U. S. Term Limits, Inc. v. Thornton

Justice Thomas, with whom The Chief Justice, Justice O’Connor, and Justice Scalia join,

dissenting.

It is ironic that the Court bases today’s decision on the right of the people to “choose whom they please to govern them.” See ante, at 783, 793, 795, 819. Under our Constitution, there is only one State whose people have the right to “choose whom they please” to represent Arkansas in Congress. The Court holds, however, that neither the elected legislature of that State nor the people themselves (acting by ballot initiative) may prescribe any qualifications for those representatives. The majority therefore defends the right of the people of Arkansas to “choose whom they please to govern them” by invalidating a provision that won nearly 60% of the votes cast in a direct election and that carried every congressional district in the State.

I dissent. Nothing in the Constitution deprives the people of each State of the power to prescribe eligibility requirements for the candidates who seek to represent them in Congress. The Constitution is simply silent on this question. And where the Constitution is silent, it raises no bar to action by the States or the people.

*846I

Because the majority fundamentally misunderstands the notion of “reserved” powers, I start with some first principles. Contrary to the majority’s suggestion, the people of the States need not point to any affirmative grant of power in the Constitution in order to prescribe qualifications for their representatives in Congress, or to authorize their elected state legislators to do so.

A

Our system of government rests on one overriding principle: All power stems from the consent of the people. To phrase the principle in this way, however, is to be imprecise about something important to the notion of “reserved” powers. The ultimate source of the Constitution’s authority is the consent of the people of each individual State, not the consent of the undifferentiated people of the Nation as a whole.

The ratification procedure erected by Article VII makes this point clear. The Constitution took effect once it had been ratified by the people gathered in convention in nine different States. But the Constitution went into effect only “between the States so ratifying the same,” Art. VII; it did not bind the people of North Carolina until, they had accepted it. In Madison’s words, the popular consent upon which the Constitution’s authority rests was “given by the people, not as individuals composing one entire nation, but as composing the distinct and independent States to which they respectively belong.” The Federalist No. 39, p. 243 (C. Rossiter ed. 1961) (hereinafter The Federalist). Accord, 3 Debates in the Several State Conventions on the Adoption of the Federal Constitution 94 (J. Elliot 2d ed. 1876) (hereinafter Elliot) (remarks of James Madison at the Virginia Convention).1

*847When they adopted the Federal Constitution, of course, the people of each State surrendered some of their authority to the United States (and hence to entities accountable to the people of other States as well as to themselves). They affirmatively deprived their States of certain powers, see, e.g., Art. I, §10, and they affirmatively conferred certain powers upon the Federal Government, see, e. g., Art. I, § 8. Because the people of the several States are the only true source of power, however, the Federal Government enjoys no authority beyond what the Constitution confers: The Federal Government’s powers are limited and enumerated. In the words of Justice Black: “The United States is entirely a creature of the Constitution. Its power and authority have no other source.” Reid v. Covert, 354 U. S. 1, 5-6 (1957) (plurality opinion) (footnote omitted).

In each State, the remainder of the people’s powers— “[t]he powers not delegated to the United States by the Constitution, nor prohibited by it to the States,” Arndt. 10 — are either delegated to the state government or retained by the people. The Federal Constitution does not specify which of these two possibilities obtains; it is up to the various state constitutions to declare which powers the people of each State have delegated to their state government. As far as *848the Federal Constitution is concerned, then, the States can exercise all powers that the Constitution does not withhold from them. The Federal Government and the States thus face different default rules: Where the Constitution is silent about the exercise of a particular power — that is, where the Constitution does not speak either expressly or by necessary implication — the Federal Government lacks that power and the States enjoy it.

These basic principles are enshrined in the Tenth Amendment, which declares that all powers neither delegated to the Federal Government nor prohibited to the States “are reserved to the States respectively, or to the people.” With this careful last phrase, the Amendment avoids taking any position on the division of power between the state governments and the people of the States: It is up to the people of each State to determine which “reserved” powers their state government may exercise. But the Amendment does make clear that powers reside at the state level except where the Constitution removes them from that level. All powers that the Constitution neither delegates to the Federal Government nor prohibits to the States are controlled by the people of each State.

To be sure, when the Tenth Amendment uses the phrase “the people,” it does not specify whether it is referring to the people of each State or the people of the Nation as a whole. But the latter interpretation would make the Amendment pointless: There would have been no reason to provide that where the Constitution is silent about whether a particular power resides at the state level, it might or might not do so. In addition, it would make no sense to speak of powers as being reserved to the undifferentiated people of the Nation as a whole, because the Constitution does not contemplate that those people will either exercise power or delegate it. The Constitution simply does not recognize any mechanism for action by the undifferentiated people of the Nation. Thus, the amendment provision of Article *849V calls for amendments to be ratified not by a convention of the national people, but by conventions of the people in each State or by the state legislatures elected by those people. Likewise, the Constitution calls for Members of Congress to be chosen State by State, rather than in nationwide elections. Even the selection of the President — surely the most national of national figures — is accomplished by an electoral college made up of delegates chosen by the various States, and candidates can lose a Presidential election despite winning a majority of the votes cast in the Nation as a whole. See also Art. II, § 1, cl. 3 (providing that when no candidate secures a majority of electoral votes, the election of the President is thrown into the House of Representatives, where “the Votes shall be taken by States, the Representatives from each State having one Vote”); Arndt. 12 (same).

In short, the notion of popular sovereignty that undergirds the Constitution does not erase state boundaries, but rather tracks them. The people of each State obviously did trust their fate to the people of the several States when they consented to the Constitution; not only did they empower the governmental institutions of the United States, but they also agreed to be bound by constitutional amendments that they themselves refused to ratify. See Art. V (providing that proposed amendments shall take effect upon ratification by three-quarters of the States). At the same time, however, the people of each State retained their separate political identities. As Chief Justice Marshall put it, “[n]o political dreamer was ever wild enough to think of breaking down the lines which separate the States, and of compounding the American people into one common mass.” McCulloch v. Maryland, 4 Wheat. 316, 403 (1819).2

*850Any ambiguity in the Tenth Amendment’s use of the phrase “the people” is cleared up by the body of the Constitution itself. Article I begins by providing that the Congress of the United States enjoys “[a]ll legislative Powers herein granted,” § 1, and goes on to give a careful enumeration of Congress’ powers, § 8. It then concludes by enumerating certain powers that are prohibited to the States. The import of this structure is the same as the import of the Tenth Amendment: If we are to invalidate Arkansas’ Amendment 73, we must point to something in the Federal Constitution that deprives the people of Arkansas of the power to enact such measures.

B

The majority disagrees that it bears this burden. But its arguments are unpersuasive.

1

The majority begins by announcing an enormous and untenable limitation on the principle expressed by the Tenth Amendment. According to the majority, the States possess only those powers that the Constitution affirmatively grants to them or that they enjoyed before the Constitution was adopted; the Tenth Amendment “could only ‘reserve’ that *851which existed before.” Ante, at 802. From the fact that the States had not previously enjoyed any powers over the particular institutions of the Federal Government established by the Constitution,3 the majority derives a rule precisely opposite to the one that the Amendment actually prescribes: “‘[T]he states can exercise no powers whatsoever, which exclusively spring out of the existence of the national government, which the constitution does not delegate to them.’ ” Ibid, (quoting 1 J. Story, Commentaries on the Constitution of the United States § 627 (3d ed. 1858)).

The majority’s essential logic is that the state governments could not “reserve” any powers that they did not control at the time the Constitution was drafted. But it was not the state governments that were doing the reserving. The Constitution derives its authority instead from the consent of the people of the States. Given the fundamental principle that all governmental powers stem from the people of the States, it would simply be incoherent to assert that the people of the States could not reserve any powers that they had not previously controlled.

The Tenth Amendment’s use of the word “reserved” does not help the majority’s position. If someone says that the power to use a particular facility is reserved to some group, he is not saying anything about whether that group has previously used the facility. He is merely saying that the peo*852pie who control the facility have designated that group as the entity with authority to use it. The Tenth Amendment is similar: The people of the States, from whom all governmental powers stem, have specified that all powers not prohibited to the States by the Federal Constitution are reserved “to the States respectively, or to the people.”

The majority is therefore quite wrong to conclude that the people of the States cannot authorize their state governments to exercise any powers that were unknown to the States when the Federal Constitution was drafted. Indeed, the majority’s position frustrates the apparent purpose of the Amendment’s final phrase. The Amendment does not preempt any limitations on state power found in the state constitutions, as it might have done if it simply had said that the powers not delegated to the Federal Government are reserved to the States. But the Amendment also does not prevent the people of the States from amending their state constitutions to remove limitations that were in effect when the Federal Constitution and the Bill of Rights were ratified.

In an effort to defend its position, the majority points to language in Garcia v. San Antonio Metropolitan Transit Authority, 469 U. S. 528, 549 (1985), which it takes to indicate that the Tenth Amendment covers only “the original powers of [state] sovereignty.” Ante, at 802. But Garcia dealt with an entirely different issue: the extent to which principles of state sovereignty implicit in our federal system curtail Congress’ authority to exercise its enumerated powers. When we are asked to decide whether a congressional statute that appears to have been authorized by Article I is nonetheless unconstitutional because it invades a protected sphere of state sovereignty, it may well be appropriate for us to inquire into what we have called the “traditional aspects of state sovereignty.” See National League of Cities v. Usery, 426 U. S. 833, 841, 849 (1976); see also New York v. United States, 505 U. S. 144, 156-157 (1992). The question *853raised by the present case, however, is not whether any principle of state sovereignty implicit in the Tenth Amendment bars congressional action that Article I appears to authorize, but rather whether Article I bars state action that it does not appear to forbid. The principle necessary to answer this question is express on the Tenth Amendment’s face: Unless the Federal Constitution affirmatively prohibits an action by the States or the people, it raises no bar to such action.

The majority also seeks support for its view of the Tenth Amendment in McCulloch v. Maryland, 4 Wheat. 316 (1819). See ante, at 802. But this effort is misplaced. McCulloch did make clear that a power need not be “expressly” delegated to the United States or prohibited to the States in order to fall outside the Tenth Amendment’s reservation; delegations and prohibitions can also arise by necessary implication.4 True to the text of the Tenth Amendment, however, McCulloch indicated that all powers as to which the Constitution does not speak (whether expressly or by necessary implication) are “reserved” to the state level. Thus, in its only discussion of the Tenth Amendment, McCulloch observed that the Amendment “leav[es] the question, whether the particular power which may become the subject of contest has been delegated to the one government, or prohibited to the other, to depend on a fair construction of the whole [Constitution].” 4 Wheat., at 406. McCulloch did not qualify this observation by indicating that the question also turned on whether the States had enjoyed the power before the framing. To the contrary, McCulloch seemed to assume that the people had “conferred on the general government the power contained in the constitution, and on the States the whole residuum of power.” Id., at 410.

The structure of McCulloch’s analysis also refutes the majority’s position. The question before the Court was *854whether the State of Maryland could tax the Bank of the United States, which Congress had created in an effort to accomplish objects entrusted to it by the Constitution. Chief Justice Marshall’s opinion began by upholding the federal statute incorporating the bank. Id., at 400-425. It then held that the Constitution affirmatively prohibited Maryland’s tax on the bank created by this statute. Id., at 425-437. The Court relied principally on concepts that it deemed inherent in the Supremacy Clause of Article VI, which declares that “[t]his Constitution, and the Laws of the United States which shall be made in Pursuance thereof,... shall be the supreme Law of the Land ....” In the Court’s view, when a power has been “delegated to the United States by the Constitution,” Arndt. 10, the Supremacy Clause forbids a State to “retard, impede, burden, or in any manner control, the operations of the constitutional laws enacted by Congress to carry [that power] into execution.” McCulloch, 4 Wheat., at 436. Thus, the Court concluded that the very nature of state taxation on the bank’s operations was “incompatible with, and repugnant to,” the federal statute creating the bank. See id., at 425.

For the past 175 years, McCulloch has been understood to rest on the proposition that the Constitution affirmatively barred Maryland from imposing its tax on the Bank’s operations. See, e. g., Osborn v. Bank of United States, 9 Wheat. 738, 859-868 (1824) (reaffirming McCulloch’s conclusion that by operation of the Supremacy Clause, the federal statute incorporating the bank impliedly pre-empted state laws attempting to tax the bank’s operations); Maryland v. Louisiana, 451 U. S. 725, 746 (1981) (citing McCulloch for the proposition that the Supremacy Clause deprives the States of the power to pass laws that conflict with federal statutes); see also North Dakota v. United States, 495 U. S. 423, 434 (1990) (plurality opinion) (citing McCulloch for the proposition that state laws may violate the Supremacy Clause when they “regulate the Government directly or discriminate against *855it”).5 For the majority, however, McCulloch apparently turned on the fact that before the Constitution was adopted, the States had possessed no power to tax the instrumentalities of the governmental institutions that the Constitution created. This understanding of McCulloch makes most of Chief Justice Marshall’s opinion irrelevant; according to the majority, there was no need to inquire into whether federal law deprived Maryland of the power in question, because the power could not fall into the category of “reserved” powers anyway.6

*856Despite the majority’s citation of Garcia and McCulloch, the only true support for its view of the Tenth Amendment comes from Joseph Story’s 1833 treatise on constitutional law. See 2 J. Story, Commentaries on the Constitution of the United States §§623-628. Justice Story was a brilliant and accomplished man, and one cannot casually dismiss his views. On the other hand, he was not a member of the Founding generation, and his Commentaries on the Constitution were written a half century after the framing. Rather than representing the original understanding of the Constitution, they represent only his own understanding. In a range of cases concerning the federal/state relation, moreover, this Court has deemed positions taken in Story’s commentaries to be more nationalist than the Constitution warrants. Compare, e.g., id., §§1063-1069 (arguing that the Commerce Clause deprives the States of the power to regulate any commerce within Congress’ reach), with Cooley v. Board of Wardens of Port of Philadelphia ex rel. Soc. for Relief of Distressed Pilots, 12 How. 299 (1852) (holding that Congress’ Commerce Clause powers are not exclusive). See also 1 Life and Letters of Joseph Story 296 (W. Story ed. 1851) (extract of manuscript written by Story) (“I hold it to be a maxim, which should never be lost sight of by a great statesman, that the Government of the United States is *857intrinsically too weak, and the powers of the State Governments too strong”). In this case too, Story’s position that the only powers reserved to the States are those that the States enjoyed before the framing conflicts with both the plain language of the Tenth Amendment and the underlying theory of the Constitution.

2

The majority also sketches out what may be an alternative (and narrower) argument. Again citing Story, the majority suggests that it would be inconsistent with the notion of “national sovereignty” for the States or the people of the States to have any reserved powers over the selection of Members of Congress. See ante, at 803, 805. The majority apparently reaches this conclusion in two steps. First, it asserts that because Congress as a whole is an institution of the National Government, the individual Members of Congress “owe primary allegiance not to the people of a State, but to the people of the Nation.” See ante, at 803. Second, it concludes that because each Member of Congress has a nationwide constituency once he takes office, it would be inconsistent with the Framers’ scheme to let a single State prescribe qualifications for him. See ante, at 803-804, 837-838.

Political scientists can debate about who commands the “primary allegiance” of Members of Congress once they reach Washington. From the framing to the present, however, the selection of the Representatives and Senators from each State has been left entirely to the people of that State or to their state legislature. See Art. I, § 2, cl. 1 (providing that Members of the House of Representatives are chosen “by the People of the several States”); Art. I, § 3, cl. 1 (originally providing that the Senators from each State are “chosen by the Legislature thereof”); Arndt. 17 (amending § 3 to provide that the Senators from each State are “elected by the people thereof”). The very name “congress” suggests a *858coming together of representatives from distinct entities.7 In keeping with the complexity of our federal system, once the representatives chosen by the people of each State assemble in Congress, they form a national body and are beyond the control of the individual States until the next election. But the selection of representatives in Congress is indisputably an act of the people of each State, not some abstract people of the Nation as a whole.

The concurring opinion suggests that this cannot be so, because it is the Federal Constitution that guarantees the right of the people of each State (so long as they are qualified electors under state law) to take part in choosing the Members of Congress from that State. See ante, at 842. But the presence of a federally guaranteed right hardly means that the selection of those representatives constitutes “the exercise of federal authority.” See ante, at 841. When the people of Georgia pick their representatives in Congress, they are acting as the people of Georgia, not as the corporate agents for the undifferentiated people of the Nation as a whole. See In re Green, 134 U. S. 377, 379 (1890) (“Although [Presidential] electors are appointed and act under and pursuant to the Constitution of the United States, they are no more officers or agents of the United States than are the members of the state legislatures when acting as electors of federal senators, or the people of the States when acting as electors of representatives in Congress”). The concurring opinion protests that the exercise of “reserved” powers in the area of congressional elections would constitute “state interference with the most basic relation between the Na*859tional Government and its citizens, the selection of legislative representatives.” See ante, at 842. But when one strips away its abstractions, the concurring opinion is simply saying that the people of Arkansas cannot be permitted to inject themselves into the process by which they themselves select Arkansas’ representatives in Congress.

The concurring opinion attempts to defend this surprising proposition by pointing out that Americans are “citizens of the United States” as well as “of the State wherein they reside,” Arndt. 14, § 1, and that national citizenship (particularly after the ratification of the Fourteenth Amendment) “has privileges and immunities protected from state abridgment by the force of the Constitution itself,” ante, at 842. These facts are indeed “beyond dispute,” ante, at 844, but they do not contradict anything that I have said. Although the United States obviously is a Nation, and although it obviously has citizens, the Constitution does not call for Members of Congress to be elected by the undifferentiated national citizenry; indeed, it does not recognize any mechanism at all (such as a national referendum) for action by the undifferentiated people of the Nation as a whole. See supra, at 848-849. Even at the level of national politics, then, there always remains a meaningful distinction between someone who is a citizen of the United States and of Georgia and someone who is a citizen of the United States and of Massachusetts. The Georgia citizen who is unaware of this distinction will have it pointed out to him as soon as he tries to vote in a Massachusetts congressional election.

In short, while the majority is correct that the Framers expected the selection process to create a “direct link” between Members of the House of Representatives and the people, ante, at 803, the link was between the Representatives from each State and the people of that State; the people of Georgia have no say over whom the people of Massachusetts select to represent them in Congress. This arrange*860ment must baffle the majority,8 whose understanding of Congress would surely fit more comfortably within a system of nationwide elections. But the fact remains that when it comes to the selection of Members of Congress, the people of each State have retained their independent political identity. As a result, there is absolutely nothing strange about the notion that the people of the States or their state legislatures possess “reserved” powers in this area.

The majority seeks support from the Constitution’s specification that Members of Congress “shall receive a Compensation for their Services, to be ascertained by Law, and paid out of the Treasury of the United States.” Art. I, §6, cl. 1; see ante, at 804. But the fact that Members of Congress draw a federal salary once they have assembled hardly means that the people of the States lack reserved powers over the selection of their representatives. Indeed, the historical evidence about the compensation provision suggests that the States’ reserved powers may even extend beyond the selection stage. The majority itself indicates that if the Constitution had made no provision for congressional compensation, this topic would have been “left to state legislatures.” Ante, at 809; accord, 1 Farrand 215-216 (remarks of James Madison and George Mason); id., at 219, n. *. Likewise, Madison specifically indicated that even with the compensation provision in place, the individual States still *861enjoyed the reserved power to supplement the federal salary. 3 id., at 315 (remarks at the Virginia ratifying convention).

As for the fact that a State has no reserved power to establish qualifications for the office of President, see ante, at 803-804, it surely need not follow that a State has no reserved power to establish qualifications for the Members of Congress who represent the people of that State. Because powers are reserved to the States “respectively,” it is clear that no State may legislate for another State: Even though the Arkansas Legislature enjoys the reserved power to pass a minimum-wage law for Arkansas, it has no power to pass a minimum-wage law for Vermont. For the same reason, Arkansas may not decree that only Arkansas citizens are eligible to be President of the United States; the selection of the President is not up to Arkansas alone, and Arkansas can no more prescribe the qualifications for that office than it can set the qualifications for Members of Congress from Florida. But none of this suggests that Arkansas cannot set qualifications for Members of Congress from Arkansas.

In fact, the Constitution’s treatment of Presidential elections actively contradicts the majority’s position. While the individual States have no “reserved” power to set qualifications for the office of President, we have long understood that they do have the power (as far as the Federal Constitution is concerned) to set qualifications for their Presidential electors — the delegates that each State selects to represent it in the electoral college that actually chooses the Nation’s chief executive. Even respondents do not dispute that the States may establish qualifications for their delegates to the electoral college, as long as those qualifications pass muster under other constitutional provisions (primarily the First and Fourteenth Amendments). See Williams v. Rhodes, 393 U. S. 23, 29 (1968); McPherson v. Blacker, 146 U. S. 1, 27-36 (1892). As the majority cannot argue that the Consti*862tution affirmatively grants this power,9 the power must be one that is “reserved” to the States. It necessarily follows that the majority’s understanding of the Tenth Amendment is incorrect, for the position of Presidential elector surely “ ‘spring[s] out of the existence of the national government.’ ” See ante, at 802.

3

In a final effort to deny that the people of the States enjoy “reserved” powers over the selection of their representatives in Congress, the majority suggests that the Constitution expressly delegates to the States certain powers over congressional elections. See ante, at 805. Such delegations of power, the majority argues, would be superfluous if the people of the States enjoyed reserved powers in this area.

Only one constitutional provision — the Times, Places and Manner Clause of Article I, § 4 — even arguably supports the majority’s suggestion. It reads:

“The Times, Places and Manner of holding Elections for Senators and Representatives, shall be prescribed in each State by the Legislature thereof; but the Congress may at any time by Law make or alter such Regulations, except as to the Places of chusing Senators.”

Contrary to the majority’s assumption, however, this Clause does not delegate any authority to the States. Instead, it simply imposes a duty upon them. The majority gets it exactly right: By specifying that the state legislatures “shall” prescribe the details necessary to hold congressional elections, the Clause “expressly requires action by the States.” *863See ante, at 804. This command meshes with one of the principal purposes of Congress’ “make or alter” power: to ensure that the States hold congressional elections in the first place, so that Congress continues to exist. As one reporter summarized a speech made by John Jay at the New York ratifying convention:

“[E]very government was imperfect, unless it had a power of preserving itself. Suppose that, by design or accident, the states should neglect to appoint representatives; certainly there should be some constitutional remedy for this evil. The obvious meaning of the paragraph was, that, if this neglect should take place, Congress should have power, by law, to support the government, and prevent the dissolution of the Union. [Jay] believed this was the design of the federal Convention.” 2 Elliot 326 (emphasis in original).10

Constitutional provisions that impose affirmative duties on the States are hardly inconsistent with the notion of reserved powers.

*864Of course, the second part of the Times, Places and Manner Clause does grant a power rather than impose a duty. As its contrasting uses of the words “shall” and “may” confirm, however, the Clause grants power exclusively to Congress, not to the States. If the Clause did not exist at all, the States would still be able to prescribe the times, places, and manner of holding congressional elections; the deletion of the provision would simply deprive Congress of the power to override these state regulations.

The majority also mentions Article II, §1, cl. 2: “Each State shall appoint, in such Manner as the Legislature thereof may direct, a Number of [Presidential] Electors, equal to the whole Number of Senators and Representatives to which the State may be entitled in the Congress . . . .” But this Clause has nothing to do with congressional elections, and in any event it, too, imposes an affirmative obligation on the States. In fact, some such barebones provision was essential in order to coordinate the creation of the electoral college. As mentioned above, moreover, it is uncontested that the States enjoy the reserved power to specify qualifications for the Presidential electors who are chosen pursuant to this Clause. See supra, at 861-862.

Respondent Thornton seeks to buttress the majority’s position with Article I, § 2, cl. 1, which provides:

“The House of Representatives shall be composed of Members chosen every second Year by the People of the several States, and the Electors in each State shall have the Qualifications requisite for Electors of the most numerous Branch of the State Legislature.”

According to respondent Thornton, this provision “grants States authority to prescribe the qualifications of [voters]” in congressional elections. Brief for Respondent Congressman Ray Thornton 4. If anything, however, the Clause limits the power that the States would otherwise enjoy. Though it does leave States with the ability to control who may vote *865in congressional elections, it has the effect of restricting their authority to establish special requirements that do not apply in elections for the state legislature.

Our case law interpreting the Clause affirmatively supports the view that the States enjoy reserved powers over congressional elections. We have treated the Clause as a one-way ratchet: While the requirements for voting in congressional elections cannot be more onerous than the requirements for voting in elections for the most numerous branch of the state legislature, they can be less so. See Tashjian v. Republican Party of Conn., 479 U. S. 208, 225-229 (1986). If this interpretation of the Clause is correct, it means that even with the Clause in place, States still have partial freedom to set special voting requirements for congressional elections. As this power is not granted in Article I, it must be among the “reserved” powers.

II

I take it to be established, then, that the people of Arkansas do enjoy “reserved” powers over the selection of their representatives in Congress. Purporting to exercise those reserved powers, they have agreed among themselves that the candidates covered by §3 of Amendment 73 — those whom they have already elected to three or more terms in the House of Representatives or to two or more terms in the Senate — should not be eligible to appear on the ballot for reelection, but should nonetheless be returned to Congress if enough voters are sufficiently enthusiastic about their candidacy to write in their names. Whatever one might think of the wisdom of this arrangement, we may not override the decision of the people of Arkansas unless something in the Federal Constitution deprives them of the power to enact such measures.

The majority settles on “the Qualifications Clauses” as the constitutional provisions that Amendment 73 violates. See ante, at 806. Because I do not read those provisions to im*866pose any unstated prohibitions on the States, it is unnecessary for me to decide whether the majority is correct to identify Arkansas’ ballot-access restriction with laws fixing true term limits or otherwise prescribing “qualifications” for congressional office. As I discuss in Part A below, the Qualifications Clauses are merely straightforward recitations of the minimum eligibility requirements that the Framers thought it essential for every Member of Congress to meet. They restrict state power only in that they prevent the States from abolishing all eligibility requirements for membership in Congress.

Because the text of the Qualifications Clauses does not support its position, the majority turns instead to its vision of the democratic principles that animated the Framers. But the majority’s analysis goes to a question that is not before us: whether Congress has the power to prescribe qualifications for its own members. As I discuss in Part B, the democratic principles that contributed to the Framers’ decision to withhold this power from Congress do not prove that the Framers also deprived the people of the States of their reserved authority to set eligibility requirements for their own representatives.

In Part C, I review the majority’s more specific historical evidence. To the extent that they bear on this case, the records of the Philadelphia Convention affirmatively support my unwillingness to find hidden meaning in the Qualifications Clauses, while the surviving records from the ratification debates help neither side. As for the postratification period, five States supplemented the constitutional disqualifications in their very first election laws. The historical evidence thus refutes any notion that the Qualifications Clauses were generally understood to be exclusive. Yet the majority must establish just such an understanding in order to justify its position that the Clauses impose unstated prohibitions on the States and the people. In my view, the historical evidence is simply inadequate to warrant the majority’s *867conclusion that the Qualifications Clauses mean anything more than what they say.

A

The provisions that are generally known as the Qualifications Clauses read as follows:

“No Person shall be a Representative who shall not have attained to the age of twenty five Years, and been seven Years a Citizen of the United States, and who shall not, when elected, be an Inhabitant of that State in which he shall be chosen.” Art. I, §2, cl. 2.
“No Person shall be a Senator who shall not have attained to the Age of thirty Years, and been nine Years a Citizen of the United States, and who shall not, when elected, be an Inhabitant of that State for which he shall be chosen.” Art. I, §3, cl. 3.

Later in Article I, the “Ineligibility Clause” imposes another nationwide disqualification from congressional office: “[N]o Person holding any Office under the United States, shall be a Member of either House during his Continuance in Office.” §6, cl. 2.

The majority is quite correct that the “negative phrasing” of these Clauses has little relevance. See ante, at 792, n. 8. The Qualifications Clauses would mean the same thing had they been enacted in the form that the Philadelphia Convention referred them to the Committee of Style:

“Every Member of the House of Representatives shall be of the age of twenty-five years at least; shall have been a citizen of the United States for at least seven years before his election; and shall be, at the time of his election, an inhabitant of the State in which he shall be chosen.” 2 Farrand 565.

See also id., at 567 (same phrasing for Senate Qualifications Clause). But these different formulations — whether negative or affirmative — merely establish minimum qualifica*868tions. They are quite different from an exclusive formulation, such as the following:

“Every Person who shall have attained to the age of twenty five Years, and been seven Years a Citizen of the United States, and who shall, when elected, be an Inhabitant of that State in which he shall be chosen, shall be eligible to be a Representative.”

At least on their face, then, the Qualifications Clauses do nothing to prohibit the people of a State from establishing additional eligibility requirements for their own representatives.

Joseph Story thought that such a prohibition was nonetheless implicit in the constitutional list of qualifications, because “[f]rom the very nature of such a provision, the affirmation of these qualifications would seem to imply a negative of all others.” 1 Commentaries on the Constitution of the United States § 624 (1833); see also ante, at 793, n. 9. This argument rests on the maxim expressio unius est exclusio alterius. When the Framers decided which qualifications to include in the Constitution, they also decided not to include any other qualifications in the Constitution. In Story’s view, it would conflict with this latter decision for the people of the individual States to decide, as a matter of state law, that they would like their own representatives in Congress to meet additional eligibility requirements.

To spell out the logic underlying this argument is to expose its weakness. Even if one were willing to ignore the distinction between requirements enshrined in the Constitution and other requirements that the Framers were content to leave within the reach of ordinary law, Story’s application of the expressio unius maxim takes no account of federalism. At most, the specification of certain nationwide disqualifications in the Constitution implies the negation of other nationwide disqualifications; it does not imply that individual States or their people are barred from adopting their own *869disqualifications on a state-by-state basis. Thus, the one delegate to the Philadelphia Convention who voiced anything approaching Story’s argument said only that a recital of qualifications in the Constitution would imply that Congress lacked any qualification-setting power. See 2 Farrand 123 (remarks of John Dickinson); cf. ante, at 793, n. 9, and 815-816, n. 27.

The Qualifications Clauses do prevent the individual States from abolishing all eligibility requirements for Congress. This restriction on state power reflects the fact that when the people of one State send immature, disloyal, or unknowledgeable representatives to Congress, they jeopardize not only their own interests but also the interests of the people of other States. Because Congress wields power over all the States, the people of each State need some guarantee that the legislators elected by the people of other States will meet minimum standards of competence. The Qualifications Clauses provide that guarantee: They list the requirements that the Framers considered essential to protect the competence of the National Legislature.11

If the people of a State decide that they would like their representatives to possess additional qualifications, however, they have done nothing to frustrate the policy behind the Qualifications Clauses. Anyone who possesses all of the constitutional qualifications, plus some qualifications required by state law, still has all of the federal qualifications. *870Accordingly, the fact that the Constitution specifies certain qualifications that the Framers deemed necessary to protect the competence of the National Legislature does not imply that it strips the people of the individual States of the power to protect their own interests by adding other requirements for their own representatives.

The people of other States could legitimately complain if the people of Arkansas decide, in a particular election, to send a 6-year-old to Congress. But the Constitution gives the people of other States no basis to complain if the people of Arkansas elect a freshman representative in preference to a long-term incumbent. That being the case, it is hard to see why the rights of the people of other States have been violated when the people of Arkansas decide to enact a more general disqualification of long-term incumbents. Such a disqualification certainly is subject to scrutiny under other constitutional provisions, such as the First and Fourteenth Amendments. But as long as the candidate whom they send to Congress meets the constitutional age, citizenship, and inhabitancy requirements, the people of Arkansas have not violated the Qualifications Clauses.

This conclusion is buttressed by our reluctance to read constitutional provisions to preclude state power by negative implication. The very structure of the Constitution counsels such hesitation. After all, § 10 of Article I contains a brief list of express prohibitions on the States. Cf. Cipollone v. Liggett Group, Inc., 505 U. S. 504, 517-519 (1992) (Stevens, J.) (applying the expressio unius maxim to conclude that Congress’ inclusion of an express pre-emption clause in a federal statute implies that state laws beyond the reach of that clause are not pre-empted); Nevada v. Hall, 440 U. S. 410, 425 (1979) (Stevens, J.) (suggesting that in light of the Tenth Amendment and the Constitution’s express prohibitions on the States, “caution should be exercised before concluding that unstated limitations on state power were intended by the Framers”). Many of the prohibitions listed in *871§ 10, moreover, might have been thought to be implicit in other constitutional provisions or in the very nature of our federal system. Compare, e. g., Art. II, § 2, cl. 2 (“[The President] shall have Power, by and with the Advice and Consent of the Senate, to make Treaties”), and Art. I, §8, cl. 5 (“The Congress shall have Power . . . [t]o coin Money”), with Art. I, § 10, cl. 1 (“No State shall enter into any Treaty” and “No State shall... coin Money”); see also Art. VI, cl. 2 (explicitly declaring that state law cannot override the Constitution). The fact that the Framers nonetheless made these prohibitions express confirms that one should not lightly read provisions like the Qualifications Clauses as implicit deprivations of state power. See generally Barron ex rel. Tiernan v. Mayor of Baltimore, 7 Pet. 243, 249 (1833).12

The majority responds that “a patchwork of state qualifications” would “undermin[e] the uniformity and the national character that the Framers envisioned and sought to ensure.” Ante, at 822. Yet the Framers thought it perfectly consistent with the “national character” of Congress for the Senators and Representatives from each State to be chosen by the legislature or the people of that State. The majority never explains why Congress’ fundamental character permits this state-centered system, but nonetheless prohibits *872the people of the States and their state legislatures from setting any eligibility requirements for the candidates who seek to represent them.

As for the majority’s related assertion that the Framers intended qualification requirements to be uniform, this is a conclusion, not an argument. Indeed, it is a conclusion that the Qualifications Clauses themselves contradict. At the time of the framing, and for some years thereafter, the Clauses’ citizenship requirements incorporated laws that varied from State to State. Thus, the Qualifications Clauses themselves made it possible that a person would be qualified to represent State A in Congress even though a similarly situated person would not be qualified to represent State B.

To understand this point requires some background. Before the Constitution was adopted, citizenship was controlled entirely by state law, and the different States established different criteria. See J. Kettner, Development of American Citizenship, 1608-1870, pp. 213-218 (1978). Even after the Constitution gave Congress the power to “establish an uniform Rule of Naturalization . . . throughout the United States,” Art. I, § 8, cl. 4, Congress was under no obligation to do so, and the Framers surely expected state law to continue in full force unless and until Congress acted. Cf. Sturges v. Crowninshield, 4 Wheat. 122, 196 (1819) (so interpreting the other part of § 8, cl. 4, which empowers Congress to establish “uniform Laws on the subject of Bankruptcies”).13 Accordingly, the constitutional requirement that *873Members of Congress be United States citizens meant different things in different States. The very first contested-election case in the House of Representatives, which involved the citizenship of a would-be Congressman from South Carolina, illustrates this principle. As Representative James Madison told his colleagues, “I take it to be a clear point, that we are to be guided, in our decision, by the laws and constitution of South Carolina, so far as they can guide us; and where the laws do not expressly guide us, we must be guided by principles of a general nature . . . .” Cases of Contested Elections in Congress 32 (M. Clarke & D. Hall eds. 1834) (reporting proceedings from May 22, 1789).

Even after Congress chose to exercise its power to prescribe a uniform route to naturalization, the durational element of the citizenship requirement in the Qualifications Clauses ensured that variances in state law would continue to matter. Thus, in 1794 the Senate refused to seat Albert Gallatin because, owing to the individual peculiarities of the laws of the two relevant States, he had not been a citizen for the required nine years. Id,., at 859-862, 867 (reporting proceedings from February 20 and 28, 1794).

Even if the Qualifications Clauses had not themselves incorporated nonuniform requirements, of course, there would still be no basis for the assertion of the plurality below that they mandate “uniformity in qualifications.” See 316 Ark. 251, 265, 872 S. W. 2d 349, 356 (1994). The Clauses wholly omit the exclusivity provision that, according to both the plurality below and today’s majority, was their central focus. In fact, neither the text nor the apparent purpose of the Qualifications Clauses does anything to refute Thomas Jefferson’s elegant legal analysis:

*874“Had the Constitution been silent, nobody can doubt but that the right to prescribe all the qualifications and disqualifications of those they would send to represent them, would have belonged to the State. So also the Constitution might have prescribed the whole, and excluded all others. It seems to have preferred the middle way. It has exercised the power in part, by declaring some disqualifications .... But it does not declare, itself, that the member shall not be a lunatic, a pauper, a convict of treason, of murder, of felony, or other infamous crime, or a non-resident of his district; nor does it prohibit to the State the power of declaring these, or any other disqualifications which its particular circumstances may call for; and these may be different in different States. Of course, then, by the tenth amendment, the power is reserved to the State.” Letter to Joseph C. Cabell (Jan. 31, 1814), in 14 Writings of Thomas Jefferson 82-83 (A. Lipscomb ed. 1904).14

B

Although the Qualifications Clauses neither state nor imply the prohibition that it finds in them, the majority infers from the Framers’ “democratic principles” that the Clauses must have been generally understood to preclude the people of the States and their state legislatures from prescribing any additional qualifications for their representatives in Congress. But the majority’s evidence on this point establishes only two more modest propositions: (1) the Framers did not want the Federal Constitution itself to impose a *875broad set of disqualifications for congressional office, and (2) the Framers did not want the Federal Congress to be able to supplement the few disqualifications that the Constitution does set forth. The logical conclusion is simply that the Framers did not want the people of the States and their state legislatures to be constrained by too many qualifications imposed at the national level. The evidence does not support the majority’s more sweeping conclusion that the Framers intended to bar the people of the States and their state legislatures from adopting additional eligibility requirements to help narrow their own choices.

I agree with the majority that Congress has no power to prescribe qualifications for its own Members. This fact, however, does not show that the Qualifications Clauses contain a hidden exclusivity provision. The reason for Congress’ incapacity is not that the Qualifications Clauses deprive Congress of the authority to set qualifications, but rather that nothing in the Constitution grants Congress this power. In the absence of such a grant, Congress may not act. But deciding whether the Constitution denies the qualification-setting power to the States and the people of the States requires a fundamentally different legal analysis.

Despite the majority’s claims to the contrary, see ante, at 796-797, n. 12, this explanation for Congress’ incapacity to supplement the Qualifications Clauses is perfectly consistent with the reasoning of Powell v. McCormack, 395 U. S. 486 (1969). Powell concerned the scope of Article I, § 5, which provides that “[e]ach House [of Congress] shall be the Judge of the Elections, Returns and Qualifications of its own Members.” As the majority itself recognizes, “[t]he principal issue [in Powell] was whether the power granted to each House in Art. I, § 5,... includes the power to impose qualifications other than those set forth in the text of the Constitution.” Ante, at 788. Contrary to the majority’s suggestion, then, the critical question in Powell was whether §5 conferred a qualification-setting power — not whether the Quali*876fications Clauses took it away. Compare Powell, supra, at 519 (describing the question before the Court as “what power the Constitution confers upon the House through Art. I, § 5”), and 536 (describing the Court’s task as “determining the meaning of Art. I, § 5”) with ante, at 789, and 792, n. 8 (suggesting that Powell held that the Qualifications Clauses “limit the power of the House to impose additional qualifications”). See also Buckley v. Valeo, 424 U. S. 1, 133 (1976) (taking my view of Powell).

Powell’s analysis confirms this point. After summarizing a large quantity of historical material bearing on the original understanding of what it meant for a legislature to act as “the Judge” of the qualifications of its members, see 395 U. S., at 521-531, Powell went on to stress that the Philadelphia Convention specifically rejected proposals to grant Congress the power to pass laws prescribing additional qualifications for its Members, and that the Convention rejected these proposals on the very same day that it approved the precursor of § 5. See id., at 533-536. Given this historical evidence, the Powell Court refused to read § 5 as empowering the House to prescribe such additional qualifications in its capacity as “Judge.” And if nothing in the Constitution gave the House this power, it inevitably followed that the House could not exercise it. Despite the majority’s claims, then, Powell itself rested on the proposition that the institutions of the Federal Government enjoy only the powers that are granted to them. See also ante, at 793, n. 9 (describing the Qualifications Clauses merely as an independent basis for the result reached in Powell)15

*877The fact that the Framers did not grant a qualification-setting power to Congress does not imply that they wanted to bar its exercise at the state level. One reason why the Framers decided not to let Congress prescribe the qualifications of its own Members was that incumbents could have used this power to perpetuate themselves or their ilk in office. As Madison pointed out at the Philadelphia Convention, Members of Congress would have an obvious conflict of interest if they could determine who may run against them. 2 Farrand 250; see also ante, at 793-794, n. 10. But neither the people of the States nor the state legislatures would labor under the same conflict of interest when prescribing qualifications for Members of Congress, and so the Framers would have had to use a different calculus in determining whether to deprive them of this power.

As the majority argues, democratic principles also contributed to the Framers’ decision to withhold the qualification-setting power from Congress. But the majority is wrong to suggest that the same principles must also have led the Framers to deny this power to the people of the States and the state legislatures. In particular, it simply is not true that “the source of the qualification is of little moment in assessing the qualification’s restrictive impact.” Ante, at 820. There is a world of difference between a self-imposed constraint and a constraint imposed from above.

Congressional power over qualifications would have enabled the representatives from some States, acting collectively in the National Legislature, to prevent the people of another State from electing their preferred candidates. The John Wilkes episode in 18th-century England illustrates the problems that might result. As the majority mentions, Wilkes’ district repeatedly elected him to the House of Commons, only to have a majority of the representatives of other *878districts frustrate their will by voting to exclude him. See ante, at 790. Americans who remembered these events might well have wanted to prevent the National Legislature from fettering the choices of the people of any individual State (for the House of Representatives) or their state legislators (for the Senate).

Yet this is simply to say that qualifications should not be set at the national level for offices whose occupants are selected at the state level. The majority never identifies the democratic principles that would have been violated if a state legislature, in the days before the Constitution was amended to provide for the direct election of Senators, had imposed some limits of its own on the field of candidates that it would consider for appointment.16 Likewise, the majority does not explain why democratic principles prohibit the people of a State from adopting additional eligibility requirements to help narrow their choices among candidates seeking to represent them in the House of Representatives. Indeed, the invocation of democratic principles to invalidate Amendment 73 seems particularly difficult in the present case, because Amendment 73 remains fully within the control of the people of Arkansas. If they wanted to repeal it (despite the 20-point margin by which they enacted it less than three years ago), they could do so by a simple majority vote. See Ark. Const., Arndt. 7.

The majority appears to believe that restrictions on eligibility for office are inherently undemocratic. But the Qualifications Clauses themselves prove that the Framers did not share this view; eligibility requirements to which the people of the States consent are perfectly consistent with the Fram*879ers’ scheme. In fact, we have described “the authority of the people of the States to determine the qualifications of their most important government officials” as “an authority that lies at the heart of representative government.” Gregory v. Ashcroft, 501 U. S. 452, 463 (1991) (internal quotation marks omitted) (refusing to read federal law to preclude States from imposing a mandatory retirement age on state judges who are subject to periodic retention elections). When the people of a State themselves decide to restrict the field of candidates whom they are willing to send to Washington as their representatives, they simply have not violated the principle that “the people should choose whom they please to govern them.” See 2 Elliot 257 (remarks of Alexander Hamilton at the New York Convention).

At one point, the majority suggests that the principle identified by Hamilton encompasses not only the electorate’s right to choose, but also “the egalitarian concept that the opportunity to be elected [is] open to all.” See ante, at 794; see also ante, at 819-820. To the extent that the second idea has any content independent of the first, the majority apparently would read the Qualifications Clauses to create a personal right to be a candidate for Congress, and then to set that right above the authority of the people of the States to prescribe eligibility requirements for public office. But we have never suggested that “the opportunity to be elected” is open even to those whom the voters have decided not to elect. On that rationale, a candidate might have a right to appear on the ballot in the general election even though he lost in the primary. But see Storer v. Brown, 415 U. S. 724, 726, n. 16 (1974); see also Bullock v. Carter, 405 U. S. 134, 142-143 (1972) (rejecting the proposition that there is any fundamental right to be a candidate, separate and apart from the electorate’s right to vote). Thus, the majority ultimately concedes that its “egalitarian concept” derives entirely from the electorate’s right to choose. See ante, at 794, n. 11; see also ante, at 819 (deriving the “egalitarian *880ideal” from the proposition that the Qualifications Clauses do not unduly “ ‘fetter the judgment... of the people’ ” (quoting The Federalist No. 57, at 351)). If the latter is not violated, then neither is the former.

In seeking ratification of the Constitution, James Madison did assert that “[ujnder these reasonable limitations [set out in the House Qualifications Clause], the door of this part of the federal government is open to merit of every description . . . .” The Federalist No. 52, at 326. The majority stresses this assertion, and others to the same effect, in support of its “egalitarian concept.” See ante, at 794, 819-820, and n. 30. But there is no reason to interpret these statements as anything more than claims that the Constitution itself imposes relatively few disqualifications for congressional office.17 One should not lightly assume that Madi*881son and his colleagues, who were attempting to win support at the state level for the new Constitution, were proclaiming the inability of the people of the States or their state legislatures to prescribe any eligibility requirements for their own Representatives or Senators. Instead, they were merely responding to the charge that the Constitution was undemocratic and would lead to aristocracies in office. Cf. ante, at 791 (referring to “the antifederalist charge that the new Constitution favored the wealthy and well born”). The statement that the qualifications imposed in the Constitution are not unduly restrictive hardly implies that the Constitution withdrew the power of the people of each State to prescribe additional eligibility requirements for their own Representatives if they so desired.

In fact, the authority to narrow the field of candidates in this way may be part and parcel of the right to elect Members of Congress. That is, the right to choose may include the right to winnow. See Hills, A Defense of State Constitutional Limits on Federal Congressional Terms, 53 U. Pitt. L. Rev. 97, 107-109 (1991).

To appreciate this point, it is useful to consider the Constitution as it existed before the Seventeenth Amendment was adopted in 1913. The Framers’ scheme called for the legislature of each State to choose the Senators from that State. Art. I, §3, cl. 1. The majority offers no reason to believe that state legislatures could not adopt prospective rules to guide themselves in carrying out this responsibility; not only is there no express language in the Constitution barring legislatures from passing laws to narrow their choices, but there also is absolutely no basis for inferring such a prohibition. Imagine the worst-case scenario: a state legislature, wishing *882to punish one of the Senators from its State for his vote on some bill, enacts a qualifications law that the Senator does not satisfy. The Senator would still be able to serve out his term; the Constitution provides for Senators to be chosen for 6-year terms, Art. I, §3, cl. 1, and a person who has been seated in Congress can be removed only if two-thirds of the Members of his House vote to expel him, § 5, cl. 2. While the Senator would be disqualified from seeking reappointment, under the Framers’ Constitution the state legislature already enjoyed unfettered discretion to deny him reappointment anyway. Instead of passing a qualifications law, the legislature could simply have passed a resolution declaring its intention to appoint someone else the next time around. Thus, the legislature’s power to adopt laws to narrow its own choices added nothing to its general appointment power.

While it is easier to coordinate a majority of state legislators than to coordinate a majority of qualified voters, the basic principle should be the same in both contexts. Just as the state legislature enjoyed virtually unfettered discretion over whom to appoint to the Senate under Art. I, §3, so the qualified voters of the State enjoyed virtually unfettered discretion over whom to elect to the House of Representatives under Art. I, § 2. If there is no reason to believe that the Framers’ Constitution barred state legislatures from adopting prospective rules to narrow their choices for Senator, then there is also no reason to believe that it barred the people of the States from adopting prospective rules to narrow their choices for Representative. In addition, there surely is no reason to believe that the Senate Qualifications Clause suddenly acquired an exclusivity provision in 1913, when the Seventeenth Amendment was adopted. Now that the people of the States are charged with choosing both Senators and Representatives, it follows that they may adopt eligibility requirements for Senators as well as for Representatives.

*883I would go further, for I see nothing in the Constitution that precludes the people of each State (if they so desire) from authorizing their elected state legislators to prescribe qualifications on their behalf. If the people of a State decide that they do not trust their state legislature with this power, they are free to amend their state constitution to withdraw it. This arrangement seems perfectly consistent with the Framers’ scheme. From the time of the framing until after the Civil War, for example, the Federal Constitution did not bar state governments from abridging the freedom of speech or the freedom of the press, even when those freedoms were being exercised in connection with congressional elections. It was the state constitutions that determined whether state governments could silence the supporters of disfavored congressional candidates, just as it was the state constitutions that determined whether the States could persecute people who held disfavored religious beliefs or could expropriate property without providing just compensation. It would not be at all odd if the state constitutions also determined whether the state legislature could pass qualifications statutes.

But one need not agree with me that the people of each State may delegate their qualification-setting power in order to uphold Arkansas’ Amendment 73. Amendment 73 is not the act of a state legislature; it is the act of the people of Arkansas, adopted at a direct election and inserted into the State Constitution. The majority never explains why giving effect to the people’s decision would violate the “democratic principles” that undergird the Constitution. Instead, the majority’s discussion of democratic principles is directed entirely to attacking eligibility requirements imposed on the people of a State by an entity other than themselves.

The majority protests that any distinction between the people of the States and the state legislatures is “untenable” and “astonishing.” See ante, at 809, n. 19. In the limited area of congressional elections, however, the Framers them*884selves drew this distinction: They specifically provided for Senators to be chosen by the state legislatures and for Representatives to be chosen by the people. In the context of congressional elections, the Framers obviously saw a meaningful difference between direct action by the people of each State and action by their state legislatures.

Thus, even if one believed that the Framers intended to bar state legislatures from adopting qualifications laws that restrict the people’s choices, it would not follow that the people themselves are precluded from agreeing upon eligibility requirements to help narrow their own choices. To be sure, if the Qualifications Clauses were exclusive, they would bar all additional qualifications, whether adopted by popular initiative or by statute. But the majority simply assumes that if state legislatures are barred from prescribing qualifications, it must be because the Qualifications Clauses are exclusive. It would strain the text of the Constitution far less to locate the bar in Article I, §2, and the Seventeenth Amendment instead: One could plausibly maintain that qualification requirements imposed by state legislatures violate the constitutional provisions entrusting the selection of Members of Congress to the people of the States, even while one acknowledges that qualification requirements imposed by the people themselves are perfectly constitutional. The majority never justifies its conclusion that “democratic principles” require it to reject even this intermediate position.

C

In addition to its arguments about democratic principles, the majority asserts that more specific historical evidence supports its view that the Framers did not intend to permit supplementation of the Qualifications Clauses. But when one focuses on the distinction between congressional power to add qualifications for congressional office and the power of the people or their state legislatures to add such qualifications, one realizes that this assertion has little basis.

*885In particular, the detail with which the majority recites the historical evidence set forth in Powell v. McCormack, 395 U. S. 486 (1969), should not obscure the fact that this evidence has no bearing on the question now before the Court. As the majority ultimately concedes, see ante, at 792-793, 796, 798, it does not establish “the Framers’ intent that the qualifications in the Constitution be fixed and exclusive,” ante, at 790; it shows only that the Framers did not intend Congress to be able to enact qualifications laws.18 If any*886thing, the solidity of the evidence supporting Powell’s view that Congress lacks the power to supplement the constitutional disqualifications merely highlights the weakness of the majority’s evidence that the States and the people of the States also lack this power.

1

To the extent that the records from the Philadelphia Convention itself shed light on this case, they tend to hurt the majority’s case. The only evidence that directly bears on the question now before the Court comes from the Committee of Detail, a five-member body that the Convention charged with the crucial task of drafting a Constitution to reflect the decisions that the Convention had reached during its first two months of work. A document that Max Farrand described as “[a]n early, perhaps the first, draft of the committee’s work” survived among the papers of George Mason. 1 Farrand xxiii, n. 36. The draft is in the handwriting of *887Edmund Randolph, the chairman of the Committee, with emendations in the hand of John Rutledge, another member of the Committee. As Professor Farrand noted, “[e]ach item in this document ... is either checked off or crossed out, showing that it was used in the preparation of subsequent drafts.” 2 id., at 137, n. 6; see also W. Meigs, The Growth of the Constitution in the Federal Convention of 1787, pp. I-IX (1900) (providing a facsimile of the document).

The document is an extensive outline of the Constitution. Its treatment of the National Legislature is divided, into two parts, one for the “House of Delegates” and one for the Senate. The Qualifications Clause for the House of Delegates originally read as follows: “The qualifications of a delegate shall be the age of twenty five years at least, and citizenship: and any person possessing these qualifications may be elected except [blank space].” Id., at II (emphasis added). The drafter(s) of this language apparently contemplated that the Committee might want to insert some exceptions to the exclusivity provision. But rather than simply deleting the word “except” — as it might have done if it had decided to have no exceptions at all to the exclusivity provision — the Committee deleted the exclusivity provision itself. In the document that has come down to us, all the words after the colon are crossed out. Ibid.

The majority speculates that the exclusivity provision may have been deleted as superfluous. See ante, at 815-816, n. 27.19 But the same draft that contained the exclusivity language in the House Qualifications Clause contained no *888such language in the Senate Qualifications Clause. See 2 Farrand 141. Thus, the draft appears to reflect a deliberate judgment to distinguish between the House qualifications and the Senate qualifications, and to make only the former exclusive. If so, then the deletion of the exclusivity provision indicates that the Committee expected neither list of qualifications to be exclusive.

The majority responds that the absence of any exclusivity provision in the Committee’s draft of the Senate Qualifications Clause merely reflected the fact that “senators, unlike Representatives, would not be chosen by popular election.” Ante, at 815, n. 27. I am perfectly prepared to accept this explanation: The drafter(s) may well have thought that state legislatures should be prohibited from constricting the people’s choices for the House of Representatives, but that no exclusivity provision was necessary on the Senate side because state legislatures would already have unfettered control over the appointment of Senators. To accept this explanation, however, is to acknowledge that the exclusivity provision in the Committee’s draft of the House Qualifications Clause was not thought to be mere surplusage. It is also to acknowledge that the Senate Qualifications Clause in the Committee’s draft — “the qualification of a senator shall be the age of 25 years at least: citizenship in the united states: and property to the amount of [blank space],” 2 Farrand 141 — did not carry any implicit connotation of exclusivity. In short, the majority’s own explanation for the difference between the two Qualifications Clauses in the Committee’s draft is fundamentally at odds with the expressio unius argument on which the majority rests its holding.

2

Unable to glean from the Philadelphia Convention any direct evidence that helps its position, the majority seeks signs of the Framers’ unstated intent in the Framers’ comments about four other constitutional provisions. See ante, at 808-*889811 (citing Art. I, §2, cl. 1; §4, cl. 1; §5, cl. 1; and §6, cl. 1). The majority infers from these provisions that the Framers wanted “to minimize the possibility of state interference with federal elections.” Ante, at 808. But even if the majority’s reading of its evidence were correct, the most that one could infer is that the Framers did not want state legislatures to be able to prescribe qualifications that would narrow the people’s choices. See supra, at 883-888. However wary the Framers might have been of permitting state legislatures to exercise such power, there is absolutely no reason to believe that the Framers feared letting the people themselves exercise this power. Cf. The Federalist No. 52, at 326 (Madison) (“It cannot be feared that the people of the States will alter this [electoral-qualification] part of their constitutions in such a manner as to abridge the rights secured to them by the federal Constitution”).

In any event, none of the provisions cited by the majority is inconsistent with state power to add qualifications for congressional office. First, the majority cites the constitutional requirement that congressional salaries be “ascertained by Law, and paid out of the Treasury of the United States.” Art. I, § 6, cl. 1. Like the Qualifications Clauses themselves, however, the salary provision can be seen as simply another means of protecting the competence of the National Legislature. As reflected in the majority’s own evidence, see ante, at 809-810; see also 1 Farrand 373 (remarks of James Madison), one of the recurring themes of the debate over this provision was that if congressional compensation were left up to the States, parsimonious States might reduce salaries so low that only incapable people would be willing to serve in Congress.

As the majority stresses, some delegates to the Philadelphia Convention did argue that leaving congressional compensation up to the various States would give Members of Congress “an improper dependence” upon the States. Id., at 216 (remarks of James Madison); ante, at 809-810. These *890delegates presumably did not want state legislatures to be able to tell the Members of Congress from their State, “Vote against Bill A or we will slash your salary”; such a power would approximate a power of recall, which the Framers denied to the States when they specified the terms of Members of Congress. The Framers may well have thought that state power over salary, like state power to recall, would be inconsistent with the notion that Congress was a national legislature once it assembled. But state power over initial eligibility requirements does not raise the same concerns: It was perfectly coherent for the Framers to leave selection matters to the state level while providing for Members of Congress to draw a federal salary once they took office. Thus, the Compensation Clause seems wholly irrelevant; contrary to the majority’s suggestion, see ante, at 811, n. 21, it does not address elections at all.

Second, the majority gives passing mention to the Elector-Qualifications Clause of Article I, § 2, which specifies that in each State, the voters in House elections “shall have the qualifications requisite for Electors of the most numerous Branch of the State Legislature.” But the records of the Philadelphia Convention provide no evidence for the majority’s assertion that the purpose of this Clause was “to prevent discrimination against federal electors.” See ante, at 808.20 *891In fact, the Clause may simply have been a natural concomitant of one of the Framers’ most famous decisions. At the Convention, there was considerable debate about whether Members of the House of Representatives should be selected by the state legislatures or directly by the voters of each State. Taken as a whole, the first Clause of Article I, § 2— including the elector-qualifications provision — implements the Framers’ decision. It specifies that the Representatives from each State are to be chosen by the State’s voters (that is, the people eligible to participate in elections for the most numerous branch of the state legislature).

Third, the majority emphasizes that under Article I, §5, “[e]ach House [of Congress] shall be the Judge of the Elections, Returns and Qualifications of its own Members.” See ante, at 804, 811, 822. There was no recorded discussion of this provision in the Philadelphia Convention, and it appears simply to adopt the practice of England’s Parliament. See n. 18, supra. According to the majority, however, § 5 implies *892that the Framers could not have intended state law ever to “provide the standard for judging a Member’s eligibility.” Ante, at 812.

My conclusion that States may prescribe eligibility requirements for their Members of Congress does not necessarily mean that the term “Qualifications,” as used in Article I, § 5, includes such state-imposed requirements. One surely could read the term simply to refer back to the requirements that the Framers had just listed in the Qualifications Clauses, and not to encompass whatever requirements States might add on their own. See Nixon v. United States, 506 U. S. 224, 237 (1993) (dictum) (asserting that the context of § 5 demonstrates that “the word ‘[qualifications’ . . . was of a precise, limited nature” and referred only to the qualifications previously “set forth in Art. I, § 2”). The Framers had deemed the constitutional qualifications essential to protect the competence of Congress, and hence the national interest. It is quite plausible that the Framers would have wanted each House to make sure that its Members possessed these qualifications, but would have left it to the States to enforce whatever qualifications were imposed at the state level to protect state interests.

But even if this understanding of §5 is incorrect, I see nothing odd in the notion that a House of Congress might have to consider state law in judging the “Qualifications” of its Members. In fact, §5 itself refutes the majority’s argument. Because it generally is state law that determines what is necessary to win an election and whether any particular ballot is valid, each House of Congress clearly must look to state law in judging the “Elections” and “Returns” of its Members. It would hardly be strange if each House had to do precisely the same thing in judging “Qualifications.” Indeed, even on the majority’s understanding of the Constitution, at the time of the framing all “Qualifications” questions that turned on issues of citizenship would have been governed by state, law. See supra, at 872-873.

*893More generally, there is no basis for the majority’s assertion that the Framers would not have charged “federal tribunals” with the task of “judging . . . questions concerning rights which depend on state law.” See ante, at 812. Cases involving questions of federal law hardly exhaust the categories of cases that the Framers authorized the federal courts to decide. See Art. Ill, §2, cl. 1. The founding generation, moreover, seemed to assign relatively little importance to the constitutional grant of jurisdiction over “all Cases . . . arising under this Constitution, the Laws of the United States, and Treaties made ... under their Authority.” Ibid. The First Congress never even implemented this jurisdictional grant at the trial level; it was not until 1875 that Congress “revolutionized the concept of the federal judiciary” by giving federal courts broad jurisdiction over suits arising under federal law. See P. Bator, D. Meltzer, P. Mishkin, & D. Shapiro, Hart and Wechsler’s The Federal Courts and the Federal System 962 (3d ed. 1988). By contrast, the founding generation thought it important to implement immediately the constitutional grant of diversity jurisdiction, in which the rules of decision generally come entirely from state law. See Judiciary Act of 1789, 1 Stat. 73, 78, 92; Erie R. Co. v. Tompkins, 304 U. S. 64, 77-80 (1938).

The fourth and final provision relied upon by the majority is the Clause giving Congress the power to override state regulations of “[t]he Times, Places and Manner of holding [congressional] Elections.” Art. I, §4, cl. 1. From the fact that the Framers gave Congress the power to “make or alter” these state rules of election procedure, the majority infers that the Framers would also have wanted Congress to enjoy override authority with respect to any matters of substance that were left to the States. See ante, at 810-811. As Congress enjoys no “make or alter” powers in this area, the majority concludes that the Framers must not have thought that state legislatures would be able to enact qualifications laws.

*894But the Framers provided for congressional override only where they trusted Congress more than the States. Even respondents acknowledge that “the primary reason” for the “make or alter” power was to enable Congress to ensure that States held elections in the first place. See Tr. of Oral Arg. 51; see also supra, at 863, and n. 10. The Framers did trust Congress more than the States when it came to preserving the Federal Government’s own existence; to advance this interest, they had to give Congress the capacity to prescribe both the date and the mechanics of congressional elections. As discussed above, however, the Framers trusted the States more than Congress when it came to setting qualifications for Members of Congress. See supra, at 877. Indeed, the majority itself accepts this proposition. See ante, at 832 (acknowledging that the Framers were “particularly concerned” about congressional power to set qualifications).

To judge from comments made at the state ratifying conventions, Congress’ “make or alter” power was designed to serve a coordination function in addition to ensuring that the States had at least rudimentary election laws. For instance, George Nicholas argued at the Virginia Convention that if regulation of the time of congressional elections had been left exclusively to the States, “there might have been as many times of choosing as there are States,” and “such intervals might elapse between the first and last election, as to prevent there being a sufficient number to form a House.” 9 Documentary History of the Ratification of the Constitution 920 (J. Kaminski and G. Saladino eds. 1990). For this reason too, if the National Legislature lacked the “make or alter” power, “it might happen that there should be no Congress^] . . . and this might happen at a time when the most urgent business rendered their session necessary.” Ibid.; cf. 2 Elliot 535 (remarks of Thomas McKean at the Pennsylvania ratifying convention) (defending §4 on the ground that congressional elections should be “held on the same day throughout the United States, to prevent corruption or *895undue influence”). Again, however, the desire to coordinate state election procedures did not require giving Congress power over qualifications laws.

The structure of the Constitution also undermines the majority’s suggestion that it would have been bizarre for the Framers to give Congress supervisory authority over state time, place, and manner regulations but not over state qualifications laws. Although the Constitution does set forth a few nationwide disqualifications for the office of Presidential elector, see Art. II, § 1, cl. 2 (“no Senator or Representative, or Person holding an Office of Trust or Profit under the United States, shall be appointed an Elector”), no one contends that these disqualifications implicitly prohibit the States from adding any other eligibility requirements; instead, Article II leaves the States free to establish qualifications for their delegates to the electoral college. See supra, at 861-862. Nothing in the Constitution, moreover, gives Congress any say over the additional eligibility requirements that the people of the States or their state legislatures may choose to set. Yet under Article II, “[t]he Congress may determine the Time of chusing the Electors . . . .” Art. II, § 1, cl. 4.

The majority thus creates an unwarranted divergence between Article I’s provisions for the selection of Members of Congress and Article II’s provisions for the selection of members of the electoral college. Properly understood, the treatment of congressional elections in Article I parallels the treatment of Presidential elections in Article II. Under Article I as under Article II, the States and the people of the States do enjoy the reserved power to establish substantive eligibility requirements for candidates, and Congress has no power to override these requirements. But just as Article II authorizes Congress to prescribe when the States must select their Presidential electors, so Article I gives Congress the ultimate authority over the times, places, and manner of holding congressional elections.

*896The majority’s only response is that my reading of the Constitution would permit States to use their qualification-setting power to achieve the very result that Congress’ “make or alter” power was designed to avoid. According to the majority, States could set qualifications so high that no candidate could meet them, and Congress would be powerless to do anything about it. Ante, at 811.

Even if the majority were correct that Congress could not nullify impossible qualifications, however, the Constitution itself proscribes such state laws. The majority surely would concede that under the Framers’ Constitution, each state legislature had an affirmative duty to appoint two people to the Senate. See Art. I, §3, cl. 1 (“The Senate of the United States shall be composed of two Senators from each State, chosen by the Legislature thereof ...” (emphasis added)); cf. Art. I, §3, cl. 2 (“if Vacancies happen by Resignation, or otherwise, during the Recess of the Legislature of any State, the Executive thereof may make temporary Appointments until the next Meeting of the Legislature, which shall then fill such Vacancies”). In exactly the same way that §3 requires the States to send people to the Senate, § 2 also requires the States to send people to the House. See Art. I, § 2, cl. 1 (“The House of Representatives shall be composed of Members chosen every second Year by the People of the several States . . .”); cf. Art. I, §2, cl. 4 (“When vacancies happen in the Representation from any State, the Executive Authority thereof shall issue Writs of Election to fill such Vacancies”).

The majority apparently is concerned that (on its reading of the “make or alter” power) Congress would not be able to enforce the constitutional proscription on impossible qualifications; enforcement would instead be relegated to the courts, the Executive Branch, or the political process. But this concern is equally applicable whether one adopts my view of the Qualifications Clauses or the majority’s view. Both the majority and I agree that it is unconstitutional for *897States to establish impossible qualifications for congressional office. Both the majority and I also agree that it is theoretically conceivable that a State might defy this proscription by erecting an impossible qualification. Whether Congress may use its “make or alter” power to override such laws turns entirely on how one reads the “make or alter” power; it has nothing to do with whether one believes that the Qualifications Clauses are exclusive.

It would not necessarily be unusual if the Framers had decided against using Congress’ “make or alter” power to guard against state laws that disqualify everyone from service in the House. After all, although this power extended to the times and manner of selecting Senators as well as Representatives, it did not authorize Congress to pick the Senators from a State whose legislature defied its constitutional obligations and refused to appoint anyone. This does not mean that the States had no duty to appoint Senators, or that the States retained the power to destroy the Federal Government by the simple expedient of refusing to meet this duty. It merely means that the Framers did not place the remedy with Congress.21

But the flaws in the majority’s argument go deeper. Contrary to the majority’s basic premise, Congress can nullify state laws that establish impossible qualifications. If a State actually holds an election and only afterwards purports to disqualify the winner for failure to meet an impossible condition, Congress certainly would not be bound by the purported disqualification. It is up to each House of Congress to judge the “[qualifications” of its Members for itself. See Art. I, § 5, cl. 1. Even if this task includes the responsibility of judging qualifications imposed by state law, see supra, at 892-893, Congress obviously would have not only *898the power but the duty to treat the unconstitutional state law as a nullity. Thus, Congress could provide the appropriate remedy for the State’s defiance, simply by seating the winner of the election.

It follows that the situation feared by the majority would arise only if the State refused to hold an election in the first place, on the ground that no candidate could meet the impossible qualification. But Congress unquestionably has the power to override such a refusal. Under the plain terms of § 4, Congress can make a regulation providing for the State to hold a congressional election at a particular time and place, and in a particular manner.22

3

In discussing the ratification period, the majority stresses two principal data. One of these pieces of evidence is no evidence at all — literally. The majority devotes considerable space to the fact that the recorded ratification debates do not contain any affirmative statement that the States can supplement the constitutional qualifications. See ante, at 812-815. For the majority, this void is “compelling” evidence that “unquestionably reflects the Framers’ common understanding that States lacked that power.” Ante, at 812, 814. The majority reasons that delegates at several of the ratifying conventions attacked the Constitution for failing to require Members of Congress to rotate out of office.23 If *899supporters of ratification had believed that the individual States could supplement the constitutional qualifications, the majority argues, they would have blunted these attacks by pointing out that rotation requirements could still be added State by State. See ante, at 814.

But the majority’s argument cuts both ways. The recorded ratification debates also contain no affirmative statement that the States cannot supplement the constitutional qualifications. While ratification was being debated, the existing rule in America was that the States could prescribe eligibility requirements for their delegates to Congress, see n. 3, swpra, even though the Articles of Confederation gave Congress itself no power to impose such qualifications. If *900the Federal Constitution had been understood to deprive the States of this significant power, one might well have expected its opponents to seize on this point in arguing against ratification.

The fact is that arguments based on the absence of recorded debate at the ratification conventions are suspect, because the surviving records of those debates are fragmentary. We have no records at all of the debates in several of the conventions, 3 Documentary History of the Ratification of the Constitution 7 (M. Jensen ed. 1978), and only spotty records from most of the others, see ibid.; 1 id., at 34-35; 4 Elliot 342; Hutson, The Creation of the Constitution: The Integrity of the Documentary Record, 65 Texas L. Rev. 1, 21-23 (1986).

If one concedes that the absence of relevant records from the ratification debates is not strong evidence for either side, then the majority’s only significant piece of evidence from the ratification period is The Federalist No. 52. Contrary to the majority’s assertion, however, this essay simply does not talk about “the lack of state control over the qualifications of the elected,” whether “explicitly” or otherwise. See ante, at 806.

It is true that The Federalist No. 52 contrasts the Constitution’s treatment of the qualifications of voters in elections for the House of Representatives with its treatment of the qualifications of the Representatives themselves. As Madison noted, the Framers did not specify any uniform qualifications for the franchise in the Constitution; instead, they simply incorporated each State’s rules about eligibility to vote in elections for the most numerous branch of the state legislature. By contrast, Madison continued, the Framers chose to impose some particular qualifications that all Members of the House had to satisfy. But while Madison did say that the qualifications of the elected were “more susceptible of uniformity” than the qualifications of electors, The Federalist No. 52, at 326, he did not say that the Constitution *901prescribes anything but uniform minimum qualifications for congressmen. That, after all, is more than it does for congressional electors.

Nor do I see any reason to infer from The Federalist No. 52 that the Framers intended to deprive the States of the power to add to these minimum qualifications. Madison did note that the existing state constitutions defined the qualifications of “the elected” — a phrase that the essay used to refer to Members of Congress — “less carefully and properly” than they defined the qualifications of voters. But Madison could not possibly have been rebuking the States for setting unduly high qualifications for their representatives in Congress, because they actually had established only the sketchiest of qualifications. At the time that Madison wrote, the various state constitutions generally provided for the state legislature to appoint the State’s delegates to the Federal Congress.24 Four State Constitutions had added a term-limits provision that tracked the one in the Articles of Confederation,25 and some of the Constitutions also specified that people who held certain salaried offices under the United States were ineligible to represent the State in Congress.26 But only two State Constitutions had prescribed any other *902qualifications for delegates to Congress.27 In this context, when Madison wrote that the state constitutions defined the qualifications of Members of Congress “less carefully and properly” than they defined the qualifications of voters, he could only have meant that the existing state qualifications did not do enough to safeguard Congress’ competence: The state constitutions had not adopted the age, citizenship, and inhabitancy requirements that the Framers considered essential. Madison’s comments readily explain why the Framers did not merely incorporate the state qualifications for Congress. But they do not imply that the Framers intended to withdraw from the States the power to supplement the list of qualifications contained in the Federal Constitution.28

Though The Federalist No. 52 did not address this question, one might wonder why the Qualifications Clauses did not simply incorporate the existing qualifications for members of the state legislatures (as opposed to delegates to Congress). Again, however, the Framers’ failure to do so cannot be taken as an implicit criticism of the States for setting unduly high entrance barriers. To the contrary, the age and citizenship qualifications set out in the Federal Constitution are considerably higher than the corresponding qualifications contained in the state constitutions that were then in force. At the time, no state constitution required members of the lower house of the state legislature to be more than 21 years old, and only two required members of the upper house to be 30. See N. H. Const. of 1784, Pt. II, in 4 Thorpe 2460; S. C. Const. of 1778, Art. XII, in 6 Thorpe 3250. Many *903States, moreover, permitted naturalized aliens to take seats in the state legislature within one or two years of becoming citizens. See Kettner, Development of American Citizenship, at 214-219.

The majority responds that at the time of the framing, most States imposed property qualifications on members of the state legislature. See ante, at 807-808, n. 18. But the fact that the Framers did not believe that a uniform minimum property requirement was necessary to protect the competence of Congress surely need not mean that the Framers intended to preclude States from setting their own property qualifications.

In fact, the constitutional text supports the contrary inference. As the majority observes, see ibid,., and ante, at 825, n. 35, at the time of the framing some States also imposed religious qualifications on state legislators. The Framers evidently did not want States to impose such qualifications on federal legislators, for the Constitution specifically provides that “no religious Test shall ever be required as a Qualification to any Office or public Trust under the United States.” Art. VI, cl. 3. Both the context29 and the plain language of the Clause show that it bars the States as well as the Federal Government from imposing religious disqualifications on federal offices. But the only reason for extending the Clause to the States would be to protect Senators and Representatives from state-imposed religious qualifications; I know of no one else who holds a “public Trust under the United States” yet who might be subject to state disqualifications. If the expressw unius maxim cuts in any direction in this case, then, it undermines the majority’s position: The Framers’ prohibition on state-imposed religious disqual*904ifications for Members of Congress suggests that other types of state-imposed disqualifications are permissible. See Rotunda, Rethinking Term Limits for Federal Legislators in Light of the Structure of the Constitution, 73 Ore. L. Rev. 561, 574 (1994).

4

More than a century ago, this Court was asked to invalidate a Michigan election law because it called for Presidential electors to be elected on a district-by-district basis rather than being chosen by “the State” as a whole. See Art. II, § 1, cl. 2. Conceding that the Constitution might be ambiguous on this score, the Court asserted that “where there is ambiguity or doubt, or where two views may well be entertained, contemporaneous and subsequent practical construction^] are entitled to the greatest weight.” McPherson v. Blacker, 146 U. S., at 27. The Court then described the district-based selection processes used in 2 of the 10 States that participated in the first Presidential election in 1788,3 of the 15 States that participated in 1792, and 5 of the 16 States that participated in 1796. Id., at 29-31. Though acknowledging that in subsequent years “most of the States adopted the general ticket system,” id., at 32, the Court nonetheless found this history “decisive” proof of the constitutionality of the district method, id., at 36. Thus, the Court resolved its doubts in favor of the state law, “the contemporaneous practical exposition of the Constitution being too strong and obstinate to be shaken . . ..” Id., at 27.

Here, too, state practice immediately after the ratification of the Constitution refutes the majority’s suggestion that the Qualifications Clauses were commonly understood as being exclusive. Five States supplemented the constitutional disqualifications in their very first election laws, and the surviving records suggest that the legislatures of these States considered and rejected the interpretation of the Constitution that the majority adopts today.

*905As the majority concedes, the first Virginia election law erected a property qualification for Virginia’s contingent in the Federal House of Representatives. See Virginia Election Law (Nov. 20, 1788), in 2 Documentary History of the First Federal Elections, 1788-1790, pp. 293, 294 (G. DenBoer ed. 1984) (hereinafter First Federal Elections) (restricting possible candidates to “freeholder[s]”). What is more, while the Constitution merely requires representatives to be inhabitants of their State, the legislatures of five of the seven States that divided themselves into districts for House elections 30 added that representatives also had to be inhabitants of the district that elected them. Three of these States adopted durational residency requirements too, insisting that representatives have resided within their districts for at least a year (or, in one case, three years) before being elected.31

*906In an attempt to neutralize the significance of the district residency requirements, respondent Hill asserts that “there is no evidence that any state legislature focused, when it created these requirements, on the fact that it was adding to the constitutional qualifications.” Brief for Respondents Bobbie E. Hill et al. 20. But this claim is simply false.

In Massachusetts, for instance, the legislature charged a committee with drafting a report on election methods. The fourth article of the resulting report called for the State to be divided into eight districts that would each elect one representative, but did not require that the representatives be residents of the districts that elected them. Joint Committee Report (Nov. 4, 1788), in 1 First Federal Elections 481. When the members of the State House of Representatives discussed this report, those who proposed adding a district residency requirement were met with the claim that the Federal Constitution barred the legislature from specifying additional qualifications. See Massachusetts Centinel (Nov. 8, 1788) (reporting proceedings), in 1 First Federal Elections 489. After “considerable debate,” the House approved the committee’s version of the fourth article by a vote of 89 to 72. Ibid. But the State Senate approved a district residency amendment, 1 First Federal Elections 502, and the House then voted to retain it, id., at 504.

Although we have no record of the legislative debates over Virginia’s election law, a letter written by one of the members of the House of Delegates during the relevant period indicates that in that State, too, the legislature considered the possible constitutional objection to additional disqualifications. In that letter, Edward Carrington (an opponent of the district residency requirement) expressed his view that the requirement “may exceed the powers of the Assembly,” *907but acknowledged that there was “no prospect of its being struck out” because Federalists as well as Anti-Federalists at least professed to “think it right.” 2 id., at 367 (letter from Carrington to Madison, Nov. 9-10, 1788). Carrington was correct about the views of his colleagues: By a vote of 80 to 32, the House of Delegates rejected a motion to delete the added qualifications, while a similar motion in the State Senate lost by a vote of 12 to 3. Id., at 287, 293.32

The surviving records from Maryland and Georgia are less informative, but they, too, show that the legislatures of those States gave special attention to the district residency requirements that they enacted.33 Out of the five original *908States that adopted district residency requirements, in fact, only in North Carolina were the records so poor that it is impossible to draw any inferences about whether the legislature gave careful attention to the implications of the requirement.34

*909The majority asserts that “state practice with respect to residency requirements does not necessarily indicate that States believed that they had a broad power to add restrictions,” because the States “may simply have viewed district residency requirements as the necessary analog to state residency requirements.” Ante, at 827, n. 41. This argument fails even on its own terms. If the States had considered district residency requirements necessary for the success of a district election system, but had agreed with the majority that the Constitution prohibited them from supplementing the constitutional list of qualifications, then they simply would have rejected the district system and used statewide elections. After all, the majority deems district residency requirements just as unconstitutional as other added qualifications. See ante, at 799.

The majority’s argument also fails to account for the durational element of the residency requirements adopted in Georgia, North Carolina, and Virginia (and soon thereafter in Tennessee). These States obliged Congressmen not only to be district residents when elected but also to have been district residents for at least a year before then. See n. 31, supra.

Finally, the majority’s argument cannot explain the election schemes of Maryland and Georgia. Though these States did divide themselves into congressional districts, they allowed every voter to vote for one candidate from each *910district. See Georgia Election Law (Jan. 23, 1789), in 2 First Federal Elections 456, 457; Maryland Election Law (Dec. 22, 1788), in 2 First Federal Elections 136, 138. In other words, Maryland and Georgia imposed district residency requirements despite permitting every voter in the State to vote for every representative from the State. Neither of these States could possibly have seen district residency requirements as the “necessary analog” to anything; they imposed these requirements solely for their own sake.

The majority nonetheless suggests that the initial election laws adopted by the States actually support its position because the States did not enact very many disqualifications. See ante, at 826-827, n. 41. In this context, the majority alludes to the fact that no State imposed a religious qualification on federal legislators, even though New Hampshire continued to require state legislators to be Protestants and North Carolina imposed a similar requirement on people holding places of trust in the State’s “civil department.” See ante, at 826-827, n. 41, and 825, n. 35. But the majority concedes that “Article VI of the Federal Constitution . . . prohibited States from imposing similar qualifications on federal legislators.” Ante, at 825, n. 35. As discussed above, the constitutional treatment of religious qualifications tends to undermine rather than support the majority’s case. See supra, at 903-904.

The majority also points out that no State required its own federal representatives to rotate out of office after serving one or more terms. Ante, at 826. At the time of the framing, however, such requirements were increasingly disfavored on policy grounds. The advantages of incumbency were substantially fewer then than now, and turnover in office was naturally quite high. The perceived advantages of term limits were therefore smaller than they are today. But the perceived disadvantages were just as great: Term limits prevented the States or the people of the States from keeping good legislators in office, even if they wanted to do so. *911See G. Wood, Creation of the American Republic, 1776-1787, p. 439 (1969).

It is true that under the Articles of Confederation, four States had imposed term limits on their delegates to Congress. See ante, at 826. But three of these provisions added nothing to the limits in the Articles themselves, see Md. Const. of 1776, Form of Government, Art. XXVII (echoing Article of Confederation V), in 3 Thorpe 1695; N. H. Const. of 1784, Pt. II (same), in 4 Thorpe 2467; N. C. Const. of 1776, Art. XXXVII (similar), in 5 Thorpe 2793, and the other one contained only a minor variation on the provision in the Articles, see Pa. Const. of 1776, Frame of Government, §11, in 5 Thorpe 3085. Indeed, though the majority says that “many States imposed term limits on state officers,” ante, at 825-826, it appears that at the time of the framing only Pennsylvania imposed any restriction on the reelection of members of the state legislature, and Pennsylvania deleted this restriction when it adopted a new Constitution in 1790. Compare Pa. Const. of 1776, Frame of Government, §8, in 5 Thorpe 3084, with Pa. Const. of 1790, in 5 Thorpe 3092-3103; cf. Va. Const. of 1776, Form of Government (perhaps imposing term limits on members of the upper house of the state legislature), in 7 Thorpe 3816. It seems likely, then, that the failure of any State to impose term limits on its senators and representatives simply reflected policy-based decisions against such restrictions.

The majority counters that the delegates at three state ratifying conventions — in Virginia, New York, and North Carolina — “proposed amendments that would have required rotation.” Ante, at 813; cf. ante, at 826, and n. 40. But the amendments proposed by both the North Carolina Convention and the Virginia Convention would have imposed term limits only on the President, not on Members of Congress. See 4 Elliot 245 (North Carolina) (“[N]o person shall be capable of being President of the United States for more than eight years in any term of fifteen years”); 3 id,., at 660 *912(Virginia) (similar). If the majority is correct that these conventions also “voiced support for term limits for Members of Congress,” see ante, at 826,35 then the evidence from these conventions supports my position rather than the majority’s: the conventions deemed it necessary for the Constitution itself to impose term limits on the President (because no State could do that on its own), but they did not think it necessary for the Constitution to impose term limits on Members of Congress. This understanding at the Virginia and North Carolina conventions meshes with the election laws adopted by both States, which reflected the view that States could supplement the Qualifications Clauses. See supra, at 905, and n. 31, 909.36

*913If the majority can draw no support from state treatment of religious qualifications and rotation requirements, we are left only with state treatment of property qualifications. It is true that nine of the State Constitutions in effect at the time of the framing required members of the lower house of the state legislature to possess some property, see ante, at 823-824, n. 33, and that four of these Constitutions were revised shortly after the framing but continued to impose such requirements, see ante, at 824-825, and n. 35. Only one State, by contrast, established a property qualification for the Federal House of Representatives. But the fact that more States did not adopt congressional property qualifications does not mean that the Qualifications Clauses were commonly understood to be exclusive; there are a host of other explanations for the relative liberality of state election laws.37 And whatever the explanation, the fact remains that *914five of the election laws enacted immediately after ratification of the Constitution imposed additional qualifications that would clearly be unconstitutional under today’s holding. This history of state practice — which is every bit as strong as the history we deemed “decisive” in McPherson v. Blacker, 146 U. S., at 36 — refutes the majority’s position that the Qualifications Clauses were generally understood to include an unstated exclusivity provision.

5

The same is true of the final category of historical evidence discussed by the majority: controversies in the House and the Senate over seating candidates who were duly elected but who arguably failed to satisfy qualifications imposed by state, law.

*915As the majority concedes, “‘congressional practice has been erratic’ ” and is of limited relevance anyway. Ante, at 819 (quoting Powell v. McCormack, 395 U. S., at 545). Actions taken by a single House of Congress in 1887 or in 1964 shed little light on the original understanding of the Constitution. Presumably for that reason, the majority puts its chief emphasis on the 1807 debate in the House of Representatives about whether to seat Maryland’s William Mc-Creery. See ante, at 816-818. I agree with the majority that this debate might lend some support to the majority’s position if it had transpired as reported in Powell v. McCormack. See ante, at 816-817. But the Court’s discussion— both in Powell and today — is misleading.

A Maryland statute dating from 1802 had created a district entitled to send two representatives to the House, one of whom had to be a resident of Baltimore County and the other of whom had to be a resident of Baltimore City. McCreery was elected to the Ninth Congress as a resident of Baltimore City. After his reelection to the Tenth Congress, however, his qualifications were challenged on the ground that because he divided his time between his summer estate in Baltimore County and his residence in Washington, D. C., he was no longer a resident of Baltimore City at all.

As the majority notes, a report of the House Committee of Elections recommended that McCreery be seated on the ground that state legislatures have no authority to add to the qualifications set forth in the Constitution. See 17 Annals of Cong. 871 (1807); ante, at 816-817. But the committee’s submission of this initial report sparked a heated debate that spanned four days, with many speeches on both sides of the issue. See 17 Annals of Cong. 871-919, 927-947 (reporting proceedings from Nov. 12, 13, 16, and 18, 1807). Finally, a large majority of the House voted to recommit the report to the Committee of Elections. Id., at 950 (Nov. 19, 1807). The committee thereupon deleted all references to the *916constitutional issue and issued a revised report that focused entirely on the factual question whether McCreery satisfied the state residency requirement. Id., at 1059-1061 (Dec. 7, 1807). After receiving the new report, the House seated McCreery with a resolution simply saying: “Resolved, That William McCreery is entitled to his seat in this House.” Id., at 1237 (Dec. 24, 1807). By overwhelming majorities, the House rejected both a proposal to specify that McCreery possessed “the qualifications required by the law of Maryland,” ibid., and a proposal to declare only that he was “duly qualified, agreeably to the constitution of the United States,” id., at 1231. Far from supporting the majority’s position, the McCreery episode merely demonstrates that the 10th House of Representatives was deeply divided over whether state legislatures may add to the qualifications set forth in the Constitution.38

The majority needs more than that. The prohibition that today’s majority enforces is found nowhere in the text of the Qualifications Clauses. In the absence of evidence that the Clauses nonetheless were generally understood at the time of the framing to imply such a prohibition, we may not use the Clauses to invalidate the decisions of a State or its people.

Ill

It is radical enough for the majority to hold that the Constitution implicitly precludes the people of the States from prescribing any eligibility requirements for the congres*917sional candidates who seek their votes. This holding, after all, does not stop with negating the term limits that many States have seen fit to impose on their Senators and Representatives.39 Today’s decision also means that no State may disqualify congressional candidates whom a court has found to be mentally incompetent, see, e. g., Fla. Stat. §§97.041(2), 99.021(1)(a) (1991), who are currently in prison, see, e. g., Ill. Comp. Stat. Ann., ch. 10, §§5/3-5, 5/7-10, 5/10-5 (1993 and West Supp. 1995), or who have past vote-fraud convictions, see, e.g., Ga. Code Ann. §§21-2-2(25), 21-2-8 (1993 and Supp. 1994). Likewise, after today’s decision, the people of each State must leave open the possibility that they will trust someone with their vote in Congress even though they do not trust him with a vote in the election for Congress. See, e. g., R. I. Gen. Laws § 17-14-1.2 (1988) (restricting candidacy to people “qualified to vote”).

In order to invalidate § 3 of Amendment 73, however, the majority must go further. The bulk of the majority’s analysis — like Part II of my dissent — addresses the issues that would be raised if Arkansas had prescribed “genuine, unadulterated, undiluted term limits.” See Rotunda, 73 Ore. L. Rev., at 570. But as the parties have agreed, Amendment 73 does not actually create this kind of disqualification. See *918Tr. of Oral Arg. 53-54; cf. ante, at 828. It does not say that covered candidates may not serve any more terms in Congress if reelected, and it does not indirectly achieve the same result by barring those candidates from seeking reelection. It says only that if they are to win reelection, they must do so by write-in votes.

One might think that this is a distinction without a difference. As the majority notes, “[t]he uncontested data submitted to the Arkansas Supreme Court” show that write-in candidates have won only six congressional elections in this century. Ante, at 830, n. 43. But while the data’s accuracy is indeed “uncontested,” petitioners filed an equally uncontested affidavit challenging the data’s relevance. As political science professor James S. Fay swore to the Arkansas Supreme Court, “[m]ost write-in candidacies in the past have been waged by fringe candidates, with little public support and extremely low name identification.” App. 201. To the best of Professor Fay’s knowledge, in modern times only two incumbent Congressmen have ever sought reelection as write-in candidates. One of them was Dale Alford of Arkansas, who had first entered the House of Representatives by winning 51% of the vote as a write-in candidate in 1958; Alford then waged a write-in campaign for reelection in 1960, winning a landslide 83% of the vote against an opponent who enjoyed a place on the ballot. Id., at 201-202. The other incumbent write-in candidate was Philip J. Philbin of Massachusetts, who — despite losing his party primary and thus his spot on the ballot — won 27% of the vote in his unsuccessful write-in candidacy. See id., at 203. According to Professor Fay, these results — coupled with other examples of successful write-in campaigns, such as Ross Perot’s victory in North Dakota’s 1992 Democratic Presidential primary — “demonstrate that when a write-in candidate is well-known and well-funded, it is quite possible for him or her to win an election.” Ibid.

*919The majority responds that whether “the Arkansas amendment has the likely effect of creating a qualification” is “simply irrelevant to our holding today.” Ante, at 836. But the majority — which, after all, bases its holding on the asserted exclusivity of the Qualifications Clauses — never adequately explains how it can take this position and still reach its conclusion.

One possible explanation for why the actual effect of the Arkansas amendment might be irrelevant is that the Arkansas Supreme Court has already issued a binding determination of fact on this point. Thus, the majority notes that “the state court” has advised us that “there is nothing more than a faint glimmer of possibility that the excluded candidate will win:” Ante, at 830. But the majority is referring to a mere plurality opinion, signed by only three of the seven justices who decided the case below. One of the two justices who concurred in the plurality’s holding that Amendment 73 violates the Qualifications Clauses did write that “as a practical matter, the amendment would place term limits on service in the Congress,” but he immediately followed this comment with the concession that write-in candidacies are not entirely hopeless; his point was simply that “as a practical matter, write-in candidates are at a distinct disadvantage.” 316 Ark., at 276; 872 S. W. 2d, at 364 (Dudley, J., concurring in part and dissenting in part). As a result, the majority may rely upon the state court only for the proposition that Amendment 73 makes the specified candidates “distinctly]” worse off than they would be in its absence— an unassailable proposition that petitioners have conceded.

In the current posture of these cases, indeed, it would have been extremely irregular for the Arkansas Supreme Court to have gone any further. Disputed questions of fact, in Arkansas as elsewhere, generally are resolved at trial rather than on appeal from the entry of summary judgment. See *920Ark. Rule Civ. Proc. 56.40 Accordingly, the majority explicitly disclaims any reliance on the state court’s purported finding about the effect of Amendment 73. See ante, at 830, n. 44.

Instead, the majority emphasizes another purported conclusion of the Arkansas Supreme Court. As the majority notes, the plurality below asserted that “[t]he intent” of Amendment 73 was “to disqualify congressional incumbents from further service.” 316 Ark., at 266, 872 S. W. 2d, at 357. According to the majority, “[w]e must, of course, accept the state court’s view of the purpose of its own law: We are thus authoritatively informed that the sole purpose of §3 of Amendment 73 was to attempt to achieve a result that is forbidden by the Federal Constitution.” Ante, at 829.

I am not sure why the intent behind a law should affect our analysis under the Qualifications Clauses. If a law does not in fact add to the constitutional qualifications, the mistaken expectations of the people who enacted it would not seem to affect whether it violates the alleged exclusivity of those Clauses. But in any event, the majority is wrong about what “the state court” has told us. Even the plurality *921below did not flatly assert that the desire to “disqualify” congressional incumbents was the sole purpose behind §3 of Amendment 73. More important, neither of the justices who concurred in the plurality’s holding said anything at all about the intent behind Amendment 73. As a result, we cannot attribute any findings on this issue to the Arkansas Supreme Court.

The majority suggests that this does not matter, because Amendment 73 itself says that it has the purpose of “evading the requirements of the Qualifications Clauses.” See ante, at 831 (referring to the “avowed purpose” of Amendment 73). The majority bases this assertion on the amendment’s preamble, which speaks of “limiting] the terms of elected officials.” See ante, at 830. But this statement may be referring only to §§ 1 and 2 of Amendment 73, which impose true term limits on state officeholders. Even if the statement refers to § 3 as well, it may simply reflect the limiting effects that the drafters of the preamble expected to flow from what they perceived as the restoration of electoral competition to congressional races. See infra, at 924. In any event, inquiries into legislative intent are even more difficult than usual when the legislative body whose unified intent must be determined consists of 825,162 Arkansas voters.

The majority nonetheless thinks it clear that the goal of §3 is “to prevent the election of incumbents.” See ante, at 830, 836. In reaching this conclusion at the summary-judgment stage, however, the majority has given short shrift to petitioners’ contrary claim. Petitioners do not deny that § 3 of Amendment 73 intentionally handicaps a class of candidates, in the sense that it decreases their pre-existing electoral chances. But petitioners do deny that § 3 is intended to (or will in fact) “prevent” the covered candidates from winning reelection, or “disqualify” them from further service. One of petitioners’ central arguments is that congressionally conferred advantages have artificially inflated the pre-existing electoral chances of the covered candidates, and *922that Amendment 73 is merely designed to level the playing field on which challengers compete with them.

To understand this argument requires some background. Current federal law (enacted, of course, by congressional incumbents) confers numerous advantages on incumbents, and these advantages are widely thought to make it “significantly more difficult” for challengers to defeat them. Cf. ante, at 831. For instance, federal law gives incumbents enormous advantages in building name recognition and good will in their home districts. See, e. g., 39 U. S. C. § 3210 (permitting Members of Congress to send “franked” mail free of charge); 2 U. S. C. §§61-1, 72a, 332 (permitting Members to have sizable taxpayer-funded staffs); 2 U. S. C. § 123b (establishing the House Recording Studio and the Senate Recording and Photographic Studios).41 At the same time that incumbent Members of Congress enjoy these in-kind benefits, Congress imposes spending and contribution limits in congressional campaigns that “can prevent challengers from spending more . . . to overcome their disadvantage in name recognition.” App. to Brief for State of Washington as Amicus Curiae A-4 (statement of former 10-term Representative William E. Frenzel, referring to 2 U. S. C. § 441a). Many observers believe that the campaign-finance laws also give incumbents an “enormous fund-raising edge” over their challengers by giving a large financing role to entities with incentives to curry favor with incumbents. Wertheimer & Manes, Campaign Finance Reform: A Key to Restoring the Health of Our Democracy, 94 Colum. L. Rev. 1126, 1133 (1994). In *923addition, the internal rules of Congress put a substantial premium on seniority, with the result that each Member’s already plentiful opportunities to distribute benefits to his constituents increase with the length of his tenure. In this manner, Congress effectively “fines” the electorate for voting against incumbents. Hills, 53 U. Pitt. L. Rev., at 144-145.

Cynics see no accident in any of this. As former Representative Frenzel puts it: “The practice ... is for incumbents to devise institutional structures and systems that favor incumbents.” App. to Brief for State of Washington as Amicus Curiae A-3. In fact, despite his service from 1971 to 1989 on the House Administration Committee (which has jurisdiction over election laws), Representative Frenzel can identify no instance in which Congress “changed election laws in such a way as to lessen the chances of re-election for incumbents, or to improve the election opportunities for challengers.” Ibid.

At the same time that incumbents enjoy the electoral advantages that they have conferred upon themselves, they also enjoy astonishingly high reelection rates. As Lloyd Cutler reported in 1989, “over the past thirty years a weighted average of ninety percent of all House and Senate incumbents of both parties who ran for reelection were reelected, even at times when their own party lost control of the Presidency itself.” Cutler, Now is the Time for All Good Men ... , 30 Wm. & Mary L. Rev. 387, 395; see also Kristol, Term Limitations: Breaking Up the Iron Triangle, 16 Harv. J. L. & Pub. Policy 95, 97, and n. 11 (1993) (reporting that in the 100th Congress, as many Representatives died as were defeated at the polls). Even in the November 1994 elections, which are widely considered to have effected the most sweeping change in Congress in recent memory, 90% of the incumbents who sought reelection to the House were successful, and nearly half of the losers were completing only their first terms. Reply Brief for Petitioners U. S. Term Limits, Inc., et al. 4, n. 5. Only 2 of the 26 Senate incumbents seeking reelection were defeated, see ibid., and one of *924them had been elected for the first time in a special election only a few years earlier.

The voters of Arkansas evidently believe that incumbents would not enjoy such overwhelming success if electoral contests were truly fair — that is, if the government did not put its thumb on either side of the scale. The majority offers no reason to question the accuracy of this belief. Given this context, petitioners portray § 3 of Amendment 73 as an effort at the state level to offset the electoral advantages that congressional incumbents have conferred upon themselves at the federal level.

To be sure, the offset is only rough and approximate; no one knows exactly how large an electoral benefit comes with having been a long-term Member of Congress, and no one knows exactly how large an electoral disadvantage comes from forcing a well-funded candidate with high name recognition to run a write-in campaign. But the majority does not base its holding on the premise that Arkansas has struck the wrong balance. Instead, the majority holds that the Qualifications Clauses preclude Arkansas from trying to strike any balance at all; the majority simply says that “an amendment with the avowed purpose and obvious effect of evading the requirements of the Qualifications Clauses by handicapping a class of candidates cannot stand.” Ante, at 831. Thus, the majority apparently would reach the same result even if one could demonstrate at trial that the electoral advantage conferred by Amendment 73 upon challengers precisely counterbalances the electoral advantages conferred by federal law upon long-term Members of Congress.

For me, this suggests only two possibilities. Either the majority’s holding is wrong and Amendment 73 does not violate the Qualifications Clauses, or (assuming the accuracy of petitioners’ factual claims) the electoral system that exists without Amendment 73 is no less unconstitutional than the electoral system that exists with Amendment 73.

*925I do not mean to suggest that States have unbridled power to handicap particular classes of candidates, even when those candidates enjoy federally conferred advantages that may threaten to skew the electoral process. But laws that allegedly have the purpose and effect of handicapping a particular class of candidates traditionally are reviewed under the First and Fourteenth Amendments rather than the Qualifications Clauses. Compare Storer v. Brown, 415 U. S., at 728-736 (undertaking a lengthy First and Fourteenth Amendment analysis of a California rule that denied ballot access to any independent candidate for Congress who had not severed his ties to a political party at least one year prior to the immediately preceding primary election, or 17 months before the general election), with id., at 746, n. 16 (dismissing as “wholly without merit” the notion that this rule might violate the Qualifications Clauses). Term-limit measures have tended to survive such review without difficulty. See, e. g., Moore v. McCartney, 425 U. S. 946 (1976) (dismissing an appeal from State ex rel. Maloney v. McCartney, 159 W. Va. 513, 223 S. E. 2d 607, on the ground that limits on the terms of state officeholders do not even raise a substantial federal question under the First and Fourteenth Amendments).

To analyze such laws under the Qualifications Clauses may open up whole new vistas for courts. If it is true that “the current congressional campaign finance system . . . has created an electoral system so stacked against challengers that in many elections voters have no real choices,” Wertheimer & Manes, 94 Colum. L. Rev., at 1133, are the Federal Election Campaign Act Amendments of 1974 unconstitutional under (of all things) the Qualifications Clauses? Cf. Buckley v. Valeo, 424 U. S. 1 (1976) (upholding the current system against First Amendment challenge). If it can be shown that nonminorities are at a significant disadvantage when they seek election in districts dominated by minority voters, would the intentional creation of “majority-minority *926districts” violate the Qualifications Clauses even if it were to survive scrutiny under the Fourteenth Amendment? Cf. Shaw v. Reno, 509 U. S. 630, 649 (1993) (“[W]e express no view as to whether [the intentional creation of such districts] always gives rise to an equal protection claim”); id., at 677 (Stevens, J., dissenting) (arguing that States may draw district lines for the “sole purpose” of helping blacks or members of certain other groups win election to Congress). More generally, if “[district lines are rarely neutral phenomena” and if “districting inevitably has and is intended to have substantial political consequences,” Gaffney v. Cummings, 412 U. S. 735, 753 (1973), will plausible Qualifications Clause challenges greet virtually every redistricting decision? Cf. id., at 754 (noting our general refusal to use the Equal Protection Clause to “attemp[t] the impossible task of extirpating politics from what are the essentially political processes of the sovereign States”); see also Burns v. Richardson, 384 U. S. 73, 89, n. 16 (1966) (finding nothing invidious in the practice of drawing district lines in a way that helps current incumbents by avoiding contests between them).

The majority’s opinion may not go so far, although it does not itself suggest any principled stopping point. No matter how narrowly construed, however, today’s decision reads the Qualifications Clauses to impose substantial implicit prohibitions on the States and the people of the States. I would not draw such an expansive negative inference from the fact that the Constitution requires Members of Congress to be a certain age, to be inhabitants of the States that they represent, and to have been United States citizens for a specified period. Rather, I would read the Qualifications Clauses to do no more than what they say. I respectfully dissent.

The ringing initial words of the Constitution — “We the People of the United States” — convey something of the same idea. (In the Constitution, after all, “the United States” is consistently a plural noun. See *847Art. I, § 9, cl. 8; Art. II, § 1, cl. 7; Art. Ill, § 2, cl. 1; Art. Ill, § 3, cl. 1; cf. Amar, Of Sovereignty and Federalism, 96 Yale L. J. 1425,1455 (1987) (noting this fact, though reaching other conclusions).) The Preamble that the Philadelphia Convention approved before sending the Constitution to the Committee of Style is even clearer. It began: “We the people of the States of New-Hampshire, Massachusetts, Rhode-Island and Providence Plantations, Connecticut, New-York, New-Jersey, Pennsylvania, Delaware, Maryland, Virginia, North-Carolina, South-Carolina, and Georgia____” 2 Records of the Federal Convention of 1787, p. 565 (M. Farrand ed. 1911) (hereinafter Farrand). Scholars have suggested that the Committee of Style adopted the current language because it was not clear that all the States would actually ratify the Constitution. M. Farrand, The Framing of the Constitution of the United States 190-191 (1913). In this instance, at least, I agree with the majority that the Committee’s edits did not work a substantive change in the Constitution. Cf. ante, at 792, n. 8.

The concurring opinion appears to draw precisely the opposite conclusion from the passage in McCulloch that contains this sentence. See ante, at 840-841. But while the concurring opinion seizes on Marshall’s references to “the people,” Marshall was merely using that phrase in contradistinction to “the State governments.” Counsel for Maryland had *850noted that “the constitution was formed and adopted, not by the people of the United States at large, but by the people of the respective States. To suppose that the mere proposition of this fundamental law threw the American people into one aggregate mass, would be to assume what the instrument itself does not profess to establish.” McCulloch, 4 Wheat., at 363 (argument of counsel). Marshall’s opinion accepted this premise, even borrowing some of counsel’s language. See id., at 403. What Marshall rejected was counsel’s conclusion that the Constitution therefore was merely “a compact between the States.” See id., at 363 (argument of counsel). As Marshall explained, the acts of “the people themselves” in the various ratifying conventions should not be confused with “the measures of the State governments.” Id., at 403; see also id., at 404 (noting that no state government could control whether the people of that State decided to adopt the Constitution).

At the time of the framing, of course, a Federal Congress had been operating under the Articles of Confederation for some 10 years. The States unquestionably had enjoyed the power to establish qualifications for their delegates to this body, above and beyond the qualifications created by the Articles themselves. See Brief for Respondents Bobbie E. Hill et al. 39, n. 79 (conceding this point); see also, e. g., Md. Const. of 1776, Art. XXVII (prescribing such qualifications), in 3 Federal and State Constitutions 1695-1696 (F. Thorpe ed. 1909) (hereinafter Thorpe); N. H. Const. of 1784, Pt. II (same), in 4 Thorpe 2467. It is surprising, then, that the concurring opinion seeks to buttress the majority’s case by stressing the continuing applicability of “the same republican principles” that had prevailed under the Articles. See ante, at 839.

Despite the majority’s odd suggestion to the contrary, see ante, at 796-797, n. 12,1 fully agree with this sensible position. See supra, at 848.

Though cited by the majority, see ante, at 802, Crandall v. Nevada, 6 Wall. 35 (1868), did not deviate from this accepted view of McCulloch. See Crandall, supra, at 48 (observing that McCulloch and a number of other cases “distinctly placed the invalidity of the State taxes on the ground that they interfered with an authority of the Federal government”).

To support its decision to attribute such surplusage to McCulloch, the majority quotes Marshall’s observation that his opinion “ ‘does not deprive the States of any resources which they originally possessed,’ ” because the power to tax federal instrumentalities was not encompassed by the States’ “‘original right to tax.’” Ante, at 802 (quoting McCulloch, 4 Wheat., at 436, 430). In part, Marshall was simply refuting counsel’s argument that it would constitute an “overwhelming invasion of State sovereignty” for Congress to establish a bank that operated within a State but that nonetheless was exempt from state taxes. See id., at 337-339 (argument of counsel) (stressing that “the right to raise revenue” is “the highest attribute of sovereignty” and indeed amounts to “the right to exist”). While Marshall acknowledged that “this original right of taxation” was an “essential” attribute of state sovereignty that Congress could not constitutionally control or invade, he focused more precisely than counsel on “the nature and extent of this original right,” id., at 428, and concluded that it did not include the right “to tax the means employed by the government of the Union, for the execution of its powers.” Id., at 430. In this respect, then, the Court was referring to the States’ “original” powers in much the same context as Garcia v. San Antonio Metropolitan Transit Authority, 469 U. S. 528 (1985): The Court was examining whether Congress’ exercise of the “privilege of exempting its own measures from State taxation,” McCulloch, supra, at 434, had invaded a protected sphere of state sovereignty.

Marshall did go on to argue that the power to tax the operations of the Bank of the United States simply was not susceptible to control by the *856people of a single State. See 4 Wheat., at 430. But that theory is perfectly consistent with my position. Marshall reasoned that the people of a single State may not tax the instrumentalities employed by the people of all the States through the National Government, because such taxation would effectively subject the people of the several States to the taxing power of a single State. See id., at 428. This sort of argument proves that the people of a single State may not prescribe qualifications for the President of the United States; the selection of the President, like the operation of the Bank of the United States, is not up to the people of any single State. See infra, at 862. It does not follow, however, that the people of a single State may not prescribe qualifications for their own representatives in Congress.

See 1 S. Johnson, A Dictionary of the English Language 393 (4th ed. 1773) (defining “congress” as “[a]n appointed meeting for settlement of affairs between different nations: as, the congress of Cambray”); T. Sheridan, A Complete Dictionary of the English Language (6th ed. 1796) (“an appointed meeting for settlement of affairs between different nations; the assembly which governs the United States of America”).

The majority even suggests that congressional elections do not really work in this way, because each House of Congress has the power to judge its Members’ qualifications. See ante, at 804 (citing Art. I, § 5, cl. 1). But the power to act as “Judge” under Art. I, §5, is merely the power to apply pre-existing qualifications to which the people of each State have consented. See Powell v. McCormack, 395 U. S. 486 (1969). Whether or not § 5 directs each House to judge state-law disqualifications as well as those contained in the Constitution, see infra, at 895, it is clear' that neither House may exclude a representative from Massachusetts for failure to meet a qualification that the people of Massachusetts have not accepted.

The only provision that might conceivably do so is Article II, § 1, which recognizes the authority of state legislatures to specify the “Manner” in which a State appoints its Presidential electors. But if a qualifications law is a “Manner” regulation for purposes of this Clause, then it is also a “Manner” regulation for purposes of Article I, §4 — which would mean that the Constitution specifically recognizes the power of both the States and the Congress to set qualifications for Senators and Representatives.

Accord, e. g., 2 Elliot 24 (remarks of Caleb Strong at the Massachusetts ratifying convention) (“[I]f the legislature of a state should refuse to make such regulations, the consequence will be, that the representatives will not be chosen, and the general government will be dissolved. In such case, can gentlemen say that a power to remedy the evil is not necessary to be lodged somewhere? And where can it be lodged but in Congress?”); 2 Documentary History of the Ratification of the Constitution 400 (M. Jensen ed. 1976) (notes of Anthony Wayne at the Pennsylvania ratifying convention) (“4th section occasioned by an eventual invasion, insurrection, etc.”); The Federalist No. 59, at 363 (Hamilton) (observing that if not subject to any checks, the States “could at any moment annihilate [the Federal Government] by neglecting to provide for the choice of persons to administer its affairs”).

These statements about the Clause’s purposes also help refute the majority’s claim that it was bizarre for the Framers to leave the States relatively free to enact qualifications for congressional office while simultaneously giving Congress “make or alter” power over the States’ time, place, and manner regulations. See infra, at 896-898.

Thus, the age requirement was intended to ensure that Members of Congress were people of mature judgment and experience. See, e. g., 1 Farrand 375 (remarks of George Mason at the Philadelphia Convention); 3 id., at 147 (remarks of James McHenry before the Maryland House of Delegates). The citizenship requirement was intended both to ensure that Members of Congress were familiar with the country and that they were not unduly susceptible to foreign influence. See, e. g., 2 id., at 216 (remarks of George Mason). The inhabitancy requirement was intended to produce a National Legislature whose Members, collectively, had a local knowledge of all the States. See, e. g., The Federalist No. 56 (Madison). The Ineligibility Clause was intended to guard against corruption. See, e. g., 1 Farrand 381 (remarks of Alexander Hamilton).

The principle that the Constitution rests on the consent of the people of the States points in the same direction. Both the process of selecting delegates to the Philadelphia Convention and the ratification procedure erected by Article VII were designed to let the States and the people of the States protect their interests. Lest those protections be evaded, one should not be quick to read the Qualifications Clauses as imposing unstated prohibitions that pre-empt all state qualifications laws. Cf. L. Tribe, American Constitutional Law §6-25, p. 480 (2d ed. 1988) (arguing that courts should hesitate to read federal statutes to pre-empt state law, because “to give the state-displacing weight of federal law to mere congressional ambiguity would evade the very procedure for lawmaking on which Garcia [v. San Antonio Metropolitan Transit Authority, 469 U. S. 528 (1985)] relied to protect states’ interests”); Gregory v. Ashcroft, 501 U. S. 452, 464 (1991) (applying this argument).

Even when Congress enacted the first federal naturalization law in 1790, it left open the possibility that the individual States could establish more lenient standards of their own for admitting people to citizenship. While Hamilton had suggested that Congress’ power to “establish an uniform Rule” logically precluded the States from deviating downward from the rule that Congress established, see The Federalist No. 32, at 199, the early cases on this question took the opposite view. See Collet v. Collet, 2 Dall. 294, 296 (CC Pa. 1792) (Wilson, Blair, and Peters, JJ.). States therefore continued to enact naturalization laws of their own until 1795, *873when Congress passed an exclusive naturalization law. See J. Kettner, Development of American Citizenship, 1608-1870, pp. 242-243 (1978).

The majority notes Jefferson’s concession that state power to supplement the Qualifications Clauses was “one of the doubtful questions on which honest men may differ with the purest of motives.” See ante, at 813, n. 24; 14 Writings of Thomas Jefferson 83 (A. Lipscomb ed. 1904). But while Jefferson cautioned against impugning the motives of people who might disagree with his position, his use of the phrase “[o]f course” suggests that he himself did not entertain serious doubts of its correctness.

The majority also errs in its interpretation of Nixon v. United States, 506 U. S. 224 (1993). See ante, at 796, n. 12. In dictum, Nixon did refer to “the fixed meaning of ‘[q]ualifications’ set forth in Art. I, §2.” 506 U. S., at 237. But as both the surrounding context and the internal punctuation of this passage make clear, Nixon was referring to the meaning of the word “Qualifications” in § 5; that term, after all, does not even appear in the House Qualifications Clause of §2. Thus, Nixon merely said that §5 directs the House to judge the qualifications “set forth in Art. I, §2,” and not qualifications of its own invention. See also infra, at 895. There *877would have been no occasion for Nixon to extend Powell: The only point of its discussion was to explain why the question at issue in Powell was justiciable, while the question at issue in Nixon (which concerned impeachment) was not.

Oregon, for instance, pioneered a system in which the state legislature bound itself to appoint the candidates chosen in a statewide vote of the people. See Hills, A Defense of State Constitutional Limits on Federal Congressional Terms, 53 U. Pitt. L. Rev. 97, 108 (1991). The majority is in the uncomfortable position of. suggesting that this system violated “democratic principles.”

For instance, the majority quotes Noah Webster’s observation that under the Constitution, “the places of senators are wisely left open to all persons of suitable age and merit, and who have been citizens of the United States for nine years.” See ante, at 820, n. 80 (citing “A Citizen of America” (Oct. 17, 1787), in 1 Debate on the Constitution 129, 142 (B. Bailyn ed. 1993) (hereinafter Bailyn)). But there is no reason to read Webster as denying the power of state legislatures to pass resolutions limiting the field of potential candidates that they would consider for appointment to the Senate. Indeed, it seems implausible that Webster would have been invoking the majority’s vision of “democratic principles” in support of the constitutional provisions calling for Senators to be appointed by the various state legislatures rather than being elected directly by the people of the States.

Similarly, the majority quotes a newspaper piece written by John Stevens, Jr., to the people of New York. See ante, at 819-820. But Stevens gave the following explanation for his assertion that “[n]o man who has real merit... need despair” under the system erected by the Constitution: “He first distinguishes himself amongst his neighbours at township and county meeting; he is next sent to the State Legislature. In this theatre his abilities ... are . .. displayed to the views of every man in the State: from hence his ascent to a seat in Congress becomes easy and sure.” “Americanus,” Daily Advertiser, Dec. 12, 1787, in 1 Bailyn 487, 492. As the States indisputably controlled eligibility requirements for membership in the various state legislatures, and indeed had established some disquali*881fications, I do not read Stevens to be saying that they were barred from doing the same thing with respect to Congress. Without addressing whether the people of the States may supplement the Qualifications Clauses, Stevens was merely praising the Constitution for imposing few such requirements of its own.

For instance, the majority quotes at length from the debate that arose in the Philadelphia Convention when the Committee of Detail proposed the following clause: “The Legislature of the United States shall have authority to establish such uniform qualifications of the members of each House, with regard to property, as to the said Legislature shall seem expedient.” See 2 Farrand 179, 248-251; ante, at 790-791. The defeat of this proposal — like the defeat of Gouverneur Morris’ motion to drop the words “with regard to property” from the clause, so as to empower Congress to enact qualifications of any sort — simply reflects the Framers’ decision not to grant Congress the power to supplement the constitutional qualifications. Considered out of context, some of James Madison’s comments during the debate might be thought to go farther. See ibid. But the majority itself properly dispels this false impression. See ante, at 793, n. 10; see also Powell v. McCormack, 395 U. S., at 534.

Likewise, Powell drew support from Alexander Hamilton’s comments in The Federalist No. 60, which the majority also quotes. See ante, at 791. But as the majority concedes, when Hamilton wrote that “[t]he qualifications of the persons who may choose or be chosen [for Congress] . . . are defined and fixed in the Constitution, and are unalterable by the legislature,” he was merely restating his prior observation that the power to set qualifications “forms no part of the power to be conferred upon the national government.” See The Federalist No. 60, at 371 (emphasis added). Indeed, only if “the legislature” to which Hamilton was referring is Congress can one make sense of his remark that the qualifications of voters as well as Congressmen are “fixed in the Constitution” and “unalterable by the legislature.” Hamilton surely knew that the States or the people of the States control eligibility for the franchise. See Art. I, §2, cl. 1.

The majority does omit the context necessary to understand one aspect of the historical evidence presented in Powell. The majority quotes Powell’s observation that “on the eve of the Constitutional Convention, English precedent stood for the proposition that ‘the law of the land had regulated *886the qualifications of members to serve in parliament’ and those qualifications were ‘not occasional but fixed.’ ” 395 U. S., at 528 (quoting 16 Parliamentary History of England 589, 590 (1769)); see ante, at 790. The English rule seems of only marginal relevance: The pre-existing rule in America — that States could add qualifications for their representatives in Congress, see n. 3, supra, while Congress itself could not — is surely more important. But in any event, Powell did not claim that the English rule deemed parliamentary qualifications to be fixed in the country’s (unwritten) constitution, beyond the reach of a properly enacted law. Instead, qualifications were “fixed” rather than “occasional” only in the sense that neither House of Parliament could “exclude members-elect for general misconduct not within standing qualifications.” Powell, 395 U. S., at 528. The English rule, in other words, was simply that when sitting as the judge of its members’ qualifications, each House of Parliament could do no more than administer the pre-existing laws that defined those qualifications, see id., at 529, for “one House of Parliament cannot create a disability unknown to the law.” T. Plucknett, Taswell-Langmead’s English Constitutional History 585 (11th ed. 1960); cf. INS v. Chadha, 462 U. S. 919 (1983). This history was relevant to Powell (which dealt with the grounds on which one House of Congress could exclude a Member-elect), but it is not relevant to this case.

The majority also argues that in any event, the views of the members of the Committee “tel[l] us little about the views of the Convention as a whole.” Ante, at 815, n. 27. But our task is simply to determine whether at the time of the framing, the language of the Qualifications Clauses would have been commonly understood to contain an exclusivity provision. The surviving records suggest that the members of the Committee of Detail did not understand the final Qualifications Clauses to be exclusive, and the majority offers no reason to think that their understanding of the language was unusual for their time.

The majority inaccurately reports James Madison’s explanation of the Elector-Qualifications Clause in The Federalist No. 52. Madison neither mentioned nor addressed the consequences of “allowing States to differentiate between the qualifications for state and federal electors.” See ante, at 808. Instead, he addressed the problems that would have arisen if the Constitution had assigned control over the qualifications of voters in House elections to the state legislatures rather than to the people of each State. It was such an arrangement that, in Madison’s view, “would have rendered too dependent on the State governments that branch of the federal government which ought to be dependent on the people alone.” The Federalist No. 52, at 326; cf. ante, at 808. The Elector-Qualifications Clause avoided this problem because the various state constitutions con*891trolled who could vote in elections for the most numerous branch of the state legislature, and no state government could alter these requirements unless the people of the State (through the state constitution) decided to let it do so. See The Federalist No. 52, at 326.

Though one obviously could uphold the action of the people of Arkansas without reaching this issue, Madison’s comments should not be read to suggest that the Elector-Qualifications Clause bars the people of a State from delegating their control over voter qualifications to the state legislature. The Clause itself refutes this reading; if a state constitution permits the state legislature to set voter qualifications, and if eligibility for the franchise in the State therefore turns on statutory rather than constitutional law, federal electors in the State still must meet the same qualifications as electors for the most numerous branch of the state legislature. Madison could not possibly have disagreed with this understanding of the Clause. Instead, he was simply explaining why, when it came to voter qualifications for House elections, the Framers had not followed the model of Article I, §3, cl. 1, and vested ultimate control with the state legislatures (regardless of what the people of a State might provide in their state constitutions).

Likewise, the Constitution requires the States to appoint Presidential electors, Art. II, § 1, cl. 2, but it does not provide for any congressional override if the States refuse to do so (or if the States set impossibly high qualifications and then announce that no one meets them).

Even if there is anything left of the majority’s argument on this point, it would still have no bearing on whether the Framers intended to preclude the people of each State from supplementing the constitutional qualifications. Just as the Framers had no fear that the people of a State would destroy congressional elections by entirely disenfranchising themselves, see The Federalist No. 52, at 326, so the Framers surely had no fear that the people of the States would destroy congressional elections by entirely disqualifying all candidates.

As the majority notes, see ante, at 837, and 812, n. 22, the Philadelphia Convention had dropped without discussion a portion of the original Randolph Resolutions calling for Members of the House of Representatives *899“to be incapable of re-election for the space of [blank space] after the expiration of their term of service.” 1 Farrand 20. This provision, which at a minimum would have barred all Members of the House from serving consecutive terms, was abandoned without objection when the Convention voted to require House Members to stand for election every three years. See id., at 214-217; see also id., at 362 (opting for 2-year terms instead). Subsequently, indeed, some members of the Convention appeared to be unaware that a rotation requirement had ever been proposed. See 2 id., at 120 (remarks of Gouverneur Morris).

The majority properly does not cite the omission of this nationwide rotation requirement as evidence that the Framers meant to preclude individual States from adopting rotation requirements of their own. Just as individual States could extend the vote to women before the adoption of the Nineteenth Amendment, could prohibit poll taxes before the adoption of the Twenty-fourth Amendment, and could lower the voting age before the adoption of the Twenty-sixth Amendment, so the Framers’ decision not to impose a nationwide limit on congressional terms did not itself bar States from adopting limits of their own. See, e. g., Ga. Const. of 1877, § 2-602 (adopted Aug. 3, 1943) (reducing voting age to 18 nearly three decades before the Twenty-sixth Amendment was proposed); Harman v. Forssenius, 380 U. S. 528, 539 (1965) (noting that by the time the Twenty-fourth Amendment was proposed, “only five States retained the poll tax as a voting requirement”); Congressional Research Service, The Constitution of the United States of America: Analysis and Interpretation 1571 (1973) (reporting that 11 States had adopted women’s suffrage by the time the Nineteenth Amendment was proposed). Cf. ante, at 837, and n. 50.

See Del. Const. of 1776, Art. 11, in 1 Thorpe 564; Md. Const. of 1776, Form of Government, Art. XXVII, in 3 Thorpe 1695; Mass. Const. of 1780, Pt. 2, Ch. IV, in 3 Thorpe 1906; N. H. Const. of 1784, Pt. II, in 4 Thorpe 2467; N. Y. Const. of 1777, Art. XXX, in 5 Thorpe 2634-2635; N. C. Const. of 1776, Form of Government, Art. XXXVII, in 5 Thorpe 2793; Pa. Const. of 1776, Frame of Government, § 11, in 5 Thorpe 3085; S. C. Const. of 1778, Art. XXII, in 6 Thorpe 3253; Va. Const. of 1776, in 7 Thorpe 3817.

Md. Const. of 1776, Form of Government, Art. XXVII, in 3 Thorpe 1695; N. H. Const. of 1784, Pt. II, in 4 Thorpe 2467; N. C. Const. of 1776, Art. XXXVII, in 5 Thorpe 2793; Pa. Const. of 1776, Frame of Government, § 11, in 5 Thorpe 3085.

Md. Const. of 1776, Form of Government, Art. XXVII, in 3 Thorpe 1695; N. H. Const. of 1784, Pt. II, in 4 Thorpe 2467; Pa. Const. of 1776, Frame of Government, § 11, in 5 Thorpe 3085.

See Md. Const. of 1776, Art. XXVII, in 3 Thorpe 1695; N. H. Const. of 1784, Pt. II, in 4 Thorpe 2467.

The majority suggests that I have overlooked Madison’s observation that subject to the “reasonable limitations” spelled out in the House Qualifications Clause, the Constitution left the House’s door “open to merit of every description.” See ante, at 807-808, n. 18; see also ante, at 808 (quoting a similar passage from The Federalist No. 57). As discussed above, however, such statements do not advance the majority’s case. See supra, at 880-881.

The immediately preceding portion of the Clause requires not only “[t]he Senators and Representatives before mentioned” but also “the Members of the several State Legislatures, and all executive and judicial Officers, both of the United States and of the several States,” to take an “Oath or Affirmation” to support the Constitution. Art. VI, cl. 3.

Despite the majority’s emphasis on the Framers’ supposed desire for uniformity in congressional elections, even the majority does not dispute that the Framers wanted to let States decide for themselves whether to use district elections in selecting Members of the House of Representatives. The Framers fully expected that in some States each Member of the House would be chosen by the people of the whole State, while in other States each Member would be directly accountable only to the people of a single district. See, e. g., 14 Papers of Thomas Jefferson 3 (J. Boyd ed. 1958) (letter from Madison to Jefferson, Oct. 8, 1788).

See Georgia Election Law (Jan. 23,1789) (restricting representatives from each district to “resident[s] of three years standing in the district”), in 2 First Federal Elections 456, 457; Maryland Election Law (Dec. 22, 1788) (simple district residency requirement), in 2 First Federal Elections 136, 138; Massachusetts Election Resolutions (Nov. 20, 1788) (same), in 1 First Federal Elections 508, 509 (M. Jensen & R. Becker eds. 1976); North Carolina Election Law (Dec. 16, 1789) (requiring the person elected from each district to have been “a Resident or Inhabitant of that Division for which he is elected, during the Space or Term of one Year before, and at the Time of Election”), in 4 First Federal Elections 347; Virginia Election Law (Nov. 20, 1788) (requiring each candidate to have been “a bona fide resident for twelve months within such District”), in 2 First Federal Elections 293, 294. Upon being admitted to the Union in 1796, Tennessee also required its Members in the Federal House of Representatives to have *906been Tennessee residents for three years and district residents for one year before their election. Act of Apr. 20, 1796, ch. 10, in Laws of the State of Tennessee 81 (1803).

After the Virginia Legislature had enacted this bill, some of James Madison’s friends suggested that he might find it harder to win election in his own district than in certain other areas of the State. They believed that if Madison won the popular vote in one of those other districts, the House of Representatives could seat him on the theory that States cannot add to the constitutional qualifications. See 11 Papers of James Madison 378-379 (R. Rutland & C. Hobson eds. 1977) (letter from Carrington to Madison, Dec. 2,1788). Other advisers, however, warned that the people of Virginia might not share this understanding of the Constitution. As Alexander White wrote in a letter to Madison:

“Some Gentlemen suppose you may be elected in other Districts, and that Congress would disregard the Act which requires Residence in a particular District. I will not undertake to decide that question, but this I know, such a determination would afford much ground of clamour, and enable the opposers of the Government to inflame the Minds of the People beyond anything which has yet happened.” Id., at 380 (Dec. 4, 1788).

Madison himself apparently never endorsed the idea that he should test the district residency requirement. Instead, he ran from his own district (where he overcame a stiff challenge from another future President, James Monroe).

The records show that Maryland’s House of Delegates put the district residency requirement to a separate vote and approved it by a margin of 41 to 24. 2 First Federal Elections 129-130 (summarizing proceedings from Dec. 3, 1788). A subsequent effort to jettison the requirement lost by a vote of 39 to 28. Id., at 132-133 (summarizing proceedings from Dec. 10,1788). Language in Maryland’s second election law confirms that *908the state legislature knew that it was supplementing the Qualifications Clauses. The Act of December 10, 1790, stipulated that each candidate must “b[e] a resident of his district at the time of the election, and hav[e] resided therein twelve calendar months immediately before, and [be] other ways qualified according to the constitution of the United States.” 1790 Laws of Maryland, ch. XVI, art. VIII.

In Georgia, too, the State House of Assembly called special attention to the district residency requirement. Shortly before Georgia held its first federal elections, the House adopted a resolution to stress that if the top votegetter in any district had not been “an actual resident of three years standing” in that district, then “such person shall not be considered as eligible nor shall he be commissioned.” 2 First Federal Elections 459 (resolution of Feb. 4,1789).

Even the experience in New York and South Carolina — the only States that opted for district elections without requiring district residency — does not support the majority’s position. While the records from South Carolina are sketchy, those from New York affirmatively undermine the majority’s suggestion that the Qualifications Clauses were commonly understood to be exclusive. When the topic was first broached in the State Assembly, the assemblymen defeated a district residency proposal amid comments that “to add any other qualification [to those listed in the Constitution] would be unconstitutional.” 3 First Federal Elections 232 (Dec. 18, 1788). But the State Senate took a different view, adding a district residency requirement when it considered the election bill. Id., at 320. The Assembly then approved the requirement by a vote of 36 to 12, id., at 325-326 (Jan. 19, 1789), but reconsidered the requirement the following day (apparently with more assemblymen in attendance). After a sophisticated debate on the constitutional question, with some assemblymen arguing that the district residency requirement was unconstitutional and others responding that the Constitution merely erected minimum qualifications, the Assembly divided evenly over the requirement: 28 voted in favor of it and 28 voted against it. Id., at 328-335 (Jan. 20,1789). The chairman broke the tie with a vote against the requirement. Id., at 335. Still, there clearly was no consensus in the New York Assembly. What is more, some of the votes against the district residency requirement may well have been cast by assemblymen who simply opposed the requirement *909on policy grounds, as an undue restriction on the people’s ability to elect nonresidents if they wanted to do so. In any event,.the New York Senate obviously considered the requirement constitutional.

There is evidence that some members of the Pennsylvania Legislature considered the Qualifications Clauses to be exclusive. See 1 id., at 282-288. Of course, they also believed that §2 of Article I — which calls for Members of the Federal House of Representatives to be “chosen ... by the People of the several States” — forbade Pennsylvania to elect its representatives by districts. See id., at 283. The legislatures of the five States that adopted district residency requirements, who had the Pennsylvania example before them, disagreed with the Pennsylvania legislators.

The majority correctly notes that each convention, in addition to proposing a list of specific “Amendments to the Constitution,” proposed a “Declaration of Rights” to be appended to the Constitution. In both States, this “Declaration” contained the general exhortation that members of both the Legislative and Executive Branches “should, at fixed periods, be reduced to a private station, return into the mass of the people, and the vacancies be supplied by certain and regular elections.” 4 Elliot 243; 3 id., at 657-658. But both Declarations went on to state that at these elections, the previous occupants of the office in question should “be eligible or ineligible [for reelection], as the rules of the constitution of government and the laws shall direct.” 4 id., at 243; 3 id., at 658. Accordingly, it is hard to describe either Declaration as a “proposed . .. constitutional amendment supporting term limits for Members of Congress.” See ante, at 826, n. 40.

As for New York, the State’s ratifying convention did propose amending the Federal Constitution to provide “[t]hat no person be eligible as a senator for more than six years in any term of twelve years.” 1 Elliot 329-330. The majority finds it significant that when this suggestion fell on deaf ears, New Yorkers did not amend their State Constitution to impose this restriction on their state legislature’s appointment authority. Before the Seventeenth Amendment was adopted, however, the Federal Constitution vested the choice of Senators in the state legislatures rather than the people. See Art. I, § 3, cl. 1. At least without a delegation of this authority from the legislature, cf. supra, at 878-882, and n. 16, the people of New York may well have thought that they could no more amend *913the State Constitution to narrow the legislature’s choices for Senator than they could amend the State Constitution to take the appointment of Senators entirely away from the legislature. It obviously would not follow that they doubted their ability to amend the State Constitution to impose constraints on their own choice of Representatives. The ratifying convention’s proposal thus sheds absolutely no light on whether New Yorkers considered the Qualifications Clauses to be exclusive.

Property qualifications may simply have seemed unnecessary. For instance, it surely was far more likely that a pauper would secure one of the 202 seats in the South Carolina House of Representatives than that he would secure one of South Carolina’s five seats in the United States House of Representatives. Compare S. C. Const. of 1778, Art. XIII, in 6 Thorpe 3251, with U. S. Const., Art. I, §2, cl. 3; cf. S. C. Const. of 1790, Art. I, §3 (providing for a 122-seat State House of Representatives), in 6 Thorpe 3258. It may be significant, then, that the one State that saw fit to enact a congressional property qualification was also the State that had the largest congressional delegation. See U. S. Const., Art. I, §2, cl. 3 (allocating 10 seats to Virginia). In addition, people of the day expected that “[t]he representatives of each State [in the federal House] . . . will probably in all cases have been members ... of the State legislature.” The Federalist No. 56, at 348 (Madison); see also n. 17, supra (quoting article by John Stevens, Jr.). Because most States had property re*914quirements for their state legislators, there may have been little perceived need for a separate property qualification for their Members of Congress.

Even States that wanted to create such a qualification, and that considered it within their constitutional authority to do so, might have been deterred by the possibility that the Federal House of Representatives would take a different view. As I have shown, there certainly was no general understanding that the Qualifications Clauses included an unstated exclusivity provision. But people of the day did consider this to be “one of the doubtful questions on which honest men may differ with the purest motives.” 14 Writings of Thomas Jefferson, at 83 (letter to Joseph C. Cabell, Jan. 31, 1814); see n. 14, supra. If some States feared that the “honest men” in the House might throw out the results of an election because of a qualifications law, they might well have thought that any policy benefits of such laws were outweighed by the risk that they would temporarily be deprived of representation in Congress. Alternatively, they may simply have wanted to stay away from difficult constitutional questions. Cf. Ashwander v. TVA, 297 U. S. 288, 347 (1936) (Brandéis, J., concurring). Thus, despite concluding that the States do enjoy the power to prescribe qualifications, Thomas Jefferson questioned whether the advantages of added qualifications were sufficient to justify enacting a law whose constitutionality could be disputed. See 14 Writings of Thomas Jefferson, at 84.

Though obliquely acknowledging this fact, the majority thinks it relevant that some subsequent commentators have mistakenly accepted the gloss put on the McCreery case by two editors in 1834. See ante, at 817-818 (citing treatises, each of which relies upon Cases of Contested Elections in Congress (M. Clarke & D. Hall eds. 1834)). But surely we need not accept an inaccurate view of history merely because it has appeared in print. The majority also cites Thomas Jefferson’s hazy recollection of the McCreery case, see ante, at 817, without acknowledging Jefferson’s conclusion that the States were free to supplement the Qualifications Clauses. See supra, at 873-874.

Going into the November 1994 elections, eight States had adopted “pure” term limits of one sort or another. See Colo. Const., Art. XVIII, § 9a; Mich. Const., Art. II, § 10; Mo. Const., Art. III, § 45(a); Mont. Const., Art. IV, § 8; Ohio Const., Art. V, § 8; Ore. Const., Art. II, § 20; S. D. Const., Art. III, §32; Utah Code Ann. §20A-10-301. Eight other States had enacted “ballot access” provisions triggered by long-term incumbency or multiple prior terms in Congress. See Ariz. Const., Art. VII, § 18; Ark. Const., Arndt. 73, §3; Calif. Elec. Code Ann. §25003 (West Supp. 1994); Fla. Const., Art. VI, §§ 4(b)(5), (6); N. D. Cent. Code §16.1-01-13.1 (Supp. 1993); Okla. Const., Art. II, §12A; Wash. Rev. Code §§29.68.015, 29.68.016 (1994); Wyo. Stat. §22-5-104 (Supp. 1994). In the 1994 elections, six more States — Alaska, Idaho, Maine, Massachusetts, Nebraska, and Nevada— enacted term-limit or ballot-access measures, bringing to 22 the total number of States with such provisions. See Pear, The 1994 Elections, N. Y. Times, Nov. 10, 1994, p. B7, col. 4. In 21 of these States, the measures have been enacted by direct vote of the people.

Even if one were inclined to believe that the Arkansas Supreme Court had departed from the usual practice and had purported to make a binding determination on a disputed issue of fact, we would not be foreclosed from examining the basis for that determination. To be sure, on direct review of a state court’s judgment, we will not “conduct a more searching review of findings made in state trial court than we conduct with respect to federal district court findings.” Hernandez v. New York, 500 U. S. 352, 369 (1991) (plurality opinion). But that is only to say that we will review state-court findings under the “clear error” standard. Ibid.; accord, id., at 372 (O’Connor, J., concurring in judgment); cf. id., at 379 (Stevens, J., dissenting) (identifying no standard of review, but arguing that the state court’s decision should be reversed because its underlying factual findings were erroneous). In certain areas, indeed, this Court apparently gives quite little deference to the initial factfinder, but rather “exercisefs] its own independent judgment” about the factual conclusions that should be drawn from the record. See Bose Corp. v. Consumers Union of United States, Inc., 466 U. S. 485, 501, and n. 17 (1984) (Stevens, J.).

Former Representative William E. Frenzel describes the House Recording Studio as a sophisticated operation used “to prepare tapes of speeches and messages to voters.” Frenzel explains: “Taxpayers pay for the facilities, the personnel that run them, the production costs, and the costs of distributing, by mail or otherwise, the tapes that members supply (from their taxpayer-funded expense accounts). These messages are widely disseminated by broadcasters, who can use them to fill air time at no cost to themselves.” App. to Brief for State of Washington as Amicus Curiae A-5 to A-6.