Links to author's official homepage and unofficial homepage.

This page has been accessed times since Nov. 13, 1996. Originally published at 88 Nw. U. L. Rev. 1346 (1994).

Copyright (c) 1994 A. Michael Froomkin. All rights reserved. Permission to reproduce granted for any non-profit, educational, use.



A. Michael Froomkin<*>

All but two officials in the executive branch occupy posts and exercise powers that owe their existence to an act of Congress,<1> yet the extent of Congress's power to give top executive branch officers protection from dismissal for policy differences with the President remains contested.<2> Justice Scalia has suggested in forceful dissents that the Constitution gives the President the power to make all policy choices delegated by Congress to the executive branch, and that restrictions on the President's power to fire policy-making persons are therefore unconstitutional.<3> This view has recently been restated by Professor Steven Calabresi and Mr. Kevin Rhodes in The Structural Constitution,<4> which purports to find structural constitutional support for Justice Scalia's view that Congress may not vest any policy-making discretion in the hands of executive branch officials who are not subject either to presidential commands or to presidential dismissal.

Part I of this Article summarizes the debate about the extent of *1347 Congress's power to restrict the President's power over officials in the executive branch. Part II summarizes and critiques Professor Calabresi and Mr. Rhodes's main contentions regarding the importance of certain clauses in the Constitution, and the interplay between Article II and Article III. In Part II, I suggest that The Structural Constitution's argument for absolute presidential control over the executive branch asks the wrong question and, even on its own highly textualist terms, comes up with implausible answers.

Having criticized The Structural Constitution for not being structural enough, Part III offers the outline of a truly structural approach to the constraints on Congress's ability to design the executive branch. Contrary to what Professor Calabresi and Mr. Rhodes suggest, a proper structural analysis of the Constitution undermines the constitutional case for an executive branch with a chain of command organized along military lines and instead emphasizes the existence of a discernible balance between Congress's role in structuring the executive and the President's inherent and default powers. I argue that this more contextual approach better comports with both constitutional text and the holdings (but not always the reasoning) of all the relevant Supreme Court decisions and furthers the most important objective that the separation of powers is designed to serve, that of hindering tyranny.

I. Unitarians, Trinitarians, and Other Heretics

It is rarely disputed that Congress can abolish the job of any unelected official in the executive branch if it chooses.<5> Counterbalancing this power, it is now also clear that, unless the Constitution specifically provides otherwise, Congress itself may affect the rights, duties, or responsibilities of persons outside the legislative branch only by legislation duly presented to the President for signature or by impeachment.<6> But 200 years of administrative practice and judicial analysis have failed to produce a consensus regarding the extent to which the Constitution requires that policy- making ( i.e., "nonministerial"<7>) officers in the executive *1348 branch act as the President instructs. Instead, there are two camps with diametrically opposed views.

"Unitarians" believe that the Constitution makes the President the Chief Administrator to whom all other members of the executive branch are fully subordinate.<8> Unitarians further believe that any statute which does not allow the President to fire a subordinate, or at least to countermand her decisions as to policy matters, is for that reason unconstitutional. Unitarians thus argue that independent agencies violate the separation of powers. The unitarian style of constitutional interpretation tends to be formalist, although some unitarians also rely on their understanding of the Framers' intentions, on one reading of early U.S. administrative practice, and on particular Supreme Court decisions such as Myers v. United States,<9> in which Chief Justice Taft set out a strong case for presidential supremacy over the executive branch. Today, Justice Scalia is the leading exponent of the unitarian view, but many administrations, particularly the Reagan Administration, have also taken forceful unitarian positions in litigation.<10>

Nonunitarians, whom one might call trinitarians,<11> do not agree that the only means by which Congress can create a statutory program is to vest full control of it in the hands of the President. Nonunitarians agree that (most or all) existing independent agencies are constitutional and usually agree that Myers should be limited to its facts, but they have had some difficulty producing a consensus theory of why the Constitution permits Congress to create so-called independent agencies. These difficulties are in part attributable to Humphrey's Executor v. United States,<12> which effectively overturned much of the decision in Myers. By confirming that Congress had the power to insulate at least some policy-making officials from presidential dismissal, Humphrey's Executor permitted the modern independent agency. But it did so in a manner that created a proliferation of theories suggesting that agency independence turns on *1349 identifying the agency's powers as "quasi- executive" or "quasi-judicial"--terms whose meanings are at best quasi-clear. This lack of clarity even prompted some errant nonunitarians to consider abandoning the trinity and admitting a "fourth branch" to the pantheon. Although the trinitarian mainstream has avoided that error and thus recognizes that so-called independent agencies are in fact a part of the executive branch, the trinitarian mainstream's attempts to reconcile Humphrey's Executor with the separation of powers have been troubled by confusion, much of it caused by the Supreme Court.

Most modern Supreme Court separation of powers decisions have tended to produce outcomes consistent with the nonunitarian position, but the unitarians nonetheless have had reasons to believe that they were not defeated.<13> Even when the formalist/unitarian position lost, the Court displayed some unitarian tendencies. For example, on several occasions the Court has suggested that excessive intermingling of functions in hybrid entities might make those bodies unconstitutional.<14> Worse--from the trinitarian viewpoint--many of the Court's most trinitarian opinions rely on an uneasy mixture of fundamentally inconsistent doctrines.<15> Thus, although nonunitarians have been ascendant since Humphrey's Executor, and the most recent decisions tend toward a functional approach in which Congress can limit the removal power when the Supreme Court is persuaded that the office requires a degree of political independence,<16> the unitarians have never admitted defeat, and the trinitarians have felt compelled to take the unitarians seriously. Perhaps mindful that history repeats itself, the unitarians have been able to argue that next time Humphrey's Executor may be put to rest.<17>

*1350 II. The New Unitarian Challenge

The latest and most comprehensive unitarian challenge can be found in The Structural Constitution.<18> The article is significant because it proffers what the unitarian vision has mostly lacked--a structural underpinning for what has tended to be a purely textual, even formalistic, account of interbranch relations.<19>

Even more significantly, Professor Calabresi and Mr. Rhodes also assert that the unitarian view of the executive branch follows naturally from any reasonable reading of the judiciary's powers under Article III. This is a powerful claim, and it makes The Structural Constitution the most ambitious and most comprehensive modern attempt to justify the unitarian vision of Article II. Unfortunately, upon close examination the vision proves to be little more than elegant textualism; worse, it is unpersuasive textualism.<20>

A. The Vesting Thesis

Each of the first three articles of the Constitution begins with a Vesting Clause:

"All legislative Powers herein granted shall be vested in a Congress . . . ."<21>

"The executive Power shall be vested in a President of the United States of America."<22>

"The 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."<23>

The Structural Constitution's central claim echoes Justice Scalia's assertion that the "shall be vested" clauses of Articles II and III constitute significant and discernible grants of power to the President and the courts in a way which the "narrower" Vesting Clause of Article I, qualified *1351 as it is by the words "herein granted," does not.<24> I will call this the "Vesting thesis."

Professor Calabresi and Mr. Rhodes assert that the Vesting thesis is the most plausible reading of the Vesting Clause in Article III and that because the words are so similar, the Vesting thesis must apply to Article II also. The authors also argue that the similarity between the Vesting Clauses of Articles II and III suggests that any reading of Article II which fails to establish the President as Chief Administrator would require an interpretation of Article III that would undermine the independence of the federal judiciary. Since a captive judiciary is plainly undesirable, they conclude that the Vesting thesis must be correct. From this foundation, and drawing on a series of articles by Professor Akhil Reed Amar,<25> the authors proceed to dismiss arguments that Congress may in some circumstances limit the President's removal power over executive branch officials.<26>

B. "The judicial Power shall be vested . . ."

Professor Calabresi and Mr. Rhodes introduce their expansive view of the President's inherent Article II powers by analogy to their interpretation of the judiciary's Article III powers. They assert that the Vesting Clause of Article III "must" be a major grant of power to the federal courts because it "appears to be the only explicit constitutional source of the federal judiciary's authority to act."<27> This assertion is critical to their enterprise, for it sows the seed of their consistent interpretation of the Vesting Clause in Article II as an expansive, unenumerated empowerment *1352 of the President.<28> Certainly, if we believe (or adopt as a rule of construction) that the Framers were careful drafters, it would be strange to read the article of the Constitution establishing the judiciary as neglecting to give it authority to act. Thus, if Professor Calabresi and Mr. Rhodes are correct that the federal judiciary would lack the authority to act without the Vesting Clause of Article III, then they have made a strong case for their reading of both Article II and Article III. Rebutting that case would require a persuasive account of why the structures of Article II and Article III differ.<29>

The linchpin claim, therefore, is that there is no other language in the text of Article III that gives the judiciary the power to decide cases, and therefore, that power can be derived only from the Vesting Clause, which, after all, "vests the judicial power." Certainly, on its own terms there is nothing untoward about a reading of Article III that finds the judicial power emanating from the Vesting Clause. This is, however, far from the only plausible reading, whether one approaches the problem structurally or textually.

From a structural point of view, Professor Calabresi and Mr. Rhodes's approach is all trees and no forest; it is no less (or more) than an exacting form of textualism--comparing similar language from different clauses and articles. In fact, it is structurally incoherent to suggest that the judiciary owes its power to decide cases to the Vesting Clause of Article III. A structural approach to the Constitution would reason from the existence of three "co-equal" branches of government rather than from the words "shall be vested"-- or any other clause. The recognition of the judiciary's equal constitutional stature, deriving primarily from the powers granted to the Supreme Court in the text of Article III, suffices to demonstrate to a structuralist that the judiciary must have the power to decide cases. The existence of a Supreme Court staffed with Justices who have life tenure and both original jurisdiction and powers of appellate review suffices to show that the institution must have something important to do. The failure to confront this uncomplicated and fundamentally structural insight illustrates how far the new unitarians remain from actually viewing the Constitution structurally.<30> The existence of a full article defining the judicial branch in a short text just drives home the point.<31>

As textualism, The Structural Constitution also falls short.<32> Even on purely textual grounds it cannot seriously be suggested that the federal *1353 judiciary would lack the authority to decide cases if there were no Vesting Clause in Article III. One does not have to search very hard or far in Article III to find a firm textual basis for the federal courts' authority to act: it lies in the Heads of Jurisdiction.<33> Jurisdiction has long been understood as the authority to decide, which is what courts do when they act.<34> In defining the conditions that must be met in order for a court to exercise its judicial power, a jurisdictional provision is a grant of that power when those conditions are met.<35> Professor Calabresi and Mr. Rhodes dismiss the idea that the jurisdictional grants in Article III are substantive grants of power, saying only that "these jurisdictional 'grants' . . . merely describe the categories of cases or controversies to which 'the judicial power shall extend' (and thereby circumscribe it as well)."<36> They do not explain how it might be that the judicial power can extend to a given category of cases without necessarily requiring that the judiciary have the power to decide those cases.

In view of its centrality to their overall argument, Professor Calabresi and Mr. Rhodes's denigration of the substantive import of the Heads of Jurisdiction deserves careful attention. First, contrary authority is dismissed in a footnote which acknowledges that "some courts" read the jurisdictional grants as affirmative grants of authority to the federal courts rather than limitations on a general grant in the Vesting Clause.<37> As the "some courts" are the Supreme Court in two decisions spanning this century, this viewpoint is perhaps worth more than a footnote.<38> Kansas v. Colorado, decided in 1907, accepts that the Vesting *1354 Clause of Article III vests the "entire judicial power of the Nation," unlike the Vesting Clause in Article I, which "does not make a general grant of legislative power."<39> Were it not for what follows, this would support the unitarian view. However, Justice Brewer also stated that the enumeration in Article III, Section 2 "is a definite declaration, a provision that the judicial power shall extend to--that is, shall include--the several matters particularly mentioned."<40> Thus, to the limited extent that the 1907 Court had considered the matter, it rejected the suggestion put forward in The Structural Constitution that the Vesting Clause is the sole source of the Supreme Court's jurisdiction.<41> By 1965, in Texas v. New Jersey, the Supreme Court appears to have found the idea that the heads of jurisdiction are grants of power so obvious as to merit no discussion.<42>

Professor Calabresi and Mr. Rhodes state that the Supreme Court's acceptance of the Heads of Jurisdiction as actual grants flies in the face of "criticism by knowledgeable commentators."<43> The sole commentator on whom they rely, Professor Martin Redish, indeed argues that the courts erred in reading the Heads of Jurisdiction as grants of power to the judiciary; he forthrightly takes the matter to its logical conclusion and suggests that the federal common law of admiralty should be abandoned as illegitimate.<44> One wonders whether Professor Calabresi and Mr. Rhodes agree and whether they expect their readers to agree. For example, the Calabresi - Rhodes view would seem to foreclose the federal courts from a common-law-making role over wrongful death actions at sea in admiralty, despite Moragne v. States Marine Lines, Inc.<45> One wonders whether the authors would vest this power in the states-- creating inconsistent regulation--and whether they would read the constitutional *1355 clauses giving Congress power over seaborne war and felony<46> as meaning that there is no power vested in Congress to regulate nonfelonious peacetime civilian life on the high seas.

In fact, Professor Calabresi and Mr. Rhodes do little more than assert that the Heads of Jurisdiction cannot serve as a textual foundation for the federal judiciary's authority to decide cases.<47> They appear to believe that the Heads of Jurisdiction cannot be substantive grants of power because the Constitution states that "The judicial Power shall extend to"<48> them rather than "The judiciary shall have Power to decide."<49> As I understand it from discussions with Professor Amar, the distinction is thought to be significant because the (absent) phrase "The judiciary shall have Power to decide" would not only identify the power to decide, but explain who would have it--the federal judiciary. The (extant) phrase "Judicial Power shall extend to" defines a power, but fails to make clear who may exercise that power, thus leaving open the possibility that the judicial power in question could be exercised by the states.<50> The argument is that the crucial task of giving this power to the federal courts as opposed to the state courts is one that is, and can only be, performed by the Vesting Clause.

Professor Amar makes an appealing argument, particularly when one considers that the Supreme Court's jurisdiction consists of both original and appellate matters.<51> An appellate jurisdiction implies the existence of a court to appeal from: if there are no inferior federal courts, then that would have to be a state court. Even if one adopts Professor Amar's argument, however, it does not support the Vesting thesis, which is the weight that Professor Calabresi and Mr. Rhodes seek to make it carry.<52> Professor Amar accepts that Congress had no obligation to create inferior federal courts so long as the Supreme Court retains the final power of appeal in certain areas.<53> In the absence of inferior federal courts, those federal judicial powers which are not reserved to the Supreme Court, or *1356 to federal courts in general, can be exercised by state courts alone, Vesting Clause or no Vesting Clause.<54> The issue then becomes one of identifying which federal judicial powers are reserved to the federal courts in general or the Supreme Court alone and which other ones can be exercised by the state courts.<55> The Constitution's guarantee of a right to a fair forum for certain classes of cases can be satisfied by trial in state courts and appeal to the Supreme Court. Similarly, the requirement that the Supreme Court exercise part of its authority in a purely appellate capacity is met by having appeals from the states.<56> It would not be beyond Congress's power to give the state courts original jurisdiction over most issues that statutes currently make exclusively federal. The residue might not be trivial, but it would be relatively small.<57> So long as the state courts can hear constitutional claims, and the Supreme Court can hear the appeals, it is difficult to see the constitutional problem, even though it is easy to imagine the chaos that might result.

In short, if one adopts Professor Amar's "two-tiered" approach, Article III divides the judicial power into two categories: unrestrictable federal *1357 judicial powers and powers assigned to the federal courts subject to those exceptions that Congress might choose to make. The danger of state court usurpation that the Vesting thesis is supposed to prevent is greatly lessened, perhaps eliminated entirely. The distinction between the first tier and the second is founded in the Heads of Jurisdiction (not the Vesting Clause), which are the words that define and give content to the judicial power that is vested in the judicial branch.<58> The existence of the first tier guarantees the essential decision making of the federal courts, if only in an appellate capacity.

I have concentrated on Professor Amar's approach because Professor Calabresi and Mr. Rhodes rely on it, and because it fits more neatly with the views they reject.<59> It is worth noting, however, that in addition to the structural approach outlined earlier,<60> alternate textual approaches also demonstrate the chimerical nature of the threat of excessive state court power or the missing "vesting" of federal court jurisdiction. For example, focusing on the Exceptions Clause<61> might lead one to the conclusion that Congress has great, or even unlimited, power to reduce the jurisdiction of the federal courts. Professor Calabresi and Mr. Rhodes would probably say that this demonstrates the value of the Vesting thesis, for it protects the federal courts from such threats. The threat, however, is fairly remote. Congressional weakening of federal courts means that the state courts gain, not Congress itself, and the balance of power among the branches is only tilted, not destroyed.<62> The federal nature of the republic suggests that Congress is unlikely to act in this manner, and indeed it has rarely even threatened to do so. Thus, federalism works to support the balance of power among the branches.<63>

*1358 C. "The executive Power shall be vested . . ."

In addition to their argument by analogy to Article III, Professor Calabresi and Mr. Rhodes argue that the expansive view of presidential power flows naturally from either position in the jurisdiction-stripping debate.<64> Indeed, Professor Calabresi and Mr. Rhodes assert that theories of limited jurisdiction-stripping power (which I shall call mandatory jurisdiction) require that the President have the power to remove all executive branch officials, and that theories of broad jurisdiction-stripping power "strongly suggest" the same result.

1. Mandatory Jurisdiction.--Professor Calabresi and Mr. Rhodes argue that if one accepts Professor Amar's categories of mandatory federal jurisdiction, it follows that Congress may never impose limits on the President's power to remove executive branch officials. Professor Amar's categories mean that Congress may create exceptions to the Supreme Court's appellate jurisdiction in the "mandatory" areas, but only if the mandatory jurisdiction is vested in a lower federal court.<65> Further, although state courts may have jurisdiction over federal claims, Professor Amar argues that the Article III Vesting Clause requires that the final power of decision in a mandatory case rest with an Article III federal court. Any other conclusion, Professor Amar argues, "contravenes the Vesting Clause by vesting federal judicial power in a state court."<66> Professor Calabresi and Mr. Rhodes push the point further and draw an analogy between Congress's Damoclean threat to "undermine the independence of the federal courts" by reducing their jurisdiction and what they see as Congress's similar threat "to the President's independence" that would be caused by a broad power to insulate executive officers from removal. They claim that, just as the mandatory jurisdiction thesis blocks Congress's putative ability to undermine the independence of the federal courts,<67> so too should a comparable evil by which Congress restricts *1359 the President's independence (President stripping<68>) be blocked in Article II. Professor Calabresi and Mr. Rhodes are somewhat vague, however, as to the dangers to the President's "independence," as opposed to the President's power.

The thrust of the argument against Congress's power to constrain unenumerated presidential powers, such as the removal power, seems to be an argument for strong symmetry. Even so, it is not persuasive. Unless one believes that all categories of federal jurisdiction are mandatory, symmetry fails to compel the conclusion that all presidential powers, presumably including removal, are immune to dilution--or balance--by Congress. Even Professor Calabresi and Mr. Rhodes do not suggest that all the Heads of Jurisdiction in Article III are mandatory, much less that all the jurisdiction vested in the federal courts (which they believe exceeds the Heads of Jurisdiction) is mandatory.<69> If some categories of enumerated federal jurisdiction are optional, strong symmetry certainly does not forbid a distinction between presidential powers enumerated in Article II<70> and unenumerated powers such as the removal power.<71> Both the nonmandatory Article III powers and the nonenumerated Article II powers are therefore potentially subject to some form of congressional check or diminution. Yet, Professor Calabresi and Mr. Rhodes *1360 shy away from claiming that all categories of federal jurisdiction in Article III are mandatory, despite the indispensable role of that claim in their argument.<72> To repeat, if all categories of Article III jurisdiction are not mandatory, then the suggestion that some are mandatory is completely compatible with the notion that unenumerated presidential powers such as the removal power can be at least partly restricted by Congress.

2. Nonmandatory Jurisdiction.--Professor Calabresi and Mr. Rhodes assert that theories of broad congressional power to restrict jurisdiction of the federal courts imply limited congressional power to restructure the Executive. The suggestion that the Vesting Clauses of Articles II and III are significant grants of power<73> is only a floor for a more sweeping argument in which Professor Calabresi and Mr. Rhodes suggest that the differences between Article II and Article III demonstrate that the power vested in Article II is actually broader than the power vested by the identical clause in Article III because Article II lacks any reference to congressional powers analogous to Congress's power to make exceptions to the federal courts' jurisdiction. The absence of similar congressional powers in Article II, we are told, makes Article II's Vesting Clause "look sweeping indeed."<74>

Again, this is completely nonstructural and unpersuasive textualism. Structurally, once one accepts that the Heads of Jurisdiction in Article III are in fact substantive allocations of power to the judiciary (at least until the other branches, or a supermajority of Congress, reduce it),<75> then the absence of a clause in Article II corresponding to the Exceptions Clause of Article III is less surprising than Professor Calabresi and Mr. Rhodes make it appear because the President's enumerated powers, particularly with regard to domestic affairs, are rather limited.<76> And the powers that are enumerated, such as the veto and the pardon power cannot, by design, be reduced by congressional action or even delegated by the President. Structural considerations support this view, for were these powers to be reduced or shifted there is no state-level actor or body that could take them up. Second, there is an important difference between the business of the executive and that of the federal courts. Both branches derive a significant portion of their duties and functions from federal legislation. The executive receives funds and directions as to how to spend them, and also rule-making and regulatory authority along with standards *1361 to guide its exercise of discretion. A large part of the federal judiciary's business also arises under federal statutes, but unlike the President (whose powers and duties in the absence of statutes are primarily concerned with the military and foreign affairs), the Supreme Court has significant duties (whether these are seen as mandatory or default rules) arising from the Constitution itself. Were it not for legislation creating and funding departments, subordinates, and programs, the President would have relatively little to do, particularly in the domestic civilian sphere.<77> There simply were no executive or presidential powers granted in Article II which required a counterweight in that article of the type placed in Article III.<78> In short, in the domestic arena at least, the Presidency has so few inherent powers that there is almost nothing to strip. The modern Presidency's vast powers are creations of statute.

3. The Hypothetical Horrible.--Professor Calabresi and Mr. Rhodes seek to buttress their expansive reading of the President's powers under Article II by suggesting that if narrower readings are accepted, consistency across articles would force unacceptable readings of Article III. Indeed, Professor Calabresi and Mr. Rhodes assert that the Supreme Court has already erred by deciding separation of powers cases regarding the President's Article II powers which, if taken to their logical conclusion, would require the Supreme Court to recognize excessive congressional power to restrict Article III jurisdiction and would ultimately undermine the independence of the federal courts.<79> The chief hypothetical evil--that the Court may have to recognize a congressional jurisdiction-stripping power--is a menace somewhat dulled by familiarity. Correctly or not, Congress has had at least some reason to believe that it has a full-blooded jurisdiction-stripping power, but it has made very sparing use of it. Jurisdiction stripping may take power from the federal courts, but it gives it to state courts, not Congress, a fact which is no doubt partly responsible for Congress's continuing lack of interest in exploring the outer limits of that power.<80>

*1362 Furthermore, it is not at all obvious that Congress can prevent the federal courts from hearing cases that implicate basic constitutional rights, particularly if no adequate state forum exists to hear the claim. Even if Congress did not provide the jurisdiction by statute, the authority to hear basic constitutional cases, particularly habeas corpus petitions, might be derived from the Constitution itself, if only on the theory that to read the Constitution to create rights without remedies would amount to nullifying it. There is also reason to doubt the constitutionality of a congressional attempt to exclude fundamental constitutional cases from being heard in any federal court, including the Supreme Court.<81>

D. Reading and Using the Vesting Clauses

Professor Calabresi and Mr. Rhodes assert that the Vesting Clauses of Articles II and III are robust if "somewhat nebulous" grants of power to the executive and judiciary, while Article I's Vesting Clause is not because it refers only to powers "herein granted."<82> Readers who reject the Vesting thesis are said to be guilty of ignoring the Constitution's text and structure, and in particular of misunderstanding Article II,<83> or of writing the Vesting Clause of Article II out of the Constitution. As a matter of constitutional construction, The Structural Constitution poses a simple challenge: If the Vesting Clauses of Articles II and III are not in and of themselves significant grants of power, what do they do?<84>

True structural analysis would have been far more persuasive. Professor Calabresi and Mr. Rhodes are right to point out that Articles II and III are very similar in structure and differ somewhat from Article I and that this similarity and difference is also present in the wording of their respective Vesting Clauses.<85> The significance of the Vesting Clauses can best be appreciated by contrasting the federal constitution *1363 with its predecessor, the Articles of Confederation. The Articles provided for only one branch of government--Congress. Such executive and judicial powers as that government possessed were exercised directly by Congress or through its agents, who were appointed by and in theory responsible to it.<86> As we all know, the federal constitution represented a sharp and intentional break with this past. The three Vesting Clauses reflect this. Article I speaks of the powers "herein granted," which makes clear that the federal Congress would have only the powers granted in the Constitution, not the powers held by its predecessor body. By contrast, the Vesting Clauses of Articles II and III create new bodies where there had been none before. To use Professor Calabresi and Mr. Rhodes's phrase, they are "performatives"<87>--they do the necessary work of firing the "empty vessels" that the remainder of those articles then fill. Article I is different because there was already a (con)federal body called "Congress" in existence, while in the case of Articles II and III the bodies had to be created.<88>

Professor Calabresi and Mr. Rhodes's treatment of the "judicial Power" is unpersuasive on its own terms. Having dismissed the Heads of Jurisdiction as substantive grants of power,<89> Professor Calabresi and Mr. Rhodes nonetheless turn to them to define the boundaries of the judicial power, in order to avoid having to argue in favor of judicial activism. The Heads of Jurisdiction are, they say, "an exclusive list" of the contents of the judicial power because the list is headed by a clause saying that the judicial power "shall extend to" those categories while this language is absent from Article II.<90> By reading "shall extend to" in Article III to mean effectively the same as "herein granted" in Article I, Professor Calabresi and Mr. Rhodes have implicitly undermined their argument that the structure of Article III shares a fundamental property with Article II that is lacking in Article I. To have admitted that powers granted by both Articles I and III are only those enumerated within those articles would not only have revealed that the authors seek to apply an expansive interpretation to Article II only, but would have fatally undercut the consistency between Articles II and III on which the authors *1364 put so much stress.<91>

While imperfect, the structural view is more persuasive than the Vesting thesis.<92> The chief problem with reading the Vesting Clauses as grants of "executive" and "judicial" power is that we have no clear idea what those words mean in the context of the Constitution other than from the text of the articles that follow. Professor Calabresi and Mr. Rhodes's response to my argument that if "it were possible to give content to the President's power without recourse to the text of the constitutional article that purports to define that office, then presumably the first three articles of the Constitution could have been expressed in a sentence"<93> is highly unsatisfactory. They argue that the clauses which follow the first sentence in Article II "are not offered to us as an exclusive list,"<94> but give insufficient explanation as to why this might be or what it means for a government of limited powers.<95> Of course, if one accepts *1365 the idea that Article II's Vesting Clause is followed by an implicit " e.g.," it follows naturally that one need not be overly concerned with finding enumerated justifications for the President's assertions of power, so long as one can find a justification for labeling them "executive."<96> Where one looks for such justifications is barely addressed in The Structural Constitution, although surely a good test of a theory of a branch's powers in a system of balanced and limited government is the adequacy of the theory's limiting principles.<97>

The authors' claim that the Heads of Jurisdiction in Article III are preceded by an implicit " i.e.," while the enumerated presidential powers in Article II are prefaced by an implicit " e.g.," has the virtue of allowing the authors' conclusion. It has the defects of making textual analysis of those clauses of the Constitution irrelevant, indeed of reducing several clauses of the Constitution to surplusage,<98> and of undermining the concept *1366 of a government of limited powers. These defects outweigh any advantages.

Given the great stress that Professor Calabresi and Mr. Rhodes place on the parallel between Articles II and III, one might have expected an argument that the judicial power should be treated as expansively as the executive power.<99> And since some concept of tradition seems to lie at the root of the Professor Calabresi and Mr. Rhodes view of the meaning of executive power, one might expect that the same would apply to their exegesis of the judicial power. This line of reasoning would likely push one toward the conclusion that the federal courts must have the power to decide cases under common law, since that power was a feature of all law courts that the Framers could have had before them as examples, thus informing their understanding of the judicial power. It follows that Swift v. Tyson<100> was correctly decided and that Erie R.R. v. Tompkins<101> and United States v. Hudson & Goodwin<102> were wrong. Strangely, Professor Calabresi and Mr. Rhodes do not follow their argument through to this conclusion. Indeed, Professor Calabresi and Mr. Rhodes's suggestion that the judiciary is to have "all" the federal judicial power could, if taken to its logical extreme, be read to threaten concurrent state court jurisdiction over claims based on federal law, although it is unlikely that the authors contemplated or would advocate a "dormant" jurisdiction clause.

III. Structure and Politics in Constitutional Law

There is general agreement that the Supreme Court's separation of powers decisions are hopelessly contradictory:<103> Myers<104> and Springer *1367 v. Government of Philippine Islands<105> give way to Humphrey's Executor<106> and Weiner v. United States;<107> INS v. Chadha,<108> Buckley v. Valeo,<109> Bowsher v. United States,<110> and Metropolitan Wash. Airports Auth. v. Citizens for the Abatement of Aircraft Noise, Inc.<111> sit uneasily beside Morrison v. Olson,<112> Commodity Futures Trading Comm'n v. Schor,<113> and Mistretta v. United States.<114> Lurking in the background are The Steel Seizure Case<115> and Nixon v. Administrator of GSA.<116> If one can bring oneself to ignore the doctrinal temptations posed by the irreconcilable rationales of these cases, there is a general rule which explains all the results and even offers some hope of predicting outcomes in future cases: just as Congress may increase the other two branches' powers by statutes conferring jurisdiction or rule-making/action- taking power, so too may Congress constrain the other two branches' exercise of their powers so long as Congress strictly obeys two fundamental constraints. First, Congress must obey the strictures set out in Buckley and Chadha, that neither it nor other persons in the legislative branch (broadly defined<117>) may constitutionally take any action that affects the *1368 rights, duties, or responsibilities of any person outside that branch, except via the procedures of bicameralism and presentment set out in Article I, or that falls under a small number of carefully limited exceptions.<118> Second, all three branches' powers can be divided into the two classes of enumerated and unenumerated powers. Congress's power to limit the President's enumerated powers is relatively slight. It exists only when Congress can assert a directly relevant enumerated power of its own.<119> But, Congress's power to check other, unenumerated, presidential powers is somewhat greater.

Overall, the Court's decisions fit a pattern in which Congress's power to check the other branches by determining their structure is very great, but Congress is checked by the requirements that it act through persons outside the legislature (which usually means persons in the executive or the judiciary<120>) and that Congress not aggrandize its own powers. Thus, in Myers, Buckley, Chadha, Bowsher, and Metropolitan Airports, separation of powers was violated by Congress seeking to reserve an executive power for itself.<121> Humphrey's Executor, Wiener, Morrison, Schor, and Mistretta all concerned cases in which Congress had lessened the President's power (or increased the judiciary's) without reserving a corresponding power for itself.<122> Indeed, when the issue is an unenumerated presidential power, such as the power to remove executive branch officials, the Supreme Court has, since at least 1838, consistently focused on whether Congress has impermissibly aggrandized itself, not on whether the President's "nebulous" executive power is being *1369 undermined.<123> When Congress removes powers from the President by defunding or canceling departments, the result is usually to give increased power to states or to decrease the President's powers while not increasing those of Congress or the judiciary, something which the Court's decisions permit.<124>

Professor Calabresi and Mr. Rhodes's major structural argument for their expansive reading of unenumerated presidential power is that the judicial branch, or "judiciary," is plural and the President unitary.<125> As a purely formal matter, this fails to compare like with like: in the context of the Constitution, the proper structural analogue to "judicial" is "executive" not "President." The proper analogue of "President" is "Supreme Court."<126> In any case, the claim that the vast executive bureaucracy headed by a single president is more fundamentally "unitary" than a small judiciary headed by a single "supreme" albeit multimember court is far from being the self- evident proposition Professor Calabresi and Mr. Rhodes make it out to be,<127> but it is buttressed by a clear decision at the Constitutional Convention to reject a Council of State. We know at least that the Framers did not want the executive headed by either a king or a committee, just as we know they wanted neither a parliament along English lines nor a congress along the lines of the Articles of Confederation.<128> *1370 The decision to stop at this perfectly reasonable starting point explains why Professor Calabresi and Mr. Rhodes find the recognition of heads of departments in the Appointment Clause to be "mysterious " and assert--without argument other than the Vesting thesis--that inferior officers cannot exercise executive power without the President's permission.<129> It also explains Professor Calabresi and Mr. Rhodes's rejection of my analogy between Congress's power to structure the lower federal courts and Congress's power to structure the executive.<130> What it does not explain is why, in what is otherwise a highly textualist analysis, Professor Calabresi and Mr. Rhodes feel free to denigrate parts of the text they find troubling.

In fact, however, the judiciary is not as plural as Professor Calabresi and Mr. Rhodes suggest:<131> it is highly debatable whether Congress is or was under any obligation to create lower federal courts any more than it is under an obligation to create executive agencies or staff them with heads of departments.<132> Lower courts and executive departments are *1371 practical, not constitutional, necessities.<133> Article II enumerates certain powers as belonging to the President. Other executive powers lack this enumeration. Congress cannot constitutionally diminish the President's enumerated powers--including, it should be noted, the duty to "take Care that the Laws be faithfully executed"--unless Congress can point to a relevant enumerated power of its own.<134> Other executive powers, however, are less well insulated from Congress's power to structure the executive branch. In those cases, where the President's authority is only implied from the enumerated powers, Congress may have countervailing enumerated or implied powers. If so, as Justice Jackson suggested in his Youngstown concurrence<135> and has recently been reaffirmed by Chief Justice Rehnquist<136> and by Justices Kennedy and O'Connor,<137> the field does not inevitably belong to the President.

Professor Calabresi and Mr. Rhodes also suggest that the Jackson-Rehnquist- Kennedy-O'Connor view of the President's powers amounts to rewriting the Article II Vesting Clause to read "the executive power shall be vested in a President, and in such inferior Entities as the Congress may from time to time ordain and establish," a criticism they aim primarily *1372 at my student Note.<138> Of all their many criticisms of the Note, this one comes the closest to being well-founded, but it misses the mark because it is predicated on the Vesting thesis, an unpersuasive view of the nature and purpose of the Vesting Clause.<139> In addition, this criticism ignores the distinction between enumerated and unenumerated powers.<140> The charge would have more merit if the Note argued that Congress could assign the commander-in-chief power to the Chairman of the Joint Chiefs of Staff; the Note of course suggested no such thing.

But I do assert--and thus the force in Professor Calabresi and Mr. Rhodes's exaggeration of my view--that both structural considerations and the frequency and manner of the references to inferior officers and heads of departments in the Constitution mean that the holders of those offices are more than constitutional irrelevances.<141> The sheer frequency of references to executive branch officers is telling enough, but I would place greater reliance on structural conditions. For example, Congress has the power to impeach all civil officers, suggesting that they are persons of a stature which makes it reasonable for Congress (which creates their posts) to have some say in the conditions of their tenure. Together, these structural and textual factors lead to the conclusion that the civilian side of the executive branch is less monolithically hierarchical than the armed services, and that Congress can legitimately give some inferior officers and heads of departments at least limited protection from dismissal and countermand by the President.<142>

*1373 In the absence of a very robust theory of what tradition is relevant, the expansive vision of executive power implied by the Vesting thesis gives little if any guidance as to where one finds limits on presidential power.<143> By contrast, the view set out above provides better, if still imprecise, guidelines. Focusing on the political branches' enumerated powers provides a principled way to begin to define precisely which powers Congress can assign to executive agents shielded from the presidential removal at will.<144> For enumeration to be an effective guide, however, one must determine the scope of the least specific, and thus potentially most far-reaching, enumerated presidential power that the President "shall take Care that the Laws be faithfully executed."<145> Many thoughtful commentators have argued that the Take Care Clause (rather than the Vesting Clause) requires that the President have complete control over the executive branch.<146> I previously offered textual and structural arguments suggesting that the Take Care Clause, while not meaningless, means something less in the civilian sphere than the power the Commander in Chief has over the military.<147> Although Professor Calabresi and Mr. Rhodes apparently do not accept the distinction, it does not amount to suggesting that Congress can necessarily insulate every executive officer completely from the President.<148> Instead, it suggests *1374 that for some executive officers, Congress can institute a requirement that the President have "cause" before exercising the removal power. It further accepts that a reasonable reading of the Take Care Clause may require that the President retain the power to dismiss for "cause" over all persons in the executive branch<149> and further recognizes that precisely what constitutes "cause" in the constitutional sense remains incompletely defined.<150> Admittedly, enumeration alone does not suffice to identify precisely which types of jobs in the executive branch are presidential (and thus serve at the President's pleasure) and which are only executive (and thus can, if Congress chooses, be given protection from dismissal without cause).<151> Rather than seek to sacrifice almost every separation of powers decision before and since on the altar of Myers, we would all be better served if the academic debate took up the thorny problem of defining sufficient "cause" to trigger the bedrock removal power and to whom it applies.<152>

Reading the Constitution structurally means more than reading, whenever reasonably possible, like words and clauses in different articles alike. While this is desirable, it is only a beginning. Ambiguities and uncertainties remain, and these must be resolved in concrete cases. A structural perspective offers to resolve them by reference to the structure of the government the Constitution creates instead of sentence structure. If the method is applied honestly, it provides a limited but significant range of normative considerations that can be invoked to inform constitutional interpretation.

One semistructural value that is appropriately, and frequently, deployed to help interpret the Appointment Clause is the idea of "balance" between the branches. Balance is only semistructural because it is not a constitutional end in itself but rather a means to the end of preventing tyranny.<153> Balance is a useful proxy for this underlying value. And if balance is what really concerns most parties to the Appointments Clause debate (and I think it is), then a structural interpretation at least *1375 has the value of concentrating minds on the real matter at issue. Unfortunately, "balance" has proved to be of limited value in resolving separation of powers controversies persuasively because of a lack of consensus as to where the balance should be, what constitutes a balance, and what "weight" should be accorded various powers.<154> The authors of The Structural Constitution appear to believe that recent Supreme Court decisions create an imbalance of power in favor of Congress. Others, myself included, continue to see an imperial Presidency. In the recent past, while the President has tended to be a Republican and Congress has tended to have Democratic majorities, more liberal judges and commentators have appeared more likely to side with the view that the balance had tilted against Congress, while more conservative judges and commentators have tended to argue that the President was being unconstitutionally constrained.<155>

Professor Calabresi and Mr. Rhodes would have been on stronger structural grounds if they had sought to root their argument in the structural value of the democratic process and of democratic accountability or in the modern version of it that has come to call itself republicanism. Such an argument would, however, require an account of the degree to which the original Constitution was supposed to be democratic--or antidemocratic--and the degree to which subsequent amendments, whether textual or otherwise, have imbued the document with a more democratic ethos. We could then talk directly about some things that really matter, for which arguments about constitutional text seem so often to be only proxies: How much democracy, and of what type, does the Constitution require? If we conclude that the Constitution today embodies a pluralist conception of politics, is that vision better served by allowing the President full administrative authority or by allowing Congress to constrain it?<156> Which promotes true democracy, Congress, the President, or both (or neither)? To what extent should we allow our constitutional views to be shaped by current (or 1790s, or 1850s, or 1960s) social science? Should the President's removal authority depend on our views as to how *1376 bureaucrats behave? On Montesquieu's? To what extent does one's opinion about facts determine one's views about how to read the Constitution? What happens when facts change?

It will be interesting to see whether many positions change now that we have a Democratic administration.<157>


<*>Associate Professor, University of Miami School of Law. M.Phil. (Cantab) 1984, J.D. 1987, Yale Law School. Internet: As the unnamed, albeit not anonymous, author of the " Yale Law Journal Note" (In Defense of Administrative Agency Autonomy, 96 Yale L.J. 787 (1987)), which comes in for particularly harsh criticism in Steven G. Calabresi & Kevin H. Rhodes, The Structural Constitution: Unitary Executive, Plural Judiciary, 105 Harv. L. Rev. 1155 (1992), I am grateful to the editors of the Northwestern University Law Review for allowing me a few pages for this reply and to Akhil Reed Amar, Caroline Bradley, Erwin Chemerinsky, Sharon Keller, Bernard Oxman, Robert Pushaw, Jonathan Simon, Alan Swan, and Steve Winter for their help as I grappled with earlier drafts. As at least one of them disagrees strongly with my views, none should be blamed for my failures to listen to them. Larry Lessig and Cass Sunstein kindly shared their article in progress with me, Julie Owen corrected innumerable Bluebooking errors, and I received superb library support from Nora de la Garza and Sue Ann Campbell.

<1>The exceptions are the President and Vice President. I use the term "executive branch" (as distinguished from the Presidency) to include all executive departments in the federal government, including so-called independent agencies and government-owned corporations.

<2>The absence of consensus is due to the opacity and paucity of the relevant constitutional text. This may be a blessing: Harold Bruff has argued that "the optimal level of specificity for constitutional rules that organize the government is low." Harold H. Bruff, On the Constitutional Status of the Administrative Agencies, 36 Am. U. L. Rev . 491, 493 (1987).

<3>See Mistretta v. United States, 488 U.S. 361, 424 (1989) (Scalia, J., dissenting); Morrison v. Olson, 487 U.S. 654, 705-27 (1988) (Scalia, J., dissenting).

<4>Steven G. Calabresi & Kevin H. Rhodes, The Structural Constitution: Unitary Executive, Plural Judiciary, 105 Harv. L. Rev. 1155 (1992).

<5>But see Comment, Abolition of Federal Offices as an Infringement on the President's Power to Remove Federal Executive Officers: A Reassessment of Constitutional Doctrines, 42 Fordham L. Rev. 562 (1974) (arguing that Congress may not abolish agencies under some circumstances).

<6>See INS v. Chadha, 462 U.S. 919 (1983) (holding legislative veto unconstitutional); Metropolitan Wash. Airports Auth. v. Citizens for the Abatement of Aircraft Noise, Inc., 501 U.S. 252 (1991); Buckley v. Valeo, 424 U.S. 1 (1976). The major constitutional provision avoiding this rule is the Senate's power of advice and consent over appointments and treaties. U.S. Const. art. II, s 2, cl. 2. Legislative hearings, investigations, and subpoenas are other important exceptions to the general rule.

<7>A ministerial task is one "where the duty ... is so plainly prescribed as to be free from doubt"; a nonministerial task is one which involves "the character of judgment or discretion." Wilbur v. United States, 281 U.S. 206, 218-19 (1930). It has long been held that officers who perform purely ministerial tasks can be given tenure in office. See, e.g., Kendall v. United States ex rel. Stokes, 37 U.S. (12 Pet.) 524 (1838) (Postmaster not subject to presidential direction when executing purely ministerial task); United States v. Perkins, 116 U.S. 483 (1886) (Congress may regulate removal of inferior officers, whose appointment it vests in the heads of departments). Limits to the President's power to fire ministerial officials are less controversial than are limits to the power to fire nonministerial officials because the President's power to set policy is, by definition, unaffected.

<8>For a fuller description of the unitarian view, see Calabresi & Rhodes, supra note 4, at 1165-71.

<9>272 U.S. 52 (1926).

<10>See, e.g., Defendant's Memorandum in Support of Cross Motion for Summary Judgment and Response by the United States to Motions by the Various Other Parties at 26ff., Synar v. United States, 626 F. Supp. 1374 (D.D.C. 1986) (No. 85-3495) (per curiam) (three-judge court), aff'd sub nom. Bowsher v. Synar, 478 U.S. 714 (1986).

<11>By "trinitarian" I do not mean to suggest that proponents believe there are three sources of power within the executive branch, or that there is more than one executive branch. Rather, the name is designed to suggest a more holistic approach to separation of powers problems that focuses on the existence of three equal branches rather than treating each of the three branches in isolation.

<12>295 U.S. 602 (1935) (upholding statute preventing President from firing FTC Commissioner without cause against constitutional challenge).

<13>Among the cases that have provided the most comfort for the unitarian view are Public Citizen v. United States Dep't of Justice, 491 U.S. 440, 488-89 (1989) (Kennedy, J., concurring in judgment) (congressional interference "with the President's exclusive responsibility" is per se constitutional violation); INS v. Chadha, 462 U.S. 919 (1983) (holding legislative veto unconstitutional); Northern Pipeline Constr. Co. v. Marathon Pipe Line Co., 458 U.S. 50 (1982) (plurality opinion) (holding that Bankruptcy Act impermissibly granted Article III powers to Article I judges); Buckley v. Valeo, 424 U.S. 1 (1976) (per curiam) (holding that Federal Elections Commission had executive functions and thus could not include members appointed by Congress).

<14>See supra note 13 (citing cases); see also Peter L. Strauss, The Place of Agencies in Government: Separation of Powers and the Fourth Branch, 84 Colum. L. Rev. 573 (1984).

<15>Academic commentators accuse the Court, with much justice, of zigzagging between formalism and functionalism. See, e.g., Stephen L. Carter, From Sick Chicken to Synar : The Evolution and Subsequent De-Evolution of the Separation of Powers, 1987 B.Y.U. L. Rev. 719; Stephen L. Carter, The Independent Counsel Mess, 102 Harv. L. Rev. 105 (1988); Erwin Chemerinsky, A Paradox Without a Principle: A Comment on the Burger Court's Jurisprudence in Separation of Powers Cases, 60 S. Cal. L. Rev. 1083 (1987); E. Donald Elliott, Why Our Separation of Powers Jurisprudence Is So Abysmal, 57 Geo. Wash. L. Rev. 506 (1989); Strauss, supra note 14; Peter L. Strauss, Was There A Baby In the Bathwater? A Comment on the Supreme Court's Legislative Veto Decision, 1983 Duke L.J. 789.

<16>See, e.g., Morrison v. Olson, 487 U.S. 654 (1988).

<17>See, e.g., Peter P. Swire, Note, Incorporation of Independent Agencies into the Executive Branch, 94 Yale L.J. 1766 (1985) (arguing that Humphrey's Executor should be reversed). The Reagan Administration aggressively advocated a unitarian approach to separation of powers. See, e.g., Brief for United States at 44-51, Bowsher v. Synar, 478 U.S. 714 (1986) (No. 85-1377).

<18>Calabresi & Rhodes, supra note 4.

<19>The unitarians have also relied on a historical account that is open to challenge. See Lawrence Lessig & Cass R. Sunstein, The President and the Administration, 94 Colum. L. Rev. 1, 12-84 (1994); A. Michael Froomkin, Note, In Defense of Administrative Agency Autonomy, 96 Yale L.J. 787, 804- 08 (1987).

<20>"The question is not whether the text shall be respected, but rather how one goes about respecting a text of that high generality and consequent ambiguity which marks so many crucial constitutional texts." Charles L. Black, Jr., Structure and Relationship in Constitutional Law 30 (Ox Bow Press 1985) (1969). By focusing on structural interpretation I do not mean to denigrate textualism, but only to distinguish between a structural view and the type of analysis in The Structural Constitution.

<21>U.S. Const. art. I, s 1.

<22>U.S. Const. art. II, s 1, cl. 1.

<23>U.S. Const. art. III, s 1.

<24>See supra note 4, at 1175-79.

<25>Akhil Reed Amar, A Neo-Federalist View of Article III: Separating the Two Tiers of Federal Jurisdiction, 65 B.U. L. Rev. 205 (1985) hereinafter Amar, Neo-Federalist View; Akhil Reed Amar, The Two-Tiered Structure of the Judiciary Act of 1789, 138 U. Pa. L. Rev. 1499 (1990) hereinafter Amar, Structure; Akhil Reed Amar, The Bill of Rights as a Constitution, 100 Yale L.J. 1131 (1991); Akhil Reed Amar, Reports of My Death Are Greatly Exaggerated: A Reply, 138 U. Pa. L. Rev. 1651 (1990); Akhil Reed Amar, Marbury, Section 13, and the Original Jurisdiction of the Supreme Court, 56 U. Chi. L. Rev. 443 (1989) hereinafter Amar, Original Jurisdiction ; Akhil Reed Amar, Taking Article III Seriously: A Reply to Professor Friedman, 85 Nw. U. L. Rev. 442 (1991). For alternate readings of Article III, which reject Justice Story's account in Martin v. Hunter's Lessee, see, e.g., Paul M. Bator, Congressional Power Over the Jurisdiction of the Federal Courts, 27 Vill. L. Rev. 1030 (1982); Gerald Gunther, Congressional Power to Curtail Jurisdiction: An Opinionated Guide to the Ongoing Debate, 36 Stan. L. Rev. 895 (1984).

<26>See Amar, Neo-Federalist View, supra note 25. In this Article I adopt the convention of speaking of the President's "removal" authority over an official as shorthand for the President's ability to direct that official's use of her discretion--despite some reason to doubt the strength of this connection as an empirical matter. See Michael Herz, United States v. United States : When Can the Federal Government Sue Itself ?, 32 Wm. & Mary L. Rev. 893, 952-53 (1991) (discussing political, legal, and practical impediments to President's ability to control subordinates' discretion).

<27>Calabresi & Rhodes, supra note 4, at 1176. Article III, s 1 is quoted supra text accompanying note 23. Contrast Amar, Neo-Federalist View, supra note 25, at 232 (recognizing coordinate nature of all branches).

<28>See Calabresi & Rhodes, supra note 4, at 1187.

<29>Such an account is far from impossible. See infra text accompanying note 74.

<30>Compare Black, supra note 20, at 30.

<31>Textualist arguments based on the location of clauses and phrases can be overdone. If the Constitution created the judiciary in exactly the same words but located the full text of Article III in Article I or II, the structural argument would remain equally valid.

<32>In what follows, I give disproportionate attention to the textualist arguments in order to take on Professor Calabresi and Mr. Rhodes on their own grounds. I believe structural arguments fully demonstrate the error of the Vesting thesis.

<33>U.S. Const. art. III, s 2, cls. 1-2:

The judicial Power shall extend to all Cases, in Law and Equity, arising under this Constitution, the Laws of the United States, and Treaties made, or which shall be made, under their Authority;--to all Cases affecting Ambassadors, other public Ministers and Consuls;--to all Cases of admiralty and maritime Jurisdiction;--to Controversies to which the United States shall be a Party;-- to Controversies between two or more States;--between a State and Citizens of another State;--between Citizens of different States;--between Citizens of the same State claiming Lands under Grants of different States, and between a State, or the Citizens thereof, and foreign States, Citizens or Subjects.

In all Cases affecting Ambassadors, other public Ministers and Consuls, and those in which a State shall be a Party, the supreme Court shall have original Jurisdiction. In all the other Cases before mentioned, the supreme Court shall have appellate Jurisdiction, both as to Law and Fact, with such Exceptions, and under such Regulations as the Congress shall make.

<34>8 The Oxford English Dictionary 320 (J.A. Simpson & E.S.C. Weiner eds., 2d ed. 1989) includes in its first definition of jurisdiction, "power of declaring and administering law or justice; legal authority or power" and supports this definition with examples from the 13th through 19th centuries. For evidence that modern American usage is the same, see Black's Law Dictionary 853 (6th ed. 1990).

<35>The Eleventh Amendment, which withdraws jurisdiction with the words "The Judicial power ... shall not ... extend to" suggests that the words "shall extend to" were understood to be a grant of jurisdiction. I am indebted to Erwin Chemerinsky for this observation.

<36>Calabresi & Rhodes, supra note 4, at 1177 (footnotes omitted).

<37>Id. at 1177 n.118.

<38>The cases cited by Professor Calabresi and Mr. Rhodes are Texas v. New Jersey, 379 U.S. 674, 675 (1965) (exercising, without discussion, "original jurisdiction under Art. III, s 2, of the Constitution" to decide a controversy over escheat of intangibles between states) and Kansas v. Colorado, 206 U.S. 46 (1907) (holding Supreme Court had original jurisdiction to hear suit between two states). Other cases might be invoked as well, such as Cohens v. Virginia, 19 U.S. (6 Wheat.) 264, 404-06 (1821), in which Chief Justice Marshall relied on the Heads of Jurisdiction in order to determine whether the Supreme Court could exercise jurisdiction over decisions of state courts in criminal cases.

<39>Kansas v. Colorado, 206 U.S. at 81-82 (Brewer, J.).

<40>Id. at 82.

<41>To be fair, the passage also undercuts the suggestion advanced below, see infra text accompanying note 87, that the Vesting Clause in Article III is only a performative and not a grant of jurisdiction.

<42>See Texas v. New Jersey, 379 U.S. at 675.

<43>Calabresi & Rhodes, supra note 4, at 1177 n.118.

<44>See Martin H. Redish, Federal Jurisdiction: Tensions in the Allocation of Judicial Power 122, 146-47 (2d ed. 1990). Actually, Professor Redish believes that all federal common law is illegitimate. See Martin H. Redish, Federal Common Law, Political Legitimacy, and the Interpretive Process, 83 Nw. U. L. Rev. 853 (1989).

<45>398 U.S. 375 (1970) (asserting admiralty jurisdiction over wrongful death actions).

<46>See, e.g., U.S. Const. , art. I, s 8, cl. 10 (Congress shall have power to "define and punish Piracies and Felonies committed on the high Seas, and Offenses against the Law of Nations"); id. at cl. 11 (power to "declare War, grant Letters of Marque and Reprisal, and make Rules concerning Captures on Land and Water"); id. at cl. 13 (power to "provide and maintain a Navy").

<47>See, e.g., Calabresi & Rhodes, supra note 4, at 1186-87.

<48>U.S. Const. art. III, s 2, cl. 1.

<49>See Calabresi & Rhodes, supra note 4, at 1209 (emphasis omitted).

<50>Although I do not share this concern, it deserves to be taken seriously. An originalist might resolve it by turning to The Federalist, in which Hamilton reads the Constitution to empower the federal courts, while the state courts retained concurrent jurisdiction. The Federalist Nos. 81, 82 (Alexander Hamilton).

<51>See supra note 33 (quoting U.S. Const. art. II, s 2, cl. 2).

<52>Amar initially appeared to lean towards the Vesting thesis. Compare Amar, Neo-Federalist View, supra note 25, at 229, 231-32, 239 with subsequent articles cited supra note 25.

<53>See Amar, Neo-Federalist View, supra note 25, at 233; Amar, Structure, supra note 25, at 1504 n.10. Amar concludes that where the judicial power extends to "all Cases"--federal question, ambassador, and admiralty--a federal court must have the ultimate say either directly or by appeal. Amar, Original Jurisdiction , supra note 25, at 445. By contrast, for the Vesting thesis to work, one would have to argue that the lower courts are constitutionally required.

<54>Indeed, I am uncertain whether Congress has any obligation even today to give the lower courts any work to do, although it clearly has an obligation to pay federal judges whether or not they have a docket. See U.S. Const. art. III, s 1. Whether federal courts are constitutionally required remains controversial. Compare Glidden Co. v. Zdanok, 370 U.S. 530, 551 (1962) (Harlan, J.) (no requirement) with Theodore Eisenberg, Congressional Authority To Restrict Lower Federal Court Jurisdiction, 83 Yale L.J. 498 (1974) (arguing inferior federal courts are now required by Constitution).

<55>A textualist would resolve this by examining the Heads of Jurisdiction, not the Vesting clause of Article III.

<56>Were Congress so provoked by the actions of the inferior federal courts that it eliminated their jurisdiction, much of their caseload would fall to the state courts automatically, as these have concurrent jurisdiction over many of the issues that form the federal docket.

<57>There is a class of cases with constitutional implications that may need to be heard in federal court, such as a habeas corpus petition filed by a U.S. citizen detained abroad by U.S. forces. Either an inferior federal court must exist to hear such claims or, if no other adequate court exists to hear them, the petitions must be heard by the Supreme Court itself. The latter would be difficult to square with Marbury v. Madison, 5 U.S. (1 Cranch) 137 (1803) (holding, inter alia, that Congress may not increase Supreme Court's original jurisdiction beyond three categories specified in Article III, s 2, cl. 2). Before anyone cries heresy, note that Marbury's suggestion that Congress may not give the lower courts jurisdiction over cases falling within the original jurisdiction of the Supreme Court has already been rejected. See, e.g., Illinois v. City of Milwaukee, 406 U.S. 91 (1972). Cf. Steven G. Calabresi & Gary Lawson, Equity and Hierarchy: Reflections on the Harris Execution, 102 Yale L.J. 255, 274 (1992) (suggesting that all judicial power described in U.S. Const. art. III, s 2 must be vested in Supreme Court whether or not inferior courts are created).

Interestingly, Marbury itself concerned the appointment of a Justice of the Peace for the District of Columbia who served for five years during which he was "not removable at the will of the executive." 5 U.S. (1 Cranch) at 172. If Chief Justice Marshall had adhered to the Vesting thesis, Marbury would not even be a footnote to history, since the Court would have dismissed the case as moot, stating that whether or not Marbury received his commission was irrelevant, as the President had the vested right to remove him at once.

<58>The "two-tiered" view of Article III thus dovetails well with my argument below that there are two tiers of executive power (enumerated and unenumerated). See infra p. 1359. By contrast, the Vesting thesis recognizes only one tier for Article III and seeks to apply this vision, by analogy, to Article II in order to conclude that (some? all?) unenumerated powers are beyond Congress's power to curtail.

<59>Throughout this Article, I give more space to textualist arguments than I feel they deserve because I seek to meet Professor Calabresi and Mr. Rhodes on their own ground. The text is important, but there is more than one way to respect the text. See supra note 20.

<60>See supra text accompanying note 30.

<61>U.S. Const. art. III, s 2, cl. 2.

<62>The Framers presumably understood this. In addition, they had a political perspective that eliminated the importance of the issue. The Framers expected that Congress would create a national court system, complete with inferior federal courts. In order to ensure that federal courts had the power that they needed, Congress need do only exactly what the first Congress actually did--create inferior federal courts and grant them significant jurisdiction. The Judiciary Act of 1789 ensured that the structural danger of triumphant state courts would remain an academic issue. There was and is a clear political solution to the problem of insufficient federal jurisdiction-- one that the Framers anticipated would be used when they gave Congress the power to create inferior federal courts to define the parameters of federal jurisdiction and that was in fact used.

<63>A full-throated federalist (in the 1980s sense of the term), or antifederalist (in the 1790s sense of the term) might add that if Congress so mistrusted the federal courts that it preferred to have critical national issues determined by the states, it probably had a good reason, and the Constitution should be read to allow this.

<64>Calabresi & Rhodes, supra note 4, at 1159. For an even-handed summary of the debate over Congress's power to structure the executive and the judiciary, see id. at 1159-61.

<65>As Professor Calabresi and Mr. Rhodes note, Professor Amar's construction differs from those of Justice Story and Henry M. Hart Jr. Id. at 1161-63.

<66>Id. at 1164. Those who believe that the states are not created by the Constitution, but are somehow prior to or independent of it, and that their courts have jurisdiction to decide federal claims under the state constitutions rather than because of the federal constitution, might have some difficulty with this argument. If the states are prior and independent, then the Constitution is not vesting anything in the state courts, but merely failing to take away what they already have, or sharing it around in a new and interesting way. Further, as Professor Calabresi and Mr. Rhodes note, the concept of a Supreme court and inferior courts can easily be understood to imply a more hierarchical relation between the Supreme Court and other federal courts than Professor Amar would require. More complex views of the polity, such as that rooted on national consent, see, e.g., Bruce Ackerman, We The People (1991), are more compatible with Professor Amar's view.

<67>Professor Calabresi and Mr. Rhodes argue that their interpretation of Article II must be correct, because an alternate interpretation would drive an interpretation of Article III that would overly weaken the federal courts. See Calabresi & Rhodes, supra note 4, at 1173-75, 1214-15. However, Professor Calabresi and Mr. Rhodes's assertion that it is "widely perceived ... and indeed may" be the case that jurisdiction stripping in favor of Article I courts weakens Article III courts ignores the fact that decisions of the Article I bodies are in all cases subject to judicial review, at least on questions of law. See id. at 1174; see also Amar, Structure, supra note 25, at 1547-59 (arguing that Article I courts are constitutional so long as there is Article III appellate review).

<68>See Calabresi & Rhodes, note 4, at 1192.

<69>Professor Calabresi and Mr. Rhodes do suggest that more limited theories of mandatory federal jurisdiction, such as those advanced by Henry M. Hart, Jr. and Lawrence G. Sager, also support their thesis. The claim is that if Hart and Sager are correct that Congress cannot divest the Supreme Court of its "essential functions," even in the face of all the language in Article III which suggests that Congress can control federal jurisdiction, then the same construction applied to Article II's Vesting Clause, which is not counterbalanced by similar procongressional language, means that "the Article II Vesting Clause must command a more than 'essential' role for the president. The only conceivable role that is more than 'essential' is the 'exclusive' role advocated by Justice Scalia." Calabresi & Rhodes, supra note 4, at 1205-06. This argument would be more persuasive if Article II and Article III allocated both branches equivalent power in the absence of congressional action. But absent congressional action, the President has few powers while the judiciary has many, destroying the underlying symmetry between the two branches' relationship with Congress and explaining why Congress has the power to make (at least some) "exceptions" to the jurisdiction of the federal courts. See infra text accompanying notes 75-78. Professor Amar suggests that Congress may make exceptions only to the Supreme Court's appellate jurisdiction. See, e.g., Amar, supra note 25, at 1654. If this is correct, strong symmetry suggests that there must be some executive powers that can be reduced by Congress.

<70>These are the power to receive ambassadors, the power to take care that the laws be faithfully executed, the power of commander in chief, the veto power, and the power to grant pardons. See U.S. Const. art. II. For an attempt to draw out the distinction between presidential and executive powers, see Froomkin, supra note 19.

<71>See Froomkin, supra note 19, at 803-04.

<72>Professor Calabresi and Mr. Rhodes do note that Robert N. Clinton, A Mandatory View of Federal Court Jurisdiction: A Guided Quest for the Original Understanding of Article III, 132 U. Pa. L. Rev. 741 (1984), views all Nine Heads of Jurisdiction in Article III as mandatory, but dismiss his position as "at odds with 200 years of practice." Calabresi & Rhodes, supra note 4, at 1164 n.45.

<73>See supra Part II.B

<74>Calabresi & Rhodes, supra note 4, at 1186; id. at 1192.

<75>See supra text accompanying note 30; infra text accompanying note 87.

<76>See Charles L. Black, Jr., The Working Balance of the American Political Departments, 1 Hastings Const. L.Q. 13 (1974).

<77>See Froomkin, supra note 19, at 803-04 (arguing by analogy that Congress's power to limit the President's control of the domestic administrative machinery must be at least as great as its power to control the President's Commander in Chief power).

<78>Which is not to say that no counterweights exist, since the legislative powers clearly granted to Congress in Article I include a number of specific powers relating to the military and the conduct of foreign affairs.

<79>See, e.g., Calabresi & Rhodes, supra note 4, at 1174-75, 1209, 1212-13.

<80>Whatever its scope, Congress's power to reduce the jurisdiction of Article III courts is a far less fearsome power than its power to create legislative courts or independent agencies, because the latter allow Congress potentially to bring the functions of other departments under its own domination. See id. at 1174 (Professor Calabresi and Mr. Rhodes agree with this proposition "as a formal matter"), 1192 ("the President-stripping power is far more extreme than the jurisdiction-stripping power--the former allows for a formal violation of the separation of powers, but the latter does not. Accordingly, President stripping requires a far more explicit textual basis than does jurisdiction stripping.") (footnote omitted).

Those who wish to augment the political balance of the Constitution with a Constitutional prohibition should find comfort in Professor Amar's two-tiered view of Congress's power to reduce the Supreme Court's appellate jurisdiction.

<81>See generally Amar, supra note 25. The case is strongest for habeas petitions in peacetime, since the Constitution states that the writ may be suspended only in cases of "Rebellion or Invasion." U.S. Const . art. I, s 9, cl. 2.

<82>Calabresi & Rhodes, supra note 4, at 1175-76, 1195, 1198.

<83>See, e.g., id. at 1199 (contending that construction of Article II Vesting Clause advocated here and in my Note is "utterly wrong").

<84>One could argue, as Erwin Chemerinsky suggested to me, that the three Vesting Clauses simply serve a naming function. They tell us what the three institutions created by the Constitution should be called. See Edward S. Corwin, The Steel Seizure Case: A Judicial Brick Without Straw, 53 Colum. L. Rev. 53 (1953). Names are important; naming things is neither a grant of power nor is it surplusage.

Surplusage is another possibility, but those who share Professor Calabresi and Mr. Rhodes's intuition that readings of the Constitution which produce surplusage are to be avoided when possible, will not find this an attractive alternative. There is enough effective surplusage in the Constitution already. Cf. Jacobson v. Massachusetts, 197 U.S. 11, 22 (1905) (noting Preamble to Constitution is not a grant of power to any department of the federal government).

<85>See Calabresi & Rhodes, supra note 4, at 1199.

<86>See Articles of Confederation art. IX (appointment of Commissioners as agents of Congress to try disputes between states and territorial disputes involving grants from multiple states; appointment of all officers) (1778).

<87>See Calabresi & Rhodes, supra note 4, at 1179 n.125 (noting, but not discussing, view of clauses as performatives).

<88>For a survey of the various ways in which modern Supreme Court decisions undermine the Vesting thesis, see Joseph P. Verdon, Note, The Vesting Clauses, The Nixon Test, and the Pharaoh's Dreams, 78 Va. L. Rev. 1253, 1258-67 (1992).

<89>See supra text accompanying note 34.

<90>Robert Pushaw suggested an additional problem with the restrictive view of the federal judicial power in commenting on an earlier draft of this article: the judicial power extends to "all Cases in Law and Equity." Given the expansive power of equity courts in the 18th century, this grant is broad indeed.

<91>In an attempt to explain why Article II deserves a more sweeping construction than Article III, Professor Calabresi and Mr. Rhodes find significance in the fact that Article III "vests" the "judicial power of the United States," while Article II, which "vests" the "executive power," lacks the "qualifier" of "of the United States." See Calabresi & Rhodes, supra note 4, at 1195 n.214, 1196. One could be forgiven for thinking that if the words "of the United States" are a "qualifier," then the absence of that "qualifier" in Article II means that because the President's grant of power is not "limited" by "of the United States," it extends to other powers not "of the United States" ( e.g. , that of state governors?).

<92>Two other readings of the Vesting Clauses deserve mention. To begin with Article I ("All legislative Powers herein granted"), the words "all" and "herein" can easily refer to the entire constitutional text, not just Article I. This reading explains and reflects the fact that congressional powers appear in the other articles of the Constitution. E.g., U.S. Const. art. II, s 1, cl. 3-4 (amended 1804) (congressional role in presidential election); id. at art. II, s 2, cl. 2 (Senate advice and consent power); id. at art. III, s 1 (congressional power to create inferior federal courts); id. at art. IV, s 3, cl. 1 (congressional power to admit new states); id. at art. IV, s 3, cl. 2 (congressional power of federal property); id. at art. V (congressional power to initiate constitutional amendments). Admittedly, this reading of "herein" may be subject to challenge because the other branches also have a few, but significant, enumerated powers which appear outside their respective articles. See U.S. Const. art. I, s 3, cl. 4 (Vice President to preside over Senate); id. at art. I, s 3, cl. 6 (Chief Justice to preside over impeachment trial of President); id. at art. I, s 7, cl. 2 (President to sign or veto bills); id. at art. II, s 2, cl. 2 (judiciary's appointment of inferior officers); id. at art. IV, s 4 (President's emergency power, when Congress in recess, to quell domestic violence). By contrast, Congress's non-Article I powers include significant powers that Professor Calabresi and Mr. Rhodes might concede are legislative. E.g., U.S. Const. art. III, s 1 (congressional power to create inferior federal courts); id. at art. IV, s 3, cl. 2 (congressional power over federal property).

Second, the presence of the word "all" in Article I's Vesting Clause may be enough by itself to make this a broader grant of power than the Vesting Clauses in Article II and III. Whether or not one adopts this view, it throws a new light on the Professor Calabresi and Mr. Rhodes's reading of the Vesting Clauses in Articles II and III, which the authors feel compelled to explicate as meaning "the President is to have all of the executive power and the Article III judiciary is to have all of the judicial power." Calabresi & Rhodes, supra note 4, at 1176. See also Morrison v. Olson, 487 U.S. 654, 705 (1988) (Scalia, J., dissenting) (focusing on "all").

<93>Froomkin, supra note 19, at 800.

<94>Calabresi & Rhodes, supra note 4, at 1196 (emphasis omitted).

<95>Professor Calabresi and Mr. Rhodes limit the scope of their article to "whether Congress may divest the President of powers that are concededly 'executive' " and say that "a meaningful discussion of the definition of the elusive term 'executive power' is beyond the scope" of their article. Calabresi & Rhodes, supra note 4, at 1165 n.52. Professor Calabresi and Mr. Rhodes claim to focus on whether Congress may divest the President of powers that are concededly "executive," such as the (unenumerated) prosecution power, but then note that even this is disputed. Id. While it may appear that I am criticizing Professor Calabresi and Mr. Rhodes for not writing a different article, I believe that the two issues are actually one: "to say that the first sentence of Article II vests executive power in the President begs the question of what those powers are." Froomkin, supra note 19, at 799-800.

<96>See Calabresi & Rhodes, supra note 4, at 1165 n.52, 1196.

<97>Professor Calabresi and Mr. Rhodes do invoke tradition, see, e.g., id. at 1196-97, but surely would not deny that there is frequently controversy about what our tradition says. In any case, a reference to tradition in this context begs the question of which tradition. That of the state governments? Which? When? Surely one tradition which is not relevant is the example of George III. Indeed, Professor Calabresi and Mr. Rhodes are clear that they are not arguing that the President's powers are equal to a king's. See id. at 1196- 97. Tradition may in any case turn out to be a weak reedfor the unitarian cause. See Henry P. Monaghan, The Protective Power of the Presidency, 93 Colum. L. Rev. 1, 12-20 (1993) (surveying 17th and 18th century English and American conceptions of executive power). In 1787, no state constitution gave the Executive illimitable removal powers. Myers v. United States, 272 U.S. 52, 247-48 (1926) (Brandeis, J., dissenting). Cf. Lessig & Sunstein, supra note 19, at 84 (disputing unitarian claims of historical support).

<98>Consider, for example, the pardon power. The specification of the pardon power in Article II makes it clear that pardons do not undo an impeachment, and to this extent the reference to pardons in Article II, s 2, cl. 1 is presumably not surplusage to Professor Calabresi and Mr. Rhodes. If, however, the pardon power were not mentioned in the Constitution, one can easily imagine the authors invoking a tradition of executive power to justify pardons, perhaps even in the case of pardons undoing impeachments.

In this connection, it is instructive to consider Ex Parte Grossman, 267 U.S. 87 (1925), which discusses the related problem of pardons before impeachments. Grossman was authored by Chief Justice Taft shortly before he wrote Myers v. United States, 272 U.S. 52 (1926), the high-water mark of judicial endorsement of expansive inherent presidential powers, and probably the most recent Supreme Court majority to speak approvingly of the Vesting thesis. Grossman goes to great pains to demonstrate that the textual enumeration of the President's pardon power is coterminous with that enjoyed by the King of England at the time the Constitution was adopted and thus extends to pardoning criminal contempt, and that at common law the King's pardon was no bar to impeachment. Had Taft accepted the Vesting thesis as set out in The Structural Constitution, he should have written a short opinion explaining that the plenary pardon power is part of the executive power vested in the President, and that the purpose of the pardon clause, U.S. Const. art. II, s 2, cl. 1, was merely to freeze the common law exception for cases of pardons after impeachment.

The Grossman opinion is also instructive for its discussion of the meaning of the words "of the United States" in the context of Article II, s 2, cl. 1. Contrary to the expansive gloss offered by Professor Calabresi and Mr. Rhodes on the same words in Article III, s 1, cl. 1, see supra note 91, Taft understood "offenses against the United States" in Article II to mean "as distinguished from offenses against the States." Grossman, 267 U.S. at 113-14.

<99>See supra text following note 89 (Professor Calabresi and Mr. Rhodes's argument that the words "shall extend to" in Article III, s 2 limit judicial power to those listed in that clause).

Many of the Framers found it "axiomatic" that the judicial power must be co- extensive with the legislative and the executive power. See, e.g., The Federalist Papers No. 80 (Hamilton).

<100>41 U.S. (16 Pet.) 1 (1842) (relying on federal common law to decide suit whose federal jurisdiction was founded on parties' diversity of citizenship).

<101>304 U.S. 64 (1938) (overruling Swift to hold that federal courts sitting in diversity should apply the common law of the states).

<102>11 U.S. (7 Cranch) 32 (1812) (holding federal courts lack jurisdiction over common law crimes).

<103>See supra note 15.

<104>Myers v. United States, 272 U.S. 52, 176 (1926) (holding statute requiring Senate consent to presidential removal of postmaster unconstitutional).

<105>277 U.S. 189 (1928) (holding that statute permitting certain members of Philippine legislature to vote government-owned shares in corporations ex officio violated principle of separation of powers).

<106>Humphrey's Ex'r v. United States, 295 U.S. 602 (1935) (upholding statute preventing President from firing FTC Commissioner without cause against constitutional challenge).

<107>Wiener v. United States, 357 U.S. 349 (1958) (holding that statute which was silent as to tenure in office of member of "quasi-judicial" War Claims Commission barred President from dismissing member without cause).

<108>462 U.S. 919 (1983) (holding that a two-house legislative veto violates constitutional requirement that Congress act in accordance with Presentment Clause of Constitution).

<109>424 U.S. 1 (1976) (holding that members of Congress could not constitutionally serve on FEC).

<110>478 U.S. 714 (1986) (holding that Comptroller General could not be an executive branch official because Congress reserved the right to remove him by legislation and that therefore he could not constitutionally exercise budget- cutting powers given to him by Deficit Reduction Act).

<111>111 S. Ct. 2298 (1991) (holding that participation of members of Congress in committee with power to veto decisions regarding local airports violated separation of powers).

<112>487 U.S. 654 (1988) (upholding constitutionality of independent special prosecutor).

<113>478 U.S. 833 (1986) (holding that CFTC's jurisdiction to hear state law common law counterclaims did not violate separation of powers or infringe courts' Article III jurisdiction).

<114>488 U.S. 361 (1989) (upholding constitutionality of Sentencing Commission against challenge predicated on participation of Article III judges on commission).

<115>Youngstown Sheet & Tube Co. v. Sawyer, 343 U.S. 579 (1952) (upholding grant of injunction against President's assertion of power to seize steel mills in peacetime; rejecting President's claim of inherent or implied power to seize property absent specific statutory or constitutional authorization).

<116>433 U.S. 425 (1977) (upholding constitutionality of Presidential Recordings and Materials Preservation Act, on grounds that President had no relevant enumerated power, against challenge by former President predicated, inter alia, on separation of powers and presidential privilege).

<117>The Supreme Court clearly believed that the congressional removal power in Bowsher was a congressional attempt to acquire undue leverage over an official arguably in the executive branch, and thus held that the existence of the removal provision alone sufficed to classify the Comptroller as a legislative branch official. See Bowsher v. Synar, 478 U.S. 714, 730-31 (1986).

<118>See supra note 6.

<119>See, e.g., Public Citizen v. United States Dep't of Justice, 491 U.S. 440, 467, 483-89 (1989) (Kennedy, J., concurring) (holding application of open meetings law to ABA consultation with Justice Department on judicial appointments would impermissibly infringe President's enumerated appointment power); Nixon, 433 U.S. at 443 (upholding statutory intrusion into disposal of presidential papers on grounds that President had no relevant enumerated power); United States v. Klein, 80 U.S. (13 Wall.) 128 (1871) (statute seeking to alter effect of presidential pardon unconstitutional interference with enumerated presidential power).

<120>Congress can also give powers to the states. In addition, Congress can charter a federal corporation, which directly affects private parties and requires little or no action by the other branches. The rule proposed in the text remains valid so long as the actors are outside the legislative branch.

<121>Myers concerned a congressional attempt to require senatorial approval before a postmaster was fired. Myers v. United States, 272 U.S. 52 (1926). Bowsher held that the Comptroller General could not be an executive branch official because Congress reserved the right to remove him by legislation. Bowsher, 478 U.S. at 727-32. Chadha struck down the one-house veto because it impermissibly inserted Congress into an executive function. INS v. Chadha, 462 U.S. 919, 958-59 (1983). Metropolitan Airports held that Congresspersons could not serve on a commission that had the power to reverse certain executive decisions. Metropolitan Wash. Airports Auth. v. Citizens for the Abatement of Aircraft Noise, Inc., 111 S. Ct. 2298, 2306 (1991).

<122>In Humphrey's Executor, Weiner, and Morrison the statute in question reduced the President's removal power, but did not inject Congress into the procedure. By contrast, in Commodity Futures Trading Comm'n v. Schor, 478 U.S. 833, 856 (1986), the Court noted that Congress was not seeking to increase its own powers.

<123>See Kendall v. United States ex rel. Stokes, 37 U.S. (12 Pet.) 524, 610 (1838) (stating that although the President has a special constitutional position, "it by no means follows that every officer in every branch of a department is under the exclusive direction of the President").

From a functionalist perspective, however, an executive branch official shielded from presidential control may indeed be more responsive to the dictates of the subcommittees that control her budget than a comparable official who serves at the pleasure of the President. Thus, the Supreme Court's sometime rejection of functional analysis, see supra note 15, may ultimately weaken the Presidency.

<124>Formally, the creation of executive branch officers shielded in some manner from the President's power to fire makes the President relatively weaker than the President would be had Congress not imposed the shield, but this is permitted so long as Congress is not actually any stronger in an absolute sense.

<125>Calabresi & Rhodes, supra note 4, at 1191.

<126>The executive power is vested in a President; the judicial power is vested in "one supreme Court, and such inferior Courts as the Congress may from time to time ordain and establish." U.S. Const. art. III s 1. Professor Calabresi and Mr. Rhodes would have it that this makes the entire federal judiciary the logical correlative of President. Word order can take you only so far; sometimes you have to consider what the words mean. If inferior federal courts are not constitutionally required, the correct analogue of President is indeed "Supreme Court," for both the President and the Supreme Court are actors with constitutional duties and powers and stand at the apex of a constitutional department whose inferior components owe their existence and duties to acts of Congress. See William S. Dodge, Note, Congressional Control of Supreme Court Appellate Jurisdiction: Why the Original Jurisdiction Clause Suggests an "Essential Role," 100 Yale L.J. 1013 (1991) (discussing hierarchical relationship between Supreme Court and inferior federal courts). Once created, of course, federal judges enjoy the constitutional privilege of life tenure which differentiates them from nonministerial officers in the executive.

<127>For a very thoughtful and thorough discussion of the modern executive branch, which recognizes both unitary and pluralistic tendencies, see Herz, supra note 26, at 914-30.

<128>Further, the 18th-century criticism that plurality in the Executive tends to blur responsibility and thus democratic accountability is as politically valid today as then. The problem with elevating this into a structural constitutional maxim is that so much of the pre-amendment Constitution was designed to be undemocratic ( e.g., electoral college, appointment of senators, judicial review). It is difficult to be both an originalist and a raging democrat at the same time (except when speaking of the House of Representatives) without invoking a sophisticated theory of constitutional evolution caused either by changes in national mores, see Ackerman , supra note 66, or by amendments to the Constitution.

<129>See Calabresi & Rhodes, supra note 4, at 1181-82. Saikrishna B. Prakash, Note, Hail To the Chief Administrator: The Framers and the President's Administrative Powers, 102 Yale L.J. 991 (1993), marshals historical evidence to argue that the legislative history of the Constitution shows that the inferior appointments clause was intended to strengthen the President by taking the burden of making minor appointments off his hands. Even assuming this to be correct, but see Lessig & Sunstein, supra note 19, at 22-44, there is ample other evidence from the text of the Constitution alone, see infra note 140, to reject the unitarian thesis.

<130>Professor Calabresi and Mr. Rhodes accuse my Note of reading Article II to create "a parity among executive officials" when in fact the Constitution gives "all executive power to 'a President,' " Calabresi & Rhodes, supra note 4, at 1204 (italics in original), as compared to a plural Judiciary. I did not suggest in my note that there is "a parity" between the President and other executive branch officials; there clearly is not. There is, however, a parity between the President and Congress; the question is to what extent that parity allows Congress to constrain an unenumerated presidential power

<131>Since the Constitution does not specify how many Justices constitute a Supreme Court, there is no textual obstacle to a Supreme Court consisting of one person. See Robert Cover, Violence and the Word, 95 Yale L.J. 1601, 1626 (1986), which illustrates the danger of relying on purely textual arguments in Constitutional interpretation.

<132>See, e.g., Sheldon v. Sill, 49 U.S. (8 How.) 441, 448-49 (1850) (stating that Congress had no obligation to create inferior courts and that it could restrict their jurisdiction). For an argument that lower courts are constitutionally required, see Eisenberg, supra note 54.

Professor Calabresi and Mr. Rhodes's assertion that the executive is unitary while the courts are plural is most plausible if one adopts Amar's view that Congress can give inferior federal courts jurisdiction which is shielded from Supreme Court review. Anyone who believes, as I do, that the supremacy of the Supreme Court requires that it have the power to hear appeals from inferior courts is unlikely to agree. Cf. Calabresi & Lawson, supra note 57, at 273-74 (noting hierarchical relationship between Supreme Court and inferior federal courts).

<133>Professor Calabresi and Mr. Rhodes see a United States in which Congress has "general legislative power whereas the executive and judicial powers have been splintered among independent agencies or officers and legislative courts." Calabresi & Rhodes, supra note 4, at 1176 (footnote omitted). In fact, the legislative power is equally "fragmented" by delegations to officers serving at the pleasure of the President who make administrative regulations. So-called independent agencies undoubtedly blur accountability in government (so too do ordinary agencies); even so, the case that the modern administrative state is unconstitutionally founded on the twin evils of improper insulation and excessive delegation is unpersuasive if admittedly not unthinkable.

<134>U.S. Const. art. II, s 3. The extent to which the Take Care Clause creates powers is slightly less clear. If one saw the President's removal power as deriving from the Take Care Clause, it would stand on a stronger textual footing than its current status as an unenumerated derivation from the appointments power. See infra text accompanying note 145. Congress is under no obligation to give the President any assistants to help discharge the Take Care function; but once Congress creates inferior federal officials, whatever duties are implied by the Take Care Clause attach to those offices.

<135>See Youngstown Sheet & Tube Co. v. Sawyer, 343 U.S. 579, 640-41 (1952) (Jackson, J., concurring).

<136>See Morrison v. Olson, 487 U.S. 654, 685-90 (1988).

<137>See Public Citizen v. United States Dep't of Justice, 491 U.S. 440, 484-86 (1989) (Kennedy, J., concurring in judgment). Professor Calabresi and Mr. Rhodes note this opinion in a footnote and seek to dismiss it by reiterating their mantra that "it is also true that the Article II Vesting Clause explicitly commits the entire executive power to the control of the President." Calabresi & Rhodes, supra note 4, at 1204 n.248. The prevalence of such assertions in The Structural Constitution makes it particularly unfortunate that the 63-page article does not indicate the authors' view of what the executive power might be. See supra text accompanying note 96.

Verdon, supra note 88, at 1256, suggests that the Jackson-Rehnquist-Kennedy- O'Connor approach amounts to using a "formalist" approach to enumerated power, and a "functionalist" approach to unenumerated powers. Cf. Peter L. Strauss, Formal and Functional Approaches to Separation-of-Powers Questions-- A Foolish Inconsistency?, 72 Cornell L. Rev. 488, 522-24 (1987) (arguing functionalism can work).

<138>Calabresi & Rhodes, supra note 4, at 1204.

<139>See supra text accompanying notes 30, 87.

<140>This government is acknowledged by all to be one of enumerated powers. The principle, that it can exercise only the powers granted to it, would seem too apparent to have required to be enforced by all those arguments which its enlightened friends, while it was depending before the people, found it necessary to urge. That principle is now universally admitted. McCulloch v. Maryland, 17 U.S. (4 Wheat.) 316, 405 (1819).

<141>Officers of the United States are mentioned in U.S. Const. art. I, s 6, cl. 2 (the Incompatibility Clause); art. I, s 8, cl. 18 (Necessary and Proper clause); art. I, s 9, cl. 8 (prohibition on acceptance of foreign Emolument, Office, or Title); art. II, s 1, cl. 6 (Congress may by law provide for devolution of Presidency to other Officer if neither President nor Vice President can serve); art. II, s 2, cl. 1 (Opinions in Writing Clause); art. II, s 2, cl. 2 (Appointments Clause) (twice); art. II, s 4 (impeachment of officers). Officers of the United States are also mentioned in id. amend. XIV, s 3.

<142>This argument does not rely on the vanishing distinction between "purely" executive powers and those not "purely" executive. See Morrison v. Olson, 487 U.S. 654, 689-90 (1988) (abandoning classifications used in Humphrey's Executor and Wiener); see also Froomkin, supra note 19, at 809-14 (proposing distinction between "presidential" and "executive" powers). As Chief Justice Taft stated in Grossman,

The Federal Constitution nowhere expressly declares that the three branches of the Government shall be kept separate and independent. All legislative powers are vested in a Congress. The executive power is vested in a President. The judicial power is vested in one Supreme Court and in such inferior courts as Congress may from time to time establish.... Complete independence and separation between the three branches, however, are not attained, or intended, as other provisions of the Constitution and the normal operation of government under it easily demonstrate....

There are instances of positive and negative restraints possibly available under the Constitution to each branch of the government in defeat of the action of the other. They show that the independence of each of the others is qualified and is so subject to exception as not to constitute a broadly positive injunction or a necessarily controlling rule of construction. The fact is that the Judiciary, quite as much as Congress and the Executive, is dependent on the cooperation of the other two, that government may go on. Ex Parte Grossman, 267 U.S. 87, 119-20 (1925).

<143>See supra text accompanying note 97. Some scholars question the relevance, or possibility, of basing legal arguments on an idea of traditions which they find to be no more than a social construct--and an inaccurate idea at that. See, e.g., Stephen J. Schnably, Property and Pragmatism: A Critique of Radin's Theory of Property and Personhood, 45 Stan. L. Rev. 347 (1993).

<144>Froomkin, supra note 19, at 809-11.

<145>U.S. Const. art. II, s 3. Justice Scalia recognized the importance of the Take Care Clause in James B. Beam Distilling Co. v. Georgia, 111 S. Ct. 2439, 2450 (1991) (Scalia, J., concurring), when he stated that it was the only power that gave the President "power to bind private conduct in areas not specifically committed to his control by Constitution or statute."

<146>See, e.g., Lee S. Liberman, Morrison v. Olson : A Formalistic Perspective on Why the Court Was Wrong, 38 Am. U. L. Rev. 313, 325-41 (1989). In deference to the fact that the Appointments Clause provides an alternate mechanism for the appointment of inferior officers by judges and heads of departments, the argument is usually restricted to the President's power to fire superior officers.

<147>See Froomkin, supra note 19, at 799-808. Professor Calabresi and Mr. Rhodes dismiss this as a "tactic of adding together inferences against presidential removal power (from the appointment vesting power and surrounding historical data) and against presidential control (from the Opinion in Writing Clause) to trump the Vesting Clause of Article II." Calabresi & Rhodes, supra note 4, at 1208. Professor Calabresi and Mr. Rhodes undermine their analysis of the Take Care Clause as a grant of plenary supervisory power to the President with the admission that "in many circumstances, the Take Care Clause may obligate the President to vest inferior officers with executive powers." Id. at 1183 n.151. The doctrinal tangle this causes is revealed by subsequent statements that these officers may be " statutorily enabled to act" but are only " constitutionally enabled to do so" when "the President (perhaps tacitly) chooses to delegate his executive power." Id. at 1185 n.158. To which one might respond, How does the President do this? Has it ever not been done?

<148>Despite the statements to the contrary made by Professor Calabresi and Mr. Rhodes, which unfortunately are repeated uncritically in Prakash, supra note 129, at 1008, my note argues only that Congress can put a "for cause" limit on the President's power to fire certain classes of civil officers. I did not, and do not, argue that every civil officer can necessarily be insulated from the President's removal power, especially if there is "cause" for removal. See Froomkin, supra note 19, at 809-14.

<149>Chief Justice Rehnquist, writing for six other Justices, adopted this view in Morrison v. Olson, 487 U.S. 654, 691-92 (1988).

<150>See Froomkin, supra note 19, at 812-14.

<151>My Note tentatively offered some additional principles based on existing removal statutes, see id. at 813-14, but these left much to be desired. A better version appears in Lessig & Sunstein, supra note 19, at 106-18.

<152>For one such attempt, see Verdon, supra note 88. Relying in part on the language of the three Vesting Clauses, Verdon proposes that the Court generally defer to Congress when questions cannot be resolved by reference to enumerated powers, and he argues that this is in fact what the Court has done. Id. at 1277-89.

<153>See Rebecca L. Brown, Separated Powers and Ordered Liberty, 139 U. Pa. L. Rev. 1513, 1530-31 (1991); Froomkin, supra note 19, at 804.

<154>On how the weighing process should work, see Steven J. Burton , Judging in Good Faith s 2.3 (1992). On the dangers, and cognitive necessity, of metaphor, see Steven L. Winter, The Metaphor of Standing and the Problem of Self-Governance, 40 Stan. L. Rev. 1371, 1382-93, 1458-1503 (1988).

<155>This may be a recent phenomenon. See Morrison v. Olson, 487 U.S. 654, 724 (1988) (Scalia, J., dissenting) (noting that Humphrey's Executor was once seen as a reactionary decision by an anti-New Deal Court).

<156>For one attempt to engage these questions directly, see Harold J. Krent & Ethan G. Shenkman, Of Citizen Suits and Citizen Sunstein, 91 Mich. L. Rev. 1793 (1993), which responds to Cass R. Sunstein, What's Standing After Lujan ? Of Citizen Suits, "Injuries," and Article III, 91 Mich. L. Rev. 163 (1992). The authors argue against citizen standing by setting out a bright- line vision of the limits of Congress's power to delegate law enforcement power. Cf. Cass R. Sunstein, Correspondence, Article II Revisionism, 92 Mich. L. Rev. 131 (1993) (response). But see Winter, supra note 154 (reviewing historical evidence and republican arguments).

<157>This article was accepted for publication shortly after the Clinton election. Since then, Minority Leader Robert Dole and other Senate Republicans have discovered a new interest in restricting the President's power over foreign policy. See Helen Dewar, Now It's the GOP Asserting Role For Congress on Foreign Policy, Wash. Post , Oct. 26, 1993, at A20. Republican senators who had objected to the creation of an Independent Counsel when the President was a Republican backed a very similar measure once the Counsel's target would be a Democratic administration. See, e.g., Adam Clymer, Congress Ready to Revive Independent Counsel Law, and With It a Conflict, N.Y. Times , Jan. 26, 1994, at A11.