Edward Renshaw
Professor of Economics
State University of New York at Albany
On February 4, 1994 the DJIA plunged 96.24 points after the Federal Reserve revealed it had nudged the federal funds rate upward in an effort to prevent future acceleration in the inflation rate. The preemptive strike against an unobservable phenomenon not only created turbulence in the stock and bond markets but soon came under heavy fire from critics who believed that Chairman Greenspan "may have moved too soon in pushing up interest rates." (Crutsinger, February 22, 1994) The inflation rate for the all item CPI had actually declined by two-tenths of a percentage point during 1993 and during January of 1994 there was no reported increase in the CPI. The tide of informed opinion on Wall Street, however, soon began to claim that "the Fed moved too slow on increasing rates". (Murray April 11, 1994)
In this essay we will briefly examine the success of past efforts by the Fed to prevent an acceleration of the inflation rate and then examine some statistical indicators which may have supported the decision to increase short term interest rates. Forecasting what will happen to the inflation rate is tricky, however, since history would suggest that preemptive strikes by the Fed may make a difference. Food and energy price shocks are very difficult to predict and there have also been some changes in the construction of the CPI index that may make a difference.
There have been two other cases when the Fed moved early and aggressively to "head off" a possible increase in the CPI inflation rate. See those years identified with an asterisk in Table 12.1. The first case occurred in 1959 when the US economy was recovering in a vigorous way from the economic recession of 1957-58. Stephen McNees has noted (1992) that "an acceleration in the inflation in the mid '50s may have been the source of two recessions, the 1957-58 recession born of the necessity to roll back an actual acceleration in inflation, and the 1960-61 recession born out of fear of having to repeat that experience".
In 1983, when the US economy was again recovering "too fast" from a severe recession the Fed allowed the yield on T-bills to increase on a December-December basis by one full percentage point. This preemptive strike was more successful. There was no increase in the CPI inflation rate in 1983 and only a .1 percentage point increase in 1984, when real GDP increased at the fastest year-to-year rate since the Korean War build-up years of 1950-51. The US economy slowed to a more sustainable pace in the last half of 1984 and there was no recession.
In 1988, after a partial recovery of imported oil prices from the collapse of 1986 had caused a 3.3 percentage point rebound in the CPI inflation rate for 1987 and when bad weather and the largest reduction in US crop production per acre since the drought of 1936 was threatening a food price shock, the Fed allowed the Treasury bill rate to rise 2.3 percentage points. An earlier effort to keep inflation under control, by allowing short term interest rates to rise at a rapid rate, had been partially abandoned after the stock market crash of October 1987. The aggressive attempt to halt the inflationary spiral in 1988 was more successful. There was no increase in the CPI inflation rate in that year and the increase for 1989 (when imported oil prices jumped upward by 24.2 percent) was only .2 percentage points. The higher interest rates may have been responsible for a slight decline in industrial production (on a Dec.-Dec. basis) during 1989 but there was no recession until after Iraq's invasion of Kuwait in the summer of 1990.
There was no change in the CPI inflation rate associated with the 2.6 percentage point increase in the T-bill rate in 1994 and in the following year the inflation rate actually declined by .2 percentage points. In a speech before the Mortgage Bankers Association in 1996, William McDonough, President of the Federal Reserve Bank of New York, concluded that, "Much of our success in containing inflation in recent years reflects monetary policy actions that preempted inflationary pressures before they actually showed up in general prices."
The case for a forward looking interest rate reaction function based on inflation indicators, rather than an actual increase in the CPI inflation rate, is strengthened by those years in Table 12.1 when the Fed waited for an actual increase in the inflation rate or did not let the T-bill rate increase considerably faster than the inflation rate. In most of these cases the US economy experienced even worse inflation in the following year or went through a recessionary cure for inflation (as was the case following the large increases in the inflation rate in 1956, 1969 and 1979).
In (1989) Stephen McNees, a Vice President of the Federal Reserve Bank of Boston, noted that several events have combined to produce disillusionment with model-based approaches to forecasting the inflation rate. "The Phillips curve approach is widely regarded to have been discredited by the positive relationship between inflation and unemployment rates during the supply shocks in the 1970s. The monetarists approach, that inflation is a purely monetary phenomenon, was undermined by institutional changes permitting the payment of interest on demand deposits and the associated proliferation of 'near monies.' Both of these traditional approaches to modeling inflation were further discredited by the emergence of the rational expectations hypothesis, which emphasized the forward- looking nature of inflation expectations, providing further reason to doubt all strictly 'adaptive' models."
Disillusionment with conventional approaches to inflation forecasting has inspired a more inductive search for inflation indicators with properties similar to the components of the Conference Board's composite index of leading economic indicators. As of January 1994 there were several indicators pointing in the direction of greater inflation than was experienced during 1993.
When the goal is to predict an acceleration in the inflation rate before it occurs, considerable attention should be paid to first differences in the growth rate for hourly earnings. See Table 12.2. Since 1968 there have (so far) only been three years (1974, 1980 and 1994) that were followed by a decline in the CPI inflation rate after an increase in the hourly earning growth rate. The first two declines in the inflation rate were more nearly the result of economic recessions rather than a successful effort to prevent inflation from getting out of hand.
From 1989-92 the December-December growth rates for the hourly earnings of non supervisory workers in nonagricultural industries drifted downward from 4.1 percent to only 2.3 percent and helped to produce the lowest core inflation rate (excluding food and energy) in the US since the implementation of wage and price controls in 1971. In 1993, however, the growth of hourly earnings accelerated by .2 percentage points to 2.5 percent. A fear that this tightening of the labor market would cause business enterprises to raise prices at a faster rate was heightened by shortages of experienced truck drivers and skilled labor in some construction industries.
Concern about the possibility of a wage-price spiral was also amplified at the Fed by a new (1993) estimate of the natural unemployment rate by Stuart Weiner, an assistant vice president and economist at the Federal Reserve Bank of Kansas City. His estimates suggested that the unemployment rate which would keep inflation from accelerating or decelerating, in the absence of adverse developments such as crop failures and oil price shocks, was equal to 6.25 percent and could go even higher in future years depending on the impact of structural changes occurring in the labor market. This rather controversial assessment was made before the release by the Labor Department In January 1994 of new estimating procedures which some analysts feared would raise the reported unemployment rates for 1994 and beyond.
In 1994 the preliminary Dec-Dec. percentage change in average hourly earnings in private nonagricultural industries increased by .1 percentage point. In 1995 there was a more noteworthy increase amounting to .5 percentage points and in 1996 the preliminary increase was equal to .7 percentage points.
In a February 1997 presentation before the Senate Banking Committee Fed Chairman Greenspan said he sees signs that the widespread fear over job security is abating. That is significant because worker insecurity has been restraining wages at a time of low unemployment and allowing the Fed to delay an increase in short-term rates. He vowed the Fed won't wait for inflation to actually accelerate before boosting short term interest rates saying that "a pre-emptive policy tightening may become appropriate before any sign of actual higher inflation becomes evident".
Another inflation indicator of concern to some analysts at the beginning of 1994 was farm prices. Flooding in the nation's corn belt and drought in the Southeast caused the prices received by farmers for all crops to increase 13.7 percent during 1993. It should be noted, however, that rising grain prices have a history of not being a very reliable inflation indicator. See Table 12.3. The initial impact of higher grain prices will sometimes be offset by lower meat prices as farmers send cattle and hogs to market early to avoid the high cost of fattening them up.
Some members of the Board of Governors of the Federal Reserve have suggested, in public, that the price of Gold may be a good inflation indicator. Double digit increases in the December-December price of gold have often been followed by accelerating inflation if the economy was not bogged down in a recession or about to enter a recession as was the case in 1974, 1979 and 1980. See Table 12.4.
Economists at the New York Fed, however, have become more skeptical about the usefulness of commodity prices as inflation indicators. After testing eight commonly used indexes Bloomberg and Harris (1995) have concluded that "although commodities had some predictive power in the past, the commodity-consumer price connection has broken down in the more recent period." They argue that this shift primarily reflects the diminished role of traditional commodities in U.S. production and the "sterilization" of some inflation signals by offsetting monetary policy actions.
Wall Street, however, has continued to respond vigorously to commodity inflation. During the first five months of 1996, when drought in the winter wheat belt and low inventories of liquid hydrocarbons and grain stocks caused commodity prices to soar upward at a rapid rate, investors in fixed income securities allowed long term interest rates to increase by more than one percentage point.
Also of concern to some inflation watchers at the end of 1993 was the behavior of the capacity utilization rate in manufacturing. In the post 1947 period inflation has sometimes accelerated after the capacity utilization rate has risen above 82 percent (Kahn 1994 and Garner 1994). In November 1993 the capacity utilization rate crossed the lower end of the 82 percent threshold that Kahn, and other staff members associated with the Federal Reserve, considered indicative of inflationary pressure.
Fierce competition from foreign goods producers and the success of the Fed's efforts at preventing an acceleration in the inflation rate in 1994, at a time when the capacity utilization rate was continuing to increase, however, has reduced its credibility as an inflation indicator (Bleakley 1995). For an informal look at capacity utilization and its possible usefulness as a predictor of large changes in the CPI inflation see Table 12.5.
The case for higher short term interest rates at the end of 1993, however, was not entirely dependent upon the behavior of inflation indicators. In April 1994 Alan Murray of the Wall Street Journal noted that "The reason the Fed needs to raise interest rates now isn't to combat incipient inflation, but to stop stimulating an economy that needs no stimulus. It isn't time to slam on the brakes; but it is past time to take the foot off the accelerator."
At the end of 1993 the yield on CDs and money market funds was only slightly higher than the inflation rate. In a world where the after tax return on savings deposits is negative, it makes sense for risk averse consumers to invest heavily in new houses, automobiles and other types consumer durables rather than put their savings in a money market account and help to provide business enterprises with some of the additional capital that is needed to boost productivity and make the US economy more competitive and prosperous in the long run.
In his (1979) analysis of annual and quarterly forecasts Zarnowitz concluded that: "Forecasts of inflation are not much better than projections of the most recently observe inflation rates, and they lag behind the actual rates much like such projections." Since this conclusion was published there have only been three years when the President's Council of Economic Advisers' 4th quarter to 4th quarter forecast of the implicit price deflator for real GNP or GDP was able to outperform a simple no change in the preceding year's preliminary growth rate prediction for the IPD. See Table 12.6.
Forecasting what will happen to the inflation rate is particularly tricky since history would suggest that preemptive strikes by the Fed may make a difference. Food and energy price shocks are very difficult to predict and there have also been some changes in the construction of the CPI index that may make a difference.
Food, energy, shelter and medical care have been the most volatile components of the CPI in recent decades. The good news with regard to food inflation is that the weights given to this item in the CPI have been significantly reduced over time. In December 1972, before the world wide drought and food shortage of 1973-74, the relative importance of food was 22.4 percent. In January 1987, when the CPI was revised to reflect expenditure patterns for 1982-84, the weight given to the food component was reduced to nearly 16 percent. Since a higher proportion of the remaining 16 percent is now spent on packaging and advanced food processing a large increase in the price of basic food stuffs will not raise the all item CPI as much as was formerly the case.
The bad news with regard to food inflation is that we seem to have returned to a less benign era where large reductions in crop yield are occurring more frequently. From 1910-36 there were six declines in crop yields per acre amounting to ten percent or more. In the next 37 year period from 1936-73 there were no declines of this magnitude. From 1973-93 there were five years when total crop production in the United States declined by more than nine percent.
In 1983, 1988 and 1993, however, the US experienced greater crop losses than those sustained in 1974 and 1980 but was able to avoid significant food inflation as a result of a large build-up in grain stocks in this country and around the world. That surplus had largely disappeared at the beginning of the 1996 crop year.
December-December increases in the food component of the CPI which have been one percentage point or more in excess of the inflation rate for the all item CPI, have the distinction of usually being followed by increases in the inflation rate. Food inflation of this magnitude has almost always been bad news for investors in long term government bonds. The only exception so far, as far as bond holders are concerned, followed 1971 when the prices of most other goods and services were still subject to wage and price controls. See Table 12.7.
The good news with regard to the wage-price spiral is that cost of living escalator clauses have gone out of fashion making wage inflation less responsive to food and energy price shocks. In 1990, for the first time since 1960, the growth of average hourly earnings slowed during a year containing a peak in business activity. High unemployment in Europe and most developing countries and the globalization of manufacturing have made many industrial unions more concerned about job security than a possible resumption of wage-price inflation.
While consumers are now spending a higher fraction of their income on housing some much needed stability has been imparted to the home ownership component of the CPI cost of shelter index as a result of a 1983 shift to a rental equivalent measure of home owner costs. The old procedure was to measure the price of new houses, mortgage interest rates and other cost elements that are borne by those relatively few families that are fortunate enough to be able to afford a new home during the month in question. This made the cost of home ownership extremely sensitive to changes in monetary policy. In 1970, 1975, 1978, 1979, 1980, 1981 and 1982 the official CPI increased from one to 2.5 percentage points more rapidly than an experimental rental equivalent CPI. Tight money, instead of being a cure for inflation, probably made the official inflation rate worse in some of these years.
The bad news with regard to the inflation rate for shelter is that it is more sticky and not likely to come down as much in response to short lived recessions as other prices. This component of the consumer price index, moreover, is likely to continue to increase more rapidly, on the average, than the CPI as a whole.
Medical care services are another inflation prone component of the CPI that ought, in principle, to be more controllable in the 1990s with many older persons quitting smoking, drinking more moderately, getting more exercise and about to shrink in numbers as a result of the baby bust which occurred during World War II and the great depression of the 1930s. The AIDS epidemic and many new and more costly medical procedures, however, are likely to make health care cost containment one of our most enduring and frustrating problems.
Since food and energy price shocks are less likely to be amplified now by cost of living adjustment clauses in labor contracts and by changes in mortgage interest rates, however, it is reasonable to hope that future changes in the CPI inflation rate will be smaller, on the average, than was the case from 1945-91.
Economists have long known that the all item CPI overstates the true increase in the cost of living. The fixed weight character of the index doesn't allow for improvements in welfare as consumers shift from expensive to cheaper goods in response to changes in relative prices. The Bureau of Labor Statistics, moreover, does not fully adjust the index for improvements in the quality of goods and services. Adjustments for such bias would not only help to balance the Federal Budget, by reducing cost of living increases for retired workers, but would make it easier for the Federal Reserve to achieve its goal of stable prices without raising interest rates to the point of tipping the US economy into so many recessions.
Since 1972 the growth of hourly earnings for production workers on private non-agricultural payrolls has been about equal, on the average, to the chain-type price index used to deflate gross domestic product which does make an effort to adjust the growth of GDP for changes in relative prices and some improvements in the quality of goods and services. See Table 12.8. This would suggest that workers might not be disadvantaged by having the Labor Department incorporate similar adjustments in the more widely publicized CPI.
Since the publication of Muth's (1961) article on rational expectations a considerable amount of effort has gone into trying to determine whether professional forecasters make the best use of available information. Where AVt is the actual value of the variable to be forecasted, FVt is the forecasted value, Xt is any information available to the forecaster at the time the forecast was made and et is an error term, the usual test for rationality is:
AVt = a + bFVt + cXt + et (1)
For a set of forecasts to be considered rational the "a" coefficient should not be significantly different from zero; the "b" coefficient should not be significantly different from one; and the "c" coefficient should not be significantly different from zero. There are lots of problems with this test. If there is an unanticipated change in the inflation environment, for example, variables that were not judged to very useful inflation indicators at the time a forecast was made might turnout to be highly significant in retrospect based on tests that benefit from additional information (that was not available at the time of the forecast) and 20-20 hindsight.
Another problem that is often ignored in trying to determine whether forecasts are rational is that the evaluators usually don't know what kind of model was used to make the forecast. Without detailed knowledge of what went into the forecast it is hard know whether the forecaster failed to make the best use of other information.
It can be a fun type of exercise, however, to see how well professional forecasts pass the "abc" tests associated with equation (1). Using a standard regression package containing autoregressive parameters one can quickly show that the inflation forecasts of the President's Council of Economic Advisors (CEA) have been subject to a "c" type bias. Since the Volcker revolution of 1979 almost all of the CEA's inflation forecasts have been too high. See columns (3) and (4) of Table 12.6. For the 1978-95 period as a whole the CEA could have obtained better forecasts, in most years, by simply assuming that there would be no change in the inflation rate.
A similar bias in the direction of being overly fearful of inflation is also apparent in the consensus forecasts of the year-year inflation rates for the GNP or GDP deflator compiled by the Blue Chip Economic Indicators newsletter as of early January for the year in question. See Table 12.9. Some analysts, such as Ben Martin Bly of the Harvard Psychology Department, have suggested that this type of bias may be rational if the clients of professional forecasters are more forgiving with regard to an over estimate of the inflation rate than an under estimate.
This type of forecasting bias is a little disturbing, however, since it might encourage the Fed to be overly aggressive in its war on inflation. If professional forecasters cannot be trusted to examine their own record and adjust their inflation forecasts for upward bias, is it reasonable to suppose that investors and Board of Governors of the Federal Reserve will do that for them? Support for the notion that investors in fixed income securities may have become overly fearful of accelerating inflation is most apparent when one examines annual changes in long term bond yields and the CPI inflation rate.
An extension of Irving Fisher's theory of interest identifies three determinants of long-term nominal bond yields: long-term real interest rates, risk premia, and long-term inflationary expectations. An econometric study of ten year Treasury bonds by Ireland (1996) suggests that the long- term real rate is quite stable and that the risk premium is small. If these findings are correct movements in long-term bond yields should provide a reliable measure of changes in expected inflation. An unanswered question is whether the inflationary expectations that can be inferred from changes in long term bond yields are "rational".
In Table 12.10. Dec.-Dec. changes in the yield on ten year Treasury bonds are rank ordered in terms of their magnitude from 1959-95 so that one can more easily compare them with both the current and following year changes in the Dec.-Dec. inflation rate for the all item consumer price index. There does seem to be a positive relationship between these two variables in the current year but not in the following year.
The regression coefficients associated with the first equation in Table 12.11 are not inconsistent with the conclusion that changes in the CPI inflation rate in the current year have been about equal, on the average, to the change in the yield on ten year Treasury bonds. The adjusted R-squared can be increased in a statistically significant way, however, by also including the Nov.-Nov. percentage change in payroll employment in the regression to help "explain" current year changes in the CPI inflation rate. See equation (2).
When the objective is to explain what happens to the CPI inflation rate in the following year, however, one observes a negative coefficient for the preceding change in the Treasury bond yield which becomes statistically significant for the 1959-94 period as a whole when the preceding growth rate for payroll employment is also included in the regression. See equations (3) and (4) in Table 12.11.
During his period the increases in long term interest rates fail all three parts of the "abc" test for rational following year forecasts. The implication is that the inflation forecasts to be derived from an expectational interpretation of changes in long term bond yields are more adaptive (backward looking) than forward looking, at a one year horizon.
The negative coefficients for the preceding change in the CPI inflation rate are not inconsistent with the hypothesis that the bond market may have a tendency to over react to the threat of inflation.
An "irrational" bond market, however, does not necessarily mean that changes in long term interest rates are totally useless from a forecasting point of view. At the end of 1995 equation (3) was predicting a .4 percentage point increase in the all item CPI inflation rate for 1996 and equation (4) an increase of .8 percentage points which turned out to be correct.
There is not much doubt that bond traders and speculators do pay close attention to commodity prices and the behavior of payroll employment. When the February, May and June 1996 increases in employees on nonagricultural payrolls turned out to be far in excess of what was expected by professional forecasters both the bond and stock markets were clobbered.
It should be noted, however, that changes in the CPI inflation rate are not as closely linked to preceding changes in payroll employment and long term interest rates as was once the case. When equation (4) is refit to data for the more recent period from 1982-94 the coefficients for changes in T-bond yields and payroll employment are smaller (-.272 and .194 respectively) and are not very significant from a statistical point of view. The predicted increase in the CPI inflation rate for 1996 is also reduced from .8 percentage points to only .3 percentage points.
*An earlier version of this paper was published in The Journal of Fixed Income, June 1995, 61-70.
Bleakley, Fred (1995). "Capacity Utilization Is Losing Credibility," The Wall Street Journal, February 14, p. A2 & A10.
Bloomberg, S. Brock and Ethan Harris (1995). "The Commodity--Consumer Price Connection: Fact or Fable?" FRBNY Economic Policy Review, October, 21-38.
Crutsinger, Martin (1994). "Greenspan Accused of Hastiness," The Times Union, February 22, A1 & A7.
Garner, C. Alan (1994). "Capacity Utilization and U.S. Inflation," Federal Reserve Bank of Kansas City Economic Review, Fourth Quarter, 5-21.
Ireland, Peter (1996). Long-Term Interest Rates and Inflation: A Fisherian Approach" Federal Reserve Bank of Richmond, Economic Quarterly, 82 Winter) 21-35.
Kahn, George (1994). "Achieving Price Stability: A 1993 Report Card," Federal Reserve Bank of Kansas City Economic Review, First Quarter, 5-18.
McNees, Stephen (1989). "How Well Do Financial Markets Predict the Inflation Rate?" New England Economic Review, Sept./Oct., 31- 46.
-----, (1992). "The 1990-91 Recession in Historical Perspective," New England Economic Review, Jan./Feb., 3-22.
Murray, Alan (1994). "Fed Moved Too Slow On Increasing Rates," The Wall Street Journal, April 11, 1A.
Muth, J. (1961). "Rational Expectations and the Theory of Price Movements," Econometrica, 29(3), 315-35.
Renshaw, Edward (1992). The Practical Forecasters' Almanac(Business One Irwin).
-----, "The Fed's Preemptive Strike Against Accelerating Inflation," The Journal of Fixed Income, June 1995, 61-70.
Weiner, Stuart (1993). "New Estimates of the Natural Rate of Unemployment," Federal Reserve Bank of Kansas City Economic Review, 4th Quarter, 53-69.
Zarnowitz, Victor (1979). "An Analysis of Annual and Multiperiod Quarterly Forecasts of Aggregate Income, Output, and the Price Level," Journal of Business, January, 1-33.
Year Change in CPI First Differences CPI Inflation Rate
T-Bill Inflation Current Following
Yield Rate Year Year
1980 3.6 12.5 - .8 -3.6
1978 3.0 9.0 2.3 4.3
1979 3.0 13.3 4.3 - .8
1994* 2.6 2.7 .0 - .2
1973 2.3 8.7 5.3 3.6
1988 2.3 4.4 .0 .2
1959* 1.8 1.7 - .1 - .3
1969 1.8 6.2 1.5 - .6
1977 1.7 6.7 1.8 2.3
1955 1.4 .4 1.1 2.6
1972 1.1 3.4 .1 5.3
1983* 1.0 3.8 .0 .1
1968 .9 4.7 1.7 1.5
1956 .6 3.0 2.6 - .1
1963 .6 1.6 .3 - .6
1966 .6 3.5 1.6 - .5
1965 .5 1.9 .9 1.6
*Identifies successful strikes against the possibility of accelerating inflation which occurred fairly early in a business expansion.
Year Difference CPI First Differences CPI Inflation Rate
in Wage Inflation Current Following
Inflation Rate Year Year
1968 2.7 4.7 1.7 1.5
1974 1.5 12.3 3.6 -5.4
1978 1.3 9.0 2.3 4.3
1987 1.3 4.4 3.3 .0
1973 1.2 8.7 5.3 3.6
1977 .8 6.7 1.8 2.3
1988 .8 4.4 .0 .2
1980 .7 12.5 - .8 -3.6
1996----- .7 3.3 .7 ?
1995----- .5 2.5 - .2 .8
1976 .4 4.9 -2.0 1.8
1989 .4 4.6 .2 1.5
1971 .3 3.3 -2.3 .1
1993----- .2 2.7 - .2 .0
1994----- .1 2.7 .0 - .2
1972 - .1 3.4 .1 5.3
Year Crop CPI First Differences CPI Inflation Rate
Inflation Inflation Current Following
Rate Rate Year Year
1973 52.0 8.7 5.3 3.6
1988 21.4 4.4 .0 .2
1980 19.8 12.5 - .8T -3.6
1983 19.3 3.8 .0 .1
1972 16.5 3.4 .1 5.3
1987 15.3 4.4 3.3 .0
1993---- 13.7 2.7 - .2 .0
1995---- 11.3 2.5 - .2 ?
1978 10.9 9.0 2.3 4.3
1971 10.1 3.3 -2.3 .1
Year % Change CPI First Differences CPI Inflation Rate
Price Inflation Current Following
Gold Rate Year Year
1979 122.3 13.3 4.3 - .8
1974 70.9 12.3 3.6 -5.4
1973 66.8 8.7 5.3 3.6
1978 29.2 9.0 2.3 4.3
1980 28.6 12.5 - .8 -3.6
1987 24.2 4.4 3.3 .0
1986 21.6 1.1 -2.7 3.3
1977 19.9 6.7 1.8 2.3
1993----- 14.8 2.7 - .2 .0
Year Capacity CPI First Differences CPI Inflation Rate
Utilization Inflation Current Following
Rate Rate Year Year
1966 91.1 3.5 1.6 - .5
1965 89.5 1.9 .9 1.6
1953 89.3 .7 - .1 -1.4
1973 88.1 8.7 5.3 3.6
1967 87.2 3.0 - .5 1.7
1968 87.2 4.7 1.7 1.5
1955 87.0 .4 1.1 2.6
1969 86.8 6.2 1.5 - .6
1956 86.1 3.0 2.6 - .1
1951 85.8 6.0 .1 -5.2
1964 85.6 1.0 - .6 .9
1952 85.4 .8 -5.2 - .1
1979 85.4 13.3 4.3Oil - .8
1978 85.1 9.0 2.3Food 4.3
1988 83.6 4.4 .0 .2
1989 83.2 4.6 .2Oil 1.5
1974 83.8 12.3 3.6Oil -5.4
1972 83.7 3.4 .1 5.3
1957 83.6 2.9 - .1 -1.1
1963 83.5 1.6 .3 - .6
1994----- 83.3 2.7 .0 - .2
1996----- 83.2 3.3 .8 ?
1995----- 83.0 2.5 - .2 .8
1950 82.8 5.9 8.0 .1
1977 82.8 6.7 1.8 2.3
1990 81.3 6.1 1.5Oil -3.0
1959 81.6 1.7 - .1 - .3
1962 81.4 1.3 .6 .3
1987 81.6 4.4 3.3Oil .0
1993 80.6 2.7 - .2 .0
1984 80.4 3.9 .1 - .1
Source of basic data: Economic Report of the President.
Year Forecasted Preliminary Forecasted Preliminary
Growth Rate Growth Rate Growth Rate Growth Rate
for Real GNP for Real GNP for the Implicit for the Implicit
or GDP or GDP Price Deflator Price Deflator
(1) (2)n (3) (4)n
1978 4.8(- .5)B 4.3(-1.4) 6.0( 2.3)B 8.3( 2.4)
1979 2.2(-1.4)B .8(-3.5) 7.4( 1.6) 9.0( .7)
1980 -1.0( .7)B - .3(-1.1) 9.0( 1.0) 10.0( 1.0)
1981 1.8(-1.1) .7( 1.0) 10.2(-1.6) 8.6(-1.4)
1982 3.0(-4.2) -1.2(-1.9) 7.2(-2.6)B 4.6(-4.0)
1983 3.1( 3.0)B 6.1( 7.3) 5.6(-1.5) 4.1(- .5)
1984 4.5( 1.1) 5.6(- .5) 5.0(-1.5) 3.5(- .6)
1985 4.0(-1.5)B 2.5(-3.1) 4.3(-1.1) 3.2(- .3)
1986 4.0(-1.8) 2.2(- .3) 3.8(-1.6) 2.2(-1.0)
1987 3.2( .6)B 3.8( 1.6) 3.6(- .3)B 3.3( 1.1)
1988 2.4( .2)B 2.6(-1.2) 3.9( .0)B 3.9( .6)
1989 3.5(-1.1) 2.4(- .2) 3.7( .1) 3.8(- .1)
1990 2.6(-2.3) .3(-2.1) 4.2(- .2) 4.0( .2)
1991 .9(- .7) .2(- .1) 4.3(-1.3) 3.0(-1.0)
1992 1.9( 1.0)B 2.9( 2.7) 3.2(- .8) 2.4(- .6)
1993 2.9(- .1) 2.8(- .1) 2.6(- .4) 2.2(- .2)
1994 3.0( 1.0)B 4.0( 1.0) 2.7(- .4) 2.3( .1)
1995 2.4(- .9)B 1.5(-2.5) 2.9(- .4) 2.5( .2)
1996 2.2( 1.2)B 3.4( 1.9) 2.8(-1.0) 1.8(- .7)
1997 2.0 2.5
The forecasted growth rates for real GNP or GDP and its implicit price deflator are from the Economic Report of the President. The figures in parentheses are the error terms obtained by subtracting the forecasted growth rate from the actual preliminary growth rate in columns (2) and (4) from the next Economic Report. In those cases where a forecast range was presented, a mid point is used to evaluate the CEA's forecast.
(2)n and (4)n. The actual preliminary growth rates in the following year's Economic Report of the President. The figures in parentheses are the first differences in these preliminary growth rates. The evaluations for 1995 and the forecasts for 1996 are based on chain-type indexes.
B identifies years when the Council of Economic Adviser's forecasting error turned out to be smaller absolutely than the "no change in the growth rate forecasts" implied by the first differences in the preliminary growth rates.
Year Relative ----------Following Year---------------
Food First Difference Long Term Government
Inflation CPI Inflation Bond Return Minus
Rate Return on T-Bills
1973 11.6 3.6 - 3.6
1950 3.9 .1 - 5.4
1978 2.8 4.3 -11.6
1986 2.7 3.3 - 8.2
1960 1.7 - .7 - 1.1
1965 1.6 1.6 - 1.2
1977 1.4 2.3 - 8.4
1972 1.2 5.3 - 8.0
1951 1.1 -5.2 - .5
1971 1.0 .1 1.9
1989 1.0 1.5 - 9.6
1996---- 1.0 ? ?
Source of bond return data: The financial returns from 1951-87 are from Roger Ibbotson and Rex Sinquefield, Stocks, Bonds, Bills, and Inflation (Charlottesville, Virginia: The Research Foundation of the Institute of Chartered Financial Analysts, 1989). Since 1989 the financial returns for government securities are obtained by computing the price appreciation associated with end of the year asked and bid prices for stripped coupons maturing in February which are reported in the Wall Street Journal.
Chained
Hourly Deflator Actual Minus the
Year Earnings Real GDP Predicted Inflation Rate
(1)n (2) (3)n (4)n
1973 6.2P 5.6 - .6* 1.4
1974 8.0 8.9 .9* 3.3
1975 8.4 9.4 1.0* .5**
1976 7.2 5.8 -1.4 -3.6
1977 7.6 6.4 -1.2 .6**
1978 8.2 7.3 - .9 .9
1979 7.9 8.5 .6* 1.2
1980 9.0P 9.3 .3* .8
1981 9.1P 9.4 .3* .1**
1982 6.9 6.3 - .6 -3.1
1983 4.6 4.2 - .4 -2.1
1984 3.2 3.8 .6 - .4**
1985 3.1 3.4 .3 - .4
1986 2.5 2.6 .1 - .8
1987 2.4 3.1 .7* .5
1988 3.2 3.7 .5 .6
1989 4.1 4.2 .1 .5
1990 3.6P 4.4 .8* .2**
1991 3.1 3.9 .8 - .5**
1992 2.4 2.8 .4 -1.1
1993 2.5 2.6 .1 - .2
1994 2.7 2.3 - .4 - .3**
1995 2.9 2.5 - .4 .2**
1996 3.3 2.1 -1.2 - .4**
(1)n. The earnings of production workers have been adjusted for overtime in manufacturing and for interindustry shifts for the years 1954-88 and are unadjusted earnings thereafter.
(3)n. The predicted inflation rate is equal to the growth of hourly earnings in column (1) minus a constant term equal to 1.3 percentage points for the years 1953-72 and zero for the years since 1972.
(4)n. The predicted inflation rate is equal to the growth rate for the implicit price deflator in the preceding year.
*Indicates years when the December-to-December percentage change in the energy component of the CPI was 2.5 percentage points or more in excess of the comparable percentage change in the all item CPI.
**Identifies years when the naive forecasting error in column (4) is smaller absolutely than the forecasting error in column (3).
P equals a year containing an official NBER peak in business activity.
Source of basic data: Economic Report of the President.
Year Blue Chip Consensus Forecasts 4th Quarter Actual
January February Projection Inflation
Survey Survey Rate
(1) (2) (3)n (4)n
1977 5.3( .2)B 5.4( .1)B 4.7( .8) 5.5( .3)
1978 6.1( 1.3)B 6.0( 1.4)B 5.8( 1.6) 7.4( 1.9)
1979 7.8( 1.1) 7.9( 1.0) 8.2( .7) 8.9( 1.5)
1980 9.1(- .1) 9.2(- .2) 8.9( .1) 9.0( .1)
1981 9.6(- .5)B 9.8(- .7) 9.8(- .7) 9.1( .1)
1982 7.7(-1.7)B 7.6(-1.6)B 8.6(-2.6) 6.0(-3.1)
1983 5.1(- .9) 4.9(- .7) 4.5(- .3) 4.2(-1.8)
1984 4.7(- .9) 4.5(- .7) 4.0(- .2) 3.8(- .4)
1985 4.1(- .8) 3.7(- .4) 3.5(- .2) 3.3(- .5)
1986 3.5(- .8) 3.4(- .7) 3.1(- .4) 2.7(- .6)
1987 3.2(- .2)B 3.0( .0)B 2.2( .8) 3.0( .3)
1988 3.6(- .2) 3.4( .0)B 3.3( .1) 3.4( .4)
1989 4.3(- .2) 4.5(- .4) 4.1( .0) 4.1( .7)
1990 4.0( .1)B 4.0( .1)B 3.8( .3) 4.1( .0)
1991 4.2(- .6) 3.7(- .1)B 4.0(- .4) 3.6(- .5)
1992 3.0(- .4) 2.7(- .1)B 3.0(- .4) 2.6(-1.0)
1993 2.7(- .2) 2.5( .0)B 2.4( .1) 2.5(- .1)
1994 2.5(- .4) 2.4(- .3) 2.2(- .1) 2.1(- .4)
1995 2.9(- .5) 2.7(- .3) 2.3( .1) 2.4( .3)
---------Mean Absolute Forecasting Errors, 1977-95-------
.58 .46 .52 .74
The figures in parentheses are the actual minus predicted forecasting error terms expressed in percentage points. The January and February surveys are conducted very early in the month.
(3)n. The preliminary fourth quarter to fourth quarter growth rate for the implicit price deflator for the preceding year which is published in the January issue of the Survey of Current Business.
(4)n. The preliminary year-to-year growth rate for the implicit price deflator which is published in the Survey of Current Business in January of the following year. The figures in parentheses are first differences in these preliminary inflation rates.
B represents years when the forecasting error for the Blue Chip Consensus was smaller than the forecasting error for the fourth quarter projection in column (3).
Year Yield Change CPI First Differences CPI Inflation Rate
Ten Year Inflation Current Following
T-Bond Rate Year Year
1980 2.45 12.5 - .8 -3.6
1994---- 2.04 2.7 .0 - .2
1987 1.88 4.4 3.3 .0
1969 1.62 6.2 1.5 - .6
1979 1.38 13.3 4.3 - .8
1978 1.32 9.0 2.3 4.3
1983 1.29 3.8 .0 .1
1981 .88 8.9 -3.6 -5.1
1967 .86 3.0 - .5 1.7
1959 .83 1.7 - .1 - .3
1977 .82 6.7 1.8 2.3
1974 .69 12.3 3.6 -5.4
1996----- .59 3.3 .8 ?
1975 .57 6.9 -5.4 -2.0
1965 .44 1.9 .9 1.6
1972 .43 3.4 .1 5.3
1973 .38 8.7 5.3 3.6
1968 .33 4.7 1.7 1.5
1963 .27 1.6 .3 - .6
1990 .24 6.1 1.5 -3.0
1966 .22 3.5 1.6 - .5
1961 .22 .7 - .7 .6
1988 .12 4.4 .0 .2
1964 .05 1.0 - .6 .9
1962 - .20 1.3 .6 .3
1992 - .32 2.9 - .2 - .2
1984 - .33 3.9 .1 - .1
1971 - .46 3.3 -2.3 .1
1960 - .85 1.4 - .3 - .7
1991 - .99 3.1 -3.0 - .2
1993---- -1.00 2.7 - .2 .0
1976 -1.13 4.9 -2.0 1.8
1970 -1.26 5.6 - .6 -2.3
1989 -1.27 4.6 .2 1.5
1995---- -2.10 2.5 - .2 .8
1986 -2.15 1.1 -2.7 3.3
1985 -2.24 3.8 - .1 -2.7
1982 -3.18 3.8 -5.1 .0
Variables ---------Regression Coefficients--------
--
Dependent Variable: CPI Inflation Current Year Change Following Year
Change
Equation Number (1) (2) (3) (4)
Independent Variables
Constant term. -.074 -1.327 .041 -1.878
(-.213) (-2.621) (.105) (-3.854)
Dec.-Dec. change in the yield on .854 .538 -.171 -.655
ten year Treasury bonds. (2.989) (1.965) (-.535) (-2.485)
Nov.-Nov. percentage change in .568 .870
total payroll employment. (3.133)
(4.986)
Summary Statistics
Adjusted R-squared .18 .35 -.02 .40
Standard Error of the Regression 2.08 1.86 2.33 1.79
Durbin Watson Statistic 1.81 1.99 1.65 2.23
The parentheses contain t-statistics for the hypothesis that the coefficient's true value is zero.
Go on to Essay 13:
Return to the Introduction