I’m glad to be here at the Cato Monetary Conference, a conference that over many years has made an important contribution to monetary policy. We come together at an important moment for monetary policy, a period of significant inflation after 40 years of relative price stability.

I’d like to make three observations broadly on monetary policy challenges going forward. First, under just about any theory that we have in economics, inflation at this moment was quite predictably an outcome of the policies that were being pursued.

For those who, like me, take a perspective that’s more based on levels of aggregate demand relative to levels of supply, you had a combination of substantially negative real interest rates, a two-trillion-dollar savings overhang from resources people were not able to spend during the COVID-19 pandemic, and massive fiscal expansion in the United States totaling some 14 percent of gross domestic product (GDP) in 2021. Against an output gap of perhaps 3 percent of GDP, that made a substantial overflow inevitable.

It’s true that there were bottlenecks and supply chain issues, but many of those disturbances weren’t really disturbances on the supply side. They were bottlenecks that arose from very strong aggregate demand. That there would be bottlenecks and constraints on supply in the aftermath of the pandemic should not have been surprising. So for me, the inflation that has arisen was an entirely predictable consequence of excess demand. But I don’t think that inflation hitting 9 percent can be blamed entirely on excess demand.

I don’t think the central problem was one of conceptual understanding. I think the central problem was a failure to act on the conceptual understandings that are already established in economics. You saw it in rhetoric that came out of the central banking community in late 2020 and early 2021, emphasizing their social responsibilities. They were emphasizing issues that monetary policy can’t really address, like the particularly high unemployment rates facing minority groups and disadvantaged populations.

My own reading of the evidence is that restraint applied more vigorously and more credibly is ultimately less costly in terms of unemployment—in terms of lost output— than monetary restraint that is not credible. Just as nations that are better prepared for war are often less likely to have to fight war, central banks that are credible in their anti-inflation commitments will find fulfilling those commitments to be less costly.

I welcome the evolution of the Fed’s rhetoric over the past six months in this direction, most notably in Jerome Powell’s recent remarks in Jackson Hole. But I continue to believe that it is very important to recognize that disinflation has costs and that it is important to signal a willingness to accept those costs rather than undermine credibility by projecting serenity about the ability to achieve a soft landing. The grim truth is that ideas about soft landings are, in my view, like what Samuel Johnson said of second marriages: the triumph of hope over experience.

My second observation goes to policy over the substantially longer run, and it has to do with the period after the 2008 financial crisis in the United States. I am convinced that there is a broad range of forces that are operating to reduce neutral real interest rates. Some are on the savings side, including increased uncertainty, increased life expectancy, and an increased share of income going to those with very high incomes. Some are on the investment side, including slower population growth, which means there is less need for education and training, less need to house new workers, etc.

Some of the causes relate to the price of capital goods. A cellphone today has more computing power than a Cray supercomputer did in 1993, and yet it only costs $600. Oil drillers tell me that one rig can account for twice as many wells as was the case five years ago, for example. These sorts of examples have proliferated. They all have in common an increase in the supply of capital relative to the demand. That means neutral real interest rates are going to be substantially lower in the future than they have been in the past, to put it differently. There will be a tendency for savings to flow into existing capital goods rather than new capital goods because of reduced investment opportunities. That means higher asset prices, which means more tendency to take on leverage. I regard these tendencies in the pre-COVID-19 period as being driven by economic fundamentals rather than by monetary policy.

I believe that the likely consequence of maintaining low rates would have been even more sluggish growth and even more of a deflationary tendency than what we observed in the industrial world. I think there is a significant chance that as the extreme fiscal stimulus associated with COVID-19 leaves the system, and that if inflation is brought down, we will return to such a situation where the absorption of savings will be a crucial macroeconomic challenge. One of the great debates will be whether that absorption of savings is best achieved through policy-determined very low real rates, possibly even negative real rates, or whether it is best pursued by fiscal and other structural policies directed and operating on the level of saving and the level of investment.

My instincts, but I cannot prove that I am right, are on the latter side. I suspect that very low real interest rates court financial instability, make excess leverage and bubbles more likely, and also slow the process of creative destruction of failed or should-be-failed enterprises. If I am right, I think we will need to consider direct policies, such as changes in the level of deficits, redistributions from richer to poorer or from younger to older, measures to inhibit precautionary saving by providing better insurance, or measures to directly stimulate investment in key areas such as the environment, where the social return exceeds the private return. These may be better strategies for dealing with a savings absorption problem, but I think this reality is one we may well need to face again within a few years.

My third observation is of a different sort. I have become increasingly convinced in recent years that central bank communication has lost its way. I believe that an earlier generation of central bankers such as Alan Greenspan and particularly Paul Volcker understood what the Delphic oracles understood: that if you are perceived as omnipotent and omniscient, it is a good idea to keep your statements vague so as to not destroy an illusion that gives people confidence going forward.

The current habit in the central banking community of frequent forecasts, frequent press conferences, and reliance on forward guidance seems to me to be quite problematic because inevitably forecasts turn out to be wrong. Promises turn out to not be doable, and the result is a substantial loss of credibility. The predictions made by the Fed in the spring of 2021, that interest rates would remain at zero well into 2024, seem to me to be a particularly extreme example of this phenomenon. The central difficulty with forward guidance in my view is that the markets don’t believe it, so it doesn’t have much of an impact on longer-term rates and therefore not much of an impact on market behavior to stabilize the economy.

On the other hand, central banks take their own forward guidance seriously and therefore feel constrained to adhere to the policies they promised, even when those policies prove unwise in light of subsequent conditions. So my hope would be for central banks to adopt policies that are humbler in recognizing inherent uncertainties and that try less to precisely forecast or guide than they have in recent years.