Anonymous Negativity

While there is much doubt as to the effects of negative advertising, there is no doubt that political scientists love to study the topic. Some have concluded that negative advertisements have little to no effectFootnote ⁶ or that their effect is not substantially different from positive ads.Footnote ⁷ Still, others suggest negative ads work within certain contexts.Footnote ⁸ Psychologists have argued that negative information is better remembered than positive information,Footnote ⁹ which would support the findings of political advertising scholars who have concluded voters remember negative ads better.Footnote ¹⁰ There is evidence that negative ads generate unfavorable attitudes toward the object of an attack and more positive attitudes toward the sponsor, if the voters perceive the ad to be truthful.Footnote ¹¹ There is also far from unanimous agreement that negative ads are entirely bad. Geer's (Reference Geer2006) well-cited study notes that the most common complaint of negative ads is that they are misleading; however, it is not clear that negative ads are any more misleading than positive ones.

Regardless of whether negative ads work, it is obvious that campaign practitioners believe they are effective, and they remain a fixture of the American electoral system. Negative ads are as popular as ever.Footnote ¹² Going negative, however, can create negative consequences for both the target and the sponsor of an ad. Negative advertisements can create a backlash against candidates who sponsor them.Footnote ¹³ Yet, voters are less likely to hold a candidate accountable for an attack run on his or her behalf by an outside interest group.Footnote ¹⁴ This may help explain why there has been a dramatic increase in attack ads by outside groups, which are technically independent of candidates and the parties.Footnote ¹⁵ Candidates often quickly distance themselves from attack ads by outside groups that benefit themselves, which can minimize or even eliminate the potential for backlash.Footnote ¹⁶

The increase in outside group attack ads has been even more intense following the Supreme Court's Citizens United v. Federal Election Commission (2010) ruling, allowing outside groups to spend “non-hard money” (funds not regulated by the FEC) on direct advocacy in favor or opposition of candidates for office.Footnote ¹⁷ Ironically, rules that technically prohibit candidates from coordinating with outside groups post-Citizens United can provide candidates with cover from being held responsible for attack ads. When Mitt Romney was asked to call for the removal of misleading attack ads purchased by a pro-Romney super PAC in 2012, for example, he overstated the potential for punishment for “coordinating” with an outside group. “My goodness,” he claimed, “if we coordinate in any…we go to the big house.”Footnote ¹⁸

Candidates would prefer to distance themselves from negative attack ads, but would the sponsors of negative advertisements ever seek to dissociate themselves with the negative messages they helped to create? To put it another way, are individuals and organizations more likely to fund and produce negative advertisements if the public never knows who was behind the attack?

Outside of politics, it is well established that the potential for negative attacks is increased when one's identity is concealed. Anonymity increases the likelihood to post rude, nasty, or critical comments on newspaper discussion boards.Footnote ¹⁹ Additionally, anonymity is a major, and typically necessary, condition for racist comments advocating violence in Internet chatrooms.Footnote ²⁰

Within the realm of politics, there are a number of examples of where wealthy funders sought to keep their support for outside group attacks anonymous. Senator John McCain was attacked by a front group dubbed “Republicans for Clean Air” during the 2000 primaries, for example. Only after the primary was decided did the 527 organization's backers, brothers Sam and Charles Wyly, energy company billionaire Bush supporters, acknowledge funding the ads.Footnote ²¹

Business and Anonymous Politicking

When it comes to influencing politics, no sector has sought to avoid being associated with political activities more than for-profit businesses.Footnote ²² For-profit corporations have always maintained an uneasy relationship with political candidates. Scholars have long known that, generally speaking, businesses prefer to keep their efforts to exert political influence private,Footnote ²³ but traditional electoral activities, such as making PAC contributions, are highly visible.

Some correctly foretold of the rise in anonymous political campaign spending by for-profit corporations post Citizens United. Torres-Spelliscy (Reference Torres-Spelliscy2011, 61) noted, “History shows that for-profit corporations spend through non-profits to enjoy their anonymity while spending without accountability from shareholders or customers.” Although a lack of disclosure rules prevents knowing precisely how much for-profit corporations have contributed to outside groups since Citizens United, there is evidence that corporate campaign intervention is increasing, mostly to the benefit of one side of the ideological spectrum. For example, a reporting mistake on the part of the Republican Governors Public Policy Committee, a 501(c)(4) nonprofit, disclosed that large corporations such as Coca-Cola, Exxon Mobil, Microsoft, Pfizer, and Walmart were funding the group's efforts to elect Republican governors.Footnote ²⁴

There is even evidence from the state level of businesses attempting to avoid disclosure of their donations by funneling money through various nonprofits before it is eventually spent for its intended purpose. In Utah, for example, Republican attorney general candidate John Swallow used a network of political committees and “vaguely named nonprofit organizations” in multiple states to secretly raise money from payday loan businesses. With the help of a consultant, he successfully directed more than $150,000 from payday loan businesses and executives to outside groups. Outside ads helped Swallow defeat two primary candidates, one of whom supported efforts to regulate the payday loan industry.Footnote ²⁵ Although Swallow was elected, he is now awaiting trial for more than a dozen charges related to his efforts to cover up the fundraising scheme.Footnote ²⁶

The Swallow saga is hardly the only state scandal involving businesses and dark-money. During a couple of California 2012 ballot elections, Republican activists running a nonprofit called Small Business Action sought to avoid state donor disclosure rules by first funneling donations through separate out-of-state nonprofits. California campaign officials uncovered an elaborate network of businesses, conservative nonprofits, and large-money individual donors, led by billionaires Charles and David Koch.Footnote ²⁷ Republican activists admitted that the money funneling idea was hatched as means of protecting business donors from public retribution.Footnote ²⁸

These recent examples of for-profit corporations and business executives trying to influence elections anonymously should come as little surprise. Business executives have long understood the potential benefit of contributing to a candidate's electoral efforts,Footnote ²⁹ but they also recognize the potential for backlash if the public perceives they are behaving too politically.Footnote ³⁰ Evidence suggests that corporations do curtail their electoral activities, when they are the targets of protests or large-scale public criticism.Footnote ³¹ Corporate political spending may also create tension between corporate managers and shareholders.Footnote ³² Several corporations have faced backlash for their involvement in campaigns post Citizens United. In 2010 Target and Best Buy came under intense criticism from gay rights activists for contributions to a state political committee called MN Forward, which supported a socially conservative candidate for governor. Minnesota's state disclosure rules, passed immediately after Citizens United, required MN Forward to disclose the names of its donors, causing Target and Best Buy customers and shareholders to criticize the groups, with some staging protests in Target stores.Footnote ³³

Following the MN Forward incident, few publicly traded corporations have been willing to make contributions directly to super PACs and state political committees, although the Citizens United ruling allows them to do so. Instead, politically active corporations prefer to contribute to 501(c) nonprofits that are not required to disclose donor information.Footnote ³⁴ This practice of spending on elections (by purchasing independent ads mentioning candidates) through organizations that do not disclose their donors has been dubbed “dark money.”Footnote ³⁵ According to the Center for Responsive Politics (CRP), dark-money nonprofits have spent more than $500 million since Citizens United, not counting the millions to be spent in the 2016 election.Footnote ³⁶

Additionally, it also has become increasingly common for some political nonprofits to serve primarily as a conduit for campaign contributions by raising money anonymously and then contributing the money to super PACs or other political nonprofits, which then use the money for elections.Footnote ³⁷ The rise of dark money and funneling contributions through nonprofits to other groups, a practice labeled “campaign money laundering,”Footnote ³⁸ has fueled calls for new transparency rules on election spending. The most important transparency proposal at the federal level has been the DISCLOSE Act, which, among other things, requires nonprofits that spend on elections to disclose the names of large donors.Footnote ³⁹ In fact, the version of the bill passed by the House in 2010 would require a group to list the names of top donors in television and radio ads.Footnote ⁴⁰

Determining if dark-money groups are more likely to engage in negative advertising, is important for a variety of reasons. As dark-money spending becomes more common, it is likely that anonymously funded attack ads will become more common as well. This is especially true if groups are more likely to run negative ads when they are permitted to conceal their donors, as this study reveals. Many of these dark-money groups take on misleading names, such as Main Street Advocacy, that mask the group's and donors' true agenda. A better understanding of how these groups operate and specifically determining if anonymity incentivizes these groups to engage in attack politics, improves our understanding of the possible consequences of disclosure rules such as the DISCLOSE Act.

Data and Methods

This study uses data on expenditures in congressional campaigns from the 2010, 2012, and 2014 election cycles. These expenditures constitute nearly all outside spending after the Citizens United ruling.Footnote ⁴¹ Data for outside spending was collected from the Center for Responsive Politics (CRP).Footnote ⁴² The CRP notes whether the expenditure was “in support,” “in opposition,” or “support and opposition” of a candidate. The last category is sometimes called a “contrast ad.”Footnote ⁴³ This study uses two dependent variables. For the binary logistic regression, a dummy variable was created indicating whether the purchase was for purely an opposition ad (not including the latter contrast ads). A final model uses multinomial logit with a three-category dependent variable predicting both negative and contrast ads, using positive expenditures as the reference category.Footnote ⁴⁴ This tells us whether dark-money is also predictive of contrast ads, in addition to purely opposition ones. The CRP also reports whether the group making the expenditure fully discloses its donors. Expenditures made from groups that provided no disclosure of donor information make up the dark-money independent variable.Footnote ⁴⁵

Several other key variables are also included. One such variable is ideology of the sponsoring group. Conservative groups are significantly more likely to engage in anonymous political spending.Footnote ⁴⁶ This finding is certainly supported by observations of the press and open-government groups, who have consistently found dark-money groups to be, more often than not, conservative.Footnote ⁴⁷ There has been surprisingly little research into whether one side of the ideological spectrum is any more likely to use negative advertising. In recent elections, Republicans have slightly more often engaged in negative advertising,Footnote ⁴⁸ and in one study of campaign consultants, Republican consultants were slightly more likely to support going negative in close races.Footnote ⁴⁹ But this hardly counts as proof of an ideological imbalance when it comes to negative advertising. Furthermore, this question has yet to be applied to conservative vs. liberal groups, as opposed to negative ads purchased by the candidates or parties. To correct this dearth of research on the role of ideology and negative advertising, dummy variables noting whether the group purchasing the expenditure is conservative or liberal are included.Footnote ⁵⁰

Numerous scholars and news organizations have remarked on the increasing amounts of money spent on outside expenditures, especially in the wake of Citizens United. While few have attempted to determine whether groups are spending more on negative advertising than ads in favor of candidates, some recent research suggest that groups do spend more on negative ads.Footnote ⁵¹ When it comes to ballot measures, there seems to be parody in pro and anti-measure spending by interest groups.Footnote ⁵² However, it seems reasonable to assume that groups will spend more on negative campaigns than positive ones, especially considering that negative ads are more often a fixture in competitive races.Footnote ⁵³ Therefore, the cost of the expenditure expressed in $1,000 increments is an independent variable.

Some scholars have observed that outside-group spending has been dominated by a relatively small number of groups who have spent in only a hand-full of competitive congressional races;Footnote ⁵⁴ thus, a dummy variable, labeled “10 or more expenditures,” was included to determine whether the most politically active groups are more likely to engage in negative ads.

U.S. Senate races clearly attract more outside group attention and obviously tend to be more expensive than U.S. House races.Footnote ⁵⁵ Therefore, a dummy variable noting whether the expenditure was made for a Senate race is used.

Finally, as mentioned earlier, Citizens United has resulted in two new campaign-spending vehicles for outside groups: super PACs and 501(c) nonprofits. Super PACs are technically titled independent expenditure only committees and have only existed since 2010. Technically, they are required to disclose their donors,Footnote ⁵⁶ although, as mentioned earlier, groups sometimes avoid disclosing the original sources of funds by funneling contributions through nonprofits before they are spent by a super PAC.Footnote ⁵⁷ In fact, I recently demonstrated that organizations that operate both super PACs and 501(c)s are far less likely to disclose their donor information (Chand Reference Chand2015). Thus, it is important to note that a number of super PAC expenditures in this study were made by groups that CRP identifies as providing no donor disclosure.

Political 501(c) nonprofits have existed for years, but Citizens United allowed such groups to engage in “express advocacy,” ads that explicitly advocate the “election or defeat of a clearly identified candidate for federal office.”Footnote ⁵⁸ Unlike super PACs, they are not required to disclose donor information. The analysis in the following section indicates whether groups use super PACs or 501(c)s to purchase their outside ads, with traditional PACs and “soft 527s” serving as the reference categories.Footnote ⁵⁹

The following section presents analysis of outside spending in the last three election cycles, revealing that there is indeed a relationship between going negative and anonymous spending.

Analyzing Dark and Dirty Ads

Analysis for this study includes 13,274 expenditures made during the 2010 through the 2014 elections, which constitutes 99 percent of all expenditures made over the three cycles.Footnote ⁶⁰ The vast majority of expenditures during the most recent election cycles were positive and purchased by groups that disclosed their donors, although it is worth noting that a small minority of positive and fully disclosed expenditures were probably not advertisements.Footnote ⁶¹ Positive expenditures made up 67.8, 72.9, and 66.8 percent of total expenditures in 2010, 2012, and 2014, respectively. Negative ads made up 12, 12.2, and 15.9 percent for each of the respective cycles. The data also confirm prior studies, which have observed most outside spending is generated by a small number of highly active organizations.Footnote ⁶² Groups that made ten or more expenditures during individual election cycles accounted for the majority of expenditures in each cycle. Such groups made up 78.7 (2010), 76.2 (2012), and 63.7 (2014) percent of expenditures. Percentages are presented in Table 1.

Table 1 also shows the sum total spent for each type of expenditure. While most individual expenditures by outside groups were positive, positive expenditures constituted a minority of the overall amount spent in each cycle. These positive expenditures were only 16.7 percent in 2010, 11.4 percent in 2012, and 13.2 percent in 2014. Negative ads made up 18.2 percent of spending in 2010, 30.5 percent in 2012, and 20.5 percent in 2014. The majority of spending in each cycle went to contrast ads: 65.1 percent in 2010, 58 percent in 2012, and 66.2 percent in 2014. Based on these descriptive statistics, it seems obvious that groups do spend more on negative and contrast advertisements than on positive advertisements, as observed in recent studies.Footnote ⁶³

Clearly, negative advertising and dark money are two separate phenomena, but are they related? The binary logistic regression models in Table 2 strongly indicate that there is a relationship between negative advertising and the level of disclosure by the group purchasing the ad. Three models are presented testing this assumption, one for each cycle. In all three cycles, dark money is a clear predictor of negative advertising. In 2010, the dark-money odds ratio shows that dark-money expenditures were nearly one and a half times more likely to be for negative ads, than for positive or contrast ads (both of which serve as the reference). In 2012, they were more than three times more likely to be negative, and in 2014 they were nearly two times (1.99) more likely.

Results for the other independent variables are both surprising and largely expected. In all three cycles, liberal and conservative groups were more likely to run negative ads than groups that supported both Democrats and Republicans. Liberal expenditures ranged from three to more than three and a half times more likely to be for negative ads than positive or contrast ads. Conservative expenditures ranged from being more than one and a half times (1.76) in 2012 to more than nine times more likely to be for negative ads in 2010. Senate expenditures were only significant in the 2014 elections, with Senate expenditures being one and a half times more likely to go to negative ads. Unsurprisingly, the cost of the expenditure was a significant predictor for two of the election cycles, both the 2010 and 2012.

Three of the variables did produce somewhat surprising results. First, it is clear that outside groups are far more likely to use super PACs than 501(c) nonprofits when engaging in negative advertising. Expenditures from 501(c) nonprofits were not predictive of negative ads. In fact, in the 2012 elections, 501(c) expenditures were actually 44 percent less likely (odds ratio .56) to go to negative ads. Because the dark-money variable is predictive of negative ads, it would seem, at first glance, counterintuitive that super PAC expenditures would be predictive of negative ads and 501(c) expenditures not be, considering the former are technically required to disclose their donors and the latter are not. However, as mentioned earlier, one must keep in mind that super PACs raise money from organizations—not just individuals. Thus, it has become very common for super PACs to raise and spend money that was previously given to one, or in some cases many, other nonprofit organizations for the purposes of concealing the original donor's identity.Footnote ⁶⁴ This is precisely why CRP identifies some super PAC expenditures as being made by dark-money organizations.

The other somewhat surprising variable is the dummy indicating whether the expenditure came from a group that made ten or more expenditures in the election cycle. These highly active groups were no more likely to engage in negative advertising. In fact, in the 2010 and 2012 election cycles, expenditures for these groups were negative predictors of negative ads.

The models in Table 2 clearly show groups are more likely to engage in negative advertising when they do not disclose any donor information. But they provide no clue as to whether one side of the political spectrum is more likely to use dark money in negative ads. The models in Table 3, which use interaction terms indicating whether the expenditures were made by liberal or conservative dark-money groups, can provide this insight. These results suggest that there is indeed an ideological difference in the use of dark-money funded attack ads, and the results are as one would expect. In all three of the liberal models, while the liberal main effect remained a significant predictor of negative ads, the liberal dark-money interaction terms were not positive, significant predictors of negative advertising. In fact, in the 2012 and 2014 election cycles, liberal dark-money expenditures were actually significant, negative predictors of negative advertising. Liberal dark-money expenditures were 67 percent less like to go toward negative ads in 2012 and 48 percent less likely in 2014.Footnote ⁶⁵

Conservative dark money was a clearly significant predictor of negative advertising in two of the three election cycles. During the 2012 cycle, conservative dark-money expenditures were four and half times (4.8) more likely to be used for negative ads; in the 2014 cycle, they were two and a half times (2.57) more likely to be used for negative ads. Interestingly, the conservative main effect coefficient was negative, although insignificant, when including the conservative dark-money interaction term in 2012. These findings make sense given prior research on dark money and the political activities of organized business. We know that conservative organizations have been far more likely to use dark money in recent elections;Footnote ⁶⁶ if we accept the premise of research on politically active businesses, we would presume that the business community has the greatest incentive to conceal its involvement with negative politics.Footnote ⁶⁷ The conservative dark-money coefficient and odds ratio are both large in the 2010 model, but are not significant due to the fairly large associated standard error.

What if we take contrast ads into the account? The binary models just presented have predicted negative ads, using both positive and contrast expenditures as the reference categories for the dependent variable. However, does dark money also predict the use of contrast ads? To answer this, while still predicting the effects of dark money on negative ads, one needs to run multinomial logistic regression using both contrast ads and negative ads as the predicted outputs and positive ads as the reference (base) category. Table 4 presents three such models (one for each category) using the same independent variables as in the binary models.

Interestingly, in all three of the multinomial models, the effect of dark money on the likelihood of contrast ads is negative, and in one model, the 2010 election, the coefficient is actually significant. In 2010, dark-money expenditures were 21 percent less likely to go toward contrast ads than to positive ones. This is made clearer when we now look at the effects of dark money on negative ads. With contrast ads no longer part of the reference category, dark-money expenditures are still more likely to be spent on negative ads. However, the effects of dark money are slightly smaller and only significant for two of the three election cycles. In 2012, dark-money expenditures were nearly three times more likely to go toward negative ads than to positive ones. In 2014, the odds were more than one and a half times. Both of these effects are large and significant, but the coefficients are smaller than in the binary models presented in Table 2. Also, notably, the effects of dark money are not a significant predictor of negative ads in the 2010 multinomial model, although the coefficient and odds ratio are still fairly large. Clearly, based on these multinomial models, outside groups prefer dark money to go toward negative advertising. However, dark money is not preferred for contrast ads. In fact, given that all of the dark-money coefficients are negative for contrast ads, groups may actually prefer to use dark money for positive ads (the reference category) more than they do for contrast ads.

The rest of the independent variables behaved mostly the same as in the binary models, with a few new insights provided. The 501(c) variable, while not a predictor of negative advertising, is a good predictor of contrast ads. The 501(c) coefficient is significant for all three elections. The effect is quite large in the 2010 model, where 501(c) expenditures were nearly four times more likely to go to contrast ads, as opposed to positive ones. The 501(c) coefficient is also significant for negative ads in 2010. However, this was also the case in the binary model and the 501(c) coefficients for negative ads in the other two models are not significant. The Senate variable also produced some surprising results. The Senate coefficients, while only modestly large, were significant positive predictors of contrast ads in all three election cycles. The largest effect was in 2012, when Senate expenditures were 1.86 times more likely to go toward contrast ads. As with the binary models, the Senate expenditures were only a significant predictor of negative ads in the 2014 election.

The super PAC variable is significant at predicting both types of ads for all three models. The liberal and conservative coefficients are also significant for both types of ads. Liberal and conservative groups are more likely to use negative ads and contrast ads than non-ideological groups. Predictably, the amount spent is a significant predictor for both types of ads. The Senate coefficients for both ads in all election cycles were significant, except for negative ads in 2010. Coefficients for the most active groups (the ten or more groups), reveal little new information. The coefficients were significant inverse predictors of negative ads for both the 2010 and 2012 models. The multiple expenditure coefficient was close to being a significant predictor of negative ads in 2014.Footnote ⁶⁸

Discussions of Disclosure

Negativity in elections is certainly nothing new. In the early days of the republic, candidates hurled both professional and personal attacks against their opponents in newspapers. It should come as little shock that the first race to use television advertising (the presidential campaign of 1952) also contained the first television attack ad.Footnote ⁶⁹ However, anonymous political attacks ads have become a more frequent issue in modern elections.

This study shows that negative and anonymous advertising are related, despite being separate issues. Anonymously funded ads were significantly more likely to be opposition ads than ads by groups that disclosed their donors. Furthermore, while anonymous ads overall are significantly more likely to be negative, these anonymously funded attack ads are a decidedly conservative electioneering tool. While neither side has cornered the market on negative advertising, this study shows that conservative dark-money expenditures are significant predictors of attack ads. Liberal dark-money purchases are not. This finding, in addition to previous research showing that conservative groups are far more likely to use dark money, clearly indicates that conservative organizations are placing a greater emphasis on concealing their donors when they criticize candidates.

The multinomial models in the prior section also strongly indicate that groups prefer using dark money for negative ads, but not for contrast ads. Clearly, if groups are concerned about backlash from negative advertising, that concern does not extend to the negativity within contrast ads.

As with all studies, this one has its limitations. For one, it makes no distinction between the types of advertisements purchased with the expenditures (e.g., television, radio, direct mail). It also examines outside expenditures only in congressional races and in only three election cycles. Still, the data used for this study does constitute nearly all outside spending in congressional races by outside groups since the Citizens United ruling.

On a related point, it should be noted that dark money, as defined within this study, was a relatively new concept on election day in 2010. The Supreme Court issued Citizens United on 21 January 2010, less than ten months before election day. Furthermore, super PACs did not develop until after a D.C. Circuit Court of Appeals decision in the case SpeechNow.org v. FECFootnote ⁷⁰ on 26 March. Thus, the full of effects of Citizens United were in place for less than half of the 2010 election cycle. This may, partly, explain why the dark-money coefficient in the 2010 multinomial model, while large, is not quite significant for predicting negative ads. This may also explain why the overall amount of variance explained in the 2010 multinomial model (16.7 percent) is less than half of the other two multinomial models. Further, it could also clarify why the conservative dark-money interaction term was not significant in 2010 but was clearly a predictor of negative ads in the last two election cycles. Those points aside, it is important to note that dark money was a significant predictor of negative advertising for the 2010 election in the first binary model.

The biggest limitation, however, is obviously the speculation that conservative dark-money groups are funded by for-profit businesses. Because dark-money groups do not disclose their donors, it is clearly difficult to prove that such groups receive funding by for-profit business interests seeking to avoid a public backlash. It is equally possible that liberal groups are less likely to use dark money in attack ads because liberal groups and Democratic candidates have publicly endorsed greater transparency.

Still, there is ample evidence to show that businesses typically seek, when possible, to conceal their political activities.Footnote ⁷¹ Business associations also overwhelmingly support Republicans. Furthermore, there have been numerous incidents of businesses supporting dark-money conservative groups, which have been exposed by state investigators,Footnote ⁷² open-government groups,Footnote ⁷³ and simple errors on the part of dark-money groups when recording their finances.Footnote ⁷⁴ Taking all of these facts together, it seems safe to assume for-profit businesses are at least partially responsible for the increasing use of dark-money negative advertising.

There are some additional conclusions to be drawn from the analysis presented here. First, groups clearly spend more criticizing candidates they oppose rather than praising those they support. Spending on negative and contrast ads far exceeds expenditures groups make in support of candidates. Second, outside groups have eagerly embraced the use of super PACs and 501(c)s post Citizens United; however, groups appear to be using these campaign vehicles differently. Super PACs are popular tools for both negative and contrast expenditures. Groups appear to be using 501(c)s more strategically. These nonprofits are not any more popular as tools for negative advertising than traditional PACs, but they were strong predictors of contrast ads.

Since the Citizens United ruling, the primary goal of campaign finance advocates has been to increase transparency for outside groups. Advocates have been unsuccessful at the federal level. While the DISCLOSE Act passed the Democratic-controlled U.S. House of Representatives in 2010, a filibuster in the Senate prevented the bill from becoming law.Footnote ⁷⁵ These transparency activists suffered additional setbacks with passage of the 2016 budget. In the appropriations act, Republican members of Congress successfully barred the Securities and Exchange Commission and Internal Revenue Service from creating rules requiring publicly traded corporations and political nonprofits that spend on elections from disclosing their donor information.Footnote ⁷⁶

Still, disclosure rules rest on strong constitutional grounds. Unlike the prohibitions on outside expenditures contained in the Bipartisan Campaign Reform Act (BCRA) of 2002, which were first modified in FEC v. Wisconsin Right to Life Committee (2007) and ultimately struck down in Citizens United, disclosure regulations have generally performed well during legal challenges.Footnote ⁷⁷ Challenges to Minnesota's disclosure rules, after the Target and Best Buy controversy, were upheld.Footnote ⁷⁸ Even a clear majority in Citizens United upheld disclosure requirements, stating that “[t]he First Amendment protects political speech; and disclosure permits citizens and shareholders to react to the speech of corporate entities in a proper way.”Footnote ⁷⁹

Furthermore, the benefits of disclosure have been well established.Footnote ⁸⁰ In many races, there is little information about the candidates and few cues on which voters can base their decisions. The knowledge that a sector of business interests is running ads on behalf of a candidate is useful information to voters.Footnote ⁸¹ In fact, this has been empirically demonstrated with donor disclosure information on ballot initiatives,Footnote ⁸² which is a major reason why all states that use ballot initiatives require interest groups to disclose their advertising expenditures.Footnote ⁸³

Knowing that donor information and who is responsible for ads will be disclosed may also create greater incentive for groups to be honest. In fact, a rationale behind disclosure requirements in federal lobbying laws such as the Lobbying Disclosure Act (LDA) of 1995 is to make sure lobbyists accurately represent clients and present information.Footnote ⁸⁴ As some scholars have noted, requiring lobbyists to disclose their clients encourages them to behave truthfully.Footnote ⁸⁵ The disclosure requirements in lobbying regulations have also passed constitutional challenges in the courts. In February 2008, the National Association of Manufacturers (NAM) challenged provisions in the Honest Leadership and Open Government Act of 2007 (which strengthened the LDA). Under the law, NAM was required to disclose the names of member organizations that funded and actively participated in the group's congressional lobbying campaigns. The case was dismissed in district court, and in 2009, the D.C. Circuit Court of Appeals affirmed the constitutionality of lobbying disclosure rules.Footnote ⁸⁶

Advocates also push disclosure rules for the purpose of curbing corruption. Preventing corruption, and even the appearance of it, is yet another reason the courts have endorsed disclosure rules. In Buckley v. Valeo (1976), the Supreme Court stated that “disclosure requirements deter actual corruption and avoid the appearance of corruption by exposing large contributions and expenditures to the light of publicity.”Footnote ⁸⁷

One may question whether disclosure rules would discourage small-money donors from contributing to outside groups, given the findings of this study.Footnote ⁸⁸ In fact, there is some evidence to suggest that disclosure rules for contributions directly to candidates may deter some individuals from contributing.Footnote ⁸⁹ However, the disclosure rules for outside groups discussed in this paper are highly unlikely to impact the decisions of small-money individual contributors. Rules such as the DISCLOSE Act only apply to contributions of more than $10,000 to outside groups that spend on elections. Furthermore, donors can still make large contributions to such groups anonymously by specifying that their donations should not be used for campaign purposes.Footnote ⁹⁰ Moreover, disclosure rules will not discourage all publicly traded companies from donating. Many corporations in the United Kingdom continue to donate to parities and leaders after the U.K. adopted disclosure rules in the 1980s.Footnote ⁹¹

Republicans in Congress have prevented the passage of stronger transparency rules; however, there are some reasons to expect changes in transparency laws. Transparency supporters have enjoyed some progress passing stronger disclosure rules at the state level.Footnote ⁹² Additionally, disclosure requirements enjoy bi-partisan support from voters. More than 80 percent of Democratic voters—and more than 80 percent of Republican voters—say they support donor disclosure requirements for groups that attempt to influence elections.Footnote ⁹³ In fact, disclosure requirements were even popular with Republican members of Congress until Citizens United. After the Republicans for Clean Air debacle, Congress, led by Senator McCain, passed disclosure rules for soft 527 organizations, then known as “stealth PACs.”Footnote ⁹⁴ Later, Congress bi-partisanly passed even stronger disclosure requirements for 527s, which required groups to report the names, addresses, occupations, and even employers of all persons to the IRS that contributed more than $200 to the group,Footnote ⁹⁵ which is the functional equivalent to reporting requirements for PACs.

Whether stronger disclosure rules should be adopted is a question beyond the ability of this study to answer. However, this study does suggest that requiring greater disclosure will reduce the incentive to go negative, at least for some groups.