dissenting.
It is the proud boast of our democracy that we have “a government of laws and not of men.” Many Americans are familiar with that phrase; not many know its derivation. It comes from Part the First, Article XXX, of the Massachusetts Constitution of 1780, which reads in full as follows:
“In the government of this Commonwealth, the legislative department shall never exercise the executive and judicial powers, or either of them: The executive shall never exercise the legislative and judicial powers, or either of them: The judicial shall never exercise the legislative and executive powers, or either of them: to the end it may be a government of laws and not of men.”
The Framers of the Federal Constitution similarly viewed the principle of separation of powers as the absolutely central guarantee of a just Government. In No. 47 of The Federalist; Madison wrote that “[n]o political truth is certainly of greater intrinsic value, or is stamped with the authority of more enlightened patrons of liberty.” The Federalist No. 47, p. 301 (C. Rossiter ed. 1961) (hereinafter Federalist). Without a secure structure of separated powers, our Bill of Rights would be worthless, as are the bills of rights of many nations of the world that have adopted, or even improved upon, the mere words of ours.
The principle of separation of powers is expressed in our Constitution in the first section of each of the first three Articles. Article I, § 1, provides that “[a]ll legislative Powers herein granted shall be vested in a Congress of the United *698States, which shall consist of a Senate and House of Representatives.” Article III, §1, provides that “[t]he judicial Power of the United States, shall be vested in one supreme Court, and in such inferior Courts as the Congress may from time to time ordain and establish.” And the provision at issue here, Art. II, § 1, cl. 1, provides that “[t]he executive Power shall be vested in a President of the United States of America.”
But just as the mere words of a Bill of Rights are not self-effectuating, the Framers recognized “[t]he insufficiency of a mere parchment delineation of the boundaries” to achieve the separation of powers. Federalist No. 73, p. 442 (A. Hamilton). “[T]he great security,” wrote Madison, “against a gradual concentration of the several powers in the same department consists in giving to those who administer each department the necessary constitutional means and personal motives to resist encroachments of the others. The provision for defense must in this, as in all other cases, be made commensurate to the danger of attack.” Federalist No. 51, pp. 321-322. Madison continued:
“But it is not possible to give to each department an equal power of self-defense. In republican government, the legislative authority necessarily predominates. The remedy for this inconveniency is to divide the legislature into different branches; and to render them, by different modes of election and different principles of action, as little connected with each other as the nature of their common functions and their common dependence on the society will admit. ... As the weight of the legislative authority requires that it should be thus divided, the weakness of the executive may require, on the other hand, that it should be fortified.” Id., at 322-323.
The major “fortification” provided, of course, was the veto power. But in addition to providing fortification, the Founders conspicuously and very consciously declined to sap the Executive’s strength in the same way they had weakened *699the Legislature: by dividing the executive power. Proposals to have multiple executives, or a council of advisers with separate authority were rejected. See 1 M. Farrand, Records of the Federal Convention of 1787, pp. 66, 71-74, 88, 91-92 (rev. ed. 1966); 2 id'., at 335-337, 533, 537, 542. Thus, while “[a]ll legislative Powers herein granted shall be vested in a Congress- of the United States, which shall consist of a Senate and House of Representatives,” U. S. Const., Art. I, § 1 (emphasis added), “[t]he executive Power shall be vested in a President of the United States,” Art. II, § 1, cl. 1 (emphasis added).
That is what this suit is about. Power. The allocation of power among Congress, the President, and the courts in such fashion as to preserve the equilibrium the Constitution sought to establish — so that “a gradual concentration of the several powers in the same department,” Federalist No. 51, p. 321 (J. Madison), can effectively be resisted. Frequently an issue of this sort will come before the Court clad, so to speak, in sheep’s clothing: the potential of the asserted principle to effect important change in the equilibrium of power is not immediately evident, and must be discerned by a careful and perceptive analysis. But this wolf comes as a wolf.
I
The present case began when the Legislative and Executive Branches became “embroiled in a dispute concerning the scope of the congressional investigatory power,” United States v. House of Representatives of United States, 556 F. Supp. 150, 152 (DC 1983), which — as is often the case with such interbranch conflicts — became quite acrimonious. In the course of oversight hearings into the administration of the Superfund by the Environmental Protection Agency (EPA), two Subcommittees of the House of Representatives requested and then subpoenaed numerous internal EPA documents. The President responded by personally directing the EPA Administrator not to turn over certain of the docu*700ments, see Memorandum of November 30, 1982, from President Reagan for the Administrator, Environmental Protection Agency, reprinted in H. R. Rep. No. 99-435, pp. 1166-1167 (1985), and by having the Attorney General notify the congressional Subcommittees of this assertion of executive privilege, see Letters of November 30, 1982, from Attorney General William French Smith to Hon. John D. Dingell and Hon. Elliott H. Levitas, reprinted, id., at 1168-1177. In his decision to assert executive privilege, the President was counseled by appellee Olson, who was then Assistant Attorney General of the Department of Justice for the Office of Legal Counsel, a post that has traditionally had responsibility for providing legal advice to the President (subject to approval of the Attorney General). The House’s response was to pass a resolution citing the EPA Administrator, who had possession of the documents, for contempt. Contempt of Congress is a criminal offense. See 2 U. S. C. § 192. The United States Attorney, however, a member of the Executive Branch, initially took no steps to prosecute the contempt citation. Instead, the Executive Branch sought the immediate assistance of the Third Branch by filing a civil action asking the District Court to declare that the EPA Administrator had acted lawfully in withholding the documents under a claim of executive privilege. See ibid. The District Court declined (in my view correctly) to get involved in the controversy, and urged the other two branches to try “[cjompro-mise and cooperation, rather than confrontation.” 556 F. Supp., at 153. After further haggling, the two branches eventually reached an agreement giving the House Subcommittees limited access to the contested documents.
Congress did not, however, leave things there. Certain Members of the House remained angered by the confrontation, particularly by the role played by the Department of Justice. Specifically, the Judiciary Committee remained disturbed by the possibility that the Department had persuaded the President to assert executive privilege despite reservations by the *701EPA; that the Department had “deliberately and unnecessarily precipitated a constitutional confrontation with Congress”; that the Department had not properly reviewed and selected the documents as to which executive privilege was asserted; that the Department had directed the United States Attorney not to present the contempt certification involving the EPA Administrator to a grand jury for prosecution; that the Department had made the decision to sue the House of Representatives; and that the Department had not adequately advised and represented the President, the EPA, and the EPA Administrator. H. R. Rep. No. 99-435, p. 3 (1985) (describing unresolved “questions” that were the basis of the Judiciary Committee’s investigation). Accordingly, staff counsel of the House Judiciary Committee were commissioned (apparently without the knowledge of many of the Committee’s members, see id., at 731) to investigate the Justice Department’s role in the controversy. That investigation lasted 214 years, and produced a 3,000-page report issued by the Committee over the vigorous dissent of all but one of its minority-party members. That report, which among other charges questioned the truthfulness of certain statements made by Assistant Attorney General Olson during testimony in front of the Committee during the early stages of its investigation, was sent to the Attorney General along with a formal request that he appoint an independent counsel to investigate Mr. Olson and others.
As a general matter, the Act before us here requires the Attorney General to apply for the appointment of an independent counsel within 90 days after receiving a request to do so, unless he determines within that period that “there are no reasonable grounds to believe that further investigation or prosecution is warranted.” 28 U. S. C. § 592(b)(1). As a practical matter, it would be surprising if the Attorney General had any choice (assuming this statute is constitutional) but to seek appointment of an independent counsel to pursue the charges against the principal object of the congressional *702request, Mr. Olson. Merely the political consequences (to him and the President) of seeming to break the law by refusing to do so would have been substantial. How could it not be, the public would ask, that a 3,000-page indictment drawn by our representatives over 2!4 years does not even establish “reasonable grounds to believe” that further investigation or prosecution is warranted with respect to at least the principal alleged culprit? But the Act establishes more than just practical compulsion. Although the Court’s opinion asserts that the Attorney General had “no duty to comply with the [congressional] request,” ante, at 694, that is not entirely accurate. He had a duty to comply unless he could conclude that there were “no reasonable grounds to believe,” not that prosecution was warranted, but merely that “further investigation” was warranted, 28 U. S. C. § 592(b)(1) (1982 ed., Supp. V) (emphasis added), after a 90-day investigation in which he was prohibited from using such routine investigative techniques as grand juries, plea bargaining, grants of immunity, or even subpoenas, see § 592(a)(2). The Court also makes much of the fact that “the courts are specifically prevented from reviewing the Attorney General’s decision not to seek appointment, § 592(f).” Ante, at 695. Yes,1 but Congress is not prevented from reviewing it. The context of this statute is acrid with the smell of threatened impeachment. Where, as here, a request for appointment of an inde*703pendent counsel has come from the Judiciary Committee of either House of Congress, the Attorney General must, if he decides not to seek appointment, explain to that Committee why. See also 28 U. S. C. § 595(c) (1982 ed., Supp. V) (independent counsel must report to the House of Representatives information “that may constitute grounds for an impeachment”).
Thus, by the application of this statute in the present case, Congress has effectively compelled a criminal investigation of a high-level appointee of the President in connection with his actions arising out of a bitter power dispute between the President and the Legislative Branch. Mr. Olson may or may not be guilty of a crime; we do not know. But we do know that the investigation of him has been commenced, not necessarily because the President or his authorized subordinates believe it is in the interest of the United States, in the sense that it warrants the diversion of resources from other efforts, and is worth the cost in money and in possible damage to other governmental interests; and not even, leaving aside those normally considered factors, because the President or his authorized subordinates necessarily believe that an investigation is likely to unearth a violation worth prosecuting; but only because the Attorney General cannot affirm, as Congress demands, that there are no reasonable grounds to believe that further investigation is warranted. The decisions regarding the seope of that further investigation, its duration, and, finally, whether or not prosecution should ensue, are likewise beyond the control of the President and his subordinates.
II
If to describe this case is not to decide it, the concept of a government of separate and coordinate powers no longer has meaning. The Court devotes most of its attention to such relatively technical details as the Appointments Clause and the removal power, addressing briefly and only at the end of its opinion the separation of powers. As my prologue sug*704gests, I think that has it backwards. Our opinions are full of the recognition that it is the principle of separation of powers, and the inseparable corollary that each department’s “defense must ... be made commensurate to the danger of attack,” Federalist No. 51, p. 322 (J. Madison), which gives comprehensible content to the Appointments Clause, and determines the appropriate scope of the removal power. Thus, while I will subsequently discuss why our appointments and removal jurisprudence does not support today’s holding, I begin with a consideration of the fountainhead of that jurisprudence, the separation and equilibration of powers.
First, however, I think it well to call to mind an important and unusual premise that underlies our deliberations, a premise not expressly contradicted by the Court’s opinion, but in my view not faithfully observed. It is rare in a case dealing, as this one does, with the constitutionality of a statute passed by the Congress of the United States, not to find anywhere in the Court’s opinion the usual, almost formulary caution that we owe great deference to Congress’ view that what it has done is constitutional, see, e. g., Rostker v. Goldberg, 453 U. S. 57, 64 (1981); Fullilove v. Klutznick, 448 U. S. 448, 472 (1980) (opinion of Burger, C. J.); Columbia Broadcasting System, Inc. v. Democratic National Committee, 412 U. S. 94, 102 (1973); United States v. National Dairy Products Corp., 372 U. S. 29, 32 (1963), and that we will decline to apply the statute only if the presumption of constitutionality can be overcome, see Fullilove, supra, at 473; Columbia Broadcasting, supra, at 103. That caution is not recited by the Court in the present case because it does not apply. Where a private citizen challenges action of the Government on grounds unrelated to separation of powers, harmonious functioning of the system demands that we ordinarily give some deference, or a presumption of validity, to the actions of the political branches in what is agreed, between themselves at least, to be within their respective spheres. But where the issue pertains to separation of pow*705ers, and the political branches are (as here) in disagreement, neither can be presumed correct. The reason is stated concisely by Madison: “The several departments being perfectly co-ordinate by the terms of their common commission, neither of them, it is evident, can pretend to an exclusive or superior right of settling the boundaries between their respective powers . . . Federalist No. 49, p. 314. The playing field for the present case, in other words, is a level one. As one of the interested and coordinate parties to the underlying constitutional dispute, Congress, no more than the President, is entitled to the benefit of the doubt.
To repeat, Article II, § 1, cl. 1, of the Constitution provides:
“The executive Power shall be vested in a President of the United States.”
As I described at the outset of this opinion, this does not mean some of the executive power, but all of the executive power. It seems to me, therefore, that the decision of the Court of Appeals invalidating the present statute must be upheld on fundamental separation-of-powers principles if the following two questions are answered affirmatively: (1) Is the conduct of a criminal prosecution (and of an investigation' to decide whether to prosecute) the exercise of purely executive power? (2) Does the statute deprive the President of the United States of exclusive control over the exercise of that power? Surprising to say, the Court appears to concede an affirmative answer to both questions, but seeks to avoid the inevitable conclusion that since the statute vests some purely executive power in a person who is not the President of the United States it is void.
The Court concedes that “[t]here is no real dispute that the functions performed by the independent counsel are ‘executive’,” though it qualifies that concession by adding “in the sense that they are law enforcement functions that typically have been undertaken by officials within the Executive Branch.” Ante, at 691. The qualifier adds nothing but at*706mosphere. In what other sense can one identify “the executive Power” that is supposed to be vested in the President (unless it includes everything the Executive Branch is given to do) except by reference to what has always and everywhere — if conducted by government at all — been conducted never by the legislature, never by the courts, and always by the executive. There is no possible doubt that the independent counsel’s functions fit this description. She is vested with the “full power and independent authority to exercise all investigative and prosecutorial functions and powers- of the Department of Justice [and] the Attorney General.” 28 U. S. C. § 594(a) (1982 ed., Supp. V) (emphasis added). Governmental investigation and prosecution of crimes is a quintessentially executive function. See Heckler v. Chaney, 470 U. S. 821, 832 (1985); Buckley v. Valeo, 424 U. S. 1, 138 (1976); United States v. Nixon, 418 U. S. 683, 693 (1974).
As for the second question, whether the statute before us deprives the President of exclusive control over that quintessentially executive activity: The Court does not, and could not possibly, assert that it does not. That is indeed the whole object of the statute. Instead, the Court points out that the President, through his Attorney General, has at least some control. That concession is alone enough to invalidate the statute, but I cannot refrain from pointing out that the Court greatly exaggerates the extent of that “some” Presidential control. ■ “Most importan[t]” among these controls, the Court asserts, is the Attorney General’s “power to remove the counsel for ‘good cause.’” Ante, at 696. This is somewhat like referring to shackles as an effective means of locomotion. As we recognized in Humphrey's Executor v. United States, 295 U. S. 602 (1935) — indeed, what Humphrey’s Executor was all about — limiting removal power to “good cause” is an impediment to, not an effective grant of, Presidential control. We said that limitation was necessary with respect to members of the Federal Trade Commission, which we found to be “an agency of the legislative and judicial *707departments,” and “wholly disconnected from the executive department,” id., at 630, because “it is quite evident that one who holds his office only during the pleasure of another, cannot be depended upon to maintain an attitude of independence against the latter’s will.” Id., at 629. What we in Humphrey’s Executor found to be a means of eliminating Presidential control, the Court today considers the “most important]” means of assuring Presidential control. Congress, of course, operated under no such illusion when it enacted this statute, describing the “good cause” limitation as “protecting the independent counsel’s ability to act independently of the President’s direct control” since it permits removal only for “misconduct.” H. R. Conf. Rep. 100-452, p. 37 (1987).
Moving on to the presumably “less important” controls that the President retains, the Court notes that no independent counsel may be appointed without a specific request from the Attorney General. As I have discussed above, the condition that renders such a request mandatory (inability to find “no reasonable grounds to believe” that further investigation is warranted) is so insubstantial that the Attorney General’s discretion is severely confined. And once the referral is made, it is for the Special Division to determine the scope and duration of the investigation. See 28 U. S. C. § 593(b) (1982 ed., Supp. V). And in any event, the limited power over referral is irrelevant to the question whether, once appointed, the independent counsel exercises executive power free from the President’s control. Finally, the Court points out that the Act directs the independent counsel to abide by general Justice Department policy, except when not “possible.” See 28 U. S. C. § 594(f) (1982 ed., Supp. V). The exception alone shows this to be an empty promise. Even without that, however, one would be hard put to come up with many investigative or prosecutorial “policies” (other than those imposed by the Constitution or by Congress through law) that are absolute. Almost all investigative and prosecutorial deci*708sions — including the ultimate decision whether, after a technical violation of the law has been found, prosecution is warranted — involve the balancing of innumerable legal and practical considerations. Indeed, even political considerations (in the nonpartisan sense) must be considered, as exemplified by the recent decision of an independent counsel to subpoena the former Ambassador of Canada, producing considerable tension in our relations with that country. See N. Y. Times, May 29, 1987, p. A12, col. 1. Another preeminently political decision is whether getting a conviction in a particular case is worth the disclosure of national security information that would be necessary. The Justice Department and our intelligence agencies are often in disagreement on this point, and the Justice Department does not always win. The present Act even goes so far as specifically to take the resolution of that dispute away from the President and give it to the independent counsel. 28 U. S. C. § 594(a)(6) (1982 ed., Supp. V). In sum, the balancing of various legal, practical, and political considerations, none of which is absolute, is the very essence of prosecutorial discretion. To take this away is to remove the core of the prosecutorial function, and not merely “some” Presidential control.
As I have said, however, it is ultimately irrelevant how much the statute reduces Presidential control. The case is over when the Court acknowledges, as it must, that “[i]t is undeniable that the Act reduces the amount of control or supervision that the Attorney General and, through him, .the President exercises over the investigation and prosecution of a certain class of alleged criminal activity.” Ante, at 695. It effects a revolution in our constitutional jurisprudence for the Court, once it has determined that (1) purely executive functions are at issue here, and (2) those functions have been given to a person whose actions are not fully within the supervision and control of the President, nonetheless to proceed further to sit in judgment of whether “the President’s need to control the exercise of [the independent counsel’s] *709discretion is so central to the functioning of the Executive Branch” as to require complete control, ante, at 691 (emphasis added), whether the conferral of his powers upon someone else “sufficiently deprives the President of control over the independent counsel to interfere impermissibly with [his] constitutional obligation to ensure the faithful execution of the laws,” ante, at 69S (emphasis added), and whether “the Act give[s] the Executive Branch sufficient control over the independent counsel to ensure that the President is able to perform his constitutionally assigned duties,” ante, at 696 (emphasis added). It is not for us to determine, and we have never presumed to determine, how much of the purely executive powers of government must be within the full control of the President. The Constitution prescribes that they all are.
The utter incompatibility of the Court’s approach with our constitutional traditions can be made more clear, perhaps, by applying it to the powers of the other two branches. Is it conceivable that if Congress passed a statute depriving itself of less than full and entire control over some insignificant area of legislation, we would inquire whether the matter was “so central to the functioning of the Legislative Branch” as really to require complete control, or whether the statute gives Congress “sufficient control over the surrogate legislator to ensure that Congress is able to perform its constitutionally assigned duties”? Of course we would have none of that. Once we determined that a purely legislative power was at issue we would require it to be exercised, wholly and entirely, by Congress. Or to bring the point closer to home, consider a statute giving to non-Article III judges just a tiny bit of purely judicial power in a relatively insignificant field, with substantial control, though not total control, in the courts — perhaps “clear error” review, which would be a fair judicial equivalent of the Attorney General’s “for cause” removal power here. Is there any doubt that we would not pause to inquire whether the matter was “so central to the *710functioning of the Judicial Branch” as really to require complete control, or whether we retained “sufficient control over the matters to be decided that we are able to perform our constitutionally assigned duties”? We would say that our “constitutionally assigned duties” include complete control over all exercises of the judicial power — or, as the plurality opinion said in Northern Pipeline Construction Co. v. Marathon Pipe Line Co., 458 U. S. 50, 58-59 (1982): “The inexorable command of [Article III] is clear and definite: The judicial power of the United States must be exercised by courts having the attributes prescribed in Art. III.” We should say here that the President’s constitutionally assigned duties include complete control over investigation and prosecution of violations of the law, and that the inexorable command of Article II is clear and definite: the executive power must be vested in the President of the United States.
Is it unthinkable that the President should have such exclusive power, even when alleged crimes by him or his close associates are at issue? No more so than that Congress should have the exclusive power of legislation, even when what is at issue is its own exemption from the burdens of certain laws. See Civil Rights Act of 1964, Title VII, 42 U. S. C. §2000e et seq. (prohibiting “employers,” not defined to include the United States, from discriminating on the basis of race, color, religion, sex, or national origin). No more so than that this Court should have the exclusive power to pronounce the final decision on justiciable cases and controversies, even those pertaining to the constitutionality of a statute reducing the salaries of the Justices. See United States v. Will, 449 U. S. 200, 211-217 (1980). A system of separate and coordinate powers necessarily involves an acceptance of exclusive power that can theoretically be abused. As we reiterate this very day, “[i]t is a truism that constitutional protections have costs.” Coy v. Iowa, post, at 1020. While the separation of powers may prevent us from righting every wrong, it does so in order to ensure that we do not lose lib*711erty. The checks against any branch’s abuse of its exclusive powers are twofold: First, retaliation by one of the other branch’s use of its exclusive powers: Congress, for example, can impeach the executive who willfully fails to enforce the laws; the executive can decline to prosecute under unconstitutional statutes, cf. United States v. Lovett, 328 U. S. 303 (1946); and the courts can dismiss malicious prosecutions. Second, and ultimately, there is the political check that the people will replace those in the political branches (the branches more “dangerous to the political rights' of the Constitution,” Federalist No. 78, p. 465) who are guilty of abuse. Political pressures produced special prosecutors — for Teapot Dome and for Watergate, for example — long before this statute created the independent counsel. See Act of Feb. 8, 1924, ch. 16, 43 Stat. 5-6; 38 Fed. Reg. 30738 (1973).
The Court has, nonetheless, replaced the clear constitutional prescription that the executive power belongs to the President with a “balancing test.” What are the standards to determine how the balance is to be struck, that is, how much removal of Presidential power is too much? Many countries of the world get along with an executive that is much weaker than ours — in fact, entirely dependent upon the continued support of the legislature. Once we depart from the text of the Constitution, just where short of that do we stop? The most amazing feature of the Court’s opinion is that it does not even purport to give an answer. It simply announces, with no analysis, that the ability to control the decision whether to investigate and prosecute the President’s closest advisers, and indeed the President himself, is not “so central to the functioning of the Executive Branch” as to be constitutionally required to be within the President’s control. Apparently that is so because we say it is so. Having abandoned as the basis for our decisionmaking the text of Article II that “the executive Power” must be vested in the President, the Court does not even attempt to craft a substitute criterion — a “justiciable standard,” see, e. g., Baker v. Carr, *712369 U. S. 186, 210 (1962); Coleman v. Miller, 307 U. S. 433, 454-455 (1939), however remote from the Constitution — that today governs, and in the future will govern, the decision of such questions. Evidently, the governing standard is to be what might be called the unfettered wisdom of a majority of this Court, revealed to an obedient people on a case-by-case basis. This is not only not the government of laws that the Constitution established; it is not a government of laws at all.
In my view, moreover, even as an ad hoc, standardless judgment the Court’s conclusion must be wrong. Before this statute was passed, the President, in taking action disagreeable to the Congress, or an executive officer giving advice to the President or testifying before Congress concerning one of those many matters on which the two branches are from time to time at odds, could be assured that his acts and motives would be adjudged — insofar as the decision whether to conduct a criminal investigation and to prosecute is concerned— in the Executive Branch, that is, in a forum attuned to the interests and the policies of the Presidency. That was one of the natural advantages the Constitution gave to the Presidency, just as it gave Members of Congress (and their staffs) the advantage of not being prosecutable for anything said or done in their legislative capacities. See U. S. Const., Art. I, §6, cl. 1; Gravel v. United States, 408 U. S. 606 (1972). It is the very object of this legislation to eliminate that assurance of a sympathetic forum. Unless it can honestly be said that there are “no reasonable grounds to believe” that further investigation is warranted, further investigation must ensue; and the conduct of the investigation, and determination of whether to prosecute, will be given to a person neither selected by nor subject to the control of the. President — who will in turn assemble a staff by finding out, presumably, who is willing to put aside whatever else they are doing, for an indeterminate period of time, in order to investigate and prosecute the President or a particular named individual in his administration. The prospect is frightening (as I will dis*713cuss at some greater length at the conclusion of this opinion) even outside the context of a bitter, interbranch political dispute. Perhaps the boldness of the President himself will not be affected — though I am not even sure of that. (How much easier it is for Congress, instead of accepting the political damage attendant to the commencement of impeachment proceedings against the President on trivial grounds — or, for that matter, how easy it is for one of the President’s political foes outside of Congress — simply to trigger a debilitating criminal investigation of the Chief Executive under this law.) But as for the President’s high-level assistants, who typically have no political base of support, it is as utterly unrealistic to think that they will not be intimidated by this prospect, and that their advice to him and their advocacy of his interests before a hostile Congress will not be affected, as it would be to think that the Members of Congress and their staffs would be unaffected by replacing the Speech or Debate Clause with a similar provision. It deeply wounds the President, by substantially reducing the President’s ability to protect himself and his staff. That is the whole object of the law, of course, and I cannot imagine why the Court believes it does not succeed.
Besides weakening the Presidency by reducing the zeal of his staff, it must also be obvious that the institution of the independent counsel enfeebles him more directly in his constant confrontations with Congress, by eroding his public support. Nothing is so politically effective as the ability to charge that one’s opponent and his associates are not merely wrongheaded, naive, ineffective, but, in all probability, “crooks.” And nothing so effectively gives an appearance of validity to such charges as a Justice Department investigation and, even better, prosecution. The present statute provides ample means for that sort of attack, assuring that massive and lengthy investigations will occur, not merely when the Justice Department in the application of its usual standards believes they are called for, but whenever it *714cannot be said that there are “no reasonable grounds to believe” they are called for. The statute’s highly visible procedures assure, moreover, that unlike most investigations these will be widely known and prominently displayed. Thus, in the 10 years since the institution of the independent counsel was established by law, there have been nine highly publicized investigations, a source of constant political damage to two administrations. That they could not remotely be described as merely the application of “normal” investigatory and prosecutory standards is demonstrated by, in addition to the language of the statute (“no reasonable grounds to believe”), the following facts; Congress appropriates approximately $50 million annually for general legal activities, salaries, and expenses of the Criminal Division of the Department of Justice. See 1989 Budget Request of the Department of Justice, Hearings before a Subcommittee of the House Committee on Appropriations, 100th Cong., 2d Sess., pt. 6, pp. 284-285 (1988) (DOJ Budget Request). This money is used to support “[f]ederal appellate activity,’’-“[organized crime prosecution,” “[p]ublic integrity” and “[f]raud” matters, “[njarcotic & dangerous drug prosecution,” “[internal security,” “[g]eneral litigation and legal advice,” “special investigations,” “[pjrosecution support,” “[o]rganized crime drug enforcement,” and “[mjanagement & administration.” Id., at 284. By comparison, between May 1986 and August 1987, four independent counsel (not all of whom were operating for that entire period of time) spent almost $5 million (one-tenth of the amount annually appropriated to the entire Criminal Division), spending almost $1 million in the month of August 1987 alone. See Washington Post, Oct. 21, 1987, p. A21, col. 5. For fiscal year 1989, the Department of Justice has requested $52 million for the entire Criminal Division, DOJ Budget Request 285, and $7 million to support the activities of independent counsel, id., at 25.
In sum, this statute does deprive the President of substantial control over the prosecutory functions performed by the *715independent counsel, and it does substantially affect the balance of powers. That the Court could possibly conclude otherwise demonstrates both the wisdom of our former constitutional system, in which the degree of reduced control and political impairment were irrelevant, since all purely executive power had to be in the President; and the folly of the new system of standardless judicial allocation of powers we adopt today.
Ill
As I indicated earlier, the basic separation-of-powers principles I have discussed are what give life and content to our jurisprudence concerning the President’s power to appoint and remove officers. The same result of unconstitutionality is therefore plainly indicated by our case law in these areas.
Article II, §2, cl. 2, of the Constitution provides as follows:
“[The President] shall nominate, and by and with the Advice and Consent of the the Senate, shall appoint Ambassadors, other public Ministers and Consuls, Judges of the supreme Court, and all other Officers of the United States, whose Appointments are not herein otherwise provided for, and which shall be established by Law: but the Congress may by Law vest the Appointment of such inferior Officers, as they think proper, in the President alone, in the Courts of Law, or in the Heads of Departments.”
Because appellant (who all parties and the Court agree is an officer of the United States, ante, at 671, n. 12) was not appointed by the President with the advice and consent of the Senate, but rather by the Special Division of the United States Court of Appeals, her appointment is constitutional only if (1) she is an “inferior” officer within the meaning of the above Clause, and (2) Congress may vest her appointment in a court of law.
As to the first of these inquiries, the Court does not attempt to “decide exactly” what estáblishes the line between *716principal and “inferior” officers, but is confident that, whatever the line may be, appellant “clearly falls on the ‘inferior officer’ side” of it. Ante, at 671. The Court gives three reasons: First, she “is subject to removal by a higher Executive Branch official,” namely, the Attorney General. Ibid. Second, she is “empowered by the Act to perform only certain, limited duties.” Ibid. Third, her office is “limited in jurisdiction” and “limited in tenure.” Ante, at 672.
The first of these lends no support to the view that appellant is an inferior officer. Appellant is removable only for “good cause” or physical or mental incapacity. 28 U. S. C. §596(a)(1) (1982 ed., Supp. V). By contrast, most (if not all) principal officers in the Executive Branch may be removed by the President at will. I fail to see how the fact that appellant is more difficult to remove than most principal officers helps to establish that she is an inferior officer. And I do not see how it could possibly make any difference to her superior or inferior status that the President’s limited power to remove her must be exercised through the Attorney General. If she were removable at will by the Attorney General, then she would be subordinate to him and thus properly designated as inferior; but the Court essentially admits that she is not subordinate. See ante, at 671. If it were- common usage to refer to someone as “inferior” who is subject to removal for cause by another, then one-would say that the President is “inferior” to Congress.
The second reason offered by the Court — that appellant performs only certain, limited duties — may be relevant to whether she is an inferior officer, but it mischaracterizes the extent of her powers. As the Court states: “Admittedly, the Act delegates to appellant [the] ‘full power and independent authority to exercise all investigative and prosecutorial functions and powers of the Department of Justice.Ibid., quoting 28 U. S. C. §594(a) (1982 ed., Supp. V) (emphasis *717added).2 Moreover, in addition to this general grant of power she is given a broad range of specifically enumerated powers, including a power not even the Attorney General possesses: to “contes[t] in court. . . any claim of privilege or attempt to withhold evidence on grounds of national security.” § 594(a)(6).3 Once all of this is “admitted,” it seems *718to me impossible to maintain that appellant’s authority is so “limited” as to render her an inferior officer. The Court seeks to brush this away by asserting that the independent counsel’s power does not include any authority to “formulate policy for the Government or the Executive Branch.” Ante, at 671. But the same could be said for all officers of the Government, with the single exception of the President. All of them only formulate policy within their respective spheres of responsibility — as does the independent counsel, who must comply with the policies of the Department of Justice only to the extent possible. § 594(f).
The final set of reasons given by the Court for why the independent counsel clearly is an inferior officer emphasizes the limited nature of her jurisdiction and tenure. Taking the latter first, I find nothing unusually limited about the independent counsel’s tenure. To the contrary, unlike most high ranking Executive Branch officials, she continues to serve until she (or the Special Division) decides that her work is substantially completed. See §§ 596(b)(1), (b)(2). This particular independent prosecutor has already served more than two years, which is at least as long as many Cabinet officials. As to the scope of her jurisdiction, there can be no doubt that is small (though far from unimportant). But within it she exercises more than the full power of the Attorney General. The Ambassador to Luxembourg is not anything less than a principal officer, simply because Luxembourg is small. And the federal judge who sits in a small district is not for that reason “inferior in rank and authority.” If the mere fragmentation of executive responsibilities into small compartments suffices to render the heads of each of those compartments inferior officers, then Congress could deprive the President of the right to appoint his chief law enforcement officer by dividing up the Attorney General’s responsibilities among a number of “lesser” functionaries.
*719More fundamentally, however, it is not clear from the Court’s opinion why the factors it discusses — even if applied correctly to the facts of this case — are determinative of the question of inferior officer status. The apparent source of these factors is a statement in United States v. Germaine, 99 U. S. 508, 511 (1879) (discussing United States v. Hartwell, 6 Wall. 385, 393 (1868)), that “the term [officer] embraces the ideas of tenure, duration, emolument, and duties.” See ante, at 672. Besides the fact that this was dictum, it was dictum in a case where the distinguishing characteristics of inferior officers versus superior officers were in no way relevant, but rather only the distinguishing characteristics of an “officer of the United States” (to which the criminal statute at issue applied) as opposed to a mere employee. Rather than erect a theory of who is an inferior officer on the foundation of such an irrelevancy, I think it preferable to look to the text of the Constitution and the division of power that it establishes. These demonstrate, I think, that the independent counsel is not an inferior officer because she is not subordinate to any officer in the Executive Branch (indeed, not even to the President). Dictionaries in use at the time of the Constitutional Convention gave the word “inferiour” two meanings which it still bears today: (1) “[l]ower in place, . . . station, . . . rank of life, . . . value or excellency,” and (2) “[subordinate.” S. Johnson, Dictionary of the English Language (6th ed. 1785). In a document dealing with the structure (the constitution) of a government, one would naturally expect the word to bear the latter meaning — indeed, in such a context it would be unpardonably careless to use the word unless a relationship of subordination was intended. If what was meant was merely “lower in station or rank,” one would use instead a term such as “lesser officers.” At the only other point in the Constitution at which the word “inferior” appears, it plainly connotes a relationship of subordination. Article III vests the judicial power of the United States in “one supreme Court, and in such inferior Courts as *720the Congress may from time to time ordain and establish.” U. S. Const., Art. Ill, §1 (emphasis added). In Federalist No. 81, Hamilton pauses to describe the “inferior” courts authorized by Article III as inferior in the sense that they are “subordinate” to the Supreme Court. See id., at 485, n., 490, n.
That “inferior” means “subordinate” is also consistent with what little we know about the evolution of the Appointments Clause. As originally reported to the Committee on Style, the Appointments Clause provided no “exception” from the standard manner of appointment (President with the advice and consent of the Senate) for inferior officers. 2 M. Farrand, Records of the Federal Convention of 1787, pp. 498-499, 599 (rev. ed. 1966). On September 15, 1787, the last day of the Convention before the proposed Constitution was signed, in the midst of a host of minor changes that were being considered, Gouverneur Morris moved to add the exceptions clause. Id., at 627. No great debate ensued; the only disagreement was over whether it was necessary at all. Id., at 627-628. Nobody thought that it was a fundamental change, excluding from the President’s appointment power and the Senate’s confirmation power a category of officers who might function on their own, outside the supervision of those appointed in the more cumbersome fashion. And it is significant that in the very brief discussion Madison mentions (as in apparent contrast to the “inferior officers” covered by the provision) “Superior Officers.” Id., at 637. Of course one is not a “superior officer” without some supervisory . responsibility, just as, I suggest, one is not an “inferior officer” within the meaning of the provision under discussion unless one is subject to supervision by a “superior officer.” It is perfectly obvious, therefore, both from the relative brevity of the discussion this addition received, and from the content of that discussion, that it was intended merely to make clear (what Madison thought already was clear, see id., at 627) that those officers appointed by the President with Senate *721approval could on their own appoint their subordinates, who would, of course, by chain of command still be under the direct control of the President. ■
This interpretation is, moreover, consistent with our admittedly sketchy precedent in this area. For example, in United. States v. Eaton, 169 U. S. 331 (1898), we held that the appointment by an Executive Branch official other than the President of a “vice-consul,” charged with the. duty of temporarily performing the function of the consul, did not violate the Appointments Clause. In doing so, we repeatedly referred to the “vice-consul” as a “subordinate” officer. Id., at 343. See also United States v. Germaine, supra, at 511 (comparing “inferior” commissioners and bureau officers to heads of department, describing the former as “mere . . . subordinates”) (dicta); United States v. Hartwell, supra, at 394 (describing clerk appointed by Assistant Treasurer with approval of Secretary of the Treasury as a “subordinate office[r]”) (dicta). More recently, in United States v. Nixon, 418 U. S. 683 (1974), we noted that the Attorney General’s appointment of the Watergate Special Prosecutor was made pursuant to the Attorney General’s “power to appoint subordinate officers to assist him in the discharge of his duties.” Id., at 694 (emphasis added). The Court’s citation of Nixon as support for its view that the independent counsel is an inferior officer is simply not supported by a reading of the case. We explicitly stated that the Special Prosecutor was a “subordinate office[r],” ibid., because, in the end, the President or the Attorney General could have removed him at any time, if by no other means than amending or revoking the regulation defining his authority. Id., at 696. Nor are any of the other cases cited by the Court in support of its view inconsistent with the natural reading that an inferior officer must at least be subordinate to another officer of the United States. In Ex parte Siebold, 100 U. S. 371 (1880), we upheld the appointment by a court of federal “Judges of Election,” who were charged with various duties involving the oversee*722ing of local congressional elections. Contrary to the Court’s assertion, see ante, at 673, we did not specifically find that these officials were inferior officers for purposes of the Appointments Clause, probably because no one had contended that they were principal officers. Nor can the case be said to represent even an assumption on our part that they were inferior without being subordinate. The power of assisting in the judging of elections that they were exercising was assuredly not a purely executive power, and if we entertained any assumption it was probably that they, like the marshals who assisted them, see Siebold, 100 U. S., at 380, were subordinate to the courts, see id., at 397. Similarly, in Go-Bart Importing Co. v. United States, 282 U. S. 344 (1931), where we held that United States commissioners were inferior officers, we made plain that they were subordinate to the district courts which appointed them: “The commissioner acted not as a court, or as a judge of any court, but as a mere officer of the district court in proceedings of which that court had authority to take control at any time.” Id., at 354.
To be sure, it is not a sufficient condition for “inferior”, officer status that one be subordinate to a principal officer. Even an officer who is subordinate to a department head can be a principal officer. That is clear from the brief exchange following Gouverneur Morris’ suggestion of the addition of the exceptions clause for inferior officers. Madison responded:
“It does not go far enough if it be necessary at all — Superior Officers below Heads of Departments ought in some cases to have the appointment of the lesser offices.” 2 M. Farrand, Records of the Federal Convention, of 1787, p. 627 (rev. ed. 1966) (emphasis added).
But it is surely a necessary condition for inferior officer status that the officer be subordinate to another officer.
The independent counsel is not even subordinate to the President. The Court essentially admits as much, noting that “appellant may not be ‘subordinate’ to the Attorney Gen*723eral (and the President) insofar as she possesses a degree of independent discretion to exercise the powers delegated to her under the Act.” Ante, at 671. In fact, there is no doubt about it. As noted earlier, the Act specifically grants her the “full power and independent authority to exercise all investigative and prosecutorial functions of the Department of Justice,” 28 U. S. C. § 594(a) (1982 ed., Supp. V), and makes her removable only for “good cause,” a limitation specifically intended to ensure that she be independent of, not subordinate to, the President and the Attorney General. See H. R. Conf. Rep. No. 100-452, p. 37 (1987).
Because appellant is not subordinate to another officer, she is not an “inferior” officer and her appointment other than by the President with the advice and consent of the Senate is unconstitutional.
IV
I will not discuss at any length why the restrictions upon the removal of the independent counsel also violate our established precedent dealing with that specific subject. For most of it, I simply refer the reader to the scholarly opinion of Judge Silberman for the Court of Appeals below. See In re Sealed Case, 267 U. S. App. D. C. 178, 838 F. 2d 476 (1988). I cannot avoid commenting, however, about the essence of what the Court has done to our removal jurisprudence today.
There is, of course, no provision in the Constitution stating who may remove executive officers, except the provisions for removal by impeachment. Before the present decision it was established, however, (1) that the President’s power to remove principal officers who exercise purely executive powers could not be restricted, see Myers v. United States, 272 U. S. 52, 127 (1926), and (2) that his power to remove inferior officers who exercise purely executive powers, and whose appointment Congress had removed from the usual procedure of Presidential appointment with Senate consent, could be restricted, at least where the appointment had been made by *724an officer of the Executive Branch, see ibid.; United States v. Perkins, 116 U. S. 483, 485 (1886).4
The Court could have resolved the removal power issue in this case by simply relying upon its erroneous conclusion that the independent counsel was an inferior officer, and then extending our holding that the removal of inferior officers appointed by the Executive can be restricted, to a new holding that even the removal of inferior officers appointed by the courts can be restricted. That would in my view be a considerable and unjustified extension, giving the Executive full discretion in neither the selection nor the removal of a purely executive officer. The course the Court has chosen, however, is even worse.
Since our 1935 decision in Humphrey’s Executor v. United States, 295 U. S. 602 — which was considered by many at the time the product of an activist, anti-New Deal Court bent on reducing the power of President Franklin Roosevelt — it has been established that the line of permissible restriction upon removal of principal officers lies at the point at which the powers exercised by those officers are no longer purely executive. Thus, removal restrictions have been generally regarded as lawful for so-called “independent regulatory *725agencies,” such as the Federal Trade Commission, see ibid,.; 15 U. S. C. §41, the Interstate Commerce Commission, see 49 U. S. C. § 10301(c) (1982 ed., Supp. IV), and the Consumer Product Safety Commission, see 15 U. S. C. § 2053(a), which engage substantially in what has been called the “quasi-legislative activity” of rulemaking, and for members of Article I courts, such as the Court of Military Appeals, see 10 U. S. C. § 867(a)(2), who engage in the “quasi-judicial” function of adjudication. It has often been observed, correctly in my view, that the line between “purely executive” functions and “quasi-legislative” or “quasi-judicial” functions is not a clear one or even a rational one. See ante, at 689-691; Bowsher v. Synar, 478 U. S. 714, 761, n. 3 (1986) (White, J., dissenting); FTC v. Ruberoid Co., 343 U. S. 470, 487-488 (1952) (Jackson, J., dissenting). But at least it permitted the identification of certain officers, and certain agencies, whose functions were entirely within the control of the President. Congress had to be aware of that restriction in its legislation. Today, however, Humphrey’s Executor is swept into the dustbin of repudiated constitutional principles. “[0]ur present considered view,” the Court says, “is that the determination of whether the Constitution allows Congress to impose a ‘good cause’-type restriction on the President’s power to remove an official cannot be made to turn on whether or not that official is classified as ‘purely executive.’” Ante, at 689. What Humphrey’s Executor (and presumably Myers) really means, we are now told, is not that there are any “rigid categories of those officials who may or may not be removed at will by the President,” but simply that Congress cannot “interefere with the President’s exercise of the ‘executive power’ and his constitutionally appointed duty to ‘take care that the laws be faithfully executed,’” ante, at 689-690.
One can hardly grieve for the shoddy treatment given today to Humphrey’s Executor, which, after all, accorded the same indignity (with much less justification) to Chief Justice *726Taft’s opinion 10 years earlier in Myers v. United States, 272 U. S. 52 (1926) — gutting, in six quick pages devoid of textual or historical precedent for the novel principle it set forth, a carefully researched and reasoned 70-page opinion. It is in fact comforting to witness the reality that he who lives by the ipse dixit dies by the ipse dixit. But one must grieve for the Constitution. Humphrey's Executor at least had the decency formally to observe the constitutional principle that the President had to be the repository of «(/ 'executive power, see 295 U. S., at 627-628, which, as Myers carefully explained, necessarily means that he must be able to discharge those who do not perform executive functions according to his liking. As we noted in Bowsher, once an officer is appointed ‘“it is only the authority that can remove him, and not the authority that appointed him, that he must fear and, in the performance of his functions, obey.’” 478 U. S., at 726, quoting Synar v. United States, 626 F. Supp. 1374, 1401 (DC 1986) (Scalia, Johnson, and Gasch, JJ.). By contrast, “our present considered view” is simply that any executive officer’s removal can be restricted, so long as the President remains “able to accomplish his constitutional role.” Ante, at 690. There are now no lines. If the removal of a prosecutor, the virtual embodiment of the power to “take care that the laws be faithfully executed,” can be restricted, what officer’s removal cannot? This is an open invitation for Congress to experiment. What about a special Assistant Secretary of State, with responsibility for one very narrow area of foreign policy, who would not only have to be confirmed by the Senate but could also be removed only pursuant to certain carefully designed restrictions? Could this possibly render the President “Lunjable to accomplish his constitutional role”? Or a special Assistant Secretary of Defense for Procurement? The possibilities are endless, and the Court does not understand what the separation of powers, what “[ajmbition . . . counteract[ing] ambition,” Federalist No. 51, p. 322 (Madison), is all about, if it does not expect Congress to try them. As far as I can discern from the Court's opinion, it is now *727open season upon the President’s removal power for all executive officers, with not even the superficially principled restriction of Humphrey’s Executor as cover. The Court essentially says to the President: “Trust us. We will make sure that you are able to accomplish your constitutional role.” I think the Constitution gives the President — and the people-more protection than that.
V
The purpose of the separation and equilibration of powers in general, and of the unitary Executive in particular, was not merely to assure effective government but to preserve individual freedom. Those who hold or have held offices covered by the Ethics in Government Act are entitled to that protection as much as the rest of us, and I conclude my discussion by considering the effect of the Act upon the fairness of the process they receive.
Only someone who has worked in the field of law enforcement can fully appreciate the vast power and the immense discretion that are placed in the hands of a prosecutor with respect to the objects of his investigation. Justice Robert Jackson, when he was Attorney General under President Franklin Roosevelt, described it in a memorable speech to United States Attorneys, as follows:
“There is a most important reason why the prosecutor should have, as nearly as possible, a detached and impartial view of all groups in his community. Law enforcement is not automatic. It isn’t blind. One of the greatest difficulties of the position of prosecutor is that he must pick his cases, because no prosecutor can even investigate all of the cases in which he receives complaints. If the Department of Justice were to make even a pretense of reaching every probable violation of federal law, ten times its present staff will be inadequate. We know that no local police force can strictly enforce the traffic laws, or it would arrest half the driving population on *728any given morning. What every prosecutor is practically required to do is to select the cases for prosecution and to select those in which the offense is the most flagrant, the public harm the greatest, and the proof the most certain.
“If the prosecutor is obliged to choose his case, it follows that he can choose his defendants. Therein is the most dangerous power of the prosecutor: that he will pick people that he thinks he should get, rather than cases that need to be prosecuted. With the law books filled with a great assortment of crimes, a prosecutor stands a fair chance of finding at least a technical violation of some act on the part of almost anyone. In such a case, it is not a question of discovering the commission of a crime and then looking for the man who has committed it, it is a question of picking the man and then searching the law books, or putting investigators to work, to pin some offense on him. It is in this realm — in which the prosecutor picks some person whom he dislikes or desires to embarrass, or selects some group of unpopular persons and then looks for an offense, that the greatest danger of abuse of prosecuting power lies. It is here that law enforcement becomes personal, and the real crime becomes that of being unpopular with the predominant or governing group, being attached to the wrong political views, or being personally obnoxious to or in the way of the prosecutor himself.” R. Jackson, The Federal Prosecutor, Address Delivered at the Second Annual Conference of United States Attornevs, April 1, 1940.
Under our system of government, the primary check against prosecutorial abuse is a political one. The prosecutors who exercise this awesome discretion are selected and can be removed by a President, whom the people have trusted enough to elect. Moreover, when crimes are not investigated and prosecuted fairly, nonselectively, with a rea*729sonable sense of proportion, the President pays the cost in political damage to his administration. If federal prosecutors “pick people that [they] thin[k] [they] should get, rather than cases that need to be prosecuted,” if they amass many more resources against a particular prominent individual, or against a particular class of political protesters, or against members of a particular political party, than the gravity of the alleged offenses or the record of successful prosecutions seems to warrant, the unfairness will come home to roost in the Oval Office. I leave it. to the reader to recall the examples of this in recent years. That result, of course, was precisely what the Founders had in mind when they provided that all executive powers would be exercised by a single Chief Executive. As Hamilton put it, “[t]he ingredients which constitute safety in the republican sense are a due dependence on the people, and a due responsibility.” Federalist No. 70, p. 424. The President is directly dependent on the people, and since there is only one President, he is responsible. The people know whom to blame, whereas “one of the weightiest objections to a plurality in the executive ... is that it tends to conceal faults and destroy responsibility.” Id., at 427.
That is the system of justice the rest of us are entitled to, but what of that select class consisting of present or former high-level Executive Branch officials? If an allegation is made against them of any violation of any federal criminal law (except Class B or C misdemeanors or infractions) the Attorney General must give it his attention. That in itself is not objectionable. But if, after a 90-day investigation without the benefit of normal investigatory tools, the Attorney General is unable to say that there are “no reasonable grounds to believe” that further investigation is warranted, a process is set in motion that is not in the full control of persons “dependent on the people,” and whose flaws cannot be blamed on the President. An independent counsel is selected, and the scope of his or her authority prescribed, by a *730panel of judges. What if they are politically partisan, as judges have been known to be, and select a prosecutor antagonistic to the administration, or even to the particular individual who has been selected for this special treatment? There is no remedy for that, not even a political one. Judges, after all, have life tenure, and appointing a surefire enthusiastic prosecutor could hardly be considered an impeachable offense. So if there is anything wrong with the selection, there is effectively no one to blame. The independent counsel thus selected proceeds to assemble a staff. As I observed earlier, in the nature of things this has to be done by finding lawyers who are willing to lay aside their current careers for an indeterminate amount of time, to take on a job that has no prospect of permanence and little prospect for promotion. One thing is certain, however: it involves investigating and perhaps prosecuting a particular individual. Can one imagine a less equitable manner of fulfilling the executive responsibility to investigate and prosecute? What would be the reaction if, in an area not covered by this statute, the Justice Department posted a public notice inviting applicants to assist in an investigation and possible prosecution of a certain prominent person? Does this not invite what Justice Jackson described as “picking the man and then searching the law books, or putting investigators to work, to pin some offense on him”? To be sure, the investigation must relate to the area of criminal offense specified by the life-tenured judges. But that has often been (and nothing prevents it from being) very broad — and should the independent counsel or his or her staff come up with something beyond that scope, nothing prevents him or her from asking the judges to expand his or her authority or, if that does not work, referring it to the Attorney General, whereupon the whole process would recommence and, if there was “reasonable basis to believe” that further investigation was warranted, that new offense would be referred to the Special Division, which would in all likelihood assign it to the same *731independent counsel. It seems to me not conducive to fairness. But even if it were entirely evident that unfairness was in fact the result — the judges hostile to the administration, the independent counsel an old foe of the President, the staff refugees from the recently defeated administration— there would, he no one accountable to the public to whom the blame could be assigned.
I do not mean to suggest that anything of this sort (other than the inevitable self-selection of the prosecutory staff) occurred in the present case. I know and have the highest regard for the judges on the Special Division, and the independent counsel herself is a woman of accomplishment, impartiality, and integrity. But the fairness of a process must be adjudged on the basis of what it permits to happen, not what it produced in a particular case. It is true, of course, that a similar list of horribles could be attributed to an ordinary Justice Department prosecution — a vindictive prosecutor, an antagonistic staff, etc. But the difference is the difference that the Founders envisioned when they established a single Chief Executive accountable to the people: the blame can- be assigned to someone who can be punished.
The above described possibilities of irresponsible conduct must, as I say, be considered in judging the constitutional acceptability of this process. But they will rarely occur, and in the average case the threat to fairness is quite different. As described in the brief filed on behalf of three ex-Attorneys General from each of the last three administrations:
“The problem is less spectacular but much more worrisome. It is that the institutional environment of the Independent Counsel — specifically, her isolation from the Executive Branch and the internal cheeks and balances it supplies — is designed to heighten, not to check, all of the occupational hazards of the dedicated prosecutor; the danger of too narrow a focus, of the loss of perspective, of preoccupation with the pursuit of one alleged suspect to the exclusion of other interests.” Brief for Edward *732H. Levi, Griffin B. Bell, and William French Smith as Amici Curiae 11.
It is, in other words, an additional advantage of the unitary Executive that it can achieve a more uniform application of the law. Perhaps that is not always achieved, but the mechanism to achieve it is there. The mini-Executive that is the independent counsel, however, operating in an area where so little is law and so much is discretion, is intentionally cut off from the unifying influence of the Justice Department, and from the perspective that multiple responsibilities provide. What would normally be regarded as a technical violation (there are no rules defining such things), may in his or her small world assume the proportions of an indictable offense. What would normally be regarded as an investigation that has reached the level of pursuing such picayune matters that it should be concluded, may to him or her be an investigation that ought to go on for another year. How frightening it must be to have your own independent counsel and staff appointed, with nothing else to do but to investigate you until investigation is no longer worthwhile — with whether it is worthwhile not depending upon what such judgments usually hinge on, competing responsibilities. And to have that counsel and staff decide, with no. basis for comparison, whether what you have done is bad enough, willful enough, and provable enough, to warrant an indictment. How admirable the constitutional system that provides the means to avoid such a distortion. And how unfortunate the judicial decision that has permitted it.
* * *
The notion that every violation of law should be' prosecuted, including — indeed, especially — every violation by those in high places, is an attractive one, and it would be risky to argue in an election campaign that that is not an absolutely overriding value. Fiat justitia, ruat coelum. Let justice be done, though the heavens may fall. The reality is, however, that it is not an absolutely overriding value, and it *733was with the hope that we would be able to acknowledge and apply such realities that the Constitution spared us, by life tenure, the necessity of election campaigns. I cannot imagine that there are not many thoughtful men and women in Congress who realize that the benefits of this legislation are far outweighed by its harmful effect upon our system of government, and even upon the nature of justice received by those men and women who agree to serve in the Executive Branch. But it is difficult to vote not to enact, and even more difficult to vote to repeal, a statute called, appropriately enough, the Ethics in Government Act. If Congress is controlled by the party other than the one to which the President belongs, it has little incentive to repeal it; if it is controlled by the same party, it dare not. By its shortsighted action today, I fear the Court has permanently encumbered the Republic with an institution that will do it great harm.
. Worse than what it has done, however, is the manner in which it has done it. A government of laws means a government of rules. Today’s decision on the basic issue of fragmentation of executive power is ungoverned by rule, and hence ungoverned by law. It extends into the very heart of our most significant constitutional function the “totality of the circumstances” mode of analysis that this Court has in recent years become fond of. Taking all things into account, we conclude that the power taken away from the President he.re is not really too much. The next time executive power is assigned to someone other than the President we may conclude, taking all things into account, that it is too much. That opinion, like this one, will not be confined by any rule. We will describe, as we have today (though I hope more accurately) the effects of the provision in question, and will authoritatively announce: “The President’s need to control the exercise of the [subject officer’s] discretion is so central to the functioning of the Executive Branch as to require complete control.” This is not analysis; it is ad hoc judgment. And it fails to explain why it is not true that — as the text of *734the Constitution seems to require, as the Founders seemed to expect, and as our past cases have uniformly assumed — all purely executive power must be under the control of the President.
The ad hoc approach to constitutional adjudication has real attraction, even apart from its work-saving potential. It is guaranteed to produce a result, in every case, that will make a majority of the Court happy with the law. The law is, by definition, precisely what the majority thinks, taking all things into account, it ought to be. I prefer to rely upon the judgment of the wise men who constructed our system, and of the people who approved it, and of two centuries of history that have shown it to be sound. Like it or not, that judgment says, quite plainly, that “[t]he executive Power shall be vested in a President of the United States.”
I agree with the Court on this point, but not because of the section of the statute that it cites, § 592(f). What that provides is that “[t]he Attorney General’s determination ... to apply to the division of the court for the appointment of an independent counsel shall not be reviewable in any court. ” Quite obviously, the determination to apply is not the same as the determination not to apply. In other contexts, we have sternly avoided “construing” a statute to mean what it plainly does not say, merely in order to avoid constitutional problems. See Commodity Futures Trading Comm’n v. Schor, 478 U. S. 833, 841 (1986). In my view, however, the Attorney General’s decision not to refer would in any event be nonreviewable as the exercise of prosecutorial discretion. See Heckler v. Chaney, 470 U. S. 821 (1985).
The Court omits the further provision that the independent counsel exercises within her sphere the “full power” of “the Attorney General, [with one minor exception relating to wiretap authorizations] and any other officer or employee of the Department of Justice[.]” § 594(a). This is, of course, quite difficult to square with the Court’s assertion that appellant is “ ‘inferior’ in rank and authority” to the Attorney General. Ante, at 671.
The independent counsel’s specifically enumerated powers include the following:
“(1) conducting proceedings before grand juries and other investigations;
“(2) participating in court proceedings and engaging in any litigation, including civil and criminal matters, that [the] independent counsel deems necessary;
“(3) appealing any decision of a court in any case or proceeding in which [the] independent counsel participates in an official capacity;
“(4) reviewing all documentary evidence available from any source;
“(5) determining whether to contest the assertion of any testimonial privilege;
“(6) receiving appropriate national security clearances and, if necessary contesting in court. . . any claim of privilege or attempt to withhold evidence on grounds of national security;
“(7) making applications to any Federal court for a grant of immunity to any witness ... or for warrants, subpoenas, or other court orders, and for purposes of sections 6003, 6004, and 6005 of title 18, exercising the authority vested in a United States attorney or the Attorney General;
“(8) inspecting, obtaining, or using the original or a copy of any tax return . . . ;
“(9) initiating and conducting prosecutions in any court of competent jurisdiction, framing and signing indictments, filing informations, and handling all aspects of any case filed in the name of the United States; and
“(10) consulting with the United States Attorney for the district in which the violation was alleged to have occurred.” §§ 594(a)(1) — (10).
In addition, the statute empowers the independent counsel to hire a staff of a size as large as she “deems necessary,” § 594(c), and to enlist and re*718ceive “where necessary to perform [her] duties” the assistance, personnel and resources of the Department of Justice, § 594(d).
The Court misunderstands my opinion to say that “every officer of the United States exercising any part of [the executive] power must serve at the pleasure of the President and be removable by him at will.” Ante, at 690, n. 29. Of course, as my discussion here demonstrates, that has never been the law and I do not assert otherwise. What I do assert — and what the Constitution seems plainly to prescribe — is that the President must have control over all exercises of the executive power. See supra, at 705. That requires that he have plenary power to remove principal officers such as the independent counsel, but it does not require that he have plenary power to remove inferior officers. Since the latter are, as I have described, subordinate to, i. e., subject to the supervision of, principal officers who (being removable at will) have the President’s complete confidence, it is enough — at least if they have been appointed by the President or by a principal officer — that they be removable for cause, which would include, of course, the failure to accept supervision. Thus, Perkins is in no way inconsistent with my views.