36. "Caution: Industrial Policy Is Coming," Challenge, (September/October 1992), pp. 58-61.

Whoever the next President may be, industrial policy (by one name or another) is coming.

If President Bush is reelected he is widely expected at least to continue with his previous efforts on the behalf of the computer industry (such as investment in a supercomputer that could conduct a trillion operations per second), auto makers (for which he donned a salesman's hat in Japan), and big financial institutions (which his administration is bailing out rather than have them exposed to the rigors of the market). If the economic recovery continues to be sluggish, Republicans may well resort to much more extensive industrial policy-making.

Robert Reich, a leading champion of industrial policy, is a major economic advisor of Governor Bill Clinton. Indeed, Clinton's economic plan calls for a massive investment of $200 billion in the infrastructure (communications, education, and so forth) over the next four years.


While there is much to be concerned with about industrial policy, neoclassical economists tend to bemoan it for the wrong reasons. Hence, they fail to see the measures that can be undertaken to safeguard us from the pitfalls of industrial policy. The standard neoclassical economic line is that industrial policy is nothing but a thinly veiled form of government intervention in the natural beautiful workings of the free market. (Former Secretary of Defense Casper W. Weinberger, calls it "socialism with a business face--part of the same socialism which has failed wherever tried.") Others argue that industrial policy is doomed to failure, adds billions to the deficit, and yields nothing but economic "dislocations." Chrysler should have been allowed to fail. The government's agreement with Japan on computer chips is said to have caused nothing but mischief. The suggestion that the Japanese economy has thrived under its industrial policy is termed misleading; William Niskanen, head of the libertarian Cato Institute holds that the Japanese government has contributed much less through industrial policy than is often argued.

Liberal economists, such as Robert Reich and Robert Kuttner, counter that industrial policy is already in place in the United States. The only remaining question is will we openly admit to it and use it systematically to our shared benefit, or allow it to be uncoordinated, sporadic, and confused.

Otis Graham Jr., in his new book The Industrial Policy Debate, quotes one study that states "it is often forgotten that our government is already intervening like a brain-damaged octopus." Graham also cites a Congressional committee which concludes that "the country currently has an extensive and expensive array of industrial policies... neither coordinated, cohesive, nor consistent...a melange of stop-gap measures. Among the industries that are benefiting are those that develop gasohol, the builders of windmills, nuclear utilities, agricultural businesses, and Savings and Loans Associations (and through them, real estate companies). Kuttner points out that the Pentagon acts like an American counterpart to the Japanese Ministry of International Trade and Industry (MITI), promoting many of those industries that play a military but also a major civilian manufacturing role, such as aircraft manufacturing and robotics.

Liberal economists are undoubtedly correct in observing that despite much of the rhetoric about the evils of "big government," the Republican Administrations of the last twelve years have themselves adopted a fair number of industrial policies, not unlike, say, those of the Carter Administration. Recently, for example, the Bush Administration singled out twenty-two technologies that the Pentagon would support. And the Export-Import bank continues to favor select large corporations, especially Boeing, McDonald Douglas, Westinghouse, and General Electric.

While these observations are valid, they in themselves do not answer the question of whether such interventions are in the public interest. Are their ill-effects the result of there not being sufficient and systematic industrial policy, or is industrial policy but one more form of "distorting" government intervention?


To understand industrial policy one needs first to lay aside the model of perfect competition (free markets) that drives so many neoclassical economists and laissez-faire conservatives. One must recognize that there never was, at any place, any time, a free market; moreover, such a blessed state was rarely even approximated and when it was, only for one rather short period of time. Most important, one can learn surprisingly little about the conditions we always face, of politicized mixed economies, from what might have happened in the never-never land of perfect competition. The model of perfect competition is unlike the laws of zero friction, which actually tell us something about movement when there is some measure of friction. It is a radically different world.

To suggest, in this context, that we face "imperfect competition" is doubly misleading. First, the term implies that what is missing in a given economy for it to be fully competitive are but a few details, like specks of dust on an otherwise perfectly pure surface. The fact is, economies are far from this state. Partial competition, or competition under moral and legal rules and in the face of significant power differentials among the competitors--or mixed economies--continues to be the rule.

Second, it is well known that the rules of perfect competition cannot be adapted to partial competition. Hence, we need to learn much more about the mechanics and dynamics of mixed economies, a task that has been much delayed by the many economists preoccupied with models of perfect competition or extremely simplistic models of monopolistic competition ("assume one seller, and one buyer...").

The place to start building models of mixed economies is to take as given that there is a significant degree of public involvement in the economy, which is interwoven with other factors so that lexographic context-setting cannot be assumed. With these two cardinal assumptions in place, one is ready to explore the benefits vs. damages caused by the various forms of public-private mixes on a variety of economic outcomes. (This process is akin to comparing quotas to tariffs, instead of extolling free trade).

It is in this context that industrial policy finds its place. Rather than bemoaning its existence or treating it as a miracle cure, it should be compared to other forms of government guidance, other modes of private-public combination, and other elements of the politicized economy.

Viewed in this way, one realizes that industrial policy belongs in the larger category of public policies that are often called "structural." These policies are needed because, the record shows, drawing only on macro policies, such as changing the levels of interest or tax rates (monetary and fiscal policies), is often insufficient to serve the economic goals that the society seeks to promote. For example, if pushing the economy to work at a higher rate of growth does not produce enough "trickle down," there must be special intervention on the behalf of ghettos or Appalachia, if these economies are to be lifted up as the national economy rises.


From this socio-economic vantage point, the right question is not should one have an industrial policy, but what kind? The main consideration here is not whether industrial policy is systematic or spottish, meager or encompassing, but how to protect it from undue partisan, political influence.

Industrial policy entails reallocation of resources by the government. In the American context this means reallocation that is subject to gross and detailed control by a Congress, rather than carried out by some executive Czar. Congress, in turn, is, most of the time, on many issues, deeply indebted to special-interest groups. Often, when a structural economic policy is passed through the hands of Congress, it is twisted to suit narrow special-interests, in the process undermining the shared and public goals the policy is supposed to serve.

Thus, one may argue whether or not there should have been an Economic Development Administration (EDA), whose purpose was to help lagging parts of the country by granting poverty stricken areas credit below the market rate. However, when 85.5 percent of the counties in the United States, including some of the richest, were qualified by Congress for EDA loans, the program costs soared out of control and its goal was defeated.

Similarly, one may argue whether workers who are laid off because of imports should be retrained to ease the transitions involved (and to maintain political support for relatively freer trade). However, once Congress has written the law in such a way that numerous groups, favored by labor unions, receive preferential treatment, the program costs soar out of hand, and it turns from a trade adjustment measure to one more program of entitlement.

A recent Wall Street Journal article describes the expansion of an enterprize zone in Kentucky under political pressure. The zone, originally designed to reinvigorate an economically distressed area and which covered 3.8 square miles of Louisville's inner city, has grown to encompass over forty-five square miles and is home to a new airport, swank lounge, and car rental companies. The same story is retold in other parts of the country, threatening to turn this program into an expansion tax boondoggle. The same holds for scores of other programs.

In the last fifteen years, the conditions have grown less favorable to programs that Congress controls, simply because the monumental rise in the costs of campaigning has made special-interest groups and their political action committees (PACs) even more controlling of Congress. The need to reward those who make contributions to their campaign chests is, of course, a major reason that Congress often insists on the right to determine, in effect, which specific corporations get funds (say, to build a given weapons system), rather than budget X billions of dollars for a general program (say, for bombers, submarines, and missiles), and let the proper agencies spend where they can get the best bang for their bucks.

As a result, there is little doubt, as was pointed out years ago by economist George Eads (in testimony before Congress) and by myself (in my book, Capital Corruption) and recently reaffirmed by Kevin Phillips (in the Harvard Business Review), that given half a chance Congress would expand industrial policy to a point where its costs would become a major burden on the economy. Moreover, it would allocate the funds to industries that carry the most political clout rather than those that are particularly innovative, able to export, or are otherwise promising for the future of the American economy. Worse, because old industries (such as coal and steel), are by nature more politically entrenched than fledgling new ones (such as artifical intelligence AI! and biotechnology), Congress would strongly favor the former over the latter, even if it could be determined that coal and steel should be scaled back and AI and bio-technology spurred--a point that is widely contested. The question for the next administration is hence not whether to have industrial policy but how to shield it from the politicians and the special-interest groups. (It might be suggested, quite properly, that it would better to defang the interest groups, say by effectively reducing campaigning costs and campaign contributions. However, regrettably this is not realistically what one can expect.)


There are basically two ways to proceed. Although neither is a panacea for every corruption or abuse of the public trust, both proposals would substantially reduce the favors given to private interests at public expense. One could set up a reindustrialization authority, to be appointed by the President and Congress, and governed by a nonpartisan board. Board members would need to have long terms (fifteen years) like the head of the Government Accounting Office (GAO). They would determine who gains the various funds and other governmental benefits (loan guarantees, export insurance, and so forth). Congress would only set the annual (preferably tri-annual) budget for the reindustrialization authority.

While, this well may work, it has the disadvantage of locking one more part of the political economy into an institution that would be largely independent from public policies. We already have the Fed which, although insulated from pork-barreling political pressures, often responds only grudgingly and with much delay to changes in economic policies favored by the "rest" of the government, the President and Congress, even when they pull in tandem. For this reason, the Fed is often compared to one of four airplane engines, out of sync with the others.

Establishing a reindustrialization authority would be tantamount to turning over the controls of one more engine of the plane to some other semi-independent authority. While this might be necessary, it is an unsatisfactory way to fly an airplane or guide an economy.


The other possibility is to limit industrial policy to semi-guided policies. These are policies that aim to send a signal to the economy that certain broad-gauged activities or sectors are favored (hence, the "guidance")without allowing Congress to determine the specific beneficiaries (hence the "semi").

We semi-guide when we allow people to enlarge their investment retirement accounts (IRAs) and Keogh plans because we use tax deferment to encourage saving, without determining the specific investments into which the funds must go. (We set some limits, such as that these funds cannot be invested in antiques, but the range is quite wide. IRAs and Keogh plans would be guided if we told IRA holders they must invest, say, in office real estate to qualify for their tax benefits). We semi-guide when we grant corporations tax credit for investments in research and development; we guide when the NSF and NIH reviews thousands of specific research proposals and when endless peer committees decide who gets which grant. We semi-guide when we allow accelerated depreciation for investments in equipment and plants; we guide when Congress writes what in effect amounts to tax concessions to specific corporations. (Although the text is often written as if it is generic, such as a tax exemption for "all" corporations that were charted in July 1985, that deal with forest products and are in Northwest towns larger than a given size but not smaller than some other size. The description fits only one favored corporation.)

In one of those unfortunate twists of terminology, where concepts confuse more than they clarify "industrial policy" is widely used to refer both to a guided approach and to semi-guided policies. Kevin Phillips, for instance, uses the term industrial policy to refer both to increased funding for basic research, increased tax credits for R&D, as well as to supporting the building of a specific kind of supercomputer. (Once more, one cannot say one of these usages is "wrong" as there is no agreed upon definition of the term. However, as these examples show, the term can encompass a very wide array of policies that differ significantly in their nature. Hence, the merit of the suggested distinction.)

The intensifying debate as to where economic policy should be headed would benefit greatly if a clear distinction were made between semi-guided industrial policies and guided industrial policies (and between both of these forms of structural industrial policies and nonstructural macro policies that are unguided).

With this cardinal distinction, one sees strong reasons to firmly oppose industrial policies of the guided kind, and good cause to endorse semi-guided industrial policies that provide a high degree of protection from special-interest groups.


To spur the economy to serve the goals that the society is committed to advancing, macro policies may well not suffice and structural policies may be needed, not to replace financial and fiscal policies but to supplement them. Industrial policy is one such structural policy. The problem is not that it constitutes a form of government guidance; all economies have those. Difficulties arise because our legislators (national and state) are often under the thumb of special-interest groups. Until their role is sharply curtailed, an industrial policy is extremely likely to become a new form of pork-barrelling rather than serve the public interest. Hence, we need to find ways to protect industrial policy from special-interest groups. Creating an autonomous reindustrialization authority might suffice; however, crafting a semi-guided industrial policy, aimed at sectors or whole modes of economic activities (rather than specific industries not to mention particular corporations) is preferable. Such semi-guided approaches would not entail locking in the guidance of yet one more segment of economic policy in a control tower, insulated from the general public and the national policy making apparatus.

Amitai Etzioni is author of The Moral Dimension: Toward A New Economics and founder of the Society for the Advancement of Socio-Economics (SASE).


The Communitarian Network
2130 H Street, NW, Suite 703
Washington, DC 20052