experience, which gives them a "sense of what is practically enforceable" and enables them to "plan a program of development." 10

Third, because of the necessary and proper slowness of legislative procedure and the fact that Congress is not in continuous session, it results that its laws cannot be changed so readily, to correct mistakes and to meet changing conditions, as can the regulations of administrators.

Fourth, if a legislative rule is both specific and unworkable, the administrator is in a dilemma. He must either try to work the unworkable, and thereby invite litigation and defeat of the real purpose of the statute, or he must evade or ignore the letter of the law. Legislative details often multiply administrative difficulties and tie the hands of the administrator in red tape. On the other hand, the administrator in working out the details can at first make flexible rules, increase their rigidity by degrees, and meanwhile win over the interests to be regulated by tact and diplomacy, by mixing "suasion with command." Such a policy, for example, has been followed by the Securities and Exchange Commission in the regulation of the stock exchanges.

11

Fifth, although the administrator should not become a slave to precedent, if he is "sufficiently detached from the strife of interest and imbued with a sense of professionalism," he will develop "objective responsibility" to precedent, to expert opinion, and to professional ethics. In a matter like quarantine, he is limited also by the techniques of his science. These "inherent checks which in professional organs evolve principle out of constantly recurring action" enable the permanent official to "substitute principle for discretion", whether he is evolving regulations or making case-to-case decisions. Specifically, under a proper personnel system, an administrative agency is better able than Congress to utilize technical information and guidance, both in matters of substance and in matters of form and technique. This is no reflection upon Congress, for its proper role is not to deal with such matters but to translate into law the objectives of public opinion which its administrative agents are then to effectuate in action. It is only when Congress invades the administrative field that the special disadvantages of statutory prescriptions as compared with executive determinations apply.

Executive Regulations and the Rule of Law 12

By the rule of law is here meant the system whereby governmental action in the individual

10 Ernst Freund, "Substitution of Rule for Discretion in Public Law," American Political Science Review, IX (November 1915), pp. 666-76. Except where otherwise indicated, quotations in this section are from Freund.

11 Carl Friedrich, "Responsible Government Service Under the American Constitution," in Problems of the American Public Service (New York: McGraw-Hill, 1935), pp. 37-38.

12 Cf. generally Hart, The Ordinance Making Powers of the President, Ch. II.

case is guided by uniform rules to which the administrator may be required to conform by appeal to a tribunal which is impartial in the sense of being independent of the administrator.

The rule of law is frequently contrasted with discretion; but this is misleading. The rulemaking power is itself discretionary; it involves discretion as to a uniform rule. Nor can discretion be eliminated from the application of such a rule to particular cases. The effort would be impracticable and, if it were attempted, would stifle initiative by red tape.

The purpose of the rule of law is rather to minimize arbitrary, discriminatory, or hit-ormiss exercise of authority. If uniform rules work hardships, there will be general criticism, and the rules will be changed. Even if such hardships endure for a while, they apply equally to all, and hence seem less unjust than discriminatory action. The rule-making authority thus has inherent checks that do not operate in the exercise of power in individual cases. It is there that arbitrariness creeps in, personal spite finds expression, and favoritism and discrimination are readily possible. For these dangers there are only two safeguards: First, the provision of uniform rules to guide, canalize, and to some extent narrow down the otherwise untrammeled discretion of the administrator and, second, the provision of appeal to an independent tribunal authorized to protect the individual against orders that do not conform to those rules.

On the whole, then, discretion as to uniform rules is less dangerous than discretion as to individual cases and discretion as to individual cases is less dangerous if guided by preexisting rules than if untrammeled.

It is worth repeating that it does not follow that such rules can or should so narrow discretion in the issuance of particular orders as to eliminate it. Human experience is notable for the recurring emergence of new or even unique. situations; hence, there must always be room, under proper safeguards, for flexibility and individualization, for the assumption of responsibility for the exercise of sound judgment. To be sure, a new situation may call merely for the definition of a new class and the drafting of a new rule suited to that class. As experience develops, furthermore, cases that at first seemed unique may begin to fall into some general classes. But human experience is too varied and variable to be reducible to formulas without curtailment of progress. This is especially true in an age of transition.

The appropriate degree of discretion in particular cases depends upon circumstances. Thus, complete executive discretion is needed for matters like pardons, where no known rules are considered applicable and each case must be judged on its own merits. Even in the administration of rules, varying degrees of discretion must be permitted in particular cases.

Especially in the early stages of regulation, it may be necessary to proceed from case to case under a very vague mandate, after the traditional manner of the common law courts.

Nevertheless, the rule of law has important general bearing upon the rule-making power. In the first place, it means that Congress should make every effort to define a policy rather than merely to dump a regulatory problem in the lap of an administrative agency with no more guidance than that it shall act in accordance with public interest, convenience, or necessity. This conclusion is not inconsistent with the idea that Congress should state a general policy rather than freeze a mass of specific details into the statute. An example of what is meant is found in the Wagner-Connery Labor Relations Act. In this act Congress did not merely tell the Labor Relations Board it should decide disputes concerning the rights of labor to collective bargaining in accordance with the public interest. It defined in general terms several unfair labor practices and empowered the Board to eradicate them. In so doing, it steered a middle course between attempting to anticipate every possible situation and expressing no policy except that in an empty formula. This is sound legislative technique, quite aside from the merits of the policy itself, for the rule of law means first of all that the legislature should seek to express the objectives which the administrator is to seek.

This conclusion cannot be applied, however, as a rigid formula. When the definition of the functions of the Federal Trade Commission was before Congress, one view was that the statute should list a long series of unfair methods of competition; while the other view, which finally prevailed, was that Congress should not define the term. Without assuming to pass judgment on this issue as of 1914, it may be admitted that when a new type of regulation is undertaken, there may be general agreement only on the necessity of tackling the problem and not on its essential features, nor on any very clearly defined objectives. It may seem best to set up an agency to explore the problem by trial and error methods. All that can be said in such a case is that Congress should first explore carefully the alternatives, and that after some years of experience it should reconsider the whole problem.

There is still another difficulty. At some crucial point it may be politically inexpedient for Congress to define a clear-cut policy. Thus when the Labor Relations Act was passed, a split was emerging within the ranks of labor itself on the issue of craft versus industrial unions. Congress left to the Board the task of selecting the appropriate bargaining unit in each case as it arose. To do more would have been impracticable at a time when the issue was still an open one.

It is thus frankly admitted that the ideal which has been set forth must yield to practical

circumstances, but this does not destroy its value as a goal to be sought within the limits of practicality.

The same conclusion applies to the second point at which the rule of law bears upon the rule-making power. The rule of law suggests that when Congress defines a general policy and delegates discretionary power to execute this policy, it should always at least consider granting such power in the form of authority to make uniform rules and regulations which will make its general policies more specific before they are applied to individual cases. For in this way the administrator generalizes his own policies under the congressional mandate, and the citizen has more specific guidance as to what his obligations are.

Viewed in this light, the rule-making power is not in conflict, but rather in conformity, with the rule of law as applied to the actual problems of present-day government. It is a means of introducing the rule of law at the administrative level.

The extent to which the administrator can concretize legislative abstractions in the form of rules and regulations, which have the force of law and bind him until they are modified, must necessarily vary with the circumstances of each type of regulation. Some problems involve the constant recurrence of the same kind of situation. Again, at many points in the law, what is above all desired is a rule that is definite, even though necessarily arbitrary, and which thus furnishes a certain guide to businessmen. This is especially the case where the rule may readily be changed by an administrator if it proves unsatisfactory. Abstract justice has nothing to say about the rule of the road, but public convenience requires that all drivers of vehicles be habituated to driving on the right side of the road, and be required to do so. Similarly, it may be that the Federal Trade Commission should establish a definite percentage of wool (with a definite tolerance) as necessary to keep the advertising of cloth as "all wool" from being an unfair method of competition. If so, it would be well for Congress to delegate the power to issue regulations having the force and effect of law in regard to such matters as could be handled in this way.

On the other hand, full weight must be given to the fact that a matter like rate making is of a different order. It may be that Congress could define the factors which constitute a reasonable rate, and could delegate to the Interstate Commerce Commission power to render these factors even more concrete, not only for its own guidance but for that of the railroads. But in the last analysis it is probably wise, and indeed necessary, to leave much to the sound judgment of the Commission in each case.

In any event, it is more important to guide and to canalize discretion in the particular case than to narrow it down. To leave the application of rules to sound discretion is safer in the

long run than to hem in the administrator by a network of rules that reduce him to an unthinking automaton. What is desired is the golden mean, and that varies with circum

stances.

It is worth repeating, however, that the rulemaking power is in entire conformity with the idea of the rule of law as applied to twentieth century American government. That idea can no longer mean the rule of detailed statutes. It means rather, first, that Congress should guide the administrator by as careful a definition of general policy as is possible, and, second, that the gap between congressional mandate and administrative orders in particular cases be bridged, though by no means narrowed to mere red tape, by the rule-making power. It is the problem of Congress to decide in each case the scope of rule-making and hence the scope within which particular discretion may operate most efficiently. This cannot be decided by any rule of thumb.

One final consideration may conclude the discussion of executive regulations and the rule of law. If Congress gives an administrative orgán a mere empty formula for its mandate, and then simply tells it to apply this formula by the case-to-case method, the courts are apt to interfere; for interpretation of this formula is a question of law on which the courts have the last word. The famous Gratz case 13 is an example. But if Congress delegates to the administrative organ the power to supplement its policy by complementary rules and regulations having the force and effect of law, then the courts will interfere only if the delegation is unconstitutional or the rules and regulations are ultra vires.

This is an especial reason why it is well for Congress to consider extension of the rulemaking power into still other areas. It is not in conformity with the rule of law, as properly understood, to have the meaning of the law uncertain until the courts finally interpret it retroactively. Furthermore, judicial interference with what is really an area of development of new public policy is certainly undesirable. The courts are properly the guardians of private rights and the protectors of the individual against arbitrary administrative action; but this means that they are not suited to the determination of public policy. Their primary function makes them tend to maintain the status quo even where new standards of conduct need to be evolved in the public interest. When they dictate public policy, they necessarily do so in a negative, hampering way. This is a burden which should not be imposed upon them, if they are to be protected against harsh criticism and prevented from hamstringing the development of new social standards.14

13 Federal Trade Commission v. Gratz, 253 U. S. 421, 427 (1920).

14 Frederick F. Blachly and Miriam Oatman, Administrative Legisla tion and Adjudication (Washington: Brookings Institution, 1934), pp. 237-50.

Interpretative Regulations as a

Means of Increasing the Certainty of the Law

The traditional technique of regulation in the United States has been the enactment of a statutory rule and enforcement of that rule by the courts. Enforcement was initiated by the plaintiff in a private suit or by the Department of Justice in seeking a criminal indictment.

In the simple days of the agricultural era this procedure worked well enough. The statute was apt to be specific and the language understandable by all. Even an an abstract standard got content from being related to situations that came within the range of experience of the average citizen. In this it resembled such a common law standard as reasonable care, which got its meaning from the customs of the community. Under these circumstances, the jury system was admirably adapted to the function of holding a man to the generally accepted standards of the time and place.

In the industrial era, however, this traditional technique has developed grave defects. Statutory requirements relate more and more frequently to specialized relations that are outside the range of the layman's experience; and such requirements are often necessarily stated in highly technical language. A striking example is the Securities Act of 1933.

No statute can be self-explanatory. The framers word it with certain types of situation in mind. They cannot anticipate all variations from type, especially since new situations are constantly emerging from the kaleidoscopic pattern of events that constitutes the twentieth century economy.

The process of statutory amendment is not adequate for clarification, for it is slow and intermittent. Each occasion for amendment, moreover, opens the gates to competing pressures as well as to changes carefully devised in the light of administrative experience. In the interim between amendments, uncertainty prevails; and after an amendment is passed, the first question that arises is: What does it mean? Does it cover this or that situation?

Nor is this uncertainty overcome by having the courts years later decide retroactively what the statute means as applied to private or administrative action already taken. Judicial interpretation is as inadequate as statutory amendment in telling those affected what their obligations are.

In such cases, moreover, the jury system may give a hit-or-miss application of a legal ruleone which may be not only discriminatory in its net effect but also productive of uncertainty. What one jury may do is no certain indication of what the next will do, especially when the legal standard relates to something about which the jurors know very little.

This is no light matter. The public interest may require regulation; but the businessman needs to be able to discount the future. Cer

tainty in the law is essential to the conduct of his business, for under a system of private enterprise uncertainty as to future governmental action has a depressing or deflationary effect upon trade. There thus emerges from the technique of regulation this serious dilemma: Governmental controls are necessary in the public interest, but they slow down business activity and thus hamper reemployment.

This seemed the heart of the reasonable complaints about the Securities Act of 1933, as originally enacted: The issuance of securities was alleged to be hampered by uncertainty as to the liabilities of those responsible for issuance. It is the most frequent complaint about all new regulatory statutes. Indeed, there is some reason to believe that such statutes tend to simmer down, especially when conservatives are at the helm, to rather ineffective forms of control, for the simple reason that it seems impracticable to hold business to standards which it can learn only retroactively, after there has been litigation.

This cannot be the final answer when control is required by the public interest. What is needed is a technique that will break through the dilemma with a minimum of change in traditional methods.

The original technique of regulation did not require elaborate administrative machinery, except for tax collection, for a service like the postal system, and for a few other matters. But as regulatory problems increased in number and complexity, judicial enforcement proved ineffective. The courts acted only when cases were brought before them. The procedure of the courts was slow. Juries did not produce uniform application of statutory standards.

As a result, Congress began to create agencies which would function continuously to enforce its regulations. Even this, however, did not solve the problem of uncertainty. An agency directly charged with the administration of a statute can issue rules and regulations having the force of law only within the limits of specific statutory delegation. Though it must necessarily interpret the statute in the process of applying it, such interpretations do not have the force of law. The courts give them great weight as a form of administrative practice; but they may find that as a matter of law the interpretations are incorrect. The practical effect is to create confusion and to keep the interpretations from guaranteeing certainty to businessmen.

15

What is needed is a satisfactory middle term between legislation and final adjudication, both of which are essential parts of the American system. The problem of finding such a middle term arose in connection with amendment of the Securities Act of 1933. As a result, Congress adopted a new technique which seemed so sensible that it was later incorporated in the

15 Cf. United States v. Hill, 120 U. S. 169 (1887).

amendments to the Securities Exchange Act of 1934. In this form it reads:

The Commission and the Board of Governors of the Federal Reserve System shall each have power to make such rules and regulations as may be necessary for the execution of the functions vested in them by this title, and may for such purpose classify issuers, securities, exchanges, and other persons or matters within their respective jurisdictions. No provision of this title imposing any liability shall apply to any act done or omitted in good faith in conformity with any rule or regulation of the Commission or the Board of Governors of the Federal Reserve System, notwithstanding that such rule or regulation may, after such act or omission, be amended or rescinded or be determined by judicial or other authority to be invalid for any reason. 16

It is proposed that this technique be considered in connection with every similar regulatory statute. Such a statute might well provide that the agency charged with its administration may, upon its own initiative, or upon application, issue interpretative regulations defining and interpreting any and all the terms and provisions of the act and their applicability to any generally defined situation; and that no provision of the act imposing any liability, whether civil or criminal, shall apply, in the absence of collusion, to any act done or omitted in good faith in conformity with any such interpretative regulation, notwithstanding that such regulation may, after such act or omission, be amended or rescinded or be determined by judicial or other authority to be invalid for any reason.

At first glance this may appear to be a startling delegation of power. But several things are to be remembered.17 This is no blanket delegation of power to make regulations having the effect of law such as was contained in the National Industrial Recovery Act. Congress may legislate in as much detail as it sees fit, and still the problem of uncertainty arises. Furthermore, the proposal merely delegates the power to do by general rulings what the agency charged with the administration of the statute must actually do every time it applies the act to a particular case. "Of late", says Comer, "the judiciary has come out with the definite pronouncement that the authority to administer an act gives the Executive the implied authority to interpret it, since interpretation is necessary to the performance of its duty of administration." 18 General interpretations are certainly as permissible as particular interpretations, especially when specifically authorized by statute. It may be added that the courts look with more favor upon delegations which involve the privileges and immunities of citizens than upon those which involve duties and liabilities.

The only novel feature of this new technique is the formal authorization of general interpretations merely upon application or upon the

16 Public, No. 621, 74th Cong. 2d sess.

17 See Walter Wheeler Cook, "Certainty in the Construction of the Law", American Bar Association Journal, XXI (January 1935), p. 19. 18 Comer, op. cit., p. 140.

18

agency's initiative-merely, that is, to give certainty to one who is in doubt about his obligations under the statute-plus the statutory guaranty of immunity from the statutory liabilities for those who, in good faith, comply with this official guidance.

The proposal does not oust the courts from jurisdiction finally to interpret the statute. It merely relieves a man from liability for having, prior to the decision, followed in good faith a mistaken interpretation of the administrator. Indeed, in the absence of collusion, there would be no criminal intent to violate the statute on the part of one who, in good faith, conformed to authorized interpretative regulations.

A similar reasoning applies to the proposed immunity from civil liability for one who conforms to authorized interpretative regulations. The plaintiff in a civil suit has a claim created by the statute, and he is entitled to no more than the statute gives him. Congress may

certainly make the right to damages, which it grants him, subject to a qualification which is so reasonable and is, indeed, only commonsense justice to the defendant. 19

It is evident that this simple technique is in entire accord with the American system of government, and that at the same time it furnishes the needed middle term between legislation and adjudication. It enables the businessman to secure advance information of what his liabilities under the statute are, and, more important, it guarantees that this advance information can safely be relied upon. It is submitted that if this new technique, which has already proved satisfactory to the Securities and Exchange Commission and those affected by it, were made a regular feature of regulatory statutes, it would help in resolving the dilemma between reform and recovery.

19 In his article Cook cites the case of La Bourgoyne, 210 U. S. 95, 134 (1908).