A very detailed walkthrough of the big new left economic idea.
Illustrations by Christina Animashaun
Apr 16, 2019, 5:00 PM UTC
Dylan Matthews is a senior correspondent and head writer for Vox’s Future Perfect section and has worked at Vox since 2014. He is particularly interested in global health and pandemic prevention, anti-poverty efforts, economic policy and theory, and conflicts about the right way to do philanthropy.
The theory, in brief, argues that countries that issue their own currencies can never “run out of money” the way people or businesses can. But what was once an obscure “heterodox” branch of economics has now become a major topic of debate among Democrats and economists with astonishing speed.
For that, we can thank Rep. Alexandria Ocasio-Cortez (D-NY), who told Business Insider in January that MMT “absolutely” needs to be “a larger part of our conversation.” That was the most vocal mainstream support MMT had gotten, which for years had been championed by economists like Stephanie Kelton (a former adviser to Bernie Sanders), L. Randall Wray, Bill Mitchell (who coined the name Modern Monetary Theory), and Warren Mosler — as well as a growing number of economists at Wall Street institutions.
With AOC on board, a wave of denunciations from mainstream economists and others followed. Fed Chair Jerome Powell, Bill Gates, former Treasury Secretary Larry Summers, and former IMF chief economist Kenneth Rogoff all attacked the theory.
Or, more accurately, they attacked what they thought the theory to be. MMT is more nuanced than the “governments never have to pay for stuff” caricature it’s earned among other economists, and MMT advocates are famously (and often understandably) ornery when they sense they’re being misrepresented.
At the same, that caricature gets at what may ultimately be the most important effect of MMT as an idea: It could convince some Democrats to break away from the view that spending always has to be “paid for” with tax increases. How many Democrats buy that conclusion, and how far they’re willing to take it, remains to be seen. But some are already moving in that direction: While emphasizing that “debt matters,” Sen. Elizabeth Warren (D-MA) recently noted, “we need to rethink our system in a way that is genuinely about investments that pay off over time.”
The rise of MMT could allow Democrats to embrace the de facto fiscal policy of Republican presidents, who tend to explode the deficit to finance pet initiatives like tax cuts and defense spending, leaving Democrats to clean up afterward. MMT could be Democrats’ way of saying, “We don’t want to be suckers anymore.”
That would be a big deal. Getting comfortable with new deficit-financed programs would help Democrats overcome the single biggest impediment to their agenda: raising taxes to fund their programs. MMT could offer a way to justify passing big priorities like single-payer health care or free college without resorting to major middle-class tax hikes.
And if the idea behind MMT is wrong, that shift could be a false promise, one that offers short-term political benefits at the expense of hard to foresee economic costs.
So let’s dive into the wonky details of MMT. And I do mean wonky — this is a pretty technical article that gets into the nitty-gritty of why MMT is different from mainstream economics. But I think those details are important, and they’re easy for even very smart, very informed people to get wrong.
I’ll explain MMT theories about deficits, inflation, and employment, and what it all means for Democratic Party politics in 2020 and beyond.
If you ask a mainstream economist why budget deficits can be harmful, they’ll probably tell you a story about interest rates and investment.
In the standard story, the government levies taxes and then uses them to pay for what it can. To pay for the rest of its expenses, it then borrows money by issuing bonds that investors can buy up. But such borrowing has a big downside. Budget deficits increase demand for loans, because the government needs loans on top of all the loans that private individuals and businesses are demanding.
And just as a surge in demand for, say, tickets to a newly cool band should increase the going price of those tickets (at least on StubHub), a surge in demand for loans makes loans more expensive: The average interest charged goes up.
For the government, this is an additional expense it has to incur. But the higher interest rate applies to private companies and individuals too. And that means fewer families taking out mortgages and student loans, fewer businesses taking out loans to build new factories, and just generally slower economic growth (this is called “crowding out”).
If things get really bad and the government is struggling to cover its interest payments, it has a few options, none of which mainstream economists typically like: financial repression (using regulation to force down interest rates); paying for the interest by printing more money (which risks hyperinflation); and defaulting on the debt and saying that lenders just won’t get all their money back (which makes interest rates permanently higher in the future, because investors demand to be compensated for the risk that they won’t be paid back).
MMTers think this is all, essentially, confused. (Because MMT is a school of thought with many distinct thinkers, I will be using a recent textbook by MMT-supportive economists Mitchell, Wray, and Martin Watts as my main source when describing the school as a whole. But do keep in mind that individual MMT thinkers may depart from the textbook’s analysis at some points.)
For one thing, they adopt an older view, known as the endogenous money theory, that rejects the idea that there’s a supply of loanable funds out there that private businesses and governments compete over. Instead, they believe that loans by banks themselves create money in accordance with market demands for money, meaning there isn’t a firm trade-off between loaning to governments and loaning to businesses of a kind that forces interest rates to rise when governments borrow too much.
MMTers go beyond endogenous money theory, however, and argue that government should never have to default so long as it’s sovereign in its currency: that is, so long as it issues and controls the kind of money it taxes and spends. The US government, for instance, can’t go bankrupt because that would mean it ran out of dollars to pay creditors; but it can’t run out of dollars, because it is the only agency allowed to create dollars. It would be like a bowling alley running out of points to give players.
A consequence of this view, and of MMTers’ understanding of how the mechanics of government taxing and spending work, is that taxes and bonds do not and indeed cannot directly pay for spending. Instead, the government creates money whenever it spends.
So why, then, does the government tax, under the MMT view? Two big reasons: One, taxation gets people in the country to use the government-issued currency. Because they have to pay income taxes in dollars, Americans have a reason to earn dollars, spend dollars, and otherwise use dollars as opposed to, say, bitcoins or euros. Second, taxes are one tool governments can use to control inflation. They take money out of the economy, which keeps people from bidding up prices.
And why does the government issue bonds? According to MMT, government-issued bonds aren’t strictly necessary. The US government could, instead of issuing $1 in Treasury bonds for every $1 in deficit spending, just create the money directly without issuing bonds.
The Mitchell/Wray/Watts MMT textbook argues that the purpose of these bond issuances is to prevent interest rates in the private economy from falling too low. When the government spends, they argue, that adds more money to private bank accounts and increases the amount of “reserves” (cash the bank has stocked away, not lent out) in the banking system. The reserves earn a very low interest rate, pushing down interest rates overall. If the Fed wants higher interest rates, it will sell Treasury bonds to banks. Those Treasury bonds earn higher interest than the reserves, pushing overall interest rates higher.
“These activities are coordinated with the treasury, which will usually issue new bonds more or less in step with its deficit spending,” Mitchell, Wray, and Watts write. “This is because the central bank would run out of bonds to sell to drain the excess reserves created by deficit spending.”
But the basic upshot of all this is that taxing less than the government spends, and issuing bonds in tandem, isn’t a problem under most prevailing circumstances, per MMT. The main constraint on government deficits is inflation, but at a time like now when inflation is low, that’s not a serious concern.
Indeed, MMT has incorporated an approach to analyzing deficits — the “sectoral balances” framework — developed by the late British economist Wynne Godley, which implies that government deficits are often necessary to boost savings in the private sector. Godley’s insight was that when the government is in debt, that necessarily means another segment of the economy is running a surplus, either the domestic US economy or the external economy.
So when the US is importing more stuff than it exports (as is normally the case), and the domestic US economy is overwhelmed with debt that it’s trying to get rid of (as was the case after the 2008 crash, as private homeowners and others were left underwater), the government, as a matter of arithmetic, has to run deficits if it wants to help the private sector recover. Indeed, in their textbook Mitchell, Wray, and Watts suggest that the 2001 recession was the result of the US fiscal surplus at that time forcing the private sector into deficit: “In most advanced economies, sharp, severe economic downturns typically follow a period when fiscal surpluses are accompanied by large private sector deficits.”
“In the long term,” they conclude, “the only sustainable position is for the private domestic sector to be in surplus.” As long as the US runs a current account deficit with other countries, that means the government budget has to be in deficit. It isn’t “crowding out” investment in the private sector, but enabling it.
When you lay out the MMT view on deficits, non-MMTers typically have one of two reactions:
The first reaction flows from MMT’s rhetoric about the government always being able to print more money. The image of a government creating infinite piles of cash to finance whatever it wants to spend brings to mind Weimar-era wheelbarrows of cash, as Larry Summers wrote in his critique of MMT:
[i]t is not true that governments can simply create new money to pay all liabilities coming due and avoid default. As the experience of any number of emerging markets demonstrates, past a certain point, this approach leads to hyperinflation. Indeed, in emerging markets that have practiced modern monetary theory, situations could arise where people could buy two drinks at bars at once to avoid the hourly price increases. As with any tax, there is a limit to the amount of revenue that can be raised via such an inflation tax. If this limit is exceeded, hyperinflation will result.
The MMT reply to this is simple: No, our approach won’t lead to hyperinflation, because we take inflation incredibly seriously. Taxes are, they concede, sometimes necessary to stave off inflation, and as a consequence, preventing inflation can require cutting back on deficit spending by hiking taxes. But the lower inflation caused by higher taxes is not an effect of “lowering the deficit”; the lower deficit is just an artifact of the choice to raise taxes to fight inflation.
Like most strands of economics, MMT thinks that inflation can result when aggregate demand (all the purchasing being done in the economy) outstrips the real stuff (consumer goods, factories for corporations, etc.) available for purchase. If there are a lot of dollars out there trying to purchase stuff, and not enough real stuff to purchase, that stuff becomes more expensive — so, inflation.
“The second reason [after making people use the currency] to have taxes … is to reduce aggregate demand,” the Mitchell, Wray, and Watts textbook states. Eliminating all taxes while spending 30 percent of GDP on government functions, they note, would spur a massive increase in aggregate demand, one that might cause dangerous inflation.
This leads into the second argument: that MMT isn’t all that different from standard econ. The most complete expression of this view is in a piece by economists Arjun Jayadev and J.W. Mason for the Institute for New Economic Thinking, a lefty research funder that has backed MMTers as well as more mainstream economists.
Jayadev and Mason argue that MMT, as they understand it, swaps the roles of fiscal and monetary policy. Under standard macroeconomics, ensuring that the economy is at full employment and that prices are stable are the responsibilities of the monetary policy — the Federal Reserve — which can achieve both goals by manipulating interest rates. If the Fed hits a 0 percent interest rate, then fiscal authorities (Congress and the president) can come in to boost aggregate demand and get the economy moving again, as the 2008 and 2009 stimulus measures attempted. But normally, it’s all the Fed’s job.
In MMT, the fiscal authority is in charge of both. Most MMTers are of the view that the interest rate set by the Federal Reserve should always be 0 percent — in part because they think the use of government-issued bonds that bear interest is a mostly pointless practice. “Our preferred position is a natural rate of zero and no bond sales. Then allow fiscal policy to make all the adjustments,” Mitchell wrote in a 2009 blog post. “It is much cleaner that way.”
To Jayadev and Mason, this looked a lot like a normal economic model, with the roles switched. Instead of raising interest rates to fight inflation, you raise taxes.
MMTers were not pleased with this characterization, with three prominent MMT writers (Scott Fullwiler, Rohan Grey, and Nathan Tankus) explaining in a letter to the Financial Times:
When we suggest that a budget constraint be replaced by an inflation constraint, we are not suggesting that all inflation is caused by excess demand. Indeed, from our view, excess demand is rarely the cause of inflation. Whether it’s businesses raising profit margins or passing on costs, or it’s Wall Street speculating on commodities or houses, there are a range of sources of inflation that aren’t caused by the general state of demand and aren’t best regulated by aggregate demand policies.
Thus, if inflation is rising because large corporations have decided to use their pricing power to increase profit margins at the expense of the public, reducing demand may not be the most appropriate tool.
In other words: Inflation doesn’t usually result from too-high aggregate demand, which taxes can help cool. Instead, it comes from monopolists and other predatory capitalists using their market power to push prices higher, and it can be tackled by directly regulating those capitalists.
But even when too much demand does result in inflation, Fulwiller, Grey, and Tankus say we shouldn’t necessarily jump to taxes as a solution. “When MMT says that a major role of taxes is to help offset demand rather than generate revenue, we are recognizing that taxes are a critical part of a whole suite of potential demand offsets, which also includes things like tightening financial and credit regulations to reduce bank lending, market finance, speculation and fraud,” they write.
Grey has pointed, for example, to France’s credit regulations in the post-WWII era as a potential inspiration. Those limited and redirected bank lending, which is one way to lower aggregate demand without new taxes. If it’s harder for companies and individuals to get loans, they’ll take out fewer loans and buy less stuff.
So if MMT prescribes various regulations (and, where necessary, taxes) to control inflation, while keeping interest rates at zero, how does it plan to achieve full employment?
This is an idea that predates and transcends MMT as a school of thought, with advocates among non-MMT economists like William Darity Jr. and Darrick Hamilton, and a history of support from American labor unions and civil rights leaders. The basic concept is that the government would offer, as a right of citizenship, a job at minimum wage (usually $15 an hour for these purposes) with benefits, working for the government or a nonprofit, to any adult who wants one.
This is different from subsidized employment, which exists in limited forms now, and even from the massive public works programs of the New Deal like the Civilian Conservation Corps and the Works Progress Administration, which employed millions but did not guarantee jobs to all.
The idea behind such a sweeping and universal program, in the context of MMT, is to ensure full employment no matter what policies the government is adopting to fight inflation. Indeed, the job guarantee is in part a way to keep wages down, or at least keep them from continually rising, to prevent an inflationary spiral.
Absent a job guarantee, raising taxes excessively could slow economic activity and cost jobs, as could regulations that attempt to crack down on certain industries. A job guarantee would be able to enroll anyone hurt by those measures and make sure they’re still employed somewhere.
In the Mitchell/Wray/Watts textbook, the authors argue that both the MMT approach and the mainstream approach fight inflation in ways that generate “buffer stocks” of workers. In the mainstream approach, inflation is controlled by raising interest rates, which slows economic growth (sometimes to the point of recession) and puts people out of work, creating a buffer stock of unemployed people. That buffer stock, that increase in unemployment, is the cost of fighting inflation. This trade-off is often represented through a relationship known as the Phillips curve.
In MMT, people in the job guarantee serve as a similar buffer stock. When the government slows aggregate demand, through higher taxes or regulations or some other means, that forces people out of private sector work and onto the job guarantee — not the unemployment rolls.
“Instead of a person becoming unemployed when aggregate demand falls below the level required to maintain full employment, that person would enter the JG workforce,” the authors write.
By contrast, during downturns, a JG would work as an automatic stabilizer, putting spending money in the pockets of laid-off workers and helping mitigate recessions.
Setting the JG wage at the minimum wage is important for anchoring inflation. In tight labor markets, employers sometimes choose to increase wages and pay for the change with higher prices, setting off inflation. But if the JG wage is tethered to the minimum, then employers always have the option of hiring workers from the JG pool, who, under the theory, can be hired at the low fixed wage given to them in the JG program. That gives them a way to avoid raising wages and setting off price increases. “There can be no inflationary pressures arising directly from a policy where the government offers a fixed wage to any labor not wanted by other employers,” the textbook authors write.
It may be surprising to think of the job guarantee as a way to control, rather than bid up, wages, but this is the explicit intention described in the textbook. The authors write, “Would the incumbent workers use the decreased threat of unemployment to pursue higher wage demands? That is unlikely. … [T]here might be little perceived difference between unemployment and a JG job for a highly paid worker, which means that they will still be cautious in making wage demands.”
This vision of the job guarantee as a tool for controlling workers’ wages is somewhat at odds, at least rhetorically, with MMT’s messaging that a job guarantee is a humanitarian measure. JG jobs are probably better than involuntary unemployment, sure — but the macroeconomic role they’re playing here, in part, is in the interest of price stability, not worker well-being.
Matt Bruenig, a vocal MMT critic from the left, has argued that using a job guarantee to discipline worker wages bears an uncomfortable resemblance to the “workfare” efforts of the 1990s, a characterization that MMT advocates have vocally disputed. “The program is based on the principle of ‘fair work’ not ‘workfare,” Pavlina Tcherneva, a Bard economist and arguably the leading MMT researcher on job guarantee policy, writes. “It does not require people to work for their benefits. It is instead an alternative to existing workfare programs.” But there’s nonetheless a tension between using the job guarantee to provide good, desirable jobs and ensuring that it sets a low enough fixed wage that it’s not inflationary.
That was a lot of theory, and frankly, a lot of it is much more nuanced than how MMT is likely to be employed in practice. Barring a radical shift in the culture of central banking, and the dominant views of both major political parties, I don’t see some of the key operational recommendations of MMT being adopted anytime soon.
Committing to a zero interest rate policy permanently, for instance, would be a dramatic move by the Fed, effectively a repudiation of its statutory commitments to ensure price stability and full employment. Indeed, it’s unimaginable to me that that could happen without an act of Congress repealing those statutory obligations and mandating a zero rate.
Similarly, a US decision to stop issuing Treasury bonds would disrupt a key part of the international financial system, where US government bonds are used as a go-to risk-free asset to which other bond interest rates are linked. That feels similarly inconceivable.
Where I could see MMT having an impact is in the realm of domestic policymaking. Already, multiple 2020 candidates, including Sens. Bernie Sanders, Cory Booker, and Kirsten Gillibrand, have embraced a job guarantee, in various forms.
And more generally, I think it’s likely that MMT will help give intellectual respectability to the notion that Democrats don’t have to pay for everything they want to do, be that a Green New Deal or Medicare-for-all or a big middle-class tax cut.
To be sure, it is not the only force pushing in that direction. Perhaps the most important influence is the behavior of the Republican Party. Ronald Reagan exploded the budget deficit by enacting massive tax cuts and defense spending increases, which his cuts to welfare spending couldn’t hope to match. George W. Bush blew up the first balanced budget in a generation with two rounds of tax cuts and two immensely expensive foreign wars — as well as a massive financial crisis at the end of his tenure. And in barely two years in office, Donald Trump has passed his trillion-plus-dollar tax cut package, with proposals for lower spending existing mostly as an annual pledge in his budget proposal, never to be actually enacted.
Game theorists have known for decades that one of the best ways to generate cooperative behavior in a prisoner’s dilemma-type game is a tit-for-tat strategy: If your opponent cooperated last time, you cooperate, and if they defected last time, you defect.
Democrats have effectively been offering to cooperate and pay for all their budget proposals, or even entertain (as under Obama) and enact (as under Clinton) big bipartisan balanced budget deals — even as Republicans repeatedly defect and show no interest in paying for anything. The rational move in such a game is to start defecting yourself, and declare that you’re not going to pay for anything either.
So even if you want to generate balanced budgets in the future, Democratic deficit spending might be a way to get Republicans more on board with that going forward. And MMT just strengthens Democrats’ bargaining position in this regard, as it lets them send a credible signal that they don’t even think it’s a good idea to pay for everything.
