Why Food Stamps Live in the Farm Bill and Whether That Marriage Should Last

Ask a stranger what a farm bill is for and they will say farms. Reasonable. Then tell them that the bulk of the money in the farm bill goes to grocery assistance for families in cities, suburbs and small towns, and watch them check their pocket to see whether you are joking. You are not. The "farm bill" is a marriage of two programs that have almost nothing in common except a committee room, a calendar and a very old political bargain.

I want to explain how that marriage came about, why it keeps producing trouble, and whether conservatives should defend it, annul it, or negotiate a separation. I will give you the best case for keeping them together, the best case for splitting them, and then my verdict. You are free to disagree with the verdict. You should not be free of the facts.

How we got here

Food assistance in this country began as a way to move surplus. In 1939 the Agriculture Department started a food stamp program that let people on relief buy food with stamps, partly to help the unemployed and partly to help farmers by raising demand for their crops. It ended in 1943, when the war ended unemployment and farm surpluses.

In 1961 President Kennedy started pilot projects, and in 1964 Congress passed the Food Stamp Act, signed by President Johnson that August, which made the program permanent. Notice who ran it. Food stamps lived at the Department of Agriculture, because the department that managed the surplus was the department that managed the stamps. The logic was farm-first: food aid was a way to use up what farmers grew.

Through the 1960s the program was small and voluntary for counties. A county could choose to take part. Then, in the early 1970s, two things came together.

First, pressure grew to expand food aid. A Senate committee led by George McGovern of South Dakota drew attention to hunger, and Republican Senator Bob Dole of Kansas became an unlikely ally of McGovern on the subject, a bipartisan alliance that mattered for years afterward. Second, farm legislation needed votes. A farm bill that served only farm interests had a problem, because the number of farm districts in Congress was shrinking as the country urbanized.

The solution was a trade. Urban members would vote for farm commodity programs. Rural members would vote for food stamps. The nutrition program was carried by the farm bill, and by the mid-1970s food stamps were a permanent, expanding passenger on the same train. The 1973 farm bill is often cited as the key moment, since it widened the program and pushed it toward every county in the country. The Food and Agriculture Act of 1977 then eliminated the requirement that recipients buy their stamps, a reform that opened the program wider. Political scientists call this logrolling: you vote for my bill, I vote for yours. It is as old as legislatures. It has also lasted fifty years, which is unusually long for a bargain made in a hurry.

In 2008, the food stamp program was renamed the Supplemental Nutrition Assistance Program, or SNAP, in that year's farm bill. The stamps had by then become electronic cards, and the name finally caught up with the paperwork.

Why the five-year bill keeps stalling

The farm bill is meant to be rewritten about every five years. In practice, it has become a hard bill to pass, and the reason is the marriage.

The modern pattern goes like this. The farm-state members want commodity programs and crop insurance. Some of them dislike the cost of food aid. The urban members want food aid and do not love farm subsidies. Each side's bill is only as strong as the other side's votes. When either side's support weakens, the whole package wobbles.

Congress treats the expiration date on a farm bill the way a college student treats a paper deadline: as a suggestion to be revisited after one more extension. It has dealt with this repeatedly by doing nothing: it lets the old law expire and passes an extension to buy time. The 2012 farm bill authority expired and was extended into 2013. The 2018 law was due to run out in 2023, and Congress extended it rather than rewrite it. There is a built-in pressure to act in the form of permanent law: if the temporary provisions lapse, the 1938 and 1949 farm acts, which were never repealed, return, with price supports far above market levels. The threat of that "dairy cliff" is meant to focus minds. It has sometimes focused them just enough to pass another extension.

The most dramatic recent test of the marriage came in 2013. In June of that year, the House rejected the farm bill on the floor, in part because many Republicans thought the nutrition title was too generous and many Democrats thought proposed cuts to it went too far. House leaders then did something that had never been done in the modern era: in July 2013 they passed a farm-only bill with the nutrition title stripped out, and in September they passed a separate nutrition bill. The two halves never became law separately. The final bill, the Agricultural Act of 2014, put the two back together, with a smaller cut to nutrition. The experiment ended with the parts reunited.

The 2018 bill told a similar story. In May of that year the House version failed, partly because it included stricter work requirements for food aid and partly because of an unrelated fight on immigration. It passed a month later. The Senate version dropped the stricter requirements, and the final law, signed in December 2018, followed the Senate approach. Every time, the debate about what to require of food aid recipients nearly sank the support for farmers, and the farm programs nearly sank the food aid.

The best case for keeping them together

Here is the strongest argument I can build for the marriage, in a few numbered points, because it is a serious one.

1. The votes. A farm bill with no urban support does not pass. Rural America is a smaller share of the country than it used to be. Cut the link, and farm programs must survive on farm-district votes alone, which are a fraction of the House.
2. The shared interest. SNAP benefits get spent at grocery stores, which buy from processors, who buy from farmers. Demand for food is demand for farm products. The programs are not as unrelated as they look.
3. The stability. A single large bill is a negotiated package. Break it, and each piece gets picked off by its enemies.
4. The discipline. Putting both in the same bill forces farm-state conservatives and urban Democrats to look at each other's priorities. Compromise is hard to build any other way.

That is an honest argument and I do not think it is silly. In particular, point one matters. A divorce could leave the farmer worse off than the marriage did.

The best case for splitting them

And here is the other side, equally serious.

1. Honest scrutiny. The nutrition title is by far the largest part of the farm bill, according to the Congressional Budget Office's scoring, which you can check yourself on the CBO's website. When it sits inside a bill named for farms, it gets less independent attention. Every program deserves its own debate, on its own merits, with its own price tag in the headline.
2. Truth in labeling. If you call a bill a farm bill and most of the spending is something else, you are misleading the taxpayer about what he is paying for. A conservative should favor plain names.
3. The work question. Work requirements and eligibility rules for food aid are a legitimate policy fight. The 1996 welfare reform law limited able-bodied adults without children to a few months of benefits in a three-year period unless they worked or trained, with exceptions and waivers. Whether to tighten, loosen or enforce those rules should be settled on its own, not as the price of keeping a wheat program alive.
4. Farmers deserve better. A farmer's risk coverage should not depend on an unrelated fight about welfare. It should be debated by people who understand crop insurance, not held hostage.
5. Accountability. If a program is worth keeping, it can win its own vote.

Point four is the one that interests me most. In 2013, the farm programs were nearly sunk by a fight that had nothing to do with farming.

The verdict

I come down on the side of splitting them, with eyes open about the risk.

Here is why. The marriage made sense in 1973, when urban votes were the only way to pass a farm bill and the food aid program was young. It makes less sense now, for three reasons. The nutrition title has become the main event, and a conservative who believes in scrutiny cannot be comfortable with the biggest spending item riding in a bill named after the smallest. The work-and-eligibility debate deserves its own vote, with the members who care most about it voting on the question straight. And the marriage rewards stalling: since every reauthorization is a hostage negotiation, Congress prefers to extend rather than decide.

Yes, a divorce lawyer would call this a bold move, and yes, in Washington even the divorces come with a fifty-page rider. But the risk is real. If the farm title has to stand on its own, rural members must persuade a wider Congress that crop insurance and commodity programs deserve support. That is a harder sell than a trade. My answer is that it should be a harder sell. A program that cannot persuade a majority on its merits should be reformed until it can. The honest case for a safety net for farmers is strong, and I laid it out in From the 1933 Farm Act to Freedom to Farm, How the Safety Net Got Built. It does not need a hostage to survive.

And if the votes are not there for a clean split, the next best thing is to insist on separate treatment inside the bill: separate titles with separate scoring and separate floor votes. Make the cost of each visible.

What a split would and would not change

Let me clear up a misunderstanding, because it muddies the debate. Splitting the bill would not move food aid out of the Agriculture Department, and it would not take it away from the agriculture committees. Those committees, the House Committee on Agriculture and the Senate Committee on Agriculture, Nutrition, and Forestry, would still write the rules for SNAP. The money would still flow through the annual agriculture appropriations bill, which is a different track from the farm bill. What would change is the vehicle: two bills, two scores, two floor votes, and two sets of amendments.

Would that matter? I think it would, for the same reason that a restaurant with two separate checks gets a more honest conversation than one with a single bill and a "split it however you like" shrug. When the nutrition title is on its own, a member who wants to tighten work rules, or expand eligibility, has to say so on the record in a vote about food aid and nothing else. When the farm title is on its own, a member who wants to cap payments to the largest operations, or defend them, has to do the same. The marriage let both sides vote for the bill and blame the other half. A divorce takes away the excuse.

What a family should watch

If you are a farmer, watch for the next farm bill vote and ask how your crop insurance coverage fares, and read Hail Hits Your Wheat Field and a Crop Insurance Claim Moves to a Check to see how that coverage works in practice. If you are a taxpayer, ask your member what share of the farm bill is nutrition, and look up the CBO number yourself. If you want the budget mechanics behind all of this, How a budget becomes law, from the President's request to the checks explains where authorization ends and appropriations begin, and how a minority party fights a spending bill when it does not have the votes shows what a minority can do when the majority writes the bill. The 1990s welfare debate is told in divided government gave us welfare reform and a balanced budget, and the myths on both sides of the subsidy argument are tested in Eight Myths About Farm Subsidies That Both Parties Keep Repeating. When the money does run out, the question of who gets paid is covered in what a government shutdown really shuts down and who gets paid.

One more thing. Do not let anybody tell you this fight is about hating farmers or hating the poor. Both are caricatures. The real question is how to run two very different programs honestly, so that voters can see what they pay for and judge it.

The bottom line

• Food stamps entered the farm bill through a vote-trading deal in the early 1970s, after the 1964 Food Stamp Act. Today the nutrition title is the largest part of the bill by CBO's scoring.
• The marriage has made the five-year bill hard to pass, and Congress has relied on extensions to avoid decisions. The one real test, the 2013 split, ended with the pieces reunited.
• My verdict: split them, or at least give each its own title, score and vote, so that nutrition and farm programs each face scrutiny on their merits.
• Tell your representative that you want a farm bill with the real cost of each program on the label, and a vote on food aid rules that does not depend on the fate of crop insurance.