“I think it means the end of deflation is just around the corner.”
— Hirokaza Yuihama, Daiwa Institute of Research, on the Japanese economy (2006)
“The contractionary mindset of the Japanese economy caused by 15 years of the deflationary mindset is now beginning to change.”
— Shinzo Abe, Prime Minister of Japan (2013)
“Unlike snow, Japan’s deflationary mindset won’t melt easily.”
— Haruhiko Kuroda, Governor of the Bank of Japan (2018)
“Tomorrow when I wake or think I do, what shall I say of today? That with Estragon my friend, at this place, until the fall of night, I waited for Godot?”
— Vladimir in Waiting for Godot
The economic impact of COVID-19 has raised the stakes for policy makers everywhere, and central bankers are all in. But perhaps none more so than Jerome Powell, Chairman of the U.S. Federal Reserve. From rate cuts to liquidity support and from swap lines to asset purchase schemes, the Fed has opened up the floodgates. It’s still not clear when they’ll be closed again. But with all this new money gushing into the economy, shouldn’t Chairman Powell, the rest of the FOMC, and everyone else be worried about inflation?
It’s just ECON 101, right? Too much money chasing too few goods results in inflation. Double the supply of money and prices inevitably follow, if not immediately then at some point in the not-too-distant future. It’s only a matter of time. But when? One day. One day, inflation shall come. We are now just passing the time, waiting for inflation.
The Quantity Theory of Money
“Inflation is on its way,” say the apostles of the quantity theory of money (QToM). “We can prove it, a priori,” they say, and point to a version of the equation of exchange: MV = PY. Both sides of that equation represent nominal GDP, which is the monetary value of the total amount of output (i.e. goods and services) produced within an economy over a given period of time. Each side of the equation just uses different variables to arrive at the same quantity.
On the left-hand side, we have the total supply of money (M) multiplied by the rate at which money switches hands, or the velocity of money (V). On the right-hand side, we have the average price (P) of a unit of output multiplied by total output (Y). Both sides are a measure of nominal GDP, and the assumptions made by the QToM about this equation provide us with a rationale for why money printing might lead to inflation.
Money’s velocity, or the rate at which money circulates, is assumed to be fixed, determined by entrenched spending habits and social practices. Paychecks arrive every two weeks, bills must be paid every month, beer on Fridays and groceries on Saturdays, etc. The business cycle may induce certain deviations in money’s velocity — rising in booms and falling in busts — but the general trend is flat, fixed by that great deadener known as habit.
Total output is also assumed to be fixed, at least in the short run. The factors of production — land, labor, and capital — are operating at close-to-full capacity. If more output is to be produced, then more land must be acquired, more workers must be birthed or immigrate from abroad, more factories and equipment must be built, or at the very least, the respective productivities of each factor must be increased. All of this takes time.
If V and Y are fixed, then a rising M must lead to a rise in P in order to maintain the equality, MV = PY, so the reasoning goes. In other words, if the total supply of money is increased, then we should expect inflation. Unlike money’s velocity and total output, whose changes depend on either external shocks or the progression of time, the supply of money is at the discretion of those who run the printing press, which can be booted up at a moment’s notice.
The Digital Printing Press
“We print it digitally,” says Fed Chairman Pozzo, …er Powell. “As the central bank, we have the ability to create money, digitally.” Translation: the Fed can increase M. But how exactly does a digital printing press work? It’s not a secret. Powell says, “we [create money] by buying Treasury bills, or bonds, or other guaranteed government securities.”
In ordinary times, the purchase of securities (and also their sale) is one tool by which the Fed implements monetary policy, a tool which goes by the name of open market operations (something discussed in more detail here). In extraordinary times, the Fed expands its security purchases, big time. The total number, the maturity of the securities, and the range of specific types of securities purchased, are all ratcheted up. This is called quantitative easing (QE).
We’ve been here before, and not that long ago. The last time the Fed went on a shopping spree for assets was in response to the 2007–2008 global financial crisis and Great Recession that followed. The chief architect of that program was former Fed Chairman Ben Bernanke. Here’s a few words from him on the subject.
Bernanke starts with the unprovocative claim that the Fed “lends” money to banks. If you think of the securities the Fed purchases as collateral for the money credited to banks’ reserve accounts, then it is a kind of lending (especially if at some future date the transaction is reversed or unwound). But then he confesses that this is much more “akin, although not exactly the same, to printing money than it is to borrowing.” Why? The money that is lent never existed prior to the loan being made.
Bernanke then explains why the Fed is doing this: “we need to do that, because our economy is very weak and inflation is very low.” Emphasis on the low inflation. “When the economy begins to recover, that’ll be the time when we need to unwind those programs, raise interest rates, reduce the money supply, and make sure that we have a recovery that does not involve inflation.” Bernanke’s thoughts are haunted by the QToM, but his first priority is recovery. Inflation is only a secondary concern at this point.
That interview was done in 2009, not long after the Fed began its first round of QE. Two more rounds followed, the last of which ended in October 2014. The graph below illustrates those three rounds of QE and their impact on the size of the Fed’s balance sheet. The blue line represents the number of assets held by the Fed.
The Fed’s balance sheet stopped increasing after the end of QE3 and remained fairly steady over the next three years. Around the beginning of 2018, the Fed began to reduce the size of its balance sheet, selling securities instead of buying them. But it didn’t get very far before it had to reverse course, resuming asset purchases in 2019, and well before anyone even knew what COVID-19 was. Then, the blue line goes right vertical — that you can blame on the pandemic. Money printer go brrr…
Bernanke’s non-Inflationary (non-)Recovery
The Fed’s preferred measure of inflation is the Core Personal Consumption Expenditures (Core PCE) price index (note that Core PCE excludes food and energy prices as these tend to be much more volatile). The Fed’s goal is for this price index to rise at an annual rate of 2%. That means an inflation rate of 2% each year.
Since the big drop in 2009, core inflation has only edged above the Fed’s 2% target a few times, once around the start of 2012 and once more around the middle of 2018. Most of the rest of the time, it’s been below target. Now it’s called a target for a reason, and not an upper limit — persistently low inflation risks turning into deflation, which can be just as harmful, if not more so, than inflation. For a blatantly severe case, think Great Depression. For something more insidious and current, think Japan’s deflated generation (here).
But it looks like Bernanke got what he wanted, a non-inflationary…hmm, recovery? Some would question the strength of that recovery, but put that question aside for the moment. The present concern is how to think about the QToM considering that despite all the money printing, inflation has been essentially nonexistent. The Fed more than quadrupled its balance sheet in a span of about five years and all we got is sub-2% inflation. And it’s worth noting that the Fed never did “unwind” its balance sheet as Bernanke claimed would be necessary to avoid inflation.
The QToM is just one version of how to interpret the quantity of exchange, MV = PY. While that equation implies an equality, it says nothing about the behaviour of the individual variables, nor does it imply any causal relationship between them. If there is any causal relationship, that relationship derives not from the equation itself but from assumptions about how the world works. There are at least three other versions of how the world works, and which aren’t necessarily mutually exclusive.
1. Output in the Short Run
The QToM assumes that output, or Y, is fixed in the short run and that’s because the factors of production that produce output are assumed to be operating at full capacity. If there is an increase in the supply of money that leads to an increase in spending, output cannot increase because the factors of production are already being fully utilized. Prices will rise to curtail the demand. But what if the factors of production, and thus the businesses that employ them, are not actually operating at full capacity?
One way we could test this assumption is by looking at the capacity utilization rate, which measures the proportion of total productive capacity being utilized. For example, if the capacity utilization rate is 100%, then the economy is fully utilizing its capacity to produce output. Anything less than 100% means there’s some room to increase output before capacity is maxed out. It’s a measure of just how much productive slack there actually is in the economy.
The capacity utilization rate was about 90% around the middle of the 1960s. Over the past 50 years there has been a lot of variation. Capacity utilization plunges during recessions (the shaded areas) and then gradually recovers. However, each recovery period has generally led to a capacity utilization rate that is lower than it was prior to the recession. The period following the Great Recession is no different. Is that what recovery looks like? Ben? Oh, and then April 2020. The economy just fell off a cliff.
Another way to measure slack in the economy is through the labor market. This one’s a bit tricky. The unemployment rate fell to 3.5% in the latter half of last year. It hasn’t been that low in 50 years, suggesting tightness rather than slack. A tight labor market generally means workers have more bargaining power to demand higher wages and other benefits. If firms want the workers, they have to fork over the cash. In order to maintain profit margins, they might increase the prices of the goods and services they supply. Now that’s a recipe for inflation. But we’re still waiting.
The problem with the unemployment rate as a measure of labor-market slack is its bluntness. To be counted among the employed, you only have to work one hour per week. One. Hour. Per. Week. That doesn’t sound like anyone’s full labor capacity. Of course, one hour per week is probably not that common. But the point of the matter is that the headline unemployment rate makes no distinction between the fully employed and those working part time, and who may be wanting to work full time.
In early 2018, the San Francisco Fed found that the level of involuntary part-time employment was about 40% higher than one would normally expect it to be at the current stage of the recovery (here). Part-time, not to mention gig-type work is usually not associated with strong labor bargaining power, especially if it’s involuntary. Nobody demands higher wages when there is already a scarcity of hours to work. If there’s any labor struggle going on, it’s likely to be between employees battling for hours. Meanwhile, new grads take unpaid internships just to get some experience. Such conditions may generate fissures of resentment, which sometimes split open along racial, generational, or other lines of difference.
Also, the labor force participation rate — the proportion of the population that is considered part of the labor force, both unemployed and employed — is significantly lower than it was prior to the Great Recession. Part of this could be due to the greying of the population, leading to more people retiring than are entering the labor force. But it is significant how much it slumped immediately after the Great Recession, and never recovered, an involuntary early retirement of sorts.
The unemployment rate can fall not just when the unemployed find work, but also when the unemployed drop out of the labor force altogether. If five out of a hundred labor-force participants are unemployed, the unemployment rate is 5% (5 / 100 = 0.05 * 100%); if one of the five who are unemployed, tired of looking for a job that isn’t there, drops out of the labor force, the employment situation “improves” as the unemployment rate falls to 4% (4/99 = 0.0404 * 100%). Is that what recovery looks like? Ben?
All this is not to say that the employment situation had not improved at all over the past decade or that it was still in a dismal state. One might even go so far as to say that it had recovered, ten years later. But it probably was not as hot as some were trumpeting. Powell said it best: “to call something hot, you need to see some heat” (here). The problem, in his opinion, was that wages weren’t responding (probably due to a lack of bargaining power). He continued, “While we hear reports of companies finding it hard to find qualified labor, we don’t see wages responding.”
Companies saying they are finding it hard to find qualified labor is just another way of saying they don’t want to pay to train people, they want the government to do it. If the labor market was tight and the economy was booming, firms would be throwing out all kinds of incentives to attract anybody with two feet and heart beat. Firms would not only train workers, they’d actually offer paid training. Imagine that was the way the world worked. Alas, we’re still waiting.
That was last year. As of April 2020, the unemployment rate was sitting around 15%, capacity utilization at about 65%, and we just saw what happened to the labor force participation rate. That’s a lot of slack.
Barring forced lockdowns on economic activity, there’s no reason to think that the supply of output is constrained in the short run. Since it’s not constrained, any new money creation that stimulates demand for goods and services is more likely to be met by increases in productive activity. That means new goods and services will be produced to soak up all that new money. In other words, Y increases, not P.
If anything is constrained at the moment it is demand for output, including demand which arises from business investment spending. But don’t expect businesses to go on a shopping spree for new plant and equipment any time soon with the capacity utilization rate as low as it is. If corporates go on any kind of spending binge, it’s more likely to be on buying back their own stock.
2. Total Output or Total Transactions?
MV = PY is a common version of the equation of exchange. But there is another version that looks like this, MV = PT, where T stands for the total number of transactions and P is just the price of a typical transaction. If the only transactions that exist are GDP transactions, then T is effectively Y and the two equations are identical. But if we can identify transactions that are not GDP transactions, then the equations are distinct and newly-created money does not necessarily go towards the purchase of newly-produced output.
Enter financial assets. Newly-created money can be used to purchase financial assets like stocks and bonds. This can lead to a certain kind of inflation, but its a different sort of inflation than the fear-mongers usually warn about when they warn about inflation. It’s called asset-price inflation, referring to a significant rise in the prices of financial assets.
Take a glance back at the graph above showing the Fed’s QE programs and rise in the assets held on the Fed’s balance sheet. See that yellow line? It represents the S&P 500 Index, a market index of large-cap U.S. stock prices. Those prices have soared. They have also soared relative to GDP, suggesting that newly-created money has been disproportionately used for financial asset transactions by comparison with transactions for domestically produced goods and services.
As of the end of 2019, the market capitalization (i.e. number of shares outstanding multiplied by the share price) of U.S. equities reached above 200% of GDP for an all-time high. Aha! So this is what recovery looks like, eh Ben?— the value of so-called fictitious capital is soaring while real economic output limps merrily along. It looks like people are more interested in purchasing ownership shares of U.S. companies than the goods and services that those companies provide.
Of course, “people” might not be the right word. Some of the biggest purchasers of equities in the past year have been the corporations themselves. Share buybacks have exploded over the past decade and especially in recent years. At least somebody’s finding a use for all that cheap money the Fed’s been pumping into the financial system.
Bond prices have also been rising, which means bond yields have been falling (bond prices and bond yields are inversely related). Indeed, bond prices have risen so much that we now live in a world where negative yielding debt is a commonplace. That’s right. Negative yielding debt, the kind where the lender pays the borrower. As of August 2019, approximately 25% of government bonds worldwide were trading at negative yields (here).
Now why would anybody want to buy an asset with a negative yield? The answer: capital gains, safety, liquidity, and central banks. Capital gains can still be earned on negative-yielding bonds if their prices continue to rise. As for safety, the average person can achieve this with a bank deposit account. But bank deposits are the liabilities of private banks, and they are only insured up to $250,000 by the government. For institutional investors with billion-dollar cash pools, the uninsured liabilities of private banks don’t cut it; only the sovereign liabilities of the most credit-worthy governments do. Deep financial markets for the liabilities of these credit-worthy sovereigns provide the liquidity. And of course, we already know why central banks are buying them.
The fact that a significant portion of sovereign debt now carries negative yields suggests the demand for these safe assets is outstripping the supply. Governments just aren’t issuing debt fast enough. Huh. What? Did somebody just say there’s not enough government debt?! Or maybe, there’s just too much cash floating around looking for a safe home and, not finding one, is forced into riskier assets like privately-issued mortgage backed securities. Meanwhile, there’s always somebody that could use some cash just to buy some food, clothing, or shelter — that is, real goods and services. But looks like they’ll have to wait, too.
All that to say that we may just be looking for inflation in the wrong places. All this time we’ve been looking for it in markets for real goods and services when perhaps we should’ve been looking for it in financial markets. But before ending things here, there is at least one more version of how the world works, which provides one more reason why we might not get too worried about inflation in the current environment. Shall we go for it? Sure. Why not? It’ll pass the time (of course, it’ll pass anyways).
3. You Can’t Force Households or Businesses to Borrow
When you hear arguments that money printing necessarily causes inflation, the precise mechanism as to how this actually occurs is often left a bit vague, if it is acknowledged at all. It’s like, yeah, you know, if you print money, prices, they go like — boom! You know? Let’s try to fill in some of those intermediate steps, the “yeah”s and “you know”s.
Step one: print the money. We demonstrated above how the Fed digitally prints money by purchasing debt securities and crediting private banks with reserve accounts. Reserves are the newly created money. So what? Banks now have more reserves. How do you go from reserves sitting on the balance sheets of private banks to rising prices of goods and services?
Step two: banks lend more. The assumption here is that more reserves provide banks with a greater capacity to lend to businesses and households. More reserves equals more credit availability.
Step three: buy stuff with the newly-available credit. The additional private credit is used by firms to fund current business activities or expand into new ones, and by households to increase their consumption. The resulting increased demand for goods and services, which are limited in supply by factors of production operating at full capacity, leads to rising prices — inflation.
That’s the mechanism, and now that it’s been clearly delineated, it’s much easier to see what the assumptions are. First, we have the assumption that bank lending is constrained by the amount of reserves in the banking system, and second, the assumption that there is some unlimited demand for credit by firms and households. For a critique of the first assumption, see here. The second assumption is more pertinent to our present concern.
Based on the mechanism just presented, given the claim that money printing necessarily leads to inflation, it must be the case that households and businesses will always soak up any new credit being offered by banks. We might add that the claim also relies on the assumption that banks are always willing to lend in order to meet this insatiable demand for credit. Both assumptions are doubtful.
Banks are willing to lend when they believe they can make a profit by doing so, and this depends on whether or not they believe the borrower will repay the principal of the loan plus interest. The less likely the borrower appears to be able to make those payments, the riskier that borrower is. The bank may still be willing to lend but only at an increased premium, which means a higher interest rate.
Interest rates are the cost of borrowing money. The borrowing decisions of both households and businesses are sensitive to this cost. If interest rates rise, borrowing money will be less appealing. But businesses and households are also sensitive to their perceived ability to repay the money they have borrowed at some future point in time. Failure to repay comes at the expense of lost credit-worthiness and the consequent inability to borrow in the future.
The ability to repay is itself dependent on economic conditions. When business is booming and jobs are secure, both households and businesses will find it easier to repay their loans. Confidence in the ability to take on more debt is high. But when the economy slows and unemployment rises, that confidence falters. There’s also less incentive for businesses to invest in new projects since the economy is slowing. Fearing job losses, households will also limit any unnecessary consumption in favor of replenishing their rainy day funds. The appetite for debt will wane under such conditions.
But it’s precisely under such conditions that central banks begin to inject reserves into the banking system with the aim of lowering interest rates. Yet this is also the precise moment at which the risk of defaults on loans begins to rise. Business revenues decline and households whose members have lost their jobs start missing or deferring loan payments. Risk premiums rise, pushing interest rates up. The more severe the downturn the more the central bank will have to do in order to lower interest rates, like introducing QE.
The central bank’s lowering of interest rates is aimed at stimulating demand for credit, because credit can then be used to purchase goods and services. If the stimulus is effective, the economy will begin to recover. But the basic problem facing central banks is that while they can entice businesses and households to borrow more through lower interest rates, they can’t force them to do it. If the economic downturn is bad enough, monetary stimulus will be like pushing on a string.
Conclusion: Still Waiting for Inflation
It’s been more than a decade now since the Fed unleashed the first round of QE. The Fed’s balance sheet exploded after three rounds. It never did get back to pre-crisis levels, and we never got any inflation, at least not the kind that should worry anybody. Since the start of the pandemic, the Fed’s balance sheet has nearly doubled in size, shooting past the $7 trillion mark in May. Here we are again, waiting for inflation.
In April, with much of the U.S. economy in lockdown mode, inflation fell to a rate of 1.0% year-over-year. That’s the lowest level since December 2010 (here). It’s not surprising. Demand for output has been decimated. Soaring unemployment and lost paychecks mean consumption demand is down. Cratered capacity utilization means demand from business investment is down. The rest of the world is in the same conundrum, meaning demand for exports is down. It’s hard for prices to increase when there’s no demand to pull them up.
This is not to deny that the fallout of the pandemic may produce some inflationary pressures. Decades of globalization, trade cooperation, and increasing integration of the world’s economies has provided abundant sources of cheap labour and greater economies of scale. That has helped to keep prices of output down. If in response to the crisis countries become more insular, global trade may dwindle and globalization may begin to reverse course. A negative supply shock like that could be inflationary. In that case, rising costs, rather than money printing, would push prices up.
But all else equal, the bigger threat right now is deflation. Is it contagious, the deflationary mindset? Will the U.S. economy catch that Japanese virus? Given that demand from consumption, business investment, and exports are likely to be weak for the foreseeable future, the deflationary mindset is a serious risk, but it has nothing to do with Japan, the Japanese, or their economy. It could happen to any economy.
It’s a macroeconomic problem and requires a macroeconomic solution. John Maynard Keynes offered us a solution in response to the Great Depression of the 1930s. The current pandemic-induced crisis could be just as bad, worse even. Keynes showed us that in the face of decimated demand from households, businesses, and abroad, there is only one other source of demand — government spending. Recent years have seen the call for a Green New Deal, a call which appears more pressing than ever.
The danger is that the inflation mongers, not to mention the deficit hawks, instil enough fear to usher in a new era of austerity. They will surely draw on the QToM and the equation of exchange to support their arguments. But remember, the QToM is just one version of how one might interpret the quantity of exchange. We’ve shown there are at least three other ways to think about it. Why do people tend to flock to the one version? Who knows — people are bloody ignorant knuckleheads, or apes, or however the damn line goes. Can we do that one over again?