Jeff Gates is right to look for institutional arrangements that combine markets with equality, and his proposals for popular ownership of industry have a lot of promise. However, I think they suffer from a tendency to romanticize ownership and hence to exaggerate the potential contribution of the proposals to egalitarian democracy. Moreover, his vague notion of ownership occasionally diverts attention from some critical issues of institutional design.

We should note initially that corporate profits represent only about 12 percent of national income. So even if Gates's reforms were to accomplish a dramatic redistribution of ownership income, they would not, by themselves, make a strong dent in the overall inequality that Gates deplores. In strictly egalitarian terms, the most important reforms focus on the labor market-education designed to raise labor productivity, minimum wage and benefit standards, and legislation encouraging unionization and collective bargaining.

Nevertheless, as Gates emphasizes, the concentration of capital has especially powerful political and social effects. If the "ownership solution" could mitigate them without unduly sacrificing economic efficiency, it would be a great achievement.

Gates's proposal should be compared to current pension and retirement policies, which are largely oriented toward inducing savings through subsidy or fiat. The savings that these policies generate take the form of creditor claims. By contrast, Gates proposes subsidies and fiats to generate ownership claims. Ownership claims differ from creditor claims in two major respects. First, while creditor claims are senior and fixed, ownership claims are junior and residual. Owners get what's left over after the creditors are paid off. Second, while creditors get only limited control over the enterprise as long its is solvent, owners have major control rights.

On average, ownership claims earn more than creditor claims. One of Gates's goals in equalizing ownership is to equalize these relatively large returns to ownership. However, these returns are functionally related to certain functions that owners perform. For ownership to perform its economic role, these functions have to accompany the rewards of ownership. There are three principal ownership functions. First, owners or their agents allocate capital toward the most promising investments. Second, owners act as entrepreneurs, using their control to make the enterprise more profitable. Third, owners bear risk. They are most exposed to loss from business failure.

To what extent is it desirable to extend these functions to ordinary citizens? It seems doubtful that average citizens are well qualified to evaluate the comparative merits of investments. Current allocative practices are arguably biased against certain types of investment and certain groups. A variety of programs, including those of the Small Business Administration and Community Loan Funds, attempt to rectify this bias, but most depend on technical judgments by professional managers. Most ordinary citizens who invest in stocks don't try to "pick winners"; they diversify by buying a representative sampling of the market, which is what most experts think they should do. As for entrepreneurship, it would be a good thing to expand opportunities for this role to a broader range of people-but again, this is not a role that one would expect a large fraction of citizens to play.

The diversified, passive owner, who is neither a stock picker nor an entrepreneur, gets rewarded for bearing risk. Returns to ownership claims are higher than for debt claims because ownership claims are riskier. But, of course, returns are higher only on average and over time. The relative riskiness of ownership claims means that many of them suffer losses. If the economy is to continue along capitalist premises, to increase the participation of ordinary citizens in profits probably means increasing their exposure to the risks of corporate losses. For a lot of reasons, this is probably a good idea, but it's not an uncontroversial one, even among progressives. For example, unions have traditionally preferred defined benefit pension funds, which give workers what are basically fixed, low-risk debt claims, to defined contribution plans, which give them riskier claims dependent on the profitability of the underlying investments.

Gates has a fourth function in mind for his expanded "ownership solution"-reducing what are now external costs of corporate activities on workers, customers, and local communities. As long as businesses are owned by absentee owners, the costs of unnecessary plant closings, defective products, and pollution are externalities-since they're born by nonowners, the businesses have no direct incentives to minimize them. To the extent that we extend ownership claims to these affected constituencies, they can use their control rights to force more accountability for such costs.

The best illustrations of the poten-tial for efficiency gains from such cost-internalization-through-ownership extension comes from companies like Chrysler and United Airlines, where giving ownership claims to workers has facilitated restructurings that have improved the general health of the enterprise. Without the ownership claims, the workers would not have trusted managers enough to make the concessions in pay and work practices necessary to restructure. Ownership rights gave them some assurance that managers would not take opportunistic advantage of their concessions.

Nevertheless, there are limits to the generalization of such approaches. For one thing, increased control in any particular enterprise comes at the expense of decreased diversification and hence increased risk. Workers already have a lot of human capital tied up in their workplaces. If they take major financial stakes, they increase their exposure. To be sure, they reduce certain risks, such as managerial opportunism. But they become more exposed to risks outside the enterprise. A general downturn in the industry could wipe out their investment. Diversified investors are much less exposed to such risk. This is the reason that, when workers take ownership stakes in restructuring corporations, they frequently sell out their stakes after the restructuring is complete. That's what the Chrysler workers did after the successful bailout of the 1970s, and it's what the United workers are in the process of doing now. When the enterprise is in crisis, risk-taking is unavoidable. But once it is stabilized, workers would like to reduce risk by diversifying, which means giving up control.

Another consideration is that, for many externalities, government regulation is a superior response to expanded ownership. For example, for many decades, many banks and insurance companies were customer-owned. They were-and a few surviving ones still are-called "mutuals." Mutuals are owned by their depositors or policyholders. However, in recent decades, the mutuals have been converting to conventional corporations, and even in the surviving ones, there is no indication that the customer owners are very active. As Henry Hansmann has pointed out, the most likely explanation is that government regulation has made customer ownership obsolete. When customers had to fear that the companies would behave recklessly and opportunistically with their money to the point they would be unable to deliver the promised benefits when due, ownership was an important check. But now, most depositors and policyholders seem to believe that regulation will protect them adequately.

An underdeveloped component of Gates's proposal concerns how the exercise of ownership rights is to be institutionalized. Not all current forms of corporate ownership involve the kind of power and accountability that would be needed to accomplish Gates's goals. Consider two wholly inadequate current models:

First, there's the typical passive individual corporate shareholder. She is "rationally apathetic." Given her small stake in any particular company in her portfolio, it doesn't pay her to inform herself or become active in corporate governance. Except in the rare instances when some financial entrepreneur mounts a proxy fight, she routinely gives her proxy to management. In the event of a hostile takeover, she always sells her shares to the highest bidder.

Second, there's the "public interest director," a person chosen by management to speak for some external interest, such as workers, minorities, the environment, or the local community. Such directors occasionally play useful roles, but more often than not, their roles are cosmetic. Since they have no financial investment of their own and don't represent any constituency that does, they tend not to be taken seriously. And it's not clear why they should be, since they're usually not accountable to their putative constituencies in any organized fashion.

There's a real danger that the "ownership solution" might turn out to mean merely more of one or the other type of fake participation. The important point is that the idea of ownership is not self-defining, and its current manifestations are not necessarily well adapted for a program of economic democracy. If ordinary citizens are to participate in corporate governance, they will need to do so through representation. This means that they need representatives who not only have clout in the boardroom, but can be held meaningfully accountable to them. The most promising route is probably to connect corporate ownership to institutions performing other functions-labor unions, local governments, community development corporations. For the "ownership solution" to work, we need more of these institutions, and they need to be stronger.