Friedman's Four Ways to Spend Money Explain Every Government Budget

You already know the four ways to spend money. You learned them without a textbook the first time someone took you shopping with a company card, or let you pick out a gift using a relative's cash, or told you to grab whatever you liked because it was on the house. You noticed how your behavior changed. That is the whole of Milton Friedman's matrix, and it explains more about Washington than a shelf of budget reports.

The idea comes from Free to Choose, the 1980 book Friedman wrote with his wife and fellow economist Rose Friedman, published alongside a ten-part public television series. It is a simple two-by-two box, and it is one of the most useful things a citizen can carry around in his head. Let me draw it, run it through a few households, and then answer the best objection, which is that elections and oversight fix the problem.

The box

Friedman asks two questions about any act of spending. Whose money is it? And for whom is it being spent? Two answers each, four boxes.

Now the incentives, box by box. Friedman states them in terms of two concerns: how economically you spend, meaning how careful you are about cost, and how effectively you spend, meaning how careful you are about whether the result serves its purpose.

Box 1: your own money, on yourself. You have every reason to be thrifty and every reason to get what you actually want. Think of buying a pair of work boots with your own pay. You compare prices, you test the fit, and you get something that will last. You are careful about cost and careful about value.

Box 2: your own money, on someone else. You still watch the cost, since it is your money. But you are less sure the gift fits the recipient, so you are less careful about effectiveness. Anyone who has bought a sweater for a relative has wasted a little money on the wrong size, and has also learned why a gift card exists.

Box 3: someone else's money, on yourself. Now the cost stops mattering much to you, since it is not yours, but you still care greatly about what you get. This is the expense account lunch. You will find an extravagant restaurant, and you will make sure the meal is good.

Box 4: someone else's money, on someone else. Neither concern is anchored. You have no reason to economize, since it is not your money, and no reason to check the result, since it is not for you. Friedman's observation is that nobody has an interest in being careful about either thing. It is the fourth box where waste lives.

Where government spending sits

Here is the punchline. Friedman's point is that when a government program takes money from taxpayers and spends it to provide goods or services for other people, it is operating in Box 4. Not always, and not only. But as a structural matter, the officials who design and run the program are spending money they did not earn on people they do not know.

I am not claiming that government officials are lazy or dishonest. Many are conscientious and hardworking. The argument is about incentives, not characters. A conscientious official in Box 4 still lacks the signals that Box 1 supplies automatically: the pain of paying and the satisfaction of getting what you wanted. No price tells him he is overspending; no unhappy customer walks away to a competitor.

Compare two households, labelled as hypotheticals.

• Household A has a family budget of its own. The parents decide to replace a broken furnace. They get three quotes, ask about warranties and negotiate. This is Box 1.
• Household B is asked to run a fund to replace furnaces for families in another city, using money raised by a levy. The administrator has no personal stake in the price, and the families who receive the furnaces have no say in the bid. Even if the administrator is honest and careful, there is no price signal, no competitor and no customer who can walk away. This is Box 4.

Nothing here requires cynicism. It just requires noticing which signals are missing.

The Washington examples, in general terms

I want to stay with what we can state exactly rather than invent figures. The federal budget includes enormous transfers and programs, and the mechanism of Box 4 applies to the part that is a program run by one set of people for the benefit of another with money taken from a third. We can say, without any statistics, three things that follow from the box.

1. Costs creep. Where no one is the owner of the money, a program's budget tends to be defended by its beneficiaries and its administrators, who see the benefits clearly, while the cost is spread across millions of taxpayers, each of whom feels a small part of it. That asymmetry is the central finding of public choice economics, discussed in Public Choice Theory Studies Politicians as People, Not as Angels.
2. Success is measured by input. Programs are commonly judged by how much money they spend, not by what they achieve. A program with a bigger budget is called a stronger commitment. Box 4 produces no price that says otherwise. Real per-pupil spending in the schools, which we examine in Real Spending Per Pupil More Than Doubled Since 1970, So Where Are the Gains, is a useful case of the difference between spending and results.
3. Programs outlive their purpose. Because no one owns the money, no one has a reason to stop spending it. The program keeps going, and the people who run it have every reason to find new reasons for it. Baseline budgeting makes the habit worse, as we explain in baseline budgeting, where a smaller increase gets reported as a cut.

For the largest numbers, use the primary sources. The Treasury's debt figures and the CBO's budget reports are public, and the national debt clock on this site carries the official debt totals. The mechanism is the thing to carry out of this essay, not any particular figure.

The subtle fourth box and the third

One more observation, because Friedman made it and it is easy to miss. Box 3, spending someone else's money on yourself, has its own problem: you spend lavishly and care a great deal about quality. Where does government do this? It does it whenever officials and legislators spend public money on their own offices, perks and travel, or when a program is designed to benefit the people who administer it. The temptation to spend without limit runs highest where you get the benefit and someone else gets the bill.

Politicians have a variant of it: spending money to benefit their own voters in their own districts, with the bill sent to the whole country. Every member of Congress who brings home a project for his district is in Box 3 with respect to the district, and the rest of us are the ones paying. That is why a spending bill with something for everyone is easier to pass than a bill with something for no one, and why earmarks and logrolling are as old as legislatures themselves.

The best objection, stated fairly

Now the strongest reply. The defender of government spending will say the following.

Government is not a private household. It is accountable in ways a stranger spending a stranger's money is not. Legislators are elected. Agencies are audited. Inspectors general, the press and courts all watch. Voters can throw out a government that wastes their money. And many of the things government buys, roads, defense, courts, are public goods that no household could buy alone. So Friedman's matrix describes an abstract incentive problem that real institutions already fix.

It is a fair argument, and here is the response, in numbered points.

1. Elections are a weak substitute for prices. A price gives feedback on every purchase, instantly, in the buyer's own wallet. An election gives feedback every two, four or six years on a bundle of thousands of decisions, and the voter has no way to say yes to one program and no to another. The feedback in Box 1 is precise; the feedback in Box 4 is a rough guess.
2. The voter's incentive to inspect is tiny. A taxpayer's share of any one program is small, so he has little reason to study it. The beneficiaries, on the other hand, care intensely. The result is that those who gain from a program organize while those who pay do not. Friedman's box and the public choice school describe the same asymmetry.
3. Oversight is itself Box 4. An auditor, an inspector general or a legislative committee is also spending other people's money on other people's affairs. The overseers have the same missing signals, and the oversight body can be captured by the thing it oversees. Oversight helps. It does not restore the missing price.
4. Public goods are a real exception, but a narrower one than claimed. Defense, courts and some infrastructure are genuinely hard to supply privately, and the Constitution grants Congress those powers explicitly. Granting that, the argument does not extend to every program that has been justified by analogy. The box tells us where to look hardest: wherever the money is spent on other people.
5. The practical conclusion is to move spending toward the lower boxes. Spend at the level of government closest to the people, and wherever possible let money follow the person rather than the program, so the recipient is choosing and the spending is closer to Box 3 and Box 1. Friedman's own answer was vouchers, an idea we trace in Milton Friedman's 1955 Voucher Essay and Its Seventy-Year Road.

The verdict: the objection is right that government has accountability tools and right that some purposes require public spending. It is wrong to think the tools replace the incentive. They supplement it, imperfectly, and the weaker the supplement, the more the program drifts toward Box 4.

What to do with the box

The box is a tool, not a slogan. Use it on any proposal.

• Ask whose money and for whom. If the answer is "taxpayers' money for other people," say aloud that you are in Box 4 and ask what substitute for a price or a customer the program has.
• Prefer arrangements where the beneficiary chooses. Vouchers, tax credits and direct payments put more of the decision with the person who will live with it.
• Ask for measurements of results, not inputs. A program that cannot say what it is trying to achieve and how it will know is a program that cannot fail, and so cannot succeed either.
• Keep spending local where you can. A county commissioner who must face the same neighbors the next morning, and who knows the road in question, is closer to Box 1 than an agency three levels away. That is the reasoning behind federalism, and the Tenth Amendment is its text.
• Watch your own spending. The box applies to us. When we spend a company's money or a relative's, we should notice how our care slips.

The Founders knew this instinctively. Article I, Section 8 gives Congress the power to lay and collect taxes to pay the debts and provide for the common defence and general welfare, and then lists what it may do. Madison read that list as a limit on the spending power; Hamilton, and later the Supreme Court in United States v. Butler (1936), read the general welfare language more broadly. Whichever reading you favor, the enumeration is a built-in way to fight Box 4 if it is enforced: the fewer things the government may buy for other people, the fewer decisions made without a price.

The case in brief

1. Friedman's matrix in Free to Choose (1980) sorts spending by whose money and for whom, and shows that care about cost and about results is strongest when you spend your own money on yourself.
2. Spending other people's money on other people, Box 4, removes both incentives, and much government spending is of this kind by its structure, not because officials are bad.
3. The consequences are programs measured by inputs, costs that creep because those who benefit are organized while those who pay are dispersed, and programs that outlive their purpose.
4. Elections and oversight help but do not restore the missing price signal, since voters judge bundles rarely and overseers are themselves in Box 4.
5. The remedy is to put spending closer to the people, keep it within enumerated powers, and let the beneficiary choose where possible.

What you can do: take one line of your local school, county or city budget, find who decides it and who benefits, and ask at the next public budget hearing what result the money is meant to buy.