An Absence of Malice

Abstract

We examine whether repeated scandals within one party generate collective sanctions for fellow partisans. Do voters punish a party’s candidates because of multiple corruption scandals? Our data come from a unique survey conducted prior to the 2010 legislative elections in North Carolina, a state that had recently seen a number of high-profile corruption scandals involving Democrats exclusively. Although Republicans campaigned energetically against “the party of corruption,” we find the impact of that campaign was muted. Respondents who accurately identified at least one scandal rated the Democratic Party less favorably and thought Republicans would do better at responding to corruption. Nevertheless, vote choice was unrelated to knowledge of corruption scandals, and Republicans did not benefit from any effects on voter turnout. Importantly, respondents’ partisanship only sometimes mediated attitudes and did not affect behavior. We conclude that voters might in theory prefer “clean” parties, but their political actions are uninfluenced by that preference, a finding that has unfortunate implications for democracy.

Keywords

voting public opinion corruption parties state politics

Periodic elections are the principal mechanism for ensuring that citizens in democracies are able to hold their representatives accountable. Elections are expected to constrain politicians’ behavior because officials can be thrown out of office for failing to promote voters’ interests (Downs, 1957; Mayhew, 1974). Although research on electoral accountability tends to focus on “issue congruence,” typically measured as how well legislators’ roll call votes match their constituents’ preferences (Kingdon, 1973; Nyhan, McGhee, Sides, Masket, & Greene, 2012), ethical behavior is another important dimension (Stokes, 1963). Just as it is presumed the act of voting fosters greater substantive representation of the public’s policy interests, so the power of the franchise should also discourage elected officials’ malfeasance.¹

Indeed, scandalous politicians behave as if they anticipate censure from voters. Decades of research finds legislators caught in scandals are more likely to retire voluntarily than those who are not (Brown, 2001; Groseclose & Krehbiel, 1994; Jacobson & Dimock, 1994). Moreover, lawmakers who seek reelection after a scandal lose a substantial percentage of their expected vote share and are more likely to be defeated than their unblemished peers (Brown, 2006; Herrick, 2003; Hirano & Snyder, 2012; Peters & Welch, 1980; Praino, Stockemer, & Moscardelli, 2013; Welch & Hibbing, 1997). Basinger (2012) has calculated the cost at about 5 percentage points in U.S. House elections since the early 1970s.

Electoral sanctions have certainly been applied as a response to unscrupulous actions of individuals, but several questions remain unanswered. First, it is largely unknown whether or not cumulative scandals can damage party reputations and cost partisan colleagues votes. Almost every study to date examines how scandals affect the fate of directly involved individual elected officials, although some research indicates that parties might be affected too (Best, Ladewig, & Wong, 2013; Cobb & Taylor, 2014; Slomczynski & Shabad, 2011). Second, it is unclear which kinds of scandals matter most. All appear to harm offenders, but researchers debate whether corruption or other types have greater negative effects on perpetrators and their associates. Third, we do not really know if citizens’ political affiliations mediate their reactions to partisan scandals. Prior studies are mixed on whether voters “turn a blind eye” to their own party’s misbehavior (Anduiza, Gallego, & Muñoz, 2013; Blais et al., 2010; Cobb & Taylor, 2014). Finally, the mechanism driving electoral punishment is unsettled. Most scholars assert scandals encourage better quality candidates to challenge tainted incumbents (Abramowitz, 1991; Groseclose & Krehbiel, 1994; Lazarus, 2008). Another possibility is that voters intentionally punish unethical behavior. Unfortunately, as Hendry, Jackson, and Mondak (2008) lament, survey data needed to test this hypothesis rarely exist.

Our study addresses all of these matters, but it focuses particularly on the potential for corruption scandals to generate collective partisan accountability due to voter backlash. The issue has noteworthy implications. If, for example, citizens do not use their votes to penalize a party when several of its members have been caught trading their influence for personal gain, then “clean” parties would be advised to use other campaign strategies to gain power. From a normative perspective, it would be troubling to discover that voters are uninterested in punishing parties associated with greater corruption because the possibility of wholesale retribution is what provides an incentive for politicians to police each other’s actions. Absent public censure, effective institutional mechanisms to discourage corruption are less likely to develop (Lederman, Loayza, & Soares, 2005).

Our exploration proceeds as follows. First, we define scandals and review the theoretical and empirical importance of different types, the possibility that scandals generate collective sanctions, and the role voters play in discouraging them. To generate expectations, we construct a model that combines the literature on scandals with research on party reputations and issue voting. We then describe the case on which we test our hypotheses, the North Carolina state legislative elections of 2010, in which one party’s political corruption was covered by media and became a central part of the campaign’s rhetoric, even though the individuals accused of corruption were not on the ballot. After describing method, data, and the analysis, we discuss the implications of our findings and offer concluding remarks. Our results are clear: Most citizens are unable to cite those culpable in prominent corruption scandals, and voters who do recognize that only one party’s members have been caught in corruption fail to hold that party accountable. Party reputations are marginally affected, but there is no detectable electoral punishment.

Political Scandals

All scandals emerge as allegations or revelations about previously unknown behavior that violates social or legal norms. More concretely, Thompson (2000) writes that scandals are characterized by (a) attempts to hide the act from public view, (b) nonparticipants’ disapproval, (c) public censure of the behavior, and (d) the probability that disclosure will harm the reputations of the individuals involved. We add that scandals are political if they are about behaviors that have the potential to damage the reputation of a public institution, government official, party, or candidate for office. We also propose a fifth criterion for defining scandals: sustained media coverage. Simply put, scandalous behavior cannot matter politically if the media fail to report it (Costas-Pérez, Solé-Ollé, & Sorribas-Navarro, 2012; Nyhan, 2015; Shea, 1999).

Although the typology is admittedly crude, most scholars analyze scandals by coding them dichotomously in reference to the type of behavior exhibited. The literature ordinarily distinguishes between corruption scandals and all others that are not about corruption (but see Doherty, Dowling, & Miller, 2011). A corruption scandal occurs when a public official is accused of using the power of her office to obtain personal financial benefits, such as gifts, cash, endorsements, and campaign contributions. Some scholars define corruption as actual convictions (Meier & Holbrook, 1992), but this narrow classification misses the important role of accusation in the life cycle of a scandal, so illegality may be a characteristic but is not required. Clean though they might seem to people in other parts of the world, contemporary American politics are hardly devoid of corruption. At the federal level, members of Congress like Duke Cunningham (R-CA), Tom DeLay (R-TX), William Jefferson (D-LA), Bob Ney (R-OH), and James Traficant (D-OH) have gone to prison over the past decade or so. It is hardly surprising, then, that the public perceives widespread political corruption in this country.²

Scholars use a catchall category to describe all “non-corruption scandals.” Sometimes these are described as “moral,” because the behaviors violate intuitively different ethical standards. Publicized and recent examples of this type include former Rep. Mark Foley’s (R-FL) page scandal, Sen. Larry Craig’s (R-ID) airport bathroom incident, and Rep. Anthony Weiner’s (D-NY) “sexting.” Although these types of scandals are sometimes also about illegal behavior, lawfulness or lack thereof is not their unifying feature.

Theoretically, corruption scandals should invoke the heaviest sanctions. An extensive literature suggests they have broad negative repercussions for the polity. Anderson and Tverdova (2003) conclude, based on a survey of 16 democracies, that citizens are less trusting of public officials in countries with more corruption. Likewise, trust in governmental institutions declines when perceptions of widespread political corruption take hold (Bowler & Karp, 2004; Caillier, 2010; Dancey, 2012; Morris & Klesner, 2010). An additional consequence of this dynamic is that voter turnout is lower in countries where corruption is more prevalent (Stockemer, LaMontagne, & Scruggs, 2013) and the public is more skeptical of government’s ability to solve problems (Caillier, 2010).

It is true the two types of scandal are sometimes difficult to distinguish because moral scandals frequently involve abuses of office (Cobb & Taylor, 2014; Doherty et al., 2011). However, these episodes are expected to be less consequential than corruption because they are viewed as intrinsically personal and primarily affect the public’s evaluations of politicians as people, not assessments of their job performance and professional lives (Carlson, Ganiel, & Hyde, 2000; Funk, 1996; Miller, 1999). Corruption is also more likely to affect politicians who are at most only tangentially connected to the scandal because matters such as bribery and graft are frequently viewed as systemic, even when committed by one person, and elected officials are believed to be especially prone to malfeasance.³ The public is therefore likely to cast a broader net in an effort to impose some accountability after a corruption scandal.

The Causal Model

A number of political scientists have argued American political parties exist largely as producers of reputations that extend automatically to all members of that party (Aldrich, 1995; Cox & McCubbins, 1993; Grynaviski, 2010; Snyder & Ting, 2002, 2003). In related work, others have detected certain issues that the Democrats or Republicans “own” (Petrocik, 1996; Petrocik, Benoit, & Hansen, 2003). Issue ownership does not actually convey to the public a particular position the party has, but rather its enhanced capacity to “fix” any problems that exist in that policy area (Egan, 2013). Because reputations are not excludable, all of a party’s candidates benefit from its ownership of an issue.

The implication is that corruption scandals can shape party reputations. Studies show public views of the parties change over time (Pope & Woon, 2009), so even if a party has not traditionally had a strong reputation on corruption it can acquire one through its behavior and campaign communication (Hayes, 2008; Petrocik et al., 2003). Party brands might be affected if, for example, one party publicizes the other’s scandals assiduously. This is how all members of a party can bear collective accountability for the sins of individual members, even if none of the scandalous members are up for reelection. If voters perceive a party to be responsible for corruption, they can respond by voting for candidates of the other party. A huge literature already clearly demonstrates candidates can be punished by voters for the policy preferences and voting behavior of partisan colleagues (Egan, 2013; Petrocik, 1996; Petrocik, Benoit, & Hansen, 2003; Sides, 2007; Snyder & Ting, 2002).

Of course, for corruption to matter in elections voters must (a) know about the issue, (b) consider it to be salient, and (c) perceive important differences between the two parties about how they will handle the issue.⁴ These requirements are dependent on media coverage and party behavior. If the media do not report the behavior, voters cannot easily know about it. Likewise, if parties don’t campaign on the issue, voters will not only know less about it, but they will also find it less salient and fail to detect any partisan differences.

The literature therefore generates a multistage model of voters’ thinking and actions when presented with qualitatively and/or quantitatively meaningful cases of political scandal. First, citizens will be aware of scandals. Second, they will consider scandals to be salient election issues. Third, particularly if the cases exclusively or at least largely involve members of one party, voters will perceive material differences between the parties on the issue. These differences ought to have demonstrable effects on people’s attitudes about the parties and how they cast their votes in the following election.

Recent empirical analyses offer support for this basic model. At the presidential level, Bill Clinton’s sex scandal is thought to have cost Al Gore 3 to 4 percentage points in the popular vote—and therefore the presidency—in 2000 (Fiorina, Abrams, & Pope, 2003). Ulbig and Martorano’s (2012) study of malfeasance among state-level Republicans in Ohio indicated that Republicans running for federal offices in the state were disadvantaged. At the congressional level, Dimock and Jacobson (1995) found a strong relationship between a respondent’s views about the 1991 check-kiting scandal and how he or she voted in 1992. More directly, Best et al. (2013) and Cobb and Taylor (2014) argue that Democrats in the 2006 congressional elections benefited from multiple Republican scandals, while Slomczynski and Shabad (2011) found that voters in Polish elections punished political parties they associated with greater amounts of corruption.

Our focus here is on the last stage of the model. More specifically, we are interested in testing two hypotheses. The first, about attitudes, is as follows:

Hypothesis 1: Individuals are less likely to approve of the party they associate with corruption scandals.

The second is focused on behavior.

Hypothesis 2: Individuals are less likely to vote for candidates of the party they associate with corruption scandals.

We also investigate important partisan corollaries. Given the increasing and considerable polarization of the contemporary electorate (Abramowitz & Saunders, 2008; Layman & Carsey, 2002), partisanship might condition attitudinal or behavioral effects of knowing about corruption scandals. Interestingly, past research on this aspect of scandals generates conflicting expectations. On one hand, Anduiza et al. (2013) find that citizens identifying with the same party as a scandalous member “turn a blind eye” to their corruption, which helps explain the puzzling ability of most incumbents to survive their scandals. One reason this might occur is that partisans discount the bad news because it is incongruent with their preferences (Gerber & Green, 1999). We might therefore expect citizens sharing partisanship with scandalous members to know and care less about their corruption scandals compared with independents and those identifying with competing parties.

On the other hand, partisanship may be so strong that it might not be possible in today’s polarized environment for scandals to affect vote choice except among independents, those most able to change their behavior in response to them. Rather than vote choice, scandals might increase turnout among both independents and members of the nonoffending party, and possibly reduce it for those who identify with the scandalous party, a finding compatible with descriptions of “throwing the bums out.”

Finally, some evidence exists that partisans will sanction their own candidates. Cobb and Taylor (2014) found that Republican voters (and independents), but not Democrats, most negatively responded to the Mark Foley (R) scandal in 2006. In that case, partisans administered electoral punishment to their own team.

A Test Case: The 2010 North Carolina Legislative Elections

The 2010 state legislative elections in North Carolina provided an opportunity to test these expectations. North Carolina’s government was traditionally thought of as “clean” (Christensen, 2008; Luebke, 1998), and it did not have the sordid and colorful history of other southern states such as Louisiana.⁵ Perceptions began to change, however, when Democratic Commissioner of Agriculture, Meg Scott Phipps, resigned from her post in 2003 under an ethics cloud.⁶ Bigger scandals more germane to our study soon followed.

In 2005 and 2006, Democratic Speaker of the House, Jim Black, was swimming in controversy. Three incidents drew considerable media attention. After Black pushed through legislation setting up a state lottery, a close professional relationship he enjoyed with a likely vendor was investigated. Black effectively bribed a Republican representative, Michael Decker, to vote for him as speaker after the public gave the Republicans a 61-59 majority in the 2002 elections. Finally, Black was accused of taking bribes from chiropractors while the state association had legislation pending before the General Assembly. For this, he was convicted and spent 4 years in prison.

Around the same time, Democratic Governor Mike Easley was also caught up in an array of scandals. Media reports of two “sweetheart” deals in which he secured property for considerably less than the market rate furthered the impression that his was a corrupt administration.⁷ Later, just after the 2010 midterms, Easley pleaded guilty to a violation of campaign finance law for accepting free flights from business associates. As a result, he became the first governor of the state to be convicted of a felony.

Ironically, the issue of free flights from campaign supporters immediately plagued Easley’s Democratic successor, Beverly Perdue. In August 2010, right before the election, the State Board of Elections fined her campaign US$30,000 for failing to report 42 flights as contributions during her 2008 run for the office.⁸

Finally, although it involved activities that were not known at the time to involve corruption, former U.S. Senator and Democratic vice presidential candidate John Edwards was caught in a sex scandal following his 2008 bid for the party’s presidential nomination. Edwards had an extramarital affair with a campaign worker, Rielle Hunter, and fathered a child with her. The episode was sordid and attracted great public attention, not least because Edwards’s wife, Elizabeth, was suffering from breast cancer at the time.⁹

The media coverage of Democrats’ corruption was considerable and extended over a period of time. We searched for articles about North Carolina corruption scandals in three of the largest state newspapers using the “News Bank: America’s News” database.¹⁰ More specifically, we looked for stories about the issue involving Black, Easley, or Perdue between January 1, 2005, and October 5, 2010. These dates coincide with the clear emergence of Black’s first scandal and the first day our election survey was in the field. Using just their names and two keywords, “scandal” and “corruption,” we located 61 newspaper stories about at least one of these individuals.¹¹ Of these stories, 39 were focused primarily on Easley, 18 were about Black, and 4 covered Perdue.

There is also evidence that the state Republican Party sought to make corruption a central theme in its campaign for the General Assembly that fall—there were no contests for executive office in North Carolina in 2010 (Johnson & Niolet, 2010). As early as summer 2009, Republican delegates to the state convention were observing out loud that, “there is official after official on the Democratic side who is going to jail” (Christensen & Kiley, 2009, A3). Party chair Tom Fetzer was often in the news explaining, “One ofthe things Republicans are going to be running against is corruption” (Binker, 2010, p. 1) and asking North Carolinians to “review the Democrats’ record on ethics.”¹² Five other articles from 2010 that appeared in our data focused on these Republican efforts.¹³ Indeed, in 2010 stories about corruption, especially Easley’s, attracted considerable media attention and the issue was viewed as a problem for Democrats.¹⁴

Some Republicans were dubious of this strategy. House Minority Leader Paul Stam understood the Democrats had provided his party an opportunity but argued, “The scandals only can stir the pot a little bit.” Adding, “I don’t think people ultimately vote based on scandals unless the person you are running against was in the scandal,” he noted Republicans had made Speaker Black’s conduct a central part of their 2006 campaign (Christensen & Kiley, 2009, A3). That year, Republicans lost seats and Black held onto his by just 30 votes in a solidly Democratic district, implying that the strategy only resonated with the speaker’s constituents.

All told, the 2010 state legislative election presented North Carolina voters the information necessary to deliberately punish all candidates of one party due to their colleagues’ corruption. There were numerous important and widely publicized political corruption scandals, and there was little confusion as to which party was tainted by them.¹⁵ What is more, there were few up-ballot races to distract the public’s attention from the state Republican Party’s basic message in trying to retake the General Assembly. In 2010, there were no presidential, gubernatorial, or Council of State races, and the U.S. Senate contest between Richard Burr and Elaine Marshall, it is fair to say, generated very little public interest.

To be sure, other issues, like the recession and the Affordable Care Act for example, received greater media and voter attention than corruption. Nevertheless, given the partisan asymmetry of recent scandals, amount of media coverage, and willingness of the opposition to exploit the issue, we would be unlikely to find a similar situation where voters were presented with such a clear opportunity to hold just one party’s candidates accountable for their colleagues’ corruption.

Data and Design

To analyze the effects of these scandals on North Carolina voters, we conducted a public opinion survey before the November 2010 elections. GfK (“Knowledge Networks”) administered it over the Internet to a random sample of 655 adult residents of North Carolina between October 18 and October 30. The survey has a margin of sampling error of ±3.6%, a completion rate of 70%, and a response rate of 29% using American Association for Public Opinion Research (AAPOR) Standard Definition No. 3 (AAPOR, 2009).

Variables

We are interested in both voters’ attitudes and actions, which we measure by constructing four dependent variables. Respondents were asked (a) how favorably they viewed the two state political parties and (b) which state party they felt would do a better job at fighting corruption. These two questions capture whether Democrats’ corruption influenced how the parties were evaluated. Respondents were also asked (c) how likely they were to vote or if they had already voted.¹⁶ Finally, we asked them (d) which party’s candidate they had voted for or were planning on voting for in the upcoming state House election.¹⁷ Given the nature of the statewide random sample, it was not practical to ask respondents about voting for specific House candidates in their district. Republican votes were coded “1” in a dichotomous variable.¹⁸

Some obvious approaches to measuring people’s reaction to political scandals are likely to exaggerate their importance. Asking respondents if a scandal or the issue of corruption influences their vote, for example, assumes voters are knowledgeable about the issue. That kind of question wording also encourages affirmative answers as a result of social desirability biases.¹⁹ Alternatively, we could have asked subjects merely which party they thought was most strongly associated with North Carolina’s corruption scandals. This strategy, however, would not separate genuine knowledge from guesses. Answers would also likely be shaped by a respondent’s partisanship.²⁰

Our analytic strategy is different. We do not assume citizens have prior knowledge about scandals, nor do we ask respondents to report their judgments about them. Rather, we start from the simple premise that scandals can only directly affect attitudes and behavior if voters have accurate information about them. It is worth repeating that none of the scandalous politicians were actually on the ballot in 2010. Thus, collective sanctions for other members of the Democratic Party depend on voters knowing that each corruption scandal in the recent time frame involved a Democrat. If a respondent correctly named former Governor Mike Easley as scandalous but erroneously believed he was a Republican, for example, it would not be possible to vote against other Democrats as punishment for Easley’s behavior.

More specifically, we asked respondents, “Can you name a North Carolina politician caught in a recent political scandal?” If a respondent answered affirmatively, he or she was then asked to enter the name of that politician and to identify that politician’s partisan affiliation. Next, respondents who attempted to identify a scandalous politician were asked if they were aware of another in the state. As before, if respondents answered affirmatively they were instructed to enter the name and partisanship of the politician. For respondents who identified two scandals, we repeated this process a third time.

We coded respondents’ answers so as to construct four independent variables. The first is dichotomous, and is coded “1” if the respondent correctly named an official caught up in a corruption scandal and identified him or her as a Democrat. Here mentions of Black, Easley, or Perdue were considered recognition of corruption.²¹ All blank answers were coded as “0,” as too were names of politicians who were not involved in scandals or were from another state. A few state politicians were indeed involved in minor scandals, but because they received minimal press attention and were often not about corruption, these responses were also coded as “0.”²² A second independent variable denotes the number of corruption scandals identified correctly by the respondent. Values for this measure therefore range from 0 to 3.

Although John Edwards’s scandal was neither a state-level corruption scandal nor part of Republicans’ rhetoric about Democrats’ corruption, respondents cited it so frequently that we created a separate dichotomous variable for it. This third variable expands on the first to include mentions of Edwards as a Democrat to be an accurate identification of a scandal. Finally, a fourth independent variable separates the Edwards scandal from others about corruption and codes respondents naming Edwards as a Democrat as their only scandal “1” and all others “0.”

Following past studies, we utilize several control variables found to explain political attitudes and voting behavior. These demographic variables are age (seven categories from 18-24 to 75 and over), gender (male or female), race (White, not White), education (four categories from “less than high school” to “bachelor’s degree or higher”), and household income (19 categories from “less than $5,000” to “$175,000 or more”). We also employ measures of party identification (using the three categories of “Republican,” “independent or other,” and “Democrat”) and respondents’ approval of President Obama’s job performance (coded as “strongly disapprove,” “disapprove,” “approve,” or “strongly approve”).²³

Results

We start by presenting data about respondents’ awareness of scandals. Table 1 shows the proportion of respondents who correctly identified individuals involved in scandals and their party affiliation. These figures reveal that about 45% of the sample could accurately name a politician involved in any scandal and describe him or her as a Democrat. Republicans were more likely to identify accurately scandalous politicians and their party although not by a great deal more than could Democrats and independents.²⁴ Still, just 29% of all subjects could name an official involved in a corruption scandal while accurately identifying him or her as a Democrat. Half of all respondents who named any scandal identified Edwards’s only. Much of this might be attributable to greater news coverage given his stature as a former senator, presidential candidate, and Democratic nominee for vice president. However, the finding also suggests the public is more exposed to and better retains information about sexual scandals than corruption.

Table 1.

Respondents’ Knowledge of Scandals.

Response type	All respondents	Republicans	Democrats	Independents
Any scandal	44.9	51.4	45.8	39.8
Any corruption scandal	28.6	35.3	23.1	29.0
Edwards scandal only	23.7	27.7	18.5	26.7

Note. Figures are percentages in the response category.

It is difficult to know whether these results suggest the public lacks meaningful knowledge about corruption scandals. To put the data into perspective, we examine results of the Elon University Poll of North Carolinians’ political knowledge conducted around the same time as our survey. The poll reported about one half of the respondents knew the Democrats controlled the state’s General Assembly and Nancy Pelosi was Speaker of the U.S. House of Representatives. Only about a third could name both of the state’s U.S. Senators.²⁵ By this benchmark, North Carolinians knew roughly as much about political scandals in their state in late 2010 as they did other kinds of basic political information even if the absolute level of knowledge was low.

Multivariate Analysis: Attitudes

Table 2 shows the results of an ordered logit model in which the dependent variables measure individuals’ favorability ratings of the state Democratic and Republican parties. Respondents’ ratings were recorded on a 5-point scale with answers ranging from very unfavorably (lowest) to very favorably (neutral was the midpoint). For the purposes of space, we show only the results where the respondent was able to name at least one corruption scandal. With the exception of the “Edwards only” measure (which never produced a statistically significant coefficient in specifications), these results are replicated regardless of the indicator of scandal knowledge utilized. The table reports results of attitudes with respondents split by their partisan attachments.

Table 2.

Knowledge of Scandals and the Favorability Ratings of North Carolina’s Parties.

	Views of Republican party				Views of Democratic party
	All	Republicans	Independents	Democrats	All	Republicans	Independents	Democrats
Knowledge of scandal	0.396 (0.181)*	0.251 (0.364)	0.755 (0.296)*	0.245 (0.320)	−0.539 (0.181)*	−0.955 (0.350)*	−1.03 (0.294)*	0.338 (0.332)
Party identification	−1.13 (0.136)*				1.06 (0.134)*
Obama approval	−0.729 (0.086)*	−0.942 (0.284)*	−0.748 (0.119)*	−0.737 (0.146)*	0.974 (0.093)*	0.827 (0.291)*	1.00 (0.133)*	0.943 (0.159)*
Age	−0.071 (0.053)	−0.241 (0.122)*	−0.104 (0.085)	−0.070 (0.090)	0.048 (0.053)	0.111 (0.107)	0.000 (0.087)	0.055 (0.097)
Gender	0.163 (0.161)	0.058 (0.345)	−0.246 (0.254)	0.550 (0.288)*	0.141 (0.163)	−0.131 (0.318)	0.439 (0.258)*	−0.169 (0.308)
Race	−0.236 (0.213)	1.05 (0.945)	−0.084 (0.361)	−0.275 (0.288)	0.061 (0.217)	−0.079 (0.799)	0.738 (0.360)*	−0.217 (0.322)
Education	0.020 (0.100)	−0.121 (0.221)	−0.059 (0.154)	0.047 (0.172)	0.069 (0.101)	0.012 (0.200)	0.473 (0.159)*	−0.265 (0.188)
Income	−0.023 (0.022)	−0.147 (0.047)*	0.18 (0.036)	−0.018 (0.036)	−0.007 (0.021)	0.097 (0.043)*	−0.031 (0.036)	−0.065 (0.038)*
Cut 1	−6.93 (0.619)	−6.79 (1.64)	−4.50 (0.869)	−3.06 (1.03)	2.16 (0.565)	1.30 (1.38)	2.18 (0.858)	−3.53 (1.17)
Likelihood ratio χ²	360.74*	25.31*	58.62*	31.45*	449.59*	22.82*	95.50*	57.70*
Pseudo R²	.195	.074	.090	.055	.237	.057	.146	.113
n	645	170	252	223	646	171	252	223

Note. Method is ordered logit. Dependent variable is respondents’ views of the state political parties on a 5-point scale from very unfavorably to very favorably. “Knowledge of scandal” independent variable is whether or not respondent correctly identified at least one corruption scandal.

p < .05.

Knowledge of scandal generates more favorable views of Republicans and less favorable views of Democrats. It is clear, however, that the negative effect is greater on attitudes toward Democrats than the positive effect is for Republicans (at about 2.8, 3, and 1.1, the mean, median, and standard deviation of total responses are practically identical for both parties). Moreover, the results are different across subjects. Knowledge of scandal makes Republicans and independents more likely to view Democrats unfavorably. Democrats’ views of their party are unchanged. It is only independents’ attitudes toward Republicans that are altered in a favorable direction when the respondent is cognizant of a scandal. As theories of partisanship suggest, knowledge of scandal shifts the disposition of independents materially more than it does partisans.

To illustrate this substantively, the model predicts that with all other variables held at their means, 29.5% of White female Republicans who could not identify a corruption scandal had a neutral or some type of favorable attitude toward the Democrats, only 14.8% of those who could name a scandal held the same views. For independents with identical characteristics, these figures were 83.2% and 67.2%, respectively. The model predicts that of the same White female independents, 65.8% who could not identify a scandal and 78.8% who could had a neutral or favorable view of the Republican Party.

Table 3 presents results of a multinomial logit in which subjects were asked to identify the state party they felt was best at fighting corruption. This measure is trichotomous because it included a “no difference” choice as the midpoint of the scale. Again, it is clear that those who knew of scandals and correctly attributed them to Democrats were likely to view the Republicans as better able to combat corruption. We should note, however, that the modal reply was actually “neither.” A clear majority, 63%, expressed this belief. Only 21% who named at least one corruption scandal said Republicans would do a better job.

Table 3.

Knowledge of Scandals and the Party Thought Best at Fighting Corruption (N = 643).

	Named any corruption scandal	Number of corruption scandals named	Named any scandal	Named Edwards only
“Democrats” vs. other answers
Knowledge of scandal	−0.328 (0.344)	−0.445 (0.274)	−0.201 (0.278)	0.265 (0.298)
Party identification	1.50 (0.306)*	1.50 (0.306)*	1.52 (0.307)*	1.49 (0.306)*
Obama approval	0.711 (0.161)*	0.718 (0.162)*	0.708 (0.161)*	0.702 (0.161)*
Age	−0.002 (0.094)	0.005 (0.094)	−0.009 (0.093)	−0.012 (0.093)
Gender	0.252 (0.293)	0.230 (0.293)	0.262 (0.294)	0.303 (0.291)
Race	−.479 (0.289)*	−0.460 (0.290)	−0.483 (0.289)*	−0.479 (0.289)*
Education	−0.101 (0.169)	−0.098 (0.168)	−0.107 (0.168)	−0.113 (0.168)
Income	−0.043 (0.038)	−0.040 (0.037)	−0.044 (0.038)	−0.053 (0.037)
Constant	−6.60 (1.13)*	−6.63 (1.13)*	−6.57 (1.13)*	−6.58 (1.13)*
“Republicans” vs. other answers
Knowledge of scandal	0.745 (0.257)*	0.354 (0.152)*	0.304 (0.238)	−0.133 (0.267)
Party identification	−0.945 (0.196)*	−0.944 (0.196)*	−0.911 (0.192)*	−0.918 (0.193)*
Obama approval	−0.837 (0.167)*	−0.832 (0.167)*	−0.854 (0.166)*	−0.868 (0.167)*
Age	−0.022 (0.081)	−0.011 (0.080)	0.012 (0.079)	0.031 (0.078)
Gender	0.078 (0.243)	0.056 (0.243)	−0.025 (0.238)	−0.092 (0.233)
Race	−0.415 (0.426)	−0.387 (0.425)	−0.352 (0.423)	−0.334 (0.422)
Education	−0.027 (0.149)	−0.006 (0.148)	0.021 (0.148)	0.036 (0.147)
Income	−0.047 (0.032)	−0.045 (0.032)	−0.043 (0.032)	−0.034 (0.032)
constant	2.66 (0.942)*	2.61 (0.938)*	2.50 (0.932)*	2.58 (0.937)*
Likelihood ratio χ²	331.91*	330.84*	324.47*	323.88*
Pseudo R²	.281	.280	.274	.273

Note. Method is multinomial logit. Dependent variable is “Democrats,” “No Difference,” “Republicans.”.

p < .05.

The results of unreported specifications in which Democrats, Republicans, and independents are analyzed separately tell a similar story. Knowledge of corruption scandals made no difference for Democrats but, as with Table 2, it made both Republicans and independents more supportive of the Republican Party. In this case, independents who could identify a scandal were more likely to think Republicans were better at fighting corruption; Republicans who could were discernibly less likely to believe there was no difference between the parties on the issue.

Multivariate Analysis: Turnout and Vote Choice

Knowledge of corruption scandals does affect the views people have about the culprit’s party. Collective sanctions exist when it comes to attitudes although these are mediated somewhat by partisanship. Two other dependent variables, turnout and vote choice, allow us to investigate actual behavior. In Table 4, we present the results of a logit model of the decision of whether or not to vote made by Democrats and Republicans (unreported results of specifications containing only independents are not materially different). Overall, the findings about the determinants of turnout increase confidence in the validity of the effects of scandal we discern. These models, that is, perform in a manner we should expect if applied to any group of Americans. Consistent with the literature, older and more affluent respondents were more likely to vote.

Table 4.

Knowledge of Scandals and the Decision to Vote.

	Republicans				Democrats
	Named any corruption scandal	Number of corruption scandals named	Named any scandal	Named Edwards only	Named any corruption scandal	Number of corruption scandals named	Named any scandal	Named Edwards only
Knowledge of scandal	2.15 (0.818)*	1.90 (0.772)*	0.997 (0.494)*	0.085 (0.497)	1.70 (0.533)*	1.33 (0.462)*	0.838 (0.332)*	0.032 (0.354)
Obama approval	−0.999 (0.376)*	−0.991 (0.378)*	−0.972 (0.349)*	−1.02 (0.339)*	0.263 (0.170)	0.255 (0.170)	0.269 (0.167)	0.268 (0.164)
Age	0.331 (0.163)*	0.334 (0.163)*	0.393 (0.160)*	0.448 (0.159)*	0.410 (0.120)*	0.405 (0.119)*	0.421 (0.119)*	0.427 (0.118)*
Gender	−0.420 (0.537)	−0.388 (0.536)	−0.681 (0.522)	−0.912 (0.505)*	0.728 (0.370)*	0.692 (0.367)*	0.631 (0.360)*	0.508 (0.349)
Race	0.831 (1.08)	0.774 (1.07)	0.758 (1.06)	0.586 (1.02)	−0.368 (0.374)	−0.371 (0.373)	−0.301 (0.367)	−0.279 (0.365)
Education	−0.098 (0.287)	−0.110 (0.286)	−0.106 (0.287)	−0.015 (0.284)	0.314 (0.197)	0.324 (0.199)	0.327 (0.198)*	0.321 (0.198)
Income	−0.041 (0.062)	−0.038 (0.062)	−0.036 (0.062)	−0.014 (0.062)	0.076 (0.044)*	0.078 (0.044)*	0.080 (0.044)*	0.098 (0.044)*
constant	1.69 (2.00)	1.67 (2.00)	1.89 (1.98)	2.11 (1.94)	−5.00 (1.32)*	−4.92 (1.31)*	−5.11 (1.32)*	−4.77 (1.30)*
Likelihood ratio χ²	40.93*	42.11*	35.10*	30.86*	43.44*	42.27*	36.92*	30.33*
Pseudo R²	.240	.247	.206	.181	.154	.150	.131	.108
Percent modal category	80.1	80.1	80.1	80.1	67.9	67.9	67.9	67.9
Percent Predicted Correctly	83.6	83.0	81.9	83.0	71.9	72.3	72.3	72.8
n	171	171	171	171	224	224	224	224

Note. Method is logit. Dependent variable is “definite voter” (coded “1”) or not.

p < .05.

The findings are also similar to those about people’s attitudes and accurate knowledge of corruption scandals has measurable effects. Here, individuals who knew about a scandal were more likely to vote. Knowledge of any corruption scandal increased the likelihood of a White female Republican with average additional attributes voting by 11.6 percentage points to 98.3%; for an identical Democrat the increase was 21.1 percentage points to 87.7%.

Still, there is good reason to believe knowledge of scandal is not really a determinant of the act of voting but rather a proxy for measuring something else that does, likely the extent to which the respondent is informed about politics. This is because, as the table shows, Democrats who knew about corruption scandals were more likely to vote than those with the same party affiliation who did not. Knowledge of scandals increased Republicans’ motivation to vote, but it did not squelch the desire of Democrats to do likewise.

When correlated with education, which is a reasonably good indicator of those who know more about politics, the ability to name at least one corruption scandal generates r = .21 ( p < .001).²⁶ The idea that the knowledge of scandal is a proxy for information is also consistent with the finding that those who could name only Edwards, presumably the least informed of respondents who gave a correct answer, are no more likely to vote than those who could not accurately name a scandal of any kind. In fact, given knowledge of a political event—here scandal—is considerably more robust in the model, we might suspect it provides a better indicator of a citizen more broadly informed about politics than his or her formal education.

Our skepticism about the linkage between knowledge of scandal and political behavior is strengthened by the results shown in Table 5. In this analysis, we fail to find any evidence that knowledge of scandals significantly influences vote choice in favor of Republicans. Collective sanctioning does not occur in voter behavior. Routinely important variables like party identification, demographic factors like race and gender, and attitudes about the president’s job performance explain vote choice. Moreover, knowledge of scandals was also insignificant in (unreported) models where we examined vote choice separately for Democrats, Republicans, and independents. The effect of scandals on vote choice is not conditioned by partisan attachments.

Table 5.

Knowledge of Scandals and the Decision of Which Party to Vote for (n = 534).

	Named any corruption scandal	Number of corruption scandals named	Named any scandal	Named Edwards only
Knowledge of scandal	0.354 (0.425)	0.078 (0.279)	0.492 (0.391)	0.345 (0.467)
Party identification	−2.30 (0.312)*	−2.30 (0.312)*	−2.36 (0.320)*	−2.33 (0.318)*
Obama approval	−1.83 (0.214)*	−1.84 (0.216)*	−1.85 (0.214)*	−1.86 (0.215)*
Age	−0.023 (0.134)	−0.010 (0.134)	−0.027 (0.133)	−0.002 (0.132)
Gender	−0.648 (0.406)	−0.684 (0.409)*	−0.627 (0.406)	−0.698 (0.399)*
Race	−0.797 (0.621)	−0.765 (0.622)	−0.810 (0.624)	−0.750 (0.624)
Education	−0.203 (0.261)	−0.178 (0.260)	−0.199 (0.261)	−0.163 (0.260)
Income	0.046 (0.050)	0.046 (0.051)	0.041 (0.051)	0.042 (0.051)
constant	10.59 (1.71)*	10.56 (1.71)*	10.66 (1.73)*	10.61 (1.72)*
Likelihood ratio χ²	532.89*	532.28*	533.80*	532.75*
Pseudo R²	.720	.719	.721	.720
Percent modal category	50.6	50.6	50.6	50.6
Percent predicted correctly	92.3	91.9	92.3	92.1

Note. Method is logit. Dependent variable is Republican (coded “1”) or Democrat (coded “0”).

p < .05.

This general point is corroborated by the results of Table 6. Here we show the effects of respondents’ views of the state political parties and their ability to handle issues relevant in the 2010 campaign, including corruption. The question used in Table 3—that is, which state party the survey participant felt was better on corruption—was applied to the economy and jobs, education, health care, the budget, same-sex marriage, and illegal immigration.²⁷ Inclusion of the party identification and Obama approval controls mitigates their effects somewhat. Regardless, and as we suggested may have been the case earlier, respondents’ views on issues other than corruption had demonstrably greater influence on their vote choice in the state House elections. Moreover, people who thought the Republican Party was superior on all issues other than corruption voted for Republican candidates. The evidence is consistent with the observation that the Democrats’ national reputation was more important to voters’ selections in 2010 state legislative races than the North Carolina party brand. As before, in unreported results scandal effects did not emerge when we analyzed the data separately for Democrats, Republicans, and independents.²⁸

Table 6.

Views of the Parties on Several Issues and the Decision of Which Party to Vote for.

	Issues only model	Full model
	Issues only model	Full model
Corruption	−0.457 (0.584)	−0.591 (0.647)
Economy and jobs	0.977 (0.391)*	0.557 (0.428)
Education (the issue)	1.08 (0.407)*	0.714 (0.439)
Health care	1.32 (0.360)*	0.656 (0.433)
Budget	1.07 (0.428)*	0.867 (0.474)*
Same-sex marriage	0.731 (0.357)*	0.629 (0.405)
Illegal immigration	1.49 (0.397)*	1.18 (0.434)*
Age	−0.104 (0.136)	−0.054 (0.157)
Gender	−0.441 (0.439)	−.455 (0.509)
Race	1.02 (0.586)*	−.659 (0.786)
Education (of subject)	−0.084 (0.285)	−0.074 (0.327)
Income	0.051 (.060)	0.024 (0.067)
Party identification		−1.46 (0.404)*
Obama approval		−0.954 (0.289)*
constant	−12.83 (2.15)*	−1.71 (2.98)
Likelihood ratio χ²	551.99*	580.31*
Pseudo R²	.759	.805
Percent modal category	50.5	50.6
Percent predicted correctly	93.3	96.4
n	525	520

Note. Method is logit. Measures of attitudes have higher values if Republicans are seen as better on the issue. Dependent variable is vote for the Republican (coded “1”) or vote for the Democrat (coded “0”).

p < .05.

Conclusion

Political scandals can seem like powerful events. They generate a circus-like atmosphere and invoke public opprobrium (Sabato, 1991). Their presumed importance is bolstered by the political costs they generally impose on the individuals embroiled in them. It is telling, however, that these costs are often limited as more than half of all scandalous politicians retain their jobs (Basinger, 2012).

Our results strike a similar tone. Taking advantage of an environment where citizens were encouraged to respond to partisan scandals, we find that political corruption moderately punctuates public consciousness and shapes some political attitudes. Respondents who knew about episodes of influence peddling and linked them to Democrats viewed the Democratic Party less favorably. In addition, respondents who knew about a scandal were also more likely to say the state Republican Party was better able to combat corruption.

Nevertheless, Democrats’ multiple scandals did not result in more significant collective partisan accountability. First, despite considerable media attention and a concerted effort by Republicans to bring Democrats’ corruption scandals to light, less than one third of North Carolinians could accurately recall any single one of them. Citizens cannot punish partisan corruption if they don’t pay attention to it. Second, the effect of awareness about corruption was marginal. Most respondents viewed the two parties as being equally capable (or incapable) of combating the problem, and party favorability ratings were modestly altered. Moreover, whereas knowledge of scandal tended to alter the reputation of the parties in the expected directions among Republicans and independents, Democrats’ views of both parties remained unmoved. Third, those who accurately identified a scandal were more likely to vote, but there was no party effect. This finding is inconsistent with studies that demonstrate turnout is higher in races with incumbents embroiled in scandals and interpret the increase to a “throw the bums out” mentality of their challengers’ partisan supporters (Praino et al., 2013). Rather, it suggests awareness of scandals is highly correlated with respondents’ level of interest in politics, a variable known to promote the act of voting (Larcinese, 2007; Lassen, 2005). Finally, and of greater importance, corruption had no demonstrable influence on respondents’ vote choice. Contrary to our expectation, those who could and could not accurately identify corruption scandals were as likely to vote Democratic as Republican. This null result is consistent with the literature describing the limits of campaign effects (Sides, 2007), and is partially anticipated by scholars who argue situations exist where voters do not seek to punish political malfeasance (Alford, Teeters, Ward, & Wilson, 1994; Fackler & Lin, 1995; Rundquist, Strom, & Peters, 1977).

It is also worth repeating that North Carolinians were more aware of the Edwards scandal than any single incident of corruption. This finding could indicate greater press interest in salacious events, citizens’ heightened interest in learning about sordid information, or the former vice presidential candidate’s visibility. Furthermore, the effects of knowing about Edwards’s scandal were fewer even than those of knowing about corruption. The likely reason is that this episode was originally understood as a sex scandal, and Edwards had not held political office since 2004. According to the literature, individuals view such moral scandals as innately personal and less relevant to the transgressor’s job performance (Doherty et al., 2011).²⁹ Partisan colleagues are unlikely to become collateral damage under these circumstances.

Overall, we would characterize these results as worrisome. In a state where every corruption scandal during a spate of them was associated with Democrats and Republicans campaigned vigorously on the issue, voters were in a position to punish the Democratic Party for its members’ behavior. We find little evidence, however, that they sought to do this. To be sure, Republicans performed exceptionally well in the 2010 North Carolina state legislative elections, winning a majority of seats in both bodies, but the causes of that success appear unrelated to Democrats’ scandals. If politicians are persuaded they might suffer electorally for their colleagues’ unethical and illegal behavior, they will be more likely to invest in institutional efforts to prevent it from happening in the first place. Without the possibility of collective punishment, party leaders are less motivated to discipline their membership.

As with all studies, ours has some limitations. First, we possibly underestimate how much North Carolinians know about corruption scandals by requiring them to generate the names and partisan affiliation of those involved spontaneously. Citizens, for example, do better at identifying their representatives correctly from a list of names. However, the primary alternative strategy risked introducing information about scandals to respondents who were otherwise ignorant of them. Our preference was to avoid this kind of Type II error, where “pseudo” scandal effects can emerge through the measurement process itself. Second, relying on cross-sectional data rather than a panel limits our ability to detect the impact of scandals on partisanship itself. It is possible that some respondents had already switched their partisan identification from Democrat to Republican, for example, as a result of the numerous scandals. Unfortunately, such panel data do not exist. Third, and similarly, we do not investigate the effects scandal can have on the ability of parties to recruit competitive candidates. These effects are considered endogenous to the basic causal model. Fourth, it is possible that collective sanctions were muted precisely because none of the tainted Democrats ran for reelection. If these members had been on the ballot, then their sins presumably would have attracted even greater attention and further tough questions of all Democrats, such as why they had not called on their colleague(s) to resign.³⁰ Last, our data come from a single election cycle in just one state. Nevertheless, the 2010 North Carolina state legislative elections were uniquely situated for us to study possible collective sanctions for corruption scandals; fewer events of lower profile are even less likely to generate any positive effects.

At most, then, we can conclude from past literature that individual politicians are likely to suffer some form of personal punishment for scandals involving corruption. Ordinary citizens, however, seem uninterested in holding a party electorally accountable in the aftermath of even widespread breaches of trust. The muted effects we find are partly, though not wholly, a function of limited awareness. As Winters and Weitz-Shapiro (2013) argue, perhaps voters are willing to punish political corruption if they are made aware of it. Yet, that study, like all experimental studies, provides information about scandals as part of its treatments. In more natural settings, this kind of information is less well known (Costas-Pérez et al., 2012; Vivyan, Wagner, & Tarlov, 2012).

But many people did know about corruption scandals and still failed to punish the Democrats. Future research should, therefore, more carefully investigate the reasons for inaction when collective accountability is a plausible response to multiple scandals. One obvious barrier to generating collective punishment is partisanship (Blais et al., 2010). Out-party voters are already predisposed to voting against members of the opposing party, so even multiple scandals offer little “value added” as a reason to vote against them. Conversely, in-party corruption is judged less harshly and to be tolerated. Beyond partisanship, other possible explanations include scandal fatigue, a dulling of reactions to an increasing volume of scandalous behaviors (Kumlin & Esaisson, 2012). Alternatively, perhaps collective sanctions are absent because voters disagree that they should exist, or because they view them as ineffective. In the latter case, much like Meirowitz and Tucker (2013) theorize about protest activity in an iterative context, we should be worried that voters consider both parties so equally and inherently unethical that voting for either of them on the basis of ending corruption is a waste of time.

Footnotes

Acknowledgements

We are grateful to Nick Hatley for data collection and to Brandon Nyhan, Cheryl Boudreau, D. Sunshine Hillygus, David Doherty, and Duke and Loyola University Chicago seminar participants for their helpful suggestions on an earlier draft of the article.

Authors’ Note

An earlier version of this article was presented at the 2011 meeting of the Midwest Political Science Association. Data are available from the authors on request.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: We are grateful for funding provided by the School of Public and International Affairs that made the survey possible.

Notes

Author Biographies

Michael D. Cobb is an associate professor of political science in the School of Public and International Affairs at North Carolina State University. His research primarily address information effects on political attitudes and behavior, including the influence of misinformation, deliberation, and framing on topics such as race, judicial legitimacy, trade policies, and the use of military force abroad.

Andrew J. Taylor is professor of political science in the School of Public and International Affairs at North Carolina State University. His research is focused largely on Congress, although he has also written about North Carolina politics and scandal.

References

Abramowitz

A. I.

(1991). Incumbency, campaign spending, and the decline of competition in the U.S. House elections. Journal of Politics, 53, 34-56.

Abramowitz

A. I.

Saunders

K. L.

(2008). Is polarization a myth? Journal of Politics, 70, 542-555.

Aldrich

(1995). Why parties? The origin and transformation of party politics in America. Chicago, IL: University of Chicago Press.

Alford

Teeters

Ward

D. S.

Wilson

R. K.

(1994). Overdraft: The political cost of congressional malfeasance. Journal of Politics, 56, 788-801.

American Association for Public Opinion Research. (2009). Standard definitions: Final dispositions of case codes and outcome rates for surveys (6th ed.). Deerfield, IL: Author.

Anderson

C. J.

Tverdova

Y. V.

(2003). Corruption, political allegiances, and attitudes toward government in contemporary democracies. American Journal of Political Science, 47, 91-109.

Anduiza

Gallego

Muñoz

(2013). Turning a blind eye: Experimental evidence of partisan bias in attitudes towards corruption. Comparative Political Studies, 46, 1664-1692.

Basinger

S. J.

(2012). Scandals and congressional elections in the post-Watergate era. Political Research Quarterly, 65, 290-302.

Berelson

B. R.

Lazarsfeld

P. F.

McPhee

W. N.

(1954). Voting: A study of opinion formation in a presidential campaign. Chicago, IL: University of Chicago Press.

10.

Best

Ladewig

J. W.

Wong

D. C.

(2013). Owning valence issues: The impact of a “culture of corruption” on the 2006 midterm elections. Congress & the Presidency, 40, 129-151.

11.

Binker

(2010, January 3). Politics in ’10 are likely to be unpredictable. Greensboro News & Record, p. 1.

12.

Blais

Gidengil

Fournier

Nevitte

Everitt

Kim

(2010). Political judgments, perceptions of facts, and partisan effects. Electoral Studies, 29, 1-2.

13.

Bowler

S. B.

Karp

J. A.

(2004). Politicians, scandals, and trust in government. Political Behavior, 26, 271-287.

14.

Boylan

R. T.

Long

C. X.

(2003). Measuring public corruption in the American states: A survey of state house reporters. State Politics & Policy Quarterly, 3, 420-438.

15.

Brown

L. M.

(2001). The character of Congress: Scandals in the U.S. House of Representatives, 1966-1996 (Doctoral dissertation). University of California, Los Angeles.

16.

Brown

L. M.

(2006). Revisiting the character of Congress: Scandals in the U.S. House of Representatives, 1966-2002. Journal of Political Marketing, 5, 149-172.

17.

Caillier

(2010). Citizen trust, political corruption, and voting behavior: Connecting the dots. Politics & Policy, 38, 1015-1035.

18.

Campbell

J. E.

(1986). Presidential coattails and midterm losses in state legislative elections. American Political Science Review, 80, 45-63.

19.

Carlson

Ganiel

Hyde

M. S.

(2000). Scandal and political candidate image. Southeastern Political Review, 28, 747-757.

20.

Christensen

(2008). The paradox of Tar Heel politics: The personalities, elections, and events that shaped modern North Carolina. Chapel Hill: University of North Carolina Press.

21.

Christensen

(2010, October 27). Art Pope: A one-man Republican equalizer. News & Observer, p. A1.

22.

Christensen

Kiley

(2009, June 13). GOP makes sport of Democrats’ reign. News & Observer, p. A3.

23.

Cobb

M. D.

Taylor

A. J.

(2014). Paging congressional Democrats: It was the immorality stupid. Political Science & Politics, 47, 351-356.

24.

Costas-Pérez

Solé-Ollé

Sorribas-Navarro

(2012). Corruption scandals, voter information, and accountability. European Journal of Political Economy, 28, 469-484.

25.

Cox

G. W.

McCubbins

M. D.

(1993). Legislative leviathan: Party government in the House. Berkeley: University of California Press.

26.

Dancey

(2012). The consequences of political cynicism: How cynicism shapes citizens’ reactions to political scandals. Political Behavior, 34, 411-423.

27.

Dimock

Jacobson

G. C.

(1995). Checks and choices: The House Bank scandal’s impact on voters in 1992. Journal of Politics, 57, 1143-1159.

28.

Doherty

Dowling

C. M.

Miller

(2011). Are financial or moral scandals worse? It depends. Political Science & Politics, 44, 749-757.

29.

Downs

(1957). An economic theory of democracy. New York, NY: Harper.

30.

Egan

P. J.

(2013). Partisan priorities: How issue ownership drives and distorts American politics. New York, NY: Cambridge University Press.

31.

Fackler

Lin

T.-M.

(1995). Political corruption and presidential elections, 1929-1992. Journal of Politics, 57, 971-993.

32.

Fiorina

Abrams

Pope

(2003). The 2000 U.S. presidential election: Can retrospective voting be saved? British Journal of Political Science, 33, 163-187.

33.

Funk

C. L.

(1996). The impact of scandal on candidate evaluations: An experimental test of the role of candidate traits. Political Behavior, 18, 1-24.

34.

Gerber

Green

(1999). Misperceptions about perceptual bias. Annual Review of Political Science, 2, 189-210.

35.

Groseclose

Krehbiel

(1994). Golden parachutes, rubber checks, and strategic retirements from the 102nd House. American Journal of Political Science, 38, 75-99.

36.

Grynaviski

(2010). Partisan bonds: Political reputations and legislative accountability. New York, NY: Cambridge University Press.

37.

Hayes

(2008). Party reputations, journalistic expectations: How issue ownership influences election news. Political Communication, 25, 377-400.

38.

Hendry

D. J.

Jackson

R. A.

Mondak

J. J.

(2008). Abramoff, email, and the mistreated mistress: Scandal and character in the 2006 elections. In Mondak

J. J.

Mitchell

D. G.

(Eds.), Fault lines: Why the Republicans lost Congress (pp. 84-110). New York, NY: Routledge.

39.

Herrick

(2003). Fashioning the more ethical representative: The impact of ethics reforms in the U.S. House of Representatives. Westport, CT: Praeger.

40.

Hirano

Snyder

J. M.

(2012). What happens to incumbents in scandals? Quarterly Journal of Political Science, 7, 447-456.

41.

Jacobson

G. C.

Dimock

(1994). Checking out: The effects of bank overdrafts on the 1992 House elections. American Journal of Political Science, 38, 601-624.

42.

Johnson

Niolet

(2010, April 6). Perdue rolls out ethics proposals. News & Observer, p. A1.

43.

Kingdon

J. W.

(1973). Congressmen’s voting decisions. Ann Arbor: University of Michigan Press.

44.

Kumlin

Esaisson

(2012). Scandal fatigue? Scandal elections and satisfaction with democracy in Western Europe, 1977-2007. British Journal of Political Science, 42, 263-282.

45.

Larcinese

(2007). Does political knowledge increase turnout? Evidence from the 1997 British general election. Public Choice, 131, 387-411.

46.

Lassen

D. D.

(2005). The effect of information on voter turnout: Evidence from a natural experiment. American Journal of Political Science, 49, 103-118.

47.

Layman

G. C.

Carsey

T. M.

(2002). Party polarization and “conflict extension” in the American electorate. American Journal of Political Science, 46, 786-802.

48.

Lazarus

(2008). Buying in: Testing the rational model of candidate entry. Journal of Politics, 70, 837-850.

49.

Lederman

Loayza

N. V.

Soares

R. R.

(2005). Accountability and corruption: Political institutions matter. Economics & Politics, 17, 1-35.

50.

Luebke

(1998). Tar Heel politics 2000. Chapel Hill: University of North Carolina Press.

51.

Mayhew

D. R.

(1974). Congress: The electoral connection. New Haven, CT: Yale University Press.

52.

Meier

K. J.

Holbrook

T. M.

(1992). “I seen my opportunities and I took’em”: Political corruption in the American states. Journal of Politics, 54, 135-155.

53.

Meirowitz

Tucker

J. A.

(2013). People power or a one-shot deal? A dynamic model of protest. American Journal of Political Science, 57, 478-490.

54.

Miller

A. H.

(1999). Sex, politics, and public opinion: What political scientists really learned from the Clinton-Lewinsky scandal. Political Science & Politics, 32, 721-729.

55.

Morris

S. D.

Klesner

J. L.

(2010). Corruption and trust: Theoretical considerations and evidence from Mexico. Comparative Political Studies, 43, 1258-1285.

56.

Nyhan

(2015). Scandal potential: How political context and news congestion affect the President’s vulnerability to media scandal. British Journal of Political Science, 45,435-466

57.

Nyhan

McGhee

Sides

Masket

Greene

(2012). One vote out of step? The effects of salient roll call votes in the 2010 election. American Politics Research, 40, 844-879.

58.

Peters

J. G.

Welch

(1980). The effects of charges of corruption on voting behavior in congressional elections. American Political Science Review, 74, 697-708.

59.

Petrocik

J. R.

(1996). Issue ownership in presidential elections, with a 1980 case study. American Journal of Political Science, 40, 825-850.

60.

Petrocik

J. R.

Benoit

W. L.

Hansen

G. J.

(2003). Issue ownership and presidential campaigning. Political Science Quarterly, 118, 599-626.

61.

Pope

J. C.

Woon

(2009). Measuring changes in American party reputations, 1939-2004. Political Research Quarterly, 62, 653-661.

62.

Praino

Stockemer

Moscardelli

V. G.

(2013). The lingering effect of scandals in congressional elections: Incumbents, challengers, and voters. Social Science Quarterly, 94, 1045-1061.

63.

Rogers

(2012). Accountability in a federal system: How parties perform in office and state legislative elections. Unpublished manuscript, Princeton University, Princeton, NJ.

64.

Rundquist

B. S.

Strom

G. S.

Peters

J. G.

(1977). Corrupt politicians and their electoral support: Some experimental observations. American Political Science Review, 71, 954-963.

65.

Sabato

(1991). Feeding frenzy: How attack journalism has transformed American politics. New York, NY: Free Press.

66.

Shea

D. M.

(1999). All scandal politics is local: Ethical lapses, the media, and congressional elections. Harvard International Journal of Press Politics, 4, 45-62.

67.

Sides

(2007). The consequences of campaign agendas. American Politics Research, 35, 465-488.

68.

Slomczynski

Shabad

(2011). Perceptions of party corruption and voting behavior in Poland. Party Politics, 18, 897-917.

69.

Snyder

J. M.

Ting

M. M.

(2002). An informational rationale for political parties. American Journal of Political Science, 46, 90-110.

70.

Snyder

J. M.

Ting

M. M.

(2003). Roll calls, party labels, and elections. Political Analysis, 11, 419-444.

71.

Stockemer

LaMontagne

Scruggs

(2013). Bribes and ballots: The impact of corruption on voter turnout in democracies. International Political Science Review, 34, 74-90.

72.

Stokes

D. E.

(1963). Spatial models of party competition. American Political Science Review, 57, 368-377.

73.

Thompson

J. B.

(2000). Political scandal: Power and visibility in the media age. Cambridge, MA: Polity Press.

74.

Ulbig

S. G.

Martorano

(2012). The “Coingate” effect: The impact of scandal on attitudes toward state and federal political actors. Social Science Journal, 49, 61-71.

75.

Vivyan

Wagner

Tarlov

(2012). Representative misconduct, voter perceptions and accountability: Evidence from the 2009 House of Commons expense scandal. Electoral Studies, 31, 750-763.

76.

Welch

Hibbing

J. R.

(1997). The effects of charges of corruption on voting behavior in congressional elections. Journal of Politics, 59, 226-239.

77.

Winters

M. S.

Weitz-Shapiro

(2013). Lacking information or condoning corruption? Comparative Politics, 45, 418-436.