However, we believe that such a fast gain has little relevance when the following factors are considered:
1) tax cuts provided a temporary boost to the economy;
2) real median household income (the broadest measure of the standard of living) declined in 2002 (Chart 1), and barring an astounding resurgence in employment over the next three months, 2003 will witness the fourth consecutive annual drop in income;
3) payroll employment and manufacturing production will register a post-World War II record of three consecutive yearly decreases in 2003.
In short, never has such weakness in the standard of living, production and labor markets occurred almost two years into a recovery, as it has been defined by the National Bureau of Economic Research.
U.S. recessions are seminal events in many ways--economically, politically and globally. From an economic viewpoint two critical sectors--inflation and interest rates--are most important. Historically, a recession causes a downward trend in interest rates and inflation, and often these trends continue several years after the recession has ended. The recession-like nature of this recovery suggests that the ultimate low point in the interest rate and inflation cycle could still be years away. The large U.S. output gap, the high degree of unutilized plant and labor market capacity here and abroad, and the abnormal pattern of the recovery suggest that a high risk of deflation remains.
and interest rates--are most important. Historically, a recession causes a downward trend in interest rates and inflation, and often these trends continue several years after the recession has ended. The recession-like nature of this recovery suggests that the ultimate low point in the interest rate and inflation cycle could still be years away. The large U.S. output gap, the high degree of unutilized plant and labor market capacity here and abroad, and the abnormal pattern of the recovery suggest that a high risk of deflation remains.
The Economy's Deflationary Gap
In a deflationary gap, real GDP falls short of the potential level of real GDP, which is the level required to produce full employment. In other words, when deflationary gaps exist, the economy has idle plant and labor resources that put downward pressure on the inflation rate and, in turn, on interest rates. The opposite occurs in an inflationary gap, when real GDP exceeds the potential level of GDP. Thus, during inflationary gaps, productive resources are pushed beyond normal limits. Bottlenecks, order backlogs and slow delivery times are standard symptoms. Chart 2 shows inflationary gaps above and deflationary gaps below the zero line. The output gap has been expressed as a percent of GDP to adjust for the increased level of GDP over time.
The pattern of the output gap in the current recovery is diverging dramatically from the post-war norm. In seven of the eight previous recessions since the early 1950s (when interest rate controls ended), the low, or bottom, in the deflationary gap occurred in the final quarter of the recessions. In the 1990-91 recession, the trough in the output gap was two quarters after the official end of the recession. In the 2001 recession the deflationary gap has continued to expand over the first six quarters of the current recovery, a totally unprecedented development. Because the cyclical low in the inflation and interest rate cycle typically occurs after the low in the output gap, the continued expansion in the gap also points to the prospect that the low in inflation and interest rates lies well in the future.
Closer inspection shows that in the past four decades, the lows in the 30-year yield occurred 12.8 quarters after the trough in the output gap, while the lows in inflation were 13 quarters later (Chart 3). Going back one more decade, the lags are similar, but the Korean and Vietnam Wars, a record-setting steel strike, and a learning curve adjustment to freely floating markets distorted the broader sample. But the fact remains that since 1950, the low point in the 30-year Treasury bond yield and inflation, on average, occurred a little over two years after the trough in the output gap.
The current interest rate and inflation environment is dominated by the unprecedented lack of a reduction in the deflationary gap for the first six quarters of the current recovery. If the output gap were to start a sustained recovery from this point, a dubious assumption, the interest rate and inflation lows are three years away if the averages of the past four decades are applicable. But, other important considerations suggest that lag time for reaching the inflation and interest rate lows may be even longer.
An Alternative Deflationary Gap Measure
The output gap is heavily dependent on the measurement of potential GDP, a task of the Congressional Budget Office. Calculating potential GDP hinges critically on the issue of what constitutes full employment. If full employment is miscalculated, then so is the output gap. Also, if real GDP is incorrect, then so is the output gap. Our concern is that capital spending estimates are overstated, thus leaving real GDP overstated and the deflationary gap understated. To shed light on the output gap from a different angle, another estimate can be used by taking the sum of the unemployment rate and the amount of unutilized plant capacity (Chart 4). This alternative measure of idle resources suggests that the output gap may be even larger than reported.
The idle resources measure indicates 31.9% of our plant and labor resources were dormant in the latest quarter, the highest in 20 years. The unutilized resources rate peaked coincidentally with the seven last recessions prior to 1990, and with a four-quarter lag after the end of the 1990 recession.
The idle resources measure may have a more revealing picture than the output gap. The inflationary output gap rose after 1997, reaching a peak of 2.9% in the fourth quarter of 1999 (Chart 2). In fact, in that same quarter of 1999 the idle resource rate was 2.3% higher than in the first quarter of 1997 when the inflationary gap first began to increase (Chart 4). The idle resources measure has the distinct advantage of explaining the pattern of inflation in the post 1997 period. The inflation rate, as measured by the core personal consumption expenditures (PCE) deflator, moved slightly lower from 1997 to late 1999. The rising inflationary output gap in the late 1990s incorrectly indicated that core inflation should have risen, contrary to the idle resources measure.
Currently, the 2% deflationary gap indicated for the latest quarter appears to be understated, just as the inflationary gap of the late 1990s was overstated. The sum of the idle labor and plant capacity stood at 28% in the latest quarter, the highest since the early 1980s, while the 2% deflationary gap is not nearly as deep as the 3% gap in the early 1990s. Once again, the idle resources rate is far more consistent with the inflation experience. In the last 12 months, the core PCE deflator rose just 1.2%, the slowest rise in 40 years.
In summary, the U.S. economy has substantial idle resources, regardless of the method of measurement. But, the degree of the problem may be greater than is generally understood. In such an environment, the risk of rising core inflation is all but zero. A move to at least a mild deflation over the next several years is a distinct possibility, even before we take into consideration the excess capacity in foreign markets for the U.S. is not alone in its overcapacity.
Unrelenting Global Slack
Adding to the oversupply problem, excess slack exists in foreign markets. When the same idle resources measure is constructed for the Euro 12, excess plant and labor resources are at 28.3% in the latest quarter, the highest level since the last quarter of 1997, and only slightly below the comparable U.S. figure.
A global measure of idle resources can be estimated by weighting the U.S. and European measures equally. In the latest quarter, 30% of the globally measured resources were idle, the highest record for a series that starts in 1985 (Chart 5), which is the initial point for the European data. The previous high of 29.5% was registered in the early 1990s.
Measuring the amount of idle resources outside of the U.S. and Europe is more difficult. Japan does not have a broad plant use rate. While quantitative values of slack are unknown, some anecdotal information indicates that there are broad unused labor and plant resources in Japan. In China, 87% of manufactured product categories are in oversupply. India, other parts of Asia and many areas of Latin America also have substantial unemployment and underemployment. Accordingly, wages are very low. In China, for example, much high quality labor is available for $3.50 (equivalent) for a ten-hour day. The McKinsey Global Institute estimates that $60/hour software jobs in the U.S. can be accomplished in India for $6/hour. Thus, both domestic and global factors should continue to press inflation downward.
The recent weakness in the dollar does not necessarily mean that inflation and interest rates are going to increase. The historical record of deflationary gaps, the dollar and interest rates indicates that the deflationary gap is the controlling factor and that interest rates and inflation will fall in spite of weakness in the dollar. Five deflationary gaps have occurred since the dollar began trading in the foreign currency markets. A falling dollar during deflationary gaps has been the predominant pattern. In short, the output gap is the dog and the dollar is the tail.
The dollar fell during four of these five deflationary gaps, and interest rates and inflation declined during all five. Only during the output gap of 1980 to 1986 did the dollar rise. During the deflationary gaps of 1970-72 and 1974-78 the dollar fell by 9.9% and 8%, respectively. These declines were in line with the 7.9% drop in the dollar in the current deflationary gap. This suggests that the dollar may continue to fall, but that the downward trend in interest rates and inflation can be maintained. In the extremely steep fall of the dollar from 1970 to 1972, the Federal funds rate dropped by 400 basis points and the bond yield fell 110 basis points.
Budget Deficits, Interest Rates, and Inflation
Budget deficits tend to spread the fear of impending inflation. Without question, the U.S. budget outlook is dire. The U.S. government could reach a $600 billion budget deficit in 2004, up from $400 billion in 2003. Economic theory would say that in isolation the increased budget deficit would be associated with higher inflation and interest rates. However, we do not live in a textbook, and large budget deficits can occur while the private sector of the economy languishes.
The offset to the deficit situation is that private sector debt deflation still lies ahead. In the latest quarter, U.S. private sector debt surged to another new peak of $1.65 for every $1 of GDP, while total U.S. debt reached $3.00, the highest since 1932. The early 1990s illustrated that private debt deflation can easily offset federal debt inflation. From the end of 1990 to the end of 1994, federal debt, as a percent of GDP, rose by 5.7%, but private sector debt fell by 11.2%. During that time both inflation and interest rates moved considerably lower. The circumstances are similar today, and the outcome is likely to be similar as well.
In the second quarter, outlays at the federal, state and local sector accounted for 19.4% of GDP. The remaining 80% is the private sector. Once the private sector debt retrenchment begins, funds will be diverted from the private spending stream. This, in turn, will increase the deflationary output gap while pushing unemployment higher and factory use rates lower. As such, the weakness in the dominant private sector will swamp any pressures generated by the federal budget deficit.
Bleak Employment Prospects
The employment situation is also too weak to contribute to any inflationary pressures. In the last quarter of the previous expansion non-farm payrolls recorded 132.3 million jobs. At that time (the last time the output gap was zero), these jobs represented 62% of the working age population. In the quarter just ended, payroll employment was 2.5 million lower, dropping the payroll to working age population ratio to 59%. This employment gap, which is the number of payroll jobs needed to bring the ratio back to 62%, is likely to persist for a number of months.
To bring the payroll to working age population ratio back to 62%, payroll employment would need to rise to a level of 137.3 million. The difference between the 137.3 million and the actual level of payroll jobs of 129.8 million results in a current employment gap of 7.5 million.
Since the 4th quarter of 2000, the working age population has grown by an average of 244,000 per month. Assuming this trend will continue, it would take 75 months, or 6 1⁄4 years, to eliminate this employment gap if payroll employment rises by an average of 344,000 per month--an unlikely prospect. If payroll employment rises 294,000 a month, 50,000 more than the growth in working age population, 150 months, or more than 12 years, will be needed to eliminate the employment gap.
Since payroll employment is not likely to increase by 244,000 at any time in the foreseeable future, the current employment gap of 7.5 million is likely to continue rising in 2004. This pool of unemployed workers will serve to reinforce downward pressure on the rate of increase in wages and cause the risk of deflation to persist.
Stay with Long Treasury Bonds
The deflationary gap suggests that capital spending will remain weak for years. Although consumer spending will get a shot in the arm from larger than normal income tax refunds in the first quarter of next year, this surge will be no more beneficial than the child-care tax rebates of July and August. Foreign inroads into the U.S. economy are likely to persist. Thus, any GDP bounce in the first quarter will be no more relevant than the one registered in the quarter just ended. The trend to lower interest rates and inflation remains intact. We expect patient, long-term investors to be handsomely rewarded by sticking with their investments in long Treasury coupons and zero coupon bonds.