There is much talk of a new economy, but what aspects of the macro economy are really different from what we have always known?

The U.S. economy is in the midst of an extraordinarily long economic expansion. The expansion is now in its 100^th month since the trough of the last recession. Meanwhile, rates of both unemployment and inflation are at their lowest levels in decades. The U.S. economy is frequently described as a "jobs machine" (20 million new jobs on payrolls since March of 1991), a "miracle economy", and a "new economy".

Apart from catchy phrases, what does all of this mean? Has some fundamental change occurred that alleviated previously binding constraints on economic growth? Or is the party about to end soon, destroying the job market for the MBA class of 2000?

The Case for the New Economy

What do we mean by a new economy? Obviously, many things about the economy are different relative to 10, 20, or 30 years ago. Most striking is perhaps the change in the industrial landscape. Many new industries have emerged in the United States, most prominently, information technology. But industry structure is always evolving, so this can’t be what people mean when they speak of the New Economy.

In the popular press, probably three themes emerge in writings about the New Economy: (1) Low unemployment and high growth can now coincide with low inflation; (2) Recessions, if not a thing of the past, will at least occur with lower frequency or less intensity than in the past; and (3) Productivity growth has reached a new higher level that can persist for some time into the future. These themes are not independent, but they do have their unique aspects, so let’s examine them one at a time.

What’s the historical relationship between growth and inflation and the evidence that the relationship has changed?

There have been two important waves of thinking about the relationship between GDP growth and inflation.

In 1958, A.W. Phillips published an article in the British journal Economica that presented an empirical relationship that would come to be known as the Phillips Curve (PC). It was a relationship between unemployment and the rate of change in money wages (i.e. wage inflation), which in modern incarnations is often presented in terms of unemployment and general product price inflation. The main idea was that low unemployment signaled tight labor markets and thus foreshadowed subsequent rises in inflation. Hence, the PC predicted an inverse relationship between unemployment and inflation. Figure 1 shows the U.S. data on unemployment and inflation from 1959 to 1970. It is probably fair to say that this empirical relationship became a cornerstone of Keynesian macroeconomic policies that were implemented in the United States in the 1960s.

By the late 1960’s, there was already an intellectual backlash against Keynesian economics in general and the Phillips Curve in particular. This was truly a triumph of theory over evidence, since as Figure 1 shows, the data of the 1960’s clearly supported Phillips’ original argument. According to Milton Friedman and Edmund Phelps, the PC fit the data in the 1950’s and 60’s only because inflation increases came as a surprise. Friedman and Phelps cautioned that once the Fed tried to manipulate the perceived tradeoff by using expansionary monetary policy to keep unemployment low, inflation would not only be high, but rising. In other words, persistent attempts at monetary or fiscal stimulation of the economy would lead to accelerating inflation. Data for the 1970’s and early 80’s, shown in Figure 2, strongly supported the Friedman-Phelps critique of the PC. Whatever relationship existed previously seemed to disintegrate.

The Rational Expectations school of thought refined more basic ideas of Friedman and Phelps to suggest that the Fed could manipulate unemployment only by making unanticipated adjustments to the money supply. Since the Rational Expectations school believed that policymakers cannot systematically fool the markets, they concluded that the Fed had no ability to influence output in the long run. In their view, economic growth and the rate of unemployment are largely determined by "real" or structural factors that depend on incentives and institutions, while inflation is simply the excess of money growth over real economic growth. Thus, any "relationship" between inflation and unemployment or growth is largely coincidental. Any rate of inflation could be consistent with full employment in the economy. It just depends on how rapidly the Fed let’s the money stock grow.

Many economists now believe that unemployment and inflation are unrelated in all but the very short run. If that view is correct, it was always possible for low unemployment and high growth to coincide with low inflation. It required that you have the supply conditions necessary for low unemployment or high growth. We seem to have them now. In addition to that, you just need a disciplined monetary policy—something Alan Greenspan has delivered over the last decade.

This analysis suggests that what’s potentially new about the economy is the low unemployment and higher (or at least steady) economic growth rates. There is little dispute that low inflation could have always been achieved with a disciplined monetary policy.

Are recessions less likely and less painful?

Much of the writing about the new economy seems to focus on the spectacular duration of the expansion phases of the last two business cycles. As noted earlier, the current expansion is in its 100^th month, making it the longest peacetime expansion in U.S. economic history. The only expansion that lasted longer was the 106-month boom of the 1960s, which was fueled in part by wartime spending. But even that record is now clearly within reach. The 1980s expansion lasted 92 months, making it the second-longest peacetime expansion on record. Taken together, these observations suggest that business cycle expansions are gaining longevity at a greater rate than the population.

The data base on U.S. Business Cycle Expansions and Contractions, maintained by the National Bureau of Economic Research, the official arbiter of business cycle turning points in the United States, enables us to reach a more informed judgement about potential changes in the nature of the cycle. The basic data are presented in Table 1. Since 1854, there have been 31 business cycles in the United States (a business cycle consists of an expansion phase when real output is rising and a contraction phase when real output is falling). For all 31 cycles since 1854, the average duration of a contraction is 18 months and of an expansion is 35 months. It appears that expansions have been steadily increasing in duration, while contractions have been decreasing. For the 16 cycles that occurred through World War I, expansions averaged 27 months and contractions 22. For the six cycles between World War I and the end of World War II (which included the Great Depression), expansions averaged 35 months and contractions 18 months.  For the nine cycles since World War II (not including the current expansion since that cycle is not yet complete), expansions have averaged 50 months and contractions only 11.

Table 1. Business Cycle Duration (average months)

Number of Cycles	Contraction	Expansion
1854-1991 (31 cycles)	18	35
1854-1919 (16 cycles)	22	27
1919-1945 (6 cycles)	18	35
1945-1991 (9 cycles)	11	50

The most recent recession was also quite mild by historical standards in terms of the rise in unemployment rates and duration. However, the recession just before that in 1982-83 was the most severe in post-WWII history. The two recessions of the 1970s were also rather severe by historical standards.

I would conclude that in the U.S., recessions do seem to be occurring less frequently lately, but the severity may not be so different than in the past. However, it is worth noting that this is a trend that has been ongoing since the 1800s. It seems that the pace of change has accelerated a bit. But given the evidence, it would seem premature to declare total victory over the business cycle. Chances are, there will be another recession.

Why might recessions be less frequent? To answer that question, we would need to have a view on what causes recessions in the first place. This is a fundamental question in macroeconomics to which there is no widely accepted answer. One common view of recessions is that they result from an accumulation of imbalances in the economy. Production may not match changing demand patterns, leading to excess supply in some sectors. If the imbalances get too large, layoffs may be necessary and aggregate spending may fall. This process can set in motion forces that lead to recession if high demand sectors are not able to absorb workers who have been laid off.

New economy mavens would say that increased use of computer technology has improved the management of inventories and reduced the chance of large imbalances building up in the economy. In their view, such imbalances arising from coordination failures (millions of people making independent decisions in the market) have become less likely as computers now facilitate rapid transmission of information about demand back to suppliers. A more likely explanation of why imbalances don’t seem to build up as often is that a greater share of output is now produced in the service sector, which is less prone to cyclical fluctuations than, say agriculture, which was a dominant part of the economy in the 1800s. Agriculture has been falling and government and other services rising as a share of total output or demand for well over a century.

Others think recessions come naturally after expansions as aggregate demand inevitably overshoots aggregate supply, leading to rising inflation. Higher inflation forces monetary policy into a contractionary stance, driving up interest rates and dampening spending. Unless the Federal Reserve can manage this process in a very precise manner, there is a good chance that a recession will occur. People who have this view of recessions obviously believe that money supply can influence real output in the short run at least. This theory of recessions is not quite as easily mixed with the new economy view, unless the new economy is really just the Alan Greenspan economy. Perhaps information technology has made the Fed Chairman's job easier, but that seems a bit of a stretch. It is probably true that monetary policy is more sophisticated today than it was in the past, owing partly to more experience with and research on monetary policy.

Is There a Productivity Revolution Going On?

If there is a New Economy, it seems to be widely agreed that the reason is the information technology revolution. As firms have mastered the use of computers in the U.S., they have made dramatic improvements in efficiency and productivity. In our language, the production possibility frontier is shifting out more rapidly than at any time in recent memory.

While GDP growth is helpful in assessing whether something fundamentally different is going on, it has its limitations. There would be nothing miraculous about rapid GDP growth if it were achieved by rapid growth in the number of workers or in the number of hours worked. The real key to improved living standards is productivity growth—increases in output per unit of input.

The data on productivity growth in the U.S. economy show that the 1990s are an improvement over the late 1970s and part of the 1980s. However, productivity growth in the 1990s is nowhere near the rates we saw in the 1960s and early 70s. For the decade of the 1960s, the average annual growth rate of productivity in the non-farm business sector (a series published annually in the Economic Report of the President) was 2.8%. That growth rate declined to 1.9% in the 1970s and a mere 1.1% in the 1980s. For the period from 1990 to 1997, the rate is only 1.2%. That number seems likely to rise as the 1998 and 1999 data become available since recent growth has been quite high.

The productivity data are only as good as the output and input data on which they are based. It is pretty easy to measure bodies or hours worked, but it is difficult to measure output, especially in the rapidly growing service sector of the economy. Many people believe that because we cannot easily capture the improvement in the quality of services, inflation rates are overstated and real output growth is understated. If this is true, then productivity growth is also understated.

Some people take this to mean that even though the data suggest fairly standard behavior of productivity, the reality is that productivity has grown quite a bit. That presumes the measurement problem only applies in the 1990s, which is surely false. It may be more important in the 1990s than it was in previous decades, but measurement error would need to be quite large to make recent performance extraordinary in historical context.

The Implications

So where does this tour through some economic theory and a historical evidence leave us with respect to the New Economy? And what implications would a New Economy have for us anyway?

The notion that we can have high growth and low unemployment without high inflation may be new to some people, but it’s not new to Milton Friedman or most other professional economists. And while there is some evidence that productivity growth is rebounding a bit from a weak period in the late 1970s and part of the 1980s, it remains a far cry from the extraordinary performance of the 1960s. Progress, yes, but not exactly a paradigm shift.

This conclusion shouldn’t come as a great surprise. After all, the U.S. seems to be standing alone as a new economy. While free markets and the new industries that they can spawn are a great thing, it is hard to think that the U.S. is really so different from the rest of the world. Or that those differences have widened considerably in the last decade.

One thing that seems indisputable is that recessions are occurring with a lower frequency today than in the past. This is a trend that has been ongoing for over a century. And it may indeed be related to improvements in information technology which improve the degree of coordination that can be achieved in a decentralized economy, thereby reducing volatility. But chances are, other factors, such as the relative decline of agriculture have been more important.

You hear more talk about the New Economy from Wall Street than from academia. That may be because the Wall Street economists are closer to the action and can see things that can’t be observed from the Ivory Tower. The information technology revolution is hard for academics to see in the data, but private sector economists might have their finger on the pulse. On the other hand, the New Economy is an essential ingredient to relatively high valuations that have been reached in the U.S. stock market. Rapid growth with low inflation and no recession on the horizon would amount to a pretty favorable environment for equities. Since Wall Street economists have a vested interest in high U.S. equity valuations, it is perhaps not surprising that the New Economy has a shaky foundation.

An Impending Test for the New Economy?

The debate about the New Economy may be resolved by the data rather quickly. For the naysayers, the extraordinary performance of the U.S. economy in recent years was facilitated in large part by some chance events: the Asia crisis (which held demand and inflation in check), a possible alteration of the timing of investment spending to meet Y2K problems, and the drastic weakening of oil prices. These factors have begun to reverse course, particularly the oil prices.

While many people are aware that oil price shocks precipitated two worldwide recessions in the 1970s, few people realize how closely U.S. economic performance has followed oil prices over time. Figure 3 plots the price index for fuels and power relative to all other goods and the unemployment rate in the U.S. The correlation is unmistakable. When fuel prices rise, unemployment tends to rise as well, and vice-versa. This relationship is probably driven by two main facts: a great deal of oil is imported and it is difficult to substitute away from oil when prices rise. Thus, when oil prices go up, consumers buy roughly the same quantity, increasing total expenditure on oil and related products. This reduces the amount available to spend on other domestically-produced goods and services, unless households can reduce their saving. Reduced spending can lead to falling profits, and eventually layoffs.

If recent increases in oil prices are maintained or enhanced in the months ahead, the U.S. economy looks somewhat vulnerable. With spending already approximately equal to income (i.e. savings rates are near an all-time low), it seems likely that sustained rises in energy prices will cut into consumer spending on other items. If so, the next recession may be just around the corner. Unless, of course, we really do live in a New Economy.