Why use core inflation?

THERE is a bit of split among the central banks of the rich world over whether it is better to focus policy choices on headline inflation or "core", that is, broad inflation stripped of volatile components like energy and food prices. It seems intuitively right to use headline inflation; after all, people spend money on petrol and bread just as they do on computers and tax preparation services. The use of the core measure may seem opportunistic to laypeople—like playing games with the numbers to get the figures that justify preferred policy choices.But that's not the reason central banks like the Federal Reserve tend to focus on core inflation. Larry Meyer comments:Why do we, as forecasters, and the FOMC in its own forecasts, focus on core inflation? The question should not be whether I buy groceries and gas, but whether headline or core inflation is a better measure of where headline inflation is likely to settle once overall prices have adjusted to the higher prices of energy and food. As forecasters, we want to know not only what (headline) inflation is today, but also, and much more importantly, where headline inflation is likely to be tomorrow (the medium term). Identifying a measure of underlying inflation gives us a good head start...Our thesis, and the FOMC’s position, is that headline inflation converges to core, that is, headline inflation tomorrow will fall towards core inflation today. Whether or not this is a valid thesis is an empirical question...If you test whether higher oil prices raised core inflation over the 1970s and early 1980s (or in samples that include this period), the answer is a definitive “Yes” (higher oil prices pass through to core inflation). In this case, core inflation tomorrow will converge to headline today. However, if you test this hypothesis over the subsequent period, from the mid-1980s to today, the answer is “No” (no pass-through). In this case, headline tomorrow will converge to core today. This is the basis for our forecasts...The Fed has built credibility over the last two decades: Long-term inflation expectations are stable, have been stable for more than a decade, and are likely to remain so. This means that spikes in food and energy prices do not get translated into expectations of higher inflation down the road and, thus, do not lead to a generalized increase in prices, today or tomorrow. So the critical question is whether inflation expectations are well anchored today—we believe that they are—and, more importantly, whether they are likely to remain so...One interesting thing to point out is that in the 1970s real GDP growth was often very high, and likely above potential. Real output grew by more than 5% in 1972, 1973, 1976, and 1978. Labour markets were often tight in this period, especially in the early 1970s, and core consumer price inflation was quite high even before the oil price shock in 1973. Conditions now are obviously quite different, and so it's not necessarily a good idea to think that rising oil and food prices will deliver the stagflation we observed in the 1970s.But an interesting question does arise: what if rising food and fuel prices aren't transitory? What if emerging market growth, for example, will produce a steady, structural shift in the relative prices of scarce commodities? It's pretty clear how a credible monetary policymaker should handle a temporary shock; it's much harder to know what to do in this case. It's possible that headline inflation would persistently overshoot the desired level, even as the central bank held policy tighter than it preferred while growth lagged. In a Sumnerian, nominal GDP-targeting regime in this case, you might end up taking your 5% NGDP target in the form of 1% real GDP growth and 4% inflation for some time, while the economy adjusts.That scenario, should it emerge, will prove challenging. But it does not appear to be what most of today's inflation hawks are worried about; rather, they seem to be concerned that, as in the early 1970s, a reckless Fed is simply failing to take the necessary steps to manage garden variety inflation. This seems wrong to me, because core inflation is well in hand and because long-term inflation expectations remain quite low. And mistakenly tightening in these circumstances is likely to reduce macroeconomic stability—to knock growth off its present, near-trend (and, many would say, insufficient) rate and to send inflation expectations falling, just as they did last summer.