Skip to Content

We invite you to try out our new beta eCFR site at We’ve made big changes to make the eCFR easier to use. Be sure to leave feedback using the 'Feedback' button on the bottom right of each page!

Proposed Rule

Candidate Debates

Document Details

Information about this document as published in the Federal Register.

Document Statistics
Document page views are updated periodically throughout the day and are cumulative counts for this document. Counts are subject to sampling, reprocessing and revision (up or down) throughout the day.
Published Document

This document has been published in the Federal Register. Use the PDF linked in the document sidebar for the official electronic format.

Start Preamble Start Printed Page 15468


Federal Election Commission.


Supplemental Notice of Disposition of Petition for Rulemaking.


On February 1, 2017, the U.S. District Court for the District of Columbia ordered the Commission to reconsider its disposition of the Petition for Rulemaking filed by Level the Playing Field and to issue a new decision consistent with the Court's opinion. The Petition for Rulemaking asks the Commission to amend its regulation on candidate debates to revise the criteria governing the inclusion of candidates in presidential and vice presidential general election debates. In this supplement to the Notice of Disposition, as directed by the Court, the Commission provides further explanation of its decision to not initiate a rulemaking at this time.


March 29, 2017.


The petition and other documents relating to this matter are available on the Commission's Web site,​fosers (reference REG 2014-06), and in the Commission's Public Records Office, 999 E Street NW., Washington, DC 20463.

Start Further Info


Mr. Robert M. Knop, Assistant General Counsel, or Ms. Jessica Selinkoff, Attorney, 999 E Street NW., Washington, DC 20463, (202) 694-1650 or (800) 424-9530.

End Further Info End Preamble Start Supplemental Information


On September 11, 2014, the Commission received a Petition for Rulemaking from Level the Playing Field (“Petitioner”) regarding the Commission's regulation at 11 CFR 110.13(c). That regulation governs the criteria that debate staging organizations use for inclusion in candidate debates. The regulation, to prevent corporate spending on debates from constituting contributions to the participating candidates, requires staging organizations to “use pre-established objective criteria to determine which candidates may participate in a debate” and further specifies that, for general election debates, staging organizations “shall not use nomination by a particular political party as the sole objective criterion to determine whether to include a candidate in a debate.” 11 CFR 110.13(c). The petition asks the Commission to amend 11 CFR 110.13(c) in two respects: (1) To preclude sponsors of general election presidential and vice presidential debates from requiring that a candidate meet a polling threshold in order to be included in the debate; and (2) to require sponsors of general election presidential and vice presidential debates to have a set of objective, unbiased criteria for debate participation that do not require candidates to satisfy a polling threshold. The petition included, in addition to legal arguments, reports and other evidence in support of its position.

Procedural History

The Commission published a Notice of Availability seeking comment on the petition on November 14, 2014. Candidate Debates, 79 FR 68137. The Commission received 1264 comments in response to that notice, including one from the Petitioner that included updated and additional factual submissions. On November 20, 2015, the Commission published in the Federal Register a Notice of Disposition in which it explained why it would not initiate a rulemaking. Candidate Debates, 80 FR 72616.

The Petitioner and others sued on the basis that the Commission's failure to initiate a rulemaking was arbitrary and capricious in violation of the Administrative Procedure Act. See Level the Playing Field v. FEC, No. 15-cv-1397, 2017 WL 437400 at *1 (D.D.C. Feb. 1, 2017) (citing 5 U.S.C. 706). On February 1, 2017, the U.S. District Court for the District of Columbia concluded that the Commission acted arbitrarily and capriciously by failing to thoroughly consider the presented evidence and explain its decision; the Court ordered the Commission to reconsider its disposition of the petition and issue a new decision consistent with the Court's opinion. See id. at *13. In particular, the Court concluded that the Commission had not adequately addressed evidence concerning the 15% vote share polling threshold used by the Commission on Presidential Debates (“CPD”) as a criterion for inclusion in presidential general election debates. See id. at *12 (noting that “for thirty years [CPD] has been the only debate staging organization for presidential debates” and concluding that Commission had arbitrarily ignored evidence particular to CPD's polling criterion). The Court declined to “take the extraordinary step of ordering promulgation of a new rule,” but instead remanded for the Commission to “give the Petition the consideration it requires” and publish a new reasoned disposition or the commencement of rulemaking “if the Commission so decides.” Id. at *11, *13 (citing Shays v. FEC, 424 F. Supp. 2d 100, 116-17 (D.D.C. 2006)).

In accordance with the Court's instructions, the Commission has reconsidered the full rulemaking record. On the basis of this review, the Commission again declines to initiate a rulemaking to amend 11 CFR 110.13(c) at this time. The analysis below is intended to supplement, rather than replace, the analysis that the Commission provided in its original Notice of Disposition. 80 FR 72616.

Purpose and Requirements of Existing Candidate Debate Regulation

As the Commission stated in adopting the current candidate debate regulation in 1995, “the purpose of section 110.13 . . . is to provide a specific exception so that certain nonprofit organizations . . . and the news media may stage debates, without being deemed to have made prohibited corporate contributions to the candidates taking part in debates.” Corporate and Labor Organization Activity; Express Advocacy and Coordination with Candidates, 60 FR 64260, 64261 (Dec. 14, 1995).[1] Accordingly, the Commission has required that debate “staging organizations use pre-Start Printed Page 15469established objective criteria to avoid the real or apparent potential for a quid pro quo, and to ensure the integrity and fairness of the process.” Id. at 64262. In discussing objective selection criteria, the Commission has noted that debate staging organizations may use them to “control the number of candidates participating in . . . a meaningful debate” but must not use criteria “designed to result in the selection of certain pre-chosen participants.” Id. The Commission has further explained that while “[t]he choice of which objective criteria to use is largely left to the discretion of the staging organization,” the rule contains an implied reasonableness requirement. Id. Within the realm of reasonable criteria, the Commission has stated that it “gives great latitude in establishing the criteria for participant selection” to debate staging organizations under 11 CFR 110.13.[2] First General Counsel's Report at n.5, MUR 5530 (Commission on Presidential Debates) (May 4, 2005),​eqsdocsMUR/​000043F0.pdf.

In the first major enforcement action under this regulation almost two decades ago, the Commission found that CPD's use of polling data (among other criteria) did not result in an unlawful corporate contribution, with five Commissioners observing that it would make “little sense” if “a debate sponsor could not look at the latest poll results even though the rest of the nation could look at this as an indicator of a candidate's popularity.” MUR 4451/4473 Commission Statement of Reasons at 8 n.7 (Commission on Presidential Debates) (Apr. 6, 1998),​disclosure_​data/​mur/​4451.pdf#page=​459. Citing this statement, one court noted with respect to the use of polling thresholds as debate selection criteria that “[i]t is difficult to understand why it would be unreasonable or subjective to consider the extent of a candidate's electoral support prior to the debate to determine whether the candidate is viable enough to be included.” Buchanan v. FEC, 112 F. Supp. 2d 58, 75 (D.D.C. 2000). Nonetheless, the Commission has noted that while it cannot reasonably “question[ ] each and every . . . candidate assessment criterion,” it can evaluate “evidence that [such a] criterion was `fixed' or arranged in some manner so as to guarantee a preordained result.” MUR 4451/4473 Commission Statement of Reasons at 8-9 (Commission on Presidential Debates).

The Arguments for Changing the Regulation

The petition and many of the comments supporting it essentially argue that CPD's 15% threshold is a non-objective criterion because it is unreliable and/or intended to unfairly benefit major party candidates at the expense of independent and third-party candidates. The Court summarized the petition's arguments as attempting to establish, first, that “CPD's polling threshold is being used subjectively to exclude independent and third-party candidates” and, second, that “polling thresholds are particularly unreliable and susceptible to . . . subjective use at the presidential level, undermining the FEC's stated goal of using `objective criteria to avoid the real or apparent potential for a quid pro quo, and to ensure the integrity and fairness of the process.' ” Level the Playing Field, 2017 WL 437400 at *12.

In essence, the petition argues that there are biases against third-party and independent candidates in accurate polling, and therefore that a polling threshold requirement like CPD's presents these candidates with a Catch-22 scenario:

[A polling threshold] effectively institutionalizes the Democratic and Republican candidates as the only options with which the voters are presented. A third-party or independent candidate who is excluded from the debates loses the opportunity to take the stage against the major party nominees and demonstrate that he or she is a better alternative; the media does not cover the candidate; and the candidate does not get the public exposure necessary to compete. The “determination” that a [third-party or independent] candidate is not viable because he or she lacks a certain amount of support becomes a self-fulfilling prophecy.

Petition at 3. The petition argues that inclusion of independent and third-party candidates in presidential general election debates furthers voter education and voter turnout, which, the petition asserts, are policy purposes underlying the regulation.

Summary of Petition Evidence in Support of Changing the Regulation

In support of the argument that polling thresholds have the purpose or effect of favoring major party candidates over third-party or independent candidates, the petition presents facts and analysis regarding the name recognition required to poll at CPD's 15% threshold and the amount of money required to gain that level of name recognition. The petition provides further factual submissions that, according to the petition, show that the unreliability of polling—both generally and with respect to independent and third-party candidates—renders the 15% threshold unattainable and unreasonable for independent and third-party candidates.

The crux of the petition's factual submissions consists of two reports that purport to show that CPD's 15% threshold is designed to result in the exclusion of independent or third-party candidates. The first report, by Dr. Clifford Young, concludes that in order to reach a 15% threshold, a candidate must achieve name recognition among 60-80% of the population.[3] The second, by Douglas Schoen, estimates that the cost to a third-party or independent candidate of achieving 60% name recognition would be over $266 million, including almost $120 million for paid media content production and dissemination, which the report concludes is not a reasonably reachable figure for a non-major-party candidate.[4] Additionally, both the Young and Schoen reports conclude that polling in three-way races is inherently unreliable and not, therefore, an objective measure of the viability of third-party and independent candidates. In reaching their conclusions, both the Young and Schoen reports assert that third-party and independent candidates are disadvantaged by the fact that they do not benefit from a “party halo effect” by which Democratic and Republican candidates—regardless of name recognition—may garner a minimum vote share in polling merely for being associated with a major party, in addition to benefitting from increased name recognition from media coverage of the major party primary season.[5]

The Commission's Assessment of the Petition's Factual Submissions

1. Submissions Regarding Whether a 15% Threshold Cannot Be Attained by (and Therefore Excludes) Independent and Third-Party Candidates

The Young Report's conclusion that third-party and independent candidates require a 60-80% name recognition to meet CPD's 15% threshold does not provide a persuasive basis for changing the candidate debate regulation. Dr. Young acknowledges that his report's analysis is one-dimensional; it correlates polling results to name recognition alone, and then it draws conclusions regarding hypothetical third-party candidate performance based on that one factor. More Start Printed Page 15470specifically, Dr. Young acknowledges that polling results are not merely a function of name recognition—they are a much more complex confluence of factors. See Young Report at ¶¶ 10, 20(d) (listing other factors, beyond name recognition, affecting candidate vote share, including “fundraising, candidate positioning, election results, and idiosyncratic events”); see also Nate Silver, A Polling Based Forecast of the Republican Primary Field, FiveThirtyEight Politics (May 11, 2011) (attached to Petition as Exhibit 20) (noting that, more than name recognition, “laying the groundwork for a run quite early on,” including efforts to “hire staff, cultivate early support, brush up [ ] media skills,” predicts later vote share success). Due to the Young Report's focus on this one correlative factor, the report does not purport to establish any causative effect between name recognition and vote share, and it does not account for how external forces apart from name recognition—such as fundraising, candidate positioning, election results, and idiosyncratic events—may influence vote share. For example, the report does not take into consideration forces that might increase the vote share of an otherwise unfamiliar independent candidate—such as high unfavorable ratings among major party candidates—or forces that might decrease the vote share of an independent candidate who has become well-recognized—such as policy preferences or political missteps. Because it largely omits analysis of all other factors beyond name recognition, the Commission is not persuaded that the Young Report's conclusions are a sufficient basis on which to determine that a 15% polling threshold is so inherently unreachable by non-major-party candidates that the Commission should provide that sponsors of general election presidential debates must be prohibited as a matter of law from using it in order to fulfill the statutory prohibition on corporate contributions.

Moreover, even within the confines of name recognition, the Young Report is only weakly applicable to the debates at issue, which are presidential general election debates. The Young Report reaches its 60-80% name recognition result through three models, all of which extrapolate from data about name recognition of major party candidates at the early stages of the party primary process (i.e., before the Iowa caucuses) because, the report explains, “party halo effects” may be lower during early primary polling. Young Report at ¶ 22. The decision to measure name recognition at this extraordinarily early stage in all three models, even if only in part, may amplify polling errors, which the report notes are higher earlier in the election cycle than during the later “election salience” period—from one day to several months before election day—during which people start paying more attention to the election. Id. at ¶¶ 43(g), (i). Additionally, the use of the early party primary stage as the point of comparison for third-party or independent candidates' name recognition in September does not address or account for differences in the size of the candidate fields at those points in time. Thus, the Young Report's observations regarding early primary candidates provide little or no persuasive evidence as to the effect of a polling threshold on presidential general election candidates.

In addition, the petition appears to draw inapposite conclusions from the Young Report's data. Critically, neither the Young Report nor other evidence submitted with the petition or comments establishes that third-party or independent candidates do not or cannot meet 60-80% name recognition. In fact, at least one third-party candidate was reported to achieve over 60% name recognition in the most recent presidential campaign prior to the general election debates. See Poll Results: Third Party Candidates, YouGov (Aug. 25-26, 2016), available at​cumulus_​uploads/​document/​wc35k48hrs/​tabs_​HP_​Third_​Party_​Candidates_​20160831.pdf (showing Gary Johnson and Jill Stein having 63% and 59% name recognition among registered voters, respectively). Thus, there is no information in the rulemaking record showing that 60-80% name recognition is a prohibitively high bar for independent candidates. In other words, even if the Commission were to assume arguendo that 60-80% name recognition correlates with 15% vote share, there is no information in the record demonstrating that these thresholds inherently function to exclude third-party or independent candidates because of their party status.

Instead, the petition uses Dr. Young's name recognition threshold as a springboard to the primary argument of the Schoen Report: That the cost of achieving 15% vote share is prohibitively high for independent candidates. The Schoen Report starts from the premise that 60-80% name recognition is necessary to gain a 15% vote share and proceeds to estimate the amount of money that an independent candidate would need to spend to reach 60-80% name recognition. For the reasons stated above, the Commission does not find that this premise is adequately established by the Young Report, and therefore the Commission questions whether the Schoen Report possesses any meaningful evidentiary value. But even assuming that a candidate must reach 60-80% name recognition to achieve a 15% threshold in vote share, the Commission finds the Schoen Report not to provide a reasoned evidentiary basis for amending the rule at issue.

The Commission is unpersuaded by the Schoen Report primarily because the report builds its conclusion through an extensive series of unsupported suppositions and assertions. For example, to explain a significant portion of its calculations, the report states that “the media will not cover an independent candidate until they are certainly in the debates.” Schoen Report at 3. But the report provides no basis for this assertion other than an unexplained reference to the number of publications “follow[ing]” one particular candidate (id. at 5), and the Commission is aware of at least three non-major-party candidates who did not participate in the general election debates but received significant media attention in 2016.[6]

In another premise that the report uses to build its later conclusions, the Schoen Report asserts that independent candidates are disadvantaged because they “must resort to launching a massive national media campaign” while major party candidates “by competing in small state primaries, can build their name recognition without Start Printed Page 15471the costs of running a national campaign.” Id. In support of this statement, the report states that “Obama's 2008 victory in the Iowa caucuses catapulted him to national prominence.” Id. In fact, polling expert Nate Silver has noted that “contrary to the conventional wisdom, which holds that Barack Obama suddenly burst onto the political scene, the polling shows that he was already reasonably well-known to voters in advance of the 2008 primaries, largely as a result of his speech at the 2004 Democratic National Convention. His name was recognized by around 60 percent of primary voters by late 2006, and that figure quickly ramped up to 80 or 90 percent after he declared for the presidency in February, 2007.” Nate Silver, A Brief History of Primary Polling, Part II, FiveThirtyEight (Apr. 4, 2011),​features/​a-brief-history-of-primary-polling-part-ii/​. The only other basis that the report provides for this portion of its conclusion is the statement that Senator Rick Santorum “spent only $21,980 in [Iowa], or 73 cents per vote” in 2012. Schoen Report at 5. It is not clear how the newspaper article cited by the report derived this figure, and Schoen (despite having access to all relevant financial data through the FEC's Web site) does not appear to have assessed its accuracy. In fact, reports filed with the Commission for the period ending three days before the Iowa caucus show that Senator Santorum made disbursements of $1,906,018. Rick Santorum for President, FEC Form 3P at 4 (Jan. 31, 2012),​pdf/​317/​12950383317/​12950383317.pdf. While not all of these disbursements were targeted to Iowa, the candidate's total spending in relation to the caucuses in that state was far higher than $21,980. Even looking at only reported disbursements to Iowa payees (and, therefore, not including payments to media buyers and others outside of Iowa for activities targeted towards Iowa), the filings shows that Santorum spent over $112,000 in Iowa between October 1 and December 31, 2011, for purposes including rent, payroll, lodging, direct mail, advertising, communication consulting, and coalition building. Id. Thus, the Schoen Report's use of unexplained second-hand analysis undercuts its credibility, and the facts demonstrated by the public record give the Commission reason to doubt the Schoen Report's calculations regarding any extra benefit major party primary candidates receive from their media expenditures.

In addition, the Schoen Report states that media costs to accomplish 60% name recognition are higher in three-way races due to increased competition, and the report increases its cost estimate accordingly.[7] But the 60% figure is apparently drawn from the Young Report, which, as discussed above, addresses the very earliest stages of major party primaries. Like the Young Report, the Schoen Report does not explain why or how this 60% figure can be extrapolated from early major party primaries to three-way general elections.

The Schoen Report ultimately adopts an estimated cost of at least $100 million for a media buy that an independent candidate would require to gain the name recognition to meet the 15% threshold. Schoen Report at 6. Not only does this figure rely upon the faulty assumptions that the Commission has already noted, it is also unreliable for at least four additional reasons.

First, the $100 million figure is taken from an estimate from “a leading corporate and political media buying firm,” without any underlying data and without any explanation of the circumstances under which the firm purportedly offered that estimate. Nor does the report address (or even acknowledge) any biases in that estimate that may stem from a media buying firm's financial interest in estimating or promoting high media buy costs. The Schoen Report simply provides no evidentiary basis for the Commission to credit this third-person estimate.

Second, the $100 million estimate presumes that a candidate must go from zero percent name recognition to 60% name recognition, without noting the likelihood of a candidate starting from zero or otherwise explaining this assumption. The Schoen Report suggests, by consistently comparing the hypothetical independent candidate's position with the positions of his “two” (and only two) major party candidate competitors, that this zero percent baseline occurs at some point after the major parties have established presumptive nominees. See, e.g., Schoen Report at 10-11 (discussing “the two major party campaigns” with whom hypothetical independent candidate needing 60% name recognition will be competing for ad buy time); id. at 15 (same). A hypothetical situation in which a person with zero percent name recognition decides to run for president in approximately June of the election year and must raise name recognition from nothing to 60% within the three months before CPD looks at polls in September is unrelated to the realities of presidential elections. Presidential candidates—major party and third-party alike—generally begin campaigning a full year or more before the election, see, e.g., Jill Stein, FEC Form 2 (July 6, 2015) (declaring candidacy for president in 2016 election cycle), and they rarely start with zero name recognition, see, e.g., Petition Ex. 13 (Gallup report showing 11 candidates (including Libertarian Gary Johnson) with over 10% name recognition in January 2011). The Schoen Report's scenario—and the conclusions that the report draws from it—therefore provides no persuasive support for the petition's assertion that the candidate debate regulation must be revised.

Third, the Schoen Report bases its estimate of campaign and paid media costs on the assertion that independent candidates are unable to attract news media coverage. See Schoen Report at 4. But the report's assertion, based primarily on research published in 1999,[8] seems particularly antiquated in the age of digital and social media. See Farhad Manjoo, I Ignored Trump News for a Week. Here's What I Learned, NY Times, Feb. 22, 2017,​2017/​02/​22/​technology/​trump-news-media-ignore.html (discussing news media coverage during and since 2016 presidential election campaign in light of social media pressures). The Commission declines to promulgate rules that will govern the 2020 presidential election and beyond on the basis of opinions that are premised on such obsolete data.

Fourth, the Schoen Report's media cost estimates do not appear to take account of media purchases in support of a candidate by outside groups, including independent expenditure-only political committees (“IEOPCs”). IEOPCs may create, produce, and distribute communications in support of, but independently of, a particular candidate, and in 2016 several IEOPCs supported third-party candidate Gary Start Printed Page 15472Johnson in just that way.[9] In addition, IEOPCs may raise unlimited funds from individuals and from sources, like corporations, otherwise prohibited under the Federal Election Campaign Act, 52 U.S.C. 30101-46. Thus, the existence and rise of IEOPCs undermine the Schoen Report's assumptions about the amount of the average contribution to a candidate, as well as the report's extrapolations about the number of individual contributions needed and total sum necessary to reach Dr. Young's 60-80% name recognition threshold. See Schoen Report at 24-25 (estimating third-party candidate's “hypothetical average donation” on basis of “assumption for average donation” of “plurality” of Obama and Romney contributors under $2600 maximum).

Ultimately, the unreliability of the Schoen Report's conclusions is most clearly demonstrated by the fact that third-party candidate Gary Johnson reached 60% name recognition by August 31, 2016.[10] In the 2016 election cycle through August 31, Johnson had spent almost $5.5 million; this amount represents total disbursements for all purposes, including, but not limited to, media buys.[11] According to the Schoen Report, such a result should have been impossible: Johnson should not have been able to achieve 60% name recognition until he spent at least $266 million—fifty times more than he actually did.[12]

For all of the foregoing reasons, the Commission finds the Schoen Report unpersuasive.

Finally, the petition acknowledges that a number of third-party presidential candidates have performed sufficiently well that they were included or would have been included in debates with 15% thresholds. See Petition at 15-16. Indeed, the petition notes that as many as six candidates would apparently have satisfied this requirement at some point during their campaigns: Roosevelt in 1912, LaFolette in 1924, Thurmond in 1948, Wallace in 1968, Anderson in 1980, and Perot in 1992. Id. The petition asks the Commission to categorically disregard these examples because they predate the Internet, and in some cases, the television. Petition at 16.[13] As discussed above, the Commission agrees that pre-Internet candidacies provide only a relatively weak basis assessing how easy or difficult it would be for candidates to achieve 15% vote share in a modern election. But to the extent that the availability of Internet communication has changed this calculus, the Commission notes that advertising on the Internet can cost significantly less money than advertising in more traditional media that was available to those pre-Internet independent candidates. See, e.g. Internet Communications, 71 FR 18589, 18589 (Apr. 12, 2006) (describing Internet as “low-cost means of civic engagement and political advocacy” and noting that Internet presents minimal barriers to entry compared to “television or radio broadcasts or most other forms of mass communication”); Associated Press, Here's How Much Less than Hillary Clinton Donald Trump Spent on the Election, Fortune (Dec. 9, 2016),​2016/​12/​09/​hillary-clinton-donald-trump-campaign-spending/​ (comparing Hillary Clinton's “more traditional” television-heavy advertising strategy in campaign's last weeks—$72 million on TV ads and about $16 million on Internet ads—with Donald Trump's “nearly $39 million on last-minute TV ads and another $29 million on digital”); see also Bill Allison et al., Tracking the 2016 Presidential Money Race, Bloomberg Politics (Dec. 9, 2016),​politics/​graphics/​2016-presidential-campaign-fundraising/​ (noting that Trump's spending to “target[ ] specific groups of Clinton backers with negative ads on social media to lower Democratic turnout . . . may have been a factor in Trump's performance in battleground states”).

In sum, the Commission concludes that the petition does not present credible evidence that a 15% threshold is so unobtainable by independent or third-party candidates that it is per se subjective or intended to exclude them.

2. Submissions Regarding Whether Polls are Unreliable and Systematically Disfavor Independent and Third-Party Candidate

The Young Report's examination of polling error in three-way races with independents seeks to determine, essentially, if the threshold is drawn in the right place to identify candidates that actually have a 15% vote share. Young Report at ¶ 60. The Young Report concludes that polls in three-way races have greater errors than polls in two-way races. Specifically, the Young Report extrapolates from gubernatorial election polls taken two months before the general election (the point at which CPD uses polls as a debate inclusion criterion) where there is an 8% error rate in three-way races compared to a 5.5% error rate in two-way races. Id. at ¶¶ 52-56. Adjusting for the fact that gubernatorial race polling is “more error prone” than presidential race polling, the Young Report concludes that the applicable error rate is 6.04%. Id. at ¶¶ 57-58. The Young Report continues to extrapolate the effect of this error on candidates, such as independent or third-party candidates, that poll close to the 15% threshold; for these candidates, the Young Report concludes that there is an approximately 40% chance that a third-party or independent candidate who holds the support of 15% of the population would be excluded. Id. at ¶¶ 59-66.

The Commission is unpersuaded by this analysis for two fundamental reasons. First, as the Commission noted in its original notice of disposition, the fact that polling data can be erroneous does not mean that a debate staging organization acts subjectively in using it. 80 FR at 72618 n.6. By way of analogy, consider a school district with a policy of canceling school if a majority of local television news stations predict at least six inches of snow for the next day. That policy would be facially objective, even though such weather forecasts are known to be significantly Start Printed Page 15473inaccurate. The policy would be subjective only if the inaccuracy in the forecast were systematically biased for or against the condition being triggered (e.g., if the local weather forecasters regularly used high-end estimates of snow to drive viewer interest). But this demonstrates the second reason the Commission is unpersuaded by the petition's submissions regarding polling unreliability: The petition provides no evidence that the polling error is biased in a manner specific to party affiliation, that is, that polling is biased against third-party or independent candidates. Indeed, the petition explicitly acknowledges that “it [is] wholly unclear whether the polling over- or underestimate[s] the potential of the third party candidate.” Petition at 19 (quoting Schoen Report at 28). Thus, the Commission concludes that the petition does not demonstrate that statistical errors in polling data render the use of such data subjective or show that it is intended to exclude third-party candidates.[14]

The petition does imply that third-party and independent candidates are at a disadvantage because “there is no requirement that pollsters test third-party and independent candidates,” and therefore the CPD might “cherry pick from among the myriad polls that exist in order to engineer a specific outcome.” Petition at 17-18. But the petition presents no evidence that such manipulation has ever occurred, and the Commission is unwilling to predicate a rule change on unsupported speculation of wrongdoing. A debate sponsor who took actions to manipulate the “pre-established” and “objective” selection criteria so as to “select[ ] certain pre-chosen participants” by cherry-picking polls that excluded other candidates would violate the existing rule. Corporate and Labor Organization Activity; Express Advocacy and Coordination with Candidates, 60 FR at 64262.

The petition further argues that lowering the polling threshold is insufficient to solve polling error problems. As an initial matter, the Commission notes that the Young Report does not conclude that any and all polling thresholds are unreliable. On this point, in addition to the Young and Schoen Reports discussed above, Petitioner cites an article from Nate Silver on Republican primaries for the conclusion that “a simple poll does not capture a candidate's potential.” Petition at 17 (citing Nate Silver, A Polling Based Forecast of the Republican Primary Field, FiveThirtyEight Politics (May 11, 2011) (attached to Petition as Exhibit 20)). The cited article, though, concludes what appears to be the opposite of the point for which it is cited; it starts by explaining that it will prove the author's contention that “polls have enough predictive power to be a worthwhile starting point.” Petition, Ex. 20. In fact, that article was part four of a four part series. The second sentence of part one of that series explained that the series was intended to show that “national polls of primary voters—even [nine months] out from the Iowa caucuses and New Hampshire primary—do have a reasonable amount of predictive power in informing us as to the identity of the eventual nominee.” Nate Silver, A Brief History of Primary Polling, Part I, FiveThirtyEight (Mar. 31, 2011),​features/​a-brief-history-of-primary-polling-part-i/​. Moreover, polls like those used in September by CPD are not “inaccurate” or “unreliable” simply because their assessments of vote share do not match the final vote share on Election Day; such polls are “designed to measure the true level of public support at the time the poll is administered,” not “to measure the true level of public support on Election Day.” Commission on Presidential Debates, Comment at Ex. 2 ¶ 20 (Declaration of Frank M. Newport, Editor-in-Chief, Gallup Organization) (Dec. 15, 2014),​fosers/​showpdf.htm?​docid=​310982. As the Newport Declaration notes, “there is no doubt that properly conducted polls remain the best measure of public support for a candidate . . . at the time the polls are conducted.” Id. at Ex. 2 ¶ 21.

3. Submissions Regarding the Desirability of Expanding Debate Participation

The petition and most of the commenters who support it rely primarily on policy arguments that polling thresholds are inconsistent with the purposes of the existing regulations and that those purposes would be better served by, in essence, including more voices on the debate stage.[15] The Commission explained in its original Notice of Disposition why it was not persuaded by the petition's “arguments in favor of debate selection criteria that Start Printed Page 15474would include more candidates in general election presidential and vice presidential debates.” 80 FR at 72617. As the Commission explained, “The rule at section 110.13(c) . . . is not intended to maximize the number of debate participants; it is intended to ensure that staging organizations do not select participants in such a way that the costs of a debate constitute corporate contributions to the candidates taking part.” Id. That is the only basis on which the Commission is authorized to regulate in this area. The Commission has no independent statutory basis for regulating the number of candidates who participate in debates, and the merits or drawbacks of increasing such participation—except to the limited extent that they implicate federal campaign finance law — are policy questions outside the Commission's jurisdiction.


The evidence presented to the Commission in the petition and comments on the impracticability of independent candidates reaching the 15% threshold and on the unreliability of polling do not lead the Commission to conclude that the CPD's use of such a threshold for selecting debate participants is per se subjective, so as to require initiating a rulemaking to amend 11 CFR 110.13(c). While the reports by Dr. Young and Mr. Schoen, in addition to the historical polling and campaign finance data presented with the petition, demonstrate certain challenges that independent candidates may face when seeking the presidency, these submissions do not demonstrate either that the threshold is so high that only Democratic and Republican nominees could reasonably achieve it, or that the threshold is intended to result in the selection of those nominees to participate in the debates.

For all of the above reasons, in addition to the reasons discussed in the Notice of Disposition published in 2015, see Candidate Debates, 80 FR 72616, and because the Commission has determined that further pursuit of a rulemaking would not be a prudent use of available Commission resources, see 11 CFR 200.5(e), the Commission declines to commence a rulemaking that would amend the criteria for staging candidate debates in 11 CFR 110.13(c) to prohibit the use of a polling threshold to determine participation in presidential general election debates.

Start Signature

On behalf of the Commission,

Dated: March 23, 2017.

Steven T. Walther,

Chairman, Federal Election Commission.

End Signature End Supplemental Information


1.  See also Funding and Sponsorship of Federal Candidate Debates, 44 FR 76734 (Dec. 27, 1979) (explaining that, through candidate debate rule, costs of staging multi-candidate nonpartisan debates are not contributions or expenditures); 11 CFR 100.92 (excluding funds provided for costs of candidate debates staged under 11 CFR 110.13 from definition of “contribution”); 11 CFR 100.154 (excluding funds used for costs of candidate debates staged under 11 CFR 110.13 from definition of “expenditure”).

Back to Citation

2.  See Candidate Debates and News Stories, 61 FR 18049 (Apr. 24, 1996) (quoting H.R. Rep. No. 93-1239 at 4 (1974)).

Back to Citation

3.  Petition Ex. 3 (“Young Report”).

Back to Citation

4.  Petition Ex. 11 (“Schoen Report”).

Back to Citation

5.  See Young Report at ¶¶ 21-22.

Back to Citation

6.  Searches of the Thompson Reuters Westlaw “Newspaper” database for mentions in 2016 of independent and third-party 2016 presidential candidate names (“Gary Johnson,” “Jill Stein,” and “Evan McMullin”) show thousands of results. Moreover, the number of results for references to these independent candidates was comparable to the number of results for references to several major party candidates during comparable time periods. Using as a baseline the 277 days from the lead up to the first Republican party primary debate until Donald Trump was determined to be the presumptive nominee (August 1, 2015, to May 4, 2016), and the similar 277-day period of September 4, 2015 (before the first Democratic primary debate) to June 7, 2016 (when Hillary Clinton became the presumptive Democratic nominee), the Commission looked at mentions for independent candidates during the 277 days before the general election (February 5-November, 7, 2016). Those results show that Gary Johnson (with 3,001 results) was comparable to Bobby Jindal and Mike Huckabee (with 2,894 and 3,274 results, respectively); Jill Stein (with 1,744 results) was comparable to Rick Perry and Martin O'Malley (with 2,278 and 2,566 results, respectively); and Evan McMullin (with 353 results) was comparable to Lincoln Chafee, Jim Webb, and George Pataki (with 424, 521, and 937 results, respectively). And, while searches for Donald Trump's and Hillary Clinton's names returned significantly more results (7,451 and 7,404, respectively), those results were in line with other candidates who did not achieve high vote share in the party primaries, such as Jeb Bush with 7,102 results.

Back to Citation

7.  Schoen Report at 3; see also id. at 10 (asserting, without supporting data or sources, that costs will likely be “significantly” higher “in an election year featuring three viable candidates” and, therefore, adding 5% premium to report's earlier cost estimates).

Back to Citation

8.  Schoen Report at 4 (citing Paul Herrnson & Rob Faucheux, Outside Looking In: Views of Third Party and Independent Candidates, Campaigns & Elections (Aug. 1999)). The assertion also appears to be in tension with the statutory exclusion of the news media coverage from legal treatment as campaign spending. See 52 U.S.C. 30101(9)(B)(i) (excluding “any news story . . . distributed through the facilities of any broadcasting station, newspaper, magazine, or other periodical” from definition of “expenditure”).

Back to Citation

9.  See Open Secrets, Independent Expenditures, Gary Johnson, 2016 cycle,​pres16/​outside-spending?​id=​N00033226 (listing six “Super PACs” or IEOPCs supporting Johnson, two of which spent over $1 million in support) (last visited Feb. 24, 2017).

Back to Citation

10.  See Ariel Edwards-Levy, Third-Party Candidates are Getting a Boost in Name Recognition, Huffington Post (Aug. 31, 2016) (noting Johnson's name recognition); Poll Results: Third Party Candidates, YouGov (Aug. 25-26, 2016), available at​cumulus_​uploads/​document/​wc35k48hrs/​tabs_​HP_​Third_​Party_​Candidates_​20160831.pdf (showing Gary Johnson and Jill Stein having 63% and 59% name recognition among registered voters, respectively).

Back to Citation

11.  See Gary Johnson 2016, FEC Form 3P at 3-4 (Sept. 20, 2016),​pdf/​391/​201609209032026391/​201609209032026391.pdf (showing receipts of $7,937,608 and disbursements of $5,444,704).

Back to Citation

12.  The Young and Schoen Reports do not address a circumstance in which a candidate, like Gary Johnson, reaches at least 60% name recognition but does not reach a 15% threshold. The Commission notes, though, that this circumstance (in which name recognition does not translate to high vote share) might be explained by the other factors beyond name recognition that affect vote share, including “fundraising, candidate positioning, election results, and idiosyncratic events,” mentioned in the Young Report. See Young Report at ¶¶ 10, 20(d). Moreover, the circumstance in which name recognition does not translate to high vote share is not unique to third party candidates. See note 6, above (discussing Jeb Bush).

Back to Citation

13.  The petition also asks the Commission to disregard the strong polling results of third-party or independent candidates, like George Wallace and John Anderson, who have a prior affiliation with a major political party. Petition at 15. The Commission is not persuaded that disregarding those polling results would be reasonable in the context of assessing, as required by the court, whether the CPD's 15% threshold under the current candidate debate regulation acts “subjectively to exclude independent and third-party candidates,” since the threshold would apply to all third-party and independent candidates, regardless of prior affiliation. Level the Playing Field, 2017 WL 437400 at *12.

Back to Citation

14.  Because this data, even as cited by the petition, does not show that the regulation should be amended, the Commission need not further assess the data's validity. Nonetheless, the Commission notes that there are significant structural differences between the state polls cited by Dr. Young and national presidential polls. See, e.g., Young Report at ¶¶ 41 (explaining differences between reputable national and state or local polls, with respect to both number of interviews and margins of error), 57 (showing significant differences between state and federal polling at different points in time). Although Dr. Young adjusts the state-poll results before applying them to his national analysis, (see id. ¶ 58), the manner in which the adjustment is described leaves unexplained whether the adjustment accounts for all of the relevant differences between state and national polls.

The Petitioner also submitted in response to the Notice of Availability a comment with additional data concerning “grossly inaccurate” polling in 2014 midterm Senate and gubernatorial elections. Level the Playing Field, Comment at 1 (Nov. 26, 2014),​fosers/​showpdf.htm?​docid=​310980. However, attachments to the comment note that “midterm polling biases in Senate elections are far worse than in presidential elections.” Id. at Exhibit A. And a chart created by the Petitioner for the comment shows that, of ten races with purportedly high polling errors in races without a “viable third-party or independent candidate,” the two races included in the chart with the lowest polling error are, in fact, the only two races that include a third-party or independent candidate. Compare Level the Playing Field, Comment at 3 (showing Georgia and North Carolina Senate races with the lowest final polling errors of those entries in chart) to Level the Playing Field, Comment at Exhibit C (showing Georgia and North Carolina Senate as only races included in chart that involved three-way race polling). For all of these reasons, the Commission is not persuaded that the Petitioner's submissions regarding state and Senate polls indicate any systematic, anti-third-party flaw in the polls at issue here, which are presidential general election polls.

Back to Citation

15.  A substantial majority of the comments that the Commission received on the petition were cursory and consisted of a single sentence expressing support for the petition. See, e.g., Comment by Amanda Powell, REG 2014-06 Amendment of 11 CFR 110.13(c) (Dec. 15, 2014) (“I support the petition.”),​fosers/​showpdf.htm?​docid=​310989. Additionally, the League of Women Voters “does not support amending the FEC regulation to preclude sponsors of general election presidential and vice presidential debates from requiring that a candidate meet a polling threshold in order to be included in the debate,” but did generally support opening a rulemaking, though without supporting or proposing any specific proposal. Comment by League of Women Voters, REG 2014-06 Amendment of 11 CFR 110.13(c) (Dec. 15, 2014),​fosers/​showpdf.htm?​docid=​310985. The comment did not, however, present any substantial justification for doing so. Moreover, such an open-ended inquiry was not the focus of the petition for rulemaking.

Another commenter, FairVote, indicated that it “do[es] not oppose the use of polling as a debate selection criterion so long as candidates have an alternative means of qualifying for inclusion.” See Comment by FairVote, REG 2014-06 Amendment of 11 CFR 110.13(c) (Dec. 15, 2014),​fosers/​showpdf.htm?​docid=​310974. That commenter emphasized the Commission's recognition of the educational purpose of candidate debates and advocated that including additional candidates in debates would “broaden the substantive discussion within the debates.” Id. As explained supra, however, the main purpose of the regulation at issue is to clarify when money spent on debate sponsorship is exempt from the FECA's definition of “contribution.” The Commission's recognition of the educational value of debates does not alter its view that the determination of which candidates participate in a given debate should generally be left to the organizations sponsoring such events. See supra. In addition, while the Commenter supported Petitioner's proposed alternative to select a third debate participant based upon the number of signatures gathered to obtain ballot access, the existing rule already permits this alternative and thus amending the rule is not required to allow for that approach. See id.

Back to Citation

[FR Doc. 2017-06150 Filed 3-28-17; 8:45 am]