This article is adapted from testimony Manne delivered to the House Oversight Subcommittee on TARP, Financial Services, and Bailouts of Public and Private Programs on April 17, 2012.

Eventually that issue was put to rest with the arguments (largely pursuant to the “necessary-and-proper” clause of the Constitution) that society had become so complex that Congress had to rely on experts to do the detailed work of regulating, which Congress, by virtue of its expertise limitations, could not do; that the delegation had to have some semblance of reasonableness; and that due process, usually in the form of a right to appeal to the courts, be available. None of these justifications were premised on economic concerns. This issue was thought to be exclusively the province of lawyers and political theorists, not economists—though who was to exercise the needed expertise was never made clear.

Public Choice Theory

The next major phase in the debate over administrative powers came unexpectedly from the then-new field of Public Choice theory. Public Choice uses economics to analyze political phenomena, and the field is generally considered to have opened with the publication of James Buchanan and Gordon Tullock’s book The Calculus of Consent in 1962.

Public Choice offers two main criticisms of the regulatory agencies. The first is that the behavior of bureaucrats is more accurately seen as self-serving rather than being motivated by the public interest. This would frequently manifest itself, not simply in the older idea that bureaucrats are slothful, but in agencies’ push for ever greater budgets to fund expanded powers, first described by William Niskanen in his 1971 book Bureaucracy and Representative Government. The second criticism, in two parts, is that agencies could and very often are co-opted by the very interests that they are supposed to regulate, and that these combined interests are used for so-called “rent-seeking” purposes.

Each of these criticisms of regulatory agencies has become standard fare in political theory. To a large extent, these ideas have permeated all levels of serious discussion about the administrative state. But “permeating the discussion” is a long way from having a real political influence, and on that score Public Choice seems to have only affected intellectual and academic understanding of regulation. There has been no serious legislative check on the possibility of regulatory abuse since the federal Administrative Procedure Act of 1946. But perhaps, often with considerable lag, intellectual discussion is the source of all good government reform.

Austrian Economics

Economic criticism of central planning has a long history. A now classic debate about free markets versus central economic planning raged in the late 1920s and the 1930s in reference to ideas of “scientific socialism” being advanced by apologists for the Soviet Union’s extreme form of socialist planning. The principal criticism of this kind of planning probably originates with Ludwig von Mises, a founder of the Austrian School of Economics. Mises declared that central planning and non-market allocation of resources could not work because the only logical basis for making efficient decisions was the existence of a market price. But a market price would not be available in a socialist system because price evolved out of the voluntary interactions of individual buyers and sellers in the marketplace.

This style of criticism was developed further and elaborated by Mises’s student, Friedrich Hayek, notably in one of the most famous and influential articles in all of economic history, “The Use of Knowledge in Society.” Hayek’s basic thesis in that article and much of his later writings was that the knowledge necessary to make “correct” centralized economic decisions could never be mastered by one person or agency because the information required to make such decisions was enormous and totally diffused throughout society in the minds of countless individuals. Furthermore, the necessary knowledge changed from moment to moment as circumstances changed. Thus reliable information could never be imparted in timely fashion to central planners.

This presented a practical argument against centralized planning that is today almost undisputed. With the collapse of the Soviet economy in the late 1980s, Hayek’s explanation became a sort of gospel for anyone trying to understand the fatal weakness of such a system.

What many observers failed to notice was that the same argument applies with small modification to the kind of administrative regulation endemic today in the United States. We rarely call it “central planning,” but the types of decisions and the knowledge required for correct industry or sector-planning decisions—as, for instance, with the Securities and Exchange Commission, or the National Labor Relations Board, or the Environmental Protection Agency—are the same in a regulatory regime as in a centrally planned economy. True, the SEC does not make decisions explicitly as to which industries should receive new capital, but even mundane decisions affecting the cost of different forms of financing have allocational consequences. Indeed it is difficult to think of any significant substantive regulation that does not have some allocational consequences. The mere fact that these consequences are ignored does not mean that they are not present.

In a nutshell, Hayek’s argument is that the technical expertise necessary to make efficient allocational decisions is, of necessity, simply unavailable, whether that decision is to be made by a Soviet-style central planner or an SEC rulemaking procedure. Furthermore, there is no evidence to make us believe that a series of uninformed decisions will on balance do more good than harm. By happenstance some rules will seem to work (meaning that the market finds a new equilibrium) and a great many will be insignificant economically, but even this cannot be known for sure in advance. No sort of Darwinian survival process operates automatically to weed out bad decisions and allow good ones to survive as it does in the private sector. The bad survive along with the good, and we do not even have an apparatus in place to test which is which. Apart from a totally unjustified belief in the skills and good faith of our regulators, there is no rigorous or logical way to justify much of their work.

But alas, we are not yet at one of those defining moments in our history when we can make a choice between continuance of our present regulatory-state, quasi–central planning model and a freer, more growth-oriented, less intrusive free-market model. Laissez faire and far-reaching deregulation is not part of the 2012 American political zeitgeist.

Cost-Benefit Analysis

But that is a long was from saying that there is nothing we can do to make the system of administrative regulation work more effectively in the public interest. While a rigorous cost-benefit approach to regulation may to some degree be at odds with Hayek’s notions about “expertise” and Mises’s doubts about the practical validity of empirical evidence, we may have to live in a second-best world. That is, even if our present regulatory apparatus is doing more harm than good, it cannot in the foreseeable future be thoroughly dismantled. So we might at least try to minimize the damage that it does.

The techniques and power of so-called cost-benefit analysis have improved remarkably in the last 50 years. This reflects, in part, the huge advancement in the field of econometrics, of which cost-benefit analysis can be said to be a subfield. The quality of the data available for calculations is also much improved, largely as a result of the accessibility that computers have given to new databases and the increased reliability that computerization has added to the regression of data.

There has also been a vast improvement in the economic models we can use to test the efficacy of proposed regulations. A clear example of this is provided by the development of a field called “transactions cost economics,” for the introduction and elaboration of which Ronald Coase and later Oliver Williamson received Nobel prizes in economics. The influence of this concept can clearly be seen in Judge Douglas Ginsburg’s now-famous 2011 opinion in Business Roundtable v. SEC, where he lays out a veritable catalog of components of an acceptable cost-benefit analysis. But while the advent of these newer techniques and ideas has greatly strengthened the ability of willing administrators to make sensible empirical judgments, it has by no means vitiated the fundamental objections of Austrian School economists, or even many Chicagoans, to this kind of regulation.

In the public sector, no sort of Darwinian survival process operates automatically to weed out bad decisions and allow good ones to survive, as it does in the private sector.

Still, second-best is better than no “best” at all, and the latter is exactly what the SEC and many other agencies have been offering us for a long time by their failure to offer up any form of economic justification for their rules and decisions. This has always seemed to be particularly ironic in light of the justification legally and popularly made in the 1930s for administrative regulatory agencies. That was, of course, that the agencies’ “experts” would have the technical skills required to do “scientific” economic planning or rulemaking.

It is a little weird, then, to hear these same officials ridicule the application of such “expertise.” I don’t think that it was with reference to Hayek or Mises that a former commissioner of the SEC criticized economists who “attempt to compress the complexity of our security markets into horribly complicated formulae.” Apparently he preferred to have complex questions addressed without reference to rigorous analytical models, but more likely he—like the entire securities bar at the time—was simply unaware that such useful models even existed.

Business Roundtable v. SEC

But the SEC’s problems with economics don’t end with its failure to take seriously the basic kind of analysis one would expect of an economic regulatory agency. The agency doesn’t even do the kind of analysis that Congress has explicitly required it to do. In his decision in Business Roundtable v. SEC, Judge Ginsburg excoriated the SEC for its failure adequately to address, in connection with the recent proxy solicitation rule regarding nominations of directors, the congressional requirement that rules take into account the effects on “efficiency, competition, and capital formation.” Such failure made the SEC’s rule “arbitrary and capricious” within the meaning of the Administrative Procedure Act of 1946, and the rule was sent back to the SEC for further consideration.

Judge Ginsburg’s opinion lists many of the relevant requirements for what we normally term a “cost-benefit study.” The case seems to stand for the proposition that many agency rules (including well-established ones, under Dodd-Frank’s requirement for a cost-benefit study of old rules) will now have to stand the test of a rigorous cost-benefit analysis before they can receive the sanction of legality.

This requirement, which could be strengthened and made escape-proof by confirming congressional action, will undoubtedly have a number of salutary effects, in spite of the difficulty of getting reliable data on some of the issues that the SEC regularly faces. First, this requirement will provide an analytical template for the consideration of any new rule. That is, it will force the agency to give adequate consideration to a variety of significant economic questions that it now regularly sloughs off or to which it simply assumes the answer. Next, it will force the agency to make real-world quantitative comparisons instead of simply assuming answers or even finessing hard questions altogether. It will offer some assurance, to be sure, that the agency will not adopt rules that are economically harmful. And finally, it will make the discussion of new regulations more open to truly informed community comment as opposed to special-interest pleading. Third parties will know that their comments will be examined by sensible and knowledgeable experts and not bureaucrats interested mainly in the political implications of a new proposal. A possible incidental advantage of such an approach is that courts may now give deference to agency discretion only when it makes sense to do so—that is, on rules properly vetted by true experts in the field and not by regulatory poseurs. In time we should develop something like a common law of good practices in cost-benefit analysis and perhaps even improve the quality of economic regulation.

SEC Memorandum

On March 16, 2012, the Division of Risk Strategy and Financial Innovation (RFSI) and the Office of the General Counsel (OGC) of the SEC issued a memorandum outlining a highly sophisticated, comprehensive plan for cost-benefit analysis of SEC rules, both future and existing. This seemingly represents a revolutionary turnaround from the past practices and culture of the agency. Though it comes 80 years after this sort of thing should have become common practice, it is better late than never.

Irony of ironies, the first paragraph of the memo contains the following assertion: “High quality economic analysis is an essential part of SEC rulemaking.… The Commission has long recognized that a rule’s potential benefits and costs should be considered in making a reasoned determination that adopting a rule is in the public’s interest.” This suggests a somewhat less than enthusiastic endorsement of the use of economics to determine a rule’s propriety. But, despite my reservations about the SEC’s sincerity in this endeavor, or even of the agency’s comprehension of the relevant economics, I do not want to appear ungrateful for this obviously thorough and informed document. Perhaps it will be as revolutionary as it sounds. Perhaps. However, I do have a few—I hope—constructive questions and comments about some of the substantive details and the operation and enforcement of this new requirement.

Market failures | Footnote 16 of the memo contains the most significant substantive aspects of the economic approach the staff will be expected to observe. It correctly states that regulation should follow upon some recognized failure of the free market, though no reference appears to “government or regulatory failure” to which the market failure should always be compared. The memo then lists some examples of market failure: “externality, market power, and inadequate or asymmetric information,” with generalized examples of each of these and a couple more claimed market failures.

The problem here is that each of these alleged market failures needs to be addressed with considerable circumspection. The negative externality argument, exemplified by “spill-over financial risks,” is of a sort very uncommon—though theoretically correct—in financial markets. Further, the root causes of such problems are often well beyond the SEC’s powers to deal with, such as certain aspects of the 2008 crash that could be blamed on federal housing policies largely untouched by the SEC.

Turning to the benefits of “positive externality,” to what extent do “positive externalities” actually flow from a “disclosure” regime? One cannot simply assume the benefits of this fundamental SEC regulatory tool when trying to measure the costs and benefits of new disclosure regulation. Of course, if there were overwhelming evidence that the disclosure regime we have had in place for nearly 80 years has benefited society more than it has cost, perhaps that exercise would not be necessary for each new disclosure-type rule. But I know of no such strong evidence. There are, however, competent studies indicating the contrary.

Comment is also indicated for the inclusion of “market power” as a kind of market failure that justifies regulation. The SEC has not very often in its entire history encountered a true and significant cartel or monopoly that was not either generated or protected by government regulation of one kind or another. The now-defunct regime of a fixed commission rate structure on the New York Stock Exchange would be a good example of market power that was protected for years by SEC policy. It is hard to imagine another problem of market power that is not of this variety. Again, perhaps this notion is theoretically appropriate, but practically speaking it is a near-dead letter.

Next listed is “principal-agent problems” arising in the form of “moral hazard or in situations involving potential conflicts of interest.” Here, even at a theoretical level, the economics of the memo is wanting. Principal-agent problems and moral hazard are not indications of market failure. They simply represent market costs, and though sometimes “transactions costs” of this sort are spoken of as changing the fundamentals of market economics, we now know that this is not so. Such costs may be high, but that does not in and of itself make them into market failures. Because much of the edifice of modern corporate governance literature is built on Adolf Berle and Gardiner Means’ fallacy of the principal-agent problem as a market failure, this example in the memo might represent another shortcoming of the SEC’s underlying economics.

Asymmetric information | But then comes the most revealing and disturbing statement about market failure in the entire document: “There is asymmetric information, for example, when investors seeking to trade securities are not fully informed of all material information that could affect their investment decisions.” Investors are never “fully informed,” for if they were, there would be no risk in investment. This is the shibboleth (obviously merely the converse of “full disclosure”) under which the entire “disclosure” philosophy of the SEC has been maintained during the long years of excluding economics from its consideration. To consider less than full information to be a market failure is to misunderstand the basic idea of scarcity as part of the human condition. Information is an economic good that follows all the fundamental rules of economics, and while it does have some unique characteristics that give rise to special consideration, a simple lack of full information is not one of them. This statement is a big-enough hole in the otherwise generally appropriate document to make the entire thing an exercise in futility. A sound cost-benefit analysis of any aspect of “disclosure” regulation must not start with the question-begging assertion that asymmetric information represents a market failure.

In the most celebrated work on asymmetric information, on a “lemons” market in used cars, Nobel laureate George Akerlof showed how a lack of information by consumers could theoretically result in the collapse of an entire market for a good product. This would indeed represent a market failure par excellence. Unfortunately this theoretical demonstration (which has yet to be certified as ever existing in the real world) captured the imagination of a lot of economists searching arduously for any new market failure they could lay hands on. What most references to the Akerlof theory have failed to note is that the used car market did not disappear and that the private market had already provided all manner of solutions to the problem that Akerlof identified.

And so it is with other areas of asymmetric information, including securities markets. There is no proof of the theory, and consequently the theory itself may be lacking. This is not to suggest that some investors may not be benefited by mandated disclosures, or even that on balance this form of regulation is never beneficial. It is to say, however, that the asymmetric information form of market failure is a weak reed on which to base much regulation. And to the extent that this is simply assumed to be a market failure, it then becomes easy to justify bad regulations as being justified on a cost-benefit basis.

Congressional mandates | Footnote 19 of the memo addresses the contentious issue of whether the SEC is required to offer a cost-benefit analysis when Congress has mandated a rule. The agency has stuck to its seemingly untenable position that it should not do such an analysis when Congress has mandated a rule because Congress has left the agency with no discretion and, therefore, no need for an economic analysis in the matter. Presumably, the basis for this argument is that the analysis might contradict a stated or implicit congressional finding of a market failure.

But supporters of a cost-benefit analysis requirement are not arguing that the SEC can overrule an act of Congress. Even considering the extreme case of such a contradiction, such a finding (along with a least-cost solution) would seem to be of the essence of regulatory responsibility. After all, these agencies were created and tolerated because Congress did not have the expertise to enact detailed regulation. If Congress has made a mistake in the eyes of SEC analysts, the analysts should say so and not hide from their responsibility because of fear of some kind of retribution. Furthermore, there are few if any cases of Congress mandating a rule on which the agency in question does not still have enormous discretion about what the final product will look like. When Congress mandates that an agency adopt a rule, Congress is not writing the rule (or there would be no need to require the agency to do so), and the approach clearly implies that Congress believes there are many different ways the rule can be detailed. The devil, after all, is in the details, and it is precisely those details that need to be justified on a cost-benefit basis.

Too few economists | I should now like to turn to some practical aspects of the March 16 memorandum. While I applaud the SEC’s adoption of a more sophisticated economic approach to rulemaking than it had heretofore exhibited, there are certainly significant practical problems with the implementation of this bold plan. Last year, the SEC listed just 28 “economists” (some of whom did not have advanced degrees in economics, but instead were accountants or holders of MBA degrees) among its over‑3,000 employees, and I believe that this is an all-time high number. The agency is requesting additional money so that it may hire more economists. Given the tasks of generating new rules under the Dodd-Frank Act and its requirement for cost-benefit studies of existing rules, the number of highly trained and competent economists necessary to complete this job in several years is likely to be on the order of 100 to 150, if not more.

The foundational task of assembling the required databases for this work will in itself engage a huge number of experts for a long period of time. Each of the new staff members will have to be brought up to speed on the institutional aspects of securities regulation before they can begin this work. Where are the resources for this gargantuan task? I suggest that they already exist at the SEC in the form of what should soon be redundant lawyers and policy experts presently working on rulemaking in the “old style.” In other words, there needs to be a shift in the SEC’s orientation from law to economics, and personnel policies should reflect this new reality. This job should be able to be accomplished with no additional funding.

Implementation | The next practical question is how to make this new policy become and remain a serious reality. In other words, how is this new approach to be enforced and monitored? This is especially relevant as there will undoubtedly be agnostics inside and outside the SEC who will fight relentlessly to guard their existing intellectual and bureaucratic capital. To this end, the appropriate committee of Congress should mandate something along the line of the March 16 memo and then require regular and detailed progress reports from the SEC. These reports should also be available for public comment. For example, Congress might require such a report from the SEC three months from now, then another in six months, and another one year later, and thereafter once ever two years. The reports should make it evident whether the SEC is actually using sophisticated and objective cost-benefit analysis and other economic techniques in its rulemaking work, and they should discuss any respectable criticisms made of the SEC’s work in this regard.

As an additional safeguard, judicial review of the substance of the economic analysis should also be guaranteed and not allowed to disappear under the rubric of “agency deference.” I have no doubt that the D.C. Court of Appeals decision in Business Roundtable v. SEC had some effect on the SEC’s turnaround on the question of economic analysis. I think that the right of judicial review of allegedly faulty or insufficient analysis or other mistakes can help make these new requirements really meaningful.

Conclusion

The SEC memorandum on the use of cost-benefit analysis will not overcome the inhibiting effects of 80 years of a different intellectual culture at the SEC, but it will be a start. With congressional oversight, judicial review, and the good-faith sympathetic administration of these new rules by the SEC, a far more effective regulatory system may come about than we have had, even one with some real intellectual credibility. In time, everyone involved with the SEC may come to understand what an economic regulatory agency is supposed to be all about.

Readings

• Socialism: An Economic and Sociological Analysis, by Ludwig von Mises. Yale University Press, 1951.
• “The Market for Lemons: Quality Uncertainty and the Market Mechanism,” by George Akerlof. Quarterly Journal of Economics, Vol. 84, No. 3 (1970).
• “The Problem of Social Cost: What Problem? A Critique of the Reasoning of A. C. Pigou and R. H. Coase,” by Harold Demsetz. Review of Law and Economics, Vol. 7, No. 1 (2011).
• “The Use of Knowledge in Society,” by Friedrich Hayek. American Economic Review, Vol. 35, No. 4 (1945).