| 摘要 | Underlying many recent proposals to reform regulatory decision making is the theory of agency independence from congressional control. Regulatory agencies, it is argued, are largely autonomous, pursuing self-serving goals rather than the goals for which they were designed, with little accountability to the people’s elected representatives. Evidence typically cited for this theory emphasizes the infrequency of congressional oversight hearings and public investigations, the perfunctory consideration Congress usually gives to agency appointments and budgets, and the substantive ignorance members of Congress often display about agency operations. James Q. Wilson sums up the situation this way:
By and large, the policies of regulatory commissions are not under close scrutiny or careful control of either the White House or of Congress. … Whoever first wished to see regulation carried on by quasi-independent agencies and commissions has had his boldest dreams come true. [Regulatory agencies] operate with substantial autonomy. …
The recommendations for regulatory reform that flow from this view of agency behavior would either tie agencies more closely to Congress or force Congress to pay closer attention to agency activities. For example, the legislative veto is intended to make it easier for Congress to challenge regulatory policies. Instead of drafting, debating, and passing new legislation to direct an agency, Congress can simply veto a proposed agency rule. Sunset legislation is intended to force Congress to scrutinize agency performance in a systematic manner. Instead of having to take positive action to alter an agency’s decisions, Congress has to act positively—and regularly (say, every five years)—in order to keep a particular program or agency in business. Both devices, it is thought, would encourage regulators to justify themselves to their nominal overseers on the Hill.
An Alternate Theory
Unfortunately, if the theory is wrong, reforms based on it are not likely to be much help. Suppose that agency discretion plays only a minor role in regulatory decision making, but that congressional influence is substantial and regular. Suppose, in other words, that agencies are doing precisely what Congress wants them to do. Then schemes for giving Congress more tools would clearly have little influence on policy outcomes, or perhaps even make things worse.
But is the theory of congressional control of agency decisions a plausible one? We argue that it is. Congress need not direct agencies solely by “textbook” means—public hearings, investigations, policy pronouncements, and the like. It may also do so through a subtle bureaucratic incentive system—an “invisible (congressional) hand.” In this view, appropriations subcommittees channel budgetary rewards to agencies that successfully pursue congressional constituency interests. Oversight committees affect agency appointments by manipulating the selection process, not by making a last-minute effort at the confirmation hearings. Congressional staff members monitor agency activities regularly but informally. Finally, and perhaps most important, Congress reserves its powerful club of hearings and investigations for agencies that get out of hand. Hearings and investigations, accordingly, are not forums for the rational consideration or public airing of policy alternatives; they are tools for imposing sanctions on errant agencies.
Note that if these incentives and monitoring devices work efficiently, we should see little sign of congressional attention to the daily activities of regulatory agencies. Public hearings and investigations are, after all, resource-intensive activities that must be used sparingly. And, of course, the more effective the “club behind the door,” the less frequently it has to be used and the more ad hoc its use appears to observers who miss the underlying incentive structure.
What this means is that the evidence marshalled to support the theory of agency independence—namely, the infrequency and superficiality of congressional hearings and investigations—is also consistent with a theory of congressional control of regulatory policy. Thus we must look elsewhere for evidence to determine which theory is more appropriate. With this in mind, let us turn to a more systematic investigation of the relationship between Congress and one regulatory agency, the Federal Trade Commission (FTC).
Assembling the Evidence
In the 1970s, the FTC pursued a bold, many faceted attack on what it interpreted as abuses of economic power, earning for itself a reputation as everything from “tough cop” to “national nanny.” The “kid-vid” case, in which the FTC sought to curb allegedly deceptive advertising aimed at children, seemed to many to typify all that was wrong-headed about this behavior. Other controversial actions included antitrust suits against the nation’s largest oil producers and cereal manufacturers, and investigations of the insurance industry, used-car dealers, and a host of professions (undertakers, pharmacists, opticians).
By 1979, the commission was widely viewed as the prototypical runaway agency, and Congress—tardily, many thought—took action. It not only publicly chastised the commission in Senate hearings of September 1979, but also began to impose sanctions—halting, over the next two years, the “kid-vid” rulemaking and the FTC’s most controversial investigations and placing curbs (including the legislative veto) on the commission’s powers. In October, Senator John Durkin (Democrat, New Hampshire) described the hearings and contemplated sanctions as “shock therapy for bureaucrats,” and Representative Elliot Levitas (Democrat, Georgia) announced, “We are seeing the end of government by bureaucratic fiat.” Summarizing the situation, the National Journal noted that “The FTC [had] roamed far beyond congressional mandates with a shotgun attempt to regulate all kinds of business activities that should not be the concern of the government” (October 13, 1979). According to this view, had Congress been more watchful in the first place, the FTC would never have strayed so far from congressional interests.
Branding regulatory agencies as runaway, uncontrollable bureaucracies makes good press. But does the shoe fit the FTC case? To find out, let us first review some institutional details about Congress and the commission, and then examine statistical data on the relationship between the preferences of the members of the FTC’s Senate oversight subcommittee and the policy choices made by the commission.
A Decade of Cooperation. The late 1960s and early 1970s witnessed an explosion of consumer protection legislation, much of it associated with three influential members of the Subcommittee on Consumer Affairs of the Senate Commerce Committee: Warren G. Magnuson, Philip Hart, and Frank Moss (Democrats, from Washington, Michigan, and Utah respectively). These three consumer activists used their committee positions to hold hearings on topics that later became subjects for FTC investigations and rulemaking proceedings. Indeed, most of the rulemaking that drew criticism in 1979 had originated in Congress or resulted from congressional prodding. As Ernest Gellhorn points out, as late as 1977 Congress was regularly chastising the commission for moving too slowly in its rulemaking proceedings, not too aggressively.
Moreover, Magnuson and Moss in the Senate and John Moss (Democrat, California) in the House worked for five years to pass the Magnuson-Moss Warranties and FTC Improvement Act of 1975, which confirmed the commission’s power to promulgate generally applicable consumer protection regulations. Could Magnuson, Moss, and their consumerist colleagues have been unaware of the consequences of this legislation? This seems unlikely, since many of the FTC rulemakings that would later bring cries of regulatory excess were already under way when Magnuson-Moss passed. (The FTC had asserted, and by the time the legislation was enacted the U.S. Court of Appeals for the D.C. Circuit had held, that the agency already possessed the power to promulgate rules.) The FTC’s activism in the 1966–76 period must be seen as a coordinated effort between the agency and its oversight subcommittee.
The FTC’s activism in 1966–76 must be seen as a coordinated effort between the agency and its oversight subcommittee.
Finally, Michael Pertschuk, the controversial FTC chairman who took office in early 1977, had served as general counsel and chief of staff of the Senate Commerce Committee and had written much of the new legislation given to the FTC to administer during the 1970s. His appointment as chairman was the natural culmination of a decade of cooperation between the commission and its overseers in the Congress. The irony is that just as he became chairman, support in Congress for this type of policy began to disappear.
Statistical Analysis, 1966–1979. Following the 1976 election, there was a nearly complete turnover in the membership of the FTC’s Senate oversight subcommittee. As the table shows, in 1976 the five most senior members of that body had served an average of more than ten years each, whereas in 1977 only two of the subcommittee’s six members had served there previously and these two were among the members with the lowest seniority in 1976. Before this dramatic change, the average length of continuous service had steadily increased from the late 1960s into the mid-1970s, reaching 5.6 years by 1976. A year later, the average had fallen to 1.7.
More pertinent than the change in seniority, however, was the shift in the preferences of the subcommittee members. To test a senator’s support for consumer activism, our analysis uses the ratings that the Americans for Democratic Action (ADA) computes for measuring a legislator’s support for its program. We found that ADA scores alone predict 82 percent of the individual senators’ votes on FTC issues: the higher the ADA score, the greater the probability of voting for an activist FTC.
As the accompanying figure reveals, the subcommittee and its chairman had ADA scores that were well above the average for the full Senate throughout most of the 1966–76 period, but significantly below that average by 1979. What happened was that, while the full Senate’s average score, which had been remarkably stable for a decade, fell only a modest 8 points (from 46 to 38) between 1977 and 1979, the subcommittee’s score plunged 31 points (from 57 to 26) and the subcommittee chairman’s slid 18 points (from 50 to 32). Thus, the subcommittee and subcommittee chairman preferences shifted sharply away from consumer activism, a shift that should lead, according to the theory of congressional control, to a parallel shift in FTC policy.
MEMBERSHIP TURNOVER OF SENATE SUBCOMMITTEE ON CONSUMER AFFAIRS, 1976–77
*Subcommittee size reduced from fourteen in 1976 to six in 1977.
Source: Weingast and Moran, “Bureaucratic Discretion or Congressional Control,” p. 29.
AVERAGE ADA SCORES FOR SENATE, SENATE SUBCOMMITTEE ON CONSUMER AFFAIRS, AND THE SUBCOMMITTEE CHAIRMAN, 1966–79
[IMAGE]
Source: Weingast and Moran, p. 25.
Thus the sudden confrontation between the FTC and Congress in 1979 cannot realistically be considered—as it has served the congressmen’s political interests to portray it—the culmination of long and repeated agency disregard of congressional desires. Rather, it was the product of an unusual and temporary disharmony caused by (1) an abrupt subcommittee change which the agency, with many proceedings “in the pipeline,” could not immediately accommodate, and perhaps (2) the rare intransigence or myopia of a chairman who was the child of the recently changed committee Old Guard. In other words, the hearings and sanctions of 1979–81 were nothing more than the means by which a newly composed committee reversed the policies of its predecessor.
The high degree of congressional control before the subcommittee turnover in 1977 can be demonstrated convincingly by examining the relationship between the subcommittee’s preferences and the commission’s distribution of cases among the various laws it administers. These laws range from the Federal Trade Commission Act’s prohibitions against “unfair and deceptive trade practices” and “unfair methods of competition,” the Clayton Act’s restrictions on merger activity, and the multifarious prohibitions of new consumer protection statutes such as the Truth-in-Lending Act and the Magnuson-Moss Act to the Robinson-Patman Act’s small business protections and some statutes representing consumerist causes of bygone days like the Hobby Protection Act and the various statutes on fur, wool, and textile labeling.
To the extent that FTC policy decisions reflect congressional preferences, we should observe congressional influence over the commission’s caseload decisions. When the subcommittee is dominated by consumer activists (indicated by high ADA scores), the commission should bring a higher percentage of cases under the currently favored consumer protection statutes and a lower percentage of cases under Robinson-Patman and the other statutes that were its mainstays during the 1950s and early 1960s.
Our econometric estimates, which cover the years 1964–76, reveal that the relationship between congressional ADA scores and FTC caseload choice follows exactly this pattern. We found, for example, a strongly positive relationship between the ADA scores of subcommittee members and the proportion of cases brought under the Truth-in-Lending Act. (Indeed, our analysis allows us to predict that a major change in subcommittee scores—such as the 1977–79 drop of 31 points noted above—will yield a drastic reduction in the proportion of these cases.) A similar relationship was found between Truth-in-Lending cases and the subcommittee chairman’s ADA score, so that a fall in the latter reinforces the effect of a decrease in the subcommittee scores. The same pattern was found for the other two types of cases investigated—fur-wool-and-textile and Robinson-Patman (though there is a negative relationship between the subcommittee’s ADA rating and the proportion of these cases).
In sum, the evidence reveals substantial and regular congressional influence over FTC policy. This finding, in turn, supports our hypothesis that Congress may influence agency choices without engaging in systematic, continuous, and publicly held oversight hearings. The statistical evidence also implies that the commission is remarkably sensitive to changes in the composition of its oversight committee. In short, the evidence flatly contradicts the view that agencies are virtually independent of Congress and that agency policy making is best understood as the exercise of bureaucratic discretion.
… the evidence flatly contradicts the view that agencies are virtually independent of Congress. …
Conclusions for Regulatory Reform
An examination of one agency does not, of course, settle the question of bureaucratic control. But the FTC case is especially enlightening because the FTC was widely seen in the late 1970s as the quintessential runaway bureaucracy. Consequently, our findings suggest that the mechanics of congressional influence are both subtle and indirect, so much so that even careful observers may not perceive their influence. They further suggest that any difficulty in transmitting legislative policy into bureaucratic action lies not at the agency end of the control chain, but at the congressional end—in the committee mechanism that Congress uses to apply its influence. That is to say, the agencies reflect the interests of committees well enough, but the committees may be quite unrepresentative of Congress as a whole.
What are the implications for regulatory reform? First, reforms based on the traditional view that agencies exercise enormous bureaucratic discretion are unlikely to succeed. To the extent that congressional committees already play a regular, if subtle, role in agency decisions, reforms that tie agencies even closer to those committees may only exacerbate whatever problems now exist.
Consider, for example, the legislative veto, pushed by its supporters to curb regulatory abuse on the part of unresponsive agencies. But note that this device has already been contained in 193 statutes—without perceptibly improving the performance of the agencies involved. The only systematic investigation of existing legislative vetos, by Harold Bruff and Ernest Gellhorn, finds that their major effect is to augment committee power and control.
… the real problem … is a committee system that allows a few congressmen to protect the bureaus under their jurisdiction for the benefit of their constituencies.
Such measures not only ignore but also feed the real problem, which is a committee system that allows a few congressmen to protect the bureaus under their jurisdiction for the benefit of their constituencies. As political scientist Morris Fiorina maintains, one congressman’s out-of-control bureaucracy is another’s bread and butter. The problem is not that bureaucracies are unresponsive, but rather that the congressional committee system forces them to be responsive only to the interests represented by members of their oversight subcommittees.
The performance of regulatory agencies thus grows out of congressional institutions. Fiorina puts it this way:
In the end, the majority of our bureaucratic failures seem to have a large element of congressional failure underlying them. Wasteful, deceptive, disingenuous, paternalistic, and captive bureaucrats work in harmony with wasteful, deceptive, disingenuous, paternalistic and captive congressmen. … The bureaucrats catch a disproportionate share of the public relations flak, while the congressmen appropriate a disproportionate share of the political credit, in return for which they shelter the bureaucrats.
Intervention in the regulatory process, whether by the courts or by altering administrative procedures or structure, may thus have little beneficial effect. This explains in large measure why regulatory reform, pursued by every president since World War II, remains an elusive goal. The fundamental problem is not runaway bureaucracy but congressional politics. ■
Barry R. Weingast is assistant professor of economics and research associate at the Center for the Study of American Business at Washington University at St. Louis. Mark J. Moran is a graduate student in economics at Washington University.
Selected References
Harold H. Bruff and Ernest Gellhorn, “Congressional Control of Administrative Regulations: A Study of Legislative Vetoes,” Harvard Law Review 90, 1977, p. 1369.
Morris P. Fiorina, “Bureaucratic (?) Failures: Causes and Cures,” Formal Publication No. 43, Center for the Study of American Business, Washington University, St. Louis, October 1981.
Ernest Gellhorn, “The Wages of Zealotry: The FTC under Siege,” Regulation, January/February 1980, pp. 33-40.
Barry R. Weingast and Mark J. Moran, “Bureaucratic Discretion or Congressional Control: Regulatory Policymaking by the Federal Trade Commission,” Working Paper No. 72, Center for the Study of American Business, Washington University, St. Louis, January 1982.
James Q. Wilson, Politics of Regulation (New York: Basic Books, 1980). |
