| 摘要 | The current movement for regulatory reform has given new popularity to the legislative veto. That is the device where-by executive action authorized by statute is made subject to prior disapproval by one (the so-called one-house veto) or both (the two-house veto) houses of Congress. There is of course nothing remarkable about the ability of Congress to stay or revoke executive action which it has previously authorized; but the distinctive feature of the legislative veto is that it enables this to be accomplished by mere resolution—not formal legislation—thus avoiding the President’s veto power and (in the case of the one-house veto) the requirement of approval by both houses.
Although there are now well over a hundred examples of the legislative veto on the statute books, it is a relatively recent addition to our governmental machinery, first appearing in 1932 and only recurring a few times between then and World War II. In the past decade it has become more common—though almost always directed at a relatively narrow and specific executive action, such as the granting of pay raises to government officials. (To say it has become more common is not to imply that its validity has become accepted; presidents from Franklin Roosevelt to Jimmy Carter have questioned its constitutionality, as have many congressional leaders as well.) Several regulatory reform proposals now before Congress would attach the legislative veto to all agency rulemaking, as a means of “getting control of the bureaucracy.” Such action would elevate the device from an occasional oddity to a fundamental feature of our government. Before that occurs, it is worth considering whether the legislative veto accords with our political system and will provide the benefits that are promised.
The Role of Congress in Constitutional Interpretation
I often think that legislators—and the general public, for that matter—must be thoroughly sick of the spectacle of lawyers and legal scholars arguing that this or that feature of proposed legislation is contrary to the Constitution. We live in an age of “hair-trigger unconstitutionality,” and almost no result produced or about to be produced by the democratic process at any level of government seems immune from attack by some Scribe or Pharisee with a law degree on the ground that it contravenes the Basic Charter of our Liberties. Such facile invocation of the Constitution produces some destructive results. One is a massive increase in the power of the courts—which is why they so readily indulge it. Another is the debasement of what might be termed an erosion of the legislature’s constitutional ethic. Hearing solemn and contradictory invocation of the Founding Fathers so often, and knowing that the Supreme Court will in any event have the last word, the people’s representatives are increasingly inclined simply to produce the legislation which in their view has the most beneficial substantive effects, and leave the constitutional nit-picking to the courts.
They must resist that understandable inclination. Congress has an authority and indeed a responsibility to interpret the Constitution that are no less solemn and binding than the similar authority and responsibility of the Supreme Court—because they spring from the same source, which is the obligation to take no action that would contravene that document. Moreover, congressional interpretations are of enormous importance—of greater importance, ultimately, than those of the Supreme Court. It is not unrealistic to regard our constitutional law as divided into two strata: the higher (and more important) established by Congress, in its refusal to pass legislation contrary to our society’s understanding of the basic principles that govern our polity; and the lower established by the courts, in striking down those few congressional enactments that are, in the judges’ more restricted purview, contrary to the Constitution.
… congressional interpretations [of the Constitution] are of enormous importance—of greater importance, ultimately, than those of the Supreme Court.
And they are few. However activist the federal courts may have become with respect to state legislation, in the nearly two centuries of our republic only about 100 acts of Congress have been stricken down, in whole or in part. Congress is, in other words, the first line of constitutional defense, and the courts—even the activist modern courts—merely a backstop. Moreover, the character of what might be called congressionally applied constitutional law ultimately affects the character of the judicially applied stratum beneath it. This may or may not be desirable, but it is unquestionably true. As enduring congressional notions of constitutional limitations have changed—with respect, for example, to the scope of federal commerce-clause power, the permissible degree of governmental intrusion upon private economic activities, or the permissible extent of delegation of congressional authority—so also have the notions of the courts.
This amounts to saying no more (and no less) than that constitutional provisions subsist only as long as they remain not merely imprinted on paper, but also embedded in the thinking of the people. When our people ceased to believe in a federal government of narrowly limited powers, Congress’s constitutional interpretation disregarded such limitations, and the courts soon followed. What is involved in a consideration of the legislative veto is the question whether the doctrine of separation of powers and the principle that laws can only be passed by affirmative vote of the people’s representatives are ready for a similar fate.
The Constitutional Text
The Constitution, in Article I, section 7, clause 2, gives the President a major role in the enactment of laws—by providing that all bills passed by Congress are subject to his disapproval, which can be overridden only by two-thirds vote of each house. That “bill veto” clause is succeeded by the following provision, which is central to the issue here under discussion:
Every order, resolution, or vote, to which the concurrence of the Senate and House of Representatives may be necessary (except on a question of adjournment) shall be presented to the President of the United States; and before the same shall take effect, shall be approved by him, or being disapproved by him, shall be repassed by two thirds of the Senate and House of Representatives, according to the rules and limitations prescribed in the case of a bill.
The purpose of this provision, as confirmed by accounts of the debate at the Constitutional Convention, is to prevent Congress from evading the President’s legislative role (as some state legislatures before 1789 had evaded gubernatorial veto powers) by simply acting through measures that are not called “bills.” It was meant to ensure presidential participation in all lawmaking, under whatever form it might disguise itself. The validity of the legislative veto, then, turns quite simply upon whether it in reality constitutes lawmaking. I think it difficult to maintain that it does not.
Laws can generally be divided into two types, which sometimes overlap: The first confers rights and obligations upon private individuals. The second—which is more common at the federal level—confers powers or imposes prohibitions upon executive (or, more rarely, judicial) officials. The Corps of Engineers, for example, is given authority to construct dams or is prohibited from taking any action which would adversely affect endangered species of wildlife. Now it is undeniable, I think, that the withdrawal of executive power previously conferred—that is to say, the prohibition of executive action which is currently legitimate—can only be done by law, that is, by that process which invokes the presidential veto power. Reduced to its essence, the argument for the legislative veto is that this situation changes when the empowering law itself states, in so many words, that the power conferred may be withdrawn by Congress, without further presidential participation in the process. Then, the argument goes, the power initially conferred contains within it this built-in limitation of congressional retractability, so nothing is later taken away, and therefore no new “law” is really made.
This seems to me insupportable. If the Founding Fathers, who were so concerned about protecting the President’s participation in the legislative process that they added to the “bill veto” clause the super-cautious language of Article I, section 7, clause 3, succeeded by all their efforts in requiring no more than that Congress say in advance, for each law it passes, that the President will have no part in the amending or repeal of such law-then they were feckless indeed. Consider the possibilities: “There is hereby created a Department of Education, which shall have powers X, Y, and Z, but only for so long as Congress approves such Department and powers; and such department and powers shall cease if eliminated by vote of either house.” Since, we are told, the limitation is built into the original grant, the “law” is not “changed” when Congress abolishes—without any presidential participation—the new department or its functions.
The proponents of the legislative veto respond that, whatever may be its validity with respect to the sort of example I have given, its application to rulemaking involves different considerations, because rulemaking is really “legislative” activity. (This is sometimes softened to “quasi-legislative”—the prefix conveying nothing except a sense that the speaker does not have the courage of his convictions, or does not have the convictions of his adjectives.) There are several replies to this. The first is that the premise is quite simply wrong. Rulemaking is not, as some seem to think, a function dubiously conferred upon the executive only in very recent times because an overly pressed Congress did not have time to attend to all of its own normal business. To the contrary, rulemaking has been an executive function from the beginning. As stated by the 1941 Report of the Attorney General’s Committee on Administrative Procedure:
The promulgation of general regulations by the executive, acting under statutory authority, has been a normal feature of Federal administration ever since the Government was established. The first Congress provided that traders with the Indians should be licensed and bonded to observe “such rules, regulations, and restrictions” as might apply, including “such rules and regulations as the President shall prescribe.” In 1796 the President was given authority to establish regulations for estimating the duty upon goods, the cost of which was stated by the importers in depreciated foreign currencies. In 1809 conclusive effect as against shippers and shipowners was given to instructions and regulations of the President authorizing the collectors of customs to refuse permission for loading cargoes, to detain vessels, and to seize goods in the enforcement of the embargo acts.
The fact is that rulemaking is no more “inherently” a legislative function than is the fixing of rates (which used to be done by Congress before the Interstate Commerce Commission was established) or the granting of claims against the government (which used to be done by Congress before the Court of Claims was established). And the fact is that, with very few exceptions, all of the decisions made by the executive branch could be made instead by Congress itself (in which event they would be “legislative”) and have become “executive” functions only because Congress has chosen to commit them to the second branch. For example, the Corps of Engineers’ decision to construct a bridge of steel rather than wood is assuredly an “executive” decision—but only because Congress has left that decision to the corps instead of itself specifying the construction material. Surely a law directing the erection of a particular steel bridge would not be an “inherently unlegislative” enactment. There is, of course, a category of decision making which should not be delegated to the executive—and indeed, whose delegation would be an unconstitutional abdication of peculiarly “legislative” responsibilities. But the line defining that category has nothing to do with the line between rulemaking and other executive activities. Rulemaking authority, if conferred with adequate standards, is perfectly valid and perfectly executive; and adjudicatory or licensing authority, if conferred without such standards, is an unconstitutional delegation of legislative power.
The second reply to the argument that the distinctively “legislative” character of rulemaking somehow validates the legislative veto in that context is that, even if the premise were correct, it has no bearing upon application of the presidential veto provisions of the Constitution. The issue for that purpose is not whether the executive is “legislating” in the adoption of rules, but whether the Congress is “legislating” in preventing the effect which those rules otherwise would have. As explained earlier, it clearly is. Therefore, the lawmaking provisions of the Constitution apply—which include participation by the President under Article I, section 7.
“But,” urge the supporters of the legislative veto, “there is participation by the President. It is the President or his subordinates who propose the rules that the Congress may veto. He may in effect veto them by not proposing them. It is simply legislation in reverse!” To begin with, this argument is simply not true with respect to that substantial proportion of all rulemaking conducted by independent regulatory agencies, such as the Federal Trade Commission, the Securities and Exchange Commission, and the Interstate Commerce Commission. With respect to such agencies, the President has no say whatever in whether rules are proposed. More fundamentally, however, the action to which the President’s Article I, section 7 veto power attaches is not the agency’s issuance of regulations but congressional change of the law—and that change occurs when authorized executive action is invalidated, whether or not Congress asserted in the enabling statute the unilateral power to invalidate. And finally, I simply cannot appreciate the force of the assertion that this manner of proceeding is valid because it is simply the constitutional process in reverse. Would we say that an automobile is not defective simply because its four-speed transmission propels it backward rather than forward? Or that a plastic surgeon has done his job well if he produces perfectly regular features at the back of one’s head? The point is that the Constitution does not provide for legislation in reverse; it provides for legislation in “forward.”
The point is that the Constitution does not provide for legislation in reverse; it provides for legislation in “forward.”
There is one other argument against the applicability of Article I, section 7, clause 3, which should be met—because, although it is obvious that it is wrong, it is not obvious why it is wrong. This is the argument that the provision could not possibly stand in the way of the one-house veto, since by its terms it applies only to action “to which the concurrence of the Senate and House of Representatives may be necessary.” That argument is obviously wrong because it assumes that the Founding Fathers were careful to preserve the presidential veto as a check upon disguised legislative action by both houses of Congress, but were quite willing to let a single house proceed unchecked (or too dull to provide for that obvious eventuality). The explanation of the apparent incongruity is that, while, in the case of the one-house veto, action by the other house may not “be necessary” under the terms of the applicable statute, it is necessary under the terms of the Constitution which, in its very first section, provides that “all legislative powers herein granted shall be vested in a Congress of the United States, which shall consist of a Senate and House of Representatives.” In other words, legislative action taken by means of a one-house veto appears, superficially, to avoid the prohibition of Article I, section 7, clause 3, only because it simultaneously runs afoul of the fundamental requirement of bicamerality.
The Political Effects
More important, however, than the incompatibility of the legislative veto with the text of the Constitution are what might be termed its unconstitutional effects. Many of those who support application of the device to all rulemaking do so out of genuine commitment to democratic principles, in the belief that it will put an end to the government by bureaucracy which has come to characterize our society. They are profoundly mistaken. For the legislative veto constitutes a ratification and reinforcement, not a repudiation, of the governmental policies that have led to our present state. Its proposed extension to all rulemaking distracts attention from the real problem and offers instead a scapegoat upon which the wrath of the public can be exhausted and the zeal for genuine reform ultimately dissipated.
In the prologue of one of the pending bills, which accurately sets forth the legislation’s underlying premises, we are told that the federal agencies have been guilty of a “steady usurpation of quasi-judicial and legislative powers” and of an “often arbitrary exercise of powers granted to the executive branch.” And Congress, we are told, “which is closest and most responsive to the needs and will of the citizenry, must reassert on behalf of the people the constitutional right of the people to be secure from the unreasonable and arbitrary actions of the executive branch which result from [such] usurpation.” But in truth there has been no pattern of “arbitrary” action; for arbitrariness in this context implies departure from an established standard—and, in the relevant areas, Congress has shrunk from providing standards for many years. Nor has there been “usurpation” of power; for that implies that power has been yielded unwillingly—and Congress has in fact rushed to abdicate for many years, often conferring upon the agencies authority which the incumbent administration itself did not desire. The main culprit, in other words, has not been the agencies but Congress itself.
Consider, for example, the Consumer Product Safety Commission (CPSC), one of the newer agencies which has been the object of considerable public criticism. It was established in 1972 with jurisdiction over virtually every product distributed in commerce and with a rulemaking authority no more specific than to prescribe whatever requirements as to “performance, composition, contents, design, construction, finish, … packaging, … warnings or instructions” may be “reasonably necessary to prevent or reduce an unreasonable risk of injury associated with such product.” Or the Occupational Safety and Health Administration (OSHA), which in 1970 was given rulemaking authority no more definite for most purposes than to require “conditions, or the adoption or use of … practices, means, methods, operations, or processes, reasonably necessary or appropriate to provide safe or healthful employment and places of employment.” Or consider Title IX of the Education Amendments of 1972, which generally provides that “no person … shall, on the basis of sex, be excluded from participation in, be denied the benefits of, or be subjected to discrimination under any education program or activity receiving Federal financial assistance”; and which authorizes and directs each agency providing such assistance “to effectuate [such] provisions … by issuing rules, regulations, or orders of general applicability.” Well, who can oppose the prevention of “unreasonable risks of injury,” or the provision of “safe and healthful places of employment,” or the elimination of “sex discrimination”? We can all embrace these platitudes. But what do they mean? Does the prevention of “unreasonable risks of injury” from children’s pajamas require that they be utterly fireproof, at an enormous increase in consumer cost? Does the provision of a “safe place of employment” require split toilet seats, or the relocation of all fire extinguishers so that they are precisely X inches from the floor? And does the elimination of “sex discrimination” require that all-boy or all-girl athletic teams be abolished? These are the sorts of issues that lay beneath the platitudes when these pieces of legislation were passed, and Congress chose to leave them to the agencies (and perhaps ultimately the courts) to resolve.
… what may be termed “consensusless” programs … have increased enormously in number and are now well beyond monitoring by our political representatives, either in the White House or on Capitol Hill.
Congress has been behaving in this fashion for much of the past fifty years, assigning to the agencies vast and important decision-making authority—not only in rulemaking, but in adjudication as well—which does not require merely the application of scientific analysis, or logic, or deductive reasoning, but involves essentially political judgments, as to which there are no “right” or “wrong” answers. This may work well enough in those areas where a substantial political consensus exists. And even in the absence of such consensus, it may work well enough when the total number of programs is small enough that the President and Congress can, realistically, keep an eye upon them and correct those political judgments that are simply out of accord with the people’s desires. What has happened since the 1960s, however, is that what may be termed “consensusless” programs, such as those administered by CPSC, OSHA and, for another example, the Environmental Protection Agency, have increased enormously in number and are now well beyond monitoring by our political representatives, either in the White House or on Capitol Hill.
Even this analysis omits an aggravating element: With respect to many of the important political judgments that it has delegated, Congress would not wish to monitor even if it could do so. I refer to those judgments that were delegated in the first place simply because they were “too hot to handle.” The sex discrimination prohibition is a good example. It was apparent when that law was passed that its concrete application would arouse heated controversy over precisely such issues as all-male sports, unisex dorms, and even unisex toilets. Was it not more appealing for Congress to take the high road, winning approval from all sides by being against “sex discrimination”—and leaving to the agencies the inevitable alienation of one or another constituency which accompanies the act of giving that term content? And will not the same self-protective motivation which induced Congress to abstain from more specific legislation in the first place also induce it to abstain from revising agency action later on?
But, to come to the point, what does all this have to do with the legislative veto? It suggests, I think, that there is nothing—absolutely nothing—the legislative veto can do to solve the basic problem. There are only two powers which the legislative veto confers upon Congess that could not be exercised through the ordinary legislative process. The first of these is the ability to avoid the presidential veto. Leaving aside for the moment the fact that that is unconstitutional, is it at all relevant to the problem of controlling the agencies? Has the difficulty really been that Congress has tried repeatedly to reverse the results of agency rulemaking through legislation but has been stymied by the President? I am not aware of a single instance. And the second and last distinctive power conferred by the legislative veto—if it is a one-house veto—is the ability of both the House and the Senate separately to overturn rules without the concurrence of the other body. But again, has this been the problem? Are the files of Congress filled with dead bills overturning agency rulemaking that have been passed by one house only to be spurned by the other? Not to my knowledge. The problem has been, quite simply, that both houses have had neither the time nor the inclination to review agency rulemaking, just as they have had neither the time nor the inclination to write more detailed legislation in the first place, which would render the most significant rulemaking unnecessary. The proof of this thesis is readily available. All rulemaking of HEW’s Office of Education has been subject to legislative veto since 1974. Has that agency become the model of responsiveness after which the rest of the government should now be patterned? Hardly—because the legislative veto has had no appreciable effect in ensuring democratic control.
Instead of government by one bureaucracy, we will have government by two—hardly a step towards more democratic government.
In fact the legislative veto will be worse than ineffective in solving the problem of government by bureaucracy. It will ultimately both increase the problem and breed other threats to democratic self-government. The one scientific study of the effect of existing legislative veto provisions that I am aware of—conducted by Professors Harold Bruff and Ernest Gellhorn for the Administrative Conference of the United States—concluded that “much settlement of policy occurred in behind-the-scenes negotiations between the staffs of the [relevant congressional] committees and the agencies.” Of course that would be the result! Our elected representatives clearly will not have time to review the torrent of regulations pouring forth every month, so that task will be entrusted to a shadow rulemaking bureaucracy on Capitol Hill. That will become the real body which agency bureaucrats have to please in order to forestall congressional veto. Instead of government by one bureaucracy, we will have government by two—hardly a step towards more democratic government.
Another deleterious effect of the legitimation of the legislative veto will be an increase in that very practice of congressional delegation of vague and standardless rulemaking authority which has placed us in our current predicament. That is to say, the delusion that it will be able to control the agencies through the legislative veto will render Congress all the more ready to continue and expand the transfer of basic policy decisions to the agencies. It is significant in this regard that some of the most prominent examples of legislative vetoes enacted in the past were proposed by the executive branch itself—to induce the congressional transfer of power which would otherwise not have been accorded.
And finally, the legitimation of the legislative veto will enable continuation and expansion of the recent practice of adopting major measures by a process which preserves congressional control while relieving the people’s representatives of the embarrassment of voting. This process was exemplified by the Executive Salary Cost-of-Living Adjustment Act, which in February of 1977 enabled the members of Congress (together with high-level executive and judicial officials) to obtain a 30 percent pay raise without ever having to stand up and vote for it. This was accomplished by the process of giving the President the power to prescribe salaries (a form of rulemaking)—subject, however, to congressional veto. The latter feature was of course essential to the scheme, since Congress would not have been willing to accord the President authority over salaries without retaining such a whip-hand. What happened in 1977 was that a President who had already lost his bid for reelection (and therefore had little political capital to lose) proposed the massive increase, and Congress did not say “nay.” There was an angry public outcry, as there well should have been. Not so much, I think, because the increase was thought to be undeserved, but because of the fashion in which it had been accomplished. Such legislation by inaction is not the system our Constitution envisions. If Congress is willing to commit a matter to the executive, well and good; but if Congress wants to retain control of the matter, and thereby admits that it has not completed its legislative function—then it must act by voting, not by simply standing silent.
A similar motivation has prompted the adoption of legislative veto schemes for executive reorganization. Reorganization is terribly difficult if Congress must affirmatively vote for it. The interest group pressures which must be withstood to abolish or even restructure an agency are enormous. But that simply suggests that the issue is a very important one; and congressional subjection to those pressures, before standing up to be counted in an honest-to-goodness vote, is the very essence of the democratic process! By enabling Congress to have its way—that is, effectively to control the outcome—without affirmatively accepting responsibility in a record vote, the legislative veto is an egregious subversion of the democratic process.
Over the years, most of the criticism of the legislative veto has originated in the executive branch, and thus has focused on the propensity of the device to alter the constitutional balance of power between the first and second branches of government. In my view, that is much less fearful than its propensity to alter the balance of power between those two branches combined and the people. Its utility in ensuring agency responsiveness to the people is negligible; but it is an excellent mechanism for enabling the President and the Congress to facilitate the passage of unpopular laws by eliminating the congressional burden of having to vote for them.
[The legislative veto has been criticized for its propensity] to alter the constitutional balance … between the first and second branches of government. In my view, that is much less fearful than its propensity to alter the balance … between those two branches … and the people.
I am in the greatest sympathy with the objective of revivifying what appears to be a failing system of self-government in the United States. The legislative veto, however, is a “quick fix” which simply will not work. The same can be said for Congress’s frantic search for new procedural constraints upon agency action. Better procedures are of course always desirable. But the main illness from which we suffer is not that the agencies are acting hastily, or unintelligently, or without adequate information—nor even that they are reaching the “wrong” decisions, for most of these matters that concern us have no objective “right” or “wrong.” The main problem is that the agencies have been assigned too many tasks requiring judgments that are of an essentially political nature and that ought to be made by our elected representatives. And the only remedy, if we really want a remedy, is to take some of those tasks away and to perform them instead by legislation, or not to perform them at all.
To offer the legislative veto to our citizens as a cure for the very real alienation from government that now besets them—for their feeling of “being governed” rather than governing themselves—is to delude them. Instead, the members of Congress should tell them the hard facts, perhaps in a letter to constituents that reads something like this:
Fellow Citizens:
There is abroad in our land the feeling that we no longer control our government, but it controls us, through thousands of law-making functionaries in every field of life who are effectively beyond popular control. That feeling, I am sorry to tell you, is well founded. And the cause is quite simply that your Congress has over the years delegated so many policy judgments of the sort once made by your elected representatives to the executive agencies that by now neither the Congress nor the President can realistically monitor or supervise the results.
The system is overloaded. We are now at the point at which each major new program entails an overall diminution of democratic control.
We in Congress have done this not maliciously but with the best of intention. We have wanted to give all of you a clean environment, a safe working place, safe consumer products, protection from deceptive merchandizing—and many other protections which were in earlier years the responsibility of elected bodies in your cities, counties, and states. You have evidently approved what we have done, since you have continued to elect us on the basis of these programs.
But the time has come to tell you that all these benefits c |
