¹ For significant criticisms of lawyers by political scientists in this field, see Gaus, John, Reflections on Public Administration (University, Alabama, 1947), pp. 107–8Google Scholar; Graham, George A., “Trends in Teaching of Public Administration,” Public Administration Review, Vol. 10, pp. 69–70 (Spring, 1950)CrossRefGoogle Scholar.

² (New York, 1952).

³ Case studies like those of the Stein collection are unknown to legal education. The case method of law schools is limited almost altogether to study of opinions of appellate courts. Indeed, the appellate opinion is the obsession of today's legal education.

Yet lawyers are devoting their time less and less to litigation and more and more to avoiding litigation—advising, planning, managing, negotiating, transacting, drafting. Legal education needs a basic shift toward lawyers' problems that are unrelated to litigation; problems of what-to-do instead of problems of who-should-win; problems of choosing among practical policies, not further exercises in analytical refinements; problems involving the kind of facts that lawyers must consider in advising businessmen, not merely neat summaries of facts in appellate opinions; problems intertwined with business or politics or technology, not merely interpreting statutes or applying judge-made law. The precise need of legal education seems to me to be the successful application of the Stein method to that part of the business of a law office that never goes to court. Such a project would be exceedingly difficult. But it can be done, and it should be done.

⁴ White, Leonard D., Introduction to the Study of Public Administration, 3rd ed. (New York, 1948)Google Scholar.

⁵ I have altered somewhat the facts of Lowe v. Conroy, 120 Wis. 151, 97 N.W. 942 (1904). Contra: Raymond v. Fish, 51 Conn. 80 (1883).

⁶ 13 Wall. 335, 347 (1871).

⁷ Kendall v Stokes, 3 How. 87, 97 (1845); Spalding v. Vilas, 161 U.S. 483, 498 (1896).

⁸ Pp. xcii, lxviii. For other rejections of Dicey's rule of law, see Robson, W. A., Justice and Administrative Law, 3d ed. (London, 1951), pp. 437–44Google Scholar; Jennings, Ivor, The Law and the Constitution, 2d ed. (London, 1939), pp. 285–97Google Scholar.

⁹ Pp. 19–20.

¹⁰ A reviewer of the Stein book exalts “the rich, though more traditional, literature of public administration” and declares: “The cases so far collected do not lead toward the kind of generalizations about the administrative process that would contribute significantly toward expanding the part of public administration that might be thought of as science rather than art. Cases are necessarily particularistic.” Fesler, James W., book review, Yale Law Journal, Vol. 61, p. 1238 (1952)CrossRefGoogle Scholar. I wonder whether the literature of public administration can be either “rich” or “science” unless it is solidly based upon “particularistic” studies.

¹¹ Report of the President's Committee, p. 40 (1937).

¹² Ibid.

¹³ Commissioner Eastman of the ICC knew the particulars about these questions. See his testimony, Hearings before the Select Committee on Government Organization, on S. 2700, 75th Cong., 1st sess., p. 180 (1937). Commissioner Eastman's testimony sharply contrasts with the factual conclusions of the President's Committee.

¹⁴ Report, p. 207.

¹⁵ The closest approach in the Cushman report to factual detail is an account in a few sentences of a statement by President Hoover in 1930 which, according to Cushman, “was intended to influence them [the ICC] on a matter of policy so important that it was bound to obtrude itself upon the President's broad economic program” (p. 221). But this is still only foliage. For what was President Hoover's “broad economic program”? If it existed, did it differ from ICC policies? Did the ICC do something that impaired the President's program? Were effective means available for producing whatever coordination was needed? Wherein, if at all, would the result have been different if the ICC functions had been performed by an executive department headed by a cabinet officer?

Cushman also quoted President Hoover's statement in 1929 about the Supreme Court's decision in the O'Fallon case: “I am confident that there will be no increase in railway rates as a result of the O'Fallon case.” But Cushman expressed no opinion as to whether the President should be able to control the ICC on a problem of rate levels. That question may be appropriate for a long and penetrating study by political scientists; it can hardly be disposed of satisfactorily in an offhand paragraph.

¹⁶ At p. 221 of the Cushman report appears a heading, “Interference with Presidential Management in the Field of Administration.” But the ensuing discussion is mostly generalities. One of the most specific statements is: “Several agencies besides the Federal Trade Commission have authority over unfair trade practices, but there is no central authority that can whip conflicting policies into line.” The statement is true. But the assumption that central management is necessarily indicated may be unsound. This is precisely the point at which a study of experience would be profitable. What have been the consequences of lack of central authority in this one field? When one agency has been lethargic, has another protected the public interest? What harm, if any, has resulted from conflicting policies? A theory about need for central management will hardly answer questions of this type.

¹⁷ Report, p. 230.

¹⁸ Report, p. 41.

¹⁹ If any reader doubts this, let him examine specific cases of administrative adjudication. An easy example is the problem that came to the Supreme Court in FCC v. WOKO, 329 U. S. 223 (1946). In a proceeding on an application for renewal of a broadcasting license, the evidence showed that the applicant in a former application misrepresented the identity of its shareholders. The only relevant statute required applicants to set forth facts “as to citizenship, character, and financial, technical, and other qualifications.” The Commission has to make its own policy when particular misrepresentations are brought to light, necessarily relying upon full development of all facts and circumstances. Because the facts were undisputed, the only question for the Commission was one of policy.

No set of rules could conceivably answer all the questions of policy that may arise in adjudications. And some important agencies, notably the FTC and NLRB, have no substantive rule-making power concerning their main functions. Whether or not the mind of man may some day invent a way to escape from the need for deciding policy issues in particular cases, we may be sure that no such invention has yet appeared.

²⁰ Escape from the dilemma cannot be effected by having the Judicial Section refer questions of policy to the Administrative Section, for then the decision will be made by the Administrative Section, which will then be doing the judicial work of deciding the particular case.

Even when the specific policy issues can be anticipated, some of them can be better handled through adjudication rather than through rule making. The choice between rule making and adjudication is often very difficult and cannot be resolved in wholesale fashion; particular solutions must depend upon circumstances. For a difficult problem concerning choice between policy making through rules or through adjudication, see the majority and dissenting opinions in SEC v. Chenery Corp., 332 U.S. 194 (1947).

²¹ The staff of the President's Committee was seriously mistaken about elementary features of FTC procedure. Compare the discussion of the “eleven steps” at p. 231 of the Report with the FTC Monograph prepared for the Attorney General's Committee on Administrative Procedure. The staff's worst error was the idea that the Commission holds hearings before issuing complaints. This error substantially affected the major recommendation.

²² Hearings before the Select Committee on Government Organization, on S. 2700, 75th Cong., 1st sess., p. 99 (1937).

²³ Ibid., pp. 98–99.

²⁴ Ibid., p. 182.

²⁵ Rogers, Lindsay, “The Independent Regulatory Commissions,” Political Science Quarterly, Vol. 52, pp. 1–17 (03, 1937)CrossRefGoogle Scholar.

²⁶ Report to the Select Committee to Investigate the Executive Agencies of the Government, Sen. Committee Print No. 10, 75th Cong., 1st sess., p. 92 (1937): Regulatory cases involve “an inseparable mixture of policy-determination, fact-finding, and decision” and “often cannot be broken up into component elements of administration and adjudication.” Wallace, Schuyler, in Federal Departmentalization (New York, 1941), p. 167Google Scholar, rejected the recommendation of the President's Committee on the basis of an overstatement: “The truth of the matter is that a differentiation between the administrative, quasi-legislative and quasi-judicial aspects of the work of these commissions is almost impossible except in the realm of theory.” Although the differentiation is difficult or impossible near the borderlines, it is also often easy; for instance, all would agree that an unfair-practice case in the FTC or NLRB is judicial or quasi-judicial. A disagreement with the President's Committee on theoretical grounds also was presented by Blachly, and Oatman, , Federal Regulatory Action and Control (Washington, 1940)Google Scholar.

Thirteen years after the report of the President's Committee, a political scientist at last analyzed the Committee's proposal in the light of the detailed facts of a regulatory commission. See Hyneman, Charles, Bureaucracy in a Democracy (New York, 1950), pp. 498–519Google Scholar. He says (p. 513): “The President's Committee stated that the division of work which it proposed between judicial and nonjudicial aspects ‘would be relatively simple.’ … I find it much easier to believe that the members of the Committee and the persons on whom they relied for advice did not know what the problems were.” Hyneman's treatment has been called “as lethal a critique of the recommendations of the President's Committee on Administrative Management regarding regulatory commissions as has come to this reader's attention.” Fainsod, Merle, “The Presidency and Congress,” Public Administration Review, Vol. 11, pp. 119–24 (Spring, 1951)CrossRefGoogle Scholar. See also Redford, Emmette S., Administration of National Economic Control (New York, 1952), p. 315Google Scholar: “More significant than application of law to facts is expansion of policy through case-to-case administration. It is submitted that the argument for independent placement breaks down where policy cannot be made specific in advance of the exercise of the judicial function.”

²⁷ Pfiffner, , Public Administration, 2nd ed. (New York, 1946), p. 463Google Scholar.

²⁸ Graves, , Public Administration in a Democratic Society (Boston, 1950), p. 670Google Scholar. See also White, Leonard, Introduction to the Study of Public Administration (cited in n. 4), p. 109Google Scholar.

²⁹ The task force was made up of a law professor and an industrialist. The Hoover Commission is said to have “largely passed over … university faculties of political science, and bureaus of public administration when it selected its staff.” Charlesworth, James C., Governmental Administration (New York, 1951), p. 230Google Scholar.

³⁰ Task Force Report on Regulatory Commissions, pp. 25–26 (1949).

³¹ Ibid., at pp. 26–28.

³² The case of “The Sale of the Tankers,” at pp. 445–532 of the Stein collection, involves a problem of coordinating policy among the State Department, the Navy Department, and the Maritime Commission. The recalcitrant was not the independent agency but the Navy Department, which for a time resisted a decision by the President's Cabinet.

³³ During the summer of 1953 a third edition of this book has been published. Some of the new errors in the new edition are even more astonishing than those in the second edition. For instance, page 487: “The present requirement is that ‘substantial evidence’ must be found throughout the entire record to support administrative action…. Prior to the Administrative Procedure Act, the general practice of the courts had been to sustain agency findings if adequate evidence could be found anywhere in the record.” The phrase “throughout the record” appears five times on this one page, and numerous times on other pages. Not only does the Act contain no such phrase, but the slightest reflection will show why a reviewing court could not require substantial evidence “throughout the record.” If in a thousand-page record, 990 pages are devoted to unconvincing circumstantial evidence on one side, and ten pages to convincing direct evidence on the other side, should a reviewing court hold that the agency may not believe the direct evidence?

³⁴ E.g., 54 Stat. 1197 (1940), 38 U.S.C. §11a–2 (1946); 48 Stat. 9 (1933), 38 U.S.C. §705 (1946).

³⁵ In First Moon v. White Tail, 270 U.S. 243 (1926), the Court held that the lower court had no jurisdiction to review where the statute provided that the determination of the Secretary of the Interior should be “final and conclusive” of an Indian allotment. For other examples, see Davis, Kenneth Culp, Administrative Law, §238 (St. Paul, 1951)Google Scholar.

³⁶ An eariy such provision, copied in many later federal statutes, is in the Federal Trade Commission Act, 38 Stat. 717 (1914), 15 U.S.C. §41 (1946).

³⁷ An outstanding early example is ICC v. Union Pac. R. Co., 222 U.S. 541 (1912).

³⁸ E.g., Federal Radio Comm. v. General Elec. Co., 281 U.S. 464 (1930).

³⁹ 60 Stat. 239 (1946), 5 U.S.C. §1011 (1946).

⁴⁰ E.g., Universal Camera Corp. v. NLRB, 340 U.S. 474 (1951).

⁴¹ In Bridges v. Wixon, 326 U.S. 135 (1945), the Court's review may even have gone beyond what is generally considered proper under the substantial-evidence rule.

⁴² American School of Magnetic Healing v. McAnnulty, 187 U.S. 94 (1902). For a recent review of a fraud order, see Reilly v. Pinkus, 338 U.S. 269 (1949).

⁴³ Annual Report, Director of Administrative Office of the United States Courts, p. 103 (1950).

⁴⁴ Other statements in addition to those discussed above also seem to me erroneous. Book 1 says at page 136 that when the elements of a decision can be separated so as to identify clear-cut law, “the court will rule on the law involved”; for one of the many cases of a court's refusal to rule on the law involved, see Switchmen's Union v. National Mediation Board, 320 U.S. 397 (1943). On page 138 the author relies on Ohio V. W. Co. v. Ben Avon Borough, 253 U.S. 387 (1920), with no indication that it may no longer be law, having been superseded by such cases as FPC v. Hope Nat. Gas Co., 320 U.S. 591 (1944), and Railroad Comm. v. Rowan & Nichols Co., 310 U.S. 573 (1940). Page 142, statement that res judicata “operates against retroactive administrative action,” whereas agencies often act retroactively; the Supreme Court ordered an agency to make law and apply it retroactively in Addison v. Holly Hill, 322 U.S. 607 (1944). Page 142, citing Arizona Grocery Co. v. Atchison, T. & S. F. Ry., 284 U.S. 370, 389 (1932) in support of res judicata when the court specifically said that the Commission was “not bound by the rule of res judicata.” Page 142, statement that the Labor Board may not reopen a case; this statement is mostly false because International Union v. Eagle-Picher Co., 325 U.S. 335 (1945) rests on a special statutory provision.

⁴⁵ For a full discussion of res judicata, see Davis, , Administrative Law, pp. 563–613Google Scholar.

⁴⁶ Morgan v. United States, 298 U.S. 468, 481 (1936).

⁴⁷ In addition to errors mentioned above, the following are especially important: the statement that rules are of “general application,” ignoring section 2(c) of the Administrative Procedure Act that rules may be “of particular applicability”; the statement making “rules” broader than, “regulations,” since regulations are only one of three classes under the rules heading, contradicting the statement made elsewhere that rules and regulations “are commonly used interchangeably”; the statement that laxity in publication of rules “no longer holds in most jurisdictions,” contradicting other statements indicating that only nine states required publication of rules in 1948; as of 1953 it is not yet true that “most jurisdictions” publish rules.

⁴⁸ See the discussion of cases in Davis, , Administrative Law, pp. 676–717Google Scholar.

⁴⁹ Interpretative rules are explained, ibid., pp. 194–226.

⁵⁰ E.g., 10 East Fortieth St. Bldg., Inc. v. Callus, 325 U.S. 578 (1945).

⁵¹ In addition to errors discussed above, the following seem especially important: differentiation of “rule” from “regulation” thereby departing from courts' usage, confusing students who may misunderstand the vocabulary of lawyers and courts, and departing from section 2(c) of the Administrative Procedure Act; classification of agencies into four categories, the last three of which seriously overlap—the Packers and Stockyards Administration, for instance, falls into all of the last three categories; most regulatory agencies are licensing authorities and many are within government departments; statement that apart from constitutional bodies, agencies cannot punish for contempt, whereas statutes in eleven states confer such power, and have been upheld in at least five. (See Davis, , Administrative Law, p. 123Google Scholar.)

⁵² A quick glance at the reports of decisions of these courts will show this.

⁵³ Williams v. United States, 289 U.S. 553, 565 (1933). The Court held in Ex parte Bakelite Corp., 279 U.S. 438, 458 (1929) that the Court of Customs Appeals is a court. The Judicial Code provides that the term “court of the United States” includes the Supreme Court, courts of appeals, district courts, the Court of Claims, the Court of Customs and Patent Appeals, the Customs Court “and any court created by Act of Congress the judges of which are entitled to hold office during good behavior.” 28 U.S.C. §451 (1949). The Tax Court judges hold office for twelve years. The statute provides that the Tax Court is “in the Executive Branch of the Government.” 26 U.S.C. §1100 (1946).

⁵⁴ See Davis, , Administrative Law, pp. 778–79Google Scholar.

⁵⁵ FTC v. Ruberoid Co., 343 U.S. 470, 487 (1952)

⁵⁶ Some political scientists evidently regard this subject as something that can be easily presented to students, but I think they are misguided. The subject is sufficiently difficult that the Supreme Court has been unable to formulate a consistent body of principles. For a discussion of various aspects of the subject, see Davis, , Administrative Law, pp. 614–63Google Scholar on exhaustion and ripeness, pp. 676–717 on standing, and pp. 812–67 on non-reviewable action.

⁵⁷ The exhaustion requirement is one of general application, as even the six cases set forth in Book 5 clearly show. Those six cases involved labor relations, a building permit, immigration, an alcohol license, a tax, and a rate order.

The text presents Myers v. Bethlehem S. Corp., 303 U.S. 41 (1938) as though its doctrine were unqualified. The Supreme Court in slightly different circumstances often holds the opposite, as in Public Util. Comm. v. United Fuel Gas Co., 317 U.S. 456 (1943). No mention is made of such outstanding cases as Aircraft & Diesel v. Hirsch, 331 U.S. 752 (1947); Lichter v. United States, 334 U.S. 742 (1948).

⁵⁸ 326 U.S. 219 (1945).

⁵⁹ I am the commentator whose conclusion that the Administrative Procedure Act “supersedes Levers v. Anderson” is said to be “inescapable.” But in the article cited for this, “Administrative Law Doctrines of Exhaustion of Remedies, Ripeness for Review, and Primary Jurisdiction,” Texas Law Review, Vol. 28, pp. 187–93 (1949)Google Scholar, no such position is taken. The closest approach is a statement (p. 191) that “The Supreme Court's effort to distinguish between a meaningful rehearing and the mere formality of applying for a re-hearing has been swept away by the APA.”

⁶⁰ Davis, , Administrative Law, p. 389Google Scholar. The central problem is whether or not an internal separation of functions can be satisfactory. The President's Committee did not discuss such internal separation.

⁶¹ Many political scientists may regard Truman's, David B.The Governmental Process (New York, 1951)Google Scholar as an adequate answer to some of these questions. This book seems to me to be of outstanding excellence, but I do not see how important governmental planning could be based upon its conclusions concerning pressures on regulatory agencies. Vagueness and generality rob the basic observations of precision and reliability. Statements that are largely or mostly true are too often partly false. See for instance the following:

“Experience indicates … that the regulated groups will have more cohesion than those demanding regulation” (p. 418). The groups regulated by the NLRB from 1935 to 1947 were employers; did not those demanding regulation, the unions, have greater cohesion? Under the Motor Carrier Act of 1935, do not those demanding regulation, the railroads, have greater cohesion than the regulated? Do not scheduled airlines, which demand regulation of the non-scheduled lines, have greater cohesion?

“The significant difference between an ‘independent’ agency and one clearly part of the executive branch is that the latter may somewhat more readily reflect the relative standing of the organized and unorganized interests responsible for the passage of a statute” (p. 419). What does this mean? Does it mean that an agency in the executive branch more accurately reflects relative strength of competing groups than an independent agency? Does not a valid observation on this subject have to take into account the President's attitudes toward the regulatory program? What is the evidence, and what are the exceptions? Does the statement rest upon specific experience, or is it largely a guess?

“A commission formally independent of the chief executive and expected to assume the detachment of a judicial body is more likely to be primarily accessible to the organized elements among the regulated than is an agency in the executive branch” (p. 420). Does the statement mean that judicial detachment makes an agency more accessible to the organized elements among the regulated? Does not a court's judicial detachment make it less accessible? On the question whether an independent agency is more accessible to organized elements among the regulated than an agency in the executive branch, does not the answer depend upon agreement or disagreement between presidential policies and the positions of the organized elements among the regulated?

“A regulatory agency normally cannot operate a controversial statute effectively without the support of the chief executive” (p. 420). Is the Interstate Commerce Act a controversial statute? Did the ICC operate effectively without the support of Presidents Roosevelt and Truman? Is the Federal Trade Commission Act controversial; did the two Presidents support the Commission; and did the Commission operate effectively? Should the generalization command confidence in absence of more specific information?

“The political survival of an independent commission depends upon its reaching a modus vivendi with the regulated” (p. 420). Does this mean that the agency must satisfy the principal demands of the regulated? If so, what of the NLRB from 1935 until 1947? What of the SEC for perhaps a half dozen years after 1933?

“Because other interests may have a larger voice in the arrangements made by an executive agency, ‘independence’ for the regulators has a defensive advantage for the regulated” (p. 420). Does this mean that interests other than the regulated interests may have a larger voice in the executive agency than in the independent agency? If so, are the President's views immaterial? What of the NLRB from 1935 to 1947 when the “other interests” were the unions? Do buyers of meat have a larger voice in the policies of the Packers and Stockyards Administration than buyers of natural gas have in the policies of the FPC? If so, is it because of FPC independence? What is the evidence?

“Once a set of relationships between regulated and regulators has been established, the resources of ‘independence’ will be fully employed to defend the existing regulatory pattern against disturbance or disruption” (p. 421). Was a set of relationships established between the NLRB and regulated employers between 1935 and 1947, and did the employers “defend the existing regulatory pattern”? Did the SEC in its first few years establish such a set of relationships, and did the regulated defend them?