The leading thinker and most visible public advocate of modern monetary theory—the freshest and most important idea about economics in decades—delivers a radically different, bold, new understanding for how to build a just and prosperous society.

Stephanie Kelton’s brilliant exploration of modern monetary theory (MMT) dramatically changes our understanding of how we can best deal with crucial issues ranging from poverty and inequality to creating jobs, expanding health care coverage, climate change, and building resilient infrastructure. Any ambitious proposal, however, inevitably runs into the buzz saw of how to find the money to pay for it, rooted in myths about deficits that are hobbling us as a country.

Kelton busts through the myths that prevent us from taking action: that the federal government should budget like a household, that deficits will harm the next generation, crowd out private investment, and undermine long-term growth, and that entitlements are propelling us toward a grave fiscal crisis.

It’s an attractive vision, but it doesn’t work.

I am reminded of Einstein’s time at the Swiss Patent Office where he used to check applications to patent perpetual motion machines. They don’t work, but the fun is working out why. The same applies to proposals to bring about prosperity that depend on loosening the monetary spigots. MMT is a perfect example.

MMT is a macroeconomic school of thought in the post-Keynesian tradition. Its central tenets: fiscal deficits don’t matter; monetary policy should be subordinate to fiscal policy; and the monetary authorities should be willing to issue base money to finance government spending. MMT is associated with large-scale government spending, a focus on ending involuntary unemployment, and programs to alleviate poverty and fight climate change.

Kelton’s book builds on earlier work by Warren Mosler and Randall Wray but has its roots in Abba Lerner’s system of “functional finance,” which goes back to the 1960s. She builds on Lerner primarily by adding a federal job guarantee that would eliminate involuntary unemployment and provide an automatic economic stabilizer.

MMT makes big promises. It would “build a more just economy that works for the many and not just the few” and put “people and planet first.” “MMT’s lens enables us to see that another kind of society is possible, one in which we can afford to invest in health care, education, and resilient infrastructure. In contrast to narratives of scarcity, MMT promotes a narrative of opportunity.”

But does MMT deliver? Let’s see what she says.

“The idea that taxes pay for what the government spends is pure fantasy,” writes Kelton. Really? Let’s go back to basics. The government must finance all its expenditures. In a world in which it does not issue debt and does not issue currency, and assuming away any gifts it might receive, all its expenditures must be financed by current taxation.

If the government can issue debt but not issue currency, then it can finance its expenditures by current taxation or by issuing debt. But to issue debt is to pass on the obligation to repay that debt to future taxpayers. If that debt is to be repaid, then it must be repaid out of future tax proceeds.

If the government can issue its own currency and monopolizes the issuance of currency, then it can also pay off its debt obligations as they come due by issuing additional base money (“printing money”). Does this mean that printing money allows the government to avoid the need to raise taxes? No, because printing money lowers its value against goods and services, and so operates as a tax on money holdings and other holdings of wealth that are fixed in nominal terms (such as level annuities). So, barring gifts, all government expenditures must be financed by taxation in one form or another.

Kelton explains:

The federal government announces a wage (and benefit) package for anyone who is looking for work but unable to find suitable employment in the economy. Several MMT economists have recommended that the jobs be oriented around building a care economy … the federal government would commit to funding jobs that are aimed at caring for our people, our communities, and our planet.… Since the market price of an unemployed worker is zero … the government can create a market for these workers by setting the price it is willing to pay to hire them. Once it does, involuntary unemployment disappears. Anyone seeking paid employment has guaranteed access to a job at a rate of remuneration established by the federal government.

This sounds great: involuntary unemployment eliminated and everyone willing to work gets a high federal minimum wage or more, to the extent that market wages are forced higher to compete. But hold on. If it is such a good idea, why not raise the minimum wage beyond the $15 an hour she suggests? Why not $30 an hour? Or $50? The problem is that there are a raft of jobs that are profitable to provide at existing wages but would disappear at higher wages.¹ It is not just the existing unemployed who would end up on federal payrolls but these newly unemployed too, and many of their employers. Think of the restaurant sector. That sector and others in the same position could only survive by hiking their prices: dining out would become a lot dearer. Ordering in, too. The federal government, the employer of last resort, would find itself with the problem of what to do with all these people also turning up for guaranteed jobs. The feds would have crowded out much of the labor market and wiped out the lower paid sectors of the economy.

“Why does the financing have to come from Uncle Sam?” asks Kelton. “Simple. He can’t run out of money.” Imagine that the government pays debts coming due by handing over dollar bills that it has in a chest in the basement. If it runs out of bills, then it will default the next time a payment comes due. But then imagine if it can also print money. If it runs out of bills, it can avoid default when the next payment comes due by printing more. It does not follow, however, that the government can always meet its payment obligations by printing more money.

Suppose the government prints money at an accelerating rate and we end up with an accelerating hyperinflation. The traditional tax-collection apparatus will break down because the tax revenue will be worth almost nothing by the time it comes in. Similarly, the government will effectively be unable to borrow in its own currency because the borrowed funds would also be worth next to nothing by the time they come in. As the hyperinflation accelerates further, the real value of the revenue from printing money also goes to zero. The government then faces the prospect of default despite being able to print any amount of its own money. To give an example, by the end of the Hungarian hyperinflation of 1946, the total value of all Hungarian notes in circulation was a thousandth of a U.S. cent (see Judt 2006: 87). The Hungarian government didn’t have a cent, let alone a dime. The Hungarian government would have been unable to make repayments denominated in other currencies or make inflation-linked payments in its own.

The mistake is to presume that what is correct at the margin (i.e., that the government can avoid default by issuing a few extra dollar bills) is also correct under any circumstances, that is, at any scale. There is also the related point that issuing a small amount of money will have a negligible impact on prices but issuing a lot of money will not.

It is a “myth,” writes Kelton, “that deficits will burden the next generation.” This claim is also wrong. Suppose Congress passes a Boomers Boomtime Act to provide for a humongous 75th birthday payout to each surviving member of the first Boomer cohort born in 1946. They will reach 75 in 2021. These payments are to be financed by a zero-coupon bond with a 40-year maturity. Since none of the beneficiaries will be around to pay taxes when the bond is due to be repaid, they get a free handout.

Who bears the burden of paying for it? When the Boomer bond comes due in 2061, the government faces the following choices: (a) pay it off by raising taxes, (b) pay it off by issuing money, (c) default, (d) pay it off by rolling over, that is, by issuing a new bond.

If (a), then the burden is borne by taxpayers in 2016.

If (b), the subsequent price level is higher, so the burden takes the form of a tax on money holdings and other instruments of fixed nominal value.

If (c), default, the burden is borne by those who suffer the adverse consequences of default.

If (d), then the rollover will mean that there will more debt after 2061 than there would otherwise have been and we have the same choices again when the new payments come due. If the decision is to roll over each time, then the debt/​GDP ratio will hit a level at which the government defaults sooner than otherwise.² Thus, however the government responds when the Boomer bond matures, some group born after 1946 bears a burden from it.

More generally, any arrangement that involves one group issuing a debt that another group is expected to pay for necessarily burdens the second group. The injustice is all the worse because the second group has no say in the matter.

There are also the government’s entitlement programs, Social Security, Medicaid, etc. This takes us to Kelton’s “myth” that “entitlements are propelling us toward a long-term fiscal crisis.… There is absolutely no good reason for Social Security benefits, for example, to ever face cuts. Our government will always be able to meet future obligations because it can never run out of money.”

These programs however are just another form of debt insofar as they create obligations on the government’s part to make future payments. Consequently, my earlier argument, that programs that create future obligations burden future generations, applies here also.

Entitlements are large, so the corresponding burdens would be large as well. To illustrate, there are perhaps $210 trillion in entitlements, and possibly more.³ If these entitlements are to be paid for by future taxation, then that is a lot of future taxation. If they are to be paid for by rolling over, then we would anticipate the ratio of debt (including entitlements) to GDP rising considerably, possibly to default levels.

Then there is the option of meeting those obligations by printing money. Given that the current stock of base money is just over $5 trillion, that response implies a possible 42-fold-plus expansion of the monetary base. That, in turn, implies a considerable increase in prices. Making entitlement payments is one thing, but the purchasing power of those payments is another.

We have here another instance of the margin vs. scale issue. The government can increase entitlements a little with next to no impact on their real value. But if the government creates huge entitlements to be financed by printing money, then those entitlements are going to be greatly devalued in purchasing power terms. And what the government must absolutely not do is create huge entitlements that are inflation-linked and then rely on printing money to finance them. If it does that, it will produce both hyperinflation and default. The “myth,” that the national debt is a problem, is not a myth.

We can break down MMT into a set of policy ends (what the government spends on), and a set of policy means (how the government finances its spending), which in the case of MMT would involve large deficits and a lot of borrowing and money printing. The expenditure and the financing of that expenditure are two different issues. Bernie Sanders might use MMT to advance a more right-wing version of the Kelton agenda, but Donald Trump might seize upon the spending opportunities promised by MMT to advance his own, even more right-wing, agenda, for example, to promote policies that work for the few and not the many.

My point is that it is short-sighted and potentially counterproductive to promote a particular policy package such as MMT because you can use it to finance projects that you like, because someone else might use it to finance projects that you do not like. To the extent that MMTers persuade people that MMT-based government finance is a good idea, they can hardly restrict that message to people who share their own political views. If you think of MMT as a government-financing package, then that financing package can be used to finance any government spending program, whatever its political hue.

A deeper issue is that any policy that gives policymakers the appearance of being able to spend a lot without having to bear the unpopularity of the high taxes needed to finance that spending is a dangerous one and invites abuse. Any such policy entails a major shift in power away from the legislative branch to the executive branch, because it gives the latter additional means of finance that bypass constitutional constraints against government overspending. To elaborate, the Constitution says that fiscal policy, the power to tax and spend, is constrained by the need to obtain congressional approval. If there were no Fed, or if the Fed were genuinely independent, then Congress could deny appropriations for spending projects of which it does not approve. But if the executive branch has the power to print money, then it has a potential means to circumvent Congress. If Trump wants his wall and Congress denies the appropriations, he can then order the Treasury secretary to print the necessary money instead. From this perspective, MMT is something of a constitutional abomination.⁴

It is a fundamental principle of constitutional political economy that economic policymakers operate under rules that constrain the decisions they make, and that these rules should be designed in ways that prevent undesirable behavior on their part. Under this way of thinking, the rules operate as bulwarks that constrain policy makers in order to protect everyone else from the misuse of the powers entrusted to those policymakers.

For proponents of MMT and for many other advocates of big government, however, those rules serve no real purpose and merely constrain policymakers from achieving the lofty ends that they seek to pursue. Yet they fail to appreciate that lofty ends do not justify giving those in power unrestrained discretionary powers. As Adam Smith observed:

The statesman who should attempt to direct private people in what manner they ought to employ their capitals, would not only load himself with a most unnecessary attention, but assume an authority which could safely be trusted, not only to no single person, but to no council or senate whatever, and which would nowhere be so dangerous as in the hands of a man who had folly and presumption enough to fancy himself fit to exercise it [Smith (1776) 1994: 423].

Should MMT be adopted, then the Fed would become subservient to the Department of the Treasury, but in a more nakedly obvious way than it was during the years before the 1951 Treasury-Fed Accord, and without operating under the constraints of the Bretton Woods watered-down gold standard. Personally, I would suggest that, if MMT were adopted, then the government should make the new monetary policy arrangements transparently obvious. The Fed’s independence, such as it was, would be history and there would be no point pretending otherwise. The government should nationalize the Fed and make it a division of the Treasury, whose responsibilities would then be to issue currency and manage the national debt. Nationalizing the Fed would highlight the underlying chartalism of MMT (i.e., the idea that money is a creature of the state),⁵ and would also simplify analysis going forward because it would cut out the need to consider Fed/​Treasury interactions that only mask the underlying reality. We could then talk openly about the government printing money. If the United States is going to embark on a monetary policy worthy of a banana republic, then it should look the part.

Suppose then that the government goes full throttle MMT à la Kelton. This spending must be financed, however, and the government would have to do so by some combination of levying taxes, borrowing, and printing money. In essence, she proposes high government spending and a big deficit financed by borrowing (“debt finance”) and printing money (“monetary finance”).

If such a policy were launched at a time when there is considerable unemployment, then one would suppose that unemployment would fall. But there must eventually come a time when the economy returns to more or less “full” employment, whether because of those policies or despite them being a separate question. What happens then? To examine this question, I built a model along “unpleasant monetarist arithmetic” (see Sargent and Wallace 1981) lines and got some interesting results.

Let’s consider the following three possible MMT policies: (1) the government pursues pure debt finance; (2) the government pursues pure monetary finance; and (3) the government pursues debt finance for as long as possible, up to the point where it is about to default, and then switches to monetary finance.

Let d be the long-term growth rate of the deficit and g the long-term rate of economic growth. It is reasonable to suppose that MMTers would want d considerably in excess of g, so let us assume that this is so. Under a policy of pure debt finance, the debt/​GDP ratio would grow relentlessly and must at some point reach a level at which the government defaults. Hence, debt finance makes government default inevitable if pursued for long enough—that is, pure debt finance is unsustainable.

Under a policy of pure monetary finance, d becomes the key driver of the inflation rate. If d is steady in the long-term, then the inflation rate will converge to a long-term steady state. But if d itself grows, so the deficit grows at an accelerating rate, then long-term inflation will also accelerate.

Under the third policy, the debt/​GDP ratio rackets up to the brink of default, then the government switches to monetary finance with similar long-term outcomes as pure monetary finance.

It is important to emphasize that under monetary finance there is no way in which the government can simultaneously pursue an inflation target. Instead, the inflation rate becomes a residual outcome from the government’s fiscal policy and the government loses all control over the inflation rate. To make matters worse, the inflation rate also becomes the macroeconomy’s main shock absorber, so any shocks that produce unexpected increases in the deficit will lead to unexpected increases in inflation, which makes inflation highly uncertain too. A policy of monetary finance is therefore dangerous, because it can easily lead to runaway inflation or even hyperinflation. Externally, these effects on prices and inflation will be reflected in a falling, volatile and uncertain exchange rate.

In sum, we have a variety of possible long-term consequences, ranging from merely bad (highish and uncertain inflation, loss of control over inflation, a volatile exchange rate) to positively catastrophic (huge levels of national debt, high future taxation, national default and all that might entail, runaway inflation, hyperinflation). I could go into a long discussion of why any other fiscal-monetary policy mix would produce better long-term outcomes. On the fiscal side: a balanced budget or lower deficit financing. On the monetary side: a monetarist rule, a nominal GDP target, a Taylor Rule, a gold standard, whatever.

Suffice to say that the poor performance of MMT is not a coincidence. On the fiscal side, it encourages a much more rapid run-up of the debt/​GDP ratio than any alternative, which has got to be the worst possible fiscal policy in the long term. On the monetary side, it throws away any attempt at controlling inflation or maintaining monetary stability by making monetary policy subservient to runaway government spending. MMT performs so badly precisely because it represents the extremes of fiscal and monetary excess.

Indeed, MMT does not even work on its own terms. Kelton herself indicates that an MMT policy package is constrained by the requirement that inflation should not rise. Unfortunately, she hasn’t thought it through.

Current U.S. inflation is 2.3 percent. Is she suggesting that the government should pursue MMT subject to the constraint that the inflation rate should not rise above 2.3 percent? If so, she should advertise the fact to help dispel the concerns of those who might have gotten the impression that MMT is some sort of funny money scheme. If she did so, much of the knee-jerk opposition from sound-money people would dissipate. The problem, however, is that a commitment to maintaining inflation at no more than its current rate would severely constrain the ability of MMT to deliver on its promises.

A looser interpretation of her “inflation shouldn’t rise” constraint would apply to inflation in the long run. But then consider my unpleasant monetarist arithmetic results. Since pure debt finance is fiscally unsustainable, any MMT package must involve some element of monetary finance. But, ex hypothesi, the monetary finance must be constrained by the need to avoid rising inflation in the long run. A necessary condition to prevent rising inflation is that the rate of growth of the deficit should not itself increase. This constraint is not as severe as requiring that current inflation should never rise, but it is a severe constraint nonetheless. The problem is that it is not at all clear how much of what she promises can be delivered while satisfying this constraint. That she does not address this issue is the central failing of her book.

In effect, she offers us the prospect of a bunch of goodies but doesn’t explain why those goodies fall within the economy’s production possibility frontier, that is, are actually attainable—an intriguing oversight for an economist.⁶

She offers a revealing anecdote when she recalls a discussion between James Tobin and President Kennedy:

Tobin recalls JFK asking, “Is there any limit to the deficit? I know of course about the political limits.… But is there any economic limit?” When Tobin confessed that “the only limit is really inflation,” the president replied, “That’s right, isn’t it? The deficit can be any size, the debt can be any size, provided they don’t cause inflation. Everything else is just talk.”

“Everything else is just talk” captures it perfectly. Indeed, it undermines the entire book.

We shouldn’t forget what subsequently happened. Government spending on the Vietnam War and the Great Society overdid it, and the United States ended up with rising inflation and the monetary troubles of the 70s and early 80s. MMT’s fiscal proposals are akin to the Great Society and its monetary proposals are akin to relying on the pre-NAIRU, or fixed, Philips Curve that was discredited by Milton Friedman. From this perspective, MMT has a distinct sixties feel to it.

In both cases, the root problem is the same: the absence of a coherent theory of inflation. The Keynesians of the sixties didn’t have one and neither does Kelton. Now, as then, the solution to that problem is the same: their macro model needs some version of the quantity theory of money to connect the money supply to the price level, and thence to the inflation rate. If the central bank or the government pursue policies that lead to too much monetary growth, then the result will be a higher than desired inflation rate. The solution to that problem is also the same as it was then: to rein in the rate of monetary growth. In short, MMT is not particularly modern and the monetary theory has too much money and not enough theory.

Kelton might object that these negative outcomes would not occur under MMT because counter measures would be taken once (or even before) inflation started to rise. So, what counter measures would she pull from her MMT toolbox? The answer is a rise in taxes.⁷

One can imagine the howls that would follow a proposal to raise taxes “merely” because inflation had gone over some “arbitrary” threshold. Why abandon The Project? Why stop policies that were on the verge of making the world a better place just as the going gets tough? MMTers attempting to stick to the “if inflation rises” script on which The Project was predicated would be cast into the role of fiscal conservatives. Don’t they know that deficits don’t matter, etc.? They then reap the downside of overpromising.

If and when taxes were increased, it is doubtful that doing so would get inflation back down again. We know from monetarism that the inflation rate will only come down once the underlying monetary growth rate has slowed. But since the MMTers lack a decent model of inflation the likely policy responses would be some muddle akin to what we experienced in the late 1960s and much of the 70s, and with similar results: rising inflation followed by stagflation, a new Keynesians vs. monetarists controversy, and inflation only being brought under control again when policymakers relearn the lessons learned then—namely, the importance of the quantity theory of money.

The poor long-term performance of MMT under my simulations is a perfect illustration of why policymakers need to be constrained by rules. To illustrate the benefits of such rules, consider the following. Recall that we can think of MMT as a set of policies that break down into two subsets:

(1)MMT = {MMT spending program; MMT financing program}.

The former is about what goes out of the government’s coffers and the latter is about what goes in.

It seems to me that for most people inclined toward MMT, the big attraction is the MMT spending program. For the sake of argument, let’s hypothetically agree with that spending program and then ask if we can replace the MMT financing program with something better.

The MMT financing program consists of a combination of high deficits, tax, borrowing, and monetary finance. This financing program is a key reason why MMT performs so badly, so let’s replace it with an alternative financing program that is a combination of, let’s say, more restrained deficits, tax and borrowing, and no monetary finance. The monetary side of this program would be taken care of by some monetary policy or rule that focuses on a stable inflation rate or something similar. We then come to:

(2)Alternative to MMT = {MMT spending program; alternative financing program}.

My models indicate that this “Alternative to MMT” would produce considerably better outcomes, including a less rapid run-up of national debt and lower inflation.

Why then would you not prefer the “alternative” to MMT? The “alternative” still delivers the spending goodies you want, but in a less damaging way in the long-term. But the “alternative” is old-fashioned tax and spend!

My point is that even if you support MMT because of its spending program, there is no good reason to support MMT in preference to some tax and spend policy mix with the same spending program. Whatever your preferred government spending program, MMT is a poor way to finance it.

So, if you are a hard-left socialist who supports the Kelton government spending platform, you should support tax and spend, not MMT. And if you do not support her spending platform, say because you are not politically hard left or because you support sound money, then you would also not support MMT. Whatever your politics, MMT is not for you; MMT is just bad economics.

In the end, MMT comes down to this: the government spends a lot, issues a lot of debt, and prints a lot of money. It is not as if it hasn’t been tried before.

Haskins, R. (2015) “The Federal Debt is Worse Than You Think.” Brookings Institution blog (April 8).

Judt, T. (2006) Postwar: A History of Europe Since 1945. New York: Penguin.

Sargent, T. J., and N. Wallace (1981) “Some Unpleasant Monetarist Arithmetic.” Federal Reserve Bank of Minneapolis Quarterly Review (Fall): 1–17.

Smith, A. (1994 [1776]) An Inquiry into the Nature and Causes of the Wealth of Nations. New York: Modern Library.