No one questions that campaign finance law has struggled through multiple, agonized revisions in distinguishing issues from campaign speech and the discussion of campaign issues from advocacy for candidates or parties. The statute is little help; it speaks of the “purpose of influencing” an election,” 2 U.S.C. §431(8)(A)(i), and broader Commission glosses on the phrase, such as a test for whether a message was “electioneering” in content, eventually came to grief. The Supreme Court held the express advocacy line briefly, then gave in to a conception of the “functional equivalent” of express advocacy, and has since cast much of discussion into obsolescence by extending to corporations the right to make independent expenditures. Now tax policy-makers and tax law face pressure to work through the same issue, in limiting political intervention by 501(c)(4)s, and the results might be expected to be the same.

This disappointment will not come about for want of trying or application by the best minds in the business. The authors of the Bright Line Project have devoted considerable, expert attention to proposing improved tests for political intervention. The Bright Lines Project: Clarifying IRS Rules on Political Intervention (Interim Draft, May 23, 2013). But as noted here, the outcome of this work so far is a complex contraption, and clarity ends up giving ground to a freely roaming “facts and circumstances” analysis. Their work is also helpful in bringing out yet one more problem: the conflict between enforcement priorities and the protections we would expect under First Amendment doctrine for “controversial speech.” Controversial speech grabs attention and is laden with the potential to “influence” voters: this moves it up the list of regulatory priorities, but the very same category of speech is especially dear to the hearts of reform critics who wish to protect it from regulation.

The Bright Line Project’s “general speech rule,” designed to capture political intervention, consists of two prongs: a communication’s reference to a candidate and its expression of a “clear view” of that candidate. “Bias or favoritism of any kind” results in a “clear view,” including content that is neither wholly positive nor negative, but simply “nuanced.” Id. at 12. So the Project’s authors argue for allowing tax-exempts to engage in voter education—speech about campaign issues, not advocacy for particular candidates—on the condition that the tax-exempt speak “neutrally.” Id. at 14.

A good example of the trap that invariably opens up under this test is the question of how to treat organizations that ask candidates to endorse a policy or take a pledge, such as a no-tax pledge, and then proceed to publish the results. Tax law now classifies all such pledge requests as political intervention, and to their credit, the Bright Line authors would loosen things up. But while allowing the question, they allow only the publication of some answers. This is what they say:

Whether [the publication] results in the organization “reflecting a view” depends on the responses. If all the candidates say “yes,” there’s no bias in presenting that result to the public. If no candidate says “yes,” that could be presented neutrally, i.e. the organization is still pursuing a policy despite the fact that no candidate yet favors it. If some say yes and others no, the organization’s presentation of that result would reflect a view, and so it must refrain from publicizing the candidate’s responses…

Id.

In short, the controversial nature of the issue is the dispositive factor. For it is not controversial speech, and of the most interest to voters, if none of the candidates buy it, or if all agree with it. It is controversial and prime subject matter for voter education if there will be a division of opinion among the candidates, presumably reflecting a division within the electorate. The organization’s view is the same in all three cases—but under the test presented by the Project, the organization could not present its view in only one case—where the issue stirs up differences of opinion and the interest in voter education might be the highest. The ban on educating voters on these differences does not depend on further editorializing by the tax-exempt; it is just enough for the organization to reveal its position and the agreement or disagreement with it among the candidates whose views are presented.

This has always been the bane of the political law-making enterprise: that the speech that may count the most for First Amendment purposes is the speech that can make a difference to voter choice and stimulates the regulatory appetite to control it. The Federal Election Commission has struggled with this problem. See e.g. Faucher v. Federal Election Commission, 928 F.2d 468, 472 (1991) ( “In our view, trying to discern when issue advocacy in a voter guide crosses the threshold and becomes express advocacy invites just the sort of constitutional question the Court sought to avoid in adopting the bright-line express advocacy test in Buckley.”) The task will be no easier for the IRS, and it bears the additional burden of coming to this work as an agency that many doubt should be concerned with defining the boundaries between campaign and issue advocacy.