A deficit is a tax you did not vote for

Every budget argument runs on the same unexamined assumption: that spending which is borrowed rather than taxed has not yet been paid for. The deficit is treated as a deferred question, something the next decade will settle. Friedman’s objection was not that the number was too big. It was that the framing is wrong at the root.

What is called a deficit is a form of taxation. It is a very bad form of taxation, but it is a form of taxation.

He said that in 1993, in Why Government Is the Problem, and the sentence is doing more work than it looks. He is not saying a deficit will become a tax when the bill arrives. He is saying it already is one. The resources have been moved. Somebody’s labor and materials are being directed by the government right now, this year, and whoever would otherwise have used them has already gone without.

That reframing has a practical consequence. If a deficit is a tax, then the honest measure of what government costs is not the tax line. It is total spending. Borrowing changes who pays and when, and it changes whether anyone had to vote for it, but it does not change that it was paid.

Why he called it a bad tax, not just a tax

The quality judgment matters as much as the category. An ordinary tax has to be legislated. Someone proposes it, someone votes for it, and that person can be asked about it afterward. A deficit routes around all of that. The cost is real and the vote never happens.

This is also why the “temporary” framing deserves suspicion:

From the political point of view, increased spending may initially be designed to be temporary but few things become more permanent than temporary spending.

That was 1992, and it is not a cynical remark about politicians. It is a claim about incentives. A temporary program creates people whose work depends on it, constituencies who plan around it, and administrators whose budgets assume it. Each of those is a reason for it to continue. Nothing comparable argues for its end.

He put the asymmetry more bluntly a year later:

That illustrates a general rule: If a private enterprise is a failure, it closes down—unless it can get a government subsidy to keep it going; if a government enterprise fails, it is expanded. I challenge you to find exceptions.

The challenge is the interesting part. It is falsifiable, which is more than most political generalizations manage, and it is worth testing against cases rather than nodding along to.

The part that is easy to misread

Friedman is often cited as though he thought deficits were the engine of inflation. His actual position in 1992 was narrower and more specific:

On a technical level, I believe that there is no persuasive evidence that, given the course of monetary policy and monetary aggregates, federal government deficits have any stimulative effect.

Read that carefully. He is not saying deficits are harmless. He is saying the stimulative work people attribute to deficits is done by money creation, not by borrowing as such, and that the two get conflated. A deficit financed by genuine borrowing moves resources around. A deficit financed by printing does something else. Treating them as one phenomenon is how the argument goes wrong.

What this archive will not tell you

Friedman died on November 16, 2006. He has no view on any budget passed since, no opinion on current debt-to-GDP ratios, and nothing to say about the specific fiscal arguments of 2026. Anyone who tells you otherwise is inventing it, and this site will not.

What the passages above support is narrower and more durable than a verdict on today’s numbers: a way of reading a budget. Ask what total spending is rather than what the tax line says. Treat “temporary” as a claim requiring evidence. Keep money creation and borrowing as separate questions. Notice which failures close and which ones expand.

Those are tools for examining an argument, not conclusions to adopt, and the test of any of them is whether it survives contact with the evidence.

Every passage quoted here is reproduced word for word and carries its source below. All four have been checked against the work they came from.