T here is a common ground on economic policy that now stretches, with differences only of degree, from the radical right to Bill Clinton. Across the spectrum, all declare that the main job of government is to help markets work well. On the supply side, government can help, up to a point, by providing education, training, infrastructure, and scientific research--all public goods that markets undervalue. But when it comes to macroeconomic policy, government should do nothing except pursue budget balance, and leave the Federal Reserve alone.
To accept a balanced budget and the unchallenged monetary judgment of the Federal Reserve is, by definition, to remove macroeconomics from the political sphere. Thus, the remaining differences between Clinton and the Congress are over details. Should we head for budget balance in seven years, eight, or ten? Should we cut (or impose) this or that environmental regulation? Do Head Start, the AmeriCorps, and technology subsidies justify their cost? And so on, in long litanies that no one believes will make a fundamental difference in American lives. Even if there were substantial gains to be made by public investments on the supply side, the conservative fiscal consensus precludes them by denying the resources.
We have now seen two Democratic presidents--Carter and Clinton--deeply damaged because they did not dispute this orthodoxy in good time and therefore could not control the levers of macro policy. Macroeconomics, not microeconomics, is the active center of power. Practical conservatives understand this. It is no accident that conservatives always seek to control the high ground of deficit and interest rate policy, nor any surprise that liberals defeat themselves from the beginning when they concede it.
Yet, the economics behind this consensus is both reactionary and deeply implausible. It springs from a never-never-land of abstract theory concocted over 25 years by the disciples of Milton Friedman and purveyed through them to the whole profession. Liberals--and anyone else concerned with economic prosperity--should now reject this way of looking at the world.
THE RIGHT-WING CONSENSUS ON EMPLOYMENT AND INFLATION
The conservative macroeconomic creed is built on three basic elements. They are, first, monetarism--the idea that the Federal Reserve's monetary policy controls inflation, but has little effect on output and employment except perhaps in the very short run. Second, there is rational expectations, which is the idea, for which Robert Lucas just won the Nobel Prize, that individual economic agents are so clever, so well informed, and so well educated in economics that they do not make systematic errors in their economic decisions, especially the all-important choices of labor supply. And third, there is market clearing: the idea that all transactions, including the hiring and firing of workers, occur at prices that equate the elemental forces of supply and demand.
Taken together, these assumptions conjure an efficient labor market that yields appropriate levels of employment and wages. The employment level generated by this abstraction is the core policy concept of mainstream macroeconomics, known as the natural rate of unemployment.* If unemployment is above the natural rate, the theory dictates that prices and wages will fall. If unemployment is below the natural rate, the theory dictates that inflation will rise. Sustainable, noninflationary employment growth occurs only at the natural rate. [See Robert Eisner, "Our NAIRU Limit: The Governing Myth of Economic Policy," TAP, Spring 1995.]
Among most economists these ideas are amazingly noncontroversial. The only dispute is over a narrow point of policy--whether there is any value in attempts to steer the economy toward the natural rate if it happens to be, for a time, either above or below it. To the strictest natural-raters, doing nothing is always and everywhere the right prescription, because the economy will always return to the natural rate on its own. Policy cannot help, and the very instruments of macro policy should be abandoned.
The self-described "New Keynesian," a breed found throughout the Clinton administration, believes a vestigial role for macro policy can be preserved. Unemployment may persist above the natural rate because wages take more time than other prices to adjust to changes in supply and demand, leading to failure of the labor market to "clear." That being so, there may be no harm in policy measures--a little stimulus now and then when there is a serious recession--to speed the return to the natural rate so long as a "soft landing" is carefully engineered.
Alas, the location of the natural rate is not actually observed. Worse, the damn thing will not sit still. It is not only invisible, it moves! This is no problem for the never-do-anything crowd. But it poses painful difficulties for would-be intervenors, those few voices in the administration who call, from time to time, for summer jobs, public works programs, and lower interest rates. How can one justify a dash to the goalposts, if you don't know where they are? New Keynesians obsessively estimate and re-estimate the location of the natural rate, in order to guide their policy judgments. Sadly, they have never yet been able to predict its location, which may be one reason why there has never yet been a successful "soft landing."
WHERE IS THE NATURAL RATE?
To the (questionable) extent that the Federal Reserve has any coherent macroeconomic theory, it tends to be implicitly New Keynesian on this issue. That is, the Federal Reserve Board is an inveterate intervenor, raising interest rates when unemployment is too low, and lowering them, grudgingly, to end or sometimes to avoid recessions. And so the Federal Reserve also spends a good deal of time and effort trying to pin down the phantom and elusive natural rate.
In 1994, with the natural rate estimated by numerous astrologers at about 6 percent, monetary policymakers faced an interesting problem. Actual unemployment, now at 5.8 percent, had fallen below the estimated natural rate. So how then to interpret the rest of the data, which contrary to theory showed no evidence of accelerating inflation? Did the apparent lack of inflationary acceleration mean that the natural rate had perhaps fallen, and if so to what value? Or, had the barrier been broken and, in Robert Solow's phrase, was inflation "acceleration just around the corner"? Or again, was the whole theory rotten and fit for the garbage?
The Federal Reserve proved unwilling to change its estimate of the natural unemployment rate. So it tightened monetary policy, from Feb ruary 1994 through early 1995, as the economy broached the 6 percent unemployment barrier. But then the Federal Reserve shifted course and started cutting interest rates in July 1995, even though unemployment remained below 6 percent. Why? It will be interesting to learn, when the full minutes are released, whether the Federal Reserve formally changed its estimate of the natural rate in July of 1995, and if so, on what ground and to what number. Or we may learn that the Federal Reserve doesn't really have a natural rate theory anymore, but is only holding on to the rhetoric of these ideas, for want of any alternative that ideological conservatives might accept.
T he components of today's low inflation rate are not at all consistent with the natural rate theory. No part of present inflationary pressure, such as it is, stems from wages. Wage compensation, two-thirds of all costs, remains flat. The whole of today's modest inflation stems from a boom in profits and investment income, and from the effects of this boom on commodity prices and other incidentals of the inflation process. There has also been some contribution from the rising interest costs imposed since February 1994 by the Federal Reserve's own policy.
This problem is illustrated in "What's Driving Inflation?" (above). The old relationship between inflation and labor costs really has busted up since Reagan fired the air traffic controllers and he and Volcker overvalued the dollar. Prices may be rising at 2.7 percent annually, but real wages are scarcely moving. Indeed we find that all inflation accelerations after 1960, with the sole exception of that following Richard Nixon's election campaign in 1972 (when price controls were in force), were led by prices and not by wages.
THE NOMADIC AIRU
How is all of this to be reconciled with a theory of inflation acceleration based exclusively on the natural rate of unemployment in an aggregate labor market? It can't be done. If there is excess demand for labor, surely a good (new) classical economist must insist that real wages are rising. But they aren't--and haven't been in 20 years. Something must be wrong with the natural rate model. (Good economists at the Federal Reserve know this, and it bothers them, as it should.) In fact, something is more than wrong with the model. The model is junk, as we should have known long ago.
This is nicely shown by the two graphs above ("Follow the Bouncing Natural Rate"). The first shows how the unemployment rates at which inflation accelerated have changed over the last 40 years. In the 1950s they were low, in the 1970s, quite high. But recent data rather resemble the 1950s again, which would indicate that there is room for unemployment to come down without kicking off inflation. Depending on how you factor in the high-inflation decade of the 1970s, an honest estimate of the natural rate--even if you believe it--might be 6 percent or much lower.
But the second graph shows how fruitless the search for a natural rate really is. The graph employs centered 12-month moving averages of monthly data for both inflation and unemployment. It illustrates that rising inflation is essentially unpredictable: The shocks that cause it sometimes happen at high unemployment, sometimes not until unemployment is quite low. There is no sign in recent data of rising inflation as unemployment falls.
Indeed, the pattern of widening gyres reversed itself after the deep recession of 1982. Through the rest of the 1980s, unemployment fell without wage pressures and without sharp rises in inflation. The recession of 1989 hit while inflation was low by historic standards. And in the past four years, 1992 through 1995, there has been falling unemployment with falling unit labor costs and no rise whatever in inflation (see the horizontal ellipse). We do not know where the nomadic accelerating inflation rate of unemployment (AIRU) is today--because we don't know when the next shock might hit us. So why not go for full employment?
LIBERALS LOST ON THE SUPPLY SIDE
At the very least, New Keynesian acceptance of the New Classical theoretical structure reduces macroeconomic policy to the fringe role, that of large-scale intervention only in deep and lasting recessions. In all other circumstances, the macro authorities are warned off--as was Clinton himself during his brief Keynesian phase in early 1993.
What then can liberals do? The actual approach of the Clinton administration illustrates: Liberals can favor education, training, adjustment assistance, and other programs that upgrade skills and help workers move from one job to the next. They can support public investments in infrastructure, on the ground that these assist in the international competitiveness of the economy. They can support a combination of research and development assistance to advanced enterprises, alongside efforts to open foreign markets to American products, that help shore up the position of American companies in the world. If they are feeling brave, they can also support a higher minimum wage.
All of these are supply-side measures (except the last, which is a direct intervention in the labor market). Their purpose is to improve the long-term competitive performance of the American economy, on the thought that a more productive economy will generate higher average living standards. The further thought, that these higher averages will trickle down to low-paid production workers, is left as an assumption.
We can all agree that expenditures on education, training, research, development, and infrastructure are generally good things. But a macroeconomic commitment to full employment is the key to translating these investments into higher growth and living standards.
Education and Training. As work by Richard Rothstein in this journal and elsewhere has shown, the American system of public education is much better than its critics allege. To be sure, there are poorly educated Americans who are unqualified to fetch high wages in the job market. Yet as Katherine Newman and Chauncy Lennon have shown [see "The Job Ghetto," TAP, Summer 1995], there are far more qualified applicants in central Harlem than there are minimum wage jobs. In this macroeconomic climate, based on a false theory of the NAIRU, raising education standards will not call better jobs into being. That remains a demand-side problem. This does not mean we should give up on the task of improving inner-city schools. It only means that unless we also raise growth rates--indeed, even if we gave everyone Ph.D.s--so long as the Federal Reserve held the growth rate to 2.5 percent, it would not change the current trajectory of living standards. As a student of mine from Buenos Aires once cracked, if education alone were that powerful, Argentina would be much richer than it is.
Research and Development. A stronger case can be made for the idea that government R and D helps American companies become more technically advanced and more competitive in global markets. For four decades, our highly interventionist technology policy was a by-product of the Cold War. Now, a more explicit civilian technology policy may be needed to make up for reduced Pentagon R and D. There is surely a role in general terms for science and technology policy--ultimately many technologies lead to a better life. But they do not and cannot bring full employment, nor do they bring about a fairer and more just social order. To make science and technology policies the centerpiece of a progressive agenda, while giving up macroeconomics, is absurd.
Infrastructure. Public works expenditure is the historical cornerstone of liberal interventionism. Public works are the fastest, most direct way to put the unemployed to work. They have direct and multiplier effects on total employment. They have the side benefit that the works themselves remain useful for many decades after they are completed. They also represent, in political memory, the triumph of liberalism in the first New Deal.
But the liberal supply-siders make an entirely different claim for public works spending. Renaming it "infrastructure," (as I too have done on many occasions) they argue that it contributes in definite ways to the productivity of the private business economy. The jobs created directly, by doing the work, are immaterial to this argument. What matters is how the finished work contributes indirectly to cost reduction and increased output in the private sector.
The evidence that such effects exist is, alas, regrettably thin. Almost all of it rests on aggregative statistical relationships, essentially on the bare fact that average measured productivity growth declined during the same years that saw cutbacks in gross public investment. Almost none rests on detailed analysis of the contribution of particular projects to business efficiency.
And this is not surprising. While America might well benefit from more public investments on public-amenity grounds, export-oriented American manufacturing enterprise is evidently not hamstrung by infrastructure problems. Roads, rail, electricity, and water service are adequate to their needs. Boeing is not short of runways from which to launch its planes, nor is Silicon Valley suffering brownouts. Phones work well in this country--we even have the Internet! Pollution costs do not necessarily fall on private business producers, but on their neighbors.
Infrastructure and associated environmental spending is undoubtedly of enormous need and value. But to whom? To the American citizen, as an element in the standard of living. Roads, water, sewer, power, and communications systems are all durable public consumption goods. It is consumers and workers, not the main business shippers, who hit the potholes on the road to work. It is people who breathe the air, drink the water, and boat on the rivers and lakes. All this has little to do with international competitiveness--which is very sad, but true. This explains why business interests are not demanding higher infrastructure spending and why these items were the first to fail in the face of Republican opposition in the Congress.
W e are left with the unpleasant conclusion that the liberal mainstream has fallen into a self-deluding trap. The right has taken over the commanding heights of both fiscal and monetary policy, leaving liberals with token sums to spend on supply-side interventions. Education, training, and infrastructure are very important, but not for the reasons usually given. Business won't support funding them at levels that liberals desire, and it is wishful to argue to business that they should. We must find, instead, a language in which to defend them for the sake of the people themselves, and organize the people around them for the vital direct benefits they bring (as indeed the environmental, consumer protection, and health and safety movements have traditionally done). Otherwise they will continue to lose the budget battles.
And if we want full employment, we need something else--a full employment macroeconomics.
MACRO POLICY IN A STRUCTURALIST WORLD
Conservatives employ the myth of the market to oppose political solutions to distributive problems. But to leave things to the market is no less a political choice than any other.
Suppose the concept of an aggregate labor market and the associated metaphor of a natural rate of unemployment could be wiped away with a stroke from the professional consciousness (as it deserves to be). The policy notion that controlling the reduction of unemployment is the principal means of fighting inflation would lose its power. It would then become intellectually possible to revive the idea of giving a job to everybody who wants one. The issue becomes not how many jobs but rather who to employ and on what terms?
Investment and Consumption. Creating jobs is a matter of finding things for people to do. Investment of all kinds creates jobs, and stabilization of private investment demand is the traditional macroeconomic issue. Low and stable interest rates are essential here--more on that later. Public investment can step in where private investment will not go, and should be designed and pursued for its direct benefits, not its imaginary indirect ones. But consumption is also an important and much maligned policy objective. People should have the incomes they need to be well fed, housed, and clothed--and also to enjoy life. Public services can help: day care, education, public health, culture, and the arts all deserve far more support than they are getting.
Technology. Technological renewal should be understood as part of a strategy of maintaining investment demand. It makes sense progressively to shut down the back end of the capital stock, for environmental, safety, energy efficiency, and competitive reasons. Properly designed regulation can help, and this will open up investment opportunities for new technologies. At the same time, a flatter wage structure and bigger safety net, including retraining but also more generous early retirement for older displaced workers, would reduce the cost of job loss and the resistance from affected workers. Again, this is an adjunct of high-growth macro policy, not a substitute for it.
Inflation. Inflation policy would not go away. But the pursuit of relative price stability, rather than being the result of sluggish growth and tight money, would become concerned with the management of particular elements of cost, as the economy got closer to full employment. This includes wage pressures, and also materials prices, rent, and interest. Management of aggregate demand--an undoubted force on nonwage prices--could operate through channels with less effect on employment (a variable tax on excess profits, for example). Since wages are a major element in costs, inflation policy would be concerned with the institutional mechanisms of wage bargaining.
Distribution. This exercise returns us to the real, inevitably political questions obscured by technical mumbo jumbo about natural unemployment rates: our overall structure of incomes and opportunities. What should be the distribution of incomes? How much range, between the bottom and the top? Between capital and labor? Between skilled and not? In my view, the present course of rising inequality must be reversed, and liberals should frankly support the political steps required for this purpose. Trade unions should be strengthened and the aggressive new organizing campaigns of the AFL-CIO strongly supported. Minimum wages should be raised. And liberals should strongly defend the progressive income tax, as well as support proposals for wealth taxation, as proposed by Edward Wolff in this magazine. [See "How the Pie Is Sliced: America's Growing Concentration of Wealth," Summer 1995.]
Once the basic distribution of income has been set right, further gains in real wages can only happen, on average, at the rate of productivity growth. But to keep the distribution from getting worse again, these gains should be broadly distributed, substantially social and only slightly industrial or individual. In other words, we need to return to the principle of solidarity--that the whole society advances together.
Higher minimum wages are especially important for this purpose. In their new book, Myth and Measurement, David Card and Alan Krueger argue that raising the minimum wage within a reasonable range would not cost jobs. In fact, higher minimum wages may increase employment by reducing job turnover. This is a doubly important work, once for its direct policy relevance and again because it flatly contradicts, and deeply undercuts, standard models of the aggregate labor market.
Interest Rates. Low and stable has to be the watchword. Interest rates should lose their present macroeconomic function, which has been to guarantee stagnation. They should serve instead to arbitrate the distribution of income between debtors and creditors, financial capital and entrepreneurship. As a first approximation, real rates of return on short-term money should be zero. And there is no reason why long-term rates of interest in real terms should exceed the long-term real growth rate of the economy. Indeed they should lie below this value, effecting a gradual redistribution of wealth away from the creditor and toward the debtor class and a long-term stabilization of household and company balance sheets. Speculation in asset markets should be heavily taxed.
Deficits. Ironically, the budget deficit hardly comes up in this discussion. During the postwar boom, we were a high-employment, low-inflation, low-interest-rate society with a progressive tax structure. Such societies do not have structural-deficit problems. A peacetime military budget would also greatly help. At any rate, the present fixation on balancing the budget is nonsense, as all serious economists should loudly declare.
The above, all taken together, would be a macroeconomic policy to fight for! The liberal microeconomic supply-siders can do some useful things--or think they can--by getting a little money into education, training, infrastructure. But the point is to raise living standards, to increase security and leisure, and to provide jobs that are worth having. And that requires us to reclaim macroeconomics as a major policy tool.