American Bar Association
Forum on Communications Law
Congress, the Networks, and Exit Polls
Susan E. Seager and Laura R. Handman
In 1982, Republican Congressman Wayne Grisham of California proposed federal legislation making it a crime for any network to project the winner in any state before its polls had closed.1
Eighteen years later, W.J. "Billy" Tauzin, R-La., announced a congressional investigation into the television networks' botched Florida projections in the 2000 presidential election, but quickly conceded that he would not seek any legislation governing their coverage. "I don't think we have the legal power to constrain the networks from reporting," the congressman told journalists at his November 9 news conference. "That's why we have carefully in the past held hearings and held discussions with networks and eventually worked out what became an agreement."2 Tauzin sent a letter to the networks demanding answers to thirteen detailed questions about their newsgathering and decision- making processes on election night.3 The congressman also said he would convene public hearings in January 2001 to ask network news executives further questions about their practices, and that a system of national uniform polling hours will be considered.4
Has Congress really abandoned threats of legislation to stop networks from projecting election outcomes before all polls have closed? Probably. In the past two decades, courts have struck down as unconstitutional a series of laws aimed at stopping exit polls and choking off broadcast projections. Congress has apparently gotten the message. But Congress has not given up its attack on exit polls and election night projections, relying on the bully pulpit of congressional hearings and veiled threats to pressure the networks to delay broadcasting projections of election results.
This article reviews the legal and political developments over the past twenty years related to network projections on election night and the use of exit polls. Although courts have repeatedly struck down state laws that restrict exit poll newsgathering, members of Congress continue to use not-so-subtle coercion to persuade the networks to temporarily withhold information about elections in progress. Congressional investigations and pressure to self-regulate raise a unique set of constitutional problems. Asking media organizations to answer detailed questions about their newsgathering techniques, editorial decisions, and unpublished material raises concerns about First Amendment protections for a free press in its quest to cover a democratic government. Congressional hearings that admittedly will not lead to any legislation, coupled with veiled threats that other arms of the federal government could be used to punish noncooperative networks, could still chill the free speech rights of the networks. Indeed, the threats are not so veiled-the FCC has been asked by a Washington law firm to investigate the networks' election night "fiasco," suggesting that the networks may have violated the FCC requirement that broadcast licensees must serve the "public interest."5
Exit Polls and Voter News Service
This is not the first time that projected election results have turned out to be erroneous. The nation's first election broadcast is said to have taken place on a ham radio in 1916, when a small group of ham radio operators in New York listened to radio broadcasting pioneer Lee DeForest as he incorrectly declared that presidential hopeful Charles Evans Hughes had unseated incumbent Woodrow Wilson.6 In 1948, NBC broadcast legend H.V. Kaltenborn got it wrong when he declared that Tom Dewey would unseat Harry Truman. For the 1952 election, CBS and NBC decided to try out new technology-computers-to accurately predict that Dwight D. Eisenhower was the winner on election night.
Over the next several decades, the networks invested millions in exit polls, computers, and polling consultants, all in an effort to beat each other in declaring the winner. In early 1964, CBS pollster Lou Harris made the first use of an exit poll to enable CBS to tell viewers that Barry Goldwater had won the Republican nomination for president. Exit poll information also became rich food for academics and pollsters, who could mine the data for information on why people voted.
But the most obvious use for exit polls was to predict winners. At first, CBS explained that it was using machines and statistics to "estimate" or "project" the winner on election night. But by election night in 1964, CBS began the practice of declaring that the winner had been "elected." When Walter Cronkite announced at 9:04 p.m. EST that Lyndon Johnson had been "elected," only 20 percent of the national popular vote had been counted.7 It would turn out to be a landslide for Johnson, so the networks' early projections were never in doubt and could not have had any impact on the outcome.
Although the networks made errors in calling some elections during the 1960s and 1970s, it wasn't until the 1980 presidential election that the networks' election night practices sparked a major controversy. Using exit poll information and other indicators, NBC announced that challenger Ronald Reagan had won twenty-two states with a combined 270 electoral votes and would be the next president of the United States.8 But it was only 8:15 p.m. EST, and 5:15 p.m. in the West, where polls in at least twenty-three states remained open for another two hours and forty-five minutes.9 After ABC and CBS also declared Reagan the upset winner, Jimmy Carter made his concession speech at about 9:45 p.m. EST (6:45 p.m. PST), one hour and fifteen minutes before the polls closed in California and other Western states.
Projections and Voter Turnout
A howl of protest erupted. The loudest complaints came from Democrats, who believed that the early news of a Reagan victory depressed Democratic voter turnout in the West, causing several Democratic congressional incumbents to lose to GOP challengers. They made the same complaints in 1984 when Reagan won reelection. "If you've ever been robbed," said Al Swift, D-Wash., "you know how some West Coast voters felt in the last two presidential elections when TV networks declared a winner hours before their polls closed."10 California Secretary of State March Fong Eu declared that "Would-be voters suddenly became nonvoters after the media projections. . . . Election volunteers did not show up and voter information phones stopped ringing during a time when they are usually tied up."11
Meanwhile, the growing use of exit polls in every state was prohibitively expensive for the networks. For the 1989-1992 election cycle, the projected election unit budget for CBS alone was $21 million.12 The need to cut costs finally prompted the established networks of ABC, NBC, CBS, and CNN to pool their resources and pay for one shared exit poll data service. Eventually, Associated Press and Fox joined the consortium. The shared data service was called Voter News Service (VNS), which conducted and compiled exit poll data and actual Election Day poll returns. Individual networks still make their own election calls based on the pooled data and their own consultants, allowing room for competition in declaring the "winner" first, a practice that critics say leads to recklessness and errors such as those that occurred on election night on November 7, 2000.13 Still, until this year such errors were never on a national scale, and VNS remained a relatively obscure arm of the networks.
This year, it was the Republicans who cried foul when the major networks declared that Vice President Al Gore won the critically important State of Florida at 7:50 p.m. EST (4:50 p.m. PST), three hours and ten minutes before polls closed in the West and about an hour before they closed in the western portion of Florida's panhandle. The networks began retracting their projections for Gore at 9:55 p.m., saying their data had been faulty. Several hours later, Fox, followed by the other networks in rapid succession, declared at 2:16 a.m. that Gov. George W. Bush in fact had taken the Sunshine State and the presidency. This was followed by still another red-faced announcement just before 4 a.m. that Florida was too close to call.14
But Republicans asserted that the damage was already done, and that the illusion of an easy Gore victory chilled Republican voter desire in the West and in the Florida panhandle, causing GOP congressional candidates to lose to their Democratic challengers.15 Although some exit poll experts scoff at the charge that exit poll broadcasts can ever trigger significantly lower voter turnout, they say that the networks never should have "called" the presidential victor in Florida on election night. The data showed that the candidates were separated by less than one-quarter of 1 percent, which was too close to call.16
The Republicans are now leading the charge against the media, with Tauzin calling for a congressional investigation. "I lost some dear friends, [Congressmen Brian P.] Bilbray and [James] Rogan and perhaps even [Steven T.] Kuykendall," Tauzin said, explaining one reason for his congressional investigation of the networks' election practices. "[T]he early call in Florida may have disenfranchised, subtly, voters particularly out West, and I'm very sad about that because obviously I've lost some good friends in Congress over that; I take that very personally."17 Tauzin also complained that the early announcement about Florida "may have discouraged voters . . . in Florida," where the western panhandle straddles two time zones, and where polls were still open for another hour when the networks made their announcement at 7:50 p.m. EST that Gore had won the state.18
Given the amount of rhetoric from both major political parties, one would expect empirical studies to demonstrate that election night projections do in fact cause voter participation to plummet in Western states. But such studies are difficult to find. Tauzin admitted that he lacked "any proof that anybody didn't vote [in the 2000 election], except people who called in and said they didn't, because they thought it was over . . . it's all anecdotal."19 Martin Plissner, former political director of CBS News, said that "[n]o verifiable example has ever turned up" of any Western voter abandoning an intent to vote because of the early announcement that Carter had lost the 1980 presidential race.20 A study of Oregon voters in the November 1984 presidential election found that network projections did not influence voter turnout. Of 639 people who did not vote, fewer than 3 percent said they were influenced by early reports by the television networks.21 More recently, one Florida voter told Associated Press that he cancelled his plans to vote on November 7 after hearing the network projection that Gore would carry the state, but a local election official scoffed at his claim. "What a perfect excuse for that lazy slob," the official was quoted as saying.22
VNS Targeted for Breakup
VNS, meanwhile, has more on its hands than a congressional hearing. An antitrust advocacy group composed of academics, lawyers, and business leaders has written the Justice Department urging the breakup of the media consortium.23 A group with ties to the Republican Party filed suit in Florida against VNS and seven TV networks, seeking an injunction to block any future election night projections.24 This is somewhat ironic since it was the networks, and not VNS, that called the election for Bush. VNS went on the offensive last March when some Internet publications leaked some of its exit poll projections during the presidential primaries. VNS's lawyers fired off a "cease and desist letter," warning Slate magazine and other online publications that it would sue for copyright infringement, unlawful interference with contractual relations, and misappropriation of "hot news."25 The latter is a little-used doctrine that creates a property right in news reports. It was recognized by the Supreme Court in Associated Press v. International News Service in 191826 and given new vigor by the Second Circuit in 1997.27
Despite the paucity of hard data showing that network behavior discourages voter turnout, Congress held a series of hearings on the subject in the early 1980s. Congressman Grisham proposed his criminal statute that would bar networks from projecting the winner in any state before the polls had closed in that state.28 Colorado Senator Tim Wirth demanded that the networks suppress all announcements about election night winners until the polls had closed in the West.29 But even those with the most rudimentary First Amendment knowledge must know that laws requiring a delay in the broadcast of lawfully obtained information, in this case exit poll data, would be unconstitutional prior restraint. The U.S. Supreme Court repeatedly has held that prior restraints are "presumptively unconstitutional" and may be used, if at all, only under extremely rare and unusual circumstances.30
In the seminal Pentagon Papers decision, the Court held that the stringent constitutional requirements for justifying a prior restraint had not been satisfied, despite the government's assertion that publication of classified information about the Vietnam War would threaten national security. "[O]nly governmental allegation and proof that publication must inevitably, directly and immediately cause the occurrence of an event kindred to imperiling the safety of a [troop] transport already at sea can support even the issuance of an interim restraining order."31 The government's vague generalized interest in protecting citizens from their own failure to vote because of network projections does not come close to meeting the high standard set by the courts. Even a one-hour delay would be unconstitutional. The "loss of First Amendment rights, even for minimal periods of time, unquestionably constitutes irreparable injury."32
To avoid the insurmountable problems posed by the prior restraint doctrine, federal lawmakers have toyed with adopting proposed uniform poll hours across the nation. There has been disagreement among the states, however, over which hours would best serve both coasts, and some critics contend that Congress lacks the constitutional authority to mandate election hours for individual states.33 State legislatures, meanwhile, decided to attack the problem from a different angle by passing dozens of laws to ban exit polls.
State Bans on Exit Polls
The U.S. Supreme Court has not ruled directly on the issue, but lower federal courts and state courts have struck down every exit poll restriction challenged by the media. The Ninth Circuit issued the only federal appeals court decision in Daily Herald v. Munro.34 At issue was a statute passed by the Washington State Legislature in 1983 that prohibited anyone from coming within 300 feet of a polling place to conduct "any exit poll or public opinion poll with voters."35 The restriction was challenged by a local newspaper, the Daily Herald, as well as by the New York Times, ABC, and CBS. Judge Warren J. Ferguson, writing for the majority, began his analysis by finding that the taking of the poll, and not just dissemination of its results, was protected by the First Amendment because it involved citizens discussing their political choices. "A major purpose of the First Amendment was to protect the free discussion of governmental affairs," the court said. "This of course includes discussions of candidates."36 In the second step of his analysis, Judge Ferguson found that the First Amendment was also implicated because reporters were gathering information, and "the First Amendment protects the media's right to gather news."37
The court found that the statute was a content-based restriction on speech at a public forum (the polling place, sidewalks, and streets) and, as such, was "presumptively" unconstitutional unless it was narrowly tailored to promote a compelling government interest. Although the state had an interest38 in "maintaining peace, order, and decorum" at the polls, and "preserving the integrity of their electoral processes," the exit poll ban was not narrowly tailored to advance this interest. The state already banned disruptive conduct at the polls with another statute. It could have adopted more narrowly tailored statutes to curb disruptive exit pollsters, such as requiring polls to have separate exits and entrances or reducing the size of the restrictive area. The court declined, however, to say whether these adjustments would pass constitutional muster.
Scrutinizing the statute's legislative history, the lower court had also found that the legislature's asserted purpose of protecting decorum at the polls was in fact a pretext, and that the true purpose of the ban was to prevent the broadcasting of early election returns, which some state officials believed might influence voter turnout. The court rejected such a "general interest in insulating voters from outside influences" as "insufficient to justify speech regulation" and "impermissible."39 The appellate court cited Mills v. Alabama,40 in which the U.S. Supreme Court had struck down a criminal statute forbidding anyone from electioneering on election day as applied to the prosecution of a newspaper for endorsing a candidate for president in an election day editorial.41 Washington State's statute also flunked the strict scrutiny test because its broad reach effectively blocked all uses of the exit poll information, such as postelection newspaper stories or analyses by scholars.42
In his concurring opinion, Judge Stephen Reinhardt emphasized that it was the public dissemination of the exit poll information, not the individual discussions or newsgathering, that was the core First Amendment activity warran-ting the utmost protection. Because a "major purpose" of the First Amendment is "to protect the free discussion of governmental affairs" and ensure an "informed" public debate on politics, such a purpose would be "meaningless if the media were not allowed to obtain the information, including information of the type yielded by exit polls, on which such debate turns." Exit poll information must be protected because the data "provide[s] information not only on the outcome of the election but also on why people vote the way they did."43
The conclusions of Daily Herald v. Munro are echoed by other courts that have struck down similar statutes. Such statutes have been found to be unconstitutional because they are content-based restrictions requiring strict scrutiny, because they impermissibly restrict the media's right to gather news, and because they impermissibly restrict the public's right to talk about politics with each other and pollsters.44
Some recent U.S. Supreme Court decisions, however, signal that the Court has weakened the traditional First Amendment protections for speech in public streets, particularly on election day near the polls. In 1992, the Supreme Court voted five to three (with Justice Thomas abstaining) to uphold a Tennessee ban on all campaign activities within 100 feet of the polls in Burson v. Freeman.45 With the caveat that it was a "rare case," Justice Blackmun, writing for the plurality, deemed the statute to be a content-based restriction that survived strict scrutiny because there was a compelling state interest in remedying a history of voter intimidation and fraud outside the polls. In language that is perhaps ominous for First Amendment protection of exit polling, the Court stressed that "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."46
Writing for the dissent, Justice Stevens noted that there had been no signs of voter fraud and intimidation since the 1880s, and that the plurality "confused history with necessity" and "blithely dispensed with the need for factual findings" of alleged voter intimidation and fraud. The dissent cited the exit poll decisions by the Florida Supreme Court and lower federal courts, noting that the courts had invalidated similar restrictions on exit polling because "careful fact-finding" revealed no evidence of voter intimidation, and therefore, such prohibitions were not "necessary."47
This year, a split Court voted five to four to uphold a Colorado abortion clinic statute in Hill v. Colorado. The statute makes it a crime to approach within eight feet of a person who is entering a health care facility and communicate in any fashion without consent. Justice Stevens found the statute to be a valid content-neutral statute in terms of time, place, and matter that was justified by the patients' interests in access to the clinic and in privacy-"[t]he unwilling listener's interest in avoiding unwanted communication."48
The dissenters argued that the statute was clearly a content-based restriction that could not survive strict scrutiny. Suggesting the "deck seem[s] stacked," Justice Scalia denounced the decision as replacing "[u]ninhibited, robust and wide-open debate" with "the power of the state to protect an unheard of 'right to be left alone' on the public streets."49 In only slightly more moderate tones, Justice Kennedy stressed the First Amendment protection for the "concept of immediacy, the idea that thoughts and pleas and petitions must not be lost with passage of time," arguing that for the antiabortion speech to be "effective," it must be allowed at the time the decision is about to occur.50
Although Burson and Hill suggest that exit poll restrictions short of an outright ban might receive some support in the U.S. Supreme Court as necessary to protect the compelling interest of ballot secrecy, the Court is still likely to strike down such content-based statutes. Exit polling is conducted after a citizen has voted, and voter intimidation or fraud is no longer a danger. Overbearing behavior on the part of pollsters can be regulated by a more narrowly tailored restriction. The immediacy of the speech in exit polls is compelling: voters are readily available, their recollections are fresh, and the news is hot. Even more important, exit poll restrictions inhibit speech and information related to the workings of government and politics, a core First Amendment value implicating a broader informed citizenry. In Hill, the speech was aimed at persuading individual patients to change their minds about personal medical procedures. The media's First Amendment interest in gathering news was not a factor in either Burson or Hill. This newsgathering interest, both in the outcome and the reasons for the vote, also would weigh in favor of striking down exit poll restrictions.
In 1985, Congress decided on a new avenue of attack. Both Al Swift, D-Wash., chairman of the House Task Force on Elections, and William M. Thomas, R-Cal., the committee's ranking Republican, had a vested interest in trying to keep the networks from doing anything that might possibly discourage voter turnout in their states. The pair decided the best way to obtain cooperation from the networks was to forge a voluntary agreement with them. Congress passed a House Concurrent Resolution asking the networks to voluntarily refrain from broadcasting election projections in each state until all of its polls had closed.51 Congressional hearings were held, and network officials pledged not to broadcast the outcome in any particular state until its polls were closed.52 But this agreement left room for maneuvering. In fact, the networks had agreed to refrain from making their projections in any one state until all or most of that state's polls had closed.53 Thus, contrary to some reports, the networks did not violate their agreement when they announced that Gore had "won" Florida at 7:50 p.m. EST, because most of the state's precincts had closed by that time.
Congressman Tauzin, who chairs the Subcommittee on Telecommunications, Trade and Consumer Protection, wants to revise the agreement so that the networks agree to delay projecting the winner of the presidential contest in each state until all of the state's polls have closed. ABC and Fox reportedly have already agreed to Tauzin's request.54
This strange, informal agreement between Congress and the networks, known as "The Pledge," is arguably just a nonbinding gratuitous promise. Tauzin has "no intent to enforce it";55 and in fact, it seems unlikely that Congress could lawfully enforce the agreement.56
Unconstitutional Congressional Hearings?
Congressional hearings pose a series of potential constitutional problems. To be sure, Congress has wide latitude to hold hearings and subpoena witnesses. Courts give enormous deference to the activities of Congress and would be loath to issue an injunction prohibiting the legislative branch from holding public hearings and launching investigations. But some courts have suggested that congressional hearings and investigations could be limited if Congress acts outside its "legitimate legislative sphere," and if the First Amendment interests outweigh the asserted congressional interests.57
Judge David Bazelon of the D.C. Circuit suggested in a 1975 law review article that the use of "lifted eyebrow" tactics by Congress to pressure the broadcast media, particularly through the use of threats of action by the FCC or other federal regulators, raises "serious issues."58 In a case directly on point, a federal district judge in California invalidated a "family viewing policy" that had been adopted by broadcast networks after Congress held a series of public hearings to limit violent and sexually oriented materials. The district court found in Writers Guild of America, West v. FCC that the FCC's use of threats of more public hearings and other measures to coerce the networks into adopting the "self-regulatory reform" was "backroom bludgeoning" that violated the First Amendment.59
Similarly, in Bantam Books, Inc. v. Sullivan,60 the U.S. Supreme Court found that the use of a Rhode Island advisory commission to inform book stores of which books it considered to be "objectionable," coupled with warnings that the committee could trigger obscenity prosecutions by local police, was a form of "informal censorship" and intimidation that violated the First and Fourteenth Amendments.61 In one of the most unusual First Amendment opinions to date, District of Columbia District Judge Gerhard Gesell issued a permanent injunction barring the publication of a congressional report that he found was written "solely for the sake of expos[ing] or intimidati[ng]" progressive political groups, ranging from the Black Panther Party to Students for a Democratic Society, and had "no relationship to any existing or future proper legislative purpose." The court in Hentoff v. Ichord62 declared that it was in no way enjoining any member of Congress from speaking and used "caution and great deference." But the court issued the injunction because the report "exceeds the legislative function of Congress" and would "inhibit free speech and assembly."63
In statements about the upcoming congressional hearings on network election reports, Tauzin's press secretary, Ken Johnson, has stressed repeatedly that Congress is not embarking "on a witch hunt" and is not considering any legislation to curb network behavior. But when asked what possible enforcement tools are available to Congress when it cannot constitutionally pass a law banning or delaying the broadcast of exit poll information or predictions, Johnson replied, "Oh, we'll just see them in front of the FCC."64 He laughed and said he was "just joking," but it is doubtful that the networks would see the humor.
In fact, a Washington, D.C., law firm has filed a formal complaint with the FCC seeking an investigation of the networks for their election night errors and asking the commission to consider sanctions "up to and including" license revocation. The law firm, which filed the complaint on behalf of itself, wants the FCC to investigate whether the networks failed to act "in the public interest," which is a requirement of an FCC license renewal under 47 U.S.C. § 309(a).65 The firm recently won a D.C. Circuit decision ordering the commission to consider a citizen's group's claim that CBS's 60 Minutes aired an allegedly "distorted" report on the Ukraine. CBS has vigorously denied the claim.66
Congress could possibly run afoul of the First Amendment by asking questions and demanding unpublished material that invade the networks' behind-the-scenes newsgathering practices and editorial decisions. Most of the courts throughout the country have recognized a qualified privilege of journalists to refuse to disclose unpublished materials or testify about newsgathering activities. The privilege applies in both civil and criminal cases, even when there is no traditional confidential source to protect. The federal reporter's privilege can be overcome only where the news organization has unpublished material that is (1) highly material and relevant, and critically needed, (2) disclosure would not unduly intrude into protected First Amendment interests, and (3) the information is not available from other sources.67
Whether the reporter's qualified privilege also applies to congressional inquiries poses a fundamental question about the source of the reporter's privilege. If it is constitutionally required, Congress would be required to observe it. If it is recognized only at common law, Congress may not be required to recognize it. According to a recent Second Circuit decision, the question of the source of the privilege is still undecided, awaiting a confrontation with Congress to bring the question to the fore.68
In his letter to the networks, Congressman Tauzin asks, "What relationship and data sharing does your organization have with VNS and what analysis in-house or otherwise is used to pro-ject a winner in a State?" Another question is, "Did your organization use different models, standards or timing to call the projected winner in Massachusetts, Texas, Virginia, Georgia, California or Florida?"69 According to the congressman's office, all of the networks responded to his letter, although some have simply said they are conducting an internal investigation without answering all of his questions.70
Possible Network Responses
Network officials could arguably refuse to answer such questions by saying that such questions are not material or relevant to potential legislation that might be considered by Congress (i.e., to institute uniform national polling hours), and that disclosure would unduly intrude on protected First Amendment interest in uninhibited, uncensored media reports on elections. In 1971, CBS challenged the legality of a congressional investigation into its documentary, "The Selling of the Pentagon," and refused to comply with a congressional subpoena for outtakes from the documentary that was critical of the Pentagon. CBS contended that the material was privileged under the First Amendment. A congressional committee recommended that CBS be cited for contempt of Congress, but the full House voted down the proposed contempt citation.71
In 1992, when the Senate launched an investigation to find out who leaked Anita Hill's sexual harassment allegations against Supreme Court nominee Clarence Thomas, the special counsel subpoenaed reporters Nina Totenberg of National Public Radio and Timothy Phelps of Newsday. Both refused to reveal their sources, saying the information was protected by the First Amendment. The special counsel then subpoenaed the reporters' home phone records, sparking media protests. Ranking members of the Senate Rules Committee intervened and said the subpoenas would not be enforced, and the investigation concluded at a cost of $550,000 without producing any information on the leaks.72 Senator Ted Stevens of Alaska noted at the time that there was "no legal precedent dealing specifically with the apparent conflict between freedom of the press guaranteed by the First Amendment and Congress's inherent constitutional power to compel testimony and documents in the pursuit of an investigation," but Stevens also hailed the decision to drop the subpoenas as a move that "affirms the First Amendment, affirms the role of an independent press in a free society and affirms the Senate's commitment to freedom of inquiry and due process. . . ."73
The U.S. Supreme Court has issued various opinions on whether private citizens can assert the First Amendment as the basis for refusing to answer questions from Congress. During the height of the nation's anticommunist fervor, the Supreme Court voted five to four to uphold a contempt of Congress citation in Barenblatt v. United States,74 finding that the congressional need for information about communist activities in education outweighed First Amendment interests.
However, in another five-to-four opinion just a few years later, the Court issued an opinion that was more protective of the First Amendment in Gibson v. Florida Legislative Investigation Committee.75 The Court held that a legislative probe that intrudes into areas protected by the First Amendment must demonstrate an "overrriding and compelling interest" and a "substantial" nexus between the information sought and the compelling interest.76 The Court made it clear that it would closely scrutinize any legislative investigation implicating the First Amendment. "It is particularly important that the exercise of the power of compulsory process be carefully circumscribed when the investigative process tends to impinge upon such highly sensitive areas as freedom of speech or press, freedom of political association, and freedom of communication of ideas."77 The Court found that the Florida legislature's contempt citation against an NAACP official for refusing to produce membership lists for a communism investigation was unconstitutional because of the lack of evidence of any link between the NAACP branch and the Communist Party.78
In the upcoming Tauzin investigation, network officials might decline to answer at least some questions posed by Congress, asserting that the questions seek privileged information about unpublished and proprietary data used by the networks in reporting on election returns. Moreover, the information about network practices has little, if any, connection to any potential legislation.
The networks, already stung by their own flip-flops on election night, may not want to risk the potential public backlash by protesting the upcoming congressional hearings or protesting their agreement with Congress to withhold information until polling places close. But if the networks chose to do so, they would have some legal support to back up their challenges to congressional authority to pry into their newsgathering activities, and to use coercion to impose what would be unconstitutional censorship if the voluntary agreement were codified as law.
Now that we know that our ballot counting system is deeply flawed by machine and human errors, the networks' exit poll information serves as a check on the accuracy of state election returns. Whether the network projections were accurate or not, it is apparent that the real problems were caused by machine malfunctions, confusing "butterfly" ballots, and other voter errors. This is not the time to permit government to censor independent sources that measure the will of the voter.
1. Anthony M. Barlow, Restricting election day Exit Polling: Freedom of Expression v. the Right to Vote, Comment, 58 U. Cin. L. Rev. 1003, 1012 n.62 (1990). S ee also Note, And the Winner Is . . . Election Day Projections and the First Amendment, 4 Cardozo Arts & Ent. L.J. 373 (1985); Note, Early Election Projection, Restrictions on Exit Polling, and the First Amendment, 3 Yale L. & Pol'y Rev. 210 (1984); Note, Exit Polls and the First Amendment, 98 Harv. L. Rev. 1927 (1985); Note , Restricting the Broadcast of Election-Day Projections: A Justifiable Protection to the Right to Vote, 9 U. Dayton L. Rev. 297 (1984).
2. Transcript of Nov. 9, 2000, news conference by Billy Tauzin, R-La., FDCH Political Transcripts, available on Lexis-Nexis (Tauzin News Conference).
3. Letter from W.J. "Billy" Tauzin, Member of Congress, to Michael Eisner, The Walt Disney Co. (Nov. 9, 2000) (on file with authors). Tauzin sent copies of this letter to the chiefs of NBC, ABC, CBS, Fox, Associated Press, and CNN. Telephone Interview with Ken Johnson, press secretary to Congressman Tauzin (Dec. 4, 2000) (Johnson Interview). Some of the questions posed by Tauzin in his letter to the networks include: "What relationship and data sharing does your organization have with VNS and what analysis in house or otherwise is used to project a winner in a state?"; "What processes are employed by your organization when a discrepancy exists between VNS data and in house election data?"; "Does your organization use the same statistical models for projecting a winner in each state?"; "Did your organization use different models, standards or timing to call the projected winner in Massachusetts, Texas, Virginia, Georgia, California or Florida?"; "Does your organization still comply with the 1985 congressional agreement?"
4. Johnson Interview, supra note 3. Johnson was not more specific about any possible legislation to nationalize polling times.
5. NBC and Fox May Drop VNS, Law Firm Seeks FCC Investigation, Communications Daily, Dec. 1, 2000, available on Westlaw, 2000 WL 4696763.
6. Martin Plissner, The Control Room: How Television Calls the Shots in Presidential Elections 69 (1999).
7. Id. at 73, 77.
8. Id. at 81-85.
9. Id. at 84; Barlow, supra note 1, at 1.
10. Barlow, supra note 1, at 1.
11. Plissner, supra note 6, at 84.
12. Id. at 87.
13. Id. at 91-100.
14. Martha T. Moore, TV, Newspaper got big one wrong: Vote projections err one way, then the other, USA Today, Nov. 9, 2000.
15. Tauzin News Conference, supra note 3, at 5. The complaints were bipartisan, however. The Democrats contend that the subsequent erroneous call for Bush led to Gore's telephone call conceding to the Texas governor, Gore's awkward retraction to Bush, and Bush's sense that his victory had been snatched out from under him. Richard Morin, Bad Call in Florida, Wash. Post, Nov. 13, 2000.
16. Interview with Martin Plissner, Dec. 6, 2000.
17. Tauzin News Conference, supra note 3, at 5.
18. Id. at 1.
19. Id. at 5.
20. Plissner, supra note 6, at 84.
21. Fred Rothenberg, Source Says Networks Agree to End Exit-Poll Characterizations, Associated Press, Jan. 17, 1985.
22. Associated Press, Group Sues over Gore Projections, Wash. Post, Nov. 14, 2000.
23. Associated Press, Breakup of Voter News Service Urged, Wash. Post, Nov. 28, 2000. The group, American Antitrust Institute, said that the group should be broken up because its pooled information caused all the networks to make the same mistake.
24. Group Sues over Gore Projections, supra note 22.
25. Jack Shafer, All the News That's Fit to Suppress, Wall St. J., Mar. 15, 2000.
26. 248 U.S. 215 (1918).
27. National Basketball Ass'n and NBA Properties, Inc. v. Motorola, Inc., 105 F.3d 841, 845 (1997).
28. Barlow, supra note 1, at 1012 n.62.
29. Plissner, supra note 6, at 85-86.
30. Nebraska Press Ass'n v. Stuart, 427 U.S. 538, 558-59 (1976) ("[a]ny prior restraint on expression comes to this Court with a 'heavy presumption' against its constitutional validity").
31. New York Times Co. v. United States, 403 U.S. 713, 714 (1971) (per curiam). See also Smith v. Daily Mail Publishing Co., 443 U.S. 97, 103-04 (1979) (government statute punishing newspaper for publishing truthful information lawfully obtained about matter of public significance violates the First Amendment "absent a need to further a state interest of the highest order"); Landmark Communications, Inc. v. Virginia, 435 U.S. 829, 838, 843 (1978) (court must not accept legislature's finding that disclosure of truthful information about secret court proceedings constituted a "clear and present danger" to the administration of justice; state may not criminally punish publication of lawfully obtained information about public officials in their public duties).
32. Elrod v. Burns, 427 U.S. 347, 373 (1976).
33. Barlow, supra note 1, at 1011-12 & nn.57-61; James R. Dickenson, Networks Limit Exit-Poll Data, Wash. Post, Jan. 18, 1985.
34. 838 F.2d 380 (9th Cir. 1988).
35. Daily Herald, 838 F.2d at 382.
36. Id. at 384.
37. Id. (quoting Branzburg v. Hayes, 408 U.S. 665, 681 (1972) ("[W]ithout some protection for seeking out the news, freedom of the press could be eviscerated.")).
38. The Court did not explicitly find that the state's interest was compelling, but did so impliedly.
39. Daily Herald, 838 F.2d at 387.
40. 384 U.S. 214, 218-20 (1966).
41. Mills, 384 U.S. at 218-20.
42. Daily Herald, 838 F.2d at 387-88.
43. Id. at 389-90. Judge Reinhardt cites Owen L. Fiss, Why the State?, 100 Harv. L. Rev., 781, 785-86 (1987), for his conclusion that the state has an "affirmative duty" to promote public debate on government.
44. See CBS v. Smith , 681 F. Supp. 794, 803 (S.D. Fla. 1988) (striking down exiting poll restriction because "newsgathering is a basic right protected by the First Amendment"); Journal Broadcasting of Kentucky, Inc. v. Logsdon, 1988 U.S. Dist. LEXIS 16864, 4 (W.D. Ky. 1988); NBC v. Colburg, 699 F. Supp. 241, 242 (D. Mont. 1988). ("[g]athering and dissemination of information concerning why and how people vote constitutes speech which is protected by the First Amendment"); NBC v. Cleland, 697 F. Supp. 1204, 1210 (N.D. Ga. 1988) ("because [the statute] restricts the freedom to speak about elections, government and politics, the statute strikes at the very core of the First Amendment"); CBS Inc. v. Growe, 15 Media Law Rep. 2275, 2278 (D. Minn. 1988) (granting preliminary injunction barring enforcement of law restricting exit polls; state's interest in protecting would-be voters from possible discouragement caused by network projections was not compelling interest and not narrowly tailored); National Broadcasting Co. v. Karpan; No. C88-0320-B (D. Wyo. 1988); Firestone v. News-Press Publishing Co., 538 So. 2d 457, 459 (Fla. 1989) (state failed to substantiate claims that exit polling disrupted voting).
45. 504 U.S. 191 (1992).
46. Burson, 504 U.S. at 208-09, 210-11.
47. Id. at 222-23 (Stevens, J., dissenting).
48. 120 S. Ct. 2480, 2494, 2499 (2000). Justice Stevens also said that the statute was narrowly drawn, allowing speakers to use their voices, signs, and other forms of communication to reach the patients from outside the "minor place restriction" created by the eight-foot, no-approach zone. Id. at 2496-97.
49. Hill, 120 S. Ct. at 2503-05, 2515 (Scalia, J., dissenting).
50. Id. at 2530 (Kennedy, J., dissenting).
51. Barlow, supra note 1, at 1012 & n.63.
52. Plissner, supra note 6, at 85; Johnson Interview, supra note 3.
53. Plissner, supra note 6, at 85; Johnson Interview, supra note 3.
54. David Hatch, ABC Blinks on Projections; Net Agrees to Wait on Calling Races, Electronic Media, Nov. 27, 2000; Johnson Interview, supra note 3.
55. Johnson Interview, supra note 3.
57. Timothy B. Dyk and Ralph E. Goldberg, The First Amendment and Congressional Investigations of Broadcast Programming, 3 J.L. & Pol. 625, 640-41 (1987). See also Robert Corn-Revere, Television Violence and the Limits of Voluntarism, 12 Yale J. on Reg. 187 (1995).
58. David L. Bazelon, FCC Regulation and the Telecommunication Press, 1975 Duke L.J. 213, 215-17 (1975), also discussed in Corn-Revere, supra note 57, at 202-03.
59. Writers Guild of America, West v. FCC, 423 F. Supp. 1064, 1149-53 (C.D. Cal. 1976). Unfortunately, the district court opinion was vacated on appeal on jurisdictional grounds. Writers Guild of America, West v. ABC, 609 F.2d 355 (9th Cir. 1979).
60. 372 U.S. 58 (1963).
61. Bantam Books, 372 U.S. at 69. A key factor in the Court's decision was that the commission's notices were followed up by a policeman's visit to the bookstores.
62. 318 F. Supp. 1175, 1182 (D.C. Cir. 1970).
63. Hentoff, 318 F. Supp. 1175, 1182 (D.C. Cir. 1970).
64. Johnson Interview, supra note 3.
65. NBC and Fox May Drop VNS, Law Firm Seeks FCC Investigation, Communications Daily, Dec. 1, 2000, available on Westlaw, 2000 WL 4696763; Interview with Arthur V. Belendiuk, Dec. 11, 2000; Complaint (filed with FCC on November 27, 2000) (on file with authors).
66. The D.C. Circuit held that the FCC acted "arbitrarily and capriciously in denying [the] petition without analyzing more precisely the evidence . . . presented" and ordered the agency to set a hearing on the challenge to CBS's license application. Serafyn v. Federal Communications Comm'n, 149 F.3d 1213, 1225 (D.C. Cir. 1998).
67. See, e.g., United States v. Cuthbertson, 630 F.2d 139 (3d Cir. 1980).
68. See Gonzales v. NBC, 186 F.3d 102, 109 n.6 (2d Cir. 1999); see also Laura R. Handman and Rebecca R. Reed, Reporter's Privilege Under Federal Law Circa 2000 and Beyond, Libel Defense Resource Center Bulletin, 1999 Issue No. 3, Part II, 70-71 (1999).
69. Tauzin Letter, supra note 3, at 2.
70. Johnson Interview, supra note 3.
71. Dyk and Goldberg, supra note 57, at 639-40.
72. Reuters, Law Firm Paid $550,000 After Failing to Find Leaks in Thomas Case, Buffalo News, Dec. 1, 1992, available on Westlaw, 1992 WL 3673438; The Senate Plumbers Close Up Shop, St. Louis Post-Dispatch, Mar. 28, 1992, available on Westlaw, 1992 WL 3522483.
73. Misguided Business: Senate Committee Decides to Quit Abusing Reporters, Houston Chron., Mar. 28, 1992, available on Westlaw, 1992 WL 11460843; Television Digest, Mar. 30, 1992, available on Westlaw, 1992 WL 2245468.
74. 360 U.S. 109, 126. (1959).
75. 372 U.S. 539 (1963).
76. Gibson, 372 U.S. at 546.
77. Id. (quoting Sweezy v. New Hampshire, 354 U.S. 234, 245 (1957) (reversing the Florida Supreme Court's finding that the contempt citation was constitutional)).
78. Id. at 540-43, 558.
Susan E. Seager (susanseager@ dwt.com) is an associate in Davis Wright Tremaine LLP's Los Angeles office. Laura R. Handman (laurahandman@ dwt.com) is a partner in the firm's Washington, D.C., and New York offices.