(Bloomberg View) -- The Federal Reserve projects three interest rate hikes next year, although the market is priced for fewer.
Yet, recent public statements by senior central bank officials suggest some disagreement over the causes of the recent moderation in inflation. That means policy makers might be reluctant to continue raising rates beyond the Open Market Committee meeting this week without evidence that inflation is moving more decisively toward the 2 percent target. So, it is not surprising that markets are priced for fewer policy actions and that the yield curve implies investors expect rates to remain low for some time. These views indicate that bond investors and Fed officials are in for a rude awakening , as it is very likely that higher inflation lies ahead.
Inflation has been better behaved than history would suggest for this stage of the business cycle. Unemployment has fallen to a cycle low of 4.1 percent, a level that has been associated with some acceleration of inflation in the past. The Phillips curve and the non-accelerating inflation rate of unemployment theories, which garnered Nobel prizes in Economics for Edmund Phelps and Milton Friedman, provide solid theoretical frameworks to support the historical precedent for expecting upward pressure on inflation and rates. Somehow, all of this is being dismissed in the face of some moderation in inflation this past spring. Does this make sense?
The theory is very difficult to dismiss, since the concepts of supply and demand, the most basic of all economic concepts, are the critical foundation beneath it. A.W. Phillips offered no theoretical foundation for his empirical observation in the 1950s that declining unemployment drove up labor costs. Rather, economists simply accepted that increased labor scarcity would be associated with faster inflation and they focused on the implied trade-off between the two and the idea that policy makers could exploit the relationship to choose their preferred mix of inflation and unemployment. This simplistic view of the Phillips curve was severely undermined in the 1960s, when the curve proved incompatible with the coexistence of high inflation and high unemployment.
Phelps and Friedman repaired and enhanced the simple model by suggesting that the correct relationship was between changes in inflation and deviations in unemployment from its natural rate. Any time unemployment was below the non-accelerating inflation rate of unemployment, known as NAIRU, inflation would accelerate. And if unemployment was constant at NAIRU, inflation would remain unchanged and NAIRU could be compatible with any and all possible rates of inflation. In effect, the simple Phillips curve model becomes increasingly vertical the longer the economy deviates from NAIRU. This insight was widely lauded by economists.
Today’s confusion stems from the incorrect assumption that the trade-off between changes in inflation and deviations of the natural rate of unemployment is forever fixed and does not change. But there is absolutely no theoretical reason to support such an idea. In fact, it is easy to think of reasons why the relationship can change over time and that policy makers would be motivated to influence it.
Any institutional change that improves the workings of the labor market would be considered desirable by policy makers. For example, state employment agencies, which have been created to help the unemployed in their search for work, would reduce the cost to workers of being unemployed by reducing their search time. All other factors equal, this should lower NAIRU and also allow any given level of unemployment to be compatible with a lower level of inflation. Similarly, the increased use of job-search sites on the internet to more quickly and easily match up workers with job openings helps reduce the inflation cost of lower unemployment. Indeed, this is precisely the change that has occurred within the past decade and we are now enjoying the benefit with unemployment at levels that in the past tended to raise labor costs and inflation. And it does not mean the NAIRU model is wrong. Rather, it has merely shifted in a favorable direction.
Most critically, the evidence does not require a rejection of the NAIRU model. Such a move would be highly problematical for policy makers as they would be flying blind without any basis on which to anticipate when inflation might increase or when they must start raising rates to contain it.
Nonetheless, the shift of the curve has made their task vastly more difficult, since they cannot rely on historical low levels of unemployment to signal upward pressure on inflation. Moreover, shifts need not be restricted in one direction toward better trade-offs. For example, if unemployment insurance benefits are increased, the curve should shift in a less favorable direction. Or, incentives to avoid moving, because property taxes are reset only when someone moves, will also inhibit job shifting, which also worsens the trade-off.
Putting this all together, policy makers are unlikely to know when they need to hike rates until it is too late. After they go over the edge of the cliff and faster inflation is manifest, they will know only after the fact they missed their opportunity to increase rates to contain inflation. At best, they can rely on historical levels of unemployment to anticipate changes in inflation, but with a significantly reduced level of confidence, which opens the door to debate and disagreement within the FOMC. This is precisely the circumstance that exists within the FOMC right now.
Some implications for investors and Fed officials are clear. First, because unemployment is already very low by all historical experience, an acceleration in inflation could occur at any time. Investors ignore this risk at great peril to their bond portfolios, which are priced for an implausibly long period of tame inflation. Even at the low 2 percent growth rate of the past few years, unemployment is likely to remain on its solid downward trajectory, but gross domestic product growth has picked up and any fiscal stimulus will only reinforce growth. It is highly risky to think that inflation will remain benign just because it has been so.
Second, every labor cost and inflation report should be examined closely to determine if the economy has passed its launch point for inflation. Indeed, it is the labor cost data that should be concerning already.
Average hourly earnings, reported each month in the employment report, have increased from 2 percent last year to 2.5 percent most recently. Moreover, this measure is downward biased, because hiring over the past few years has been concentrated among the lowest-paid, least-educated workers, which depresses the average calculation. The upward trend in labor costs is corroborated by the acceleration in the Employment Cost Index, which tries to adjust for changes in the job mix. The ECI has also increased from below 2 percent to almost 3 percent over the past two years. Since labor accounts for the lion’s share of the cost of producing GDP, it is likely that the economy will soon be seeing higher inflation in the data, even if we can only guess on how quickly this becomes evident.
This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
Charles Lieberman is chief investment officer and founding member at Advisors Capital Management LLC.
For more columns from Bloomberg View, visit http://www.bloomberg.com/view.
©2017 Bloomberg L.P.