Roundtable on the American Economy

Edmund S. Phelps[1]

I remember almost as if it were yesterday my two long stays at Yale, first as a graduate student in the late 1950s and then as a researcher and teacher in the first half of the ‘60s. As a student of macroeconomics I was one of the prime beneficiaries of the presence there of James Tobin and the late Arthur Okun. They gave unstintingly to me year after year. As one of the oldest participants here I have the pleasure of remembering as well the deep contributions to Yale of William Fellner and Henry Wallich. From them I soaked up as much Continental thought as I could. It is fair to say that much of my life’s research arises from the tensions between the two contrasting perspectives on the employment and growth that these scholars taught.

I am going to take seriously the title of this Reunion by attempting to talk about America at the threshold of the next century – at least the first few years of it.

For a half-decade now, the American economy has been looking very good – better than at any time since the early ‘70s. For a while it seemed that total factor productivity in the U.S. had slowed to such an extent after 1973 that TFP in Germany and France would soon catch up, but it turned out the slowdown finally came to Europe too. Then, in the mid-‘90s, there began to be signs of faster TFP growth in the U.S. – perhaps the fastest in two decades though not a match for the ‘50s and ‘60s.[2]

There is circumstantial macroeconomic evidence of the growth revival in the pickup of real wage rates, the strong rise in corporate earnings and, of course, in GDP per unit of employment. In the rosiest estimate, based on the household survey of employment and the income measure of the GDP, output per employed person increased at the average annual rate of 1.9 per cent between 1993 and 1998. There is also an abundance of evidence at the industry level. American petroleum firms have developed uses of computers to achieve huge reductions in the cost of oil exploration. American pharmaceutical houses have used computers to speed up analyses of laboratory drugs. American retailing and wholesaling has adopted the internet to market and distribute their products, widening information and saving search costs.

Over the ‘50s and ‘60s, when I was much of the time at Yale, we would have sought to explain this revival of dynamism by looking for evidence of increased research input or bigger returns, somehow, from unchanged research effort. Science was fundamental. In the symbolism of Yale, the Gibbs building towered over the modest woodframe quarters of the social sciences on Hillhouse Avenue.

Present explanations of the revival renew the early 20th century emphasis on entrepreneurial vision in applying and developing the research findings and inventions already out there. The switch of focus from scientists to entrepreneurs reflects the fact that the revival of productivity growth has occurred in the U.S., not Europe. The entrepreneurial culture has always been more deeply rooted in the U.S. And in the past two decades the U.S. has pulled farther ahead of Europe in evolving institutional changes that improve owners’ corporate control of managers.

When deciding the causes of the growth revival macroeconomists will want to put a word in for the Clinton administration’s skillful use of what Samuelson dubbed the Neoclassical Synthesis: You can have as austere a fiscal policy as you – or society – can take and still have the equilibrium volume of employment (called “full employment” in those days) provided you couple this move contracting effective demand with an offsetting expansion of money and credit that restores effective demand. James Tobin and Franco Modigliani praised a policy of budgetary surplus, believing that the public saving it represented would speed up the growth of productivity. It may be true in the academic sense that the austere policy mix of the Rubin-Summers years stimulated saving and investment. But if it has done so it is only by serving to moderate the decline in saving that resulted when the worldwide fall of real interest rates and faster TFP growth in America powered a rise of stockmarket prices that induced households to consume more, not less as the theory predicted. The austere policy mix averted a worse rise of consumption but I don’t see how it can be credited with the pickup in productivity growth. (In any case, I believe that the mix may have had some benefits outside the theory that gave a boost to employment.[3])

The revival of productivity growth could be short lived, of course. Like the course of love, the rate of technical progress does not run smooth. But there are often waves of technical progress, since one thing leads to another. I have the feeling that the revival of productivity growth will go on for roughly another decade, barring unrelated shocks to the economy. Some of those with an historical sense have been put in mind of the growth and prosperity in the long span from the 1890s to 1914.

I want to make a few comments that may be of interest to economists on some aspects of this exhilarating period we now find ourselves in. A great many business journalists and commentators are saying that this is a new era. That is unobjectionable, as far as it goes. But then they go on to say that the old theories do not apply to the new era, that what is required is a new paradigm – one that will show, if anyone can properly formulate it, the theoretical possibility of vast increases in employment and profitability without the pathology of ever-rising inflation. Many of these business observers, appealing to higher authority, assert that the Board of Governors of the Federal Reserve System has lost confidence in the Natural Rate paradigm that I and Milton Friedman formulated in different ways back in the late 1960s. This news is disappointing to me, certainly, since until just a few days ago the ‘Fed’ was regarded as a bastion of Natural Rate thinking.

What, however, does the various dissidents from Natural Rate thinking believe? That the parameters or state variables in Natural Rate models have shifted? Or that a fundamentally different theory is required? The Board of Governors do not say. Certainly the suggestion of some observers that American producers can no longer raise their prices belongs in the latter category; it would be a fundamental departure from existing theory. But the globalization of markets has never stopped auctioneers from raising prices – they are said to do it whenever demand exceeds supply. If globalization and electronic commerce mean that an economy’s product markets are better described as auction markets than before, then the only implication for employment that I can think of is that labor’s demand curve will be pulled up by the reduced degree of monopoly power.[4] However, such a development does not seem to be important in explaining the sharp decline of the unemployment rate since mid-decade. It would also imply a sharp rise of real wages. (We have seen a faster rate of increase of real wage rates in the past couple of years but not a ‘level increase’ that would match up nicely with the ‘level increase’ in the equilibrium path of employment.) It would also imply that the rise of employment and wages would date to the beginnings of the recent era of globalization, which most scholars would put in the ‘70s or early ‘80s.

I am convinced that a large part of the sharp decline of the unemployment rate since 1995 or so and, for that matter, the trend-decline in the unemployment rate over the past 10 to 12 years is a part of a longer trend. There has been an extraordinary shrinking of the proportion of the labor force who are relatively unemployment-prone, a development that got into high gear around 1970 and has not stopped yet. This is mainly the result of the increased numbers in the labor force who obtained a college degree or at least had some years of college-level training. If we look at the education-specific unemployment rates we see that these rates were not coming down in the second half of the ‘80s and the first half of the ‘90s; the explanation is simply the compositional effect I just described.

Yet it is important to recognize that this long-term development, which was obscured by forces operating in the opposite direction in the ‘70s and, even more strongly, in the ‘80s, cannot explain the sharp drop of the general unemployment rate that began around 1995. To explain this drop two developments must be considered. One is the marked pickup in productivity growth that dates from that time, which I discussed earlier. The other is some decline of what could be called the world real interest rate, especially but not only in 1997 and 1998. These forces have shifted up the labor demand curve. In so doing, they have pulled up the natural level of employment and the natural path of real wage rates. For me, this simply story, which, by now, has quite a lot of empirical backing, largely answers the question of how unemployment has been able to fall so low without reigniting ever-rising inflation.[5]

According to my analysis, therefore, if the rapid productivity growth experienced since the mid-‘90s continues into the next decade, the likelihood is that this rapid growth will continue to be accompanied by high employment. In America we have been fortunate to inherit the right system and to have the wisdom to keep it and to sharpen it. Europe, which is less fortunate in these respects, may very well see its productivity level lose still more ground relative to the U.S. until the percentage gap has grown so large it stops growing further. When that happens, the improved growth rate of productivity (from its lower level) will provide some boost to employment in Europe. But this improvement of employment will be awfully late and it will not be sufficient to get Europe’s unemployment rates to the American level. Achieving such a target would take a major widening of higher education and fundamental reforms to bring Europe’s system, now highly statist and largely corporatist, closer to the American model of capitalism.

Notes

[1] The speaker is McVickar Professor of Political Economy at Columbia University and currently Advisor to the Strategic Project <Italy in Europe>, Consiglio Nazionale Ricerche, Rome.

[2] Recent analyses by the BLS, Robert Gordon, and Dale Jorgenson contradict this impression but that is only the result of their desire to credit to investment rather than to technical change that part of the acceleration of output attributable to the cheapening of investment goods in real terms, i.e., in consumption goods.

[3] Like a temporary cut in current tax rates on labor, a permanent cut in expected future tax rates on labor would encourage workers to reduce their propensities to quit and shirk, which would stimulate firms to take on more employees. It could also inspire working-age people to invest more in their human capital in order to make themselves more trainable and versatile in high-wage jobs.

[4] Some years ago, when Europe’s “unified market” was in the offing, I formulated this argument in terms of the dynamic theory of customer markets. See Phelps, “1992 Europe as a Unified Customer Market,” in Gerald R. Faulhaber and Gualtiero Tamburini, eds, European Economic Integration, Dordrecht: Kluwer Academic Publishers, 1991, 39-48.

[5] My work over the past decade is summarized in Edmund Phelps and Gylfi Zoega, ‘Natural-rate theory and OECD unemployment,’ Economic Journal, 108, May 1998, 782-801. On a broader front see also my small book Rewarding Work: How to Restore Participation and Self-Support to Free Enterprise, Cambridge, Mass., Harvard University Press, 1997.