The rapid economic growth of the past few years has surprised even the most optimistic forecasters, and it has made a significant difference to Americans’ sense of themselves. The volume of goods and services the economy produces, the Gross Domestic Product (GDP), has grown at an annual rate of nearly 4 percent, discounted for inflation, since 1996. Few economists projected rates of growth as high as 3 percent a year. Only a few years ago, experts said that labor productivity, the output of the economy per hour of work and the source of a nation’s rising standard of living, was barely growing. But it has grown by more than 2 percent a year on average since then, and grew by an annual rate of 3 percent in the nine months ended in June.
The fall in the rate of unemployment has had the greatest social effect. It now stands at 4.3 percent, lower than at any other time since the 1960s. A reduction in unemployment does not simply mean that a higher proportion of people have jobs. It also means that the demand for workers in general is strong, making most jobs more secure and forcing wages to rise as well. During the past three years, wages have done just that, not merely rising for better-off workers, as they did for the previous two decades, but rising by about 8 percent after inflation on average for low-income workers as well. As a result, income inequality, which widened dramatically in the 1980s and early 1990s, is at last narrowing.
Meantime, inflation has, until recently, continued to subside, defying every economic forecast that I have seen. Consumer prices rose 2 percent during the last year. This surprisingly low rate of inflation enabled the nation’s central bank, the Federal Reserve, to lower interest rates rather than raise them as had been expected with such rapid rates of growth. Long-term mortgage rates fell to around 7 percent in mid-1998, for example, making the monthly cost of financing a home purchase lower compared to average family income than it had been in nearly twenty years. Low mortgage rates have made home equity loans to finance other kinds of consumption highly attractive. Low inflation and interest rates have also been powerful stimulants for the stock market, and high stock prices in turn have been a foundation for unusually strong consumer spending that has literally outpaced increases in income in recent months.
To a growing number of observers, this exuberant performance suggests a deep change in America’s prospects. A common explanation is that computerized technological advances have at last turned the economy around. The slow rates of growth that persisted since the early 1970s and led to historically torpid growth in wages for most Americans have ended, some now say, as a new information age at last takes hold. There is a great deal of doubt among economists about this. Many argue that the economy reached a high rate of growth in the last three years that cannot be maintained, in part because it has been supported by soaring stock prices and a growing mountain of debt. Corporate borrowing rose by 10 percent last year, for example, faster than it has since the 1980s, and consumers keep borrowing more on their credit cards and against the value of their homes.
But whatever the causes of the current prosperity, much of America is clearly confident again after two decades in which its citizens continually expressed fears that institutions of all kinds—religious, economic, and political—had lost their sense of direction. The slow economic growth that began in the early 1970s accounted for the social and political confusion to a greater degree than was generally recognized. There was a widespread impression that government—and especially social programs—no longer worked as federal revenues failed to rise as rapidly as they once did. Had the economy grown by merely 1 percent a year more between 1973 and the mid-1990s, for example, federal revenues would have been more than $400 billion higher per year in the mid-1990s.
Because wages stagnated or fell, consumers felt under pressure to save less in order to keep up their standards of living. Jobs were not sufficiently available to relieve poverty in the inner cities, and this contributed to rising crime rates. White male workers in particular had falling average wages and they lost jobs at a high rate, raising tensions in the American family as wives went to work in record numbers and often became the principal breadwinners. Meantime, the gap in wages between African-American and white men remained wide, after narrowing between 1965 and 1975.
The more rapid growth of the past few years has begun to reverse these conditions. The sudden bulging surplus in the federal budget is already tangible evidence of the benefits of rapid economic growth. Tax receipts rose rapidly along with incomes and were further inflated by the increased tax collections on capital gains earned in the financial markets. If such rates of growth continue, even the future financial deficits of Social Security and Medicare will be dramatically reduced.
As a result, surveys find consumer confidence is at an all-time high. The nation shows signs of being more generous, as evidenced by the waning demand for tax cuts and the willingness of state and city governments to spend more public funds on education and transportation. Despite passing an $800 billion tax reduction that the President promises to veto, Republicans in the House are restoring spending cuts for housing, urban development, and other social programs.1 These new spending initiatives are also evidence, I think, that Amer-ica’s much-discussed divisiveness, frequently attributed to the lack of shared ideological and moral principles, has subsided noticeably (as partly suggested by the lament of conservatives that the public supported Clinton during the impeachment hearings). Similarly, it has been thought that inner-city problems were the inevitable result of an intransigent underclass that is alienated from society. But many of the poor in the inner cities are also working again, and crime rates are falling in many of these neighborhoods.2
This is exactly what economic expansions are expected to do. But the expansions of the US economy of the 1970s and 1980s were not nearly as strong as those of the past. And until only recently, the 1990s expansion was the weakest of all. In fact, the current median family income (less than that of 50 percent of the population and more than that of the other 50 percent) of about $46,000 is still only slightly higher than it was in 1989, discounted for inflation; and the median family income of 1989 was itself only a few thousand dollars higher than it was in the early 1970s. The wages of the average worker are only now just reaching their 1989 level, and are still about 10 percent below the level reached in 1973. Lost in the current enthusiasm is what a long way there still is to go.3
Moreover, the growth of real GDP and productivity since 1996 has been no faster than it was during similarly strong but short-lived periods in the 1970s and 1980s that soon exhausted themselves. For example, beginning in Reagan’s first term, between 1982 and 1986, the fastest growing years of the 1980s, real GDP (discounted for inflation) grew by 4.5 percent a year compared to 4 percent a year recently; but the rate of growth of GDP then quickly subsided to only about half as fast over the next ten years. In these same fast-growing years of the early 1980s, labor productivity increased by a rate of 2.4 percent a year (compared to about 2.1 percent a year recently), but in the subsequent ten years it grew at less than half that on average. If the performance of the past few years were truly a turning point, the rates of growth should have probably exceeded historical rates. (They have not been close to the best rates of growth of similar three- and four-year periods in the 1950s and 1960s.)
Admittedly, more rapid growth has come late in the 1990s business cycle, and this has impressed economists. The more rapid growth in the 1980s cited above followed a steep recession; at such a point, growth is usually strongest. But the 1990s cycle has reversed the normal sequence of events. The rate of growth was unusually weak early in the cycle, after the recession of the early 1990s. For example, the rate of unemployment was actually higher in 1993, when the recovery was two years old, than it was at the bottom of the recession in 1991 (and almost as high well into 1994). This was unprecedented for the first two years of a new business cycle. Pro-ductivity was almost stagnant in these years. After an initial burst in 1992, it grew by 0.1 percent in 1993, 0.5 percent in 1995, and 0.6 percent in 1996.
Thus, the late bloom of the economy, beginning in 1996, followed an extremely sluggish period of growth in which demand for consumer and perhaps even capital goods could not be fulfilled. It is possible this pent-up demand has given impetus to current high levels of consumer spending. Even Alan Greenspan, the Federal Reserve chairman, conceded in a speech in May (covering all sides of the issues as usual) that the re-cent economic performance may largely be a catch-up after the unusually slow growth of the first half of the decade.
Also, the data have been inflated significantly since 1995 by revisions which have reduced consumer prices and therefore raised business output, adding approximately 0.4 percent to reported productivity growth each year. This large adjustment will soon be made to earlier data, making the current performance less strong by comparison. Many gushing press accounts leave out this fact.
This summer, the good economic news has been tested. On the one hand, economic growth slowed to an annual rate of only 2.3 percent and productivity growth to only 1.3 percent in the three months ended in June. On the other hand, there have been increasing concerns that the rate of inflation may at last be rising. In particular, the value of the dollar has fallen against the yen and the European Union’s euro, which is helping to drive up import prices. In response to such inflationary threats, the Federal Reserve raised its key federal funds rate by a quarter of a percent in June and another quarter of a percent late this August. Long-term mortgage rates have risen to more than 8 percent, and demand for goods and services may well now be slowing.
Nevertheless, the rate of inflation has remained subdued, and a three-month slowdown in productivity growth obviously cannot be taken as a sign of a trend. Something more than simply a temporary cyclical improvement may be underway. Also, the low rate of unemployment has reminded us once again of the benefits of tight labor markets, which the nation has not experienced for twenty-five years. The rising wages for lower-income workers, for example, raise new questions about a long-held tenet among most mainstream economists. The experts have generally attributed widening income inequality to technological advances that require more sophisticated, better-educated workers. But as labor markets have tightened since 1996, even low-skilled workers have been in demand. Technological change no doubt remains an influence over the job markets, but it is increasingly clear that a low unemployment rate has remarkable curative powers.
Before we explore the possibility that the American economy can maintain low rates of unemployment and continue to grow rapidly, we should take note of the increasingly popular literature whose authors generally claim not merely that Americans have improved their standard of living significantly in the past three years but that most Americans, even the poor, have been doing well since the 1970s. Reports to the contrary, they say, are the result of misinterpreted data and the influence of the pessimistic literature of the 1980s and 1990s. The leading current example of this optimistic school is Myths of Rich and Poor: Why We’re Better Off Than We Think, by W. Michael Cox and Richard Alm. Cox is an economist for the Federal Reserve Bank of Dallas and Alm is a business reporter for the Dallas Morning News. Most of the information in their book has appeared in reports issued by the Dallas Federal Reserve Bank over the past few years. But the book is written in a lucid and sincere style.
The main point made by Cox and Alm is that the data about falling average wages do not accurately represent how well most Americans have fared since the early 1970s. They insist that such frequently cited measures as the average wage for production and non-supervisory workers, about 80 percent of all workers, is misleading, as are even broader measures such as family income. Average wages discounted for inflation, as noted earlier, are still about 10 percent lower than they were in 1973, even after the strong gains of recent years. What this means is that, say, the average thirty-year-old today is making less than the average thirty-year-old did twenty-five years ago. Such a long decline in the average wage over a twenty-five-year period has never occurred before in America’s industrial history. Also as noted, median family income has hardly risen at all.
Cox and Alm propose that a better measure of the nation’s economic health is per capita income—the total income earned in the nation per man, woman, and child. The authors maintain that this measure captures all income, including bank interest, profits, and rents. It also measures better, they say, the benefits earned on the job, such as health insurance. This is a common claim among the school of optimists. The same point is central to another recent book, The Optimism Gap, by a US News and World Report editor, David Whitman.4
Per capita income has indeed risen moderately since the mid-1970s, though at a significantly slower rate than in the 1950s and 1960s. (The addition of corporate benefits, however, when properly measured, makes almost no difference in the rate of growth.) But the authors fail to mention why: many more Americans, as a proportion of the population, are working. Baby boomers reached working age beginning in the 1970s, and 57 percent of women now work, compared to about 36 percent in the early 1970s. As a result, about 52 percent of the total population now holds a job, compared to about 38 percent throughout most of America’s industrial history. If a much higher proportion of all Americans go to work, it stands to reason that per capita income would rise faster than wages.
This doesn’t mean we are better off; it does mean we have to work more for what we want. With so many Americans working, the nation should have increased its standard of living as rapidly as it had any time in the past, or more so. This has not happened. If there hadn’t been so many two-worker families, median family income would have fallen over these years.
Moreover, the rising per capita income has mostly gone to the better-off. Income inequality for both in-dividual workers and families, even with two workers, has widened sig-nificantly since the mid-1970s. As Frank Levy, an economist from the Massachusetts Institute of Technology, notes in The New Dollars and Dreams, a valuable sequel to his influential 1987 book Dollars and Dreams, the richest 5 percent of families earned 15.6 percent of all income in 1969. In 1996, they earned 20.3 percent of all income.
Cox and Alms do not explicitly deny that income is now distributed more unequally. Their reply to concerns over inequality is that, according to their own study, income mobility is still high in America. In other words, a large proportion of those who start out poor or not very well off, they report, rise to higher income levels. To reinforce the point, the authors also cite a 1992 Treasury Department study that reached similar conclusions.
What Cox and Alm fail to mention is that their own study was harshly criticized by such respected specialists in the field as Peter Gottschalk of Boston University as soon as it was released (as was the Treasury Department study). One of several important criticisms is that the authors made no adjustment for the age of those they tracked. In other words, they included very young workers in the lower-income groups (as far as I can tell, as young as sixteen), and even those with part-time jobs while in school. As they get older, these young workers will naturally earn much more. This is not a sign of income mobility; it is a sign of growing up. Measuring income mobility is, in truth, difficult to do, but studies that make plausible adjustments conclude that income mobility is dramatically lower in America than Cox and Alm claim, and some studies show that it has declined in recent decades. (One comprehensive study by Gottschalk and Sheldon Danziger shows, in fact, that income mobility in America is no higher than it is in Europe.) Cox and Alm do not mention their critics and make no attempt to answer them. The unsuspecting reader would think there is no controversy here.5
Similar criticisms can be made of the partial nature of almost every argument the authors make. They imply that the millions of workers who were laid off over these years usually got good new jobs when the evidence is clear thatthose who found new jobs, on average, had to take substantial pay cuts. The authors note that home ownership reached 66 percent of households, “an all-time record,” in 1997. In fact, home ownership has more or less hovered around this level for twenty-five years largely because the elderly of an earlier generation who did not own homes were dying, and being replaced by a new generation of elderly who bought their homes in large volume in the 1950s and 1960s. If this had not been the case, home ownership would have fallen because a lower proportion of young people bought homes over the past thirty-five years.
But what has attracted most attention to Cox and Alm’s book is their computations of how much time the employee making an average wage must work to buy a variety of goods today compared to 1970. It took the average wage earner only seven minutes of work, they compute, to buy a half-gallon of milk in 1997, for example, compared to ten minutes for the average wage earner in 1970. It took twenty-three hours of work to buy a room air conditioner, compared to forty-five in 1970. It took 1,365 hours to buy a 1997 Ford Taurus, compared to 1,400 hours to buy a Ford Galaxie in 1970.
Thus, even Americans with only an average wage, the authors tell us, are able to buy more than they once did even as the average wage has declined. But no one has denied that some goods are produced cheaply and more productively than they once were. Nor is anyone denying that there are new, interesting products in the economy. In fact, it is only natural in a growing economy that many goods are manufactured more productively over time, and can therefore be sold more cheaply or at least don’t rise in price as fast as other goods. Even the productivity of some services has increased.
But this is not true of all goods in the economy, or, most important, of many essential services. Cox and Alm are highly selective about the goods they choose to write about. As even they note, it took an additional 1,078 hours of work to buy a 1,975-square-foot house in 1997 compared to 1970. It even took more work time to buy a pair of Levi’s jeans. Looking at their own selective table, it is also clear how much greater the productivity gains usually were before 1970 than after. If we could manufacture a comprehensive table of such reductions in hours worked in order to buy the same package of goods since the 1850s, the reductions would have been far more dramatic than in the past twenty-five years.
Most important, Cox and Alm pay very little attention to the costs of such essential services as education and health care, and for good reason. It would undermine their argument. It takes many more hours of work to get an education these days than it did before 1970. Public expenditures per pupil rose much faster than did consumer prices in general, as did the cost of private education. Few would argue that the quality of education has risen significantly. It similarly costs many more hours of work to obtain adequate health care. The quality of health care has certainly risen, but since 1970 expenditures per person have considerably more than doubled after inflation. We should remember that a microwave oven or a VCR costs at most half of the monthly cost of a family health insurance plan (VCRs are widely owned in poor nations), and their costs are tiny when compared to the costs of college tuition or the property tax that supports decent K-12 public education.6
Frank Levy’s The New Dollars and Dreams, though technical in places, provides the sense of history that Cox and Alm lack. Incomes do not have to fall absolutely to demonstrate that times have been difficult for Americans. Declining average wages aside, family incomes have grown only slightly on average. Such slow-growing family incomes have been a jarring experience for many Americans who had become accustomed to consistently and rapidly rising standards of living since the nineteenth century—especially when the costs of essential services such as education and health care have risen much faster. For example, median family income rose by 40 percent in both the 1950s and the 1960s, Levy points out, but has risen by only about 7 percent since 1973. As for incomes of individual workers, the average twenty-five-year-old in the 1980s, who is now over thirty-five, has seen a much slower increase in pay than the twenty-five-year-olds of the 1950s. What is more, the thirty-five- to forty-five-year-olds of the 1980s are on average literally earning less in the 1990s even as they are now ten years older and that much more experienced.7
Writers like Cox and Alm rightly remind us that, despite slow-growing incomes, this has also been an age of significant innovation. Progress did not stop, and many new and exciting products have been introduced at affordable prices, even for the relatively poor. Levy does not pay sufficient attention to this. (He also largely attributes inequality to educational differences in a technological age—although, as we have seen, a low unemployment rate is helping to narrow inequality.) But higher-quality products and more varied ones generally favor the well-off who can afford them. And are these products any more exciting than the products introduced in earlier ages, such as the sewing machine, the telegraph, the railroad, electricity, the automobile, the telephone, movies, the passenger jet, air conditioning, or even the fast-food chain? In truth, despite their abundance, and their much publicized promise, they are, so far, probably less so. Moreover, not just the quality of progress but the rate of progress also matters. The addition of three more bottled beers to the grocery shelves matters much less when twenty are already available than when only two or three are available. A former Bureau of Labor Statistics economist, Jack Triplett, now at the Brookings Institution, has written a particularly valuable article, “Economic Statistics, the New Economy, and the Productivity Slowdown,” showing how even sophisticated economists fail to take into account such effects.
The fact is that the requirements for joining the middle class keep rising, especially as the small proportion of Americans whose incomes have risen rapidly keep raising the standard. About this issue, Cox and Alm, as well as most other authors in this tradition, are tone-deaf. They say many Americans can now easily afford the products of the past, and that is a sign of substantial progress. But someone who buys only what a middle-class American could afford in the 1950s—the clothes, the home appliances, the vacation—would not be considered middle class today. And some things were clearly better then. We cannot regain the clean air, uncluttered highways, internationally competitive schools, and very low rates of crime of the past. Those who study such social indicators as suicide rates and health insurance coverage generally see the past twenty-five years, on balance, as a period of decline.8
In The New Dollars and Dreams, Levy notes that in the 1970s college graduates became much more interested in finding jobs that would pay well. “In the late 1960s,” he writes, “when wages were growing and the unemployment rate was below 4 percent, a white male college student could major in anything and know he would graduate with a well-paying job…. By the mid-1970s, the sense of effortless upward mobility had disappeared.” The proportion of young Americans who considered themselves middle class rose consistently in the 1950s and 1960s, but after 1970 it began to diminish.
Finally, if new products so satisfied Americans, why didn’t they buy fewer of them and save more for their retirement or their children’s education? The proportionate amount average Americans saved from their weekly paycheck has been falling steadily since the 1980s and, officially, is now negative. Why have they borrowed so much more? The proportion of the equity in their homes owned by Americans has also steadily fallen over these years because they keep borrowing against their value.9 In fact, Americans have been struggling to buy the minimal goods and services that would maintain their social status, and many of them have not been able to do so even as spouses went to work.
By claiming that Americans have done well since the 1970s, authors like Cox and Alm minimize the importance of rapid economic growth during the past few years. Judging by rising wages for nearly all workers, narrowing income inequality, falling rates of crime, and what appears to be a more confident electorate in just the last few years, there should be little doubt about the importance of maintaining a high rate of economic growth. The central questions are whether the US can continue to maintain a low rate of unemployment without generating significantly higher inflation and whether the rate of growth can be permanently raised above its slow pace since the early 1970s.
One reason for concern is that at least part of the recent reduction in inflation is caused by rapidly falling import and commodities prices that followed the Asian financial crises and is therefore probably temporary. To the extent that this is true, America’s ability to reduce the unemployment rate to current levels has been partly—and some economists would say largely—a matter of luck. The troubled Asian nations slashed the prices of their exports to help rescue their plunging economies; and the flow of capital out of Asia and back to the US pushed the dollar still higher (making imports even less expensive). Slow growth worldwide, including not only the nations in financial crisis but also Japan and Europe, also brought down the prices of key commodities, notably oil. Robert Gordon, a Northwestern University economist, estimates that these two factors accounted for a fall in US inflation of 1.4 percent. Gordon also believes there were other temporary factors contributing to inflation’s fall, including a probably unsustainable reduction in computer prices and medical costs. Slow growth also left the world with excess productive capacity. Had inflation not fallen in this period—ironically, had there been no Asian financial crises—it is likely that the Federal Reserve would have raised rather than reduced interest rates in 1997 and 1998, thus dampening economic growth and probably raising the level of unemployment. By reducing rates the Fed further stimulated growth.10
On the other hand the changing composition of the labor force may have also permanently contributed to lower unemployment. Alan Kreuger and Lawrence Katz, economists at Princeton and Harvard, respectively, point out in a forthcoming paper, “The High Pressure U.S. Labor Market of the 1990s,” that the dramatic increase in the number of imprisoned men and the falling proportion of young workers removed groups of people from the labor force who typically have above-average unemployment rates. In 1998, for example, male prisoners made up 2.3 percent of the potential male labor force. The proportion of workers under twenty-five fell from 24.5 percent in the late 1970s to 14.8 percent in 1997. According to Kreuger and Katz, these two factors alone may have allowed the unemployment rate to fall by as much as 0.6 percent without generating traditional pressures on inflation.
Another factor Kreuger and Katz believe accounts for the noninflationary fall in the unemployment rate is the high proportion of temporary workers in the economy. They make up 2.2 percent of the labor force today, compared to 0.5 percent in the early 1980s, and there has been a commensurate growth in temporary employment services. As a result, qualified workers may be easier and less costly to find—temporary workers rarely receive corporate benefits—and, with fewer workers looking for jobs, the rate of unemployment will fall. The availability of temporary workers also puts pressure on the wages of those already employed while undermining the bargaining power of labor unions. Kreuger and Katz estimate that these factors might have reduced the unemployment rate by as much as another 0.4 percent. Kreuger and Katz warn that their estimates are highly subjective, but, taken together, they could account for a full percentage point fall in the unemployment rate, or more, compared to the 1980s.
Such accounts suggest that the nation can have a lower rate of unemployment without a technological transformation of the economy. But they also suggest, now that the rate of unemployment has already fallen so low, that we cannot continue to grow as rapidly unless productivity continues to rise strongly. Because there are fewer workers to be hired, to grow fast would require more output per worker. The same accounts also imply that the current unemployment rate may already be inflationary unless productivity continues to grow rapidly. If the distressed foreign economies make a strong recovery, the fortu-itous fall in import and commodities prices could be reversed. This may already be happening. Import prices have risen recently.
Until a few months ago, arguments that the economy had entered a “new era” of rapid productivity gains based on computerized information technology were taken much more seriously by investment brokers and the financial press than by economists. Productivity grew by more than 2 percent a year on average between 1870 and 1970, and by nearly 3 percent a year in the twenty-five years after World War II. The 2 percent productivity increase from 1996 to 1997 was welcome but not as unusual as often reported.
During the past year, productivity grew by an impressive 3 percent, however. Even some previously doubtful economists are now beginning to be encouraged, though they remain cautious. The caution is appropriate. Short-term spurts in productivity could merely reflect a sharp increase in demand for goods and services, putting the economy in a sort of overdrive that may only be temporary even though it can last for a year or longer. When productivity in fact did grow rapidly in the 1950s and 1960s, it increased at an annual rate of 3.5 to 4.5 percent for two and three years at a time (and these statistics do not take account of upward revisions to the data that have been made recently).
But one of the facts that optimists find most encouraging is the persistently high rate of investment in information technology itself. Daniel Sichel, an economist with the Federal Reserve, argued in his 1997 book The Computer Revolution that investment in computers, software, and related areas had not been large enough to add as much growth to the economy as the optimists have claimed. Computers and related equipment comprised too little of the nation’s capital stock, in large part because they become outdated quickly.
In a recent paper in Business Economics, “Computers and Aggregate Economic Growth,” however, Sichel finds that the investment in computers by business in the past three years has risen dramatically, implying that returns on computers are still unusually high. If these investment decisions by business are largely correct and not merely reflective of herd behavior, computers are adding value to businesses at an unusual rate. And now, Sichel estimates, the value of capital stock has risen to the point where, even given rapid depreciation, computers may well be contributing significantly more to growth.11 The Commerce Department recently reported, for example, that information technology industries account for one third of the economic growth in the past four years.12
But Sichel proposes another explanation for the high levels of investment in information technology that has less optimistic implications for the future. The price of computer power has fallen dramatically, which means it can be profitably applied to many more activities. So it may not be that computers are producing so much more, but rather that they are much less expensive (and increasingly so) than the alternatives. Thus, high technology in general may be a dynamic sector that is attracting investment and creating hundreds of thousands of good jobs but it may not necessarily be creating a long-term transformation in the economy. If the price reductions in computer power don’t continue at the same rate, then the rate of new investment may fall dramatically. In the early 1980s, Sichel notes, there was a similar period of strong investment in computers which eventually exhausted itself.
Other writers also warn that we are making too much of technological change. Robert Brenner, a historian from the University of California at Los Angeles, argues in Turbulence in the World Economy, a book based on his recent article in New Left Review13 to be published later this year, that strong growth and resurgence of capital spending in the US may largely be the result of its ability to suppress wage growth while its competitors cannot. He thinks technological opportunities were fairly constant over the past twenty-five years and attributes the falling rates of capital spending in the 1970s and 1980s to intense competition for global markets that began in the 1960s. Competition reduced rates of return in most of the advanced nations, and therefore capital investment. The US, Brenner writes, adjusted more quickly than other nations because of the pliability of its workforce (as well as the advantages of its large marketplace). As corporate profits increased, so did the rate of capital investment, and, says Brenner, productivity rose as a result. America’s overseas competitors, unable to suppress wages, were locked into outmoded capital investment and technologies. Brenner writes that as wages now rise, the American advantage will dissipate. The high US dollar will also take a toll.
But many advocates of the new economy believe that computerization has far deeper consequences than this, and it is hard to escape the sense that some technological overhaul in the economy is now taking place. The railroads of the last century made transportation much less expensive and stimulated an investment boom in steel and machinery. But they also transformed America into an effi-cient nationwide marketplace, which benefited all kinds of other products by allowing them to be transported and marketed much more productively. Thus, the railroads produced what economists call high “social returns.”
We often hear that computers are having a similar effect. The Internet, for example, is replacing many retailers and distributors with a productive electronic marketplace. It is also creating price competition between producers in remote parts of the world, forcing all market participants to become more efficient in order to survive. And the rapid growth in productivity of the last year is at last providing enthusiasts with some hard evidence that such benefits are now spreading throughout the economy.
The question is not whether there are new social benefits from computer investments—which even critics like Brenner would not deny—but how significant they are. Robert Gordon, claims that the Internet is simply doing somewhat more efficiently what has already been done, such as advertising and retailing. Moreover, many uses of the computer merely give corporations a temporary competitive advantage that soon subsides when competitors match it—a fancy new marketing presentation, for example. Thus, they merely shift wealth between companies rather than create more wealth. For such reasons, Gordon believes that social returns for computers may be small.14
My own doubts about a new long-term trend in productivity growth are related to Gordon’s. The computer has made production and distribution faster and cheaper. But it has also changed the nature of competition. In the age of standardized products and mass production, size and speed mattered most. Fewer workers were needed to make products. But today, businesses constantly neednew, varied, and upgraded products, as well as managerial methods, to compete.15 The premium is on innovation, and this requires human skills such as creativity and imagination in tandem with the computer. They are not skills that the computer can replace as rapidly as mass production once replaced workers. Today the great-est growth in jobs is in “business services,” such as those provided by consultants.
Thus, in my view, there has indeed been an important technological transformation in America, but not necessarily one that will lead to more rapid productivity growth. The new economy is a return to a contemporary version of a crafts economy, where the most valued contribution is the skill of the worker. The new crafts workers are business managers, consultants, and marketing and financial experts as well as scientific and technical researchers and computer specialists. Or to put it differently, in an age where information is so widely available, it is not the information that is valuable but what is done with it.
Can we overcome such potential barriers to productivity growth? The social returns from computerization may still prove large, of course. But another lesson we may draw from the economic performance of the past few years is that more stimulative government policies, in this case a loosening of credit on the part of the Federal Reserve, may themselves generate more productivity. Such policies, to which the economist John Maynard Keynes would have been sympathetic, increase demand for goods and services, which stimulates capital investment and may result in higher productivity growth. In a time of low inflation and low interest rates at home, and recession abroad, it may well be appropriate to heed such advice again. Fast growth has many unrealized advantages, including the ability to reinforce itself.
If the economy continues to grow strongly despite the Fed’s interest rate increases, it is likely the central bank will raise rates again. Similarly, if the dollar keeps falling, the Fed may also believe it is obliged to raise rates to attract foreign investment to support the dollar’s value. A falling currency is inflationary because, as noted, it drives import prices up. Excessive speculation in the stock market has also been a source of concern for Alan Greenspan and the Fed’s governors because it may eventually lead to a serious correction downward.
But the US economy may also now be especially vulnerable if the Fed’s tightening goes too far. With stock price at extremely high levels, any correction could seriously dampen consumer spending, which has been the foundation of rapid growth during the past year. The nation’s trade deficit continues to expand, putting a great deal of debt in foreign hands. If foreign lenders withdraw their money from the stock market as prices fall, interest rates in America will rise further, restraining capital and consumer spending and probably sending stock prices down further. What was once a virtuous circle in which a rising dollar, rising stock prices, and low inflation and interest rates all reinforced each other could suddenly become a vicious circle involving a falling dollar, falling stock prices, and rising inflation and interest rates. If the Fed is determined to prevent even the slightest rise in inflation, a modest slowdown could become a full-blown recession.
A recession any time soon would quickly reverse the gains made by workers in the past few years. A significant slowing in the rate of GDP growth will end the rapid increases in productivity. Moreover, urgent public needs, from more spending on education and day care to new transportation and research projects, are only beginning to be addressed. If attended to, these could eventually enhance productivity; but if current prosperity fades, so will the prospects for such investment and research.
If the long-term trend rate of productivity were raised by even 0.3 percent, it would add hundreds of billions of dollars to GDP. But claims that the long-term trend of productivity growth has risen by a full percentage point a year over former rates are decidedly premature. We will have no clear answer about the future course of productivity without three or four more years of economic evidence at best. The country will have to face critical decisions before that about such matters as its public investment in education and infrastructure and reforming Social Security. More rapid productivity growth could largely eliminate the need for Social Security reforms. But as Alan Greenspan warned this June, “History is strewn with projections of technology that have fallen wide of the mark.” The Federal Reserve in particular will have to decide on the degree of stimulus the economy can withstand without secure knowledge about the future course of productivity. If it raises interest rates too much, it could well bring what otherwise may have been a more sustained rise in productivity to a premature end.
—August 24, 1999
September 23, 1999
For example, see Mitchel Benson and Jacob M. Schlesinger, “States’ Tax-cutting Binge is on the Wane,” Wall Street Journal, May 13, 1999, p. A2; and Tim Weiner, “House Speaker to Shift Tacks in Spending Bill,” The New York Times, June 5, 1999, p. A11. ↩
Richard B. Freeman and William M. Rodgers, “Area Economic Conditions and the Labor Market Outcomes of Young Men in the 1990s Expansion,” NBER Working Paper No. 7093, April, 1999. ↩
Economic Report of the President, Council of Economic Advisers, 1999. ↩
Walker and Company, 1998. ↩
See Peter Gottschalk and Sheldon Danziger, “Family Income Mobility—How Much Is There and Has It Changed?,” forthcoming, and Peter Gottschalk’s privately distributed memo on the Cox-Alm study. ↩
For an opposing view on America’s living standards to that of Cox and Alm, see The State of Working America 1998-1999 by Lawrence Mishel, Jared Bernstein, and John Schmitt (Cornell University Press, 1999). It offers the most comprehensive analysis of wage, income, and employment trends available in one volume and is updated every two years. ↩
Cox and Alm subscribe to the view that the conventional data are also misleading because the rise of consumer prices has been overstated and the growth of incomes understated. As usual, the authors do not reply to critics of this point of view, notably the Bureau of Labor Statistics, which gathers the data. I have addressed this issue in these pages (see “The New Cost of Living: A New Myth,” The New York Review, March 6, 1997) and think these claims are exaggerated. But even if Cox and Alm were right, the data have been almost equally mismeasured before the 1970s as well. So the relative decline in performance since the early 1970s would be about as great. ↩
See Marc Miringoff, Marque Luisa Miringoff, and Sandra Opdycke (contributing author), The Social Health of the Nation (Oxford University Press, 1999). (I had a minor advisory role in this book.) ↩
According to Edward N. Wolff of New York University, mortgage debt as a percent of the market value of a house has risen from about 21 percent in the early 1980s to more than 40 percent recently (forthcoming, Century Foundation). ↩
The Federal Reserve, judging by past behavior, subscribes to the widely held view that inflation will increase as the rate of unemployment falls below a specific level—known as the “non-accelerating inflation rate of unemployment,” or Nairu. When Nairu is reached, the pressure to raise wages increases. Until recently, Nairu was said to be 6 percent. See Robert J. Gordon, “Foundation of the Goldilocks Economy: Supply Shocks and the Time-Varying Nairu,” Brookings Papers on Economic Activity, 2:1998. ↩
Some economists have found that returns on computer investment have remained unusually high. Others believe they have fallen. See my “Computers: Waiting for the Revolution,” The New York Review, March 26, 1998. The economist Robert Gordon believes companies that use computers more may simply be taking profits away from those that use them less. ↩
“The Emerging Digital Economy,” Department of Commerce, June 1999. Some economists will argue that the estimate the federal government uses for deflation of information technology output is wrong and overstates this output. ↩
“The Economics of Global Turbulence,” New Left Review, May-June 1998. ↩
Robert J. Gordon, “Monetary Policy in the Age of Information Technology: Computers and the Solow Paradox” (unpublished paper, 1998). ↩
Some economists will argue that the greater variety of new products of all sorts raises real income more than has been reported by the federal government. But such innovation has always occurred, and adjustments taking account of it have never been adequately made. If such adjustments to income had been made all along, the slowdown in income growth since 1973 would still be significant. ↩