Bill Clinton
Bill Clinton; drawing by David Levine


It really was “the economy, stupid” that elected Mr. Clinton to the presidency. He also emphasized a parallel and important theme: “change.” Its potential significance is suggested by the fact that even Mr. Bush, that apostle of domestic passivity, tried to represent himself as the candidate of “change.”

The content of “change” was never made clear by anyone. If it means anything at all—which is not to be taken for granted—it must presumably mean more than just making things better: increasing employment, restraining inflation, raising SAT scores. “Change” sounds more qualitative than that. There were occasional pseudo-clarifications, like the mention of “structural” change. But “structural” often appears to be a word used to communicate that one is talking about something deep, but doesn’t actually know what it is.

Alice Rivlin is much more concrete. In Reviving the American Dream she proposes a major change in the way government relates to the economy. She would like to reshuffle the specific responsibilities of the federal government and the state governments, with the states (and cities) taking over a considerably larger share of economic policy than they have ever done since the Great Depression of the 1930s. It is a standard conservative ploy to say that the states should do more because they are closer to the people, while at the same time failing to suggest where the states are to get the financial and intellectual wherewithal to carry out their greater responsibilities. That is not an oversight: the conservative goal is not to redistribute the economic functions of government but to diminish or eliminate them. Dr. Rivlin is not in that game at all. She means what she says, and the fact that the idea comes from her lends plausibility to it in a special way.

Dr. Rivlin is President Clinton’s appointee as deputy director of the Office of Management and Budget. The director is Leon Panetta, an experienced California congressman, wise in the ways of making and negotiating the budget. There is no reason to think that he or Mr. Clinton would be willing to accept the “change” advocated skillfully by Rivlin, although it is hard to imagine that they did not know about the ideas contained in their colleague’s book. It speaks well for Mr. Clinton that he is willing to have, at the analytical heart of a centrally important agency like OMB, an articulate believer in a way of organizing the economy that may not be his own.

What makes Rivlin’s advocacy of devolution to the states and cities noteworthy is the fact that she has worked in Washington throughout her professional career. It is an eminent career—she has been president of the American Economic Association—but carried on essentially outside the universities. Her long-term association has been with the Brookings Institution, the preeminent Washington think tank, where the elite meet to eat and to write their books.

Then, when the Congressional Budget Office was established, she was its first director. In discussions and negotiations on policy, whoever has done the research and written the memo has a tactical advantage. The CBO was created so that Congress could have intellectual parity with the Executive Branch in fiscal policy. Otherwise the Executive, backed up by the Council of Economic Advisers and the OMB, is always a step ahead. Under Rivlin’s leadership, CBO was a quick success. Congress is no longer in the position, for example, of being forced to accept the president’s self-serving revenue forecasts for lack of an alternative. The directorship of CBO was intended to be an essentially nonpartisan position, serving the whole Congress. Nevertheless, at the beginning of the Reagan administration, the Republican right tried to drive Rivlin out. She dug in and remained in office long enough to establish the point that CBO is an analytical agency, the same for everyone, and not a purveyor of cooked research. (She was succeeded by Rudolph Penner, whose leanings are more conservative, no doubt, but who is also an excellent, careful economist. Rivlin had made her point. The tradition continues.)

When an eminent close-up student of economic policy, with Rivlin’s history and credentials, concludes that the federal government is in over its head, the rest of us have to take notice. We do not have to agree in the end, but the traditional presumptions cannot be taken as self-evident.

Rivlin’s diagnosis of the economic problem facing the US is squarely in the mainstream of current analysis. The most important fact of American economic life is that the real income of the average family, which had been rising at about 3 percent a year in the 1950s and 1960s, fast enough to double every twenty-five years, slowed to a crawl (or worse) about 1973. For the past twenty years, growth has been negligible. I reproduce one of her graphs because it tells the story so clearly.



It takes only a little imagination to see that a society with that history might turn mean and crabbed, limited in what it can do, worried about the future.

The story of the average family hides important diversity. During the period of slow growth or non-growth the inequality of incomes has been widening. The families at the bottom (not the same families every year, of course) have been actually getting worse off. Their incomes, corrected for inflation, have been falling. Families in the middle 60 percent have been scraping along, sharing a little in what growth there has been, but only a little. The families at the top (again not always the same people) have been doing all right; their incomes have been rising quite nicely.* If one characteristic of a good society is that the degree of inequality should not be too wide and should not be getting wider, then the years since 1977 or so have not been good for American society.

These facts about the growth and distribution of family income in the US are reflections of something going on at a deeper level. The average income in the US is governed in the long run mainly by the level of productivity, i.e., by the capacity of the economic apparatus to generate output per hour worked and by the number of hours worked per person in the population. The growth of family income slowed in the 1970s and 1980s because the growth of productivity slowed. A revival of income growth will require a revival of productivity growth. (It is true that there has been a partial improvement in productivity in the past two years. Some of it, however, comes from shedding obsolete capacity, and there is no future in that. It is too soon to know whether the remainder is more than transitory.)

The productivity story has been studied to within an inch of its life. (The slowdown is visible in Europe and Japan too, by the way; it is not strictly an American disease.) Rivlin gives a brief but clear summary of the outcome of all that research. There is no single explanation. A number of minor causal factors can be identified. When all that is done, there remains an unexplained and perhaps unexplainable puzzle. No law of nature says that the trend in productivity growth should always be the same. Some periods are better than others.

Changes in the degree of inequality can also be traced back to events in the sphere of production. The widening of inequality is clearly visible in incomes before tax and without income support payments. The Reagan-Bush achievement in taking from the poor and giving to the rich made things only slightly worse. If that policy had resulted in a resurgence of growth, the net result might have been adjudged beneficial. But the supply-side revolution was a dud. High-income earners appear to have done well because the demand for skilled, educated labor kept going up. The bottom of the heap did badly because it consists mainly of those with no skills or very limited skills. The market for those workers is shrinking even in service industries, and they are increasingly in competition with poor countries with wage levels much lower than our own.

It hardly matters whether we can identify all of the causes of the productivity slowdown and the stagnation of incomes. Everybody knows roughly what the solutions are (although it is a lot harder to prescribe just what form they should take and estimate exactly how effective they will be). Any action that incurs costs now in order to produce rewards later is, generically, an investment. A country that wants to have higher incomes in the future than it has now will have to invest more. The list of possible productive investments is fairly diverse. Spending on industrial equipment, the latest in tooling and machinery, certainly qualifies as productive investment. But so does spending on technological research and development and on the year-to-year improvement of existing goods and processes. So does spending on the country’s transportation and communication networks, water supply, and waste disposal systems, usually classified as infrastructure. And so does spending on the creation and maintenance of an educated, skilled, adaptable labor force.

This last category of investment has something else going for it: the chance that it could reverse the worsening inequality of incomes. If the disappearing demand for unskilled labor is the main force behind the deterioration of the status of the lowest income groups, then they would gain disproportionately from a serious and successful campaign to improve the quality of the labor force. Investment in education and training that was directed at the lower-income groups might both raise the average of earned incomes and reduce the gaps between the rich and poor.


These items of investment comprise what Rivlin calls “the productivity agenda,” and getting it carried out efficiently is, for her, the main goal of economic policy. The first thing to note is that investment requires savings: costs are incurred and there is no immediate output to cover them. On the national scale, the costs of an investment program have to be covered by domestic saving, either by the private sector or by governments, or else borrowed from abroad, as the US has been doing since 1981. We do not, as a nation, save enough to cover our current investment, let alone to finance a desirable increase. Why is that?

The superficial answer is that private saving in the US more or less collapsed and the federal deficit (which amounts to negative saving) increased substantially. Only the second of these can be blamed on the fiscal policy of the Reagan administration. The superficial answer is correct as far as it goes. It does not tell us why private saving tell from 9.0 percent of national income in 1974–1980 to 6.9 percent in 1981–1985 and 4.9 percent in 1986–1990. Nor does it quite explain Ronald Reagan’s policies, for that matter.

There is not much that public policy can do to induce families and businesses to save more (except to keep them employed and profitable). Of course incentives to save can be provided through the tax system. But the devices that seem worth considering, i.e., the ones that are not disruptive in some other way, are not known to be very effective. That leaves reduction of the federal deficit. When the US Treasury needs to sell fewer bonds in the market, more private funds are available to buy the private securities that finance productive investment. Deficit reduction is part of Rivlin’s agenda too. She does not make the common error of treating the budget deficit as some kind of evil-in-itself, like measles. Deficit reduction is desirable when and only when the level of federal borrowing is the effective limit to private investment.

Rivlin’s pages of diagnosis and description are exemplary: clear, documented, balanced. She describes herself as “a fanatical, card-carrying middle-of-the-roader” and that is accurate. In matters like this the middle-of-the-roader is usually right. The working of a large, complex, mixed economy is bound to be complicated, uncertain, and, in part, mysterious. A willingness to canvass a variety of quite different diagnoses and solutions will usually pay off in avoiding the worst errors. I would recommend these chapters to anyone who wants to get a feel for the complexity of the policy problems. She writes, for example.

Schools are only part of the story. Compared with Europe and Japan, the United States has paid little attention to the transition from school to work. Students not going on to college typically flounder around in the job market for several years without much idea of how to get a permanent job, what skills they need, or how to get them. Moreover, in a period of change, even those with skills may find themselves needing training for a new kind of career in midlife. A serious effort to increase productivity growth will require making retraining and additional education a normal part of working careers.

Reducing poverty and improving opportunities for those at the bottom of the income distribution is not just a matter of improving schools and training. It also requires offering better child care and health services and improving incentives to get off welfare. A major contribution could be made by reforming health financing to make health insurance available to low-income workers. The present system often penalizes welfare mothers who seek jobs by depriving them of medicaid coverage.

But there is nothing new or startling in Rivlin’s view of the nature of the problems and the general character of the solutions. It is what the clan believes, with good reason. The novelty in her analysis comes at the next step.

Her argument is that our federal system has evolved over the decades, mostly for good reasons, to a condition in which the federal policy apparatus is overloaded. It is doing things, for example in its policies to improve education and transportation, that it cannot hope to do well because local conditions differ and the differences are important for the effective planning and execution of policy. The federal government does so many things that sound policies are often discredited for the wrong reason. Moreover, because policies with specific consequences are financed by general taxes, especially the federal personal income tax, the public has the feeling that it does not know where its money goes. People lose touch with the concept that their tax payments provide public services because the connections are so indirect. This disjunction contributes to general resistance to taxation, not offset by any diminished attachment to public services. It is no wonder, then, that politicians talk about deficit reduction and do nothing.

Finally Rivlin makes a different sort of point. She argues that the continuing globalization of economic activity adds a whole new set of demands on the economic policy apparatus. The individuals and agencies dealing with economic policy now have to worry about the international implications of what they do and the implications for what they do of events and actions abroad. The job of conducting policy has become harder and consumes more time and effort. These are not concerns that the federal government can afford to neglect; they are part of a position of leadership in the world. If the federal government was not overloaded before, it is now.

Well, maybe so, but I am not convinced by this last point. The world is full of open economies. For decades they have had to worry about the international context of economic policy. Their civil servants and officials have managed to carry on. It is true that none of them, probably not even Britain in its heyday, has been quite the 900-pound gorilla that the US is, and therefore none has had to make all the calculations that a wellmeaning 900-pound gorilla has to make before it shifts its weight. Still, if the current balance of the federal-state system were about right, I doubt that having to think about exchange rates and adjusting tax rates to harmonize with those of other countries would make a big difference. This point is not essential to Rivlin’s argument anyway. She thinks the system is functioning badly even at home.

For all of these reasons, her argument continues, it would be a good idea for the federal government to shift some of its current policy responsibilities elsewhere. That can only mean to the states (and in some cases to city governments). Fortunately, according to Rivlin, the states now have the capacity to perform functions for which they once simply lacked the competence. Nearly all governors now serve four-year terms and most are eligible to succeed themselves. So able and ambitious politicians who would once have found state government unrewarding and aimed instead for federal office now become governors. They stay in office long enough and they are ambitious enough to build competent staffs. Similarly state legislators used to be part-time public officials, in session only for a fraction of the year, and tending a business or a law practice most of the time. Nowadays more and more legislatures stay in session longer, pay their members better—though probably not enough—and have developed real, specialized competence. So the devolution of economic-policy functions from Washington to the states is a genuine option.

Once again, maybe so. Mr. Clinton has surely demonstrated that even a small, poor state can produce an intelligent, energetic, and effective governor, and there are a couple more in his cabinet. It is easy to name others, past and present. But the same casual empiricism turns up the governor of Louisiana on the other side, and I imagine there are still others who are not even mediocre. I am even less confident about state legislatures. Some are excellent. My own, in Massachusetts, offers little support for Rivlin’s judgment. Maybe, if state governments had more interesting and challenging things to do, they would attract broader-gauged people. But that is a lottery, not a proved case. Leave that question aside. What division of labor between the federal government and the states does Dr. Rivlin have in mind?

She thinks that a large part of government action to improve productivity belongs at the state level, that planning and operations should be state responsibilities. Education and training programs, if they are to contribute to the revival of productivity growth, can succeed only if they are adapted to local conditions, the needs of the local employers, the desires of local citizens, the traditions of the local culture. Useful vocational education requires a degree of cooperation between schools and employers that cannot be achieved on a national scale. It may not be achievable on any scale, given our tradition of hostile relations between business and government, but the chances of success are greater if local needs are reflected in the planning process. A larger national investment in the quality of the working population, even if it comes as part of a nationally expressed commitment, will need to be realized and fine-tuned by states and cities.

Much the same case can be made for productive infrastructure. Most transportation needs tend to be local and regional. The long-distance road network already exists. In fact, it is repair, maintenance, and, above all, rational pricing that are needed rather than a lot of new construction. More than one state may be involved in a single project, but very rarely a large number. Water treatment and waste disposal are even more local matters. So here too the case for devolution of control makes sense. And here especially, but also in the education-training field, the clear association of particular tax revenues with the delivery of particular services is possible, and can be expected to increase popular support for the payment of the necessary taxes.

This leaves a fairly small part of the “productivity agenda” to the federal government. Rivlin notes correctly that public subsidies for industrial research and development cannot be left to the states because the benefits cannot be reserved to the state that happens to pay the bill. These spillovers, and the fact that the business firms themselves work in several states, virtually dictate that R&D incentives must be a federal undertaking.

Rivlin does not mention the case of public subsidies for investment in business equipment through an Investment Tax Credit, for instance. The argument is rather different in this case. Each state has an interest in inducing industry to locate new capacity within its borders and to renew and improve the capacity that is already there. So states can be expected to compete for the new jobs and higher wages that will accompany the investment they can attract. If interstate competition takes the form of improved infrastructure, superior labor quality, and other forms of economic development, no problem arises. That is exactly what Rivlin hopes will happen. Competition in the form of tax concessions, on the other hand, does nothing to advance the productivity agenda. In fact it is more likely to be self-defeating in the aggregate because the loss of revenue to the states will eventually undermine their capacity to finance the essential activities like education, training, transportation and other services that are supposed to contribute to the revival of productivity.

The federal government would not be out of the economic development business in the Rivlin scheme of things, but there is no doubt that the balance would shift dramatically. Federal economic policy would center around reform of the health care delivery system and its costs, the promotion of research and development, investment incentives, the support of higher education, and the few fields like air traffic control and some environmental protection activities where the importance of cross-border effects would make decisions by individual states inappropriate. So here too the federal government is the right instrument.

The key question arises at once: How will the states finance their new and expensive functions? Unlike many other writers on this subject, Rivlin faces the issue, although I doubt that she has solved it in a satisfactory way.

First of all, she would have the federal government abandon certain functions and shift the accompanying costs to the states. The federal budget thus moves closer to balance. Rivlin goes further and proposes that the proper target for the federal budget, including Social Security, is a surplus. Then the federal government would be a net saver, a contributor to the financing of private investment, no longer a drain on private saving. I would agree that this is desirable.

On the other hand, Rivlin assigns to the federal government the responsibility for the universal extension of health insurance and the design of a system of cost controls. These, especially the first, will incur costs of their own. Here Rivlin invokes a principle that sounds good and ought to be valid: the voting public will accept taxes that can be seen to pay for specific and valued benefits. Universal health-care insurance, for example, should be paid for by an appropriate, earmarked tax; and the net budgetary effect of the devolution of activities to the states should be a drastic improvement in the balance of the federal budget. But I think she is excessively optimistic about the willingness of voters to tax themselves, even for visible services.

How would the states cover the costs of the new functions that have been shifted to them? The same principle applies. Since the benefits will be local—that is why they are assigned to the states—they will be clearly perceived, vividly appreciated, and therefore presumably willingly paid for. The states, in other words, are expected to raise the revenues needed to pay for their newly acquired activities. Rivlin does not claim to have discovered the free lunch. She has redistributed government functions, not eliminated them. All this plus keeping the federal budget in balance must mean higher tax revenues, somewhere. Her argument is that the states will be better able to justify the use of such revenues and therefore more willing and better able to raise them.

This is not necessarily an overoptimistic view. The record shows that voters are more willing to pay for the schools their children attend than they are for abstract educational expenditures financed out of an equally abstract share of the federal income tax. But I am not sure it follows that they will be equally willing to contribute state income or sales taxes, say, to pay for a vocational education program whose direct beneficiaries are not most of the taxpayers themselves, but only a segment of the population and mostly a non-taxpaying segment at that. Both the instinct to get a free ride on the tax payments of others and suspicions of people who do so are probably as strong as ever. They can only be exacerbated by class and other divisions within the voting public.

So I am skeptical of Rivlin’s more or less explicit conviction that, in effect, the federal budget deficit can be eliminated by tax increases at the state level. Even waiving that objection for the moment, there is yet another problem. States differ widely in their capacity to pay taxes. Personal income per head in Illinois is a third higher than in Alabama. (See table opposite page.)


If each state taxed itself at the same rate to carry out the Rivlin program, and if the productivity agenda actually worked, the result would only be that Illinois would pull further ahead of Alabama. That is not an acceptable consequence of national policy. The current system of federal grants-in-aid to states circumvents this problem, at least partially. Federal tax revenues come disproportionately from the residents of the rich states; they earn more, so they pay more. If the proceeds are then redistributed back to the states on a per capita basis, as is currently the case for some programs, the result will be a transfer from richer states to poorer. This is what used to happen when federal grants-in-aid to states were larger than they are now.

Rivlin proposes that the states should finance their newly acquired responsibilities not separately, but together. They should somehow levy common taxes—on a regional or national scale—and share the revenues. If the taxes were clearly earmarked for specific purposes, she writes, they would be paid willingly. If the revenues were shared on a per capita basis, they would have the desired redistributive effect. She would have the states re-create federal revenuesharing on their own, the crucial difference being the closer connection between taxes and benefits.

This really does strain credulity. It is hard for me to imagine one of those able and ambitious governors asking voters to buy into a few pennies of sales tax to be spent on an apprenticeship program, say, of which a fraction would be spent on apprenticeship programs in other states. The fraction is not trivial. A state whose per capita retail sales were 25 percent above the national average would be giving away 20 percent of its sales-tax collections. That seems politically risky. Maybe it is just the extra remoteness of the federal government that permits this sort of redistribution to take place. It might go more smoothly, then, if the Rivlin program were financed by a national sales tax with the revenues recycled back to the states to cover the productivity agenda. Individual states could compete by piggybacking extra sums on the national tax in order to improve the local economy. This device might get around the states’ unwillingness to see the federal government poach on their traditional fiscal territory, the retail sales tax. It might not get around the general hostility to centralized taxation.

If anyone is really in search of “change,” such a program suggests an interesting direction for the middle of the road to take. It is different enough from what we have now so that it might actually work, but not so different as to be a leap in the dark. The foundation of Rivlin’s vision rests on two propositions about our political world that need to be seriously discussed. The first is that most state governments are now prepared to take on and carry out additional responsibilities without the parochialism, racism, and corruption that justified the growth of federal power in the first place. There is no general theorem that covers all cases. It is a matter of counting and classifying, and perhaps praying.

The second proposition is the claim that the American voter is willing to vote and pay taxes if only the money can be seen to be going into objects and functions that matter concretely to the people concerned. Please let it be so, I am inclined to say. History does not justify much confidence, however. Most people say they would be willing to be taxed to reduce the deficit, for instance, only not this tax, not that tax, in fact not any particular tax you can name. Maybe that tendency reflects the abstractness of deficit-reduction as a goal, and the remoteness of the federal government Maybe.

There is one further gap in Alice Rivlin’s argument that needs more discussion. She says at the beginning of her book that she is going to concentrate on the medium-to-long-run problems of the economy and ignore the policy issues connected with smoothing out the business cycle, and that is what she does. The problems of maintaining high employment, restraining inflation, and stabilizing the economy are never mentioned.

That may be a reasonable intellectual strategy. In the end, however, when all the reorganization of policy responsibilities has been done, one is entitled to ask: Who will look after maintaining high employment, restraining inflation, and stabilizing economy-wide fluctuations? The states cannot do it. The federal policy apparatus will have been reduced in size and scope. Well, there is the Federal Reserve. Indeed, monetary policy is generally the first line of defense in the short run anyway. But monetary policy cannot do it all. The very openness to the world economy that Rivlin mentions elsewhere also preoccupies the Federal Reserve and limits its freedom of action. Dr. Rivlin takes office at OMB at a moment when a new fiscal policy is clearly needed and needed for the short run. Suppose her reform was in place. Would a fanatical, card-carrying middle-of-the-roader contemplate varying the rate of the new national sales tax so as to counter changes in the business cycle? Explicitly temporary changes in sales taxes can have a powerful effect on the timing of private expenditures. Then at least Washington would still have something of central importance to do.


The outline of the Clinton economic program has become clearer since his State of the Union speech and, with more detail, in a backup document called A Vision of Change—that word again—for America. Alice Rivlin was surely a major participant in the process that shaped the program. There is no trace of her emphasis on the reform of American federalism, but that is neither here nor there. The program mainly concerns taxes and expenditures in the coming fiscal year, not governance. The diagnosis that underlies the program and the general tendency of the individual measures that have been cobbled together are very like her own, and very like those recommended by mainstream economists. By the way, I have to report a feeling of occupational pride in the quality of A Vision of Change for America: the tone is analytical, the argument is logical, the claims are realistic. Clinton has assembled a strong team of economists. We will have to see if he continues to listen to them, but at least he lets them write good stuff.

The good news is that the program is oriented to growth. Even the initial short-term stimulus package is justified by the principle that a strong recovery from the dreary recession is a necessary precondition for a surge in private investment. Not even tax incentives will induce business to build new capacity when it cannot find a use for what it already has. The stimulus of about $30 billion is modest—maybe too modest—but it provides some insurance against an anemic recovery. The notion that this is a shortsighted, expansion-happy administration is absurd.

Most of the concrete proposals come under the heading “Investing in the Future.” Even the items whose purpose is to reduce the budget deficit are handled that way, as they ought to be. President Clinton evidently understands that government deficits absorb and dissipate private saving when the economy is not in recession; and he wants to make it clear that cutting the deficit following a recession will free private savings for investment. One has to hope that he can teach as well as learn.

The specific expenditure items are a mixed bag. Many of them make excellent sense: improving transportation without extensive road-building, aiming at new technology and efficient utilization in the energy field, providing incentives for business investment in technology and new capital, and getting the federal government modestly involved in the promotion of civilian technology. Others fit less well. Improved housing for low-income families is a cause worth supporting for its own sake, and I do support it, but it will not add a cubit to productivity. That is not very important. The program was put together in a hurry. If it gathers steam, as I hope it will, the anomalies will get sorted out.

A more serious weakness is that A Vision of Change for America is skimpy on investment in “human capital.” There is not yet a concerted plan for upgrading the American labor force through a serious overhaul of education and vocational training. What the program mentions is generally good: assured funding for Head Start, a small youth apprenticeship initiative, and a few other things. But much more is called for in view of the accumulating evidence that those without up-to-date skills are doomed to earn less and work less. Maybe such an effort is in the works. It should not wait for long as inequality worsens and work habits weaken.

Here the trail comes back to Rivlin’s ideas. The federal involvement in education and training is traditionally limited in our system, in fact it is mainly limited to providing money. A serious attack on deficiencies in education and training will have to be a joint federalstate enterprise. The states must necessarily do the detailed planning and administration. They can also do what the federal government finds difficult: experiment with alternative systems of school organization and teaching. That is all the more reason to get started.

The Clinton program has consistency, direction, and some good ideas. The next step is to see if it falls victim to the usual excuses: balance the budget but don’t tax me; we demand excellence in research but every Senator must get his share; anything the government does, even if it is good, is no good. That would be sad.

This Issue

March 25, 1993