Scott S. on my inflation views

Scott Sumner writes,

I do think the Fed is more agile, but the decisive factor is that the Fed is much stronger. If you want a metaphor. . .imagine I’m driving my car and my 6-year old daughter pushes the steering wheel to try to change direction. I’d simply push back more strongly.

Read the whole thing.

Also, you should read Scott’s post on bond market vigilantes.

The 1990s were a successful period for monetary policy. . .We can infer from stable 2% inflation that the actual interest rate stayed pretty close to the natural interest rate during the 1990s.

Scott is fond of saying “Never reason from a price change.” In Fantasy Intellectual Teams, if other players use that, his owner will score M points.

The way I understand this maxim is that prices are an endogenous variable that can be affected by many things. So you cannot safely draw inferences from a price change without knowing more about what went on.

If we generalize the rule to say that you cannot safely draw inferences from changes to endogenous variables, then Scott violates his own rule promiscuously. Take the quoted passage, for example. Can we safely infer that because inflation was stable that the interest rate stayed pretty close to the natural rate?

It sounds like Scott is providing an explanation for the low inflation of the 1990s, based on a concept called the “natural rate of interest.” But the natural rate of interest is something that we cannot observe. It plays a role in macroeconomics that is analogous to the role played by “systemic racism” in Critical Race Theory.

For Scott to convince me of monetary dominance, I would like to see him reason from an observable definition of loose money or tight money. Saying that an episode of slow growth in nominal GDP proves that monetary policy was tight is convincing to someone who already believes in monetary dominance, but not to a skeptic.

A few more remarks:

1. I think of inflation as determined by the behavior of paper wealth relative to real output. If wealth grows rapidly and persistently relative to real output, then eventually consumer prices will increase markedly. The sort of deficit spending we have undertaken in 2020 and 2021 has added a lot to paper wealth and nothing to output. So watch out.

2. I think that most of the time neither fiscal policy nor monetary policy has much effect on paper wealth.

3. My story for Japan is that before the government started creating lots of paper wealth, a lot of paper wealth got destroyed in the collapse of Japanese real estate and equity prices. So Japanese consumers were not flooded with wealth on net.

4. I think that the government can monetize a fair amount of debt and get away with it, as long as it looks as though it will collect enough future taxes to pay off bondholders. It’s when the government has no choice but to monetize that things get ugly. And it’s up to bond investors to calculate when the government is going to have no choice, and to try to sell their holdings before that perception becomes widespread. That is the sort of sovereign debt crisis that you really don’t want to go through.

Q&A on my inflation views

For context, read Tyler (and Scott S.) vs. me on inflation.

1. What if the Fed chairperson explained monetary dominance out loud?

That would mean telling Republicans that tax cuts don’t stimulate and telling Democrats that more government spending doesn’t stimulate. So the Fed chairperson would alienate Congress. Retaliation would likely follow.

How could we distinguish empirically the dose-response model from the autocatalytic model of inflation?

A big challenge is to pin somebody down on how to measure the dose. There are many measures of the money supply to choose from, ranging from a narrow definition of Fedcoin plus currency to a broad definition that includes all financial assets. The current fad is to measure the dose as the difference between the interest rate and “r-star,” which is a mythical interest rate at which the economy is in perfect balance. And here, the choice of which interest rate to compare to “r-star” could vary. Finally, Scott Sumner says to heck with all of these indicators–just look at nominal income as the measure of monetary policy. That threatens to be circular. To be fair, Scott wants to use a market measure of expected future nominal GDP, which removes the circularity. But it also makes it hard to do an empirical study, since we do not have direct measures of market forecasts for nominal GDP–it is up to the investigator to try to find a proxy.

We need to agree on a way to measure dose. I want the dose to be something that the Fed controls pretty directly, as opposed to the measure of money that includes all financial assets or even Scott’s preferred measure of market expectations for nominal GDP. But then you run into the problem that the Fed’s operating procedures have changed periodically, especially with paying interest on Fedcoin.

But if you could agree on a dose measurement a priori (meaning you don’t cheat and look at historical data to try to pick a measure that best correlates with response), then in principle you could look at past experience. What you would be looking for is cases of modest but significant changes in dose. If the conventional wisdom is correct, you should see significant response. If the autocatalytic view is correct, you should see no significant response.

Do you really believe that there is nothing the Fed can do to affect the economy?

Of course I don’t go that far. The Fed could sell $400 billion on bonds tomorrow. Or Janet Yellen could walk into the New York Stock Exchange with a bomb strapped to her waist and blow the whole place up.

But if we stick to normal Fed operations, which are small and gradual, I believe that those are easily absorbed by financial markets without affecting anything important. Obviously, this is an outlier view.

Note that the fact that people believe that the Fed affects things gives it the ability to affect things, just as the fact that people believe that Elon Musk’s tweets matter gives Elon Musk the power to affect stock prices with his tweets. But I don’t think that those effects are very long-lasting in either case.

Tyler (and Scott S.) vs. me on inflation

Tyler Cowen writes,

To see why the huge increase in bank reserves did not result in inflation, consider that there has been a considerable decrease in U.S. excess bank reserves over the last five years. No one claims that this has been accompanied by a massive deflation, whether in securities markets or elsewhere. Once that point is conceded, it’s possible to see why higher levels of reserves are not necessarily inflationary.

Let me stress that his views are fairly mainstream. It is my views that you should especially doubt. I hold outlier views in two ways. Rather than argue against Tyler, I will argue against what I think Scott Sumner would say. I hope this deals with Tyler en passant.

1. Fiscal dominance vs. monetary dominance. The government tries to control the level of nominal income using fiscal and monetary policy. If you are a worker, your nominal income is your salary. If you are a self-employed yogurt maker, your nominal income is the revenue from your yogurt sales, less the cost of inputs.

If the government wants to use fiscal policy to raise nominal income, it can run a larger deficit. Congress votes to send Paul a stimulus check for $1000 by borrowing the money from Peter, giving Peter a Treasury bill. Paul feels $1000 richer, and Peter does not feel poorer, because he expects to be paid back. (As academic economists will tell you, there is actually a longstanding dispute on this point. Just Google “Barro are government bonds net wealth” or “Ricardian equivalence.”)

If the government wants to use monetary policy to raise nominal income, the Fed obtains Peter’s Treasury bill, paying for it using a digital asset, called bank reserves, or Fedcoin, if you will. The more Fedcoin that banks have, the more freely they will lend, and the more freely the public will spend.

Scott’s argument for monetary dominance is that the Fed, which sets monetary policy, is way more agile than Congress, which sets fiscal policy. It’s like a game of rock, paper, scissors in which if Congress shows rock, the Fed shows paper. Or if Congress shows scissors, the Fed shows rock. The Fed can always win.

Consider the $1.9 trillion stimulus Congress is debating. Even though it would be an adverse supply shock, as quantified by Casey B. Mulligan and Stephen Moore, it would tend to raise nominal income. If it passes, the Fed can decide to be less expansionary in order to keep nominal income on target. If it fails, the Fed can be more expansionary and still hit the target. Note that no Fed chairperson would ever say this out loud; instead, the Fed chairperson is obligated to tell Congress that whatever it plans to do is exactly what the economy needs and thank heaven for Congress, because the Fed could never do the job all by itself.

I believe in fiscal dominance. That is because I do not think that Peter cares all that much whether he hangs on to his T-bill or exchanges it for money. Scott thinks that Peter will spend more in the latter case. I am skeptical.

In this regard, my views coincide with the Modern Monetary Theory of Stephanie Kelton. (Rest assured that her views and mine differ in many other respects). She would say that Fedcoin is merely non-interest-bearing government debt (although since the financial crisis of 2008 the Fed has paid interest on Fedcoin). I might prefer to say that T-bills are interest-bearing money.

2. Inflation as an autocatalytic process

Scott, like almost all mainstream economists, sees inflation as having a continuous dose-response pattern. Give the economy a higher dose of money and it will respond with higher inflation. Other economists measure the “dose” as the employment rate.

I think of inflation as an autocatalytic process. Inflation is naturally low and stable. But it can be jarred loose from that regime and become high and variable. Then it takes a lot of force to bring it back to the low and stable regime.

Another example of an autocatalytic process is a social media platform. If you want to try to build the next Facebook, it is really hard to get started. But once enough people join, then their friends will want to join, so growth becomes automatic.

When inflation picks up to an annual rate of 8-10 percent, it changes your behavior. I know, because I remember the 1970s. When you run a business and you see your suppliers and workers demanding 10 percent more than they did a year ago, you cannot ignore that when you set your price. When you are a worker and see the cost of the stuff you buy going up 10 percent per year, you need to demand a raise just to keep up.

The real take-off point for inflation in the 1970s was the New Economic Policy of President Richard Nixon, announced in August of 1971. He let the dollar “float,” meaning that it depreciated in world markets. In a misguided attempt to stifle inflation, he imposed wage and price controls. In order to work properly, a capitalist economy must have freely moving prices. The controls were a self-inflicted adverse supply shock. Adverse supply shocks raise prices (and recall that the latest “stimulus” is an adverse supply shock on steroids). Although for a little while the price controls repressed inflation, the more enduring effect–the supply shock–went in the other direction. Note, too, that inflation itself is a supply shock, because a lot of the steps that households and businesses take to protect against inflation are steps that detract from productive activity.

Once inflation gets going, the only way to stop it is to slam on the economic brakes. Usually, this means drastically cutting government spending. But in the U.S. in the early 1980s, we slowed the economy without cutting government spending. Instead, the foreign exchange market put on the brakes by raising the value of the dollar, stimulating imports and making our exports non-competitive. And the bond market put on the brakes by raising interest rates, so that nobody could afford the monthly payment on an amortizing mortgage. After a few years of high unemployment, inflation receded.

Most economists attribute these developments to Fed policy under the sainted Paul Volcker. Scott could say that this was exhibit A for monetary dominance. The economic consensus may be right, but I would raise the possibility that the financial markets were the main drivers.

What about more recent experience? As I see it, since the 2008 crisis Congress has been undertaking ever-more-reckless deficit spending, throwing match after match on the firewood, without starting an inflation fire. Maybe that pattern will persist. But if an inflation fire does get going, I will be less surprised than the markets.

The inflation outlook

Jon Hilsenrath writes,

It is easy to find reasons for discomfort. Tesla’s stock price is up more than 300% in the past year. Copper prices are up 56%. The S&P Case-Shiller Home price index is up 9.5%. Freight prices are up 215%; soybeans, 54%, lumber, 117%.

John Greenwood and Steve H. Hanke write,

Speculative manias are in the air, as evidenced by the recent price surges for bitcoin, a digital asset with a fundamental value of zero, and GameStop, a declining retailer. Along with the other economic trends—a strong recovery, surging commodity prices and an uptick in inflation—those asset bubbles have a clear cause: the massive expansion of money and credit.

My picture of the situation right now is that the paper wealth created by various tranches of “stimulus” is flying around the asset markets, chasing Elon Musk tweets (ht Matt Levine). It is like airplanes stacked up over LaGuardia, flying in circles waiting for permission to land. At some point, people will resume spending their wealth on consumption. That is when the inflation will shift out of asset markets and into the prices of goods and services.

What will be the average annual rate of inflation between now and February 2027? I would give the following probabilities:

less than 2 percent: p = .5
between 2 and 4 percent: p = .2
between 4 and 6 percent: p = .1
over 6 percent: p = .2

Am I willing to bet? If my goal were to get rich, I would be buying an ETF that shorts long-term Treasuries. But my goal is to try to maintain what I have. I am trying as hard as I can to have a portfolio that is defensive against a high-inflation scenario. I probably should be investing in commodities.

The most likely reason that my view differs from the market’s is that I am much less confident that the Fed can stifle inflation when fiscal policy is so loose. The standard Fed response to inflation would be to buy less government debt. That would allow interest rates to rise. That causes government spending to rise by (change in interest rate)x(amount of debt to be rolled over). Because there is so much debt to be rolled over, this is a big deal. Other things equal, it puts more paper wealth into the hands of the public, so that it is less anti-inflationary than would be the case if we were starting with less outstanding debt.

And I predict intense political resistance to allowing interest rates to rise.

Deficits and inflation: a longer historical overview

Michael Bordo and Mickey D. Levy write,

the initial response combined aspects of the policy response in several overlapping crisis scenarios in the past: World Wars I and II, the Great Depression, and the Global Financial Crisis (Bordo, Levin and Levy 2020). These earlier episodes of induced fiscal and monetary expansion in the 1930s and the World Wars led to rising price levels and inflation. In this paper we survey the historical record for over two centuries on the connection between expansionary fiscal policy and inflation and find that fiscal deficits that are financed by monetary expansion tend to be inflationary.

Do not assume that the last ten years settle the issue of whether deficits are inflationary.

Dependency ratios and inflation

In a podcast with Rob Johnson of George Soros’ Institute for New Economic Thinking, Charles Goodhart and Manoj Pradhan offer an unusual explanation for secular trends in inflation. They say that a decline in the dependency ratio creates excess supply of workers relative to consumer, putting downward pressure on prices. This trend has started to reverse, so we will see upward pressure on prices. They say that this upward pressure will become visible soon after the virus crisis is over.

Think of this in worldwide terms, with the dollar as the universal unit of account. Over the last several decades, the world labor supply rose because of population trends and the inclusion of more countries, notably China, in the world production system. The share of consumers in the population remained steady, as a decline in birth rates offset population aging by reducing the growth of young dependents.

Going forward, the labor supply will grow slowly and population aging will outpace any further decline in birth rates. So the world dependency ratio will rise, and this will put upward pressure on wages and prices.

Note that money plays no role in this story. What about “Inflation is, anywhere and everywhere, a monetary phenomenon”? Perhaps if you include the hypothesis that the real mission of the central bank is to hold down government borrowing costs, you can tell the story in a way that Milton Friedman would not object. That is, as dependency ratios were falling, there was enough worldwide saving to keep down interest rates, and central banks did not have to monetize a large share of government debt, so that inflation fell. Going forward, worldwide saving will fall, and central banks will face a dilemma. As inflation appears, they will want to stop it by raising interest rates. But if they do that, governments won’t be able to afford the interest cost on their debt, so central banks will be forced to monetize a large share of debt.

Toward the latter part of the podcast, they are asked about why markets differ from their forecast. Basically, they say that markets are extrapolating forward based on the past, in which demographic pressure on inflation was downward. It will take markets a while to realize that we are in a new regime.

Finance theory and the Fed

Eugene Fama says,

Actually, the central banks don’t do anything real. They are issuing one form of debt to buy another form of debt. If you are an old Modigliani–Miller person the way I am, you think that’s a neutral activity: You’re issuing short-term debt to buy long-term debt or vice-versa. That’s not something that should have any real effects.

Pointer from Alex Tabarrok.

Two papers that influenced my view of banking in general and central banking in particular were Bank Funds Management in an Efficient Market, by Fischer Black, and Banking in the Theory of Finance, by Fama. Both take seriously the modern theory of financial markets that begins with Modigliani-Miller. One could see smoke coming out of Scott Sumner’s ears even before he responded.*

In a Modigliani-Miller world, the financial structure of one firm does not matter. Investors use the financial markets to adjust their portfolios to undo the effects of one firm’s changes to its financial structure. The public ultimately holds what it wants to hold. If it doesn’t like the mix of securities that one firm creates (by substituting bonds for equity, for example), it has ways of dealing with that. The metaphor that I would propose is that a single firm’s changes to its financial structure is like me sticking my hand in the ocean, scooping up water, and throwing it in the air: I don’t make a hole in the ocean.

Modigliani-Miller is not strictly true. But it is the best first approximation to use in looking at financial markets. That is, you should start with Modigliani-Miller and think carefully about what might cause deviations from it, rather than casually theorize under the implicit assumption that it has no validity whatsoever.

Taking this approach, I view the Fed as just another bank. Its portfolio decisions do not make a hole in the ocean. That heretical view is the basis for the analysis of inflation in my book Specialization and Trade.

*Sumner’s response is actually beside the point, in my view. The Modigliani-Miller theorem does not in any way rely on different asset classes being close substitutes. It relies on financial markets offering opportunities for people to align their portfolios to meet their needs in response to a corporate restructuring.

So you don’t have to

I read Stephanie Kelton’s The Deficit Myth and wrote a review.

It is indeed correct to say that when the government is bidding for resources, the risk of inflation is low if those resources are idle. It is also correct that unemployment is an indication of idle resources. But just because some resources are idle does not mean that the government can spend wherever it would like without affecting prices. The government would have to be an especially perspicacious and adroit entrepreneur to advance its priorities while only using idle resources.

By the way, as I read the market for U.S. Treasuries, investors are betting that Kelton is right.