Issue ownership cues and candidate support

Abstract

Issue ownership theory suggests that candidates should focus on the issues that are owned by—or associated with—their parties and mostly avoid issues that are owned by the opposing party. Doing so allows them to focus on their own party’s strengths rather than on their weaknesses. Despite this expectation, contemporary research finds that candidates discuss both their own party’s issues and trespass by talking about issues owned by the opposing parties. I argue that issue ownership cues—subtle information cues linking candidates to parties through the discussion of party-owned issues—should have heterogeneous effects across partisan groups. Using a survey experiment, I show that copartisans prefer candidates who focus on issues owned by their parties while opposing partisans prefer candidates who trespass; independents’ preferences do not appear to shift in response to these cues. It thus appears as if cuing people to connect candidates to one party or the other can inform citizens’ levels of support for those candidates. Depending on the composition of the electorate, trespassing can be an advantageous strategy for candidates.

Keywords

campaign communication candidate support issue ownership parties public opinion

The theory of issue ownership—which postulates that parties can foster long-term positive reputations on some issues—is one of the most prominent theoretical frameworks through which scholars consider the importance of issues in party and electoral politics (Budge and Farlie, 1983; Petrocik, 1996). Many proponents of issue ownership argue that it informs candidate strategy by encouraging candidates to focus on the issues owned by their parties because this drives people to think about their strengths. Some research even suggests that candidates should never trespass by discussing issues owned by a different party (Simon, 2002). Despite this, a body of research shows that candidates often talk about the same issues as one another (Banda, 2015; Kaplan et al., 2006; Sigelman and Buell, 2004) and that trespassing is common (Banda, 2013; Banda and Carsey, 2015; Damore, 2004, 2005; Sides, 2006, 2007). This line of research, while important, is largely silent on how these messaging strategies might influence the way that citizens view candidates. I address this limitation by focusing on the following research question: How does the discussion of party-owned issues by candidates—those owned both by their own and by the opposing parties—inform citizens’ evaluations and support of those candidates?

I argue that citizens use issue ownership cues—subtle information cues that tie candidates to the party that owns the issue they discuss—to inform their evaluations of candidates. An issue ownership cue is present in a political message when an issue that is owned by a party is discussed and when no position on that issue is explicitly identified. Exposure to issue ownership cues encourage citizens to link the candidate who uses them to the party that owns the issue they discuss. Prior research suggests that these cues shift citizens ideological assessments of candidates in the direction associated with the party that owns the issue they discuss (see Banda, 2016). I extend this line of research by positing that issue ownership cues also influence the way that citizens evaluate candidates and that these effects are different among those who share a candidate’s party, those who identify with the opposing party, and those who are independents.

I use a randomized survey experiment embedded on a survey of just over one thousand American citizens conducted by Qualtrics during the summer of 2018 to test my theory. The results show that (1) copartisans report warmer evaluations and become more supportive when candidates discuss their own party’s issues, (2) opposing partisans are more supportive of and report warmer evaluations of candidates who trespass by discussing the other party’s issues, and (3) independents’ evaluations of and support for candidates are unaffected by issue ownership cues. Thus, this research offers important implications about how citizens respond to the discussion of party-owned issues during campaigns and for how candidates should design their messaging strategies to best match the electorates that they face. These results help to explain why issue trespassing seems to occur in campaigns so frequently: because relative to not discussing party-owned issues at all, trespassing increases support among opposing partisans and does not influence support among copartisans.

Party ownership of issues and candidate assessment

Issue ownership theory puts forth the notion that some issues are so closely associated with parties that citizens view them as being “owned” by those parties (Budge and Farlie, 1983; Egan, 2013; Goggin and Theodoridis, 2017; Petrocik, 1996; Petrocik et al., 2003). Parties gain ownership of issues by addressing problems related to those issues in a consistent manner over time.¹ When more citizens view Party A as being better able to handle problems associated with a given issue than some other party, scholars typically consider that issue to be owned by Party A. In the context of American politics, the Democratic Party is typically viewed as owning issues that are related to civil rights, civil liberties, social welfare, and social class and group relations. Republicans, on the other hand, tend to be perceived as owning issues related to national security, taxes, and crime.² This conceptualization of issue ownership thus represents a kind of issue advantage that comes from perceptions of the issue-based strengths and weaknesses of parties.

Some research also suggests that issue ownership is a multidimensional concept (Walgrave et al., 2015). Whereas the above research focuses on the competency dimension of issue ownership, other researchers (Walgrave et al., 2012) examine a second dimension: that of spontaneous association with parties. These two dimensions of issue ownership are probably related to one another. Citizens likely spontaneously associate issues with parties because they are aware of the reputations on those issues that parties have fostered over time. If a party owns an issue, that reputation should be positive and should exist because people believe that the party has consistently handled the issue well over time. Parties thus appear to try to foster reputations on issues in order to attract and maintain the support of constituencies that hold priorities that are congruent with the parties’ own priorities.

Issue ownership cues

Prior research establishes that issue ownership cues—the discussion of party-owned issues without explicitly identified issue positions—by candidates can affect citizens’ views of those candidates’ ideological and issue positions (Banda, 2016). Issue ownership cues contain subtle bits of information that citizens can use to form or alter their attitudes about candidates. Instead of making explicit statements about their positions on one or more issues, candidates can instead discuss an issue without taking a stance. Doing so allows citizens to infer their positions in ideological and issue space by using the party that owns the issue they discussed as a proxy. Thus citizens who are exposed to an issue ownership cue should view the candidate who communicates such a cue as having associated themselves with the party that is linked to the issue through issue ownership. In other words, they should view a candidate who communicates an issue ownership cue as being more similar to the party that owns the issue than they otherwise would have been. Issue ownership cues may make the association between candidates, issues, and parties more accessible and thus more likely to be used when evaluating candidates.

As shown by Banda (2016), a citizen who is exposed to a message containing an issue ownership cue involving the discussion of a Republican-owned issue tends to assess that candidate as being more conservative—the ideology associated with the Republican Party—than when that candidate communicates no cue at all. Similarly, if the message centered on a Democratic-owned issue instead, they generally view the candidate as being more liberal than a candidate who said nothing about issues. Thus, the evidence suggests that issue ownership cues can affect the ideological assessments of candidates made by citizens. Prior research does not, however, allow researchers to observe how these cues inform the degree to which citizens might support a given candidate. I consider this possibility below, focusing specifically on the way that people evaluate candidates and on how they decide whether or not to support those candidates.

How issue ownership cues inform candidate support

I argue that issue ownership cues can influence citizens’ levels of support for candidates for at least two reasons. First, as noted above, issue ownership cues influence perceptions of candidates’ issue and ideological positions. Much of the political behavior literature rests on the notion that the issue and ideological positions of citizens and candidates affects voting behavior (e.g Downs, 1957; Kropko and Banda, 2018; Rabinowitz and Macdonald, 1989; Tomz and Van Houweling, 2008). Thus, if an issue ownership cue convinces citizens that a candidate holds a displeasing position, that perceived position should drive down their evaluations of the candidate.

Second, issue ownership cues encourage people to tie the candidate communicating the cue to the party associated with the issue they discuss. In the American context, candidates can associate themselves with either the Republican or with the Democratic Parties.³ Thus, when evaluating candidates, citizens’ views of the party to which the candidate has associated herself with should inform their opinions about the candidates. If they like the party the candidate has linked themselves to, then people should like the candidate more. If they dislike that party, then their evaluations of the candidate tied to the party should worsen.

Issue ownership cues can thus offer people two kinds of information to use when evaluating candidates. When a candidate associates herself with her own party—that is, when she communicates an in-party ownership cue—she is providing reenforcing information that should strengthen the link between her and her party’s reputation. She could also provide counter-stereotypical information by tying herself to the opposing party through the use of an out-party ownership cue. Such a cue would associate her with the opposing party, which many Americans would view as being the opposite of her own party (see Heit and Nicholson, 2010).

The effects of issue ownership cues on candidate support should not be homogenous across all citizens. I focus on three groups of people in this research: those who share their partisanship with a candidate (copartisans), those who identify with a different party than does a candidate (opposing partisans), and those who do not identify with a party (independents). I do so because partisanship tends to lead people to process information in a biased fashion (Kam, 2005; Mondak, 1993; Taber and Lodge, 2006) and is one of the most powerful forces in American political behavior.

Copartisans should on average be supportive of candidates due to their shared partisan identities. This tendency does not mean that issue ownership cues are meaningless. Copartisans should evaluate candidates who associate themselves with their own party through an in-party ownership cue more positively because the cue provides an additional bit of information tying the candidate to her own party. This additional bit of stereotypical information should be pleasing. On the other hand, the communication of counter-stereotypical information through the use of an out-party ownership cue may be displeasing because it leads copartisans to associate the candidate with the opposing party. Threat perceptions can contribute to hostility felt toward out-groups (Brewer, 1999) and being exposed to a message in which one’s own party’s candidate ties herself to the opposition may make these threat-based attitudes more salient and thus more likely to be used when evaluating a copartisan candidate who communicates an out-party ownership cue. In addition, some research suggests that members of groups tend to sanction other members who violate the norms of their groups (e.g. Mendoza et al., 2014; Shinada et al., 2004). A candidate who associates herself with the opposing party may violate such norms. These possibilities lead to the following hypotheses:

Hypothesis 1a: Copartisans will be more supportive of candidates who communicate in-party ownership cues than candidates who do not.

Hypothesis 1b: Copartisans will be less supportive of candidates who communicate out-party ownership cues than candidates who do not.

Partisans who are tasked with evaluating a candidate from the opposite party—that is, opposing partisans—should for the most part exhibit more negative candidate evaluations relative to those citizens who evaluate copartisan candidates given the power of party cues. But their evaluations may still be influenced by issue ownership cues. A candidate from the opposing party who ties herself to her own party through an in-party ownership cue should be evaluated less positively than a candidate who does not because such a message provides additional information suggesting that she holds displeasing preferences. A candidate who uses an out-party ownership cue, on the other hand, should be evaluated more positively by opposing partisans due to the association the message creates between the candidate and the opposite party. In other words, Republican identifiers should be more supportive of a Democrat who prioritizes Republican-owned issues than one who does not. This out-party ownership cue should contain some information that these Republicans should find pleasing, though likely not enough to overwhelm the strength of the Democratic label. Put more formally:

Hypothesis 2a: Opposing partisans will be less supportive of candidates who communicate in-party ownership than candidates who do not.

Hypothesis 2b: Opposing partisans will be more supportive of candidates who communicate out-party ownership cues than candidates who do not.

Next, I consider political independents. Many survey respondents claim to identify as independents before admitting to being undercover partisans upon further probing (Keith et al., 1992; Klar, 2014; Petrocik, 2009). I focus here on “pure” independents, that is, those who continue to report that they are independents even after being probed. I will refer to pure independents as “independents” throughout this article. Independents, while certainly not perfectly unbiased information processors, lack partisan lenses that systematically color their perceptions of the social world. How independents’ candidate evaluations should shift in response to issue ownership cues is less certain. I consider two sets of possibilities. First, they could respond in similar ways as I have outlined above for partisans. Independents might view candidates who communicate in-party ownership cues more positively because such cues remind them of the reputation-based strengths of the party of which that candidate is a member. Similarly, an out-party ownership cue might encourage independents to think about the strengths of the opposing party, thus driving down their evaluations of the candidate. In other words:

Hypothesis 3a: Independents will be more supportive of candidates who communicate in-party ownership cues than candidates who do not.

Hypothesis 3b: Independents will be less supportive of candidates who communicate out-party ownership cues than candidates who do not.

The second set of possibilities is directly at odds with the above expectations leading to hypothesis 3b. First, by receiving counter-stereotypical information via an out-party ownership cue, independents may be encouraged to view the candidate more positively because they are more likely to view them as being an atypical Democrat or Republican. This may make them appear more appealing in large part because out-party ownership cues lead people to view candidates as being more ideologically moderate (see Banda, 2016), which independents may prefer relative to more ideologically extreme candidates.

Second, rather than driving down evaluations by pushing people to consider the positive reputation of the opposing party, an out-party ownership cue may encourage independents to view a candidate as being competent across a wider range of issues—more specifically, among those owned by a different party—than they might otherwise. For example, a Democratic candidate who talks about crime—a Republican-owned issue—may be perceived by independents as being more capable of handling problems related to that issue than they would be had they not used such an out-party ownership cue. This leads to the following expectation:

Hypothesis 3c: Independents will be more supportive of candidates who communicate out-party ownership cues than candidates who do not.

Research design

I test these hypotheses using a survey experiment embedded in national survey conducted by Qualtrics between June 25th and July 14th of 2018. The sample is made up of 1100 participants, who are approximately representative of US citizens as a whole. Participants took the survey through the Internet.⁴ They first answered several questions about their demographic characteristics and political views. For the purposes of this research, the most important of these captured the partisan identities of participants. I use the standard seven-point scale generated by the traditional two-stage partisanship question and collapse it down to three categories: Democrats (47%), independents (19%), and Republicans (34%). I treat leaning partisans as partisans, not as independents. Thus, undercover partisans are treated as Democrats or Republicans, not as independents (Klar, 2014). Interested readers can find a table containing the key descriptive summary statistics of the sample in the Online Appendix.

Experimental treatments

After completing the pretest questionnaire, participants read a short statement attributed to a fictional Senator running for reelection in “a different state.”⁵ Using a fictional rather than actual political figure is useful because doing so allows me to isolate and observe the effects of issue ownership cues on candidate evaluations without needing to account for the views that citizens already hold about a real politician. This choice might, however, limit the degree to which the results may be generalized because they may only be indicative of how people respond to issue ownership cues when faced with candidates with whom they are unfamiliar.

The experiment began with the following text:

We would like your reaction to some comments made at a recent re-election campaign appearance by a Senator from a different state, [Democratic/Republican] Senator Franklin, who said:

The partisanship of the Senator was randomly assigned with equal probability as either a Democrat or a Republican. I categorize participants into three groups based on their own partisan identities and that of the Senator that mirror those discussed in the theory section. The “copartisan” group includes participants whose partisanship matches the party label associated with the Senator in the opening text of the treatment. In other words, Democrats matched with Democratic Senators and Republicans matched with Republican Senators are part of the copartisan group. The “opposing partisans” group contains the opposite pairings of people and Senators: Democrats matched with Republican Senators and Republicans matched with Democratic Senators. Last, the “independents” group contains all independent participants regardless of the partisanship of the Senator with whom they were matched.

Participants were then randomly exposed to one of three types of treatments. In the control treatment, no issue ownership cue was present. The Senator’s statement included no issue content, so the only piece of explicitly political information was his partisanship. The text read:

America is the greatest country in the world, and I intend to keep it that way. Send me back to Washington so I can keep fighting for you and our families. I have been a successful leader for my entire life, and I will continue to be a leader in the United States Senate well into the next decade.

Participants also could have been exposed to one of two different treatments containing issue ownership cues. In the in-party and out-party ownership cues treatments, the Senator only talked about issues owned by one party. If he talked about Democratic-owned issues, participants saw the following text:

Education is an important issue in this campaign. I will be talking about it a lot it in the coming months. Health care is also high on my agenda. I can assure you that I will devote a great deal of attention to this issue in the future as well.

If he talked about Republican-owned issues, the treatment was as follows:

National defense is an important issue in this campaign. I will be talking about it a lot in the coming months. Taxes are also high on my agenda. I can assure you that I will devote a great deal of attention to this issue in the future as well.

If the Senator was identified as a Democrat, then the first block of text constitutes an in-party ownership cue. Had the second block of text been used instead, the message would be an out-party ownership cue. Similarly, a Republican candidate connected to the first quote would have communicated an out-party ownership cue while they would have transmitted an in-party ownership cue had the second quote been attributed to them.

The above issues are typically viewed as being owned by the parties (Petrocik, 1996). More recent work by Egan (2013: 67) shows that the parties have longstanding positive reputations on these issues. These strong reputations should make the education and health care strong candidates to encourage citizens to link a candidate who discusses them to the Democratic Party. Similarly, candidates who talk about taxes and national defense should be connected to the Republican Party due to its long-term positive reputation on those issues.

Participants had an equal probability of being exposed to the control treatment or to one of two experimental treatments containing an issue ownership cue: the “in-party ownership cue” treatment or the “out-party ownership cue” treatment. Because of this random assignment, about one-third of each partisan group was shown each of the three treatments. Interested readers may see the exact distribution of the treatments by partisan group in the Online Appendix.

Key variables

Exposure to a given treatment—described above—serves as the primary independent variable of interest in my analyses. I capture participants’ levels of candidate support in two ways, both of which were collected using questions asked immediately following exposure to a treatment. First, participants were asked to evaluate the Senator using a standard feeling thermometer instrument. The text of the question read “Please rate Senator Franklin using what we call a feeling thermometer. Ratings between 50 and 100 mean that you feel favorable and warm toward Senator Franklin. Ratings between 0 and 50 mean you feel unfavorable and cool toward Senator Franklin.” This variable captures participants’ affective evaluations of the Senator. Second, participants were asked if they would vote for Senator Franklin, with “yes” coded 1 and “no” coded 0. I test my hypotheses using a series of difference in means—for the evaluations variable—and proportions—for the vote choice variable—tests in which I compare the average values of the dependent variables by treatment group within partisan categories of participants.

Results

I first examine the effects of issue ownership cues on copartisans’ support of the Senator, which are plotted in Figure 1. In each of the figures presented throughout this research, the points represent the difference in mean evaluation of the Senator (left panel) or the difference in the proportion of participants who say they would vote for the Senator (right panel) between each possible pairing of the three treatments.⁶ The vertical lines plotted through the points capture the 90% confidence intervals around these quantities of interest. If a confidence interval overlaps with the dashed horizontal line plotted at 0 on the y-axis, then it indicates that I am unable to reject the null hypothesis that a given quantity of interest differs significantly from 0. Note that I report the 90% confidence intervals because my hypotheses are directional in nature. Thus, the p values discussed throughout these results are one-tailed. Interested readers should note that in all cases, when a test produces a one-tailed p value that is less than or equal to 0.05, it also produces a two-tailed test statistic that is less than or equal to 0.05.

Figure 1.

The effects of issue ownership cues on evaluations and vote choice among copartisans. Points are difference in means (left) or proportions (right). Vertical lines represent 90% confidence intervals.

This analysis of copartisans focuses on Democrats evaluating the Democratic version of the Senator and Republicans evaluating the Republican version of the Senator. Thus, these data allow me to test hypotheses 1a and 1b. Among copartisans, the in-party ownership treatment on average leads participants to make evaluations of the Senator that are about 6.92 units more positive than are those produced by the control treatment. This difference is both substantively and statistically significant (t = 2.52 and p ≤ 0.05).⁷ The in-party ownership cue treatment produces a similar effect when examining participants’ voting choices: there is an increase in the proportion of copartisans who report that they would vote for the Senator increases of about 0.11 (z = 2.04 and p ≤ 0.05). In other words, it appears as if copartisans become more supportive of the Senator when he links himself to his own party. These results support hypothesis 1a.

These data fail to provide evidence in favor of hypothesis 1b. Note that while the mean evaluation of the Senator among those who received an out-party ownership cue was about 2 units lower than it was for those who were exposed to the control treatment, the difference fails to achieve anything approaching a traditional level of statistical significance (t = 0.72 and p = 0.24). Similarly, the proportion of participants who report that they would vote for the Senator is 0.02 units lower among out-party ownership cue recipients relative to those who received the control treatment. This difference also fails to attain a traditional level of statistical significance (z = 0.30 and p = 0.38). Thus, it does not appear as if out-party ownership cues are harmful to a candidate relative to not communicating an issue ownership cue. That said, the data do show that evaluations of and willingness to vote for the Senator are about 8.91 units (t = 3.36 and p ≤ 0.05) and 0.13 units (z = 2.37 and p ≤ 0.05) higher among participants exposed to an in-party ownership cue relative to an out-party ownership cue.

On the whole, these results suggest that partisans’ views of their own parties’ Senators are not informed by out-party ownership cues. In other words, Democrats do not appear to alter their evaluations of the Democratic version of the Senator when he discusses Republican-owned issues like taxes. Though the out-party ownership cue may link that Democrat with the Republican Party, Democratic identifiers may be inclined to downplay it when deciding how supportive they should be of their party’s candidate largely due to the strength of partisanship in informing political cognitions and behavior. In-party ownership cues, on the other hand, do appear to influence levels of support among copartisans. In general, it appears as if copartisans reward their parties’ candidates by becoming more supportive of them when they link themselves to their own party via in-party ownership cues.

Next, I report tests of hypotheses 2a and 2b among opposing partisans, that is, Democrats evaluating the Republican version of the Senator and Republicans evaluating the Democratic version of the Senator. As shown in Figure 2, the in-party ownership cues treatments does not appear to influence either of the dependent variables relative to the control treatment. Exposure to the out-party ownership cues treatment, on the other hand, does appear to lead participants to form different attitudes about the Senator. This treatment on average leads participants to evaluate the candidate about 8.76 degrees warmer (t = 2.91 and p ≤ 0.05) relative to the control. It similarly influences vote intentions; the proportion of opposing partisans who reported that they were willing to vote for the Senator was 0.19 units higher (z = 3.23 and p ≤ 0.05) among out-party ownership cue recipients than it was among those who were exposed to the control condition. Both effects are substantively large, especially the latter given that an 19 percentage point increase in the rate of defections could change the outcome of otherwise competitive elections. Readers should also note that average evaluations ( $Δ$ = 9.77, t = 3.20, and p ≤ 0.05) and vote intentions ( $Δ$ = 0.22, z = 3.70, and p ≤ 0.05) are higher among opposing partisans who received the out-party ownership cue relative to those who were exposed to the in-party cue. On the whole, I observe no evidence favoring hypothesis 2a, but the above tests provide strong support for hypothesis 2b.

Figure 2.

The effects of issue ownership cues on evaluations and vote choice among opposing partisans. Points are difference in means (left) or proportions (right). Vertical lines represent 90% confidence intervals.

Last, I test hypotheses 3a and 3b by observing how issue ownership cues influence evaluations of and vote intention for the Senator among independents. These data are plotted in Figure 3. These results provide little support for either hypothesis. First, the mean levels of independents’ affective evaluations of the Senator are essentially equal across treatments. Next, note that while both issue ownership cue treatments appear to lead to substantively meaningful increases of about 0.10 units in vote intention relative to the control condition among independents, neither effect is statistically significant. Thus, these tests fail to uncover sufficiently clear evidence that independents’ support-related attitudes are responsive to ownership cues. In other words, the above results fail to offer support for either hypotheses 3a or 3b.

Figure 3.

The effects of issue ownership cues on evaluations and vote choice among independents. Points are difference in means (left) or proportions (right). Vertical lines represent 90% confidence intervals.

In more substantive terms, these results suggest that candidates’ decisions about their messaging strategies should be informed by the partisan makeup of the electorates they face. If they are competing primarily for the support of their own partisans, then they should focus on the issues owned by their own party. A candidate who faces an electorate largely made up of supporters of the opposing party, on the other hand, should instead discuss issues that are owned by the other party. The strategic calculus becomes more complicated in more competitive environments, which may encourage candidates to talk about a mix of issues owned by both parties. It might also lead candidates to try to discuss their own issues when communicating with members of their own party and issues owned by the opposing party when addressing members of the other party.

Additional analyses

Given that my theory posits that citizens’ evaluations of candidates shift in response to issue ownership cues because those cues alter citizens’ ideological assessments of candidates’ positions, some readers may be interested in observing whether or not this process occurs. The following section contains two sets of additional analyses: (1) an observation of the degree to which exposure to issue ownership cues influence participants’ ideological assessments of the candidate and (2) an analysis designed to show how the effects of issue ownership cues on candidate support might be mediated by participants’ perceptions of their ideological distance from the candidate.

After participants were exposed to a treatment, they were asked to place the Senator on a standard seven-point ideological spectrum.⁸ On this scale, 1 meant “very liberal,” 2 meant “liberal,” 3 meant “somewhat liberal,” 4 meant “moderate,” 5 meant “somewhat conservative,” 6 meant “conservative,” and 7 meant “very conservative.” To maintain consistency with prior work (Banda, 2016), I mirrored this scale such that higher values indicate greater levels of congruence with the ideological pole associated with the parties. For example, a Democrat who was assessed as being “very liberal” would have a value of 7 while one who was identified as being “very conservative” was given a value of 1. A Republican Senator who was called “very conservative,” on the other hand, was given a value of 7 while a “very liberal” Republican was coded as 1.

Thus, the dependent variable here is a measure of congruence with the ideological stereotype associated with the Senator’s party. My expectation is that participants who are exposed to an in-party ownership cue will view the Senator as being more ideologically congruent with his party relative to the control treatment. Furthermore, exposure to the out-party ownership cue treatment should encourage participants to view the Senator as being less ideologically congruent with his party. In more concrete terms, an in-party ownership cue should push ideological assessments of a Republican Senator further to the right of the ideological spectrum because the cue provides an additional link between the Senator and the Republican Party beyond the explicit mention of his party. On the other hand, if the Republican communicates an out-party cue, then participants should view him as being less to the right because this cue produces a connection between him and the Democratic Party.

The relevant difference in means tests are plotted in Figure 4. The panel in the top-left corner shows these differences among all participants while the remaining three panels plot them for each of the three partisan groups. These results provide strong evidence that issue ownership cues inform people’s perceptions of the ideological congruence of the Senators with their own parties. Among all participants, the in-party ownership cue leads people to increase their levels of assessed congruence by 0.38 units (t = 3.24 and p < 0.05) on average while the out-party ownership cue treatment leads participants to view the Senator as 0.46 units (t = 3.89 and p < 0.05) less ideologically congruent with his party, both relative to the control treatment.

Figure 4.

The effects of issue ownership cues on ideological assessments of the Senator across the treatments. Points are differences in means. Vertical lines represent 90% confidence intervals.

I observe similar patterns among copartisans. The in-party ownership cue encourages copartisans to view the Senator as 0.47 units (t = 2.47 and p < 0.05) more ideologically congruent with his party than does the control condition. Exposure to the out-party ownership cue appears to lead participants to assess the Senator as being about 0.44 units (t = 2.43 and p < 0.05) less congruent with his party than does the control treatment. Opposing partisans also view the Senator as being 0.77 units (t = 3.81 and p < 0.05) less ideologically congruent with his party when they are exposed to an out-party ownership cue relative to when they view the control treatment, but the in-party ownership cue treatment does not on average appear to provoke a response relative to the control. Finally, I observe the opposite pattern among independents, who view the Senator as being more ideologically congruent with his party by 0.60 units (t = 2.69 and p < 0.05) when exposed to the in-party ownership cue treatment relative to the control but who do not appear to systematically respond in a way that differs from the control condition to the out-party ownership cue treatment. On the whole, these results largely mirror those reported by Banda (2016).

Given that issue ownership cues appear to influence people’s ideological assessments of the candidate and prior research suggests that spatial considerations influence voting behavior, it is reasonable to observe how the effects of issue ownership cues on candidate support might be mediated by people’s perceptions of the ideological distance between themselves and the candidate. I create a measure of ideological distance by calculating the absolute value of the difference between respondents’ seven-point self-placed ideological positions⁹ from the raw version of the seven-point ideological assessment of the candidate measure described above. This measure ranges from 0 to 6 where higher values indicate greater perceived ideological distance between a participant and the candidate. The mean and standard deviation of this measure are both approximately equal to 1.56.

Table 1 contains the results of a series of mediation tests. More specifically, these results show (1) the indirect effects of issue ownership cues on candidate support as mediated through perceived distance from the candidate, (2) the direct effects of issue ownership cues on candidate support, (3) the total indirect and direct effects of issue ownership cues on candidate support, and (4) the proportion of the total effects of issue ownership cues that is mediated. I present each of these quantities in comparisons between each of the three treatments.

Table 1.

How the effects of issue ownership cues are mediated by perceptions of ideological distance from the candidate.

	Evaluations			Vote choice
	Control vs. in-party	Control vs. out-party	In-party vs. out-party	Control vs. in-party	Control vs. out-party	In-party vs. out-party
Copartisans
Indirect effects	0.38 (0.91)	−0.59 (0.90)	0.80 (0.73)	0.01 (0.02)	−0.01 (0.02)	0.02 (0.02)
Direct effects	6.54* (2.60)	−1.41 (2.63)	8.11* (2.56)	0.11* (0.05)	−0.01 (0.05)	0.11* (0.05)
Total effects	6.92* (2.75)	−2.00 (2.78)	8.91* (2.65)	0.11* (0.06)	−0.02 (0.06)	0.13* (0.05)
Mediated proportion of total effects	0.05	0.29	0.09	0.06	0.66	0.13
Opposing partisans
Indirect effects	−1.50 (1.29)	4.20* (1.36)	−6.27* (1.46)	−0.02 (0.02)	0.06* (0.02)	−0.09* (0.03)
Direct effects	0.49 (2.54)	4.56* (2.81)	−3.51 (2.82)	−0.00 (0.05)	0.13* (0.06)	−0.12* (0.06)
Total effects	−1.01 (2.84)	8.76* (3.01)	−9.77* (3.06)	−0.02 (0.06)	0.19* (0.06)	−0.22* (0.06)
Mediated proportion of total effects	0.75	0.48	0.64	0.93	0.33	0.43
Independents
Indirect effects	1.32 (1.01)	1.84 (1.31)	−1.36 (0.96)	0.01 (0.01)	0.01 (0.02)	−0.01 (0.01)
Direct effects	−0.37 (4.29)	−1.82 (4.37)	2.29 (3.80)	0.09 (0.08)	0.09 (0.09)	0.01 (0.09)
Total effects	0.95 (4.30)	0.02 (4.24)	0.93 (3.83)	0.10 (0.08)	0.10 (0.09)	−0.00 (0.09)
Mediated proportion of total effects	0.78	0.50	0.37	0.08	0.10	0.53

Note: Standard errors are shown in parentheses.

*p ≤ 0.05 (one-tailed).

First, note that there is no evidence of mediation among copartisans. More specifically, none of the indirect effects of issue ownership cues appears to differ significantly from zero (p ≤ 0.05) in any of the paired comparisons of the treatment groups among copartisan participants. As was the case in the initial analyses, the out-party ownership cue did not affect the evaluations or the vote choice of copartisans relative to the control treatment as indicated by the lack of statistically significant direct and total effects. The in-party ownership cue, on the other hand, does appear to exhibit some noticeable influence when compared to the effects of the control and to the out-party treatments. The direct effect of exposure to the in-party cue relative to the control treatment is a 6.54 unit increase in copartisans’ evaluations of the candidate. Exposure to the in-party ownership cue also appears to directly increase copartisans’ evaluations of the candidate by 8.11 units compared to the out-party ownership cue treatment. The patterns are similar in the vote choice mediation results. Note that the proportion of the total effects of the in-party ownership cue that are mediated by perceptions of distance are quite small across the tests. This indicates that most of the effects that were described in the initial analysis of copartisans were driven by the in-party cue itself.

Next, note that the in-party ownership cue does not inform evaluations or vote choice relative to the control treatment among opposing partisans. Mirroring the initial findings, the out-party ownership cue appears to encourage opposing partisans to report higher levels of support for the candidate relative to both of the other two treatments. Furthermore, note that the patterns of mediation here are more complex than was the case among copartisans. For example, in comparison to the control treatment, exposure to the out-party ownership cue leads to a total increase of 8.76 units in evaluations of the candidate. About 4.56 of those units are driven directly by the cue itself while the remaining 4.2 unit increase is driven by the cue indirectly through opposing partisans’ perceptions of ideological distance from the candidate. About 48% of the total effect is mediated by perceptions of distance. Approximately 64% of the total effects of the in-party ownership cue on opposing partisans’ evaluations of the candidate relative to the out-party ownership cue are mediated by perceived distance. In this case, the direct effects of the ownership cue fail to achieve statistical significance. Further note that the ownership cues appear to influence vote choice both directly and indirectly through perceptions of distance.

Neither type of issue ownership cue appears to influence independents’ evaluations or support in relation to the control treatment or to each other. Furthermore, the cues do not appear to exert any statistically significant influence over these dependent variables either directly or indirectly. These mediation results mirror those presented earlier in thearticle.

I conduct a third robustness check—presented in the Online Appendix¹⁰ due to sample size concerns—wherein I split Democratic and Republican participants and observe how they respond to issue ownership cues when evaluating copartisan and opposing party candidates. Consistent with the results described in Figure 1, Democrats and Republicans respond to their own parties’ candidates communicating in-party ownership cues by evaluating the candidates more positively. Their views of their own parties’ candidates do not appear to be responsive to the presence of out-party ownership cues. Democrats also behave in ways that are consistent with the overall results shown in Figure 2: they reward candidates from the opposing party who communicate out-party cues—that is, Republicans who prioritize Democratic-owned issues—by viewing them more positively. The attitudes of Republican identifiers, on the other hand, do not appear to be informed by either type of issue ownership cue when they are tasked with evaluating a Democratic figure. This inconsistent finding could be driven by Republicans having different preferences about representation than Democrats (Barker and Carman, 2012) or by the small number of observations, especially among Republicans, when subsetting the data in this manner.

Conclusion

This research provides evidence suggesting that candidates who communicate ownership cues are evaluated by citizens in systematically different ways depending on those citizens’ partisan identities. More specifically, those who share their partisan identities with candidates tend to exhibit higher levels of support when those candidates talk about their own parties’ issues (an in-party ownership cue) relative to when they do not. People who identify with a different party than a candidate, on the other hand, appear to prefer when those candidates discuss the opposing parties’ issues (an out-party ownership cue). Independents’ support attitudes do not appear to be particularly responsive to issue ownership cues. In addition, people’s perceptions of ideological distance from the candidate appears to mediate the effects of exposure to out-party ownership cues among opposing partisans. Perceived distance does not, however, mediate the effects out-party ownership cues among other partisans or of in-party ownership cues among any partisan group. On the whole, it appears as if issue ownership cues drive cognitions of candidates, but in different ways based upon the partisan identities of citizens and the candidates they must evaluate.

These results lead to a few implications for our understanding of both public opinion and political messaging strategies. First, this research provides additional evidence that issue ownership can influence people’s cognitions, in this case by informing the way that people evaluate candidates. Issue ownership cues are subtle information cues that require relatively complex cognitive tasks on the part of citizens. People must connect the issues discussed to the party that owns the issues and finally to the candidate who communicated the cues. Citizens appear to rely on the stereotypes associated with the parties that own a given issue to form evaluations of candidates. If issue ownership cues can influence citizens’ evaluations of candidates, they may be more broadly meaningful. Perhaps citizens use these cues when forming views of the candidates’ character traits (see for example Bauer, Forthcoming).

Second, these results help to explain why candidates engage in issue trespassing: because doing so improves candidate evaluations among opposing partisans—in part by encouraging more moderate ideological assessments—and appear to lead to no costly losses among copartisans, at least in relation to the absence of messages containing issue ownership cues. Convincing even relatively small portions of the opposing party’s voters to defect can be key to winning elections in competitive election environments and for candidates competing in districts in which they face a lopsided partisan disadvantage.

This leads to a third implication, which is that the partisan composition of an electorate may inform a candidate’s messaging strategy. Candidates might be more likely to trespass as their electorates become increasingly made up of opposing partisans. In an environment with many members of the opposing party, savvy candidates should trespass in order to appeal to those potential voters. When there are comparatively few opposing partisans, trespassing may be a less useful strategy relative to a candidate focusing on her own party’s issues.

These results lead to several possible avenues of research. Future research might, as touched on above, examine the degree to which issue ownership cues affect citizens’ perceptions of candidates’ traits. Researchers might move beyond assessment and evaluation of candidates and consider how issue ownership cues affect people’s attitudes about institutions. For example, how does learning that a legislature is going to take up an issue owned by the minority party—that is, something like an out-party ownership cue—affect citizens’ attitudes toward that legislature and the parties that exist within it? Scholars might also examine how the partisan makeup of candidates’ electorates informs their proclivity to discuss issues owned by their own and the opposing party through a thorough analysis of those candidates’ speeches, mailings, and television advertisements. Finally, given that there is variance in individuals’ perceptions of party issue ownership and that these attitudes can be informed by elite communication (Dahlberg and Martinsson, 2015; Stubager and Seeberg, 2016; Tresch et al., 2015; Walgrave et al., 2009), future research might explore the degree to which citizens’ individual views about party ownership of issues condition their responsiveness to issue ownership cues.

Supplemental Material

Supplemental Material, Appendix - Issue ownership cues and candidate support

Supplemental Material, Appendix for Issue ownership cues and candidate support by Kevin K Banda in Party Politics

Footnotes

Author’s note

Replication data can be found at .

Acknowledgement

The author would like to thank Tom Carsey, Patrick Egan, George Rabinowitz, Mike MacKuen, John Petrocik, Pam Conover, Jim Stimson, Virginia Gray, and Dave Peterson for feedback on earlier versions of this research. All mistakes are of my own making.

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) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Kevin K Banda

Supplemental Material

Supplemental material for this article is available online.

Notes

References

Banda

(2013) The dynamics of campaign issue agendas. State Politics and Policy Quarterly 13(4): 446–470.

Banda

(2015) Competition and the dynamics of issue convergence. American Politics Research 43(5): 821–845.

Banda

(2016) Issue ownership, issue positions, and candidate assessment. Political Communication 33(4): 651–666.

Banda

Carsey

(2015) Two-stage elections, strategic candidates, and campaign agendas. Electoral Studies 40(1): 221–230.

Barker

Carman

(2012) Political Representation in Red and Blue America: How Cultural Differences Shape Democratic Expectations and Outcomes. Oxford: Oxford University Press.

Bauer

(Forthcoming) The effects of partisan trespassing strategies across candidate sex. Political Behavior: 1–19.

Brewer

(1999) The psychology of prejudice: ingroup love or outgroup hate? Journal of Social Issues 55(3): 429–444.

Budge

Farlie

(1983) Explaining and Predicting Elections: Issue Effects and Party Strategies in Twenty-Three Democracies. London: Unwin Hyman.

Dahlberg

Martinsson

(2015) Changing issue ownership through policy communication. West European Politics 38(4): 817–838.

10.

Damore

(2004) The dyanamics of issue ownership in presidential campaigns. Political Research Quarterly 57(3): 391–397.

11.

Damore

(2005) Issue convergence in presidential campaigns. Political Behavior 27(1): 71–97.

12.

Downs

(1957) An Economic Theory of Democracy. New York: Harper and Row.

13.

Egan

(2013) Partisan Priorities: How Issue Ownership Drives and Distorts American Politics. New York: Cambridge University Press.

14.

Goggin

Theodoridis

(2017) Disputed ownership: parties, issues, and traits in the minds of voters. Political Behavior 39(3): 675–702.

15.

Heit

Nicholson

(2010) The opposite of republican: polarization and political categorization. Cognitive Science 34(8): 1503–1516.

16.

Kam

(2005) Who toes the party line? Cues, values, and individual differences. Political Behavior 27(2): 163–182.

17.

Kaplan

Park

Ridout

(2006) Dialogue in American political campaigns? An examination of issue convergence in candidate television advertising. American Journal of Political Science 50(3): 724–736.

18.

Keith

Magleby

Nelson

, et al. (1992) The Myth of the Independent Voter. Berkeley: University of California Press.

19.

Klar

(2014) Identity and engagement among political independents in America. Political Psychology 35(4): 577–591.

20.

Kropko

Banda

(2018) Issue scales, information cues, and the proximity and directional models of voter choice. Political Research Quarterly 71(4): 772–787.

21.

Mendoza

Lane

Amodio

(2014) For members only: ingroup punishment of fairness norm violations in the ultimatum game. Social Psychological and Personality Science 5(6): 662–670.

22.

Mondak

(1993) Public opinion and heuristic processing of source cues. Political Behavior 15(2): 167–192.

23.

Petrocik

(1996) Issue ownership in presidential elections, with a 1980 case study. American Journal of Political Science 40(3): 825–850.

24.

Petrocik

(2009) Measuring party support: leaners are not independents. Electoral Studies 28(4): 562–572.

25.

Petrocik

Benoit

Hansen

(2003) Issue ownership and presidential campaigning, 1952-2000. Political Science Quarterly 118(4): 599–626.

26.

Rabinowitz

Macdonald

(1989) A directional theory of issue voting. American Political Science Review 83: 93–121.

27.

Shinada

Yamagishi

Ohmura

(2004) False friends are worse than bitter enemies: ‘altruistic’ punishment of in-group members. Evolution and Human Behavior 25(6): 379–393.

28.

Sides

(2006) The origins of campaign agendas. British Journal of Political Science 36(3): 407–436.

29.

Sides

(2007) The consequences of campaign agendas. American Politics Research 35(4): 465–488.

30.

Sigelman

Buell

(2004) Avoidance or enagagement? Issue convergence in US presidential campaigns, 1960-2000. American Journal of Political Science 48(4): 650–661.

31.

Simon

(2002) The Winning Message: Candidate Behavior, Campaign Discourse, and Democracy. Cambridge; New York: Cambridge University Press.

32.

Stubager

Seeberg

(2016) What can a party say? How parties’ communication can influence voters’ issue ownership perceptions. Electoral Studies 44(1): 162–171.

33.

Taber

Lodge

(2006) Motivated skepticism in the evaluation of political beliefs. American Journal of Political Science 50(3): 755–769.

34.

Tomz

Van Houweling

(2008) Candidate positioning and voter choice. American Political Science Review 102(3): 303–318.

35.

Tresch

Lefevere

Walgrave

(2015) “Steal me if you can!” The impact of campaign messages on associative issue ownership. Party Politics 21(2): 198–208.

36.

Walgrave

Lefevere

Tresch

(2012) The associative dimension of issue ownership. Public Opinion Quarterly 76(4): 771–782.

37.

Walgrave

Lefevere

Nuytemans

(2009) Issue ownership stability and change. How political parties claim and maintain issues through media appearances. Political Communication 26(1): 153–172.

38.

Walgrave

Tresch

Lefevere

(2015) The conceptualization and measurement of issue ownership. West European Politics 38(4): 778–796.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.19 MB