I have not worried much about inflation in recent years, but I’m beginning to. The economy is running hot, wage rates are rising, and demand growth is about to exceed supply growth. Now I predict that inflation will accelerate. The Federal Reserve won’t tolerate this, so look for them to tighten interest rates, with the risk of error leading to recession.

The numbers on the economy are good, with inflation-adjusted GDP up four percent last quarter and unemployment below four percent. That’s high for GDP, low for unemployment. But the economy has been growing and unemployment falling for years, with little inflation acceleration. The difference now is that the slack has been removed from the economy. Unemployed people are few and far between, and underutilized computers and machinery are becoming rare.

One sign of inflationary pressure is wage rates. The wages component of the employment cost index grew at its fastest pace since 2004. (This is not the old cost-push argument, that wage increases cause inflation. Rather, I use wage increases as one signal, among many, of inflationary pressures that originate elsewhere.)

As I updated my economic forecast to incorporate new GDP data and revisions to historical data, my numbers looked really good. Too good to be true, I feared. So I jumped from my demand model to my supply model, described in Economic Forecast 2018-2019: Demand Is Strong But Supply Is Weak. The demand model was very optimistic, with growth rates in the high threes. But the supply model was not impressed, limiting growth to the low twos. I resolved the difference by limiting demand growth and stimulating supply expansion. The demand revision worked across all sectors, with special emphasis on service sectors that are especially human resources dependent. On the supply side, I concluded that even more people would decide to work, and productivity growth would be even greater than I had previously estimated. I brought my GDP forecast up to three percent, from the previous projection of 2.7 percent. But the mismatch between the demand model and the supply model translates into higher inflation. It won’t soar as it did in the 1970's, but it will exceed not only the Fed’s two percent target, but three percent, too.

Plenty of caution should accompany any inflation forecast. Across the range of economic forecasters, inflation predictions have done very poorly in recent years. We’ve been looking at declining unemployment rates and expecting higher inflation. It turns out that the best forecast over the past few years would have been hard-wired at two percent, irrespective of economic growth. But now, I say nervously, things are different.

Consumers feel worse off when prices rise, but usually their incomes fully keep up. The real cost of inflation comes from families and households rearranging their affairs to reduce the risk of unforeseen changes in purchasing power. That’s pretty minor in the short run, but becomes a significant problem when higher inflation is tolerated by the Fed. Economies with higher inflation also tend to be less stable, giving the Fed more reason to act.

But monetary policy acts with time lags. If inflation accelerates as I expect, the Fed is likely to dial up interest rates rapidly. I have recently bumped up my interest rate forecast, predicting that the Fed Funds rate rises by 1.5 percentage points a year, rather than the one point per year they currently expect.

If the Fed gets the timing and magnitude of their policy changes exactly right, then we have nothing to worry about. (In my speeches, I always pause for laughter after saying this.) Tightening too hard too soon risks putting the economy into a slump, though maybe not a full-blown recession. But tightening too little too late risks inflation acceleration, which must be countered by severe tightening. That would induce a recession. Either way, recession risk is higher.

I’m not yet forecasting a recession in the next two years, but I am dialing up my risk estimate. The need for a recession contingency plan is greater now than in the last few years.