Bright Lines Project Seeks New Approach to Political Activity Regs

The Bright Lines Project applauds the Department of the Treasury and the Internal Revenue Service (hereinafter IRS and collectively, "Treasury") for undertaking this effort to clarify the definition of political activity for social welfare organizations. We strongly support the need for establishing bright lines on what constitutes political campaign intervention for tax-exempt organizations, and we welcome the IRS's much-needed attention to these important issues. However, for the many reasons identified in this letter, we believe the initial proposal released on November 29, 2013, requires very significant improvements in order to achieve its objectives, provide fair treatment of all 501(c) organizations, and avoid interference with legitimate, nonpartisan activities of nonprofit organizations. We urge Treasury to substantially rework the proposal and provide the public another opportunity to comment on a revised Notice of Proposed Rulemaking (NPRM) before moving to a final rule.

We believe a better starting point would be the rules proposed by the Bright Lines Project, which would be universally applied across all nonprofit tax categories, and offer predictability, simplicity, and ease of understanding in what constitutes political activity. Our proposal protects free speech and encourages civic engagement while still preventing abuses of the system so that nonprofits can focus on their core missions and work without fear. It also gives the Treasury the tools it needs to enforce tax exemption rules without tortuous delay or bias.

We request that Treasury sponsor a series of public hearings in different parts of the country to get more input from nonprofits before it publishes a revised NPRM, and that we have the opportunity to testify.

Finally, in addition to hearings, we recommend that Treasury find other means of seeking input from as broad an array of nonprofit leaders and advisors as possible as it moves forward to develop an alternate proposal. A similar approach was used successfully in the 1986-1990 period when Treasury developed rules to implement lobbying provisions under sections 501(h), 4911, and 4945 of the tax code.

The ambiguity about what constitutes political activity has long been a problem for 501(c)(4)s and for tax-exempt organizations more generally.² We therefore strongly support the effort "to provide greater certainty to 501(c)(4) organizations regarding their activities and reduce the need for fact-intensive determinations." The current system of determining whether a particular activity is deemed "political campaign intervention" based on Internal Revenue Service review of all of the "facts and circumstances" prevents cautious organizations from engaging in permissible activities and leads other organizations to abuse their tax-exempt status (either through error or deliberately).

The current vague and unpredictable system also makes it difficult for the IRS to effectively enforce the laws restricting electioneering by tax-exempt organizations. Readers of these comments are certainly aware of scandal that erupted in May, 2013, when IRS leadership acknowledged that organizations applying for recognition of their exemption from federal income tax under sections 501(c)(4) and 501(c)(3) were targeted for scrutiny based on inappropriate criteria and were subjected to overly burdensome and protracted review of their applications. In his report on the matter, the Treasury Inspector General for Tax Administration cited the problems IRS staff had in applying the current standard for what constitutes political activity as a significant contributor to the problems and recommended guidance on "what constitutes political campaign intervention."³

Even if Treasury were to discontinue the rulemaking process that it began last November, the IRS would still need to provide some kind of better guidelines for its agents who review exemption applications and for its examiners who conduct audits of tax-exempt organizations. If such guidance were not prepared in the form of regulations, the highest federal tax authority below the Internal Revenue Code itself, we would lose the great benefit of exactly what is occurring now -- robust input from many corners of the nonprofit sector and the political world -- mandated by the legal requirement of public commentary and public hearings.

The increased involvement of 501(c)s in political activity in the wake of the Supreme Court's decision in Citizens United v. Federal Election Commission⁴ and court rulings in other recent campaign finance cases will only intensify the need for a clear and fair definition of political activity for exempt organizations. As explained later in these comments, we are witnessing an unprecedented expansion in both the number of 501(c)(4) and other 501(c) organizations and in the use of such organizations to intervene in political campaigns. Whether one welcomes these developments or seeks to restrain this growth in 501(c) electioneering, it is clear that a workable standard of what constitutes exempt organization political activity is vital to ensuring that organizations operating in good faith can comply with the law and that the IRS can enforce the law against those operating in bad faith.

In short, recent developments require the Department of the Treasury to exercise its authority to "prescribe all needful rules and regulations for the enforcement of [the nation's tax laws]"⁵ and craft regulations to clarify Internal Revenue Code provisions related to political activity.⁶ Although we criticize the NPRM's first attempt, the significant gap between our suggestions and the currently proposed regulations should not be seen as an insurmountable barrier to a good final rule. It is worth remembering a similarly controversial past rulemaking: In the 1980s, comments on the initial draft of regulations defining 501(c)(3) lobbying were nearly all negative. Yet after reviewing those comments and working closely with the regulated community, Treasury and the IRS produced the current regulations, widely and justifiably praised for their clarity and workability.⁷ We believe that a similar effort in this case can likewise succeed, and we urge you to make the attempt.

This submission is divided into three parts: (I) the scope of proposed regulations, (II) content issues in drafting regulations to define political intervention, in which we compare the NPRM proposal to our alternative, and (III) specific definitions of candidate-related political activity proposed in the NPRM.

In composing this submission, we realize that there is a vast range of issues to discuss -- many of them technical, practical, historical, or political. We deal with some in depth in this paper and only hint at others, but believe that as the interaction between the government and the public moves forward, there will be opportunities to turn over every stone in the deliberation that will be required. Perhaps no other problem of line-drawing in the tax-exempt arena has been and will be so difficult, and so important.

The threshold question for this rulemaking is what types of tax-exempt organizations should a new definition of political activities cover. As drafted, the Treasury NPRM applies only to 501(c)(4) social welfare groups, but it seeks comments on whether the proposed definition of candidate-related political activity should apply to other 501(c)s as well.

There are many reasons why the Treasury NPRM has opened this conversation by addressing only 501(c)(4) social welfare groups. However, it would be a mistake to issue final rules defining "candid ate related political activity" limited to social welfare groups only. Such a limitation would create confusion among nonprofits and IRS regulators, hinder public involvement in the democratic process, and also push the constricted political activity to other tax-exempt or taxable organizations where the rules may be more flexible or less well-defined and donors interested in anonymity can remain so. A top priority for Treasury should be to curtail opportunities for bad actors to game the system: a first step is to develop definitions of political activity that apply universally across all tax-exempt categories in the Internal Revenue Code and, where the IRC addresses political activity by businesses and individuals, taxpayers of every kind.⁸

When the Supreme Court ruled in Citizens United v. Federal Election Commission⁹ that corporations (both for-profit and nonprofit) could spend unlimited money on electioneering through independent expenditures, it was widely expected there would be a growth in the number of, and in political spending by, social welfare organizations (like Citizens United itself) because, according to current IRS interpretations, these groups can engage in political activity without disclosure of donors as long as it is not the primary activity of the organization.¹⁰ The expectations were accurate: there was a 92 percent surge in exemption applications between FY 2009 and FY 2012 (pre- and post-Citizens United) (see Figure 1), putting significant pressure on the IRS to decide whether an organization was primarily engaged in social welfare activities. IRS agents were forced to undertake this dramatically increased workload without clear definitions of political activity.

The multi-factor, open-ended "facts and circumstances" approach used by the IRS leaves enormous room for social welfare organizations to engage in political activity without calling it as such, so long as they are willing to take a legally aggressive position.¹¹ This has led to thinly-disguised "issue ads" that most of the public recognizes as designed to engender support or opposition to candidates for public office rather than to argue for or against an issue of public concern.¹² At the same time, the IRS has not defined how much political activity these nonprofits can undertake and still qualify as tax-exempt, other than to note that it cannot be the primary activity. The ambiguity has led some attorneys to advise their clients (and some internal IRS training materials to imply)¹³ that "less-than-primary" means they can spend up to 49 percent of their annual expenses on political activity.

 

FIGURE 1.

 

 

501(c)(4) Applications:

 

92% Increase Since Citizens United

 

 

 

 

Figure 2. "Dark Money" Spent by Tax Exempt

 

Organization^s

 

 

 

 

FIGURE 3. "Dark Money" Spent

 

by Type of Tax Exempt Group¹

 

 

 

Source: Center for Responsive Politics accessed 1/12/14 at http://www.opensecrets.org/outsidespending/nonprof_summ.php?cycle=All&type=type.

Regardless of one's beliefs about the appropriate limits of political speech favored or disfavored by federal tax law, the ambiguity created by the "primary activity" standard has meant that many nonprofits, including 501(c)(4)s, 501(c)(6)s and more, have become the chosen conduits for undisclosed election-related funding and activity. As Figures 2 and 3 show, in the post-Citizens United world, the amount of political spending flowing from undisclosed donors has jumped dramatically and the major source of this hidden money comes from 501(c)(4) groups.

While the data support the notion that Treasury regulations should cover social welfare groups, common sense calls for applying the definition of political activity beyond 501(c)(4)s. For example, the Wall Street Journal pointed out that in response to the Treasury NPRM "lawyers are scouring the tax code for other financial vehicles that would allow political donors to continue to spend money on elections while remaining anonymous."¹⁴ The newspaper noted that 501(c)(6) groups and limited liability corporations could easily be used.

Case in point, on the same day as the Wall Street Journal story, the Washington Post published an analysis describing a complex network spearheaded by certain billionaire individuals, "built around a maze of groups that cloaks its donors" in secrecy.¹⁵ The analysis shows that 17 different groups made up the network that used a host of limited liability corporations and 501(c)(4) and 501(c)(6) groups to raise and distribute at least $407 million during the 2012 campaign season.

Given the substantial evidence that money intended to influence political campaigns flows through different tax-exempt and for-profit tax structures, Treasury should develop rules that define political activity applicable universally across all such entities, so that no differentials in tax treatment or public disclosure would serve to incentivize shifts in political funding to the least regulated vehicles.

We do recognize that public disclosure of campaign financing sources is not the central mission of the IRS, as it is for agencies that directly regulate campaigns and elections. However, disclosure is a mandate and a function imposed on the IRS by the amendments made in 2000 and 2002 to section 527, and inasmuch as 501(c) organizations provide alternative vehicles for political spending without such disclosure, Treasury and the IRS need to be concerned. Other differentials have also affected choices made to locate political intervention among various tax-exempt categories, including the risk of federal gift tax on donors, state charitable trust jurisdiction, access to tax-deductible funds, and ease of creation and administration.

Application to Section 501(c)(3) Organizations

Treasury's political definition, if well-designed, should also apply to 501(c)(3) organizations, the largest tax-exempt category, because charities are not permitted to engage in political activities and have no margin for error.

Under section 501(c)(3) of the tax code, charities are appropriately not permitted to support or oppose candidates running for elected public office. Although this prohibition has been on the books for sixty years, IRS regulations¹⁶ hardly begin to define political intervention except to say it can be direct or indirect: instead, the IRS evaluates "all the facts and circumstances" to determine whether a charity has intervened in a campaign. Other precedential guidance (such as Revenue Rulings) addresses a few specific types of circumstances, but leaves most questions unanswered. Case law is sparse.

The widespread uncertainty and misunderstanding among nonprofit leaders means that many charities refrain from nonpartisan activity, such as forms of genuine issue advocacy and democratic civic engagement that are not designed to influence the election or defeat of candidates for public office. The result is that their voices are silenced, or muted, in public policy debate. On the opposite side, others, often larger organizations, choose to venture boldly into the gray area beyond what has been understood as permissible but short of express speech for or against the election of identified candidates. Advocates on various sides of policy debates find themselves on an uneven playing field where more aggressive groups take advantage of uncertainty to promote their electoral agenda and more risk-averse (or carefully advised) ideological opponents steer very clear of anything related to an upcoming election. This is deeply troubling at a time when voters' rights have been challenged, nonpartisan discussion of critical issues has been stifled, polarization has worsened, and campaigns to reinvigorate American democracy -- from all participants -- are needed.

The disparate administration and enforcement that results from unclear standards further weakens the voices of law-abiding nonprofits. In the rare cases in which the details of an IRS enforcement action become known, it is often possible to find similar activities conducted by other nonprofits that apparently did not result in IRS enforcement efforts or sanctions. There are groups that can afford to incur the legal fees involved to vigorously defend questionable conduct, while others similarly situated are forced to agree to a settlement, unable to invest the resources needed to defend their choices or to understand the infraction of which they have been accused.

Even when charities are not subject to the expense and inconvenience of an IRS audit, the vague and complex legal standard they must follow means that they expend more resources on legal advice and compliance and less on actually pursuing their mission. A prudent charity is forced by the ambiguity of the rules and the uncertainty of the enforcement process to have experts vet every move that is arguably related to any candidate or election for public office. Not only does this consume significant financial resources for in-house or outside legal services, but it slows action by the charity, often making it less effective.

The Treasury proposal does nothing to help provide clarity for charities, the largest category of tax-exempt groups. In fact, the NPRM will add further confusion for charities and the result will likely be even less engagement in nonpartisan voter activities. For example, the NPRM proposes that all voter registration and get out the vote (GOTV) activities would be classified as "candidate related political activity," regardless of whether they are nonpartis an or partisan. Sophisticated charities may know about Revenue Ruling 2007-41 that makes clear that 501(c)(3)s may conduct such activities if they do so in a nonpartisan fashion. But most will not. And some may assume, rightly or wrongly, that the rules on candidate-related political activity implicitly modify or otherwise affect the IRS's prior interpretations of the campaign intervention prohibition in section 501(c)(3).

Moreover, even if charity leaders fully understand the rules, donors to charities may not. Indeed, they may react to the general rule that nonpartisan voter registration, GOTV, and other voter education activities are deemed "political" by the latest government edict and choose not to support such activities. Thus, though the NPRM does not technically cover charities, these rules are likely to have an indirect, yet consequential and adverse impact on charities.

Creating a different set of definitions for 501(c)(4) groups from 501(c)(3) standards and interpretations will wreak havoc on many tandem relationships between charities and their affiliated advocacy organization. It will also confuse grant-making from 501(c)(3) organizations to 501(c)(4) groups that conduct charitable programs.

For those charities operating with affiliated social welfare groups, the NPRM will not only add confusion but will also make implementation of such affiliated operations challenging if not burdensome. The NPRM would classify almost any mention by a 501(c)(4) group of an incumbent federal, state or local politician, who is up for re-election, within 30 days of a primary or 60 days of a general election as "candidate related political activity." It would also consider any nonpartisan candidate forums conducted by a 501(c)(4) within the blackout period as "political." These rules could effectively require all 501(c)(4) groups to remove any reference to the politicians from their website during the pre-election period or be found to have engaged in political activity.¹⁷ However, the same activities -- both mentions of incumbents and candidate forums -- when conducted by 501(c)(3)s, would likely be permissible and not considered political activity if done in a nonpartisan manner. This would create two separate standards -- a difference that, for affiliated organizations, is likely to cause serious complications.

Moreover, 501(c)(3) charities often make grants to related or unrelated 501(c)(4) social welfare groups that are well-situated to provide local community programs, pursuant to terms of grant agreements that prohibit use of grant funds for intervention in political candidate campaigns. How is that restriction to be interpreted consistently if a voter engagement program is deemed political under section 501(c)(4) while the (c)(3) funding source is fully entitled to conduct or subsidize the same activity -- so long as it is nonpartisan? Private foundations, the most highly-regulated of all charities, are specifically authorized by statute¹⁸ to make grants to public charities for nonpartisan voter registration conduct in five or more states, over more than one election period -- while the same program if conducted by a 501(c)(4) organization would be deemed political.

Even if nonprofit leaders have the best playbook clarifying all of the inconsistencies, these rules would make it very difficult for charities to engage in advocacy and other forms of civic engagement. Our laws, such as the National Voter Registration Act, have long permitted charities to urge all eligible people to vote, and provide nonpartisan resources to help people make informed judgments about policies and candidates. The Treasury proposal undermines this fundamental approach to democracy. Instead of discouraging nonpartisan civic engagement, as the Treasury NPRM does, Treasury should encourage it in order to strengthen our democracy.

Accordingly, we strongly believe that any rules published by Treasury should be clear about what nonpartisan activities charities can do that would not be classified as political intervention by any tax-exempt entity. As described in this letter, we believe Treasury should affirm in its definition of political campaign activity that nonpartisan voter registration, get out the vote, and voter education activities are not considered "political" in the absence of evidence that clearly indicates that the content of the program or the way in which it is targeted is partisan in character.

Application to Other Sections of the Internal Revenue Code

We believe that Treasury should extend this rulemaking to other tax-exempt categories including:

The NPRM also requests comments on whether to take a consistent approach regarding "activities that do not further exempt purposes under sections 501(c)(5) and 501(c)(6)." The answer is an unreserved yes. Indeed, Treasury final rules should also include, and make consistent, the standards by which political intervention are judged by all section 501(c) organizations. And for the sake of consistency, it is vital that final rules also standardize this definition for the other sections of the Internal Revenue Code, including IRC sections 162(e) (barring deduction of political expenses as business expenses), 170(c) (requirements for charitable tax deduction), 4955 (excise tax on political expenditures by public charities), 4945 (excise tax on political expenditures by private foundations), and 6033(e) (notice requirements and proxy tax for political expenditures).

Section 527(e) of the IRC presents an anomaly and a problem; for unique historical reasons it defines political candidates in a way that is markedly different from the rest of the IRC. Congressional action would be needed to align the terms that are at odds with the other IRC sections, mainly to correct the description of additional offices (appointive and political party) included under section 527. Even without a statutory fix, Treasury could certainly revisit the existing 527 regulations and at least conform the standards used to identify 527 political activities so they mesh with a new, universal set of standards designed for all tax-exempt and taxable entities. We will go into more depth on section 527 later in this submission.

The treatment of political campaign activity by section 501(c)(4) organizations is not unique in the IRC; the IRS has long been clear that section 501(c)(5) labor unions and section 501(c)(6) trade associations are subject to the same limits on political campaign intervention as 501(c)(4) organizations.¹⁹ And this is not mere theory; labor unions and trade associations have both been extremely active in federal and state elections in recent years under the existing primary activity standard that the IRS has applied to them as well as 501(c)(4)s.²⁰

As the IRS has noted, the same considerations that apply to the political activity of section 501(c)(4), (c)(5) and (c)(6) organizations "would appear to apply to other types of exempt organizations" as well .²¹ There are twenty-nine subsections to section 501(c) of the IRC, each describing a subset of organizations; many other types of section 501(c) organizations could be used for political campaign activity. The fact that few subsections are currently used for serious amounts of political activity is not necessarily prologue; if regulations dramatically restrict the political speech of some section 501(c) subsections without regulating others, activity will surely shift to less-regulated subsections. Indeed, there is evidence that this trend is already beginning; one nonprofit organization, which granted $236 million to politically-active entities in the 2012 election cycle, has claimed section 501(c)(6) status instead of section 501(c)(4) status.²² This move towards section 501(c)(6) may well be a reaction to increased controversy about and regulation of section 501(c)(4) organizations at the state and federal levels.

For the same reason, final rules should apply identical definitions to section 162(e). For-profit businesses that wish to be active in elections -- more likely since Citizens United declared invalid all federal and state election law prohibitions on corporate independent expenditures -- face a choice as to whether to engage directly do so through a contribution to a section 501(c) organization. If section 501(c) organizations were to operate under rules more restrictive, or fundamentally different, than the businesses funding those tax-exempt entities, more political activity will be done either from for-profit businesses themselves, or from entities operating as non-profits but eschewing federal tax-exempt classification. To prevent arbitrage opportunities, final rules should apply equally to section 162(e). To do so would also provide much-needed clarity to this section, which uses almost identical language to the section 501(c)(3) prohibition on political campaign intervention.

Conforming section 162(e), section 527(e), and the section 6033(e) notice and proxy tax provisions would also prevent section 501(c) organizations from having to track at least two and potentially three different definitions of political activity. Were the final rules to apply to section 501(c) tax-exempt status only, a section 501(c) organization might have to track three different, separate definitions: one for its own tax-exempt status; one that would apply to section 162(e) and the related section 6033(e) proxy tax calculation; and one to the calculation of investment income tax under section 527(f). There is no reason that these disparate standards cannot be brought under one common umbrella.

Relationship to IRC 527 (Political Organizations) and Its Regulations

Using the IRC 527 definition of political activity, which covers advocacy related to appointive public offices and political party offices, takes the NPRM regulations down an unfortunate rabbit hole. The current wording of the statute and regulations contains many elements that are at odds with rest of the IRC. Section 527 plays an important role in providing a limited tax-exempt status, with donor disclosure, for organizations devoted to the political candidate selection process, but both the statute and the regulations are in serious need of repair. The treatment of appointed officials as political candidates is only one of the problems.

Existing rules establish rather different standards defining "political activity" for purposes of sections 527 and 501(c). This discrepancy generates significant technical problems. Some of these are rooted in the statute itself, but others stem from regulatory interpretations. The proposed regulations attempt to resolve one aspect of these issues by including additional offices in the coverage of "candidate-related political activity," but in so doing create yet more problems and vastly increase the recordkeeping and compliance burden for covered organizations.

Non-overlapping Covered Offices

The current regulations provide that "the promotion of social welfare does not include direct or indirect participation or intervention in political campaigns on behalf of or in opposition to any candidate for public office."²³ This echoes the language of the statutory prohibition applicable to section 501(c)(3) organizations, which may not "participate in, or intervene in (including the publishing or distributing of statements), any political campaign on behalf of (or in opposition to) any candidate for public office." IRC § 501(c)(3). Covered activity relates to candidates for election to public office, rather than those seeking appointive office or an office that is not considered "public."

The interpretation that disqualifying 501(c)(4) political activity is the same as the activity that is impermissible for a section 501(c)(3) organization is further supported by existing rulings applying the limitation on political activity by 501(c)(4) organizations. For instance, in asserting that 501(c)(4)s may engage in political activity so long as they primarily conduct social welfare activities, Rev. Rul. 81-95 cites a number of existing 501(c)(3) rulings for examples of what constitutes campaign intervention.²⁴ Thus, Rev. Rul. 81-95 demonstrates an implicit assumption that 501(c)(4) non-social welfare political activity should be defined by the same standard used to determine what constitutes prohibited 501(c)(3) campaign intervention.²⁵ Thus, under current rules, the political activity that is defined as not furthering social welfare²⁶ is limited to advocacy with respect to elected public office.

The "exempt function" of a section 527 organization, which is the activity those organizations must primarily engage in,²⁷ is defined more broadly:

There may be historical reasons for this discrepancy -- primarily, section 527 was drafted to create a home in the tax code for an array of activities undertaken by political organizations. Because section 527 organizations enjoy only a limited exemption from tax,²⁹ there is no serious policy reason to prohibit them from engaging in peripherally-related activities or treating as allowable activity a broader set of advocacy than is treated as "political" for 501(c) organizations.

The discrepancy between campaign intervention and section 527 exempt function creates administrative headaches, however, when organizations subject to the former in measuring their operation for tax-exempt purposes under section501(c) are required to apply the latter to determine their liability for tax under I.R.C. section 527(f).³⁰ This tension has been reduced in recent decades by the IRS's practical decision to suspend imposition of the section 527(f) tax on advocacy for or against appointive office.³¹ It would be unfortunate were the IRS to re-open the gap between 527(f) tax and social welfare qualification, thus forcing organizations to effectively track "political" activity using two overlapping yet often significantly different definitions for two different legal purposes.

The proposed regulations attempt to reconcile this tension by expanding the class of offices to which the definition of "candidate-related political activity" would apply to track that included in section 527 (e). While we commend the effort to align these definitions and reduce administrative complexity, this approach does not accomplish that goal. Rather, under the proposed regulations, a 501(c)(4) organization would be required to apply two different standards to categorize the same advocacy for these different legal purposes. Though the offices involved might be the same, if nothing is done to revise the language in the existing 527 regulations describing political activity (exempt function), then the "facts and circumstances" approach would be used to determine the application of the section 527(f) tax,³² and the new regulatory standard would be used to determine whether an activity is "candidate-related political activity" that does not promote social welfare.

In addition, an organization seeking to determine whether it qualifies as a section 527 organization will also have to use the existing "facts and circumstances" test to decide when activity furthers an exempt function, which has been broadly interpreted to reach anything with a "nexus" to the candidate selection process. These divergent standards could lead to the anomalous result that an organization would be simultaneously primarily engaged in activity that is not "candidate-related political activity" and primarily engaged in section 527 exempt function activity. The treatment of such an organization might be resolved by subsequent analysis to determine whether its non-"candidate-related political activity" also furthers social welfare. If it does not, the organization would still not qualify for exemption under section 501(c)(4). However, this outcome is not guaranteed. While this anomalous result is also possible under current law, given the different offices covered by campaign intervention and exempt function activity, the approach of the proposed rules vastly expands the circumstances in which an organization could potentially be both a section 501(c)(4) and a section 527 organization. For instance, a ballot measure committee supporting an initiative designed, in part, to promote the electoral prospects of certain candidates might be within the "nexus" of section 527 exempt function and yet be pursuing perfectly appropriate 501(c)(4) social welfare legislative goals, doing nothing within the proposed definition of "candidate-related political activity."³³

Because of the limited scope of application of the proposed regulations, they could also create significant questions about the operation of a separate segregated fund ("SSF") under IRC section 527(f). Such a fund is treated as a separate taxpayer from the organization that establishes it. Thus, the sponsoring organization is not taxed on expenditures by its SSF, nor does it currently need to worry that the political expenditures of the fund will count against it in measuring whether it qualifies as primarily operated for exempt social welfare purposes. The SSF is a 527 organization, subject to the applicable disclosure requirements and investment income tax. As a 527 organization, an SSF's activities are defined by that IRC section's "exempt function," an d would continued to be determined using the "facts and circumstances" approach. Thus, under the proposed regulations, a 501(c)(4) could be faced with a situation where it sought to move certain activities to an SSF in order to protect its exempt status, yet found that those activities were not considered political for 527 purposes and thus not appropriate to move to an SSF, while they were treated as "candidate-related political activity" under the broad and sweeping definition that would apply if conducted out of the organization's general funds. This would be not only an untenable situation for the organization, but would also undermines the acknowledged public policy of encouraging exempt organizations to conduct political activities through 527 entities, which fully disclose both receipts and expenditures.

As discussed further below, we support the development of a unified standard to apply in determining qualification for section 527 status, 501(c)(4) status (and 501(c)(5) and 501(c)(6) statuses), and application of the section 527(f) tax. However, both complete harmony and an easily administrable standard cannot be achieved without legislative changes to the scope of section 527 exempt function. We urge the IRS to adopt regulations that minimize the discrepancy between the standards applied for these different tax-exempt purposes and reduce the administrative burden of compliance with applicable rules. This is best achieved by applying the same standard to all exempt organizations to determine whether advocacy qualifies as "political," even if it applies to different offices in the different statutory context of section 527 entities.

Reserved Regulations That Exclude Activities from Tax under Section 527(f)

A further technical challenge is raised by the regulations defining expenditures by section 501(c) organizations that are and are not subject to the section 527(f) investment income tax. As currently drafted, these regulations can be read to exempt from taxation any political activity a 501(c) organization may legally undertake, thus vitiating the application of the tax and creating further discrepancies between the definitions that these entities must manage.

Section 527(f) subjects 501(c) organizations to tax on the lesser of their net investment income or their exempt function expenditures. The applicable regulations create exceptions for activities that will not trigger the 527(f) tax: appearances before a legislative body for the purpose of influencing the confirmation or appointment of an individual to a public office in response to a written request from the legislative body, Treas. Reg. § 1.527-6(b)(4), and nonpartisan voter registration and get-out-the-vote (GOTV) campaigns, Treas. Reg. § 1.527-6(b)(5).³⁴

In addition, the regulations provide that two sets of expenditures will be treated as taxable only to the extent provided in two other sections, Treas. Reg. § 1.527-6(b)(2) and (3), each of which reads in its entirety, "[Reserved]." Thus, any expenditure that is described as subject to tax only to the extent provided in those contentless sections is not currently subject to tax under section 527(f).

The exclusion reads:

We recognize that historically the IRS has attempted to adopt a more limited reading of the scope of these regulations. The discussion released with the promulgation of the final section 527 regulations cites some specific types of activities expressly permitted by the statutory language of FECA. See T.D. 7744, 1981-1 C.B. 360.

Subsequent internal training materials have reiterated this limited set of activities as determining the universe of activities intended by the IRS to be covered by this reserved section: member communications, and SSF administrative and fundraising costs.³⁹ However, while the drafters of the 527 regulations may have had only these limited categories in mind, the regulations stand as written -- exempting from section 527(f) tax any activity "allowable under the Federal Election Campaign Act" except as provided in a section of regulations that remains undrafted beyond the statement "reserved."

This anomalous situation strongly suggests that any effort to define "political" activity for 501(c) organizations needs to take a look at the entire picture. We encourage the IRS to adopt a rule that will apply across the board and address the problem of these section 527 reserved regulations at the same time. Otherwise, we would be left with a startling incongruity: having borrowed the 527(e) definition of political "exempt function" for use in the (c)(4) definition of candidate-related political activity, the IRS would find that the existing 527 regulations that dictate which expenditures by a (c)(4) organization are subject to the 527(f) tax produce an even more divergent result than we have now. Filling out Schedule C on Form 990, the social welfare organization would need two very different accounting records -- one for disqualifying candidate-related political activity and another for 527(f) tax calculations. So, even though the 527 regulations provide that payments for nonpartisan voter registration and GOTV, and for testifying on appointment confirmations on request, may be made from tax-free investment income, the same activities could cost the organization its 501(c)(4) tax exemption if the NPRM rules went into effect. What counts as political for the 501(c) organization's tax-exemption should not differ so drastically from what is taxed as political under section 527(f).

Recent Congressional enactments indicate no legislative intent to re-define political activity to reach beyond elective public offices and cover appointive and political party offices; the section 527 exempt function definition of political activity stands alone as an anomaly.

We recognize that the drafters of the proposed regulations may have considered it appropriate to look to section 527 for purposes of defining political activity as the most recent Congressional pronouncement in the Internal Revenue Code. However, section 527 is by no means the last word from Congress on this subject. Its more recent passage of IRC section 4955 indicates an intent to leave the definition applicable to 501(c)(3) organizations from 1954 untouched.

IRC section 4955, enacted in 1987,⁴⁰ imposes a tax on a "political expenditure" made by a section 501(c)(3) organization. In addition to applying to "any amount paid or incurred by a section 501(c)(3) organization in any participation in, or intervention in (including the publication or distribution of statements), any political campaign on behalf of (or in opposition to) any candidate for public office,"⁴¹ it also covers a set of other expenditures when made by "an organization which is formed primarily for purposes of promoting the candidacy (or prospective candidacy) of an individual for public office (or which is effectively controlled by a candidate or prospective candidate and which is availed of primarily for such purposes)."⁴² In addition, Congress revisited IRC section 527 in both 2000 and 2002, first adding the current registration and disclosure requirements, and subsequently making minor and technical adjustments.⁴³

The latest Congressional enactment specifying political campaign activity is probably section 501(c)(29), enacted in 2010, which defines a tax-exempt health insurance issuer (under the Affordable Care Act) using the same phrasing as that in section 501(c)(3), referring to political campaigns of candidates for public office. It does not follow the 527 pattern of including appointive and party offices.

All of these legislative enactments post-date the adoption of section 527 and promulgation of regulations thereunder. In none of these cases did Congress give any indication that it was dissatisfied with the application of the campaign intervention standard to define political activity under any of the sub-sections of section 501(c), including section 501(c)(4) that pertains to social welfare purposes. Any argument that the IRS should defer to enactment of section 527 in establishing a definition of non-social welfare political activity fails in light of both the competing definition set out in section 4955 and the lack of Congressional action to extend the section 527 definitions to other organizations.

Widening the Gap between Exemption under Section 527 and under Section 501(c)

Although the proposed regulations limit their reach to the definition of "candidate-related political activity," they request comment on the question of the amount of disqualifying activity that a 501(c)(4) may engage in without losing its status. Or, alternatively, the question is how much of a 501(c)(4)'s activities must qualify as promotion of social welfare in order to maintain exemption under that IRC section. (Just as we believe the same definition of political activity should apply to all 501(c) organizations, we support applying the same standard to determine how much of such disqualifying activity is permitted.)

Current regulations require a 501(c)(4) organization to engage primarily in activities that further its social welfare purpose. In the absence of regulations more precisely defining the scope of "primarily," this has been interpreted to mean that social welfare activities must exceed 50 percent of the organization's overall activities.

An organization that primarily engages in non-social welfare political activity could be expected to qualify for exemption under section 527. (For simplicity, we set aside for now the technical discrepancies in scope of coverage between campaign intervention and section 527 exempt function under existing law and the further divergence that could result from the NPRM rules.) The organization would certainly meet the basic definition of section 527(e)(1), that a "political organization" is "organized and operated primarily for the purpose of . . . an exempt function."

However, a 527 organization's income is exempt from taxation only insofar as it qualifies as "exempt function income." That, in turn, covers certain amounts received (e.g., contributions, dues, fundraising proceeds) but only "to the extent such amount is segregated for use only for the exempt function of the political organization."⁴⁴ This segregation requirement means that an organization that spends 60 percent of its resources on political activity and 40 percent on non-political issue advocacy and lobbying (assuming, arguendo, that a clear line is drawn between those two sets of activities) would not enjoy even the limited tax exemption of section 527 unless it knew in advance what income would be required for which activity and maintained segregated accounts for each purpose. The practical upshot is that a 527 organization with a single general bank account must engage not only primarily in exempt function activity, but that other activity must always be "insubstantial."⁴⁵ Otherwise, it will be taxed on all its (unsegregated) income, including apparently contributions that would otherwise as a general principle of taxation not be included in income.

Similarly, imagine an organization that spends 45 percent of its time and resources on social welfare lobbying, 45 percent on political activity, and 10 percent on other disqualifying activity such as social events. It is not operated primarily for either social welfare or for political purposes. Yet its activities are almost entirely those deemed worthy of tax exemption, albeit in two different code sections.

As a policy matter, these outcomes make little sense. An organization willing to pay tax on its investment income and disclose contributions and expenditures in compliance with the mandates of section 527 should not be constrained in the type of advocacy it may undertake. We recognize that this is peculiar result derives from the odd statutory language of section 527, but it is a situation that IRS regulations adopted to define political activity should avoid exacerbating. Any rule that increases the set of nonpartisan advocacy activities that could disqualify an organization from exemption, such as grass roots lobbying near elections, and nonpartisan voter engagement and voter education, thereby widening the gap between 501(c)(4) and 527 status and driving these organizations into non-exempt status would be counter-productive.

While 501(c)(4) organizations (or (c)(5)s or (c)(6)s) in this situation could create a segregated 527 account for this newly-defined political activity, they may choose to simply exit the tax-exempt system. An organization that falls into this area would not qualify for tax exemption and thus would be treated as a taxable entity. If its revenue comes primarily as unrestricted gifts and contributions rather than earned income, its tax liability could be minimal. As a Form 1120 filer its tax return would not be publicly available, so even the limited information available from the Form 990 would be lost to public disclosure.

We therefore urge the IRS not to adopt any approach to defining political campaign activity that, by including nonpartisan and nonelectoral activities, would further widen the gap between section 527 and section 501(c) organizations.⁴⁶ Rather, Treasury would be better able to promulgate and enforce a system of setting qualitative and quantitative limits on 501(c) political activity by simultaneously indicating to Congress that section 527 could be amended to fit within the system by providing a tax-exempt home for ANY organization that fails to qualify under a 501(c) category due to excessive political campaign activity, even if that activity is less than primary.

The 527 Regulations Should Be Aligned with the 501(c) Rules

For all the reasons discussed above, lack of consistency between the definition (scope of coverage) and standard (method of analysis) employed under section 527 and section 501(c)(4) (as well as other sub-sections of section 501(c)) should be minimized. Ideally, one set of rules would apply for all purposes.⁴⁷

Rather than creating a definition of political activity ("candidate-related political activity") that applies only to 501(c)(4)s, and only for purposes of determining whether they qualify for exemption, the IRS should undertake a more wide-reaching rulemaking. The section 527 regulations should be re-drafted to adopt the same standard as applies to 501(c)s, both for 527 qualification and for application of the section 527(f) tax.

While the IRS may lack the authority to align the scope of offices covered by each rule, it can apply the same test in each context. As described elsewhere, we support a middle ground between the proposed regulations' absolute approach that classifies most activity based solely on when it occurs, and the current approach that weighs "all the facts and circumstances" without any precise, reliable, articulated standards to apply to those facts. A rule that embodies a reasonable standard with explicitly defined exceptions should be adopted by regulation to apply both to the determination of campaign intervention under section 501(c) and exempt function under section 527.

We would also support a limited exception to the application of the 527(f) tax for payment of the administrative and fundraising costs of an SSF, as well as for internal membership communications exempted from the definition of "expenditure" under FECA. This result can be obtained without need to draft the missing "reserved" sections of the regulations, but with minor changes to the language describing the activities not subject to tax except as provided in those sections. Thus, such expenditures would be considered political for qualification purposes if they fell within the common definition of political intervention for all 501(c) entities, but could be excused from the 527(f) investment income tax given the long history, under federal tax law and election law, of allowing nonprofits to perform such functions without requiring that they be transferred to a 527 political fund.

In the Bright Lines Project, we have endeavored to separate those activities that should be clearly regarded as political intervention from those that should clearly not be, narrowing the area in the middle where facts and circumstances determinations can be still be made but with a careful, analytical approach.

This is how we describe our drafting philosophy:

We further state a general threshold test for candidate advocacy that requires reference to and reflection of a view upon one or more candidates, so that statements about issues without reference to candidates or to voter engagement are not treated as political intervention.

However, we provide four safe harbors for speech that may reflect a view on a candidate so that organizations can still influence official actions, can do voter education in which candidates have equal opportunities to speak, can defend themselves when candidates attack them, and can allow personal expression of opinions in meetings, all without engaging in partisan campaign activity.

We believe that it is possible to move away from the past approach taken by the IRS, in which the vast majority of cases were subject to "facts and circumstances" without a commitment to clear standards. We can move toward an analytical system of definitions and exceptions to resolve the vast majority of cases one way or the other, leaving a small minority of cases to be governed by a structured examination of the facts and circumstances because the relevant factors would be difficult to systematize in advance.

The Bright Lines Project Drafting Committee has developed, in its deliberations through December, 2013, a capsule Summary of federal tax rules that would provide objective standards to encourage nonpartisan civic participation in our democracy while discouraging partisan abuse, appended to this submission as Exhibit A, with a number of footnotes providing interpretive details.

Further details of the Bright Lines Project proposal are presented in the July 2013 Explanation, available at our website, www.brightlinesproject.org. We are working on an update to the Explanation, which will be based on the December 2013 Summary and will contain further discussion of voter engagement, the need to coordinate a new universal definition with the provisions in the statute and regulations under section 527 (which we think requires Congressional attention), and other issues that have arisen during the NPRM public comment process. We hope to release the next Explanation in March, and request leave to supplement our submission to the IRS and Treasury at that time.

Our objective is to design a high-quality, fair, universal, practical and predictable way to distinguish what is political intervention from what is not. To facilitate understanding of how our approach differs from that of the NPRM, we next want to compare and contrast the features of the draft rules presented in the NPRM with the corresponding features of the alternative rules we offer, in the form of a table. Items in red indicate problems with the NPRM draft rules, and green indicates what we believe to be a better way to resolve the issue.

Issue

By far, the most significant problem with the NPRM draft regulations is the fact that its general speech test is based only on express advocacy. Yes, there are harsh treatments of voter registration, GOTV, and voter guides, as well as public communications referring to candidates and candidate appearances at events within the 30/60 day pre-election periods, even if nonpartisan in character, but practically everything else an organization might say is subject only to an express advocacy test. This would turn tax-exempt campaign messaging, during most of the election year, into a Wild West show like nothing we have seen in the past. That is why our "reflect a view on a candidate" threshold test, which captures anything said that indicates the speaker's candidate preference, is essential for the credibility of the nonprofit sector as a genuinely nonpartisan voice.

Use of Election-Law Standards Undermines Proposed Definition of Candidate-Related Political Activity

As part of its stated desire to set a clear, objective standard, the NPRM imports key elements of its definition of "candidate-related political activity" from federal election law, but, in so doing, the NPRM ignores a significant amount of political activity, seeks to regulate a significant amount of non-political activity, and arguably undermines the effort to create a bright-line standard.

The NPRM relies too heavily on federal election laws that were constructed for purposes that differ from tax law and apply only to contests for federal offices. The regulations should not be tied to the election law of any one jurisdiction, because they must apply throughout the country to state and local elections as well.

Federal Election Law Elements and Constitutional Underpinnings

Several elements of the proposed definition of candidate-related political activity are either drawn from or directly cite federal election law standards for political activity. Candidate-related political activity is defined to include:

The 30- and 60-day pre-election windows referenced in sections (2) and (8) of the NPRM are also taken directly from federal election law. In 2002, Congress passed the Bipartisan Campaign Reform Act ("BCRA"), better known as the McCain-Feingold Act, after its lead Senate sponsors. Frustrated by Buckley and subsequent rulings in federal courts that struck down efforts to regulate political speech more broadly than the magic words of express advocacy, the drafters of BCRA sought to regulate a new category of communications that the law called "electioneering communications." BCRA set out an objective, bright-line standard that sought to avoid inquiry into the intent of the communication. As defined in BCRA, electioneering communications are broadcast, cable, or satellite communications that refer to a clearly identified candidate for federal elective office that is distributed within 30 days of a primary election or nominating convention or within 60 days of a general election for that office, and, in the case of House and Senate candidates, are targeted to the district or state in which the candidate is seeking election.⁵¹ BCRA banned spending by corporations and unions for such electioneering communications and imposed disclosure requirements on electioneering communications run by others.⁵²

Although they acknowledged that communications outside of these time windows (or, for that matter, communications made via media other than television or radio and non-targeted communications) might also be attempts to influence elections, BCRA's supporters argued that the definition of electioneering communication was narrowly enough drawn and sufficiently related to the compelling government interest of preventing corruption that the provision should survive the First Amendment scrutiny that had proved fatal to earlier efforts to regulate non-express-advocacy speech. The attempt was not as successful as the proponents had hoped. While the ban on corporate and union spending for independent electioneering communications was initially upheld against constitutional challenge,⁵³ the ban was ultimately struck down as applied to communications that were not the "functional equivalent" of express advocacy, as described above in FEC v. Wisconsin Right to Life,⁵⁴ and then completely struck down in Citizens United.⁵⁵ (The electioneering communications disclosure requirements were upheld in Citizens United and still stand.)

In short, the NPRM imports a significant portion of its definition of candidate-related political activity from federal election law, and most of these are concepts developed to avoid constitutional issues associated with regulating electoral communications.

It is also important to note that the NPRM would create federal tax rules that would not only apply to federal election-related activity, but also to tax-exempt programs conducted at the state and local levels, where election laws may differ markedly. California, for instance, sets the pre-election window for reportable communications mentioning a candidate at 45 days before a primary or a general election rather than 30 or 60 days, and the span of communications affected differs from FECA and from the NPRM.⁵⁶ A number of states have no such pre-election time periods in which special rules apply. The NPRM would impose features of a system designed for federal elections on state and local jurisdictions that never agreed to such standards.

Different Constitutional Review for Restrictions on Exempt Organizations

The strict scrutiny the courts have applied in reviewing (and frequently striking down) attempts to regulate political speech under federal election law is, however, a far more stringent standard than the review that the courts have given to regulation of the activities of tax-exempt organizations. Federal courts have consistently upheld restrictions on the lobbying and political activities of tax-exempt organization against First Amendment challenges similar to those that brought down election law provisions in Buckley, Wisconsin Right to Life, and other cases.

The courts considering the restrictions on tax-exempt organizations have reasoned that the strict scrutiny normally applied in the case of First Amendment challenges is not required when the organization has voluntarily sought exemption from tax in full awareness of the limits accompanying that status. In Christian Echoes National Ministry, Inc. v. United States, for example, the court upheld the absolute ban on political campaign intervention by 501(c)(3)s, explaining:

Use of Election-Law Concepts Results in a Definition That Ignores Clear Political Activity

The NPRM's use of elements drawn from federal election law results in a definition of candidate-related political activity that fails to capture a significant amount of political campaign intervention.

A huge scope of communications and activities that would constitute political campaign intervention under section 501(c)(3) of the IRC or exempt function activity under section 527(e) would nonetheless not constitute express advocacy or the functional equivalent thereof. For example, a website could be devoted to what an organization perceives as the shortcomings of an incumbent candidate who is facing a difficult re-election challenge and ask the public to encourage the incumbent to change her ways. A direct mail piece targeting voters in a congressional district could spread accusations of a candidate's supposed scandalous behavior and ask that readers support the organization's ongoing efforts to reveal such scandals. An organization could lionize a candidate for supporting a policy position important to the organization, nominally as part of the organization's effort to support that policy position. In each example, the "magic words" of express advocacy would be absent, and there would be a putative non-electoral reason for the communication. Each could be a powerful tool in influencing an election, and each would include "facts and circumstances" that could suggest political campaign intervention under the longstanding IRS guidance. Yet none of these examples includes "words that expressly advocate" the election or defeat of a candidate and each of them is susceptible to a "reasonable interpretation other than a call for or against" a candidate.

The NPRM's proposed rule would, of course, still treat such communications referencing or featuring candidates as political if the communications were to fall within the time windows borrowed from the federal election-law definition of "electioneering communications." However, the NPRM would not treat any of the hypothetical examples above as candidate-related political activity if they occur more than 60 days before the general election and more than 30 days before the primary election or nominating convention.

It is not necessary to import federal election law standards that result in a definition of exempt-organization electioneering that fails to encompass clear political activity. Buckley's "express advocacy," Wisconsin Right to Life's "functional equivalent," and BCRA's electioneering communication time windows, were compromises that attempted to avoid First Amendment complications. The courts have consistently held that constitutional calculus is different in the context of defining political activity for tax-exempt organizations. Treasury is free to craft a bright-line rule that defines political activity more comprehensively than federal election law does.

Unnecessarily adhering to an overly narrow definition of political activity would lead to tax-exempt organizations created and operated to exploit the rules' loopholes. Were the proposed regulations to be adopted, 501(c)(4)s would engage in unrestrained partisan electoral activity (albeit without the "magic words") outside of the narrow electioneering communication time limits.

Use of Election-Law Concepts Results in a Definition That Mistakenly Captures Non-political Activity

Not only is a definition of candidate-related political activity that relies on federal election law concepts under-inclusive, as described above, but the absolute reliance on the 30- and 60-day electioneering communication time windows at the same time improperly characterizes nonpartisan activities as political.

Not all communications or events featuring candidates for office in those time windows are or should be treated as political for purposes of tax exemption. IRS guidance has long acknowledged that some exempt organization references to candidates are non-political, even close in time to an election. A communication referencing a candidate just prior to an election may be a lobbying or similar advocacy communication, not a political effort.⁵⁹ Or the communication might reference a candidate in a way clearly unrelated to influencing the outcome of the election.⁶⁰ An event at which a candidate appears just prior to an election, might be a non-partisan debate or an event in which the appearance is in a capacity unrelated to the individual's candidacy.⁶¹

Moreover, the NPRM has imported the electioneering communication time windows into the definition of candidate-related political activity without some of the additional requirements that BCRA had included in an effort to limit this overbreadth. Where BCRA limited the definition of electioneering communications to television and radio ads, the NPRM would apply to these absolute time thresholds to all forms of "public communication," including print media, communications intended to reach more than 500 people (mail, email, and large meetings), and social media and other websites (even, apparently, websites created long prior to the election that remain available for viewing during the time windows). Where BCRA had the additional requirement that an electioneering communication be targeted to the electorate from whom the referenced candidate was seeking votes, the NPRM would include communications and events not aimed at or accessible to any of the voters deciding the candidate's electoral fate.

Again, it is not necessary to adopt election law standards created to address constitutional concerns not applicable in the context of defining political activity for tax-exempt organizations. The overbreadth of such standards is just as much a reason to reject them as is their under-inclusiveness.

Use of Federal Election-Law Concept Depends on Unsettled Law

The NPRM makes it clear that the primary goal of the proposed rule is a desire to create a clear, objective standard that is easily applied by both organizations subject to the rule and IRS officials charged with enforcing it. As noted, the decision to incorporate federal election law concepts into the proposal appears to be an attempt to achieve that goal. However reliance on federal election law offers no such promise and at least one of the election law concepts the NPRM adopts could actually add subjectivity and uncertainty for years to come.

The NPRM would treat as candidate-related political activity a communication that:

In contrast, the question what communications are "susceptible of no reasonable interpretation other than. . . ." is subject to different interpretations. Some argue that "functional equivalent" adds a very small additional margin to the magic words of express advocacy; others argue that anything affecting the voters' impressions of a candidate, positively or negatively, should be covered. The Supreme Court, in Citizens United, held that "Hillary: the Movie" was the functional equivalent of express advocacy against her candidacy for President.⁶²

The difficulty with using this federal election law standard in tax-exempt law matters is graphically illustrated by IRS Private Letter Ruling 9725036. In that situation, a 501(c)(4) organization set up a 527 separate fund that was barred from engaging in express advocacy as defined under FECA. However, it planned to engage in various activities during election periods, including grass roots lobbying that had a dual character -- to influence the named public official's action on pending legislation "but also, by strong implication, raising public awareness about how the identified legislators stand on issues that X believes voters, based on opinion polling, will take into account in making judgments about the *** positions of the candidates." The IRS found that t he activities constituted political "exempt function" under section 527(e)(2). The NPRM would likely reach the opposite result, and allow the activity to be conducted outside of the 30/60 day pre-election period by the 501(c)(4) organization; lacking express advocacy and susceptible of a reasonable interpretation that the messaging had a grass roots lobbying purpose, it would not be "candidate-related political activity." Our safe harbor test for "influencing official action" would not shield the speech from treatment as political intervention because a dual-purpose communication would not have a limited relationship to the public official's prospective actions, or because it consisted of paid mass media advertising, or was targeted to close elections.

Imagine further, a closely fought election in which a challenger seeking to unseat the incumbent candidate is a college professor who has never before sought or held any public office. Suppose an advertisement published just prior to an election uses slender evidence to accuse the challenger of an abhorrent policy position and urges readers to call the challenger (providing the phone number of her college office) and to express disgust about the policy position. Is it reasonable to interpret this as an effort to get the candidate to change her views? Perhaps not, but it is almost certainly subject to a reasonable interpretation (call the professor) that is not a call to vote for or against a candidate.

In the end, no definition of political activity can or should create a robotic standard that eliminates all human judgment. But a bad definition should not be adopted based on the mistaken belief that it somehow succeeds in creating a clear standard when, in fact, it only trades old ambiguities for new ones.

A More Limited Use of Federal Election Law Concepts Is Appropriate

There are appropriate ways in which the definition of political activity for tax-exempt organizations should make reference to election law concepts.

In the Bright Lines Project, we make limited use of references to election law definitions, suggesting that as to an organization's "use of resources," it is political intervention to make a contribution (including coordinated in-kind expenses) reportable under candidate campaign finance laws in the applicable jurisdiction. The NPRM takes this approach as well in section (4)(i) of its proposed definition of candidate-related political activity.

Likewise, the Bright Lines Project agrees that express advocacy for or against a clearly identified candidate for elective office or candidate of a specific political party should be deemed to be political under any definition. (As noted elsewhere in these comments, we differ from the NPRM in its proposed broadening of the definition of "candidate" to include candidates for non-elective office.)

However, we think that there are better bright-line standards for defining political activity that don't have the defects of BCRA's 30-and 60-day electioneering communications time windows and Wisconsin Right to Life's "susceptible of no reasonable interpretation" / functional equivalent of express advocacy standard, particularly as those concepts have been imported into the proposed regulations.

In defining candidate-related political activity, the NPRM suggests a set of bright line standards. In almost every case, the suggested standard fails to distinguish efforts to influence elections that are legitimately regulated as political activity from other election and non-election efforts that historically have been and should continue to be treated as non-political charitable and educational activity. Such a misguided definition would chill vital activities that serve society as a whole. Below, we discuss each of these bright lines that the NPRM draws in the wrong place.

Definition of "Candidate"

The proposed (and current) definition of "candidate" is overly broad because it: (i) reaches anyone who may be "proposed by another" as a candid ate; and (ii) includes those who are appointed to an office as well as those elected by the public.

The NPRM defines "candidate" as "an individual who identifies himself or is proposed by another for selection, nomination, election, or appointment to any public office or office in a political organization, or to be a Presidential or Vice-Presidential elector, whether or not the individual is ultimately selected, nominated, elected, or appointed. In addition, the proposed regulations clarify that for these purposes the term "candidate" also includes any officeholder who is the subject of a recall election." (Italics added)

Part of the phrasing of the NPRM's definition of "candidate" appears to come from the current "action organization" regulations that pertain to the 501(c)(3) ban on political intervention. Treas. Reg. 1.501(c)(3)-1(c)(3)(iii) reads as follows: "The term 'candidate for public office' means an individual who offers himself, or is proposed by others, as a contestant for an elective public office, which such office be national, State, or local." That definition should be revisited, not just for 501(c)(4) entities, but for (c)(3) charities and for all tax-exempt organizations.

Capturing as a "candidate" an individual whose ascendancy is suggested by someone else is not only overly broad but impossible to anticipate. At any given time, there are dozens if not hundreds of prominent actors, business executives, athletes, governors, senators, authors, and other public figures on the national stage that one or more individuals have stated ought to run for President. The same is no doubt true in cities and municipalities where citizens, journalists and public policy interest groups recommend individuals they believe ought to be their jurisdiction's next mayor. The opportunity for opponents of an organization to use this definition to mischievous ends is boundless -- someone may be proposed as a candidate for a public office simply to prevent a social welfare (or charitable organization) from promoting or criticizing them in a particular community.

Employing a definition of "candidate" tied to the current standard contained in section 527 of the tax code is not only overbroad, but will tie the IRS's determination to absolutes susceptible of no alternative outcome. Using a definition that is not subject to rebuttal or presentation of refuting factors for further analysis by the IRS specific to the taxpayer's pertinent facts and circumstances in this instance is both unfair and a false signifier of clarity.

A better end would be to have "candidate" defined more narrowly and with a more appropriate, and brighter-line, definition. The term should reach two pools: only those who have announced themselves as candidates and those "whose election the organization expressly proposes, supports, or opposes." With such a reconfigured definition, political intervention, then, would only pertain to unannounced candidates if the organization itself expressly took a position favoring or disfavoring their election to office, rather than as a result of simple expressions of preferences by other individuals. This narrower definition also avoids widely-disparate definitions used in election laws by various jurisdictions to determine who has qualified to appear on a ballot (and thus is a "candidate") or who must file campaign finance reports because of acting to advance an individual's "candidacy" (although any candidate who is on the ballot or has registered a campaign committee will almost certainly meet the standard of "offers themselves"). This standard will thus apply consistently across all jurisdictions.

It is further problematic and troubling that the NPRM's proposed definition of "candidate" captures anyone who has been nominated for appointment to a governmental position, such as a judgeship, head of a public utilities board, cabinet secretary, or human rights commission. The current Regulation defining "candidate" for 501(c)(3) organizations, cited above, adds the word "elective" to the statutory phrasing of the charitable ban on intervening in the campaigns of candidates for "public office," and so it has been understood for decades.

While it may be appropriate to allow section 527 political organizations to embrace activities supporting or opposing various nominees for such non-elective public office, extending that broad concept to consider such effort as political intervention activity when conducted by 501(c)(4) organizations (or, potentially, other 501(c) entities) is over-reach. Social welfare organizations, 501(c)(3) charities, and other 501(c) exempt organizations commonly express their preferences for nominations and appointments to non-elective public offices as part of their permitted and appropriate lobbying activity in favor of their missions, not as political intervention or candidate-related activity. Lobbying for or against the selection or appointment of non-elected officials is a critical and appropriate exercise for many nonprofit organizations, and it is inappropriate and confounding to have such activity count as candidate-related activity and thus comprise political intervention. In a general counsel memorandum,⁶³ the IRS backed away from that position: while it could apply the 527(f) investment income tax on the expenditures of a 501(c) organization to support or oppose judicial or other executive branch appointments, it has never done so.

Definition of "Express Advocacy"

The proposed threshold definition of political speech that lays much of the groundwork for the term candidate-related political activity is unnecessarily limited to "express advocacy" and its functional equivalent and captures far too little electioneering activity.

The NPRM states that its definition of candidate-related political activity draws from FEC rules on express advocacy but expands "the concept to include communications expressing a view on the selection, nomination, or appointment of individuals, or on the election or defeat of one or more candidates or of candidates of a political party. These proposed regulations make clear that all communications -- including written, printed, electronic (including Internet), video, and oral communications -- that express a view, whether for or against, on a clearly identified candidate (or on candidates of a political party) would constitute candidate-related political activity." (Italics added)

This definition, borrowed from federal election law, initially states that it would capture any communication made by a 501(c)(4) social welfare organization that expresses a view on the selection, nomination or appointment of individuals as political intervention. However, the actual text of the proposed regulations, quoted above, limits the speech deemed to be political to (i) the magic words of express advocacy and (ii) its functional equivalent. The only expansion beyond the narrow scope of federal election law appears to be the inclusion of appointive offices.

A standard for determining candidate-related activity based on whether a communication identifies an individual candidate and expresses a view (but NOT limited to express advocacy of election or defeat) on that candidate is more workable as a starting point, as long as it is coupled with carefully and clearly specified exceptions. That is the approach taken by the Bright Lines Project.

Because of the difficulty defining the reach of express advocacy rules under election laws, we opt for a threshold definition of candidate advocacy that covers any communication referring to and reflecting a view on a candidate (similar to a tax law definition of lobbying under sections 4911 and 4945 that has worked well for 20 years), and build upon that with further definitions of voter engagement, targeting to close elections, paid mass media advertising, and safe harbor exceptions.

The Bright Lines Project proposes to employ a similar definition of political intervention ("communications that clearly identify a candidate and express a view on that candidate") as a threshold speech test, but then build upon it with well-crafted and precise safe harbors. Before using that broad standard, we apply the definition of "express advocacy" as narrowly defined under FECA (like the NPRM's proposal) as a starting point in finding political intervention. As designed by the Bright Lines Project, it is always political intervention to expressly advocate:

Under our proposal, communications that refer to and reflect a view on candidates but are not within a safe harbor are not per se political intervention, but may be defended by the organization in a structured analysis of facts and circumstances by which it may show that the speech furthered a proper exempt purpose, unrelated to intervening in any political campaign.

Communications Covered During Pre-election Periods

As previously noted, the NPRM borrows a modified version of "electioneering communications" from FECA and applies it to its new definition of "candidate-related political activity" that would replace the term "political intervention" for 501(c)(4) social welfare organizations. Under the proposal, candidate-related activity that fails to constitute social welfare activity includes public communications by a (c)(4) that clearly identify a candidate for public office and are disseminated through one of the listed means of communication that is capable of reaching 500 or more people within 60 days or a general election or 30 days of a primary election.

The NPRM's modified version of electioneering communications for defining candidate-related activity is very problematic in that it is both over-inclusive within the 60-day and 30-day timeframes and under-inclusive outside those timeframes.

We believe that use of fixed time periods (which may make sense in the context of campaign finance disclosures) is not a good way to define political intervention in tax law, because doing so inevitably suppresses non-political speech close to elections and encourages abusive methods of influencing elections outside of those periods.

It is not well-understood, but the NPRM appears to open the door to unlimited candidate-related advocacy outside of the election windows, short of express advocacy. This will surely result in a 501(c)(4) organization becoming the vehicle of choice, using undisclosed donations, for thinly-disguised "issue ads" broadcast during earlier times in the campaign season using the greater part of its resources, while using the lesser part of its resources for explicit campaign advertising during the 30 and 60 day pre-election windows.

Under federal election law, the term "electioneering communications" includes any (i) broadcast ad that (ii) refers to a clearly identified federal candidate within (iii) sixty days of a general election, or thirty days of a primary election, and which (iv) targets that candidate's constituency.⁶⁵ An ad is considered targeted to a candidate's constituency if it can be received by 50,000 or more persons in the candidate's district (as determined by the Federal Communications Commission).⁶⁶ "Broadcast ads" include television, radio, cable or satellite advertisements, but do not include Internet, telephonic or print advertisements.⁶⁷ The term in FECA was originally designed as its own bright line standard for narrowly but clearly drawing the line between communications subject to the source prohibitions and disclosure requirements of the campaign finance law, though the prohibitions against direct corporate and union financing of electioneering communications have since been voided in the 2010 Supreme Court decision, Citizens United v. Federal Election Commission.⁶⁸

As stated above, the proposed NPRM modifies the concept of electioneering communications as applied to social welfare organizations under the tax code in two critically important ways. First, it expands the concept to include all communications disseminated by any means, including Internet, telephonic or print advertisements. Second, it does not require that the public communication target a candidate's constituency. It merely requires that the communication reach 500 or more persons anywhere.

Over-Inclusive During an Election Window

This modification of the concept of electioneering communications makes its application in defining candidate-related political activity for social welfare organizations over-inclusive. Any communication disseminated to the general public through any means near an election that mentions a candidate's name for any reason would be deemed candidate-related activity, including a simple notification to the public of an upcoming candidate debate. Social welfare organizations would be constrained in sponsoring or advertising candidate debates or even issue forums in which lawmakers argue for or against legislation and public policies. Discussions in an organization's newsletter distributed to its members of a school board member's proposed changes to the district's dress code, if the member were running in an upcoming election, would qualify as candidate-related political activity.

Simply noting the sponsors of legislation on an organization's web page near an election would cross the line. The NPRM is clear on this:

Even communications that are truly nonpartisan and do not promote or oppose the election of any candidate or political party would be captured as candidate-related activity if the communications identify the name of persons seeking election or appointment to public office. This includes distribution of nonpartisan voter guides, like a ballot handbook published by the local government, that allow a forum for candidates to present themselves. All campaign news coverage by an organization that publishes a magazine or newspaper during the last month or two before the election, and all reporting of public opinion polling, would be deemed political under the NPRM. Not only would advertising candidate debates be captured, but so would organizing and staging candidate debates or forums near an election if it could be expected that 500 people would show up. In short, the NPRM's proposal would impose major, if not crippling, constraints on legitimate lobbying activity and nonpartisan educational programs that many social welfare organizations envision as part of their core mission.

Under-Inclusive Outside an Election Window

Just as troubling as capturing lobbying activity and nonpartisan educational programs as candidate-related political activity during an election window, is what the proposed NPRM does not capture outside the election window. Television ads and other public communications that clearly identify a candidate, target that candidate's voting constituency, and promote, attack, support or oppose that candidate without using the magic words of express advocacy would not count as candidate-related activity if disseminated 61 or more days before a general election or 31 or more days before a primary.

A study on television campaign advertising that served as the case record in McConnell v. Federal Election Commission, upholding the Bipartisan Campaign Reform Act, documented the general absence of express advocacy in campaign messages throughout the election cycle.⁶⁹ Of all the major categories of sponsors of campaign ads -- candidates, parties and outside groups - none of them employed the terms of express advocacy such as "vote for" or "elect" or anything comparable with much frequency. Only 2 percent of campaign ads in the 2000 federal elections sponsored by party committees and outside groups used terms of express advocacy. Even candidates rarely employed express advocacy, using such terms in only 10 percent of their ads. Sponsors of campaign ads prefer subtlety in their electioneering messages over express advocacy, focusing on a candidate's personal qualities or fitness for office.

With more money flowing into elections than ever before, and more outside groups competing with candidates and party committees to buy up expensive air time, campaigns are starting in full swing much earlier. Even in terms of campaign finance law, the 60-day and 30-day election windows for electioneering communications miss a great deal of campaign advertising. In the 2012 presidential election, for example, five and a half months before Election Day, Crossroads GPS, a 501(c)(4) nonprofit group formed by political strategists, unveiled a $25 million electioneering advertising campaign. The group opened in May 2012 with a hard-hitting commercial that cast President Obama as a failed leader, unable to deliver on his pledges to fix the country's problems.⁷⁰

More than five months before North Carolina's congressional primary in the 2014 election cycle, and almost a year before the general election, outside groups have spent more than $8 million on campaign ads designed to affect the outcome of that state's election for the office of U.S. Senator. Similar early campaign spending by outside groups in the 2014 election cycle is occurring in senatorial contests in Arkansas, Iowa, Louisiana, Michigan and Kentucky and House races in Arizona, Minnesota and New Hampshire as well.⁷¹ These are costly television ads that focus exclusively on specific candidates, reflecting positive or negative views on them, yet none of this campaign advertising would be captured as candidate-related political activity by the NPRM proposal because the ads contain no words of express advocacy and the presence of its functional equivalent is debatable at best.

While the 60-day and 30-day election windows are proving insufficient for campaign finance law, they are woefully inadequate for determining exempt status under the tax code. The tax code has very different objectives than campaign finance law -- particularly in this case, determining whether an organization is entitled to certain tax privileges as a social welfare organization. There is no timeframe under the tax code for when such a group must have social welfare as its primary purpose and when it need not, other than the somewhat arbitrary measurement of the organization's tax year. It is always supposed to have social welfare as its primary purpose. So when communications come from a nonprofit organization that conspicuously promote, attack, support or oppose candidates for public office at any time, absent other mitigating factors, they should be considered political intervention.

The Bright Lines Project provides concrete principles that define what is and what is not political intervention in a ways that avoid both the over-inclusiveness and under-inclusiveness posed by the timeframes offered in the NPRM's proposal.

The Transfer Rule in the NPRM

The NPRM treats a contribution to a 501(c) entity as political unless use of contribution for politics is prohibited and the recipient states it does not engage in politics, both in writing. This is a draconian transfer rule that would treat a contribution as political even if not a dime of it is used for political intervention. The Bright Lines Project suggests a more reasonable alternative under the heading Use of Resources.

The definition of candidate-related political activity in the NPRM includes a "contribution (including a gift, grant, subscription, loan, advance, or deposit) of money or anything of value to or the solicitation of contributions on behalf of . . . [a]ny organization described in section 501(c) that engages in candidate-related political activity within the meaning of this paragraph. . . ." The NPRM further states that a contribution to a section 501(c) organization will not be treated as a contribution to an organization engaged in candidate-related political activity if (1) "the contributor obtains a written representation from an authorized officer of the recipient organization stating that the recipient does not engage in candidate-related political activity (and the contributor organization does not know or have reason to know that the representation is inaccurate or unreliable)" and (2) "[t]he contribution is subject to a written restriction that it not be used for candidate-related political activity . . ."⁷²

Presumably this rule is intended to prevent section 501(c)(4) organizations from exceeding the limit of their permissible candidate-related activity -- whatever that limit may be -- by contributing funds to another 501(c) organization so that the recipient organization can engage in such activities. However, this rule is overbroad in pursuing that goal and will unnecessarily restrict legitimate activities of section 501(c)(4) organizations in furtherance of their missions that are unrelated to candidates and elections.

The NPRM rule as currently drafted would require that section 501(c)(4) organizations include as part of their candidate-related political activity grants to other section 501(c) organizations that are restricted for a specific purpose having nothing to do with candidate-related activity merely because the grantee engages in some amount of that activity. This might include a nation-wide section 501(c)(4) organization's grant of funds or its in-kind provision of staff to its regional affiliate or its grant to a local grass roots organization that is restricted to support the recipient's public advocacy on a policy matter, unrelated to any candidate or election, in furtherance of the grantor's mission. The granting organization may have made the determination that the most effective or only way of advancing its goals with respect to this policy matter is through supporting the regional organization's work. Even the payment of membership dues to a section 501(c) organization would be considered candidate-related activity by the member if the recipient of those dues engages in any amount of candidate-related political activity, even if funded by other revenues.

These restrictions on legitimate social welfare activity will hamper the work of section 501(c)(4) organizations. The section 501(c)(4) prospective grantor will have to count its grant as part of its own candidate-related political activity even when the grant funds are not used for any candidate-related activity under the broad definition set forth in the NPRM.⁷³ Since current Treasury regulations and IRS rulings limit the amount of candidate-related activity section 501(c)(4) organizations can do, including these contributions as part of their candidate-related activity will unfairly further limit whatever actual candidate-related political activity these organizations may be permitted to engage in. Further, it will force section 501(c)(4) organizations that are nonpartisan to report that they are engaged in political activity even if they don't engage in such activity at all, thereby causing confusion to the public and potentially damaging the organization's reputation. Further, while the NPRM does not address the question of how much candidate-related activity is too much, to the extent a section 501(c)(4) prospective grantor has already engaged in a level of activity that meets its limit, the contribution rule may leave the section 501(c)(4) granting organization with no practical means of engaging in this activity that has nothing to do with candidates or elections. Additionally, to the extent a section 501(c)(4) organization is forced to choose to not become a member of another section 501(c) organization because doing so would cause it to report activity as candidate-related or cause it to exceed its limit on candidate-related activity, the rule would hinder that organization's right to form coalitions with others that it chooses to associate with.

We strongly urge the IRS and Treasury Department to reconsider this component of the definition of candidate-related activity because it is overbroad and will unnecessarily restrict the ability of section 501(c) organizations to support legitimate work conducted by other groups, unrelated to candidates and elections, in furtherance of its tax-exempt social welfare purposes.

Voter Registration and Get-Out-the-Vote Drives

Under the proposed rules, "candidate-related political activity" includes certain specified election-related activities, including conducting voter registration and get-out-the-vote drives, distribution of material prepared by or on behalf of a candidate or section 527 organization, and preparation or distribution of a voter guide and accompanying material that refers to a candidate or a political party. In addition, an organization that hosts an event on its premises or conducts an event off-site within 30 days of a primary election or 60 days of a general election at which one or more candidates in such election appear as part of the program (whether or not such appearance was previously scheduled) would be engaged in candidate-related political activity under the proposed definition.

Voter registration and get-out-the-vote drives have long been acceptable forms of voter engagement that can be conducted by 501(c)(4) organizations to an unlimited degree (and also by 501(c)(3) organizations) -- as long as they are conducted in a nonpartisan manner. If performed following certain very clear parameters, these forms of voter engagement present no threat of political intervention. It is campaign intervention if an organization conducts a voter registration or GOTV drive by saying "register to vote for Candidate X" or "remember to vote tomorrow for Candidate X." It is equally clear that it is not campaign intervention to engage in a project that focuses broadly on the electorate and urges them to register to vote or reminds voters when the election is and urges them to vote. Under the existing IRS interpretations, there are gray areas, but the rules could be easily clarified. The Bright Lines Project proposal lays out objective standards that can easily be applied to voter engagement activities and would provide guidance in a vast majority of circumstances. The IRS could also provide new guidance expanding on the examples provided in Revenue Ruling 2007-41, and thus eliminate even more of this uncertainty.

The IRS provided in Revenue Ruling 2007-41 that it is not campaign intervention to encourage people to participate in the electoral process through voter registration and get-out-the-vote drives, conducted in a nonpartisan manner. On the other hand, registration and voter engagement activities conducted in a biased manner that favors (or opposes) one or more candidates is prohibited. Now, the devil is in the details. The IRS ruling provides two examples to explain where to draw the lines, concluding that one is nonpartisan and one is not.

In the example of a non-partisan activity, an organization that promotes community involvement sets up a booth at the state fair where citizens can register to vote. The signs and banners in and around the booth give only the name of the organization, the date of the next upcoming statewide election, and notice of the opportunity to register. No reference to any candidate or political party is made by the volunteers staffing the booth or in the materials available at the booth, other than the official voter registration forms which allow registrants to select a party affiliation. The IRS concluded that the organization is not engaged in political campaign intervention when it operates this voter registration booth. Under the NPRM, this organization would be engaged in political campaign intervention.

In the example of partisan activity, C, is an organization that educates the public on environmental issues. Candidate G is running for the state legislature and an important element of her platform is challenging the environmental policies of the incumbent. Shortly before the election, C sets up a telephone bank to call registered voters in the district in which Candidate G is seeking election. In the phone conversations, C's representative tells the voter about the importance of environmental issues and asks questions about the voter's views on these issues. If the voter appears to agree with the incumbent's position, C's representative thanks the voter and ends the call. If the voter appears to agree with Candidate G's position, C's representative reminds the voter about the upcoming election, stresses the importance of voting in the election and offers to provide transportation to the polls. The IRS concluded that C is engaged in political campaign intervention when it conducts this get-out-the-vote drive, and this would not change under the NPRM.

These examples are helpful, but more guidance is needed. There are a number of examples that are not as clear cut. For example, how should an organization classify an ad proclaiming "get out and vote for a woman's right to choose" where one candidate is pro-choice and the other is pro-life? What if the upcoming election covers many different offices and the organization's message does not identify any specific race? Similarly, it is unclear what factors can be used in targeting particular voters for GOTV or communities for voter registration efforts. Under current rules, it is campaign intervention to conduct a GOTV effort targeting only registered Democrats. But is it nonpartisan to conduct a GOTV campaign targeting an African-American community where a large percentage of the voters are registered for one party? Rather than deeming all voter registration and get-out-the-vote activities as campaign intervention, the IRS should provide guidance that answers these more difficult, but not insoluble, questions.

The Bright Lines Project proposal provides specific standards for voter engagement activities. If an organization conducts a voter engagement program, even without one of the four forms of express advocacy that we list, it would engage in political intervention if its voter-engagement communications (a) refer to a clearly identified candidate or political party and (b) reflect a view on that candidate or party. Even without speech that refers to and reflects a view on a candidate or party, under the BLP proposal, the use of an organization's resources to do selective or targeted GOTV in a federal election directed to neighborhoods or other groups suggested by a candidate's campaign would come under the federal election law's definition of an in-kind (coordinated) contribution and thus would be intervention under the BLP use of resources rule. A similar analysis would need to be done under applicable state or local law for state or local elections.

If there is no coordination with the candidate's campaign, but the organization on its own decides to target its GOTV to voters using a method to determine the voter's expressed candidate preference, that would be per se intervention. Therefore, the examples in Revenue Procedure 2007-41 (the state fair voter registration and the environmental GOTV based on issue alignment with one candidate) would not change. The state fair voter registration would not be intervention, and the environmental GOTV based on issue alignment with one candidate would be. Furthermore, the BLP proposal suggests safe-harbors for voter engagement targeted to an organization's well-defined natural constituency or to segments of the public with historically low voting participation rates. Where organizations would need to refer to the facts and circumstances, it would still be beneficial for the IRS to provide additional guidance, in particular regarding targeting of voters based on their positions on issues, without any reference to the views of candidates.

A restriction on legitimate forms of voter engagement that combat the decline in voter turnout and contain no bias for or against a candidate or party is not guidance; it is a damper on nonpartisan free speech.

Educational Voter Guides and Activities

The NPRM further defines "candidate-related politic al activity" to include the distribution of voter guides and materials prepared by or on behalf of a candidate, as well as the hosting or conduct of candidate forums, debates and other candidate appearances within 30 days of a primary election and 60 days of a general election.

Treasury and the IRS acknowledge that, under the current facts and circumstances analysis, these election-related activities may not be political campaign intervention if conducted in a non-partisan manner. However, because determining if such activities are in fact nonpartisan is fact intensive, Treasury and the IRS suggest a blanket ban on such activities for section 501(c)(4) organizations.

These sections of the proposed regulations are overly broad and restrictive and would chill legitimate voter engagement and educational activities. From the Presidential debates, each watched by approximately 60 million Americans in 2012, to the local candidate forums held in community centers, schools and churches around the country, nonprofit organizations have played a vital role in educating voters on the issues and where the candidates stand on such issues. Similarly, voter guides on a limited number or a wide-range of issues may be the most detailed, if not the only, information that voters can easily access about candidate and their views, especially in local races.

The statutory framework established by Congress for civic and political education and engagement by organizations exempt under section 501(c)(3), section 501(c)(4) and section 527 attempts to draw lines between legitimate, nonpartisan voter education, appropriate in unlimited amounts for section 501(c)(3)s and 501(c)(4)s, and political intervention on behalf of or in opposition to any candidate, appropriate for section 527 organizations and in only limited amounts for section 501(c)(4)s and prohibited for section 501(c)(3) organizations.

By arbitrarily placing voter guides and late-season candidate debates and forums on the political intervention side of the line only for section 501(c)(4)s, the proposed regulations establish a confusing and inconsistent standard that would permit section 501(c)(3) organizations to conduct certain political activities, if done in a nonpartisan manner, while restricting such activities for section 501(c)(4)s regardless of how the activities are conducted. The cure for this inconsistency is not to prohibit such activities for section 501(c)(3)s. To do so would stifle a vital form of civic engagement that is relied on by voters trying to make informed choices. Rather, Treasury and the IRS must provide consistent guidance on these issues.

The proposed NPRM voter education rules also leave open questions about other types of civic education efforts. Where would legislative report cards fall, which compare the voting records of incumbent legislators to the preferred record of a particular organization? Are such scorecards really voter guides and, if so, so they fall on the side of political intervention in all cases?

The reason given by Treasury and the IRS for creating this overly broad restriction is, in effect, that the facts and circumstances test is too difficult to apply in these areas. Treasury and the IRS have provided very helpful guidance on debates, forums and voter guides over the years. But this guidance, in many ways, is obsolete, not having been updated for the Internet and modern social media age. The IRS needs to provide clearer guidance, updated with more examples,that will help nonprofits and the IRS draw the line between nonpartisan voter education and political intervention more easily and consistently.

The Bright Lines Project has developed safe harbor provisions that will help the IRS and nonprofits balance the need for nonpartisan voter education while more effectively restricting the extent to which nonprofits support or oppose candidates for election. Under the Bright Lines proposal, a tax-exempt organization may engage in unlimited voter education that compares two or more candidates for elective office, even if the voter guide includes the organization's views on the issues discussed, and as long as the communication consists solely of content in which the time, text, and/or space is offered in equal shares to each of the participating candidates and the organization's own views and the organization's share of content is no greater than the share available to any of the participating candidates. The opportunity to participate must be given, including a full description of the opportunity and a copy of the organization's content, to all current candidates (or to those meeting an objective threshold of viability) for election (or nomination) to the office. All candidates must be given a reasonable time in advance of the final preparation of the communication to provide their views, which in no case may be less than 72 hours.

The key to the acceptance of these forms of candidate comparisons as a nonpartisan activity lies in the even-handed methodology used to produce the voter education, whether it is a debate, the publication of answers to a candidate questionnaire, or a request that each candidate record a 30-second video statement on the organization's central issue for its website. Under the safe harbor, the candidates always have a fair chance to present their positions, and this equal opportunity to speak places the power to create an unbiased voter education product in the competitive hands of the candidates. A similar approach was used by the IRS and the Justice Department to approve a nonpartisan format for the 501(c)(4) Christian Coalition's voter guides in 2005. The NPRM would declare that result to be candidate-related political activity and undo years of effort to settle the question of the Coalition's qualification as a social welfare organization.

Under the BLP safe harbor, for example, a group which supports women's reproductive health choices would be able to publish "voter guides" on these is sues, clearly expressing the organization's views on each question. Religious organizations would also be permitted to compare their views on social issues and those of the candidates. Currently, the rules would not clearly permit an organization to compare its views to those of the candidates without engaging in political intervention. The IRS also views as a negative factor (one indicating political intervention) if a voter guide focuses on a narrow set of issues. Under the Bright Lines proposal, this restriction would be lifted. An environmental group would no longer need to worry about preparing a Presidential voter guide that focuses exclusively on environmental issues and would not have to include foreign policy and social issues in the guide, which makes no sense from a programmatic or voter education perspective.

It seems even less difficult for Treasury and the IRS to create clear guidance explaining how section 501(c)(3) and section 501(c)(4) organizations may distribute candidate materials. An organization should be able to distribute materials on all legitimate candidates, on the same topics, in similar volume and using the same methods. If presented as candidate materials, they must be created by or under the control of the candidates, to ensure that the materials reflect each candidate's actual views.

Treasury and the IRS should not abandon their efforts to formulate rules that balance the need for voter education without banning nonpartisan activities that have long-been a mainstay of nonprofit civic engagement efforts because such efforts would be too difficult.

The NPRM picks out a list of programs and events and declares them to be candidate-related political activity though they may be conducted in an entirely nonpartisan manner No safe harbors are provided. The Bright Lines Project opts for broader descriptions of speech activity and uses a series of definitions and exceptions to separate the partisan from the nonpartisan in the practical world, so that democratic participation in our republic may flourish.

To be successful, a definition of political activity must apply as universally as possible throughout the tax code, applying not only to 501(c)(4)s, but also to other tax-exempt organizations and to other aspects of the tax code where a definition of political activity is relevant, such as the limitations on business deductions under I.R.C. section 162(e) and other sections cited in these comments. A successful definition will have bright-line standards, as the NPRM attempts, but those lines must be more carefully drawn and put in the right places. They must better capture true partisan political activity and explicitly exempt truly nonpartisan election-related and non-election-related activities.

Specifically, we recommend that:

• The IRS and Treasury move forward with all due speed on this critical and desperately-needed rulemaking.

• Political intervention (a better phrase than candidate-related political activity) should be defined uniformly for all nonprofit organizations, including charities, unions and business leagues, as well as for other related sections of the tax code.

• Political intervention should be defined according to bright lines standards that prevent excessive abuse of nonprofit status by political operatives, but still encourage nonpartisan democratic participation by nonprofit groups.

• Tax code definitions of political intervention should not rely too heavily on federal election law definitions, the latter of which are designed for somewhat different purposes and apply only to federal offices.

• The definition of "candidate" should not include those who are "proposed by others" nor those selected for appointive offices.

• The NPRM's adaptation of FECA's 60-day/30-day "electioneering communications" windows should not be used to define political intervention in the tax code because it is both over-inclusive of legitimate lobbying activity within the windows, and under-inclusive of candidate electioneering activity outside the windows.

• The treatment of payments to other entities needs to be narrowed so as not to capture non-electioneering grants and transfers among nonprofit organizations.

• Non-partisan voter registration and get-out-the-vote drives, voter guides and candidate forums should be viewed as educational democratic participation and not be defined absolutely as political intervention.

The unprecedented level of interest in this NPRM gives us hope that many others are prepared to engage with us and Treasury to craft a better rule. As occurred in the 1980s with the 501(c)(3) lobbying regulations, we urge Treasury and the IRS to collaborate with the nonprofit sector and those who advise the sector to create the new, improved regulations. The Bright Lines Project looks forward to participating in that effort and to submitting comments on the resulting next proposed draft.

Thank you for undertaking this long-needed and difficult work to address these important issues and for considering our comments and the thoughtful comments submitted by others. We will continue to work with you in the future -- constructively -- to bring this rulemaking to a successful conclusion.

 

Exhibit A

 

 

Bright Lines Project Summary

 

 

December 2013

 

 

[ Editor's Note: For a larger, searchable version of the

 

Exhibit A, see Doc 2014-29185 , pgs. 50-56.]

 

 

FOOTNOTES

 

²See, e.g., Comments of the Individual Members of the [American Bar Association Tax Section] Exempt Organizations Committee's Task Force on Section 501(c)(4) and Politics, May 25, 2004, available at http://www.americanbar.org/content/dam/aba/migrated/tax/pubpolicy/2004/040525exo.authcheckdam.pdf(last visited February 26, 2014), highlighting the need for guidance defining and setting permissible limits on 501(c)(4) political activity.

³ Treasury Inspector General for Tax Administration, Inappropriate Criteria Were Used to Identify Tax-Exempt Applications for Review (Reference Number: 2013-10-053).

⁴ 558 U.S. 310 (2010).

⁵ I.R.C. § 7805. We note that some commenters on the NPRM have suggested that the recent decision in Loving v. Internal Revenue Service, __F.3d. __, 2014 WL 519224 (D.C. Cir. Feb. 11, 2014), would prevent Treasury from proceeding with this rulemaking, but that is not the case. The Loving court noted that the IRS had previously expressly indicated that it lacked the authority to regulate the return preparers in question in that case. Here, the IRS has regularly issued guidance distinguishing political activity from that furthering social welfare or other exempt purposes. The NPRM seeks to amend and update that guidance rather than to wade into a previously untouched field.

⁶See, e.g., I.R.C. § 501(c)(3) (excluding from exemption organizations that "participate in, or intervene in (including the publishing or distributing of statements), any political campaign on behalf of (or in opposition to) any candidate for public office"); I.R.C. § 162(e) (prohibiting business deduction for expenses related to "participation in, or intervention in, any political campaign on behalf of (or in opposition to) any candidate for public office").

⁷ For a good short history of the 501(c)(3) lobbying rulemaking, see Judith E. Kindell & John F. Reilly, Lobbying Issues in Exempt Organizations Continuing Professional Education Technical Instruction Program for Fiscal Year 1997, 283 - 284 (1996).

⁸ We recognize that Treasury may not have authority to modify definitions in certain parts of the tax code without Congress amending the IRC; however, the phrasing in most sections is remarkably consistent (compare sections 501(c)(3), 4955(d)(1), 170(c)(2)(D), and 162(e)).

⁹Citizens United, 558 U.S. 310.

¹⁰ The NPRM has opened the question whether to alter the IRS's approach of permitting social welfare organization to engage in political activity if it is not their primary activity. The Bright Lines Project, as such, takes no position on that issue and does not address it in these comments. Any views of participants in the Bright Lines Project on the issue are presented in separate comments.

¹¹ The problem of defining political campaign activity applies to other tax-exempt categories such as 501(c)(6) trade associations, and also applies to 501(c)(3) charities with even more significance as they are prohibited from engaging in any political activity.

¹² The IRS did make efforts within the last decade to provide guidance to distinguish issue advocacy from candidate electioneering, with Revenue Rulings 2004-6 (regarding the 527(f) tax on 501(c) entities) and 2007-41 (regarding 501(c)(3) charities), but the open-ended multiple factors used to evaluate such cases has led to tens of millions spent on broadcast advertising about candidates, with no publicly-reported instances of tax law enforcement. The rulings differ from each other for reasons that have never been explained, and neither ruling addresses the kind of political activity that would count against the qualification of an organization for 501(c)(4) exempt status.

¹³ Internal Revenue Service, Rosenberg's Rules, meeting handout, available at http://www.taxanalysts.com/www/freefiles.nsf/Files/EO%20Measurement%20Term%20Definitions.pdf/$file/EO%20Measur ement%20Term%20Definitions.pdf.

¹⁴ Thomas Catan, Political Donors Chase Funding Alternatives, Wall Street Journal, Jan. 5, 2014.

¹⁵ Matea Gold Koch-backed political coalition, designed to shield donors, raised $400 million in 2012, Wash. Post, Jan. 5, 2014.

¹⁶ Treas. Reg. § 1.501(c)(3)-1(c)(3)(iii)

¹⁷ The NPRM's rules on transfers to other organizations, discussed in detail below, provide just one reason a 501(c)(4) may need to avoid engaging in any "candidate-related political activity," even if it is permitted to do so me of that activity without loss of exemption, because to do so would cut off its ability to receive grants or donations from other (c)(4) groups.

¹⁸ I.R.C. § 4945(f).

¹⁹See Gen. Couns. Mem. 34233 (Dec. 3, 1969).

²⁰See, for example, Taylor Lincoln, The New Stealth PACs: Tracking 501(c) Non-Profit Groups Active in Elections (Public Citizen, Sept. 2004), available at http://www.stealthpacs.org/documents/StealthPACs.pdf; and Public Citizen, Update: Stealth PACs (2010) available at http://www.citizen.org/Page.aspx?pid=4544.

²¹See J.E. Kindell & J.F. Reilly, Election Year Issues, Exempt Organizations Continuing Professional Education Technical Instruction Program for Fiscal Year 2002, 335-487 at 433.

²²See Public Citizen, Public Citizen Urges IRS to Investigate Tax Status of Freedom Partners, Sept. 23, 2013 available at http://www.citizen.org/pressroom/pressroomredirect.cfm?ID=3983.

²³ Treas. Reg. § 1.501(c)(4)-1(a)(2)(ii).

²⁴ Rev. Rul. 81-95, 1981-1 C.B. 332.

²⁵ This approach is consistent with other published IRS interpretations which, though not precedential, confirm the approach the IRS has taken over the years. The 2002 CPE text discusses the primary test for 501(c)(4)s and cites these authorities. See Kindell and Reilly, supra note 21 at 433. A series of letter rulings in the late 1990's includes express statements that the set of activities which are prohibited campaign intervention for 501(c)(3)s are the same activities that constitute non-social-welfare political activity which must not, under current rules, be the primary activity of a 501(c)(4). See, e.g., IRS Priv. Ltr. Rul. 9652026 (October 1, 1996) (prohibited 501(c)(3) campaign intervention is the same activity that does not promote social welfare and thus must be less than primary for a 501(c)(4)).

²⁶ And, incidentally, that has been recognized as not furthering the exempt purposes of 501(c)(5) or 501(c)(6) organizations. See Gen. Couns. Mem. 34233, supra note 20.

²⁷ The term ''political organization'' means a party, committee, association, fund, or other organization (whether or not incorporated) organized and operated primarily for the purpose of directly or indirectly accepting contributions or making expenditures, or both, for an exempt function. I.R.C. § 527(e)(1).

²⁸ I.R.C. § 527(e)(2).

²⁹ They are taxed on net investment income, I.R.C. § § 527(b)(1), (c)(1). "Exempt function income," such as proceeds of sales of campaign materials, is not taxed but only "to th e extent such amount is segregated for use only for the exempt function of the political organization." Id. § 527(c)(3).

³⁰ That IRC section imposes a tax on the greater of the 527 exempt function expenditures by a 501(c) organization and its net investment income. In effect, this tax levels the playing field by ensuring that no exempt organizations are able to use investment income tax-free for 527 activity.

³¹See IRS Announcement 88-114, 1988-37 I.R.B. 26.

³²See Rev. Rul. 2004-6 ("All the facts and circumstances must be considered to determine whether an expenditure for an advocacy communication relating to a public policy issue is for an exempt function under § 527(e)(2)") .

³³See IRS Priv. Ltr. Rul. 1999-25-051 (Mar. 29, 1999) (effort to support ballot measure is "exempt function" activity if organization has evidence to show that work would support or oppose a candidate for elective office).

³⁴ Note that the proposed regulations would treat any voter registration or GOTV campaign as "candidate-related political activity," while the section 527 regulations, if un changed, would continue to exempt much of this activity from tax. Likewise, express advocacy for or against a candidate's nomination to an appointed office would be deemed political under the NPRM while incurring no 527(f) tax if invited by a legislative body.

³⁵ That section defines indirect expenses as follows:

³⁶ Treas. Reg. § 1.527-6(b)(1)(i).

³⁷ We set aside for now the problem of determining whether a state statute is sufficiently "similar" to FECA to trigger application of this provision.

³⁸Citizens United, 558 U.S. at 336-67..

³⁹See, e.g., John Francis Reilly & Barbara A. Braig Allen, Political Campaign and Lobbying Activities of IRC 501(c)(4), (c)(5), and (c)(6) Organizations, Exempt Organizations Continuing Professional Education Technical Instruction Program for Fiscal Year 2003, L-1 - L-64 at L-10. available at http://www.irs.gov/pub/irs-tege/eotopicl03.pdf (last visited Jan. 21, 2014).

⁴⁰ Pub. L. 100-203, title X, § 10712(d), Dec. 22, 198 7, 101 Stat. 1330-468.

⁴¹ I.R.C. § 4955(d)(1).

⁴²Id. § (d)(2).

⁴³ Pub. L. 106-230, § § 1(a), 2 (a),July 1, 2000, 114 Stat. 477, 479; Pub. L. 107-276, § § 1(a), 2 (a), (b ), 5 (a), 6(a)-(c), (e)-(g), Nov. 2, 2002, 116 Stat. 1929, 1932-1934.)

⁴⁴ IRC § 527(c)(3).

⁴⁵ Treas. Reg. § 1.527-2(b)(1).

⁴⁶ We recognize that the gap between qualification for 501(c) exemption and for 527 exemption could be widened if the quantitative allowance for non-qualifying political activity were to be reduced below the "less-than-primary" level, without a corresponding reduction in the amount of political activity necessary for qualification under section 527. However, the Bright Lines Project takes no position on the question of whether, and to what level, the quantitative limit for 501(c) political activity ought to be re-set.

⁴⁷ We acknowledge that a broader definition that allows more organizations to qualify for section 527 status would be desirable. However, this could be achieved by broadening that section to allow for a range of advocacy, including both electoral and policy matters. The test for whether advocacy is considered "political" need not differ, although the outcome might be of little import to such a broadly defined 527 entity.

⁴⁸ We prefer using the term "political intervention" because it is shorter and because the word "intervention" appears in almost all the sections of the Internal Revenue Code that pertain to political candidate campaign activity. The unique character of the word thus lends itself to more precise line-drawing by alerting the reader that a special set of definitions derived from federal tax law is being advanced, rather than using words that are more common, generic, and likely to be misunderstood.

⁴⁹ 424 U.S. 1 at 43-44.

⁵⁰ 551 U.S. 449 at 451 (2007).

⁵¹ 2 U.S.C. § 434(f)(3)(A).

⁵² 2 U.S.C. § 441b(b)(2).

⁵³McConnell v. Federal Election Comm'n , 540 U.S. 93 (2003).

⁵⁴Wisconsin Right to Life, 551 U.S. at 451.

⁵⁵Citizens United, 558 U.S. at 310.

⁵⁶ California Government Code section 85310.

⁵⁷ 470 F.2d 849, 857 (10th Cir. 1972). See also, Regan v. Taxation With Representation of Washington, 461 U.S. 540 (1983) (upholding restrictions on lobbying by 501(c)(3) organizations).

⁵⁸ Kindell and Reilly, supra note 21 at 346-349.

⁵⁹See, e.g., Revenue Ruling 2004-6, situations 1, 2, and 5, Revenue Ruling 2007-41, situation 14.

⁶⁰See, e.g., Revenue Ruling 2004-6, situation 12 (mention of candidacy in alumni newsletter).

⁶¹See, e.g., Revenue Ruling 86-95 (candidate debates), Revenue Ruling 2007-41, situations 10 and 11 (non-candidate capacity appearances). The service club organizations Rotary, Kiwanis, and Lions are 501(c)(4) organizations. Under the NPRM, if a guest speaker or club officer happened to be a public official up for re-election or was otherwise a candidate, and was invited for any reason to be on the weekly luncheon program during a pre-election period, the club would be required to report the appearance as the club's political campaign activity on IRS Form 990, Schedule C.

⁶²See Citizens United, 558 U.S. at 325; Paul Ryan, A 'Reasonable Interpretation' of Wisconsin Right to Life in a New World of Uncertainty, Campaign Legal Center Blog (June 26, 2007) available at http://www.clcblog.org/blog_item-141.html (last visited February 26, 2014) (noting that "reasonable minds often disagree regarding what the 'reasonable interpretation' of an ad is").

⁶³ Gen. Couns. Mem. 39694 (Feb. 3, 1988)

⁶⁴ The NPRM does not include this "litmus test" manifestation of express advocacy, by which the organization mentions no candidates by name but expressly advocates that voters choose whom to support or oppose using criteria that clearly distinguish the candidates.

⁶⁵ Bipartisan Campaign Reform Act, title II(a), § 201, 2 U.S.C. § 434 (2002).

⁶⁶Id.

⁶⁷Id.

⁶⁸ Citizens United, 588 U.S. at 310.

⁶⁹ Craig Holman and Luke McLoughlin, Buying Time 2000 (Brennan Center for Justice, 2001).

⁷⁰ Jeremy Peters, "Blitz of Campaigns Ads Is Early and Aggressive," New York Times, May 16, 2012.

⁷¹ Kyle Trygstad, "The Most Expensive Senate Race of the Cycle -- So Far," Roll Call Dec. 15, 2013.

⁷² Arguably, meeting these two conditions is not the exclusive means of ensuring that the contributor has not engaged in candidate-related activity but rather creates a safe harbor, and the contributor that does not prohibit use of its donation for politics will not be treated as engaged in candidate-related activity when it contributes to an organization that does not engage in candidate-related activity.

⁷³ To be clear, even if a more narrow definition of candidate-related activity were developed, we propose that contributions by section 501(c)(4) organizations to another section 501(c) organization should not be treated as candidate-related activity of the grantor if the grantor takes reasonable steps to prevent such use.