Why Spending Restraint Keeps Losing in Washington, Whoever Is in Charge

Conservatives keep winning the argument about spending and losing the budget. That should bother every one of us, and I will put it plainly, because our own fundraising letters will not.

We win elections on a promise to rein in spending. Republicans have held the House, the Senate and the White House, sometimes all at once. And yet the federal budget keeps growing, whoever is in charge. If votes decided it, the matter would have been settled long ago. Voters tell pollsters they want less waste. Candidates swear they will cut it. The cutters win the primary. Then they get to Washington, and something happens between the oath and the appropriations bill.

What happens is not a personality failure. It is not that "the swamp got to him," although the swamp is real. It is a structure, and anyone who wants to win this argument needs to understand it. The structure has a name, a book and a half century of evidence behind it. Once you see it, you will never again be surprised by a spending bill.

The logic: a few win big, everyone loses a little

In 1965, an economist named Mancur Olson published The Logic of Collective Action. His insight, simple enough to fit on a napkin, upended the way people thought about politics. The ordinary assumption was that a group with a shared interest will naturally organise to pursue it, and that a bigger group will have more power. Olson argued the opposite. Small, concentrated groups organise easily. Large, scattered groups do not.

The reason is that, in a small group, each member has a large stake and can see whether his neighbour is pulling his weight. In a huge group, each member's stake is tiny, and each has every reason to let someone else do the work. Why spend a Saturday writing to your congressman about a program that costs you the price of a cup of coffee?

Combine Olson with public choice theory, which I wrote about in Yes, Markets Sometimes Fail, and Governments Fail Right Alongside Them, and you get the engine of Washington's spending habit: concentrated benefits and diffuse costs. The people who receive a subsidy know it, care about it, and organise to defend it. The people who pay for it are every taxpayer in the country, none of whom pays enough to notice.

The arithmetic, with a hypothetical

Let me show you the numbers, and I want to be very clear that this is a hypothetical, with invented figures chosen to make the arithmetic simple.

Suppose Congress considers a subsidy of $60 million a year for a particular industry. Suppose that industry has 4,000 beneficiaries, whether growers, shipbuilders, or the operators of some obscure facility.

• Benefit per beneficiary: $60,000,000 divided by 4,000 equals $15,000 each, every year.
• Now suppose 120 million households pay the bill. Cost per household: $60,000,000 divided by 120,000,000 equals fifty cents each, every year.

Think about what each side will do. The beneficiary stands to lose fifteen thousand dollars. He will hire a lobbyist, host a fundraiser for the committee chairman, show up in the district office, send a letter to every member of the delegation, and keep doing it until the program is safe. He may spend a few thousand dollars a year on defending it and consider that the best return he has ever made.

The household stands to save fifty cents. It will not write, call, donate or even learn that the subsidy exists. And a rational household shouldn't: the time it would take to find out about the program is worth more than fifty cents. The household is not stupid. It is behaving exactly the way Olson predicted.

Now multiply. Washington does not consider one such program. It considers thousands. Each one has its small, ferocious constituency and its enormous, sleeping group of payers. Each one is defended. None is individually large enough to cause the voters to rise up. And together, they are a trillion-dollar habit.

Why the incentives defeat the cutters in both parties

Let me now be specific about why the member of Congress, any member of Congress, is pushed toward spending, and why the party label matters less than you would think.

1. The reward is local, the cost is national. A member who brings home a project gets credit from the people who benefit, the local press and the local interests who will contribute to his next campaign. The cost is spread across hundreds of millions of people, of whom only a small fraction live in his district. He gets all of the credit and a tiny share of the bill.
2. Logrolling. I vote for your project, you vote for mine. Each of us can say we only voted for what helped our district, and the sum is a bill nobody would have passed standing alone. Buchanan and Tullock analysed the logic of vote trading in The Calculus of Consent in 1962 and showed why it produces more spending than a majority of voters would choose if asked directly.
3. The cutter is blamed for the cut. A member who votes to end a program gets the anger of its beneficiaries on the record, with a name attached. A member who votes to keep it gets nothing, because nobody writes an angry letter about fifty cents. The incentives are asymmetric: spending is rewarded with gratitude, and cutting is punished with fury.
4. Everyone's priority is someone else's waste. Almost everyone supports cuts to the programs that benefit other people. When the cutting begins, each constituency reveals that its own program is the exception. The farm-state conservative defends the farm programs. The defence-minded conservative defends the military bases in his district. The entitlement-minded conservative defends Social Security, because his seniors vote.
5. The packaging. Spending reaches the floor in enormous bills. Members face a choice between voting for the whole package or shutting down parts of the government. Few want the headline. Packaging makes the choice all or nothing, and the cutter, offered a choice between a bill with things he dislikes and a closure, will often vote yes.

This is why the problem survives changes in the majority. When Republicans hold power, they face the same incentives, plus the pressure of the White House's own priorities. When Democrats hold power, they have different favourites, but the same structure. And the evidence from both parties, in both centuries, shows spending growing regardless. The complaint is not that Republicans are better angels. It is that the structure punishes restraint no matter whose name is on the door.

The case a thoughtful critic would make

Let me do my duty and state the best version of the opposing argument.

"You are describing democracy. People want things, and their representatives deliver them. The programs you call wasteful are somebody's health care, somebody's job, somebody's roads. When voters tell pollsters they want less spending, they mean other people's spending. If restraint were popular, restraint would win. You are not describing a bug in the system. You are describing the system working as intended, and you simply lose the argument."

There is real truth in this. The preferences are real. Voters do want the services, and a conservative who pretends otherwise will lose elections and deserve to. But the argument proves less than it seems, for three reasons.

First, a vote on a package is not a vote on every item in it. Voters never chose the specific subsidy; the beneficiaries pushed it through while everyone else slept. If you asked taxpayers about it directly, with the cost per household printed on the ballot, you might get a different answer. That is exactly what Colorado's voters decided they wanted when they adopted the Taxpayer's Bill of Rights in 1992, which requires voter approval for tax increases; How Colorado Voters Put a Leash on Their Own Legislature tells the story.

Second, there is a cost that no one in the room is representing: the future. The child who will pay the interest on today's borrowing has no vote and no lobbyist. A fair system should include a voice for people who are not yet in the room. A Balanced Budget Is a Moral Argument Before It Is an Accounting One makes that case.

Third, the spending that is hardest to cut is not the spending that most people want. Most Americans have no direct stake in a particular corporate subsidy, a duplicative grant program or a sweetheart contract. These survive not because the majority wants them, but because the majority has no incentive to know about them.

Reforms that conservatives argue actually change the incentives

If the problem is structural, the cure must be structural. Exhortation does not work, as forty years of fundraising letters have shown. What would change the incentives? Here are the ideas that have been serious candidates, with an honest account of how each has fared.

A balanced-budget amendment. The idea is to write into the Constitution a requirement that spending not exceed revenue, with exceptions for war and emergencies. Under Article V, an amendment needs two-thirds of both chambers and ratification by three-fourths of the states. The House passed a version in January 1995, and it fell short in the Senate. Supporters say it would change the incentive by forcing a trade-off in every bill: more spending here means higher taxes or less spending there. Critics argue that it would force deep cuts in a recession, when the economy needs support, and that Congress could evade it with accounting. Both points are fair. Notably, nearly every state operates under some form of balanced-budget requirement, and Why Your State Cannot Print Money the Way Washington Does explains why states face a discipline Washington does not.

A line-item veto. The idea is to allow the President to strike individual items from a spending bill, so that the pork cannot ride along on the must-pass bill. Congress passed the Line Item Veto Act in 1996, and in Clinton v. City of New York in 1998 the Supreme Court held that it violated the Constitution's Presentment Clause, which requires the President to sign or veto a bill as a whole. Supporters have since discussed alternatives, such as a procedure under which the President proposes cuts and Congress must vote on them quickly, an approach sometimes called enhanced rescission. Critics fear that gives the President too much power over the purse, which the Constitution assigns to Congress. That is a serious objection. The debate is about where to put the pen without breaking the separation of powers.

Sunset clauses. A program that expires automatically unless Congress votes to renew it flips the burden: instead of requiring a majority to kill a program, you require one to keep it. The Budget Control Act of 2011, which followed the debt limit fight and set caps with an automatic enforcement mechanism, was an attempt at a similar idea; the 2011 debt-limit fight that produced real spending caps tells how it went. The lesson of that experience was mixed, and I will say so: the caps held for a while, and then Congress found ways around them.

Earmark limits. Congressional Republicans banned earmarks in their own conference at the end of 2010, and both chambers adopted the policy in 2011. Earmarks returned in a modified form in 2021. You can argue either side. Supporters of the ban said earmarks were the grease that helped pass larger bills; critics said it removed a small amount of spending but also removed a bargaining tool. Either way, earmarks are a small share of the budget. The large parts are elsewhere.

Rules that force trade-offs. Congress has tried pay-as-you-go rules, statutory caps and automatic cuts for decades. How Washington learned to budget, from the 1921 act to the 1997 deal follows that history. Each worked for a while and was then waived. The pattern is instructive: a rule that Congress can waive by majority vote is only as strong as the next majority.

What each of us can do

Notice what the logic implies. If spending persists because the beneficiaries are organised and the payers are not, then the answer is to organise the payers and to make the cost visible.

That means a few unglamorous things. It means insisting that every bill carry its cost per household, in plain numbers. It means reading the budget, and How to Read a Federal Budget With a Kitchen-Table Worksheet gives you a worksheet for doing it. It means supporting the state and local limits that already work. And it means choosing, in primaries, the candidates who will pay the political price of a vote to cut, not merely the ones who talk about it.

Above all, it means a change of attitude. Do not ask whether a candidate favours cutting spending in general. Everyone does. Ask him which specific program he will vote to cut, how much it saves and which of his own donors will be angry about it. If the answer is vague, you have learned what you need to know. And when the minority in Congress has to fight a spending bill without the votes, as every party eventually does, the question is what tools remain; how a minority party fights a spending bill when it does not have the votes explains them.

The bottom line

• Spending restraint loses because of arithmetic, not character: a program that gives a few thousand people $15,000 each and costs 120 million households fifty cents each will be defended by the first group and ignored by the second.
• The incentives defeat cutters in both parties, because credit for spending is local, the bill is national, and the member who cuts is blamed by name.
• Pep talks do not work. Structure does: balanced-budget rules, enforceable caps, sunset clauses, a workable line-item veto debate, and local limits like Colorado's.
• Read the actual budget, demand the cost per household on every bill, and ask candidates which specific program they will vote to cut.
• Organise the payers. The side with the fifty-cent stake will win only if somebody tells it that the stakes add up.