by Robert Gordon

The New Economy: What Productivity Miracle?

May 15, 20017 mins
CareersIT Leadership

The new economy has been compared in importance with the discovery of electricity and the invention of the internal combustion engine, and journalists frequently describe it as a new industrial revolution. The hype includes claims that the new economy is responsible for sustained productivity growth and that its very existence has eliminated the normal ups and downs of business cycles.

Before we dispute such claims, let’s define new economy. It would be much too broad to describe it as everything the electronic computer has ever achieved. Decades ago, mainframe computers eliminated clerical drudgery, but somehow that did not budge the stagnant growth rate of the U.S. economy’s productivity; it remained a mere 1.4 percent per year from 1973 to 1995. This mismatch between IT investment and stagnant productivity growth evoked Nobel Prize-winner Robert M. Solow’s famous 1987 Productivity Paradox: “We can see the computer age everywhere but in the productivity statistics.”

Because the computer began business operations in 1951 and the economic miracle dates only from 1995, a more focused definition of new economy would be “the acceleration of the rate of technical change in the computer industry after 1995.” The price of computer power (a megabyte of memory or a gigabyte of hard drive capacity) declined more than 30 percent a year from 1995 to 1998, double the 10 percent to 15 percent decline in price before 1995. Combined with this technological acceleration was the invention of Web browsers and widespread access to the Internet.

Taken together, the invention of the Web and the rapid decline in computer prices spurred a massive wave of investment in computers, peripherals and software. An important fact to remember is that economywide investment in computers grew at 20 percent before 1995 and at 40 percent after 1995. As we shall see, the fate of the U.S. economy in 2001 and the years beyond depends on whether computer investment continues to grow at 40 percent a year or slows again to the pre-1995 rate of 20 percent or even lower.

Whatever happens from here, there is no question that the U.S. economic miracle of 1995 to 2000 occurred and created enormous benefits for everyone?not just stockholders but also consumers and employees. Between 1995 and 2000, output per hour in the business sector of the economy rose at an annual rate of 2.9 percent, double the pre-1995 rate. This productivity revival became a source of pride at home in the United States and of envy abroad, and it made the 1995 to 2000 U.S. economic miracle possible. The productivity acceleration held down inflation and directly boosted GDP growth far above rates that were thought possible before 1995, thus allowing Alan Greenspan’s Federal Reserve Board to resist more than minor increases in interest rates. Soaring profits and an unheard-of surge in stock market valuations created trillions of dollars in wealth from this revival in productivity growth.

A Miracle that Can’t Hold

But there are reasons to doubt that the foundation of the Productivity Miracle can remain intact through the next five years. As a historical fact, productivity growth always contains a temporary component when the economywide output of American workers grows faster than its sustainable rate. This seems to have occurred when real GDP grew by 6.1 percent between mid-1999 and mid-2000. Even the most optimistic estimates of America’s sustained growth potential are 4 percent or somewhat lower. If output growth were to slow, productivity growth would also surely slow.

And this is what happened. As real GDP growth fell from 5.6 percent in the second quarter of 2000 to 1.1 percent in that year’s fourth quarter, productivity growth fell from 6.3 percent to 2.2 percent. As the economy slowed, productivity growth slowed with it.

There are several reasons the U.S. economy cannot grow as fast in the next few years as it did during the five-year boom between 1995 and 2000:

  • An unsustainable decline in the unemployment rate.
  • An unsustainable increase in our international trade deficit.
  • An unsustainable increase in stock market prices that in turn fueled unsustainable growth in personal consumption spending.

    But that’s only the beginning of the story. The nation’s greatest experts on the effects of computer investment, Stephen Oliner and Daniel Sichel at the Federal Reserve Board, have estimated that a full two-thirds of the 1.4 percent productivity growth revival was due to the post-1995 growth of computer investment. Part of this acceleration reflected faster growth in production of computer speed and memory per hour of work by employees making the computers, and the rest reflected the benefits enjoyed by the entire economy in having so many more computers around to make each hour of work more productive.

    Too Tied to the Computer

    That’s all well and good, but Oliner and Sichel’s research raises a scary prospect for the next few years. Because they attribute so much of the productivity revival to an acceleration in computer investment, their work implies that most of the productivity revival will disappear if the growth of computer investment dips below 40 percent to, say, the 20 percent growth rate that was typical before 1995. Yet even 20 percent seems optimistic for the next year or two. The growth rate of computer investment was down to 5 percent in the last quarter of 2000 and it looks likely to turn negative in 2001, just as it did back in 1990 to 1991.

    In short, some of the productivity revival was inherently transitory while much of it relied on a 40 percent growth rate of computer investment that could not be and has not been sustained. Think of all the reasons investment in IT hardware is slowing down.

    1. The monumental effort to build websites is largely complete, and it requires less hardware and software investment to simply maintain those sites after they are built.

    2. The timing associated with the year 2000 issue compressed the normal computer replacement cycle into a shorter than normal period during 1999 and 2000.

    3. The usual race to replace hardware to use evermore complex software seems to have petered out. Computer journalists tell their users not to upgrade beyond Windows 98 unless they are heavy users of complex games or heavy downloaders of music and DVDs, yet these types of computer leisure activities do not generate productivity gains for business companies.

    4. Most analysts agree that there has been a massive amount of overinvestment in fiber-optic telecom capacity.

    5. Many doubt that the benefits of the Web can be squeezed onto the tiny screens of next-generation mobile phones, thus raising doubts about the financial sustainability of much of the worldwide telecom industry that has created so many resources and taken on such debt in the hope that the Web-enabled mobile phone is the wave of the future.

    Coming off a high is equally difficult for an economy as for an addict. The adjustment in 2001 to 2002 will be painful. I don’t expect a depression, and we may even escape an officially defined recession. But the days of double-digit growth rates in profits and stock market returns are over, growth in both real GDP and productivity will slow significantly from the past five years, and the adjustments that everyone from Wall Street to Sand Hill Road require have only just begun.

    Robert J. Gordon is the Stanley G. Harris professor in social sciences at Northwestern University in Evanston, Ill. You can read more about his economic arguments at Which side of this debate do you favor? Let us know at