Twenty-six years ago, this Court, in a majority opinion written by Justice Hugo L. Black, struck down a state law that made it a crime for a newspaper editor to publish an editorial on election day urging readers to vote in a particular way. Mills v. Alabama, 384 U.S. 214, 86 S.Ct. 1434, 16 L.Ed.2d 484 (1966). While the Court did not hesitate to denounce the statute as an "obvious and flagrant abridgment" of First Amendment rights, id., at 219, 86 S.Ct., at 1437, it was quick to point out that its holding "in no way involve[d] the extent of a State's power to regulate conduct in and around the polls in order to maintain peace, order and decorum there." Id., at 218, 86 S.Ct., at 1437.

Today, we confront the issue carefully left open in Mills. The question presented is whether a provision of the Tennessee Code, which prohibits the solicitation of votes and the display or distribution of campaign materials within 100 feet of the entrance to a polling place, violates the First and Fourteenth Amendments.

* The State of Tennessee has carved out an election-day "campaign-free zone" through § 2-7-111(b) of its election code. That section reads in pertinent part:

"Within the appropriate boundary as established in subsection (a) [100 feet from the entrances], and the building in which the polling place is located, the display of campaign posters, signs or other campaign materials, distribution of campaign materials, and solicitation of votes for or against any person or political party or position on a question are prohibited." Tenn.Code Ann. § 2-7-111(b) (Supp.1991).1

Violation of § 2-7-111(b) is a Class C misdemeanor punishable by a term of imprisonment not greater than 30 days or a fine not to exceed $50, or both. Tenn.Code Ann. §§ 2-19-119 and 40-35-111(e)(3) (1990).

Respondent Mary Rebecca Freeman has been a candidate for office in Tennessee, has managed local campaigns, and has worked actively in state-wide elections. In 1987, she was the treasurer for the campaign of a city-council candidate in Metropolitan Nashville-Davidson County.

Asserting that §§ 2-7-111(b) and 2-19-119 limited her ability to communicate with voters, respondent brought a facial challenge to these statutes in Davidson County Chancery Court. She sought a declaratory judgment that the provisions were unconstitutional under both the United States and the Tennessee Constitutions. She also sought a permanent injunction against their enforcement.

The Chancellor ruled that the statutes did not violate the United States or Tennessee Constitutions and dismissed respondent's suit. App. 50. He determined that § 2-7-111(b) was a content-neutral and reasonable time, place, and manner restriction; that the 100-foot boundary served a compelling state interest in protecting voters from interference, harassment, and intimidation during the voting process; and that there was an alternative channel for respondent to exercise her free-speech rights outside the 100-foot boundary. App. to Pet. for Cert. 1a.

The Tennessee Supreme Court, by a 4-to-1 vote, reversed. 802 S.W.2d 210 (1990). The court first held that § 2-7-111(b) was content-based "because it regulates a specific subject matter, the solicitation of votes and the display or distribution of campaign materials, and a certain category of speakers, campaign workers." Id., at 213. The court then held that such a content-based statute could not be upheld unless (i) the burden placed on free-speech rights is justified by a compelling state interest and (ii) the means chosen bear a substantial relation to that interest and are the least intrusive to achieve the State's goals. While the Tennessee Supreme Court found that the State unquestionably had shown a compelling interest in banning solicitation of voters and distribution of campaign materials within the polling place itself, it concluded that the State had not shown a compelling interest in regulating the premises around the polling place. Accordingly, the court held that the 100-foot limit was not narrowly tailored to protect the demonstrated interest. The court also held that the statute was not the least restrictive means to serve the State's interests. The court found less restrictive the current Tennessee statutes prohibiting interference with an election or the use of violence or intimidation to prevent voting. See Tenn.Code Ann. §§ 2-19-101 and 2-19-115 (Supp.1991). Finally, the court noted that if the State were able to show a compelling interest in preventing congestion and disruption at the entrances to polling places, a shorter radius "might perhaps pass constitutional muster." 802 S.W.2d, at 214.

Because of the importance of the issue, we granted certiorari. 498 U.S. ----, 111 S.Ct. 1578, 113 L.Ed.2d 644 (1991). We now reverse the Tennessee Supreme Court's judgment that the statute violates the First Amendment of the United States Constitution.

The First Amendment provides that "Congress shall make no law . . . abridging the freedom of speech. . . ." This Court in Thornhill v. Alabama, 310 U.S. 88, 95, 60 S.Ct. 736, 741, 84 L.Ed. 1093 (1940), said: "The freedom of speech . . . which [is] secured by the First Amendment against abridgment by the United States, [is] among the fundamental personal rights and liberties which are secured to all persons by the Fourteenth Amendment against abridgment by a State."

The Tennessee statute implicates three central concerns in our First Amendment jurisprudence: regulation of political speech, regulation of speech in a public forum, and regulation based on the content of the speech. The speech restricted by § 2-7-111(b) obviously is political speech. "Whatever differences may exist about interpretations of the First Amendment, there is practically universal agreement that a major purpose of that Amendment was to protect the free discussion of governmental affairs." Mills v. Alabama, 384 U.S., at 218, 86 S.Ct., at 1437. "For speech concerning public affairs is more than self-expression; it is the essence of self-government." Garrison v. Louisiana, 379 U.S. 64, 74-75, 85 S.Ct. 209, 216, 13 L.Ed.2d 125 (1964). Accordingly, this Court has recognized that "the First Amendment 'has its fullest and most urgent application' to speech uttered during a campaign for political office." Eu v. San Francisco Democratic Comm., 489 U.S. 214, 223, 109 S.Ct. 1013, 1020, 103 L.Ed.2d 271 (1989) (quoting Monitor Patriot Co. v. Roy, 401 U.S. 265, 272, 91 S.Ct. 621, 625, 28 L.Ed.2d 35 (1971)).

The second important feature of § 2-7-111(b) is that it bars speech in quintessential public forums. These forums include those places "which by long tradition or by government fiat have been devoted to assembly and debate," such as parks, streets, and sidewalks. Perry Education Assn. v. Perry Local Educators' Assn., 460 U.S. 37, 45, 103 S.Ct. 948, 955, 74 L.Ed.2d 794 (1983).2 "Such use of the streets and public places has, from ancient times, been a part of the privileges, immunities, rights, and liberties of citizens." Hague v. CIO, 307 U.S. 496, 515, 59 S.Ct. 954, 964, 83 L.Ed. 1423 (1939) (opinion of Roberts, J.). At the same time, however, expressive activity, even in a quintessential public forum, may interfere with other important activities for which the property is used. Accordingly, this Court has held that the government may regulate the time, place, and manner of the expressive activity, so long as such restrictions are content-neutral, are narrowly tailored to serve a significant governmental interest, and leave open ample alternatives for communication. United States v. Grace, 461 U.S. 171, 177, 103 S.Ct. 1702, 1707, 75 L.Ed.2d 736 (1983). See also Ward v. Rock Against Racism, 491 U.S. 781, 791, 109 S.Ct. 2746, 2754, 105 L.Ed.2d 661 (1989).

The Tennessee restriction under consideration, however, is not a facially content-neutral time, place, or manner restriction. Whether individuals may exercise their free-speech rights near polling places depends entirely on whether their speech is related to a political campaign. The statute does not reach other categories of speech, such as commercial solicitation, distribution, and display. This Court has held that the First Amendment's hostility to content-based regulation extends not only to a restriction on a particular viewpoint, but also to a prohibition of public discussion of an entire topic. See, e.g., Consolidated Edison Co. v. Public Service Comm'n of New York, 447 U.S. 530, 537, 100 S.Ct. 2326, 2333, 65 L.Ed.2d 319 (1980). Accord, Simon & Schuster, Inc. v. New York Crime Victims Bd., --- U.S. ----, ----, 112 S.Ct. 501, ----, 116 L.Ed.2d 476 (1991) (statute restricting speech about crime is content-based).3

As a facially content-based restriction on political speech in a public forum, § 2-7-111(b) must be subjected to exacting scrutiny: The State must show that the "regulation is necessary to serve a compelling state interest and that it is narrowly drawn to achieve that end." Perry Education Assn. v. Perry Local Educators' Assn., 460 U.S., at 45, 103 S.Ct., at 955. Accord, Board of Airport Comm'rs of Los Angeles v. Jews for Jesus, Inc., 482 U.S. 569, 573, 107 S.Ct. 2568, 2571, 96 L.Ed.2d 500 (1987); Cornelius v. NAACP Legal Defense and Ed. Fund, Inc., 473 U.S. 788, 800, 105 S.Ct. 3439, 3448, 87 L.Ed.2d 567 (1985); United States v. Grace, 461 U.S., at 177, 103 S.Ct., at 1707.

Despite the ritualistic ease with which we state this now-familiar standard, its announcement does not allow us to avoid the truly difficult issues involving the First Amendment. Perhaps foremost among these serious issues are cases that force us to reconcile our commitment to free speech with our commitment to other constitutional rights embodied in government proceedings. See, e.g., Sheppard v. Maxwell, 384 U.S. 333, 361-363, 86 S.Ct. 1507, 1521-1522, 16 L.Ed.2d 600 (1966) (outlining restrictions on speech of trial participants that courts may impose to protect an accused's right to a fair trial). This case presents us with a particularly difficult reconciliation: the accommodation of the right to engage in political discourse with the right to vote—a right at the heart of our democracy.

Tennessee asserts that its campaign-free zone serves two compelling interests. First, the State argues that its regulation serves its compelling interest in protecting the right of its citizens to vote freely for the candidates of their choice.4 Second, Tennessee argues that its restriction protects the right to vote in an election conducted with integrity and reliability.5

The interests advanced by Tennessee obviously are compelling ones. This Court has recognized that the "right to vote freely for the candidate of one's choice is of the essence of a democratic society." Reynolds v. Sims, 377 U.S. 533, 555, 84 S.Ct. 1362, 1378, 12 L.Ed.2d 506 (1964). Indeed,

"No right is more precious in a free country than that of having a choice in the election of those who make the laws under which, as good citizens, they must live. Other rights, even the most basic, are illusory if the right to vote is undermined." Wesberry v. Sanders, 376 U.S. 1, 17, 84 S.Ct. 526, 535, 11 L.Ed.2d 481 (1964).

Accordingly, this Court has concluded that a State has a compelling interest in protecting voters from confusion and undue influence. See Eu, 489 U.S., at 228-229, 109 S.Ct., at 1023.

The Court also has recognized that a State "indisputably has a compelling interest in preserving the integrity of its election process." Id., at 231, 109 S.Ct., at 1024. The Court thus has "upheld generally applicable and evenhanded restrictions that protect the integrity and reliability of the electoral process itself." Anderson v. Celebrezze, 460 U.S. 780, 788, n. 9, 103 S.Ct. 1564, 1570, n. 9, 75 L.Ed.2d 547 (1983) (collecting cases). In other words, it has recognized that a State has a compelling interest in ensuring that an individual's right to vote is not undermined by fraud in the election process.

To survive strict scrutiny, however, a State must do more than assert a compelling state interest—it must demonstrate that its law is necessary to serve the asserted interest. While we readily acknowledge that a law rarely survives such scrutiny, an examination of the evolution of election reform, both in this country and abroad, demonstrates the necessity of restricted areas in or around polling places.

During the colonial period, many government officials were elected by the viva voce method or by the showing of hands, as was the custom in most parts of Europe. That voting scheme was not a private affair, but an open, public decision, witnessed by all and improperly influenced by some. The opportunities that the viva voce system gave for bribery and intimidation gradually led to its repeal. See generally, E. Evans, A History of the Australian Ballot System in the United States 1-6 (1917) (Evans); J. Harris, Election Administration in the United States 15-16 (1934) (Harris); J. Rusk, The Effect of the Australian Ballot Reform on Split Ticket Voting, 1876-1888, 8-11 (1968) (Rusk).

Within 20 years of the formation of the Union, most States had incorporated the paper ballot into their electoral system. Initially, this paper ballot was a vast improvement. Individual voters made their own handwritten ballots, marked them in the privacy of their homes, and then brought them to the polls for counting. But the effort of making out such a ballot became increasingly more complex and cumbersome. See generally, S. Albright, The American Ballot 14-19 (1942) (Albright); Evans 5; Rusk 9-14.

Wishing to gain influence, political parties began to produce their own ballots for voters. These ballots were often printed with flamboyant colors, distinctive designs, and emblems so that they could be recognized at a distance. State attempts to standardize the ballots were easily thwarted—the vote-buyer could simply place a ballot in the hands of the bribed voter and watch until he placed it in the polling box. Thus, the evils associated with the earlier viva voce system reinfected the election process; the failure of the law to secure secrecy opened the door to bribery6 and intimidation.7 See generally, Albright 19-20; Evans 7, 11; Harris 17, 151-152; V. Key, Politics, Parties, and Pressure Groups 649 (1952); J. Reynolds, Testing Democracy: Electoral Behavior and Progressive Reform in New Jersey, 1880-1920, 36 (1988); Rusk 14-23.

Approaching the polling place under this system was akin to entering an open auction place. As the elector started his journey to the polls, he was met by various party ticket peddlers "who were only too anxious to supply him with their party tickets." Evans 9. Often the competition became heated when several such peddlers found an uncommitted or wavering voter. See L. Fredman, The Australian Ballot: The Story of an American Reform 24 (1968) (Fredman); Rusk 17. Sham battles were frequently engaged in to keep away elderly and timid voters of the opposition. See Fredman 24, 26-27; 115 North American Review 628-629 (cited in Evans 16). In short, these early elections "were not a very pleasant spectacle for those who believed in democratic government." Evans 10.

The problems with voter intimidation and election fraud that the United States was experiencing were not unique. Several other countries were attempting to work out satisfactory solutions to these same problems. Some Australian provinces adopted a series of reforms intended to secure the secrecy of an elector's vote. The most famous feature of the Australian system was its provision for an official ballot, encompassing all candidates of all parties on the same ticket. But this was not the only measure adopted to preserve the secrecy of the ballot. The Australian system also provided for the erection of polling booths (containing several voting compartments) open only to election officials, two "scrutinees" for each candidate, and electors about to vote. See J. Wigmore, The Australian Ballot System as Embodied in the Legislation of Various Countries 69, 71, 78, 79 (1889) (Wigmore) (excerpting provisions adopted by South Australia and Queensland). See generally, Albright 23; Evans 17; Rusk 23-24.

The Australian system was enacted in England in 1872 after a study by the committee of election practices identified Australia's ballot as the best possible remedy for the existing situation. See Wigmore 14-16. Belgium followed England's example in 1877. Like the Australian provinces, both England and Belgium excluded the general public from the entire polling room. See Wigmore 94, 105. See generally, Albright 23-24; Evans 17-18; Rusk 24-25.

One of the earliest indications of the reform movement in this country came in 1882 when the Philadelphia Civil Service Reform Association urged its adoption in a pamphlet entitled "English Elections." Many articles were written praising its usefulness in preventing bribery, intimidation, disorder, and inefficiency at the polls. Commentators argued that it would diminish the growing evil of bribery by removing the knowledge of whether it had been successful. Another argument strongly urged in favor of the reform was that it would protect the weak and dependent against intimidation and coercion by employers and creditors. The inability to determine the effectiveness of bribery and intimidation accordingly would create order and decency at the polls. See generally, Albright 24-26; Evans 21-23; Rusk 25-29, 42-43.

After several failed attempts to adopt the Australian system in Michigan and Wisconsin, the Louisville, Kentucky, municipal government, the Commonwealth of Massachusetts and the State of New York adopted the Australian system in 1888. The Louisville law prohibited all but voters, candidates or their agents, and electors from coming within 50 feet of the voting room inclosure. The Louisville law also provided that candidates' agents within the restricted area "were not allowed to persuade, influence, or intimidate any one in the choice of his candidate, or to attempt doing so. . . ." Wigmore 120. The Massachusetts and New York laws differed somewhat from the previous acts in that they excluded the general public only from the area encompassed within a guard rail constructed six feet from the voting compartments. See id., at 47, 128. This modification was considered an improvement because it provided additional monitoring by members of the general public and independent candidates, who in most States were not allowed to be represented by separate inspectors. Otherwise, "in order to perpetrate almost every election fraud it would only be necessary to buy up the election officers of the other party." Id., at 52. Finally, New York also prohibited any person from "electioneering on election day within any polling-place, or within one hundred feet of any polling place." Id., at 131. See generally, Evans 18-21; Rusk 26.

The success achieved through these reforms was immediately noticed and widely praised. See generally, Evans 21-24; Rusk 26-31, 42-43. One commentator remarked of the New York law of 1888:

"We have secured secrecy; and intimidation by employers, party bosses, police officers, saloonkeepers and others has come to an end.

"In earlier times our polling places were

frequently, to quote the litany, 'scenes of battle, murder, and sudden death.' This also has come to an end, and until nightfall, when the jubilation begins, our election days are now as peaceful as our Sabbaths.

"The new legislation has also rendered impossible

the old methods of frank, hardy, straightforward and shameless bribery of voters at the polls." W. Ivins, The Electoral System of the State of New York, Proceedings of the 29th Annual Meeting of the New York State Bar Association 316 (1906).8

The triumphs of 1888 set off a rapid and widespread adoption of the Australian system in the United States. By 1896 almost 90 percent of the States had adopted the Australian system. This accounted for 92 percent of the national electorate. See Rusk 30-31. See also Albright 26-28; Evans 27; Justice SCALIA, Concurring in Judgment, post, at 215, n. 1 (citations to statutes passed before 1900).

The roots of Tennessee's regulation can be traced back to two provisions passed during this period of rapid reform. Tennessee passed the first relevant provision in 1890 as part of its switch to an Australian system. In its effort to "secur[e] the purity of elections," Tennessee provided that only voters and certain election officials were permitted within the room where the election was held or within 50 feet of the entrance. The act did not provide any penalty for violation and applied only in the more highly populated counties and cities. 1890 Tenn.Pub.Acts, ch. 24, §§ 12 and 13.

The second relevant provision was passed in 1901 as an amendment to Tennessee's "Act to preserve the purity of elections, and define and punish offenses against the elective franchise." The original act, passed in 1897, made it a misdemeanor to commit various election offenses, including the use of bribery, violence or intimidation in order to induce a person to vote or refrain from voting for any particular person or measure. 1897 Tenn.Pub.Acts, ch. 14. The 1901 amendment made it a misdemeanor for any person, except the officers holding the elections, to approach nearer than 30 feet to any voter or ballot box. This provision applied to all Tennessee elections. 1901 Tenn.Pub.Acts, ch. 142.

These two laws remained relatively unchanged until 1967, when Tennessee added yet another proscription to its secret ballot law. This amendment prohibited the distribution of campaign literature "on the same floor of a building, or within one hundred (100) feet thereof, where an election is in progress." 1967 Tenn.Pub.Acts, ch. 85.

In 1972, the State enacted a comprehensive code to regulate the conduct of elections. The code included a section that proscribed the display and the distribution of campaign material and the solicitation of votes within 100 feet of the entrance to a polling place. The 1972 "campaign-free zone" is the direct precursor of the restriction challenged in the present litigation.

Today, all 50 States limit access to the areas in or around polling places. See App. to Pet. for Cert. 26a-50a; Note, Defoliating the Grassroots: Election Day Restrictions on Political Speech, 77 Geo.L.J. 2137 (1989) (summarizing statutes as of 1989). The National Labor Relations Board also limits activities at or near polling places in union-representation elections.9

In sum, an examination of the history of election regulation in this country reveals a persistent battle against two evils: voter intimidation and election fraud. After an unsuccessful experiment with an unofficial ballot system, all 50 States, together with numerous other Western democracies, settled on the same solution: a secret ballot secured in part by a restricted zone around the voting compartments. We find that this wide-spread and time-tested consensus demonstrates that some restricted zone is necessary in order to serve the States' compelling interest in preventing voter intimidation and election fraud.

Respondent and the dissent advance three principal challenges to this conclusion. First, respondent argues that restricted zones are overinclusive because States could secure these same compelling interests with statutes that make it a misdemeanor to interfere with an election or to use violence or intimidation to prevent voting. See, e.g., Tenn.Code Ann. §§ 2-19-101 and 2-19-115 (Supp.1991). We are not persuaded. Intimidation and interference laws fall short of serving a State's compelling interests because they "deal with only the most blatant and specific attempts" to impede elections. Cf. Buckley v. Valeo, 424 U.S. 1, 28, 96 S.Ct. 612, 639, 46 L.Ed.2d 659 (1976) (existence of bribery statute does not preclude need for limits on contributions to political campaigns). Moreover, because law enforcement officers generally are barred from the vicinity of the polls to avoid any appearance of coercion in the electoral process, see Tenn.Code Ann. § 2-7-103 (1985), many acts of interference would go undetected. These undetected or less than blatant acts may nonetheless drive the voter away before remedial action can be taken.

Second, respondent and the dissent argue that Tennessee's statute is underinclusive because it does not restrict other types of speech, such as charitable and commercial solicitation or exit polling, within the 100-foot zone. We agree that distinguishing among types of speech requires that the statute be subjected to strict scrutiny. We do not, however, agree that the failure to regulate all speech renders the statute fatally underinclusive. In fact, as one early commentator pointed out, allowing members of the general public access to the polling place makes it more difficult for political machines to buy off all the monitors. See Wigmore 52. But regardless of the need for such additional monitoring, there is, as summarized above, ample evidence that political candidates have used campaign workers to commit voter intimidation or electoral fraud. In contrast, there is simply no evidence that political candidates have used other forms of solicitation or exit polling to commit such electoral abuses. States adopt laws to address the problems that confront them. The First Amendment does not require States to regulate for problems that do not exist.

Finally, the dissent argues that we confuse history with necessity. Yet the dissent concedes that a secret ballot was necessary to cure electoral abuses. Contrary to the dissent's contention, the link between ballot secrecy and some restricted zone surrounding the voting area is not merely timing—it is common sense. The only way to preserve the secrecy of the ballot is to limit access to the area around the voter.10 Accordingly, we hold that some restricted zone around the voting area is necessary to secure the State's compelling interest.

The real question then is how large a restricted zone is permissible or sufficiently tailored. Respondent and the dissent argue that Tennessee's 100-foot boundary is not narrowly drawn to achieve the State's compelling interest in protecting the right to vote. We disagree.

As a preliminary matter, the long, uninterrupted and prevalent use of these statutes makes it difficult for States to come forward with the sort of proof the dissent wishes to require. The majority of these laws were adopted originally in the 1890s, long before States engaged in extensive legislative hearings on election regulations. The prevalence of these laws, both here and abroad, then encouraged their reenactment without much comment. The fact that these laws have been in effect for a long period of time also makes it difficult for the States to put on witnesses who can testify as to what would happen without them. Finally, it is difficult to isolate the exact effect of these laws on voter intimidation and election fraud. Successful voter intimidation and election fraud is successful precisely because it is difficult to detect.

Furthermore, because a government has such a compelling interest in securing the right to vote freely and effectively, this Court never has held a State "to the burden of demonstrating empirically the objective effects on political stability that [are] produced" by the voting regulation in question. Munro v. Socialist Workers Party, 479 U.S. 189, 195, 107 S.Ct. 533, 537, 93 L.Ed.2d 499 (1986).11 Elections vary from year to year, and place to place. It is therefore difficult to make specific findings about the effects of a voting regulation. Moreover, the remedy for a tainted election is an imperfect one. Rerunning an election would have a negative impact on voter turnout.12 Thus, requiring proof that a 100-foot boundary is perfectly tailored to deal with voter intimidation and election fraud

"would necessitate that a State's political system sustain some level of damage before the legislature could take corrective action. Legislatures, we think, should be permitted to respond to potential deficiencies in the electoral process with foresight rather than reactively, provided that the response is reasonable and does not significantly impinge on constitutionally protected rights." Id., at 195-196, 107 S.Ct., at 537-38 (emphasis added).

We do not think that the minor geographic limitation prescribed by § 2-7-111(b) constitutes such a significant impingement. Thus, we simply do not view the question whether the 100-foot boundary line could be somewhat tighter as a question of "constitutional dimension." Munro v. Socialist Workers Party, 479 U.S., at 197, 107 S.Ct., at 538. Reducing the boundary to 25 feet, as suggested by the Tennessee Supreme Court, 802 S.W.2d, at 214, is a difference only in degree, not a less restrictive alternative in kind. Buckley v. Valeo, 424 U.S., at 30, 96 S.Ct., at 640. As was pointed out in the dissenting opinion in the Tennessee Supreme Court, it "takes approximately 15 seconds to walk 75 feet." 802 S.W.2d, at 215. The State of Tennessee has decided that these last 15 seconds before its citizens enter the polling place should be their own, as free from interference as possible. We do not find that this is an unconstitutional choice.13

At some measurable distance from the polls, of course, governmental regulation of vote solicitation could effectively become an impermissible burden akin to the statute struck down in Mills v. Alabama, supra. See also Meyer v. Grant, 486 U.S. 414, 108 S.Ct. 1886, 100 L.Ed.2d 425 (1988) (invalidating absolute bar against the use of paid circulators). In reviewing challenges to specific provisions of a State's election laws, however, this Court has not employed any " 'litmus-paper test' that will separate valid from invalid restrictions." Anderson v. Celebrezze, 460 U.S., at 789, 103 S.Ct., at 1570 (quoting Storer v. Brown, 415 U.S. 724, 730, 94 S.Ct. 1274, 1279, 39 L.Ed.2d 714 (1974)). Accordingly, it is sufficient to say that in establishing a 100-foot boundary, Tennessee is on the constitutional side of the line.

In conclusion, we reaffirm that it is the rare case in which we have held that a law survives strict scrutiny. This, however, is such a rare case. Here, the State, as recognized administrator of elections, has asserted that the exercise of free speech rights conflicts with another fundamental right, the right to cast a ballot in an election free from the taint of intimidation and fraud. A long history, a substantial consensus, and simple common sense show that some restricted zone around polling places is necessary to protect that fundamental right. Given the conflict between these two rights, we hold that requiring solicitors to stand 100 feet from the entrances to polling places does not constitute an unconstitutional compromise.

The judgment of the Tennessee Supreme Court is reversed and the case is remanded for further proceedings not inconsistent with this opinion.

It is so ordered.

Justice THOMAS took no part in the consideration or decision of this case.

Justice KENNEDY, concurring.

Earlier this Term, I questioned the validity of the Court's recent First Amendment precedents suggesting that a State may restrict speech based on its content in the pursuit of a compelling interest. Simon & Schuster, Inc. v. New York Crime Victims Bd., 502 U.S. ----, ---- - ----, 112 S.Ct. 501, ---- - ----, 116 L.Ed.2d 476 (1991) (KENNEDY, J., concurring in judgment). Under what I deem the proper approach, neither a general content-based proscription of speech nor a content-based proscription of speech in a public forum can be justified unless the speech falls within one of a limited set of well-defined categories. See ibid. Today's case warrants some elaboration on the meaning of the term "content-based" as used in our jurisprudence.

In Simon & Schuster, my concurrence pointed out the seeming paradox that notwithstanding "our repeated statement that 'above all else, the First Amendment means that government has no power to restrict expression because of its message, its ideas, its subject matter, or its content,' " id., at ---- - ----, 112 S.Ct., at 514 (quoting Police Dept. of Chicago v. Mosley, 408 U.S. 92, 95, 92 S.Ct. 2286, 2290, 33 L.Ed.2d 212 (1972)), we had fallen into the practice of suggesting that content-based limits on speech can be upheld if confined in a narrow way to serve a compelling state interest. I continue to believe that our adoption of the compelling interest test was accomplished by accident, id., 502 U.S., at ----, 112 S.Ct., at ----, and as a general matter produces a misunderstanding that has the potential to encourage attempts to suppress legitimate expression.

The test may have a legitimate role, however, in sorting out what is and what is not a content-based restriction. See Simon & Schuster, supra, at ----, 112 S.Ct., at 514 ("we cannot avoid the necessity of deciding . . . whether the regulation is in fact content-based or content-neutral"). As the Court has recognized in the context of regulations of the time, place, or manner of speech, "[g]overnment regulation of expressive activity is content neutral so long as it is 'justified without reference to the content of the regulated speech.' " Ward v. Rock Against Racism, 491 U.S. 781, 791, 109 S.Ct. 2746, 2754, 105 L.Ed.2d 661 (1989) (quoting Clark v. Community for Creative Non-Violence, 468 U.S. 288, 293, 104 S.Ct. 3065, 3069, 82 L.Ed.2d 221 (1984)) (emphasis added in Ward ). In some cases, the fact that a regulation is content-based and invalid because outside any recognized category permitting suppression will be apparent from its face. In my view that was true of the New York statute we considered in Simon & Schuster, and no further inquiry was necessary. To read the statute was sufficient to strike it down as an effort by government to restrict expression because of its content.

Discerning the justification for a restriction of expression, however, is not always so straightforward as it was, or should have been, in Simon & Schuster. In some cases, a censorial justification will not be apparent from the face of a regulation which draws distinctions based on content, and the government will tender a plausible justification unrelated to the suppression of speech or ideas. There the compelling interest test may be one analytical device to detect, in an objective way, whether the asserted justification is in fact an accurate description of the purpose and effect of the law. This explanation of the compelling interest analysis is not explicit in our decisions; yet it does appear that in time, place, and manner cases, the regulation's justification is a central inquiry. See, e.g., Ward v. Rock Against Racism, supra, 491 U.S., at 791, 109 S.Ct., at 2754; Clark v. Community for Creative Non-Violence, supra, 468 U.S., at 293, 104 S.Ct., at 3069; Heffron v. International Society for Krishna Consciousness, Inc., 452 U.S. 640, 648-649, and n. 12, 101 S.Ct. 2559, 2564, and n. 12, 69 L.Ed.2d 298 (1981). And in those matters we do not apply as strict a requirement of narrow tailoring as in other contexts, Ward v. Rock Against Racism, supra, 491 U.S., at 797, 109 S.Ct., at 2757, although this may be because in cases like Ward, Clark, and Heffron, content neutrality was evident on the face of the regulations once the justification was identified and became itself the object of examination.

The same use of the compelling interest test is adopted today, not to justify or condemn a category of suppression but to determine the accuracy of the justification the State gives for its law. The outcome of that analysis is that the justification for the speech restriction is to protect another constitutional right. As I noted in Simon & Schuster, there is a narrow area in which the First Amendment permits freedom of expression to yield to the extent necessary for the accommodation of another constitutional right. 502 U.S., at ----, ----, ----, 112 S.Ct., at ---- - ----, ----. That principle can apply here without danger that the general rule permitting no content restriction will be engulfed by the analysis; for under the statute the State acts to protect the integrity of the polling place where citizens exercise the right to vote. Voting is one of the most fundamental and cherished liberties in our democratic system of government. The State is not using this justification to suppress legitimate expression. With these observations, I concur in the opinion of Justice BLACKMUN and the judgment of the Court.

Justice SCALIA, concurring in the judgment.

If the category of "traditional public forum" is to be a tool of analysis rather than a conclusory label, it must remain faithful to its name and derive its content from tradition. Because restrictions on speech around polling places on election day are as venerable a part of the American tradition as the secret ballot, Tenn.Code Ann. § 2-7-111 (Supp.1991) does not restrict speech in a traditional public forum, and the "exacting scrutiny" that the Court purports to apply, ante, at 198, is inappropriate. Instead, I believe that § 2-7-111, though content-based, is constitutional because it is a reasonable, viewpoint-neutral regulation of a nonpublic forum. I therefore concur in the judgment of the Court.

As the Court correctly notes, the 100-foot zone established by § 2-7-111 sometimes encompasses streets and sidewalks adjacent to the polling places. Ante, at 196, n. 2. The Court's determination that § 2-7-111 is subject to strict scrutiny is premised on its view that these areas are "quintessential public forums," having " 'by long tradition . . . been devoted to assembly and debate.' " Ante, at 196 (emphasis added). Insofar as areas adjacent to functioning polling places are concerned, that is simply not so. Statutes such as § 2-7-111 have an impressively long history of general use. Ever since the widespread adoption of the secret ballot in the late 19th century, viewpoint-neutral restrictions on election-day speech within a specified distance of the polling place—or on physical presence there—have been commonplace, indeed prevalent. By 1900, at least 34 of the 45 States (including Tennessee) had enacted such restrictions.1 It is noteworthy that most of the statutes banning election-day speech near the polling place specified the same distance set forth in § 2-7-111 (100 feet),2 and it is clear that the restricted zones often encompassed streets and sidewalks. Thus, the streets and sidewalks around polling places have traditionally not been devoted to assembly and debate.

Nothing in the public forum doctrine or in this Court's precedents warrants disregard of this longstanding tradition. "Streets and sidewalks" are not public forums in all places, see Greer v. Spock, 424 U.S. 828, 96 S.Ct. 1211, 47 L.Ed.2d 505 (1976) (streets and sidewalks on military base are not a public forum), and the long usage of our people demonstrates that the portions of streets and sidewalks adjacent to polling places are not public forums at all times either. This unquestionable tradition could be accommodated, I suppose, by holding laws such as § 2-7-111 to be covered by our doctrine of permissible "time, place, and manner" restrictions upon public forum speech—which doctrine is itself no more than a reflection of our traditions, see Perry Education Assn. v. Perry Local Educators' Assn., 460 U.S. 37, 45, 103 S.Ct. 948, 955, 74 L.Ed.2d 794 (1983). The problem with this approach, however, is that it would require some expansion of (or a unique exception to) the "time, place, and manner" doctrine, which does not permit restrictions that are not content-neutral (§ 2-7-111 prohibits only electioneering speech). Ibid. It is doctrinally less confusing to acknowledge that the environs of a polling place, on election day, are simply not a "traditional public forum"—which means that they are subject to speech restrictions that are reasonable and viewpoint-neutral. Id., at 46, 103 S.Ct., at 955.

For the reasons that the Court believes § 2-7-111 survives exacting scrutiny, ante, at 198-211, I believe it is at least reasonable; and respondent does not contend that it is viewpoint-discriminatory. I therefore agree with the judgment of the Court that § 2-7-111 is constitutional.

Justice STEVENS, with whom Justice O'CONNOR and Justice SOUTER join, dissenting.

The speech and conduct prohibited in the campaign-free zone created by Tenn.Code Ann. § 2-7-111 (Supp.1991) is classic political expression. As this Court has long recognized, "[d]iscussion of public issues and debate on the qualifications of candidates are integral to the operation of the system of government established by our Constitution. The First Amendment affords the broadest protection to such political expression in order 'to assure [the] unfettered interchange of ideas for the bringing about of political and social changes desired by the people.' " Buckley v. Valeo, 424 U.S. 1, 14, 96 S.Ct. 612, 632, 46 L.Ed.2d 659 (1976) (citation omitted). Therefore, I fully agree with the plurality that Tennessee must show that its " 'regulation is necessary to serve a compelling state interest and that it is narrowly drawn to achieve that end.' " Ante, at 198 (citations omitted). I do not agree, however, that Tennessee has made anything approaching such a showing.

* Tennessee's statutory "campaign-free zone" raises constitutional concerns of the first magnitude. The statute directly regulates political expression and thus implicates a core concern of the First Amendment. Moreover, it targets only a specific subject matter (campaign speech) and a defined class of speakers (campaign workers) and thus regulates expression based on its content. In doing so, the Tennessee statute somewhat perversely disfavors speech that normally is accorded greater protection than the kinds of speech that the statute does not regulate. For these reasons, Tennessee unquestionably bears the heavy burden of demonstrating that its silencing of political expression is necessary and narrowly tailored to serve a compelling state interest.

Statutes creating campaign-free zones outside polling places serve two quite different functions—they protect orderly access to the polls and they prevent last-minute campaigning. There can be no question that the former constitutes a compelling state interest and that, in light of our decision in Mills v. Alabama, 384 U.S. 214, 86 S.Ct. 1434, 16 L.Ed.2d 484 (1966), the latter does not. Accordingly, a State must demonstrate that the particular means it has fashioned to ensure orderly access to the polls do not unnecessarily hinder last-minute campaigning.

Campaign-free zones are noteworthy for their broad, antiseptic sweep. The Tennessee zone encompasses at least 30,000 square feet around each polling place; in some States, such as Kentucky and Wisconsin, the radius of the restricted zone is 500 feet—silencing an area of over 750,000 square feet. Even under the most sanguine scenario of participatory democracy, it is difficult to imagine voter turnout so complete as to require the clearing of hundreds of thousands of square feet simply to ensure that the path to the polling-place door remains open and that the curtain that protects the secrecy of the ballot box remains closed.

The fact that campaign-free zones cover such a large area in some States unmistakably identifies censorship of election-day campaigning as an animating force behind these restrictions. That some States have no problem maintaining order with zones of 50 feet or less strongly suggests that the more expansive prohibitions are not necessary to maintain access and order. Indeed, on its face, Tennessee's statute appears informed by political concerns. Although the statute initially established a 100-foot zone, it was later amended to establish a 300-foot zone in 12 of the State's 95 counties. As the State Attorney General observed, "there is not a rational basis" for this special treatment, for there is no "discernable reason why an extension of the boundary . . . is necessary in" those 12 counties. Brief in Opposition 4a, Tenn.Att'y Gen.Op. No. 87-185.

Moreover, the Tennessee statute does not merely regulate conduct that might inhibit voting; it bars the simple "display of campaign posters, signs, or other campaign materials." § 2-7-111(b). Bumper stickers on parked cars and lapel buttons on pedestrians are taboo. The notion that such sweeping restrictions on speech are necessary to maintain the freedom to vote and the integrity of the ballot box borders on the absurd.

The evidence introduced at trial to demonstrate the necessity for Tennessee's campaign-free zone was exceptionally thin. Although the State's sole witness explained the need for special restrictions inside the polling place itself, she offered no justification for a ban on political expression outside the polling place.1 On this record it is far from surprising that the Tennessee Supreme Court—which surely is more familiar with the State's electoral practices and traditions than we are—concluded that the 100-foot ban outside the polling place was not justified by regulatory concerns. This conclusion is bolstered by Tennessee law which indicates that normal police protection is completely adequate to maintain order in the area more than 10 feet from the polling place.2

Perhaps in recognition of the poverty of the record, the plurality—without briefing, or legislative or judicial factfinding looks to history to assess whether Tennessee's statute is in fact necessary to serve the State's interests. From its review of the history of electoral reform, the plurality finds that

"all 50 States . . . settled on the same solution: a secret ballot secured in part by a restricted zone around the voting compartments. We find that this wide-spread and time-tested consensus demonstrates that some restricted zone is necessary in order to serve the States' compelling interest in preventing voter intimidation and election fraud." Ante, at 206.

This analysis is deeply flawed; it confuses history with necessity, and mistakes the traditional for the indispensable. The plurality's reasoning combines two logical errors: First, the plurality assumes that a practice's long life itself establishes its necessity; and second, the plurality assumes that a practice that was once necessary remains necessary until it is ended.3

With regard to the first, the fact that campaign-free zones were, as the plurality indicates, introduced as part of a broader package of electoral reforms does not demonstrate that such zones were necessary. The abuses that affected the electoral system could have been cured by the institution of the secret ballot and by the heightened regulation of the polling place alone, without silencing the political speech outside the polling place.4 In my opinion, more than mere timing is required to infer necessity from tradition.

We have never regarded tradition as a proxy for necessity where necessity must be demonstrated. To the contrary, our election-law jurisprudence is rich with examples of traditions that, though longstanding, were later held to be unnecessary. For example, "[m]ost of the early Colonies had [poll taxes]; many of the States have had them during much of their histories. . . ." Harper v. Virginia Bd. of Elections, 383 U.S. 663, 684, 86 S.Ct. 1079, 1091, 16 L.Ed.2d 169 (1966) (Harlan, J., dissenting). Similarly, substantial barriers to candidacy, such as stringent petition requirements, see Williams v. Rhodes, 393 U.S. 23, 89 S.Ct. 5, 21 L.Ed.2d 24 (1968), property-ownership requirements, see Turner v. Fouche, 396 U.S. 346, 90 S.Ct. 532, 24 L.Ed.2d 567 (1970), and onerous filing fees, see Lubin v. Panish, 415 U.S. 709, 94 S.Ct. 1315, 39 L.Ed.2d 702 (1974), were all longstanding features of the electoral labyrinth.

In fact, two of our most noted decisions in this area involve, as does this case, Tennessee's electoral traditions. Dunn v. Blumstein, 405 U.S. 330, 92 S.Ct. 995, 31 L.Ed.2d 274 (1972), which invalidated Tennessee's 1-year residency requirement, is particularly instructive. Tennessee's residency requirement was indisputably "traditional," having been in place since 1870. App. in Dunn v. Blumstein, O.T.1971, No. 13, p. 22. As in this case, the State defended its law on the basis of its interest in " 'secur[ing] the freedom of elections and the purity of the ballot box.' " Id., at 23. Again like this case, Dunn involved a conflict between two rights—the right to travel and the right to vote. The Court applied strict scrutiny, ruling that residency requirements are "unconstitutional unless the State can demonstrate that such laws are 'necessary to promote a compelling governmental interest.' " 405 U.S., at 342, 92 S.Ct., at 1003 (emphasis in original) (citation omitted). Although we recognized that "[p]reservation of the 'purity of the ballot box' is a formidable-sounding state interest," id., at 345, 92 S.Ct., at 1004, we rejected the State's argument that a 1-year requirement was necessary to promote that interest. In doing so, we did not even mention, let alone find determinative, the fact that Tennessee's requirement was more than 100 years old.

In Baker v. Carr, 369 U.S. 186, 82 S.Ct. 691, 7 L.Ed.2d 663 (1962), we addressed the apportionment of Tennessee's legislature. The State's apportionment regime had remained unchanged since 1901 and was such that, by the time of trial, "40% of the voters elect[ed] 63 of the 99 members of the [state] House" of Representatives. Id., at 253, 82 S.Ct., at 729 (Clark, J., concurring). Although, as Justice Frankfurter observed in dissent, " 'very unequal' representation" had been a feature of the Nation's political landscape since colonial times, id., at 307-318, 82 S.Ct., at 759-765, the Court was not bound by this long tradition. Our other cases resemble Dunn and Baker in this way: Never have we indicated that tradition was synonymous with necessity.

Even if we assume that campaign-free zones were once somehow "necessary," it would not follow that, 100 years later, those practices remain necessary. Much in our political culture, institutions, and practices has changed since the turn of the century: Our elections are far less corrupt, far more civil, and far more democratic today than 100 years ago. These salutary developments have substantially eliminated the need for what is, in my opinion, a sweeping suppression of core political speech.

Although the plurality today blithely dispenses with the need for factual findings to determine the necessity of "traditional" restrictions on speech, courts that have made such findings with regard to other campaign-free zones have, without exception, found such zones unnecessary. See, e.g., Florida Comm. for Liability Reform v. McMillan, 682 F.Supp. 1536, 1541-1542 (MD Fla.1988); Clean-Up '84 v. Heinrich, 582 F.Supp. 125 (MD Fla.1984), aff'd, 759 F.2d 1511 (CA11 1985). Likewise, courts that have invalidated similar restrictions on so-called "exit polling" by the news media have, after careful factfinding, also declined to find such prohibitions "necessary." See, e.g., Firestone v. News-Press Publishing Co., 538 So.2d 457, 459 (Fla.1989) (invalidating Florida's 50-foot zone to the extent that it reaches outside the polling room and noting that "[a]t the evidentiary hearing, no witnesses testified of any disturbances having occurred within fifty feet of the polling room. . . . The state's unsubstantiated concern of potential disturbance is not sufficient to overcome the chilling effect on first amendment rights."); Daily Herald Co. v. Munro, 838 F.2d 380, 385, n. 8 (CA9 1988) (observing with regard to Washington's 300-foot zone that " '[t]here isn't one iota of testimony about a single voter that was upset, or intimidated, or threatened' " (quoting trial transcript)); National Broadcasting Co. v. Cleland, 697 F.Supp. 1204, 1211-1212 (ND Ga.1988); CBS Inc. v. Smith, 681 F.Supp. 794, 803 (SD Fla.1988). All of these courts, having received evidence on this issue, were far better situated than we are to assess the contemporary necessity of campaign-free zones. All of these courts concluded that such suppression of expression is unnecessary, suggesting that such zones were something of a social atavism. To my mind, this recent history, developed in the context of an adversarial search for the truth, indicates that, whatever the original historical basis for campaign-free zones may have been, their continued "necessity" has not been established. Especially when we deal with the First Amendment, when the reason for a restriction disappears, the restriction should as well.

In addition to sweeping too broadly in its reach, Tennessee's campaign-free zone selectively prohibits speech based on content. Like the statute the Court found invalid in First National Bank of Boston v. Bellotti, 435 U.S. 765, 785, 98 S.Ct. 1407, 1420, 55 L.Ed.2d 707 (1978), the Tennessee statute regulates "the subjects about which persons may speak and the speakers who may address a public issue." Within the zone, § 2-7-111 silences all campaign-related expression, but allows expression on any other subject: religious, artistic, commercial speech, even political debate and solicitation concerning issues or candidates not on the day's ballot. Indeed, as I read it, § 2-7-111 does not prohibit exit polling, which surely presents at least as great a potential interference with orderly access to the polls as does the distribution of campaign leaflets, the display of campaign posters, or the wearing of campaign buttons. This discriminatory feature of the statute severely undercuts the credibility of its purported law-and-order justification.

Tennessee's content-based discrimination is particularly problematic because such a regulation will inevitably favor certain groups of candidates. As the testimony in this case illustrates, several groups of candidates rely heavily on last-minute campaigning. See App. 22-23. Candidates with fewer resources, candidates for lower visibility offices, and "grassroots" candidates benefit disproportionately from last-minute campaigning near the polling place. See Note, Defoliating the Grassroots: Election Day Restrictions on Political Speech, 77 Geo.L.J. 2137, 2158-2160 (1989) (collecting authorities).

Although the plurality recognizes that the Tennessee statute is content-based, see ante, at 197-198, it does not inquire into whether that discrimination itself is related to any purported state interest. To the contrary, the plurality makes the surprising and unsupported claim that the selective regulation of protected speech is justified because, "[t]he First Amendment does not require States to regulate for problems that do not exist." Ante, at 207. Yet earlier this Term, the Court rejected an asserted state interest because that interest "ha[d] nothing to do with the State's" content-based distinctions among expressive activities. Simon & Schuster, Inc. v. Members of New York Crime Victims Bd., 502 U.S. ----, ----, 112 S.Ct. 501, ----, 116 L.Ed.2d 476 (1991); see also Arkansas Writers' Project, Inc. v. Ragland, 481 U.S. 221, 231, 107 S.Ct. 1722, 1728, 95 L.Ed.2d 209 (1987). Similarly in Carey v. Brown, 447 U.S. 455, 464-465, 100 S.Ct. 2286, 2292, 65 L.Ed.2d 263 (1980), the Court acknowledged Illinois' interest in "residential privacy" but invalidated that State's ban on picketing because its distinction between labor and non-labor picketing could not be "justified by reference to the State's interest in maintaining domestic tranquility." In this case the same is true: Tennessee's differential treatment of campaign speech furthers no asserted state interest. Access to and order around the polls would be just as threatened by the congregation of citizens concerned about a local environmental issue not on the ballot as by the congregation of citizens urging election of their favored candidate. Similarly, assuming that disorder immediately outside the polling place could lead to the commission of errors or the perpetration of fraud, such disorder could just as easily be caused by a religious dispute sparked by a colporteur as by a campaign-related dispute sparked by a campaign worker. In short, Tennessee has failed to point to any legitimate interest that would justify its selective regulation of campaign-related expression.

Although the plurality purports to apply "exacting scrutiny," its three marked departures from that familiar standard may have greater significance for the future than its precise holding about campaign-free zones. First, the plurality declines to take a hard look at whether a state law is in fact "necessary." Under the plurality's analysis, a State need not demonstrate that contemporary demands compel its regulation of protected expression; it need only show that that regulation can be traced to a longstanding tradition.5

Second, citing Munro v. Socialist Workers Party, 479 U.S. 189, 107 S.Ct. 533, 93 L.Ed.2d 499 (1986), the plurality lightens the State's burden of proof in showing that a restriction on speech is "narrowly tailored." In Munro, we upheld a Washington ballot-access law and, in doing so, observed that we would not "requir[e] a State to make a particularized showing of the existence of voter confusion, ballot overcrowding, or the presence of frivolous candidacies prior to the imposition of reasonable restrictions on ballot access." Id., 479 U.S., at 194-195, 107 S.Ct., at 537. We stated that legislatures "should be permitted to respond to potential deficiencies in the electoral process with foresight rather than reactively, provided that the response is reasonable and does not significantly impinge on constitutionally protected rights." Id., at 195-196, 107 S.Ct., at 537-38. I have substantial doubts about the plurality's extension of Munro § reasoning to this case, most fundamentally because I question the plurality's assumption that campaign-free zones do "not significantly impinge on constitutionally protected rights." Not only is this the very question before us, but in light of the sweep of such zones and the vital First Amendment interests at stake, I do not know how that assumption can be sound.

Third, although the plurality recognizes the problematic character of Tennessee's content-based suppressive regulation, ante, at 197-198, it nonetheless upholds the statute because "there is simply no evidence" that commercial or charitable solicitation outside the polling place poses the same potential dangers as campaigning outside the polling place. Ante, at 207. This analysis contradicts a core premise of strict scrutiny namely, that the heavy burden of justification is on the State. The plurality has effectively shifted the burden of proving the necessity of content discrimination from the State to the plaintiff.

In sum, what the plurality early in its opinion calls "exacting scrutiny," ante, at 198, appears by the end of its analysis to be neither exacting nor scrutiny. To borrow a mixed metaphor, the plurality's scrutiny is "toothless." Mathews v. Lucas, 427 U.S. 495, 510, 96 S.Ct. 2755, 2764, 49 L.Ed.2d 651 (1976).

Ours is a Nation rich with traditions. Those traditions sometimes support, and sometimes are superseded by, constitutional rules. By tradition, for example, presidential campaigns end on election eve; yet Congress certainly could not enforce that tradition by enacting a law proscribing campaigning on election day. At one time as well, bans on election-day editorial endorsements were traditional in some States,6 but Mills v. Alabama, 384 U.S. 214, 86 S.Ct. 1434, 16 L.Ed.2d 484 (1966), established that such bans are incompatible with the First Amendment.

In Mills, we set aside the conviction of a newspaper editor who violated such a ban. In doing so, we declined to accept the State's analogy between the electoral process and the judicial process, and its claim that the State could, on election day, insulate voters from political sentiments and ideas much the same way as a jury is sequestered.7 We squarely rejected the State's claim that its ban was justified by the need to protect the public " 'from confusive last-minute charges and countercharges and the distribution of propaganda in an effort to influence voters on an election day.' " Id., at 219, 86 S.Ct., at 1437 (quoting State v. Mills, 278 Ala. 188, 195-196, 176 So.2d 884, 890 (1965)). To the contrary, we recognized that it is precisely on election day that advocacy and campaigning "can be most effective." Mills, 384 U.S., at 219, 86 S.Ct., at 1437. Mills stands for the simple proposition that, tradition notwithstanding, the State does not have a legitimate interest in insulating voters from election-day campaigning. Thus, in light of Mills, the fact that campaign-free zones are "traditional" tends to undermine, rather than to support, the validity of the Tennessee statute. In short, we should scrutinize the Tennessee statute for what it is—a police power regulation that also silences a substantial amount of protected political expression.

In my opinion, the presence of campaign workers outside a polling place is, in most situations, a minor nuisance. But we have long recognized that " 'the fact that society may find speech offensive is not a sufficient reason for suppressing it.' " Hustler Magazine, Inc. v. Falwell, 485 U.S. 46, 55, 108 S.Ct. 876, 882, 99 L.Ed.2d 41 (1988) (citation omitted). Although we often pay homage to the electoral process, we must be careful not to confuse sanctity with silence. The hubbub of campaign workers outside a polling place may be a nuisance, but it is also the sound of a vibrant democracy.

In silencing that sound, Tennessee "trenches upon an area in which the importance of First Amendment protections is 'at its zenith,' " Meyer v. Grant, 486 U.S. 414, 425, 108 S.Ct. 1886, 1894, 100 L.Ed.2d 425 (1988) (citation omitted). For that reason, Tennessee must shoulder the burden of demonstrating that its restrictions on political speech are no broader than necessary to protect orderly access to the polls. It has not done so.

I therefore respectfully dissent.