Medicare-for-all has a modeling problem. Any big and sweeping policy change like this needs to be modeled by economists beforehand, so policymakers can figure out how best to design it and get it right. But when it comes to Medicare-for-all, everyone involved, including its biggest advocates, are doing it wrong.

There are two distinct problems colliding here.

The first problem is pretty-straightforward: modeling something as gigantic and complex as a new single-payer health-care system is really hard. It's not just doing the macroeconomic models; the economists involved need to talk with experts in the nitty gritty of health-care policy, so they can build a model that's detailed enough to make sense. All that's going to take a lot of time and labor and research by a lot of highly-skilled people. That sort of thing doesn't come cheap.

As a result, it doesn't happen much. The right-leaning Mercatus Center did something on that scale recently, though mainly to concern troll Medicare-for-all supporters with giant federal spending numbers. Another paper came from the Political Economy Research Institute (PERI). And that's about it. The Congressional Budget Office hasn't touched the topic yet, because they're already swamped by more immediate projects.

Ezra Klein recently asked Stephanie Kelton, a Stony Brook University economist who's worked closely with Medicare-for-all champions, what she'd need to game out the policy properly. Her response? "Give me a team of economists and about six months."

In short, when it comes to the big and detailed analyses that could serve as the basis for Medicare-for-all legislation, opportunities are limited and rare. That makes it all the more important that it's done right when the opportunities do come along.

That gets us to the second problem: The entire policymaking mainstream remains convinced that every dollar spent on Medicare-for-all must be matched, dollar for dollar, for new tax revenue. Because this assumption is ubiquitous, even among Medicare-for-all supporters, it guides the modeling of the policy whenever it actually happens. The PERI paper, for instance, was very ideologically friendly to Medicare-for-all, and moved heaven and earth to find the taxes needed to match all the spending.

Now, Americans don't like taxes, so that's a major political handicap.

But more the point, that assumption that you need to fully "pay for" the program is wrong on the merits.

Here's how it actually works: The federal government controls the supply of U.S. dollars, so it can create as many as it wants. It doesn't need taxes to get the money. What it does have to do is manage the macro-economy, making sure the sum total of its fiscal policy doesn't lead to overheating and inflation. As long as the economy is below capacity, the government can spend without any sort of offset, adding more demand to the economy and creating jobs until it fills that gap. Once that gap is filled, additional spending does need to be offset, lest inflation set in.

This critique is often associated with Modern Monetary Theory, of which Kelton is a major champion. But most of the mainstream economics profession will acknowledge this description is accurate. They just think distinguishing between taxes as "paying for" spending and taxes as offsets for inflation is a matter of "tomatoe" versus "tomato."

But sometimes, the distinction does matter. In the case of Medicare-for-all, it really matters.

Let's assume the economy is already at full capacity. (It's not, but assume it is.) Then you pass Medicare-for-all, and you don't raise a cent in taxes to offset it. What happens?

First, a ton of spending by private health insurers goes away, replaced by government spending. In terms of economic capacity being used, this is a wash — no net increase, thus no additional inflationary pressure.

Next, a lot of private spending on insurance bureaucracy gets wiped out and replaced with nothing. Private providers would also presumably get squeezed for lower prices, eliminating even more spending. These are big selling points among Medicare-for-all advocates: You save a lot in overhead and cost inflation. But those changes also imply reductions in capacity usage, and thus less inflationary pressure, and thus fewer jobs.

Finally, Americans currently spend a lot on health-care premiums and out-of-pocket costs. Depending on how generous your Medicare-for-all plan is (and proposed versions are often really generous) all that money will now be free to go elsewhere. That will be new spending in the economy, using up more capacity and creating jobs and pushing up on inflation.

That's a lot of forces pushing the economy in different directions. The question is, what's the final result? "You have to net out all of this stuff," as Kelton told The Week.

She explained that if she did have her team of economists and six months, she'd first run the above scenario — Medicare-for-all with no additional taxes — and see what happens. Almost certainly, you'd overheat the economy and get inflation. Then she'd work backwards from that baseline, adding new offsets to decrease inflation as she went. The goal wouldn't be to "pay for" the spending. It would be to get the economy back to an equilibrium of full employment and stable prices. And that would require a fraction of the taxes you'd raise working off the dollar-for-dollar approach.

In other words, if the U.S. actually succeeded in passing a Medicare-for-all program with dollar-for-dollar matching taxes, it would be overkill. The country would offset the spending too much, remove too much demand, and leave the economy below full employment. The PERI study, for example, estimated over one million job losses, mainly in the private health insurance bureaucracy. That's employment that often goes to less privileged Americans, especially low-income women. And the paper didn't even get into the knock-off job losses you'd get from that in the wider economy. Where are all those workers supposed to go?

To its credit, the PERI analysis suggested a bunch of transition programs to get those workers new jobs in different sectors. But then it wanted to raise taxes to pay for those programs too, which just means digging the hole deeper.

Now, there's no way those Americans who lose their jobs in the private health insurance bureaucracy could keep working in the same sector. Permanently squashing that bureaucracy is part of the point of Medicare-for-all. But the government does need to ensure there are good jobs waiting for those people. That means ensuring the economy has reached full employment and demand is maxed out. That will keep employers desperate for labor, and they'll scramble to snatch those workers up.

But if the government fully "pays for" Medicare-for-all, it will leave those Americans in the lurch.

Exactly how much of Medicare-for-all's spending should be offset with taxes? Or what kind of taxes? No one knows, because no one's modeled it properly yet.

Someone should.