Ex-Fed chair Ben Bernanke is currently a resident fellow at the Hutchins Center on Fiscal & Monetary Policy at the Brookings Institution. The Center’s director, David Wessel, appeared on NPR’s Morning Edition yesterday to express views on inflation and deflation that are essentially indistinguishable from Bernanke’s.
In particular, like Bernanke, Wessel spoke as though inflation below 2% per annum, and a fortiori deflation, is always bad.
Asked to explain why it should be bad to have less of a bad thing like inflation, Wessel first responded: “I mean it sounds appealing, prices going down, people paying less at the store. But when inflation is low that means wages are going up very slowly too. That's of course not very popular.“ This sort of answer would earn an undergraduate a poor grade from any instructor who emphasizes the distinction between real and nominal variables. Workers should not applaud rising nominal wages per se; what matters for them is their real wages. When inflation is low, so too are nominal wage increases ordinarily. But that says nothing about the path of real wages, which in the long run are not determined by monetary policy.
Wessel continued: "There are a couple of problems with too little inflation. It can be a symptom of a lousy economy, one in which demand for goods and services and workers is anemic."
A key word here is symptom. A condition that can be a symptom of a problem is not the problem itself, and can also appear under healthy conditions. Wessel failed to note that low inflation need not be a symptom of a lousy economy, because he spoke only of variation in the aggregate demand for goods and services. Low inflation (or even deflation) can instead be the benign result of abundant growth in the real supply of goods and services. Under the gold standard, as Atkeson and Kehoe (2004) have reported, periods of lousy economy (recession) were not more common during deflation periods, nor vice-versa.
If the Fed were today targeting the path of nominal income, because the growth rate of nominal income is the sum of the inflation rate and the real income growth rate, low inflation would be a sign of a healthy economy with high real economic growth. Thus Wessel would have provided a more accurate analysis if he had spoken of problems with growth in nominal aggregate demand being weaker than anticipated, not inflation being below its target. (In terms of dynamic AD-SRAS analysis, unexpectedly low growth in AD moves the economy below the natural rate of output.)
Wessel’s second concern is harder to interpret favorably. Low inflation, he said, "can make it hard for the Fed to spur borrowing because it's hard for them to get the interest rate below the inflation rate." Two responses: (a) The Fed should not be trying to “spur borrowing.” Fed efforts to spur borrowing helped to fuel the housing bubble. (b) The Fed is not in fact currently finding it hard to get the interest rate below the inflation rate. The interest rate on 1-year T-bills has been below the CPE inflation rate for more than four years, since 2009’s dip into deflation (in that case due to weak demand for goods following the financial crisis).
Wessel went on the cite the problem of rising real debt burdens in deflation. That can indeed be a problem in an unanticipated deflation (Wessel never distinguished anticipated from unanticipated), but again, only to the extent that the deflation is driven by unexpectedly weak aggregate demand growth and not by robust aggregate supply growth. In the latter case, borrowers have more real income with which to repay.
In his answer to the interviewer’s last question, Wessel declared: “So it used to be that economists believed that a central bank can always create inflation by printing more money. But lately it's been -seems harder to do that than the textbooks had told us.” But what is the evidence that expanding the stock of money does not still generate inflation the way the old textbooks tell us? Surely not the experience of the Fed undershooting its target of 2.0% growth in the PCE by 0.4%, which only implies that the broad money growth rate was 0.4% lower than the rate that would have hit the target. Fortunately or unfortunately, it remains the case that the Fed can always raise the PCE inflation rate by engineering a higher rate of broad money growth, just as the textbooks explain.
By the way, today’s announced number for the May CPI raises year-over-year CPI inflation to 2.1%. The increase of 0.4% over April’s CPI, compounded twelve-fold, implies an annual rate of 4.9%. At this rate the “problem” of an undershooting PCE-deflator inflation rate will be “solved” before long.
(HT to David Boaz)