星期三, 六月 17, 2009
Yes, We Really Do Have Lots of Spare Capacity
· A new research note from the San Francisco Fed suggests that the economy could have much less spare capacity than traditional measures suggest. If true, this is important, as spare capacity has strong predictive power for the inflation outlook – the more spare capacity, the less pricing power for businesses, and the more downward pressure on inflation. · However, we think the SF Fed’s approach probably underestimates spare capacity. It infers the current level of spare capacity in part from the recent behavior of inflation, i.e. falling core inflation points to the existence of spare capacity; rising inflation suggests the economy is operating at or above potential. This approach seems better suited to retrospectively analyzing past output gaps than estimating the current level of spare capacity, because it takes time for spare capacity to affect inflation (it’s future inflation that will really tell us something about current spare capacity). · In any case, we have plenty of real-time measures of spare capacity available, most of which show record or near-record levels—far in excess of any possible measurement error. Firms will undoubtedly try to reduce capacity, and many are already taking actions to do so, but in some areas of the economy it could be years before potential supply more closely approximates demand.
The inflation outlook remains a hot topic for investors, with a wide range of views in the marketplace from Japan-style deflation to 1970s-style stagflation to hyperinflation. Our view has been and remains that in the near term, spare capacity—the excess of potential supply over demand levels at current pricing—is the biggest driver of inflation. When factories are running flat-out and cannot increase output, firms have more pricing power; when the unemployment rate is at record lows, workers are more likely to get raises (or get a better offer at another employer); when vacant housing is scant, rents and home prices are more likely to increase. Conversely, excess capacity in any of these areas tends to be associated with greater levels of price competition and hence lower inflation. Our analysis suggests spare capacity is important both when measured at the economy-wide level and for inflation in individual industries. The broadest measure of spare capacity is the so-called “output gap” – the difference between the economy’s actual output (GDP) and its potential. According to the Congressional Budget Office (CBO), the output gap was -6.2% in the first quarter of 2009, i.e. the economy was operating 6.2% below its potential level. By the CBO’s measure, this is the largest output gap on record (the data go back six decades) except for the Q3 1982 - Q1 1983 period (the period immediately before and after the trough of the 1981-82 recession, the deepest recession in the postwar period). Put simply, even with the current recession not yet over, the economy has developed a near-record cushion of surplus capacity. If the CBO analysis is even close to correct, the disinflationary consequences could be large; core CPI inflation fell about six percentage points from the beginning of the 1981-82 recession until 1985, by which time the output gap was mostly closed.
A new research note from the San Francisco Fed (“How Big is the Output Gap?”, FRBSF Economic Letter 2009-19, June 12, 2009, by Justin Weidner and John C. Williams) implies that the CBO might be overestimating the output gap by a substantial margin. By their calculations, it is “only” -2% of GDP—still meaningful but somewhat smaller than at the trough of the 1990-91 recession and far smaller than at the depth of the 1981-82 recession. In theory, this should still be sufficient to create some downward pressure on inflation, but the impact would be milder and any recovery in demand or cuts in capacity would close the output gap more quickly. As our benign inflation view rests to a considerable degree on the belief that the economy has lots of spare capacity, it is important to examine this analysis closely. Indeed, the Weidner/Williams comment has generated a number of questions from clients – particularly those more skeptical about the inflation outlook!
The logic of the SF Fed position is that in the environment of the financial crisis and its aftermath, potential output probably isn’t increasing as rapidly as the CBO estimates…hence the amount of spare capacity may not be as big as those estimates suggest. We certainly have sympathy for the idea that capacity growth will slow at least temporarily, and even that capacity will be destroyed or reduced outright in some areas (e.g. the auto industry). However, we think it’s important to keep three basic points in mind: 1. Even if the SF Fed’s estimate of spare capacity is right, it still implies meaningful disinflation over the next year or two. First of all, both the SF Fed and CBO estimates are for the end of the first quarter, and the economy has clearly contracted in Q2, even if the rate of contraction is slowing—so the output gap is larger now and will likely be larger still after Q3, as even the most optimistic forecasters generally do not expect above-trend growth that quickly. Closing a gap of >2% of excess capacity implies the economy needs to grow at least 2 percentage points faster than trend for one year, or 1 percentage point faster than trend for two years. If we take an estimate of 2½% trend real GDP growth (a bit more conservative than our own), this implies that even with 4½% growth over the next four quarters, we would have excess capacity until mid-2010; with 3½% growth, until mid-2011. Our own estimates are for 1%-2% growth over this period. The bottom line: even with a very optimistic growth forecast, we should have spare capacity through next year. So risks to core inflation probably should remain to the downside into 2011 even if the SF Fed’s estimate is a better one.
2. The SF Fed’s methodology is probably more appropriate for estimating past output gaps (spare capacity) than the current output gap. The SF Fed’s methodology, developed in a paper by Thomas Laubach and John C. Williams (“Measuring the Natural Rate of Interest,” Review of Economics and Statistics 85, November 4, 2003, pp. 1063-1070) infers the level of spare capacity in part by what is happening to core inflation: falling core inflation points to the existence of spare capacity (output below potential); rising inflation suggests output is above-potential. This makes sense given the hypothesized relationship between spare capacity and inflation – but there is a catch. The catch is that the output gap works with a lag: prices do not react instantaneously to spare capacity. (One way to see this is that over the postwar period, a substantial amount of disinflation has occurred during recoveries after recessions, rather than in recessions themselves.) So how can we estimate the current output gap? We need to know future inflation data to infer the current output gap, and of course we don’t have it yet. Taken to a reductio ad absurdum, applying the Laubach/Williams approach to an estimate of the current output gap could be seen as arguing that if core inflation hasn’t been falling recently, then it won’t in the future. 3. Numerous direct measures of spare capacity are at or near record levels. Anyway, we don’t need to take an indirect approach to estimating spare capacity when we have so many direct, near real-time, measures of utilization. The table below offers a wide range of measures across the economy, and the overwhelming message is that spare capacity is at or near record highs. (For more discussion of service-sector spare capacity measures, see “Spare Capacity Mostly on the Rise in Service Industries,” US Daily, April 16.) The rightmost column shows that most measures indicate that spare capacity is two or even three standard deviations above average; this seems more consistent with the CBO’s estimates of spare capacity than with those implied by the SF Fed’s analysis. Skeptics may rightly point out that any of the measures in the table could be revised, but for most measures the levels of excess capacity seem far larger than any conceivable measurement error.
Spare Capacity - At or Near Record Highs
For more detail on our inflation views, please see three recent US Economics Analyst publications: “Inflation—The Only Thing to Fear is Fear Itself” (June 5) for a more detailed analysis of the impact of spare capacity on inflation in different sectors of the economy; “Is the Fed’s Big Balance Sheet an Inflation Risk” (June 12) and “Hyperventilating About Hyperinflation (April 9) for a discussion of why we do not think Fed asset purchases and reserve growth will cause a major inflation problem. We have also discussed specific aspects of the inflation debate in a number of our US Daily publications in recent weeks.
Andrew Tilton
Goldman Sachs Financial Conditions IndexSM*(October 20, 2003=100)
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
订阅：
博文 (Atom)