George W. Bush
George W. Bush; drawing by David Levine

The Employment Act of 1946 committed the federal government to the pursuit of “maximum employment, production and purchasing power,” a commitment whose cash value has fluctuated drastically from time to time. It also established the Council of Economic Advisers (CEA), to be appointed by the president in order to provide him with objective economic analysis and advice contributing to the achievement of those goals.

In its early years the CEA stuck close to the central macroeconomic issues of “employment, production and purchasing power.” Inevitably, as the economy itself, and the federal government’s involvement in it, became more complicated, the CEA took on a much wider range of issues of economic analysis and policy.1 Today it functions as a sort of all-purpose economic consulting firm with only one client, the president. It is pretty small, as consulting firms go: the chairman and two members are still appointed by the president and confirmed by the Senate, and the professional staff numbers just twenty-one, including research assistants. The effectiveness of the CEA depends almost entirely on the extent to which each president wants to hear what the best analytical and empirical research has to say about the hot issues, and on the chemistry between the CEA chairman and his client, the chairman at present being R. Glenn Hubbard, who is on leave from the Columbia University economics department. Somehow I do not see George W. Bush as actually panting for those closely argued memos, but I hope I am wrong.

This year’s Economic Report runs to 300 pages of text, including many expository tables and graphs, plus fifteen or so pages of administrative detail, and roughly 125 pages of precious “Statistical Tables Relating to Income, Employment and Production,” most of them including data going back as far as the mid-1950s. They are not everything you always wanted to know about the US economy, but they are a lot of it.

The text itself has to walk a very thin and winding line. The three members and many of the senior staff are on temporary leave from university posts. In a few years they will go back to their departments, their teaching, and their research. They do not want to say dumb things in public, to look like fools to their students, their academic colleagues, and themselves. But they are working for the president now. No matter what they tell him in internal memos—we’ll never know—they have to follow the party line in the Report. Even if this or that favorite presidential idea or policy proposal is pointless or wrongheaded—and it has been known to happen—the Report has to take it seriously, in fact solemnly, and argue for it with whatever ingenuity can be scraped together. Nor can the CEA afford to be coyly noncommittal; the political hacks and ideologues in the White House are watching every word.

The 2002 Report has a few good moments and some bad moments, but mostly it has bland moments. To my eye it goes a few extra yards in the direction of the obsequiously political. (Is it really necessary to produce a sentence like the following one on page 64? “The President’s vision of economic security recognizes that many events impact the economy all the time.”) But this is only a jaundiced impression; the underlying problem is inescapable.


The Economic Report is published in February, just after the State of the Union message, and at about the same time as the president’s budget for the next fiscal year. The Report traditionally begins with a chapter giving the council’s and the administration’s take on the macroeconomic events of the year just past, and concludes with a forecast of the macroeconomic picture in the years to come, along with a plug for the president’s policy agenda.

The story told by this year’s Report is a defensible one. The economy is weak; it is obviously in the interest of the Bush administration that the onset of weakness be dated far enough back to be identified with the Clinton years. But the economy was in fact already slowing down before the beginning of 2001; that might possibly have been a good thing after a period of exceptionally fast growth. The particular form that the slowdown took, however, was not so favorable. Business investment in capital equipment and software had peaked in the third quarter of 2000 and started to fall. Industrial production had turned down a bit earlier. The CEA story is that it was still a toss-up on September 10, 2001, whether the US would fall into something to be labeled a recession, or was merely slowing down from an unsustainable pace.

According to this analysis, the attack of September 11 pushed the economy over the edge into recession by two routes: the direct disruption of business, especially air travel and tourism, and the indirect weakening of spending caused by the uncertainty and insecurity imposed on businesses and consumers in the aftermath.


This is one story. The National Bureau of Economic Research, which seems to care rather too much about the precise dating of successive recessions, has already given its opinion that a recession began as early as March 2001, which is when employment peaked. (The NBER is a private, nonprofit organization that publishes research by some six hundred university professors throughout the US; it has a small committee that dates business cycles.) To general surprise, the preliminary GDP data for the fourth quarter of 2001 showed the economy as steady (or actually up by a hair). That figure was later revised to show, even more surprisingly, that the economy actually grew at a 1.4 percent annual rate in the fourth quarter of last year. The NBER may be stuck with a recession that lasted only one quarter—although a recession is popularly defined as two successive quarters of negative growth. The early tea leaves for the first quarter of 2002 seem to suggest a continued gain, though slow and possibly reversible. Corporate profits and investment spending have not yet turned around, for instance; and most of the surge in consumption was clearly unsustainable bargain-grabbing by purchasers of automobiles. All this goes to show that to pose the question in terms of “recession” is not very informative, and may be misleading.

None of this was known to the CEA when it was writing the Report. What mattered then and matters now is that the economy is not yet advancing strongly, and for several reasons: businesses are conscious that they had overinvested in information technology, consumers have to adjust to the end of the stock market boom and the accumulation of debt, and potential buyers in the rest of the world are having economic troubles of their own.

When the talk turns to matters of policy, however, the CEA’s argument runs thin. The administration’s massive tax cut was enacted in June 2001 despite warnings that projections of large, continuing budget surpluses were very fragile. The tax cut was not planned with an incipient recession in mind. (The bill is called the Economic Growth and Tax Relief Reconciliation Act, or EGTRRA. If Congress ever passes a bill calling for the stoning of stray cats, it will be called the Economic Growth and Welcoming Our Furry Friends to Heaven Act.) It is true that the first small installments of tax reduction took effect in 2001 and must have helped to shore up weakening demand, especially consumption. (There is some evidence, however, that the small, one-time tax rebates, which were in any case pushed on the administration by the Democrats, have been mostly saved rather than spent; this is what economic research would have predicted anyway.) But the CEA would like to spin a similar anti-recessionary aura around the much larger coming installments of tax reduction, a kind of innocence by association with the first round of the tax cut, and that makes no sense at all. No one can possibly know which of the rate reductions scheduled to be phased in between now and 2006 will come at a useful time, and which will simply provide unwelcome aggregate demand.

With even less logic, after praising the scheduled income tax cuts, which are equivalent to an enormous reduction in national saving—the loss of revenues would add to the federal deficit or lower the surplus, and families will save only a small fraction of what they may get—the CEA goes on in the next breath to express satisfaction that the tax-cut package also included some minor incentives for private saving by the middle and upper classes. This general encomium is even extended to the reduction and eventual repeal of the estate tax, which is touted as a stimulus to the growth of entrepreneurial enterprises and the overall accumulation of wealth. These putative effects of eliminating the estate tax are irrelevant to the current situation, and dubious anyway. They certainly do not belong in a chapter on “Restoring Prosperity” because they are essentially redistributional in intent, favoring some parts of the population—those who are in the higher-income brackets and the owners of small businesses—and not others. This is a concrete application of the general principle that when they say that it’s not about the money, it’s about the money.

The need for an immediate “stimulus package” is certainly less, now that there are signs of an immediate upturn. But the short-run prospects remain tentative. Most forecasts, including the CEA’s own, predict that the economy will not make up lost ground until 2003. On the reasonable assumption that the country’s potential output has continued to grow at its long-run pace, there is now enough slack in the economy so that somewhat faster growth would not overstrain productive capacity. (The unemployment rate was 5.5 percent in February, and 72.8 percent of industrial capacity was being used, 11 points below its 1997 peak.) It would have been better if Congress had passed an acceptable stimulus package in October; it would be a reasonable thing to do even now, as the Report argues.


Of course the CEA is stuck with the Bush proposals. One of them is to advance the dates of tax reductions already scheduled under EGTRRA. For those, like me, who think those tax reductions were a fundamentally bad idea, bringing them into effect earlier is not a step in the right direction. The Report goes on to endorse the stimulus package passed by the House, as the President has done.

The original House package was an intellectual disgrace. You would ask three things of a short-run macroeconomic stimulus: it should actually stimulate private and/or public spending, it should be temporary in its effects, and it should not advance a partisan agenda. The original House bill failed abjectly on all three counts. For example, it featured permanent giveaways to large corporations, through repeal of the alternative minimum tax and rebating of past payments, with absolutely minimal incentive effects for new investment. (This was one more example of redistribution masquerading as constructive policy.) The bill that has now passed both houses and been signed by the President is worth a passing grade—although it was handed in late—but could have been better. Accelerating depreciation allowances is not the most effective way to revive investment spending; and the three-year duration is too long.

Many economists had come reluctantly to the view that, although fiscal policy is potentially a powerful tool for expanding a weak economy and contracting an overheated one, our political process is incapable of using it in a timely, constructive way; so monetary policy in the form of adjustments in interest rates and money supply is all we have for the short run. This episode unfortunately confirms that pessimism.

After this hand-wringing, I am glad to be able to say that the Report’s section on the short-run and long-run economic outlook is professional and reasonable. The CEA expects the economy to grow weakly this year, to make up, between 2003 and 2005, the slack that has accumulated, and then to settle down to something like steady growth. Of course they don’t really expect perpetual calm to prevail; this kind of smoothing away impenetrable uncertainty is conventional. Along such a path, the unemployment rate will rise above 6 percent this year, and then gradually—maybe a bit too gradually for comfort—recede to a steady 4.9 percent, compatible with stable, low inflation. This last estimate, as a bit of macroeconomic theory and practice, is nothing to be proud of, but CEA is well within the professional consensus.

For the long run, the CEA projects a “steady-state growth rate”—i.e., all parts of the economy growing in proportion—of 3.1 percent a year: employment growing at one percent annually, and productivity at 2.1 percent. I think that is entirely reasonable. The CEA does not fall for any leftover new-economy hype, if there is any left over, but they anticipate better productivity performance than the just over one percent a year that characterized the quarter-century after 1970. The Report does not mention military spending at all, perhaps because its basic concerns are too short-run for that. This might be an important omission even for the next few years.


The next chapter of the Economic Report is devoted to “retirement security.” Notice that the topic is not “Social Security”; it is precisely the administration’s goal to take as much of the “social” out of “Social Security” as it can. This was expected to be the main economic issue of the Bush years, until recession and terrorism and Enron supervened. It will return to the stage, because there are issues to be settled and, for some of them, the sooner the better. This would have been a good time for the CEA to provide readers with a little education about the fundamentals before making its pitch for Mr. Bush’s pet nostrum, individual retirement accounts, individually owned and individually invested. The CEA makes a weak pass at such education, but lapses too soon into cheerleading.

Barring an unlikely torrent of immigration, demographic projections tell us that in the years ahead there will be relatively more retirees and relatively fewer workers than there are now. In any particular future year, whatever the “retirement security” system in place, the US can consume only what it produces. So the national standard of living depends entirely on the productivity of the working population. The only way to make up for a larger number of dependents per worker is higher productivity, and the route to higher productivity in the future is more saving and investment, more technological progress, and more education and training now.

The retirement security system has traditionally had very little to do with the size of the national pie. Instead the system has determined how the pie gets divided among the working population and the retired population. This simple statement, however, now has to be amended: the rules of the retirement system are part of the set of incentives to work, and to save, that influence individual behavior. It seems inevitable and natural that some gradual lengthening of the normal working life will be part of the response to a rising ratio of dependents to workers, and the retirement system will have to accommodate this. The CEA is also full of enthusiasm for stimulating personal saving, and it touts this as an advantage of individual accounts. There is something disingenuous about an administration that has just presided—in EGTRRA—over an enormous dissipation of potential national saving now promoting a significant change in Social Security because it will have an unpredictable and probably minor favorable effect on personal saving.

It is hard to avoid the feeling that the primary attraction of the individual account system is ideological. Here is what the CEA says:

Under such a system, a worker could direct a portion of his or her payroll taxes, or possibly an additional voluntary contribution, into a personal account that he or she would legally own. The worker would then choose, from a variety of options, how the assets in the account are to be invested. Upon retirement, the worker would have access to the accumulated assets, which could be used to purchase an annuity, provide a bequest to heirs, or make withdrawals from as needed. Workers who choose to direct a portion of their existing payroll taxes into private accounts could expect a higher combined level of benefits, because an annuity funded by the personal accounts would have a higher expected value than the benefits from the traditional system that are being partially replaced by the account contributions. Personal accounts would thus represent a voluntary means by which a worker could supplement benefits from the pay-as-you- go portion of Social Security. As such, they could provide the foundation for a return to individual- based retirement security that takes advantage of the strengths of individual choice and wealth accumulation.

The Report does not confront any of the inevitable problems that would accompany a scheme of the sort described. It does not mention that any plan that diverts part of the payroll tax into individual accounts will leave the “traditional” benefits grossly underfinanced. They will have to be reduced accordingly. The CEA does not give details, but the usual device is to include a “recapture” provision that reduces the workers’ traditional benefit by something between 75 and 100 percent of the annuity value of his or her individual account. In the end, therefore, the worker’s total benefit payment will exceed the guaranteed benefit by no more than 25 percent of the monthly benefit financed by the individual account. So the combined benefit would increase by very little unless the stock market booms forever or the worker is an inspired stock-picker. Some such recapture is necessary, because the benefits under current law cannot possibly be financed if part of the payroll tax is diverted into individual accounts. According to the formula, a worker who chose not to have an individual account would receive exactly the original traditional benefit.

Nor does the Report have anything of substance to say about the large transaction costs that would go along with the (monthly?) processing of a very large number of very small accounts controlled by an equally large number of generally inexperienced investors. One recourse would be to limit workers to a very small number of mutual funds. But then it would seem much more efficient to have the investing done by the Social Security administration itself, at very low cost. Fears that this would imply government intervention into the sphere of private industry seem artificial; it would not be difficult to have the process managed by a highly visible, highly respectable board, insulated from political pressure.

Finally and probably most important is the size of the transition costs. If current workers can allocate a part of their payroll taxes to individual investment accounts, those tax revenues will not be available—as they are now—to pay benefits to current retirees and to those who will retire in the near future. The immediate financial position of the Social Security Trust Fund would necessarily be undermined, and this will continue for a long time, until most of the beneficiaries will have spent their whole working lives under the new plan.

The cumulative transition cost from the current system to the new one—the accumulated difference between the change in revenues and the change in expenditures—is not minor. In one reasonable calculation, it reaches $1.4 trillion in 2030 and then starts to fall. But that is without including the interest costs of government having to finance the gap between revenues and expenditures. If that cost is included, as most economists would think right, the transition cost would reach $9.8 trillion in 2074 and would not yet have turned downward.

I am quoting here from some calculations by Lawrence H. Thompson and John C. Wilkin involving a diversion of 2 percent of the current 12.4 percent payroll tax.2 The outcome of course depends on assumptions made about the recapture rate, the administrative costs under decentralized or centralized management, and especially on how well the stock market does. Thompson and Wilkin try a number of alternatives. Here is their conclusion:

Financial projections suggest that diverting 2 percentage points of the current Social Security tax into individual accounts is likely to worsen the program’s long-range deficit. Avoiding this result would require (1) future equity returns at least equal to those in the past, (2) investment of at least half the account balances in equities, (3) centralized administration and passive policies to assure low administrative costs, and (4) recapturing [i.e., taxing away] at least 75 percent of the account balances at retirement.

We know that equity investments are intrinsically risky. It probably makes sense, on the basis of considerations of risk and return, that part of Social Security revenues should be invested in equities, though perhaps not as much as half. But why would we want to transfer those risks to individuals and families of limited wealth and limited experience? Margaret Thatcher is said once to have announced that “there is no such thing as Society.” An alternative view is that Society functions as a universal insurance scheme that allows its members to pool part of their assets and thus partially to stabilize their incomes. This is a kind of insurance that the private market cannot produce. The Economic Report faces resolutely the other way. It is a choice, but a purely ideological choice.


The next chapter of the Report—and the last one I will discuss here—is written in an altogether different key. The subject is “Realizing the Gains from Competition,” and it is about the regulation of mergers, takeovers, joint ventures between rivals, market dominance, and other possibly anti-competitive behavior. The tone is rather above-the-fray, a bit like a textbook without the technical details, full of what “studies” show and what the lessons from “scholarship” are. The overriding theme is that the modern economics of industrial organization has revealed the crudity of many earlier beliefs about the dangers of limiting competition, and the inappropriateness of the regulatory practices derived from those beliefs.

So far as I can tell, the exposition in this chapter is accurate. It is not my home territory, so I cannot say if there are other “studies” or if the “scholarship” may be controversial. But it hangs together. This is an important field of economic research and economic policy. Competition is what makes capitalism acceptable. Monopoly tends to be not only inefficient and inequitable, but also lazy. There is plenty of evidence that competition is what drives both organizational and technological innovation, and therefore underlies the trend of rising productivity that is supposed to lift all the boats.

The Report rightly says that the delicate problem for regulation is to preserve competition without stifling the incentive to innovate. The problem arises because the successful innovator gets at least temporary market advantage and therefore at least temporary market power. Take it away too quickly and the costs and risks of innovation will not be worth bearing; take it away too slowly and the driving force of competition evaporates. Anyone could agree to that; but there is lots of room to argue about where to draw the line.

It will come as no surprise that the Report, without ever saying so directly, tends almost always to suggest that the government would do better to leave things alone. There is a charming example of this on page 130, where it is remarked that economic analysis has enhanced the capacity of government agencies

…to distinguish those cases that properly raise concerns about anticompetitive effects from those that might have raised concerns in the past, but should no longer, in light of a better understanding of competitive forces. These changes in antitrust policy are important in that they afford firms greater flexibility to lower their costs and improve their products….

One does not doubt that; but one wonders if economic analysis has uncovered no cases of behavior that was once thought to be harmless but that can be shown, with more subtle understanding, to harbor anticompetitive effects.

There is another example on pages 134–135. The text gives a clear exposition of “network effects” that make a product more valuable to users precisely because it is already used by many others. “For instance,” the Report comments, “over the past decade the number of people using e-mail has grown dramatically, making it a much more valuable means of communication for any individual user today than it was a decade ago.” Strong network effects are sometimes thought to confer the basis for monopoly power: the first seller in the field may acquire an impregnable advantage. That is possible, says the Report, but not inevitable. Who could disagree? But the Report then goes on to say that there have been instances where a network advantage has been overcome, but no “conclusive evidence that network effects have prevented the widespread adoption of a markedly superior product.” That puts an absurdly difficult burden of proof on one side of the argument. How would one identify a “markedly superior” rival that ran into early difficulty trying to overcome a competitor’s network advantage, or gave up without trying?

The chapter manages to leave the impression that monopoly power, predatory pricing, barriers to entry, and all that are peripheral or even negligible dust spots on an otherwise perfect surface. Maybe so; and the other extreme, which would propose severe and widespread regulation, makes an even worse case. One would feel better if there were evidence that the president’s economic advisers had considered that question to be a question.

The four remaining chapters cover the economics of health care, issues of “fiscal federalism” (i.e. the division of functions among federal, state, and local governments), the environment, and aspects of international trade and capital flows. I could not possibly take up such a broad range of topics here; a modern elementary economics textbook can run to eight hundred pages. But the flavor of those chapters is the same plain vanilla as the first three.

This Issue

April 11, 2002