PER CURIAM!
This matter is before us as an expedited appeal.
We must decide whether the district court erred in entering summary judgment on its own motion in this matter in favor of the State of Washington. We conclude that the district court acted improvidently because material issues of fact remain unresolved.
Under Rule 56(c) of the Federal Rules of Civil Procedure, a summary judgment may not be granted if there are genuine issues of material fact that are not resolved by the affidavits submitted by the parties. Jewel Companies, Inc. v. Payless Drug Stores Northwest, Inc., 741 F.2d 1555, slip op. at 3916 (9th Cir.1984).
In granting summary judgment, the district court declared that Washington Rev. Code § 29.51.020(l)(e) did not violate the first and fourteenth amendments to the Constitution. Section 29.51.020(l)(e) provides that no person may, “[o]n the day of any primary, general or special election, ... within a polling place, or in any public area within three hundred feet of any entrance to such polling place: ... Conduct any exit poll or public opinion poll with voters.”
We have reviewed only the briefs and the excerpts of record. Because this appeal was expedited so that we could attempt to resolve the propriety of the issuance of the summary judgment by election day, we have not yet received or reviewed the full record presented to the district court. We are satisfied from a review of the partial record presented to us by the parties, that the following issues must be resolved by the trial court before an appellate court can reach the merits of appellants’ constitutional challenge.
Disputed Issues of Fact.
We have discerned the following genuine issues of material fact that must be resolved before review is appropriate in this ease.
One. What is an exit poll as that term is used in the statute?
*1253Two. Assuming an exit poll is limited to the random handing out of questionnaires as voters leave the polls, is such conduct disruptive to peace, order, and decorum and therefore subject to regulation under Mills v. Alabama, 384 U.S. 214, 218, 86 S.Ct. 1434, 1436, 16 L.Ed.2d 484 (1965)?
Three. Can statistically reliable polls be conducted outside a 300-foot radius? The State of Washington suggests that a telephone poll would be as effective. Appellants claim an exit poll is the only feasible method.
Four. What is the minimum area around a polling place in which to ensure that potential voters are not disrupted — the voting booth, the room where the booth is located, the public sidewalk around the building, the adjacent public street?
Five. Does the presence of clusters of media pollsters — representing appellants ABC, NBC, CBS, The Daily Herald and The New York Times accosting exiting voters— discourage other persons from casting their ballots?
Six. Did the State of Washington, in enacting its ban on exit polls, intend to remedy disruption in and around polling places or was its true motive the prevention of the projection and predication of election results prior to the closing of the polls?
Seven. Does the alleged “under inclusiveness” of the statute cast a doubt concerning the evil Washington is seeking to remedy? Which was the primary purpose of the regulation — disruption of peace, order, and decorum in and around polling places to prevent harassment of potential voters — or suppression of the content of the expression?
Eight. If the State’s true motive in enacting section 29.51.020(l)(e) was to prevent disruption in and around polling places, could it have done so through regulations that are less intrusive upon activities protected by the first amendment? The parties agree that the state may enact regulations to insure peace, order and decorum in and around polling places. See Brown v. Hartlage, 456 U.S. 45, 52, 102 S.Ct. 1523, 1528, 71 L.Ed.2d 732 (1982) (“the states have a legitimate interest in preserving the integrity'of their electoral process”). We cannot determine the serious constitutional questions raised by appellants in their action for a declaration that the Washington statute is unconstitutional and that its enforcement should be enjoined without a trial on the merits of the foregoing issues.
As this court said in Kuzinich v. County of Santa Clara, 689 F.2d 1345 (9th Cir.1982) (as amended in 1983), “A summary judgment is proper here only if it can be said that viewing the evidence in the light most favorable to [the appellants] it appears as a matter of law that the purpose of the ordinance was to serve some substantial public interest unrelated to the suppression of free speech.” Id. at 1348-1349.
We cannot say that the State of Washington was entitled to a summary judgment as a matter of law. The evidence does not foreclose the possibility that the State of Washington was motivated by an intent to suppress “constitutionally protected ideas.” Board of Education v. Pico, 457 U.S. 853, 102 S.Ct. 2799, 2812, 73 L.Ed.2d 435 (1982).
Because there are genuine issues of material fact as to whether this statute was necessary to protect the peace, order, and decorum, summary judgment should not have been granted.
The judgment of the district court granting summary judgment on its own motion on .behalf of the State of Washington is reversed.
This matter is remanded to the district court for a trial on the merits of all unresolved factual issues.
NORRIS, Circuit Judge,
concurring in part and dissenting in part:
This case is an appeal from a judgment of the United States District Court for the Western District of Washington declaring Washington Revised Code § 29.51.020(l)(e) *1254to be constitutional. The statute, enacted during the 1983 session of the Washington legislature, makes it a misdemeanor for any person “[o]H the day of any primary, general or special election, ... within a polling place, or in any public-area within three hundred feet” thereof, to “[c]onduct any exit poll or public opinion poll with voters.” The question presented is whether the statute violates the First and Fourteenth Amendments to the United States Constitution.
I agree with the decision of the majority to vacate the summary judgment for the State of Washington, but I would direct entry of summary judgment for the appellants on remand. The majority’s laundry list of factual issues constitutes an amalgam of questions of law and questions of fact. The majority, however, absolutely fails to cite any evidence introduced by the State that gives rise to a triable issue of fact. Although the State did make contentions and assertions about issues of fact, it did not offer any evidence that gives rise to material issues of fact. I agree with Judge Tanner that there are no material issues of fact in this case. The constitutional question is squarely before us for decision, just as it was before him.
I
Appellants American Broadcasting Companies, Inc. (“ABC”), CBS Inc. (“CBS”), National Broadcasting Company, Inc. (“NBC”) and The New York Times Company (“The New York Times”) engage in the practice of conducting exit polls, a type of polling which involves asking voters to fill, out questionnaires as they leave the polling place. It is undisputed that this practice falls within the prohibition against “exit polls” set forth in Wash.Rev.Code § 29.51.-020(l)(e).
Appellee Ralph Munro (“Munro”) is the Secretary of State of the State of Washington and was named as a defendant in this action in his official capacity. As Secretary of State, Munro is the chief election officer of the State. Appellee Ken 0. Eikenberry (“Eikenberry”) is the Attorney General of Washington and was named as a defendant in this action in his official capacity. As Attorney General, Eikenberry is charged, inter alia, with enforcing the criminal laws of the State. (For convenience, appellees will be referred to herein as “the State” or “Washington.”)
Appellants instituted this action by filing a complaint seeking a declaratory judgment that the statutory ban on exit polling in a 300 foot zone surrounding Washington’s polling places violates the First and Fourteenth Amendments to the United States Constitution, and a permanent injunction restraining enforcement of the statute. Washington filed an answer and amended counterclaim responding to appellants’ claims and seeking an order enjoining appellants from violating the statute.
Shortly thereafter, appellants filed a motion for summary judgment pursuant to Fed.R.Civ.P. 56 supported by affidavits which explain the distinguishing character*1255istics of exit polls and how the poll data are used. In opposition, Washington filed an affidavit by Secretary of State Munro stating, “that exit polling detrimentally affects voting both through the physical activity of conducting it and through the dissemination of result predictions purportedly based upon it;” and an affidavit by Don Whiting, Supervisor of Elections in the Office of the Secretary of State, authenticating a transcript of a legislative hearing relating to the 1983 amendments expressly banning exit polls in a 300 foot zone. Washington filed no affidavits controverting the statements of fact set forth in appellants’ affidavits. Similarly, appellants submitted no affidavits controverting those offered by the State. Thus the record before Judge Tanner at the summary judgment hearing included two distinct sets of affidavits, each of which was uncontroverted.
After a two-day hearing, Judge Tanner entered a judgment declaring Wash.Rev. Code 29.51.020(l)(e) to be constitutional. His findings and conclusions of law, in their entirety, were:
1. There is no genuine issue as to any material fact that is still in controversy between the parties.
2. Wash.Rev.Code 29.51.020(l)(e) incidentally restricts the Plaintiffs’ expressive activities and inconveniences Plaintiffs’ ability to gather news.
3. The State of Washington has a legitimate and compelling interest in maintaining peace, order and decorum in and around the election polling places, in the State of Washington, because of the fundamental nature of the voting right which is to be protected.
4. The State has met its burden of showing that Wash.Rev.Code 29.51.-020(l)(e) is the least restrictive way of carrying out this legitimate interest.
5. The statute as amended is neither vague, ambiguous, nor overbroad.
Hence, Judge Tanner not only denied appellants’ motion for summary judgment, but he awarded summary judgment to Washington. In so doing, Judge Tanner acted within his authority. A district court is entitled to grant sua sponte summary judgment to the non-moving party where, as here, there are no issues of material fact and the parties had a “full and fair opportunity to ventilate the issues involved in the motion.” Scoggins v. Boeing Co., Inc., 742 F.2d 1225 at 1227 n. 1 (9th Cir.1984) (quoting Cool Fuel, Inc. v. Connett, 685 F.2d 309, 311-12 (9th Cir.1982)).
Appellants filed a timely notice of appeal, and we granted their emergency motion to expedite the appeal because of the upcoming election. The summary judgment is a “final decision” appealable under 28 U.S.C. § 1291 (1983).
In reviewing a summary judgment, our task is identical to that of the trial court. Ward v. United States Department of Labor, 726 F.2d 516, 517 (9th Cir.1984). We review the record to determine whether there is a genuine issue as to any material fact and, if not, whether a judgment may be entered as a matter of law. Fed.R.Civ.P. 56(c).
The majority concludes that this case must be remanded for a trial because material issues of fact remain unresolved. In support, they set forth a list that is in actuality a series of questions of law and questions of fact. They totally fail, however, to cite a single piece of evidence proffered by the State that gives rise to a triable issue of fact. To the contrary, the majority goes well beyond the record and the State, and engages in speculation as to evidence that the State might have offered in opposition to appellants’ motion for summary judgment but, in fact, did not. The majority’s position is untenable.
In response to appellant’s motion for summary judgment, it was the State’s burden to offer evidence controverting the affidavits of appellants and demonstrating that material issues of fact were in dispute. As noted above, the record reveals that the State proffered no such evidence and nei*1256ther the State nor the majority is able to cite any. The State simply makes a blanket statement in its brief that it controverts the facts set forth by appellants — it cites no specific facts, no evidence and nothing in the record in support of its statement. The majority’s statements are likewise conclusory and lacking in support. For example, the majority states that Washington “suggests” that a telephone poll is as effective as an exit poll. Mere suggestions are not enough. The State offered no evidence on the issue, as it was obliged to, and the majority points to none.
I am totally at a loss as to what issues of fact require a trial. The State had ample opportunity to and, in fact, did engage in extensive discovery. It did not claim below that it needed any further discovery before Judge Tanner ruled on appellants’ summary judgment motion. The State had the opportunity and obligation to proffer evidence controverting the affidavits of appellants and demonstrating that material issues of fact were unresolved. It did not do so. I agree with Judge Tanner that there are no material issues of fact in dispute. Our task on this appeal is to review de novo the question of law whether the Washington statute violates the First Amendment. United States v. McConney, 728 F.2d 1195, 1203 (9th Cir.1984) (en banc).
Before setting forth the relevant facts, I will address Washington’s contention that this court may consider only the State’s affidavits and not those filed by appellants. Washington points out that the district court never ruled on its motion to strike the affidavits of plaintiffs. Given that our task in reviewing a summary judgment is identical to that of the district court in deciding whether to grant summary judgment, see Ward v. United States Department of Labor, id., it is appropriate and indeed judicially efficient for us to rule on Washington’s motion to strike at this time. For the reasons noted in the margin, I would deny the motion, save in one respect, as indicated.
Accordingly, for the purpose of this dissent, with the one exception noted, I accept *1257as true the statements of fact set forth in all the uncontroverted affidavits.
II
Washington Revised Code § 29.51.020 prohibits, inter alia, any electioneering, circulation of cards or handbills, the solicitation of signatures for petitions, or any disruptive behavior within a restricted area surrounding the polling place. In 1983, the legislature amended the law to expand the restricted zone from 100 to 300 feet and to expressly prohibit the conducting of any exit poll within that zone. There is no dispute that to the extent exit polling is in fact disruptive, the activity would be proscribed by subsection (l)(d) of the statute even in the absence of subsection (l)(e) expressly proscribing all exit polling.
Appellants conduct exit polls by assigning a single polling reporter to ask randomly selected voters as they leave the polling place whether they are willing to participate in the poll. If the response is yes, the interviewer gives the voter a questionnaire to fill out.
Exit polls provide accurate data about voter behavior because of the near certainty that the persons interviewed have actually voted. This reliability factor is dependent upon stationing the polling reporter closer than 300 feet from the polling place. As one expert explained,
One of the most important components of exit poll reliability is the scientific selection of the sample to be surveyed. That process, consisting of the choice to question every voter, every other voter, or voters in any other pattern, cannot be accomplished 300 feet from the polling place. It is only by maintaining the predetermined sequence that the validity of the sample can be maintained.
Affidavit of Adam Clymer, Assistant to the Executive Editor of The New York Times, 1116. Appellants stress that it is simply not feasible to conduct exit polls at a distance of 300 feet from the polling place. At that distance, the voters are so dispersed and intermingled with the general population that they cannot be interviewed on a random basis. Thus, the effect of the Wash*1258ington statute is to preclude the media from conducting statistically reliable exit polls in the State of Washington. Exit polls provide information that cannot be obtained in any other way “because they involve interviews with people who have actually voted, hot people who say they will vote, who pollsters expect to vote or who say they have voted.” Clymer Affidavit, 115.
The information obtained from exit polls is used in a variety of ways. It is well known that the networks use exit polls as a basis for projecting election results even before the polls close. Such election day projections, however, represent only one use of exit poll data. The networks analyze the polling data, for example, in reporting to their large television audiences on election night. They also use the material in post-election reports. For example,
—ABC used the results of such polls in its 1982 election night coverage, inter alia, to consider whether there was any correlation between voters’- assessment of the way President Reagan was doing his job and the Congressional candidates for whom they voted. ABC has also issued reports of its polling results after each election, which have been widely distributed to members of Congress and others, addressing issues such as the “gender gap.”
Affidavit of Stanford Opotowsky, Director of Political Operations, ABC News, till 9, 11.
—NBC used the results of its election day voter polls to report that although unemployment was an important issue in the 1982 election, most voters viewed the need to curb inflation as the most pressing problem then faced by the nation. NBC has also used the information gathered from its polls to analyze the reactions, as manifested by their vote, of blacks, women and blue collar workers to the first two years of the Reagan Presidency.
Affidavit of Gordon Manning, Vice President, News Special Programming, NBC News, II7.
The New York Times, in contrast to the networks, does not use the polling data to make election day projections. Rather, the newspaper uses the data in post-election day articles reporting, for example, that the most significant issue in the 1980 election “was the nation’s economy, especially inflation;” and that there was “a general collapse of the traditional coalition that has elected Democratic Presidents since the New Deal” in that election. Clymer Affidavit, Ml 6-8, 11.
Scholars have also utilized the information obtained from exit polls to study and report on the American electoral process generally. Each of the networks and The New York Times contribute the information gathered through their election day voter polls to The Roper Center for Public Opinion Research at the University of Connecticut. The University of Michigan is also an archive for polling data gathered by appellants. The polling data maintained by The Roper Center and the University of Michigan is made available in various forms to the academic community for use in research by a broad spectrum of scholars, educators and analysts.
In his affidavit, Everett Carll Ladd, Director of The Roper Center, evaluated exit polls as follows:
Exit polls are a priceless resource in the study of voter behavior and elections. There exists no other polling technique that provides as reliable a source of information on voters and voting behavior. One advantage that exit polls have over all other sampling techniques is that they generate a huge number of interviews as compared to other polling possibilities. This large sample allows detailed analysis of subgroups of the voting population which were previously unattainable. For example, the details of the ethno-cultural *1259dimension of voting behavior can, for the first time, be closely evaluated.
Ladd Affidavit, K12.
As the State acknowledges, the Washington statute bans all exit polls within the 300 foot zone whether or not they disrupt the peace, order and decorum of the polling place and whether or not voters are willing or unwilling to participate in the polls. In stark contrast, all other journalistic activities are permitted within the 300 foot zone provided they are conducted in an orderly manner and do not interfere with the administration of the polling place. Reporters of The Daily Herald are free to interview voters as they exit the polls because the results of the interviews are used anecdotally rather than statistically. Not only newspaper reporters, but television reporters and camera crews as well are free to enter the 300 foot zone and interview voters after they have voted, as long as the interviews are not part of an exit poll. The statute effectively says that voters may be interviewed within the 300 foot zone but they may not be counted. Thus, the distinction drawn by subsection (l)(e) between permissible and impermissible journalistic activities within the 300 foot zone is not based upon the disruptive effect of the physical activity of interviewing voters, but rather is based solely on the end use of the data obtained.
Ill
Before addressing the principal First Amendment question raised by this appeal, I would consider two preliminary contentions of the State. First, the State argues as a threshold matter that appellants’ First Amendment rights are not implicated by Wash.Rev.Code § 29.51.020(l)(e) because the statute prohibits only the gathering of news, not its dissemination. Washington claims that newsgathering is not a protected First Amendment activity. I reject this claim because it is settled law that news-gathering, as well as news reporting, is protected by the First Amendment. United States v. Sherman, 581 F.2d 1358 (9th Cir.1978); In re Express-News Corp., 695 F.2d 807 (5th Cir.1982). As the Supreme Court has observed, “without some protection for seeking out the news, freedom of the press could be eviscerated.” Branzburg v. Hayes, 408 U.S. 665, 681, 92 S.Ct. 2646, 2656, 33 L.Ed.2d 626 (1972).
As authority for its contention that news-gathering falls outside the ambit of the First Amendment, the State relies upon the so-called right of access cases. These cases merely hold, however, that the First Amendment does not guarantee the media a special right of access to areas from which the public generally is excluded. See e.g., Houchins v. KQED, 438 U.S. 1, 16, 98 S.Ct. 2588, 2597, 57 L.Ed.2d 553 (1978) (the First Amendment guarantees the media “no special right of access to the Alameda County Jail different from or greater than that accorded the public generally”); Branzburg v. Hayes, 408 U.S. 665, 684-85, 92 S.Ct. 2646, 2658, 33 L.Ed.2d 626 (1971) (consistent with the First Amendment, “the press is regularly excluded from grand jury proceedings, our own conferences, [and] the meetings of other official bodies in executive session ____ [Similarly, reporters] have no constitutional right of access to the scenes of crimes or-disaster when the general public is excluded ____”). Thus, these right of access cases offer no support for the contention that newsgathering is not a protected First Amendment activity.
Finally, the right of access cases are simply inapposite because this is not a right of access case. Appellants do not seek access to an area from which either they or the general public are excluded. Washington law does not deny anyone access to the 300 foot zone surrounding Washington polling places. Rather, the statute regulates the type of newsgathering activities that journalists may engage in while within that zone.
The State’s second preliminary argument is that exit polling is, in any case, commercial activity entitled to only limited First Amendment protection. A profit motive, however, does not diminish a journalist’s First Amendment rights. The mere fact *1260that The New York Times, for example, strives to be a profitable enterprise does not render its newsgathering and reporting activities “commercial speech” entitled to less than full First Amendment protection. As the Supreme Court has stated:
If ... profit motive were determinative,, all aspects of [a newspaper’s] operations ... would be subject to regulation if it could be established that they were conducted with a view toward increased sales. Such a basis for regulation clearly would be incompatible with the First Amendment.
Pittsburgh Press Co. v. Pittsburgh Commission on Human Relations, 413 U.S. 376, 385, 93 S.Ct. 2553, 2558, 37 L.Ed.2d 669 (1973); see also, New York Times Co. v. Sullivan, 376 U.S. 254, 266, 84 S.Ct. 710, 718, 11 L.Ed.2d 686 (1964) (that the media was paid for publishing the item at issue “is as immaterial in this connection as is the fact that newspapers and books are sold”).
IV
Having concluded that the Washington statute banning exit polls within a 300 foot zone surrounding the polling place implicates appellants’ First Amendment rights, I now reach the question whether the law abridges “freedom of speech, or of the press,” U.S. Const, amend. I, in violation of the First Amendment. As in the ease of any government restriction on freedom of expression, the statute is subject to “exacting scrutiny.” See Citizens Against Rent Control v. City of Berkeley, 454 U.S. 290, 294, 102 S.Ct. 434, 436, 70 L.Ed.2d 492 (1981). To pass constitutional muster, the statute must meet two requirements. First', it must be supported by a compelling state interest. See Brown v. Hartlage, 456 U.S. 45, 53-54, 102 S.Ct. 1523, 1528-1529, 71 L.Ed.2d 732 (1982). Second, Washington must demonstrate that the regulation is “the least restrictive means available that would accomplish the legislative purpose.” Baldwin v. Redwood City, 540 F.2d 1360, 1367 (9th Cir.1976), cert. denied, 431 U.S. 913, 97 S.Ct. 2173, 53 L.Ed.2d 223 (1977); see also Village of Schaumburg v. Citizens for a Better Environment, 444 U.S. 620, 639, 100 S.Ct. 826, 837, 63 L.Ed.2d 73 (1980); Schneider v. State, 308 U.S. 147, 162, 60 S.Ct. 146, 151, 84 L.Ed. 155 (1939).
The principal interest advanced by the State and the majority in support of the statute, and the only interest relied upon by the district court in declaring it to be constitutional, is the interest in preserving order at the polling place. It is beyond question that “the States have a legitimate interest in preserving the integrity of their electoral processes.” Brown, 456 U.S. at 52, 102 S.Ct. at 1528. Indeed, appellants acknowledge that Washington has a compelling interest in preserving “peace, order and decorum” at its polling places. See Mills v. Alabama, 384 U.S. 214, 218, 86 S.Ct. 1434, 1436, 16 L.Ed.2d 484 (1966). Thus, the critical question becomes whether Washington has met its “heavy burden” of demonstrating that “no reasonable alternatives, having a lesser impact on First Amendment freedoms,” are available to serve the State’s legitimate interests. United States v. Sherman, 581 F.2d 1358, 1361 (9th Cir.1978).
The State must show that subsection (l)(e)’s specific ban on exit polls is not redundant in light of subsection (l)(d), which states that no person may, within the 300 foot zone, “[ejngage in any practice which interferes with the freedom of voters to exercise their franchise or disrupts the administration of the polling place.” Wash.Rev.Code Ann. § 29.51.020(d) (West Supp.1984-1985). Washington has simply failed to make any such showing. Far short of meeting its burden, the State actually admits in its answer to the complaint that to the extent exit polling is in fact disruptive, the activity would have been proscribed by the statute before the amendment expressly proscribing all exit polling. Answer & Amended Counterclaim U 12. Thus, on the face of it, the ban on exit polls would appear to be “unnecessarily restrictive for the purpose [it was] designed to serve.” Baldwin, 540 F.2d at 1367 (citations omitted).
*1261Washington’s attempts to generalize about the disruptive effect of exit polls fall short. The State suggests, for example, that allowing pollsters to accost voters inside the polling place before they insert their ballots in the ballot box may disrupt the voting process. Appellants do not', however, assert a right to communicate with voters as they are voting; they only assert a right to survey voters after they have voted and are leaving the polling place. In addition the State hypothesizes that pollsters might become argumentative in a way that could rise to the level of electioneering inside the polling place. Again, however, appellants do not challenge the specific statutory provision prohibiting electioneering, Wash.Rev. Code Ann. § 29.51.020(1)(a) (West Supp.1984-1985), and the State fails to explain how a 300 foot ban on all exit polls is a narrowly tailored regulation targeted at this hypothetical evil. The State also expresses concern about confusion and harassment of voters at the polls, but the State offers no explanation as to why it could not address this problem through narrowly drawn measures protecting voters from harassment or even more spécific measures designed to avoid voter confusion by requiring that pollsters explain to voters that answering questions is voluntary and not part of the official balloting process.
In addition to demonstrating that the exclusion of journalists conducting exit polls from a zone of any size is necessary to accomplish its legitimate purposes, the State must justify making the restricted zone 300 feet in size. Enlarging the zone to 300 feet in 1983 was of critical importance to appellants’ First Amendment interests because statistically reliable exit polls cannot be conducted at that distance. Yet the State offers no justification whatsoever for the size of the restricted zone.
Washington’s effort to uphold the constitutionality of the statute banning exit polls finds no support in the legislative history. One state legislator testified that some voters had called him to complain that they were confused or intimidated by exit polls. Transcript of Portions of a Hearing of the House Committee on Constitution, Elections and Governmental Ethics on February 9, 1983, attached to Affidavit of Don Whiting, at 1-2 [hereinafter cited as Transcript of Hearing], Appellee Ralph Munro, in his capacity as Washington Secretary of State, testified about an isolated incident involving a journalist who apparently tried to conduct an exit poll right next to the ballot box. Id. at 7. What is missing from this legislative history is the crucial element — an explanation of why a ban on all exit polling within a 300 foot zone is the least restrictive means of preserving “peace, order and decorum” at polling .places in Washington.
In sum, on the record before us, Washington has failed to demonstrate that the ordinance was “narrowly drawn to serve legitimate interests of the general public,” Kuszynski v. City of Oakland, 479 F.2d 1130, 1131 (9th Cir.1973) (per curiam), and that “the incidental infringement upon First Amendment rights is no greater than is essential to vindicate its subordinating interests.” Bursey v. United States, 466 F.2d 1059, 1083 (9th Cir.1972).
Appellants argue that United States v. Sherman, 581 F.2d 1358 (9th Cir.1978), is clear authority that Washington’s statutory ban of exit polls is unconstitutional. In Sherman, our court held that a judicial order prohibiting journalists from interviewing jurors after the trial was over violated the First Amendment. Id. at 1362. Citing Sherman, the Fifth Circuit reached the same result in In re Express-News Corp., 695 F.2d 807 (5th Cir.1982). I see no principled basis for distinguishing Sherman and Express-News from the case now before us. In all three cases, a compelling governmental interest was offered as support for the restriction on newsgathering: in Sherman and Express-News, the administration of justice; here the administration of elections. As the Fifth Circuit explained in Express-News, the order banning all jur- or interviews after the trial could not pass constitutional muster because the order was,
*1262unlimited in time and in scope, applying equally to jurors willing and anxious to speak and to jurors desiring privacy, forbidding both courteous as well as uncivil communications, and foreclosing questions about a juror’s general reactions as well as specific questions about other jurors’ votes that might, under at least some circumstances, be inappropriate.
695 F.2d at 810. Similarly, the Washington statute banning exit polls applies equally to voters who may wish to tell the press how and why they voted as they did and voters who do not want to reveal their decision, forbids both “courteous as well as uncivil communications,” and forecloses journalists from conducting exit polls whether or' not they do so in a disruptive manner.
Appellants also argue that the underinclusiveness of the Washington statute makes it difficult to believe that -the evil Washington is actually trying to remedy is disruption at the polls. I agree that the statute is underinclusive in that it indiscriminately bans exit polls within the 300 foot zone, but no other voter interviews. The Supreme Court has recently indicated that such underinclusiveness raises a question as to the seriousness of the government’s asserted need to restrict freedom of expression. In F.C.C. v. League of Women Voters, — U.S. -, 104 S.Ct. 3106, 82 L.Ed.2d 278 (1984), the Court rejected a claim by the United States that a ban on editorializing by public television stations was supported by an interest in preventing the use of taxpayer monies to promote views with which the taxpayers máy disagree:
Even if this were a serious interest, it is belied by the underinclusiveness of [the statute]. The Government concedes — indeed it insists — that all sorts of controversial speech are subsidized by [the Public Broadcasting Act of 1967], and yet out of all of this potentially objectionable speech, only the expression of editorial opinion by local stations is selected for suppression. If angry taxpayers were really the central, animating concern of Congress when it passed the 1967 Act, then [the ban] does not go far enough in suppressing controversial speech in this medium. That the provision is so unrelated to this asserted purpose suggests that the Government’s interest is not substantial.
Id. at - n. 16, 104 S.Ct. at 3120 n. 16. The same point was expressed by Justice Rehnquist in a somewhat different context when he said, “It is difficult to take very seriously West Virginia's asserted need to preserve the anonymity of its youthful offenders when it permits other, equally, if not more, effective means of mass communication to distribute this information without fear of punishment.” Smith v. Daily Mail Publishing Co., 443 U.S. 97, 110, 99 S.Ct. 2667, 2674, 61 L.Ed.2d 399 (1979) (Rehnquist, J., concurring). The lack of fit between Washington’s ban on exit polls and the objective of maintaining “peace, order and decorum” at the polling place, raises a serious question whether that was “really the central, animating concern” of the Washington legislature when it passed the 1983 amendments to Section 29.51.020.
In the final analysis, Washington’s ban on exit polling is not a regulation aimed primarily at disruptive behavior, but rather is one aimed at the use to which information is put. This makes defense of the statute difficult because the end use of the polling data bears no relationship to the disruptive potential of the physical act of surveying the voters as they leave the polling place. The Washington statute is thus both too restrictive — reaching exit polling that is not disruptive — and, at the same *1263time, underinclusive — failing to reach voter interviewing not part of a scientifically reliable polling process. It is “insufficiently related to the governmental interests asserted in its support to justify its interference with protected speech.” Schaumburg, 444 U.S. at 639, 100 S.Ct. at 837.
V
The State does not strenuously argue that the statutory ban on exit polls should be upheld because of the perceived adverse effect of early projections on voter turnout. Indeed, the State equivocates on the point — disclaiming on the one hand that the statute seeks to prevent the broadcasting of early projections, while arguing on the other hand that the problem is the “detrimental impact [of early projections] on voting. The projections cause a decrease in the numbers of people voting.” Appellees Brief at 16
However the State may equivocate on its legal theories in defending the constitutionality of the statute, the legislative history and the affidavit of appellee Ralph Munro make clear that the 1983 amendments were animated by concerns that election day projections had an adverse effect on voter turnout. Thus, Secretary of State Munro stresses that “dissemination of the predictions had discouraged people from voting,” Munro Affidavit at 4. His testimony at a legislative hearing reflects a specific concern that an early CBS projection of a victory for then Candidate Ronald Reagan over President Jimmy Carter resulted in a low voter turnout in the 1980 election. Transcript of Hearing at 5. An exchange between Mr. Munro and Washington Representative Ruth Fisher concerning the efforts of other Pacific Time Zone states to pass similar legislation underscores the centrality of the 1980 election to the Washington legislature. Transcript of Hearing at 11, see also id. at 23 (testimony of Thurston County Auditor Sam Reed).
In any case, the State cites no authority, and I know of none, for the proposition that government may restrict the collection and broadcasting of information about the political process out of concern for its impact on voter behavior. If the First Amendment means anything, it means that “government has no power to restrict expression because of its message, its ideas, its subject matter or its content.” Police Department v. Mosley, 408 U.S. 92, 95, 92 S.Ct. 2286, 2289, 33 L.Ed.2d 212 (1972). This is especially true when the expression relates to elections; as we said in McKinley v. City of Eloy, 705 F.2d 1110 (9th Cir.1983), “speech concerning, public'affairs is more than self-expression; it is the essence of self-government.” Id. at 1113 (quoting Garrison v. Louisiana, 379 U.S. 64, 85 S.Ct. 209, 13 L.Ed.2d 125 (1965)).
Mills v. Alabama, 384 U.S. 214, 86 S.Ct. 1434, 16 L.Ed.2d 484 (1966), is a leading case on the constitutionality of governmental interference with the flow of information about elections out of concern for its effect on voter behavior. In Mills, the Supreme Court reviewed an Alabama statute prohibiting anyone from electioneering or soliciting votes in support of or in opposition to any ballot proposition on election day. Id. at 216 n. 2, 86 S.Ct. at 1435 n. 2. *1264The Court held that the First Amendment prohibited Alabama from making “it a crime for a newspaper editor to do no more than urge people to vote one way or another in a publically held election” on election day. Id. at 220, 86 S.Ct. at 1437. In striking down the statute, the Supreme Court said, “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____ [Moreover, the] Constitution specifically selected the press ... to play an important role in the discussion ...”. Id. at 218-19, 86 S.Ct. at 1436-37 (citation omitted).
In Mills, the Supreme Court expressly rejected Alabama’s argument that the restraint on free expression was constitutional because it protected the public “ ‘from confusive last-minute charges and counter-charges and the distribution of propaganda in an effort to influence voters on an election day.’ ” 384 U.S. at 219, 86 S.Ct. at 1437. So, too, do I reject any suggestion that the Washington statute prohibiting exit polls within the 300 foot zone is constitutional because the broadcasting of early projections based upon exit polls may “influence voters on an election day.”
In the more recent decision of Brown v. Hartlage, 456 U.S. 45, 102 S.Ct. 1523, 71 L.Ed.2d 732 (1982), the Supreme Court, quoting Mills, strongly reaffirmed that the First Amendment will not tolerate suppression of information even when the state believes it useful to preserve the integrity of the election process. The Court declared,
[The First] Amendment embodies our trust in the free exchange of ideas as the means by which the people are to choose between good ideas and bad, and between candidates for political office. The State’s fear that voters might make an ill-advised choice does not provide the State with a compelling justification for limiting speech.
Id. at 60, 102 S.Ct. at 1532.
To be sure, just as in Mills v. Alabama there was a danger in permitting election day editorials, there is a danger in permitting election day projections. A voter who hears in midday that CBS has already predicted the winner may indeed be discouraged from going to the polls. It seems just as likely, however, that a midday projection that the election was too close to call — as was the case in 1960 — may spur voters to the polls who would not otherwise make the effort. In any case, the First Amendment, in guaranteeing “freedom of speech and of the press,” does not leave open to government the choice of suppressing information about the electoral process. As so aptly stated in Virginia State Board of Pharmacy v. Virginia Citizens Consumer Council, Inc., 425 U.S. 748, 770, 96 S.Ct. 1817, 1829, 48 L.Ed.2d 346 (1976), “It is precisely this kind of choice, between the dangers of suppressing information, and the dangers of its misuse if it is freely available, that the First Amendment makes for us.”
The First Amendment is a treasured legacy we all share. It reflects the wisdom of our Founding Fathers in instructing us that in a democratic society it is not the business of government to decide what we should and should not know about the political process.
I respectfully dissent.