dissenting.
The Court, acting in the name of separation of powers, takes upon itself to strike down the Gramm-Rudman-Hollings Act, one of the most novel and far-reaching legislative responses to a national crisis since the New Deal. The basis of the Court’s action is a solitary provision of another statute that was passed over 60 years ago and has lain dormant since that time. I cannot concur in the Court’s action. Like the Court, I will not purport to speak to the wisdom of the policies incorporated in the legislation the Court invalidates; that is a matter for the Congress and the Executive, both of which expressed their assent to the statute barely half a year ago. I will, however, address the wisdom of the Court’s willingness to interpose its distressingly formalistic view of separation of powers as a bar to the attainment of governmental objectives through the means chosen by the Congress and the President in the legislative process established by the Constitution. Twice in the past four years I have expressed my view that the Court’s recent efforts to police the separation of powers have rested on untenable constitutional propositions leading to regrettable results. See Northern Pipeline Construction Co. v. Marathon Pipe Line Co., 458 U. S. 50, 92-118 (1982) (White, J., dissenting); INS v. Chadha, 462 U. S. 919, 967-1003 (1983) (White, J., dissenting). Today’s result is even more misguided. As I will explain, the Court’s decision rests on a feature of the legislative scheme that is of minimal practical significance and that presents no substantial threat to the basic scheme of separation of powers. In attaching dispositive significance to what should be regarded as a triviality, the Court neglects what has *760in the past been recognized as a fundamental principle governing consideration of disputes over separation of powers:
“The actual art of governing under our Constitution does not and cannot conform to judicial definitions of the power of any of its branches based on isolated clauses or even single Articles torn from context. While the Constitution diffuses power the better to secure liberty, it also contemplates that practice will integrate the dispersed powers into a workable government.” Youngstown Sheet & Tube Co. v. Sawyer, 343 U. S. 579, 635 (1952) (Jackson, J. concurring).
f — i
The Court’s argument is straightforward: the Act vests the Comptroller General with “executive” powers, that is, powers to “[i]nterpre[t] a law enacted by Congress [in order] to implement the legislative mandate,” ante, at 733; such powers may not be vested by Congress in itself or its agents, see Buckley v. Valeo, 424 U. S. 1, 120-141 (1976), for the system of Government established by the Constitution for the most part limits Congress to a legislative rather than an executive or judicial role, see INS v. Chadha, supra; the Comptroller General is an agent of Congress by virtue of a provision in the Budget and Accounting Act of 1921, 43 Stat. 23, 31 U. S. C. § 703(e)(1), granting Congress the power to remove the Comptroller for cause through joint resolution; therefore the Comptroller General may not constitutionally exercise the executive powers granted him in the Gramm-Rudman-Hollings Act, and the Act’s automatic budget-reduction mechanism, which is premised on the Comptroller’s exercise of those powers, must be struck down.
Before examining the merits of the Court’s argument, I wish to emphasize what it is that the Court quite pointedly and correctly does not hold: namely, that “executive” powers of the sort granted the Comptroller by the Act may only be exercised by officers removable at will by the President. *761The Court’s apparent unwillingness to accept this argument,1 which has been tendered in this Court by the Solicitor General,2 is fully consistent with the Court’s longstanding recognition that it is within the power of Congress under the “Necessary and Proper” Clause, Art. I, §8, to vest authority that falls within the Court’s definition of executive power in officers who are not subject to removal at will by the President and are therefore not under the President’s direct control. See, e. g., Humphrey’s Executory. United States, 295 U. S. 602 (1935); Wiener v. United States, 357 U. S. 349 (1958).3 In an earlier day, in which simpler notions of the role of government in society prevailed, it was perhaps plausible to insist that all “executive” officers be subject to an unqualified Presidential removal power, see Myers v. United States, 272 U. S. 52 (1926); but with the advent and triumph of the administrative state and the accompanying multiplication of the tasks undertaken by the Federal Government, the *762Court has been virtually compelled to recognize that Congress may reasonably deem it “necessary and proper” to vest some among the broad new array of governmental functions in officers who are free from the partisanship that may be expected of agents wholly dependent upon the President.
The Court’s recognition of the legitimacy of legislation vesting “executive” authority in officers independent of the President does not imply derogation of the President’s own constitutional authority — indeed, duty — to “take Care that the Laws be faithfully executed,” Art. II, § 3, for any such duty is necessarily limited to a great extent by the content of the laws enacted by the Congress. As Justice Holmes put it: “The duty of the President to see that the laws be executed is a duty that does not go beyond the laws or require him to achieve more than Congress sees fit to leave within his power.” Myers v. United States, supra, at 177 (dissenting).4 Justice Holmes perhaps overstated his case, for there are undoubtedly executive functions that, regardless of the enactments of Congress, must be performed by officers subject to removal at will by the President. Whether a particular function falls within this class or within the far larger class that may be relegated to independent officers “will depend upon the character of the office.” Humphrey’s Executor, supra, at 631. In determining whether a limitation on the President’s power to remove an officer performing executive functions constitutes a violation of the constitutional scheme of separation of powers, a court must “focu[s] on the extent to which [such a limitation] prevents the Executive Branch from accomplishing its constitutionally assigned functions.” Nixon v. Administrator of General Services, 433 U. S. 425, 443 (1977). “Only where the potential for disruption is present must we then determine whether that impact is justified by an overriding need to promote objectives within the constitutional authority of Congress.” Ibid. This inquiry *763is, to be sure, not one that will beget easy answers; it provides nothing approaching a bright-line rule or set of rules. Such an inquiry, however, is necessitated by the recognition that “formalistic and unbending rules” in the area of separation of powers may “unduly constrict Congress’ ability to take needed and innovative action pursuant to its Article I powers.” Commodity Futures Trading Comm’n v. Schor, post, at 851.
It is evident (and nothing in the Court’s opinion is to the contrary) that the powers exercised by the Comptroller General under the Gramm-Rudman-Hollings Act are not such that vesting them in an officer not subject to removal at will by the President would in itself improperly interfere with Presidential powers. Determining the level of spending by the Federal Government is not by nature a function central either to the exercise of the President’s enumerated powers or to his general duty to ensure execution of the laws; rather, appropriating funds is a peculiarly legislative function, and one expressly committed to Congress by Art. I, § 9, which provides that “No Money shall be drawn from the Treasury, but in Consequence of Appropriations made by Law.” In enacting Gramm-Rudman-Hollings, Congress has chosen to exercise this legislative power to establish the level of federal spending by providing a detailed set of criteria for reducing expenditures below the level of appropriations in the event that certain conditions are met. Delegating the execution of this legislation — that is, the power to apply the Act’s criteria and make the required calculations — to an officer independent of the President’s will does not deprive the President of any power that he would otherwise have or that is essential to the performance of the duties of his office. Rather, the result of such a delegation, from the standpoint of the President, is no different from the result of more traditional forms of appropriation: under either system, the level of funds available to the Executive Branch to carry out its duties is not within the President’s discretionary control. To be sure, *764if the budget-cutting mechanism required the responsible officer to exercise a great deal of policymaking discretion, one might argue that having created such broad discretion Congress had some obligation based upon Art. II to vest it in the Chief Executive or his agents. In Gramm-Rudman-Hollings, however, Congress has done no such thing; instead, it has created a precise and articulated set of criteria designed to minimize the degree of policy choice exercised by the officer executing the statute and to ensure that the relative spending priorities established by Congress in the appropriations it passes into law remain unaltered.5 Given that the exercise of policy choice by the officer executing the statute would be inimical to Congress’ goal in enacting “automatic” budget-cutting measures, it is eminently reasonable and proper for Congress to vest the budget-cutting authority in an officer who is to the greatest degree possible nonpartisan and independent of the President and his political agenda and who therefore may be relied upon not to allow his calculations to be colored by political considerations. Such a delegation deprives the President of no authority that is rightfully his.
HH
If, as the Court seems to agree, the assignment of executive” powers under Gramm-Rudman-Hollings to an officer not removable at will by the President would not in itself represent a violation of the constitutional scheme of separated *765powers, the question remains whether, as the Court concludes, the fact that the officer to whom Congress has delegated the authority to implement the Act is removable by a joint resolution of Congress should require invalidation of the Act. The Court’s decision, as I have stated above, is based on a syllogism: the Act vests the Comptroller with “executive power”; such power may not be exercised by Congress or its agents; the Comptroller is an agent of Congress because he is removable by Congress; therefore the Act is invalid. I have no quarrel with the proposition that the powers exercised by the Comptroller under the Act may be characterized as “executive” in that they involve the interpretation and carrying out of the Act’s mandate. I can also accept the general proposition that although Congress has considerable authority in designating the officers who are to execute legislation, see supra, at 760-764, the constitutional scheme of separated powers does prevent Congress from reserving an executive role for itself or for its “agents.” Buckley v. Valeo, 424 U. S., at 120-141; id., at 267-282 (White, J., concurring in part and dissenting in part). I cannot accept, however, that the exercise of authority by an officer removable for cause by a joint resolution of Congress is analogous to the impermissible execution of the law by Congress itself, nor would I hold that the congressional role in the removal process renders the Comptroller an “agent” of the Congress, incapable of receiving “executive” power.
In Buckley v. Valeo, supra, the Court held that Congress could not reserve to itself the power to appoint members of the Federal Election Commission, a body exercising “executive” power. Buckley, however, was grounded on a textually based separation-of-powers argument whose central premise was that the Constitution requires that all “Officers of the United States” (defined as “all persons who can be said to hold an office under the government,” 424 U. S., at 126) whose appointment is not otherwise specifically provided for elsewhere in its text be appointed through the means speci*766fied by the Appointments Clause, Art. II, §2, cl. 2 — that is, either by the President with the advice and consent of the Senate or, if Congress so specifies, by the President alone, by the courts, or by the head of a department. The Buckley Court treated the Appointments Clause as reflecting the principle that “the Legislative Branch may not exercise executive authority,” 424 U. S., at 119 (citing Springer v. Philippine Islands, 277 U. S. 189 (1928)), but the Court’s holding was merely that Congress may not direct that its laws be implemented through persons who are its agents in the sense that it chose them; the Court did not pass on the legitimacy of other means by which Congress might exercise authority over those who execute its laws. Because the Comptroller is not an appointee of Congress but an officer of the United States appointed by the President with the advice and consent of the Senate, Buckley neither requires that he be characterized as an agent of the Congress nor in any other way calls into question his capacity to exercise “executive” authority. See 424 U. S., at 128, n. 165.
As the majority points out, however, the Court’s decision in INS v. Chadha, 462 U. S. 919 (1983), recognizes additional limits on the ability of Congress to participate in or influence the execution of the laws. As interpreted in Chadha, the Constitution prevents Congress from interfering with the actions of officers of the United States through means short of legislation satisfying the demands of bicameral passage and presentment to the President for approval or disapproval. Id., at 954-955. Today’s majority concludes that the same concerns that underlay Chadha indicate the invalidity of a statutory provision allowing the removal by joint resolution for specified cause of any officer performing executive functions. Such removal power, the Court contends, constitutes a “congressional veto” analogous to that struck down in Chadha, for it permits Congress to “remove, or threaten to remove, an officer for executing the laws in any fashion found to be unsatisfactory.” Ante, at 726. The Court concludes *767that it is “[t]his kind of congressional control over the execution of the laws” that Chadha condemns. Ante, at 726-727.
The deficiencies in the Court’s reasoning are apparent. First, the Court baldly mischaracterizes the removal provision when it suggests that it allows Congress to remove the Comptoller for “executing the laws in any fashion found to be unsatisfactory”; in fact, Congress may remove the Comptroller only for one or more of five specified reasons, which “although not so narrow as to deny Congress any leeway, circumscribe Congress’ power to some extent by providing a basis for judicial review of congressional removal.” Ameron, Inc. v. United States Army Corps of Engineers, 787 F. 2d 875, 895 (CA3 1986) (Becker, J., concurring in part). Second, and more to the point, the Court overlooks or deliberately ignores the decisive difference between the congressional removal provision and the legislative veto struck down in Chadha: under the Budget and Accounting Act, Congress may remove the Comptroller only through a joint resolution, which by definition must be passed by both Houses and signed by the President. See United States v. California, 332 U. S. 19, 28 (1947).6 In other words, a removal of the Comptroller under the statute satisfies the requirements of bicameralism and presentment laid down in Chadha. The majority’s citation of Chadha for the proposition that Congress may only control the acts of officers of the United States “by passing new legislation,” ante, at 734, in *768no sense easts doubt on the legitimacy of the removal provision, for that provision allows Congress to effect removal only through action that constitutes legislation as defined in Chadha.
To the extent that it has any bearing on the problem now before us, Chadha would seem to suggest the legitimacy of the statutory provision making the Comptroller removable through joint resolution, for the Court’s opinion in Chadha reflects the view that the bicameralism and presentment requirements of Art. I represent the principal assurances that Congress will remain within its legislative role in the constitutionally prescribed scheme of separated powers. Action taken in accordance with the “single, finely wrought, and exhaustively considered, procedure” established by Art. I, Chadha, supra, at 951, should be presumptively viewed as a legitimate exercise of legislative power. That such action may represent a more or less successful attempt by Congress to “control” the actions of an officer of the United States surely does not in itself indicate that it is unconstitutional, for no one would dispute that Congress has the power to “control” administration through legislation imposing duties or substantive restraints on executive officers, through legislation increasing or decreasing the funds made available to such officers, or through legislation actually abolishing a particular office. Indeed, Chadha expressly recognizes that while congressional meddling with administration of the laws outside of the legislative process is impermissible, congressional control over executive officers exercised through the legislative process is valid. 462 U. S., at 955, n. 19. Thus, if the existence of a statute permitting removal of the Comptroller through joint resolution (that is, through the legislative process) renders his exercise of executive powers unconstitutional, it is for reasons having virtually nothing to do with Chadha.7
*769That a joint resolution removing the Comptroller General would satisfy the requirements for legitimate legislative action laid down in Chadha does not fully answer the separation-of-powers argument, for it is apparent that even the results of the constitutional legislative process may be unconstitutional if those results are in fact destructive of the scheme of separation of powers. Nixon v. Administrator of General *770Services, 433 U. S. 425 (1977). The question to be answered is whether the threat of removal of the Comptroller General for cause through joint resolution as authorized by the Budget and Accounting Act renders the Comptroller sufficiently subservient to Congress that investing him with “executive” power can be realistically equated with the unlawful retention of such power by Congress itself; more generally, the question is whether there is a genuine threat of “encroachment or aggrandizement of one branch at the expense of the other,” Buckley v. Valeo, 424 U. S., at 122. Common sense indicates that the existence of the removal provision poses no such threat to the principle of separation of powers.
The statute does not permit anyone to remove the Comptroller at will; removal is permitted only for specified cause, with the existence of cause to be determined by Congress following a hearing. Any removal under the statute would presumably be subject to post-termination judicial review to ensure that a hearing had in fact been held and that the finding of cause for removal was not arbitrary. See Ameron, Inc. v. United States Army Corps of Engineers, 787 F. 2d, at 895 (Becker, J., concurring in part).8 These procedural and substantive limitations on the removal power militate strongly against the characterization of the Comptroller as a mere agent of Congress by virtue of the removal authority. Indeed, similarly qualified grants of removal power are generally deemed to protect the officers to whom they apply and to establish their independence from the domination of the possessor of the removal power. See Humphrey’s Executor v. United States, 295 U. S., at 625-626, 629-630. Removal authority limited in such a manner is more properly viewed as motivating adherence to a substantive standard established by law than as inducing subservience to the particular *771institution that enforces that standard. That the agent enforcing the standard is Congress may be of some significance to the Comptroller, but Congress’ substantively limited removal power will undoubtedly be less of a spur to subservience than Congress’ unquestionable and unqualified power to enact legislation reducing the Comptroller’s salary, cutting the funds available to his department, reducing his personnel, limiting or expanding his duties, or even abolishing his position altogether.
More importantly, the substantial role played by the President in the process of removal through joint resolution reduces to utter insignificance the possibility that the threat of removal will induce subservience to the Congress. As I have pointed out above, a joint resolution must be presented to the President and is ineffective if it is vetoed by him, unless the veto is overridden by the constitutionally prescribed two-thirds majority of both Houses of Congress. The requirement of Presidential approval obviates the possibility that the Comptroller will perceive himself as so completely at the mercy of Congress that he will function as its tool.9 If the Comptroller’s conduct in office is not so unsatisfactory to the President as to convince the latter that removal is required under the statutory standard, Congress will have no independent power to coerce the Comptroller unless it can muster a two-thirds majority in both Houses — a feat of bipartisanship more difficult than that required to impeach and convict. The incremental in terrorem effect of the possibility of congressional removal in the face of a Presidential *772veto is therefore exceedingly unlikely to have any discernible impact on the extent of congressional influence over the Comptroller.10
*773The practical result of the removal provision is not to render the Comptroller unduly dependent upon or subservient to Congress, but to render him one of the most independent officers in the entire federal establishment. Those who have studied the office agree that the procedural and substantive limits on the power of Congress and the President to remove the Comptroller make dislodging him against his will practically impossible. As one scholar put it nearly 50 years ago: “Under the statute the Comptroller General, once confirmed, is safe so long as he avoids a public exhibition of personal immorality, dishonesty, or failing mentality.” H. Mansfield, The Comptroller General 75-76 (1939).11 The passage of time has done little to cast doubt on this view: of the six Comptrollers who have served since 1921, none has been threatened with, much less subjected to, removal. Recent students of the office concur that “[b]arring resignation, death, physical or mental incapacity, or extremely bad behavior, the Comptroller General is assured his tenure if he wants it, and not a day more.” F. Mosher, The GAO 242 (1979).12 The threat of “here-and-now subservience,” ante, at 720, is obviously remote indeed.13
*774Realistic consideration of the nature of the Comptroller General’s relation to Congress thus reveals that the threat to separation of powers conjured up by the majority is wholly chimerical. The power over removal retained by the Congress is not a power that is exercised outside the legislative process as established by the Constitution, nor does it appear likely that it is a power that adds significantly to the influence Congress may exert over executive officers through other, undoubtedly constitutional exercises of legislative power and through the constitutionally guaranteed impeachment power. Indeed, the removal power is so constrained by its own substantive limits and by the requirement of Presidential ap*775proval “that, as a practical matter, Congress has not exercised, and probably will never exercise, such control over the Comptroller General that his non-legislative powers will threaten the goal of dispersion of power, and hence the goal of individual liberty, that separation of powers serves.” Ameron, Inc. v. United States Army Corps of Engineers, 787 F. 2d, at 895 (Becker, J., concurring in part).14
*776The majority’s contrary, conclusion rests on the rigid dogma that, outside of the impeachment process, any “direct congressional role in the removal of officers charged with the execution of the laws ... is inconsistent with separation of powers.” Ante, at 723. Reliance on such an unyielding principle to strike down a statute posing no real danger of aggrandizement of congressional power is extremely misguided and insensitive to our constitutional role. The wisdom of vesting “executive” powers in an officer removable by joint resolution may indeed be debatable — as may be the wisdom of the entire scheme of permitting an unelected official to revise the budget enacted by Congress — but such matters are for the most part to be worked out between the Congress and the President through the legislative process, which affords each branch ample opportunity to defend its interests. The Act vesting budget-cutting authority in the Comptroller General represents Congress’ judgment that the delegation of such authority to counteract ever-mounting deficits is “necessary and proper” to the exercise of the powers granted the Federal Government by the Constitution; and the President’s approval of the statute signifies his unwillingness to reject the choice made by Congress. Cf. Nixon v. Administrator of General Services, 433 U. S., at 441. Under such circumstances, the role of this Court should be limited to determining whether the Act so alters the balance of authority among the branches of government as to pose a genuine threat to the basic division between the lawmaking power and the power to execute the law. Because I see no such threat, I cannot join the Court in striking down the Act.
I dissent.
See ante, at 724-726, and n. 4.
The Solicitor General appeared on behalf of the “United States,” or, more properly, the Executive Departments, which intervened to attack the constitutionality of the statute that the Chief Executive had earlier endorsed and signed into law.
Although the Court in Humphrey’s Executor characterized the powers of the Federal Trade Commissioner whose tenure was at issue as “quasi-legislative” and “quasi-judicial,” it is clear that the FTC’s power to enforce and give content to the Federal Trade Commission Act’s proscription of “unfair” acts and practices and methods of competition is in fact “executive” in the same sense as is the Comptroller’s authority under Gramm-Rudman-Hollings — that is, it involves the implementation (or the interpretation and application) of an Act of Congress. Thus, although the Court in Humphrey’s Executor found the use of the labels “quasi-legislative” and “quasi-judicial” helpful in “distinguishing” its then-recent decision in Myers v. United States, 272 U. S. 62 (1926), these terms are hardly of any use in limiting the holding of the case; as Justice Jackson pointed out, “[t]he mere retreat to the qualifying ‘quasi’ is implicit with confession that all recognized classifications have broken down, and ‘quasi’ is a smooth cover which we draw over our confusion as we might use a counterpane to conceal a disordered bed.” FTC v. Ruberoid Co., 343 U. S. 470, 487-488 (1952) (dissenting).
Cf. ante, at 733 (“[UJndoubtedly the content of the Act determines the nature of the executive duty”).
That the statute provides, to the greatest extent possible, precise guidelines for the officer assigned to carry out the required budget cuts not only indicates that vesting budget-cutting authority in an officer independent of the President does not in any sense deprive the President of a significant amount of discretionary authority that should rightfully be vested in him or an officer accountable to him, but also answers the claim that the Act represents an excessive and hence unlawful delegation of legislative authority. Because the majority does not address the delegation argument, I shall not discuss it at any length, other than to refer the reader to the District Court’s persuasive demonstration that the statute is not void under the nondelegation doctrine.
The legislative history indicates that the inclusion of the President in the removal process was a deliberate choice on the part of the Congress that enacted the Budget and Accounting Act. The previous year, legislation establishing the position of Comptroller General and providing for removal by concurrent resolution — that is, by a resolution not presented to the President — had been vetoed by President Wilson on the ground that granting the sole power of removal to the Congress would be unconstitutional. See 59 Cong. Ree. 8609-8610 (1920). That Congress responded by providing for removal through joint resolution clearly evinces congressional intent that removal take place only through the legislative process, with Presidential participation.
Because a joint resolution passed by both Houses of Congress and signed by the President (or repassed over the President’s veto) is legisla*769tion having the same force as any other Act of Congress, it is somewhat mysterious why the Court focuses on the Budget and Accounting Act’s authorization of removal of the Comptroller through such a resolution as an indicator that the Comptroller may not be vested with executive powers. After all, even without such prior statutory authorization, Congress could pass, and the President sign, a joint resolution purporting to remove the Comptroller, and the validity of such legislation would seem in no way dependent on previous legislation contemplating it. Surely the fact that Congress might at any time pass and the President sign legislation purporting to remove some officer of the United States does not make the exercise of executive power by all such officers unconstitutional. Since the effect of the Budget and Accounting Act is merely to recognize the possibility of legislation that Congress might at any time attempt to enact with respect to any executive officer, it should not make the exercise of “executive” power by the Comptroller any more problematic than the exercise of such power by any other officer. A joint resolution purporting to remove the Comptroller, or any other executive officer, might be constitutionally infirm, but Congress’ advance assertion of the power to enact such legislation seems irrelevant to the question whether exercise of authority by an officer who might in the future be subject to such a possibly valid and possibly invalid resolution is permissible, since the provision contemplating a resolution of removal obviously cannot in any way add to Congress’ power to enact such a resolution.
Of course, the foregoing analysis does not imply that the removal provision of the Budget and Accounting Act is meaningless; for although that provision cannot add to any power Congress might have to pass legislation (that is, a joint resolution) removing the Comptroller, it can limit its power to do so to the circumstances specified. The reason for this is that any joint resolution purporting to remove the Comptroller in the absence of a hearing or one of the specified grounds for removal would not be deemed an implied repeal of the limits on removal in the 1921 Act (for such implied repeals are disfavored), and thus the joint resolution would only be given effect to the extent consistent with the pre-existing law (that is, to the extent that there was actually cause for removal).
Cf. Humphrey’s Executor v. United States, 295 U. S. 602 (1935), in which the Court entertained a challenge to Presidential removal under a statute that similarly limited removals to specified cause.
The Court cites statements made by supporters of the Budget and Accounting Act indicating their belief that the Act’s removal provisions would render the Comptroller subservient to Congress by giving Congress “ ‘absolute control of the man’s destiny in office.’ ” Ante, at 728. The Court’s scholarship, however, is faulty: at the time all of these statements were made — including Representative Sisson’s statement of May 3, 1921 — the proposed legislation provided for removal by concurrent resolution, with no Presidential role. See 61 Cong. Rec. 983, 989-992, 1079-1086 (1921).
Concededly, the substantive grounds for removal under the statute are broader than the grounds for impeachment specified by the Constitution, see ante, at 729-730, although given that it is unclear whether the limits on the impeachment power may be policed by any body other than Congress itself, the practical significance of the difference is hard to gauge. It seems to me most likely that the difficulty of obtaining a two-thirds vote for removal in both Houses would more than offset any increased likelihood of removal that might result from the greater liberality of the substantive grounds for removal under the statute. And even if removal by Congress alone through joint resolution passed over Presidential veto is marginally more likely than impeachment, whatever additional influence over the Comptroller Congress may thereby possess seems likely to be minimal in relation to that which Congress already possesses by virtue of its general legislative powers and its power to impeach. Of course, if it were demonstrable that the Constitution specifically limited Congress’ role in removal to the impeachment process, the insignificance of the marginal increase in congressional influence resulting from the provision authorizing removal through joint resolution would be no answer to a claim of unconstitutionality. But no such limit appears in the Constitution: the Constitution merely provides that all officers of the United States may be impeached for high crimes and misdemeanors, and nowhere suggests that impeachment is the sole means of removing such officers.
As for the Court’s observation that “no one would seriously suggest that judicial independence would be strengthened by allowing removal of federal judges only by a joint resolution finding ‘inefficiency,’ ‘neglect of duty,’ or ‘malfeasance,’ ” ante, at 730, it can only be described as a non sequitur. The issue is not whether the removal provision makes the Comptroller more independent than he would be if he were removable only through impeachment, but whether the provision so weakens the Comptroller that he may not exercise executive authority. Moreover, the Court’s reference to standards applicable to removal of Art. Ill judges is a red herring, for Art. Ill judges —unlike other officers of the United States —are specifically protected against removal for other than constitutionally specified cause. Thus, the infirmity of a statute purporting to allow removal of judges for some other reason would be that it violated the specific command of Art. III. In the absence of a similar textual limit on the removal of nonjudicial officers, the test for a violation of separation of powers should be whether an asserted congressional power to remove would constitute a real and sub*773stantial aggrandizement of congressional authority at the expense of executive power, not whether a similar removal provision would appear problematic if applied to federal judges.
The author of this statement was no apologist for the Comptroller; rather, his study of the office is premised on the desirability of Presidential control over many of the Comptroller’s functions. Nonetheless, he apparently found no reason to accuse the Comptroller of subservience to Congress, and he conceded that “[t]he political independence of the office has in fact been one of its outstanding characteristics.” H. Mansfield, The Comptroller General 75 (1939).
Professor Mosher’s reference to the fact that the Comptroller is limited to a single term highlights an additional source of independence: unlike an officer with a fixed term who may be reappointed to office, the Comptroller need not concern himself with currying favor with the Senate in order to secure its consent to his reappointment.
The majority responds to the facts indicating the practical independence of the Comptroller from congressional control by cataloging a series of *774statements and materials categorizing the Comptroller as a part of the “Legislative Branch.” Ante, at 730-732. Such meaningless labels are quite obviously irrelevant to the question whether in actuality the Comptroller is so subject to congressional domination that he may not participate in the execution of the laws.
Justice Stevens, for his part, finds that the Comptroller is an “agent” of Congress, and thus incapable of wielding the authority granted him by the Act, because his responsibilities under a variety of statutes include making reports to the Congress. Justice Stevens’ position is puzzling, to say the least. It seems to rest on the view that an officer required to perform certain duties for the benefit of Congress somehow becomes a part of Congress for all purposes. But it is by no means true that an officer who must perform specified duties for some other body is under that body’s control or acts as its agent when carrying out other, unrelated duties. As Justice Blackmun points out, see post, at 778-779, n. 1, duties toward Congress are imposed on a variety of agencies, including the Federal Trade Commission; and certainly it cannot credibly be maintained that by virtue of those duties the agencies become branches of Congress, incapable of wielding governmental power except through the legislative process. Indeed, the President himself is under numerous obligations, both statutory and constitutional, to provide information to Congress, see, e. g., Art. II, § 3, cl. 1; surely the President is not thereby transformed into an arm or agency of the Congress. If, therefore, as Justice Stevens concedes, see ante, at 737-741, the provision authorizing removal of the Comptroller by joint resolution does not suffice to establish that he may not exercise the authority granted him under Gramm-Rudman-Hollings, I see no substantial basis for concluding that his various duties toward Congress render him incapable of receiving such power.
Even if I were to concede that the exercise of executive authority by the Comptroller is inconsistent with the removal provision, I would agree with Justice Blackmun that striking down the provisions of the Gramm-Rudman-Hollings Act vesting the Comptroller with such duties is a grossly inappropriate remedy for the supposed constitutional infirmity, and that if one of the features of the statutory scheme must go, it should be the removal provision. As Justice Blackmun points out, the mere fact that the parties before the Court have standing only to seek invalidation of the Gramm-Rudman-Hollings spending limits cannot dictate that the Court resolve any constitutional incompatibility by striking down Gramm-Rudman-Hollings. Nor does the existence of the fallback provisions in Gramm-Rudman-Hollings indicate the appropriateness of the Court’s choice, for those provisions, by their terms, go into effect only if the Court finds that the primary budget-cutting mechanism established by the Act must be invalidated; they by no means answer the antecedent question whether the Court should take that step.
Given the majority’s constitutional premises, it is clear to me that the decision whether to strike down Gramm-Rudman-Hollings must depend on whether such a choice would be more or less disruptive of congressional objectives than declaring the removal provision invalid (with the result that the Comptroller would still be protected against removal at will by the President, but could also not be removed through joint resolution). When the choice is put in these terms, it is evident that it is the never-used removal provision that is far less central to the overall statutory scheme. That this is so is underscored by the fact that under the majority’s theory, the removal provision was never constitutional, as the Comptroller’s primary duties under the 1921 Act were clearly executive under the Court’s definition: the Comptroller’s most important tasks under that legislation were to dictate accounting techniques for all executive agencies, to audit all federal expenditures, and to approve or disapprove disbursement of funds. See F. Mosher, The GAO (1979). Surely the Congress in 1921 would have sacrificed its own role in removal rather than allow such duties to go unfulfilled by a Comptroller independent of the President. See 59 Cong. Rec. 8611 (1920).