, but now I've got subsidized insurance, very highly subsidized insurance. So I grow additional wheat out in the Dakotas”
Trade barriers and subsidy design create concentrated beneficiaries wielding outsized political power.
- Cargill, a privately held global commodity-trading company, opposed both U.S. subsidies and export subsidies despite short-run benefits, arguing subsidies aren't good for world agriculture generally—a stance Sumner credits to private ownership freeing them from quarterly-report pressures, making them intellectual leaders among companies on farm policy.
“it's because they are not a publicly traded company they can--it's easier, some people argue it's easier for them to look at the long haul without worrying about quarterly reports”
- U.S. cotton subsidies—when farmers received roughly half their revenue as federal checks (a 35-cent world price versus a 70-cent guarantee) and the U.S. was the largest exporter of a homogeneous product—were ruled by the WTO to have depressed world prices and harmed Brazilian farmers; the U.S. lost twice on appeal and then paid Brazil about $150 million a year for about five years as 'hush money' to preserve the cotton program.
“the claim that U.S. cotton subsidies were big enough, they stimulated U.S. cotton production enough, to drive down world cotton prices enough that it reduced the revenue opportunities for Brazilian cotton farmers. The U.S. government lost that case.”
- Sugar is heavily subsidized primarily through trade barriers (a tariff-rate quota), making the business so lucrative that many sugar-beet farms wouldn't exist without it; concentrated cane sugar players like the Fanjuls wield outsized political clout as identifiable major donors while thousands of beet growers across many congressional districts provide broad electoral pressure.
“Sugar is wonderfully well subsidized. If you like political economy stories through trade barriers, primarily.”
- The 'freedom to farm' transition payments became permanent 'direct payments'—a roughly $5 billion lump-sum-style account economists could love because it minimally distorted production—but its very transparency made it politically vulnerable, and it was eliminated in the 2014 Farm Bill as legislators preferred payments tied to (and thus distorting) incentives over checks for 'nothing.'
“it had the disadvantage from [?] point of view that it was immensely transparent. And it's remarkable to me that it lasted almost 20 years in Washington.”
- As U.S. farms grew, average managerial ability rose because flexibility allowed consolidation, so talented people found their best alternative was running large sophisticated farms rather than leaving—contrasting with Korea and Japan where U.S.-imposed land reform capped farm size, drove the smart brothers off the farm, and kept farm income low.
“it was often the case that the smarter brother stayed on the farm”
- While the theory that corn subsidies make Americans fat (cheaper corn makes cheaper hamburger that makes people fat) is logically coherent, the empirical facts do not fit that story at all.
“It turns out the facts don't fit that story at all. In theory it could be true.”
- Despite media attention in college towns and urban settings, local, artisan, and organic farming remain a tiny part of agricultural production; most organic and cage-free output comes from large diversified operations (e.g., a 1000-acre lettuce grower with a 100-acre organic division) rather than small producers, because retail-scale farming makes a good living very difficult.
“the significant organic producers are the guy with 1000 acres of lettuce, gross revenue of millions and millions of dollars who may grow 100 acres of organic lettuce”
- The romantic American Gothic image of farming—a person in overalls with a pitchfork—was never accurate; farming has long been a business, and over generations the U.S. went from perhaps 10 million people making a living farming to a few hundred thousand, plus a million-plus families who farm part-time.
“And it really was never true.”
Subsidized industries show no competitive advantage over unsubsidized ones across commodities and countries.
- The dramatic rise in agricultural productivity and the increasing concentration into larger farms over 1950-present were occurring simultaneously with the subsidy programs but were not caused by them—farm subsidies didn't make farms larger and farm consolidation didn't create the subsidies; they merely interacted politically.
“the farm subsidies didn't make farms larger; and farms getting larger didn't make the farm subsidies. I think they were going along at the same time.”
- California vegetable and almond growers, who lack significant subsidy programs, do fine and see no evidence that subsidized corn or cotton farmers are unusually wealthy by comparison—arguing the complications and impediments subsidy programs place on innovation and organization can cost more than they are worth.
“We don't see that cotton farmers have done a whole lot better than almond farmers, for example.”
- Across both commodities and countries, there is no record of subsidized industries (cotton, corn) achieving greater success than unsubsidized ones (cattle, lettuce, broccoli), so subsidies lack a compelling economic argument while retaining a compelling political one rooted in Mancur Olson's collective-action logic.
“there is no record of success that is higher in the cotton business than in the corn business, relative to, say, the cattle business or the lettuce business or broccoli.”
Political durability of farm programs rests on collective-action coordination and deliberate program opacity.
- After dismissing rationales like food security, the most successful surviving justification for farm subsidies—Sumner's 'bakers' dozen' reason number 13—is simply that we have always subsidized agriculture; multi-generational inertia makes farmers genuinely uneasy about turning programs off, and Sumner concedes this is politically compelling even though he believes it shouldn't be.
“We subsidize agriculture because we've always subsidized agriculture. And that's not as facetious as it sounds. That is to say there's lots of inertia.”
- The complexity of programs like the peanut program and milk marketing orders is partly deliberate; as a North Carolina Peanut Association official reportedly said, only two people in America understand how the peanut program works and it was his job to keep it that way—an opacity that insulates programs from scrutiny.
“There's only two people in America who understand how the peanut program works. It's my job to keep it that way.”
- Commodity groups succeed politically by uniting growers, processors, buyers, ginners, and warehousemen to 'speak with one voice,' settling internal disagreements privately; industries that achieve this (cotton, milk, corn) get subsidies, while fragmented industries that cannot paper over regional and segment differences (cattle, pork) never secure comparable programs.
“The National Cotton Council includes cotton millers and cotton buyers and cotton ginners and cotton warehousemen as well as cotton growers. Once you get all of those together, the Washington lingo is: We speak with one voice as an industry.”
Modern crop insurance is heavily subsidized, shifting insurer losses to taxpayers during widespread-loss years.
- Crop insurance historically failed due to moral hazard and adverse selection, keeping premiums high and uptake low; in the 1990s the government sharply raised premium subsidies on the logic—familiar from health insurance—that subsidizing insurance draws in more participants and dilutes adverse selection, then created the Risk Management Agency and a government-owned corporation to set rates while private companies sell the policies.
“should be familiar to people thinking about health insurance--that, gee, if we really subsidize the insurance, that will draw in more people, so we won't have as much problem with adverse selection, moral hazard.”
- Modern crop insurance costs taxpayers $10-15 billion a year via premium subsidies (farmers pay roughly 40% or less), payments to private insurers for operating costs, and—critically—reinsurance whereby taxpayers absorb the insurance companies' losses in widespread-loss years like the 2012 Midwest drought, making the taxpayer the de facto reinsurer rather than a private firm like Swiss Re or AIG.
“in the case of widespread losses, the taxpayers pick up any losses that the insurance companies have. So, we had a big drought in 2012 in the mid-West; the insurance companies would have lost a bunch of money; the taxpayer provided what in the jargon is 're-insurance.'”
- After eliminating the milk price support and export subsidy in spring 2014, the government replaced them with milk margin insurance run directly by the government (not a federally owned corporation or private insurers), offering highly subsidized coverage on the difference between an index of feed cost—half to two-thirds of dairy farm costs—and an index of milk prices.
“The government itself has something called milk revenue insurance that protects the farmers against the margin. So, it's margin protection.”
New Deal farm programs created lasting institutional structures that persist with evolutionary change.
- While agricultural subsidies date back centuries, the major U.S. farm programs began in the New Deal as part of Roosevelt's response to the Great Depression, some initially declared unconstitutional before being reinstated with edge tweaks, and have persisted in roughly that form ever since with evolutionary changes.
“here in the United States, the major farm programs, what we identify as farm programs, really started in the New Deal”
- Because the government promised to buy storable dairy products (cheese, butter, dry milk powder) rather than raw milk, it accumulated caves full of government-owned cheese outside Kansas City along with piles of dry milk powder, with the high fat content of cheese occasionally causing it to catch fire.
“outside of Kansas City, caves full of government-owned cheese at one point. As well as piles and piles of dry milk powder.”
- The New Deal used the concept of 'parity,' pegging farm product prices to their 1910-1914 levels relative to the things farmers buy, requiring a whole configuration of legislated prices by grade, quality, type, and location since a single national average price ignored real market variation.
“if we take the period from 1910-1914, that was sort of normal and farmers were doing fine. We'll call that the 'parity price.'”
- World War II did not solve agriculture's underlying problem of low commodity prices; it merely relieved the symptoms temporarily by pulling labor into the war effort and boosting export demand as Europe stopped producing, leading forecasters to reasonably predict the same depression-era problems would return after the war.
“WWII didn't solve any problem; what it did do was relieve the symptoms for a little while. We're going to have the same problem when we come back.”