High inflation is unlikely but not impossible in advanced economies
Is there deflation or inflation in our future? Some observers point to falling commodity prices, stumbling oil prices, and a depressed labor market and see low inflation, perhaps even deflation as far as forecasts go. Others point to the very large increases in fiscal deficits and central bank balance sheets and see inflation, perhaps even high inflation.
I put most of my probability mass on the low inflation forecast. But I cannot completely dismiss a small probability of high inflation.1 Let me explain.
The standard way of thinking about inflation is to look at the state of the labor market, inflation expectations, and shocks to commodity and food prices. This framework has served us decently well (not great, as attested by debates about the death of the Phillips curve) for the last 30 years. And through those lenses, it is hard to see inflation picking up any time soon.
Unemployment is exceptionally high, and even if, when the lockdown is relaxed, it will be partly matched by exceptionally high vacancies, it is hard to see a strong wage push on the horizon. Commodity prices have fallen and oil prices have collapsed, putting downward pressure on inflation.
One might have worried that the large fiscal programs to help liquidity-constrained households and firms would lead to demand exceeding the lower available supply. It has not happened. Instead there has been a large increase in saving, both because of the limits on shopping from social distancing and because of precautionary saving. While prices of some products have increased, inflation rates have, if anything, decreased since the start of the lockdown. (In the United States, the consumer price index decreased by 1.2 percent at an annual rate from February to March). One may still worry that, when social distancing is relaxed, pent up demand will lead to a burst in spending, and some inflation. If it happens, it is unlikely to be large and long enough to destabilize inflation expectations, and it is likely to disappear quickly.
Looking beyond that, it is hard to see strong demand leading to inflation. Precautionary saving is likely to play a lasting role, leading to low consumption. Uncertainty is likely to lead to low investment; unlike a regular war, there is no capital to rebuild. The challenge for monetary and fiscal policies is thus likely to be to sustain demand and avoid deflation rather than the reverse.
This is why I put most of my probability mass on low inflation for the next few years. We are, however, definitely operating in a nonstandard environment, and the standard way of thinking about inflation may be wrong. And I can think of a scenario where there is high inflation.
I believe that three elements must combine for such an outcome.
First is a very large increase in the debt-to-GDP ratio, larger than the 20 to 30 percent or so under current forecasts.
This is not a crazy hypothesis. The exit from disaster relief policies may be very slow, leading to large deficits not only this year but also next. The early lifting of lockdowns is likely to lead to a second wave of COVID-19, and perhaps more. Given the precarious state of many households and firms as a result of the first wave, each successive wave may well require more and more fiscal spending for disaster relief. Multiply the initial fiscal package by 2 or 3 and it leads to a large increase in the ratio of debt to GDP.
Second is a very large increase in the neutral rate of interest, that is, the safe real rate needed to keep the economy at potential.
This may be because the demand for sovereign bonds is downward sloping, and the increase in supply requires an increase in the rate for investors to absorb it. We do not have a precise sense of the effect, and the range of estimates is that an increase of one percentage point in the debt-to-GDP ratio increases the neutral rate by 2 to 4 basis points. So, an increase in the ratio of debt to GDP of, say, 60 percent, might lead to an increase in the neutral rate of 120 to 240 basis points, an increase that would get the neutral rate close to or above the growth rate.2 Or the neutral rate could increase for other reasons, such as a decrease in saving, an increase in investment demand, a decrease in risk aversion; none of these seems likely, but we have a sufficiently poor understanding of the determinants of the neutral rate in the past that we cannot exclude it.
Third, and perhaps most important, is fiscal dominance of monetary policy.
Faced with an increase in the neutral rate, the Federal Reserve should increase the actual policy rate in parallel, in order to avoid overheating. But this would increase debt service, requiring a potentially large fiscal adjustment to avoid a debt explosion. The government might be tempted to ask the Fed to keep the interest rate low, so as to decrease the debt burden. While today's Fed would not yield to such pressure, a future Fed, with a chair appointed by a populist president, might be more willing to bend and keep rates low for too long, leading to overheating and inflation. While some inflation is desirable, the lessons from past episodes of high inflation is that this process can end badly: Inflation expectations might de-anchor, leading to higher and higher inflation, perhaps even hyperinflation. This would reduce the real value of the debt, but not without major costs for the economy.
As I have made clear, a high inflation scenario requires the combination of three ingredients, each of which has a low probability of occurring in advanced economies. Put your own probabilities and multiply them: The resulting probability is very small. I asked some of my colleagues for their probabilities, and the product always came below 3 percent. (The probability is even smaller in the eurozone, where it is hard to see the fiscal authorities get together to impose fiscal dominance on the European Central Bank.) But it is not quite zero.
Looking at the yield curve for inflation-indexed bonds, investors do not appear to anticipate anything like this scenario. They do not see a substantial increase in the neutral rate: The yield curve for inflation-indexed bonds is negative throughout the maturity structure. They do not see an increase in inflation any time soon. The expected inflation proxied by the difference between the rate on nominal bonds and inflation-indexed bonds is about 1 percent below the Fed target of 2 percent throughout. I am on their side, but I do not completely dismiss the probability that things could turn wrong.
1. This blog post focuses on advanced economies, especially on the United States. The issues are quite different for emerging-market and developing economies and deserve a separate analysis.