Modern Monetary Theory and the Challenge of Endogenous Money Supply

For a hot minute in 2022 and 2023, MMT or Modern Monetary Theory rose to prominence in the lexicon of the mainstream media reporting on monetary policy and inflation. If you recall talk about a trillion-dollar coin, that was related to MMT. Although it’s been largely dismissed after our recent bout of inflation, MMT had one intriguing idea (missed by some folks) beneath all the nonsense: the fact that the money supply and overall spending can fluctuate independent of Federal Reserve policy.

First, a point of orientation. MMT is a fringe area of monetary economics whose advocates primarily reside at the University of Missouri Kansas City and Bard College (which hosts the Levy Institute). They are part of Post Keynesian scholarship, most prominently developed by Hyman Minsky. In fact, they consider themselves truer torch-bearers of Keynes than more mainstream Neo Keynesians like Greg Mankiw or Jason Furman.

Advocates of MMT argue that the money supply, and the price level, are determined by market activity rather than by the central bank. In the pre-2008 Fed operating framework where the Fed pushed the Fed funds rate to its target by buying and selling bonds, they had a point. While the Fed could target the fed funds interest rate, it could not simultaneously target a monetary aggregate such as the monetary base, M1, or M2 because the demand for loanable funds in the overnight bank lending market was determined by banks and market conditions.

Suppose the Fed funds rate target sits at 4 percent. If bank demand for reserves increased for some reason, maybe because they expanded lending, maybe because they wanted more liquidity, the federal funds rate will rise ceteris paribus. But the Fed doesn’t want the rate to rise, so they intervene. They buy more bonds, increasing the total quantity of reserves, to put downward pressure on the interest rate and maintain their 4 percent target. But notice that the Fed has expanded its balance sheet in response to changes in bank behavior (and indirectly in response to changes in M2).

So far, so good.

But then the MMT folks go on to argue that Fed policy doesn’t matter – that it is entirely ineffective and only market forces drive changes in money aggregates and inflation. Which leads them to suggest that the Fed can buy any amount of US debt without fueling inflation. And again, there is a slight kernel of truth to this. 

Many folks, including me, were sure in 2009 and 2010 that we would see significant inflation due to the Fed’s ballooning balance sheet. But it didn’t happen. From 2007 to 2019, the Fed’s balance sheet increased from roughly $830 billion to $3.3 trillion (an increase of about 300 percent). M1 increased about 200 percent and M2 increased about 100 percent. The Consumer Price Index only increased by 25 percent. In 2020, M1 spiked dramatically, but most of this spike was driven by a change in its calculation. The Fed decided to add savings accounts to their calculation of M1, making it much more like M2.

In money and banking textbooks, the monetary base and the money stocks (whether M1, M2, or M3) are supposed to move together closely. And, with certain assumptions, the price level is also supposed to track changes in the money supply. That’s not what we see in Figure 1, which shows different growth rates for money aggregates and inflation from June 2007 to 2019 (prior to the changed calculation of M1). 

Economists argued two things to explain the ’07 to ’19 period. First, they argue that restrictive regulations and the Fed paying interest on reserves significantly weakened the link between the monetary base and M2. Second, they argue that the velocity of money fell over this period, offsetting some of the increase in M2. That’s why we saw low inflation rather than high inflation over this period.

Both explanations are plausible. But they reinforce the MMT story about the irrelevance of a ballooning monetary base. A couple significant things changed with the response to COVID-19. The Federal Reserve increased its balance sheet dramatically — although that was not new. More important was that M2 increased significantly as well due to putting hundreds of billions of stimulus dollars directly in people’s checking accounts. 

The Fed also creates greater potential for explosions in spending that can cause inflation by flooding the market with liquidity. And despite the Federal Reserve reducing its balance sheet by $2 trillion dollars since it began tightening, its balance sheet remains enormous. There is a lot of money still out there in the economy.

The most important equation when it comes to inflation is MV=Py. This accounting identity states that the total quantity of money (M) times its velocity (V), the average number of times a year a dollar is spent, equals the total spending in the economy. On the other hand, all of the goods sold in a given year (y) multiplied by the price paid for those goods (P) also equals the total spending in the economy.

MMTers argue that monetary policy and the Fed’s balance sheet are merely accounting entries that do not affect the real economy. That is to say, they believe that velocity is the key variable. Money simply responds, through the Fed maintaining its targets, to changes in velocity — which MMTers, following Keynes, think is driven primarily by government spending.

This is why they think we do not need to worry about Federal indebtedness: the Federal Reserve can buy any amount of federal debt issued without necessarily affecting velocity. In fact, velocity will likely change to offset changes in the money supply. Instead, they claim, we should only ask the question of whether the amount of government spending is too stimulative or too restrictive. 

But they are wrong to only point the finger at government fiscal spending – though that is indeed important. What matters when it comes to economic growth and economic efficiency is resource allocation. And far from simply changing accounting entries, when the Federal Reserve buys assets, especially non-government assets, it affects the allocation of real resources by redirecting flows of financial capital and distorting the price of capital.

The question at the end of the day really becomes what will happen to the velocity of money and the growth in output? These are, indeed, endogenous variables that the Fed does not directly control. So a critical point of MMT holds. Yet Fed responses to future “crises,” reducing short-term rates to artificially low levels again, and messing with the long end of the yield curve can distort economic calculation. 

It would be far better if the Fed took as light a touch as possible and left economic calculations to those with local knowledge and actual skin in the game.