Hearing the Other Side Revisited: The Joint Workings of Cross-Cutting Discussion and Strong Tie Homogeneity in Facilitating Deliberative and Participatory Democracy

Abstract

This study seeks to shed light on the highly publicized democracy dilemma signaling that encountering disagreement tends to promote deliberative democracy, while the same experience can dampen a citizen’s motivation to participate. By assessing the processes wherein the joint workings of cross-cutting discussion and strong tie homogeneity are simultaneously associated with the outcomes of deliberative and participatory democracy, we provide a number of key insights into the puzzling quandary. First, our results indicate that cross-cutting discussion and strong tie homogeneity interact with each other to predict increased political participation. Second, the relationship between cross-cutting discussion and preference for open dialogue is stronger for those who belong to a congenial primary network. Third, efficacious individuals seem more capable of translating the benefits of an ongoing deliberative orientation into meaningful political behavior. The current research suggests that deliberation and participation can go hand in hand under a particular network context.

Keywords

democracy dilemma cross-cutting discussion strong tie homogeneity dialogic openness political participation

A sizable number of social scientists from various research domains have long debated about the effects of cross-cutting communication experiences on the health of a democratic society. Nevertheless, the literature remains overwhelmed by the puzzling dilemma succinctly explicated in Mutz’s (2006) Hearing the Other Side. In that monograph, she documented that encountering disagreement tends to promote deliberative democracy, while the same experience can dampen a citizen’s motivation to participate. In this vein, she suggested,

[t]he best social environment for cultivating political activism is one in which people are surrounded by those who agree with them, people who reinforce the sense that their own political views are the only right and proper way to proceed. (p. 3)

Ironically, the pairing of exposure to disagreement and agreement can provide a point of entry for resolving the deadlock to the extent they complement, rather than contradict, each other. Berelson (1952) early noted that a balance between political cleavage and consensus is a sine qua non of a healthy democratic system. More explicitly, Berelson, Lazarsfeld, and McPhee (1954) suggested that such a balance is best maintained when “there is a good deal of cross-group and cross-party identification and affiliation within the community” (p. 320). Although many have long appreciated the importance of maintaining a balance between consent and dissent (see also Nir, 2005; Wojcieszak, 2009), few have substantively examined what a good balance specifically refers to and how it could aid the practices of democracy. Thus we seek to fill the gap in the extant literature by showing that the proper mix of homogenous and heterogeneous ingredients can indeed enrich deliberative as well as participatory souls of democracy.

Using a nationally representative survey of adults in the United States, this study takes a particular interest in the interplay between cross-cutting discussion and strong tie homogeneity in order to resolve the democracy dilemma. In fact, there have been fruitful research efforts to address the dilemma. Some strived to identify conditions under which deliberation may coincide with participation by advancing more complex mediating (Scheufele, Hardy, Brossard, Waismel-Manor, & Nisbet, 2006; Scheufele, Nisbet, Brossard, & Nisbet, 2004) and moderating (Jang, 2009; Kim, Scheufele, & Han, 2011; Kwak, Williams, Wang, & Lee, 2005) channels. Others examined different forms of deliberation (Eveland & Hively, 2009) and participation (F. L. F. Lee, 2012), arguing that the relationship between these two democratic ideals depends on how the concepts are defined and operationalized. For the most part, this line of research has focused on the influence of cross-cutting experience on political mobilization, while paying little heed to deliberative consequences (for an exception, see Wojcieszak, Baek, & Delli Carpini, 2010). The current research differs from these previous attempts in that we assess the processes wherein exposure to agreement or disagreement is simultaneously associated with the outcomes of deliberative and participatory democracy. Furthermore, by examining the moderating role of internal efficacy in the proposed indirect process, we aim to identify a key condition under which the deliberative outcome can translate to meaningful political action. With the concurrent examination of the two types of democracy, we seek to show that the two democratic ideals can exist in a complementary rather than conflicting framework under a particular network context.

Joint Workings of Cross-Cutting Discussion and Strong Tie Homogeneity in Participatory Democracy

The influence of cross-cutting experiences on political participation is a subject of a considerable scholarly debate, with some showing a harmful effect (Mutz, 2002a, 2006; Valenzuela, Kim, & Gil de Zúñiga, 2012), while others emphasizing its salutary role in encouraging participatory democracy (Ikeda & Boase, 2011; Scheufele et al., 2004, 2006). Perhaps unraveling underlying processes may help us reconcile these conflicting results (F. L. F. Lee, 2012). On one hand, encountering disagreement can make citizens less politically active by leaving them ambivalent about complex issue stances (Huckfeldt, Mendez, & Osborn, 2004; Mutz, 2002b, 2006). Conversely, exposure to heterogeneous viewpoints could spur participation indirectly by facilitating learning about politics (McLeod et al., 1999; Scheufele et al., 2004; Scheufele, Nisbet, & Brossard, 2003). As H. Lee (2012) argued, the relative importance of these mechanisms may depend on a person’s political sophistication.

Extending on this line of research, recent studies on a similar topic aptly recognized that cross-cutting conversations can have nonmonolithic implications for different types of people. Scholars accordingly stressed that individual attributes can determine whether exposure to disagreement enhances or undermines political activism. Most pertinent to our focus, many found that the causal relationship between encountering dissimilar views and engagement in political life is conditional on various proxy indicators of attitude strength, such as political expertise (McClurg, 2006a), education (H. Lee, 2012), ideological strength (Wojcieszak et al., 2010), discussion willingness (Kim et al., 2011), and attitude certainty (Matthes, 2012). Echoing this view, Mutz (2006) also acknowledged that frequent contact with counterattitudinal positions can in fact mobilize strong opinion holders (cf. Jang, 2009).

Meanwhile, it is conceivable that citizens whose views are confirmed in their primary discussion network tend to have greater opinion strength. Evidence consistently demonstrates that an individual’s core tie network is highly homogenous, functioning as a basis for repeated exposure to agreeable beliefs and values (Granovetter, 1973; McPherson, Smith-Lovin, & Cook, 2001; Putnam, 2000). As such, agreement embedded in an immediate circle of social relations can serve as a crucial resource for consolidating one’s political views (Mutz, 2006; Price, Cappella, & Nir, 2002). Indeed, it has been shown that strong tie homogeneity is directly linked to the robustness of attitude. For example, Berelson et al. (1954) observed that those who belonged to a homogeneous primary discussion network exhibited a stronger political conviction in support of a candidate. In addition, studies indicate that frequent interactions with close associates are conducive to reinforcing group norms as well as individual opinions (Haythornthwaite, 2002; McPherson et al., 2001). In light of these considerations, it is arguable that the motivational role of cross-cutting discussion is contingent upon strong tie homogeneity, which can effectively augment attitude strength. This would mean that facing discrepant opinions is more likely to exert positive influence when individuals are plugged into a strong tie network of similar people.¹ Thus we first hypothesize (Hypothesis 1a [H1a]) that the relationship between cross-cutting discussion and political participation is conditional on strong tie homogeneity such that the positive association becomes stronger at a higher level of strong tie homogeneity.

Conceivably, a more convincing account of the positive interaction between strong tie homogeneity and cross-cutting discussion is reflected in the trend that those who are internally surrounded by likeminded others feel compelled to take meaningful political action when their shared norms are challenged by counterattitudinal contacts. This relationship may transpire because a homogenous discussion network is closely intertwined with attitude polarization (Huckfeldt, Mendez, et al., 2004), and encountered dissonant information can add more weight to this link. Polarized attitudes can then serve as an impetus for political participation, spurring higher voter turnout, more robust vote choice commitment, and earlier voting decision (Dodson, 2010; Stroud, 2011).

There are at least three reasons to believe that the influence of inherent homogeneity is accelerated by cross-cutting communication experience. First, counterattitudinal exposure, when pitted against supportive information from core ties, can constitute inoculation that could then serve to resist opposition and bolster the existing belief (McGuire, 1961; Nir, 2005; Pfau & Burgoon, 1988; Pfau et al., 1997). Second, the contrast of in-group consensus and experienced disagreement may propel the formation of collective identity, which can lead members of an internally coherent group to feel assured that “our” attitudes are more relevant than less familiar viewpoints (Mansbridge, 1983; Tajfel & Turner, 1986; Wojcieszak, 2011c). Finally, research on biased processing (Kunda, 1990; Nickerson, 1998; Petty & Cacioppo, 1990; Wojcieszak, 2011a; Wojcieszak & Price, 2010) postulates that citizens with strong prior predilections upon encountering heterogeneous views tend to assimilate consonant information without much scrutiny, but dismiss conflictual messages, thereby further strengthening their conviction (see also Gastil, Black, & Moscovitz, 2008). More germane to our interest, Wojcieszak (2010) showed that the association between active involvement in highly sympathetic online groups and resulting attitude polarization was amplified by exposure to dissimilar opinions from offline ties. Collectively, premises implicit in these theories suggest that people who are among likeminded others tend to refute dissonant messages and remain even more committed to their existing attitudes when they become cognizant of incongruent opinions (for a longitudinal study on this topic, see Binder, Dalrymple, Brossard, & Scheufele, 2009).

In terms of behavior consequences, Wojcieszak (2009) observed that the association between attachment to a decidedly uniform online group and political participation was stronger for those who confronted sizable disagreement in their discussion network. She opined, “[e]ncountering supporters and [emphasis in original] opponents, the analyzed ideologues may see that stakes are high and that their action might make a difference” (p. 578). On a somewhat different note, an analysis of a national survey by Nir (2005) suggested that less ambivalent individuals were propelled to make more decisive political action (i.e., early voting decision) in the context of cross-pressures, a result supported by Wojcieszak’s (2011b, 2011c) quasi-experimental work documenting that perceiving considerable disagreement drew political action from extreme opinion holders. Given that those who are plugged into likeminded others tend to have more robust opinions and thus less ambivalent attitudes, these studies lend further plausibility to the interplay of strong tie homogeneity and cross-cutting discussion in boosting political engagement. Moreover, Campbell and Kwak (2011) recently showed that the mobilizing influence of mobile communication was enhanced when strong tie likemindedness was coupled with a larger network size, which by extension may also denote more opportunities for cross-cutting experiences (Huckfeldt, Mendez, et al., 2004). In contrast, likemindedness did not facilitate the role of mobile-based discourse when network size was small, the result implying that strong tie coherence may not translate to participation under limited chances of exposure to disagreement. Inferred from the literature is that strong tie congeniality can provide a crucial resource for participation, but cross-cutting deliberative experience may prove to be a necessary catalyst prompting its expenditure.

Taken together, theoretical arguments discussed above suggest that individuals entrenched in a highly congenial group may be induced to reinforce their preexisting positions when they encounter substantial disagreement, and empirical evidence further indicates that they can take meaningful actions in such circumstances of cross-pressure. In line with these considerations, we hypothesize (Hypothesis 1b [H1b]) that the relationship between strong tie homogeneity and political participation is conditional on cross-cutting discussion such that the positive relationship becomes stronger at a higher level of cross-cutting discussion.

The Link Between Cross-Cutting Discussion and Deliberative Democracy as Moderated by Strong Tie Homogeneity

Advocates of deliberative democracy contend that encountering disagreement in political conversation is a key ingredient of the public sphere, given that exposure to diverse perspectives can facilitate awareness of legitimate rationales for opposing viewpoints and political tolerance (Mutz, 2002b, 2006; Mutz & Mondak, 2006; Price et al., 2002). However, these outcomes alone may not form a sufficient condition for ideal deliberative democracy; upon realizing discordant views from their cross-cutting experiences, many may opt out of future interaction with nonlikeminded others. In fact, the central tenet of the spiral of silence theory (Noelle-Neumann, 1993) asserts that citizens who experience substantial disagreement by virtue of their minority perception tend to withhold the expression of dissent. Echoing this concern, Wojcieszak and Price (2012) recently demonstrated that both perceived and actual disagreement could discourage people from voicing their opinions. As Park (2000) keenly points out, deliberation requires not only listening to but also speaking with diverse others (see also Burkhalter, Gastil, & Kelshaw, 2002). By and large, the above mentioned consequences reflect the listening dimension, as they connote an individual’s willingness to permit and put up with others’ expression of dissimilar views. Surprisingly, it is left largely unexamined whether cross-cutting experiences can lead individuals to stay enthusiastic about speaking with others who may have alternative perspectives. Motivated by this vacuum, the present study proposes “dialogic openness” as an additional but equally important component of deliberative democracy (Campbell & Kwak, 2012). The concept is designed to tap an ongoing deliberative orientation, and it specifically refers to a citizen’s propensity for public dialogue with lesser known others.

Here we seek to examine how our indicators of agreement and disagreement are linked to this benchmark of deliberative democracy. However, the extant literature offers only mixed findings about their interrelatedness and reciprocity. First, the causal relationship between cross-cutting discussion and dialogic openness remains unclear. That is, cross-cutting discussion can serve as a vital source of tolerance (Mutz, 2002b, 2006), which is then closely associated with openness (Marcus, Sullivan, Theiss-Morse, & Wood, 1995). In reverse, a tendency toward openness can afford more cross-cutting experiences (Mutz & Mondak, 2006). Nonetheless, it could well be that more frequent exposure to dissimilar viewpoints may harness people’s open tendencies to speak with unfamiliar others (Noelle-Neumann, 1993). Second, predictions pertaining to the connectedness between strong tie homogeneity and dialogic openness are no less complicated. Some theorists would argue that repeated conversations with likeminded others in a narrow circle could undermine open attitudes toward public deliberation (Calhoun, 1988; Mansbridge & Morris, 2001; Schudson, 1984). This negative relationship may operate backward through what Mutz (2006) called “affective mechanism” (p. 67) whereby greater openness may result in decreased homogeneity in a close network. Ironically though, it is those plugged in an immediate supportive group who are more likely to engage in deliberation with heterogeneous others, given that characteristics of a core tie relationship such as intimacy, passion, and conviction can empower them to prevail over the risks of embarrassment (Mansbridge, 1983; Schudson, 1984). In this case, strong tie homogeneity may boost openness.

Although the main relationships of dialogic openness with cross-cutting discussion and strong tie homogeneity remain largely ambiguous, there is some literature to suggest that cross-cutting discussion can work in concert with homogenous ingredients to predict increased deliberative tendencies. In particular, scholars of spiral of silence (e.g., Glynn & McLeod, 1984; Lasorsa, 1991; Neuwirth, Frederick, & Mayo, 2007) have well documented that people with strong attitudes (e.g., hard cores) are inclined to express their predilections even when they are aware of sizable disagreement (i.e., unfavorable opinion climate). More directly related to our focus, Matthes, Rios Morrison, and Schemer (2010) found that an unfavorable opinion climate—by extension, perceived disagreement (see, for example, Wojcieszak et al., 2010)—discouraged deliberation with less similar others only among people with weak attitude certainty. Conversely, the perception of minority opinion climate spurred those who have a firm conviction to interact with heterogeneous others. As such, an extended notion from the spiral of silence theory strongly implies that whether encountering disagreement dampens or fosters open deliberative tendencies to talk with those who may have unfamiliar viewpoints is conditional upon attitude strength.

Meanwhile, the above reviewed theoretical propositions including a classic model of voting behavior (Berelson et al., 1954) and a social network theory (McPherson et al., 2001) lend credence to the link between likemindedness in an immediate circle and opinion strength. In addition, recent theoretical developments on new media illustrate that intensive contacts with strong interpersonal ties can entail solidified attitudes (Gergen, 2008; Haythornthwaite, 2002; Norris, 2002). Thus there are considerable theoretical grounds to expect that strong tie homogeneity could significantly enhance the effects of cross-cutting discussion on deliberative orientation by providing courage and resources to interact with those holding less familiar predispositions (see also Mansbridge, 1983; Schudson, 1984).

There is also empirical evidence that could attest to the interactive role of cross-cutting discussion and strong tie homogeneity in fostering deliberative democracy. For example, Campbell and Kwak (2012) found that the combination of larger network size—by extension, greater chances of cross-cutting discussions (see, for example, Huckfeldt, Johnson, & Sprague, 2004; Mutz, 2006)—and strong tie likemindedness facilitated the civic influence of mobile communication on dialogic openness. Campbell and Kwak (2012) reasoned, “having more trusted sources accessible for anytime-anywhere affirmation solidifies views by making them more salient, thus fostering confidence in those views and the motivation to discuss them with [lesser known] others” (p. 276). This finding may also mean that the relationship between exposure to disagreement and deliberative orientation can depend upon strong tie homogeneity, which can consolidate individual attitudes. Thus, drawing on an extended notion of spiral of silence (Glynn & McLeod, 1984; Lasorsa, 1991; Matthes et al., 2010; Neuwirth et al., 2007), various theoretical propositions linking in-group likemindeness and attitude strength (Berelson et al., 1954; Gergen, 2008; Haythornthwaite, 2002; McPherson et al., 2001; Norris, 2002), and empirical evidence speaking to the collaborative workings of cross-cutting discussion and strong tie homogeneity in promotion of deliberative democracy (Campbell & Kwak, 2012), we hypothesize (Hypothesis 2 [H2]) that the relationship between cross-cutting discussion and dialogic openness is conditional on strong tie homogeneity such that the positive association becomes stronger at a higher level of strong tie homogeneity.

Internal Efficacy and the Prospect of Democracy

We have contended thus far that cross-cutting discussion and strong tie homogeneity may work in concert to enhance an individual’s ongoing deliberative orientation and motivation to participate. Yet it has been left unexamined how these two outcomes are tied to each other. In this vein, Mutz (2002b, 2006) proposed two possible processes wherein the deliberative benefits such as tolerance and awareness of oppositional rationales can yield unwanted outcomes in terms of participatory democracy: social accountability (i.e., avoidance of interconflict) and ambivalence (i.e., result from intraconflict). Her findings showed that people’s tendency to avoid conflict and ambivalent attitudes toward political options were the key factors hindering the link between deliberation and participation. Though we acknowledge that conflict avoidance and ambivalence provide convincing explanations for the demobilizing role of deliberation, we believe that not all individuals are likely to fall victim of these two processes.

In particular, we expect that neither conflict situation (i.e., inter or intra) would pose a serious hurdle to those high in internal political efficacy. After all, internal political efficacy, which refers to “beliefs about one’s own competence to understand, and to participate effectively in, politics” (Niemi, Craig, & Mattei, 1991, p. 1407), should provide enough confidence to handle either intra- or interconflict. First, politically efficacious individuals would not be hindered by social accountability because they are less prone to conflict avoidance (Ulbig & Funk, 1999). To the contrary, they would be inclined to confront a conflict since their self-appraisal of competence often excels the perceived threat (Bandura, 1997; Rimal & Real, 2003; Witte & Allen, 2000).² Indeed, a wealth of evidence indicates that politically efficacious individuals tend to cope with adverse situations that could elicit unpleasant responses (Brader, 2006; Rudolph, Gangl, & Stevens, 2000; Schuck & de Vreese, 2012; Valentino, Gregorowicz, & Groenendyk, 2009). Thus conflict avoidance is a less tenable explanation for the deleterious influence of deliberation among those high in internal efficacy. Second, efficacious individuals are less likely to experience ambivalence (Lee & Chan, 2009), perhaps because they possess more advanced cognitive ability (Chen, Casper, & Cortina, 2001) to make sense of diverse perspectives. Furthermore, Mutz (2006) found that the association between encountering diverse perspectives and taking an ambivalent stance was confined within those who were conflict avoidant. Then, efficacious individuals would be less hampered by ambivalence owing to their lower tendency to avoid conflict. In short, efficacious citizens appear to have the perceived competence that can empower them to overcome the less desirable consequences of remaining open to deliberate with unfamiliar others. Along this line, we expect (Hypothesis 3 [H3]) that internal political efficacy positively moderates the relationship between dialogic openness and political participation such that the positive association becomes stronger at a higher level of internal efficacy.

Integrating core hypotheses introduced thus far leads us to propose an overarching model of democracy depicted in Figure 1. First, while merging H1a and H2 based on the literature indicating that deliberation tends to precede participation (Kwak et al., 2005; Mutz, 2006; Scheufele, 2002; Wojcieszak et al., 2010), we suggest that counterattitudinal exposure paired with internal cohesiveness is related to dialogic openness, which is then linked to political participation. This also means that the role of the interaction between exposure to disagreement and agreement in accounting for participation is mediated by openness. On the other hand, H3 postulates that the direct relationship between openness and participation depends on internal efficacy. By adding this component to the mediating pattern above, we envision that internal efficacy moderates the overall indirect process in which the joint workings of cross-cutting discussion and strong tie homogeneity are tied to active engagement in political life via the conduit of an enduring tendency to deliberate with lesser known others. As summarized in Figure 1, we are in effect putting forth a double conditional indirect relationship (see, for example, Schuck & de Vreese, 2012), which posits that the indirect association between cross-cutting discussion and political participation through dialogic openness is conditional upon strong tie homogeneity and internal efficacy.

Figure 1.

Conceptual framework.

Method

Sample

Data for this study come from a national mail survey that was conducted during the 2008 U.S. presidential campaign. The research firm Synovate collected the data. A large body of individuals was contacted via mail and asked to indicate the willingness to participate in mail, telephone, or online surveys; individuals who were interested were asked to provide basic demographic information. A balanced sample was then drawn from among more than 500,000 people who agreed to participate in the prerecruited “mail panel.” In order to ensure representativeness, the sample was drawn to reflect demographic distributions within the five U.S. Census divisions of household income, population density, panel member’s age, gender, and region. This stratified quota sampling method was used to select 2,218 mail survey respondents. There were 66 undeliverable surveys and 1,018 usable responses were received, representing a cooperation rate of 45.9% (COOP1, AAPOR). This stratified quota sampling method largely differs from more conventional probability sample procedures yet produces highly comparable data (Putnam, 2000; Putnam & Yonish, 1999).

Demographic characteristics of the current sample resemble the profiles of the national population figures reported in U.S. Census Bureau’s 2007 American Community Survey (ACS), with respect to education attainment (the median education level for those 25 or older in both data sets is some college), household income (the median in the ACS and this study is US$50,000-US$74,999 and US$50,000-US$59,999, respectively), and age (the median age for those 18 or older in the ASC and this study is 45-54 and 53, respectively). However, there is a greater percentage of male respondents in the sample (59.7%) than in the ACS (48.6%).

Measures

Political participation

Respondents were asked about four types of involvement in traditional forms of political participation: attending a political meeting, rally, or speech; working for a candidate or a party; contacting a public official or a political party; and contributing money to a candidate or a political party. Respondents reported the frequency of involvement in each type of participation in the past month on a 7-point scale, ranging from none in the last month to everyday. The responses were averaged to form an additive index (M = 1.22, SD = .53, Cronbach’s α = .74).

Dialogic openness

Utilizing the measure developed by Campbell and Kwak (2012), the extent of dialogic openness (i.e., willingness to open political dialogue with those outside of one’s strong ties) was measured with an item asking respondents to state how much they agreed with the following statement, “I enjoy discussing political issues with others I don’t know very well,” on a 6-point scale, ranging from definitely disagree to definitely agree (M = 2.43, SD = 1.43).

Strong tie homogeneity

Two items were used to tap strong tie homogeneity. Respondents were asked to report how many of the people in their network of strong personal ties shared their political views and how many supported the same presidential candidate as the respondent. The responses were recorded on a 5-point scale, consisting of none, only some, about half, most of them, and all. The responses were averaged to form an index (M = 3.24, SD = 1.03, interitem r = .75).

Cross-cutting discussion

To measure cross-cutting discussion, respondents were asked to indicate how often in the past month they were engaged in the following two types of discursive behavior: talking to other who don’t support the candidate they favor; and having a conversation about politics or social issues that involve disagreement. Respondents reported the frequency of involvement in each type of behavior on a 7-point scale, ranging from none in the last month to everyday. The responses were averaged to form an additive index (M = 2.40, SD = 1.44, interitem r = .64).

Control variables

As control variables, this study included age, gender, education, household income, political interest, internal efficacy, party identification strength, newspaper use, television news use (local and national programs), and general discussion frequency. To measure political interest, respondents were asked to state how much they agreed with the following statement, “I am very interested in politics,” on a 6-point scale, ranging from definitely disagree to definitely agree (M = 3.47, SD = 1.65). Internal political efficacy was measured with three items asking respondents to state how much they agreed with the following statements, “I know a lot about the policies of the presidential candidate I support,” “I know a lot about the policies of political candidates I don’t support,” and “people like me have considerable influence on politics.” Reponses were measured on a 6-point scale, ranging from definitely disagree to definitely agree, and they were averaged to form an index (M = 3.31, SD = 1.15, Cronbach’s α = .70). For mass media use variables, respondents were separately asked, using a 5-point scale, about how often in the past month they used daily newspapers (M = 3.59, SD = 1.42), national nightly news, and local television news programs. The two television items were combined as an index (M = 3.57, SD = 1.17, interitem r = .57). In addition, our analysis controlled for general discussion frequency, given that adding this to a statistical model may considerably alter the pattern of the relationship between cross-cutting discussion and participation (Eveland & Hively, 2009). To measure general discussion frequency, respondents were asked to report how often in the past month they discussed politics with neighbors, friends, and family on a 7-point scale, ranging from none in the last month to everyday. Responses were averaged to form an index (M = 2.81, SD = 1.36, Cronbach’s α = .79).

Interaction terms

To carry out analyses corresponding to H1 and H2, this study created an interaction term between cross-cutting discussion and strong tie homogeneity. In addition, to investigate H3, this study created an interaction term between dialogic openness and internal efficacy. To reduce potential problems with multicollinearity between interaction terms and their components, all the component variables were standardized before the formation of the interaction terms (Cronbach, 1987; Jaccard, Turrisi, & Wan, 1990; Kwak, 1999).

Results

The first column in Table 1 (Model 1) reports the results from an OLS regression analysis predicting political participation from cross-cutting discussion and strong tie homogeneity, while accounting for a host of control variables. It is shown that cross-cutting discussion predicted a significant increase in participation (β = .121, p < .01), whereas the corresponding link for strong tie homogeneity was not significant (β = .011, ns). Meanwhile, the second column (Model 2) summarizes the results when the interaction was added, and it displays the significant interactive role of these two discussion features in predicting increased participation (β = .103, p < .01). The pattern of the relationship was plotted in Figure 2. The left panel presents support for H1a, clearly demonstrating that the relationship between cross-cutting discussion and political participation is noticeably more positive at a higher level of strong tie homogeneity. The right panel summarizes the same interactive relationship, but it displays the pattern when strong tie homogeneity is the independent variable and cross-cutting discussion is the moderator (H1b). It can be seen that the relationship between strong tie homogeneity and political participation is conditional on cross-cutting discussion such that the positive relationship becomes clearer as the frequency of cross-cutting discussion increases. Interestingly, the dotted line here signifies that cohesiveness of a primary network is less likely to be associated with greater political mobilization if it is not coupled with frequent exposure to disagreement.

Table 1.

OLS Predicting Political Participation and Dialogic Openness From Cross-Cutting Discussion, Strong Tie Homogeneity, and Their Interaction.

	Political participation		Dialogic openness
	Model 1	Model 2	Model 3	Model 4
Control variables
Age	.044	.046	.029	.030
Gender (high: female)	–.045	–.042	–.139***	–.137***
Education	.093**	.099**	.009	.013
Household income	–.044	–.043	–.036	–.036
Political interest	.046	.057	.194***	.202***
Internal efficacy	.156***	.159***	.121**	.124**
Party identification strength	.006	.005	–.015	–.017
Newspaper news use	.089**	.084_**	–.017	–.021
Television news use	.021	.016	.028	.024
General discussion frequency	.128**	.127**	.023	.023
R² (%)	16.6		21.1
Discussion heterogeneity variables
Cross-cutting discussion	.121**	.117**	.261***	.258***
Strong tie homogeneity	.011	.031	.062*	.077*
Inc. R² (%)	.9		4.0
Interaction term
Cross-cutting discussion × Strong tie homogeneity	—	.103**	—	.083**
Inc. R² (%)	—	1.0	—	.6
Total R² (%)	17.5	18.5	25.1	25.7
N	901		897

Note. Entries are standardized final regression coefficients.

p < .05. **p < .01. ***p < .001.

Figure 2.

The interplay of cross-cutting discussion and strong tie homogeneity in predicting political participation.

To probe the implications for deliberative democracy, we regressed dialogic openness on cross-cutting discussion and strong tie homogeneity, and then added the interaction term to the model. The findings summarized in the third column of Table 1 (Model 3) show that both cross-cutting discussion (β = .261, p < .001) and strong tie homogeneity (β = .062, p < .05) were significantly associated with greater dialogic openness. More importantly, the results in the last column (Model 4) lend support for H2, illustrating that the association between cross-cutting discussion and dialogical openness is conditional on strong tie homogeneity (β = .083, p < .01). The pattern of this interactive association was plotted in Figure 3. The steeper solid line indicates that the size of the association between cross-cutting discussion and dialogic openness is stronger when individuals are surrounded by analogous others in their intimate network.

Figure 3.

The link between cross-cutting discussion and dialogic openness that is conditional on strong tie homogeneity.

To assess H3, we ran a final criterion variable model predicting political participation from all of the aforementioned exogenous variables, while adding dialogic openness and its interaction term with internal efficacy as additional predictors. The abridged results are reported in Table 2. The first column (Model 5) shows that dialogic openness did not independently predict participation (β = .039, ns). Meanwhile, the second column (Model 6) demonstrates that internal efficacy played a positive moderating role between openness and participation (β = .128, p < .001). As such, these results support H3, illustrating the facilitating link between dialogic openness and political participation manifests clearly at a high level of efficacy. On the other hand, the comparable relationship at the low efficacy level is negative. These contrasting patterns of the relationship as a function of efficacy are well displayed in Figure 4.

Table 2.

Two-Way Interactive Relationship Between Dialogic Openness and Internal Efficacy in Predicting Political Participation.

	Political participation
	Model 5	Model 6
Prior blocks (R², %)	18.5
Dialogic openness	.039	.004
Inc. R² (%)	.1
Dialogic openness × Internal efficacy	—	.128***
Inc. R² (%)	—	1.6
Total R² (%)	18.6	20.2
N	897

Note. Prior blocks include age, gender, education, household income, political interest, internal efficacy (in Model 5: β = .154, p < .001; in Model 6: β = .170, p < .001), party identification strength, newspaper news use, television news use, general discussion frequency, cross-cutting discussion, strong tie homogeneity, and the interaction between cross-cutting discussion and strong tie homogeneity; Entries are standardized final regression coefficients.

p < .05. **p < .01. ***p < .001.

Figure 4.

The relationship between dialogic openness and political participation as moderated by internal efficacy.

Collectively, the findings reported thus far lend preliminary support for the overarching model of democracy that introduced a double conditional indirect relationship. Specifically, the results have evidenced the two conditional relationships; cross-cutting discussion interacted with strong tie homogeneity to predict increased dialogic openness, whereas the association between dialogic openness and political participation was significantly moderated by internal efficacy. These two conditional relationships, when adjoined, are illustrative of the double conditional indirect process embedded in our theoretical framework. With these results, though, we remain less than fully confident that the indirect path from the interplay of cross-cutting discussion and strong tie homogeneity via dialogic openness to political participation is indeed statistically significant. Furthermore, little is known about how efficacy moderates this overall indirect link. Thus, a more advanced analysis of double conditional indirect association is in order.

We constructed an analytical framework based on the coefficients for cross-cutting discussion, strong tie homogeneity, and their interaction term when predicting dialogic openness summarized in the last column of Table 1 (Model 4) and the coefficients for dialogic openness, internal efficacy, and their interaction term when predicting political participation reported in the second column of Table 2 (Model 6). Figure 5 outlines our statistical model to examine the proposed double conditional indirect association, which is adapted from PROCESS Model 22 in Hayes (2012). The central purpose of this model is to examine how the joint workings of cross-cutting discussion (X) and strong tie homogeneity (W) are related to political participation (Y) via dialogic openness (M), and further to assess the extent to which this overall indirect path is conditional upon internal efficacy (V). In effect, it is equivalent to estimating the indirect association of cross-cutting discussion with political participation through openness orientation that is conditional upon strong tie homogeneity and internal efficacy.³ This can be captured by the following: (a₁ + a₃W) (b₁ + b₃V). As a result, the indirect association between cross-cutting discussion and political participation through dialogic openness could vary across different levels of strong tie homogeneity and internal efficacy. To assess this, we computed 95% bias-corrected confidence intervals (95% bc CIs) on the basis of 5,000 bootstrap samples using Mplus.

Figure 5.

Statistical model to estimate the indirect association of X with Y through M that is doubly conditional upon W and V adapted from PROCESS Model 22 in Hayes (2012).

Table 3 summarizes the change in the degree of indirect association as dual functions of internal efficacy and strong tie homogeneity. It can be seen that when the level of efficacy is high, the indirect relationship between cross-cutting discussion and political participation via dialogic openness is significantly positive regardless of the degree of strong tie homogeneity. Furthermore, as the level of inner agreement among these efficacious individuals increases, the positive indirect path becomes even stronger, with the maximum reached when strong tie homogeneity and internal efficacy are both high (β = .045, SE = .021; 95% bc CI: [.012, .096]). On the flip side, for those who are low in efficacy, more frequent cross-cutting discussion yields a negative indirect link to participation through dialogic openness, and this undermining pattern seems further exacerbated by core tie congruency. In fact, cross-cutting discussion shows the most adverse relationship with political participation among less efficacious individuals who are mostly surrounded by likeminded others (β = –.042, SE = .019; 95% bc CI: [–.088, –.014]). Taken together, the findings show that internal political efficacy positively moderates the indirect link running from the interaction of cross-cutting discussion and strong tie homogeneity leading to political participation via the conduit of dialogic openness.

Table 3.

Double Conditional Indirect Relationships of Cross-Cutting Discussion (X) With Political Participation (Y) Through Dialogic Openness at Specific Levels of Strong Tie Homogeneity (W) and Internal Efficacy (V).

Internal efficacy (V)	Strong tie homogeneity (W)	IND	SE	BC 5000 BOOT
Internal efficacy (V)	Strong tie homogeneity (W)	IND	SE	LL95	UL95
Low	Low	–.022	.011	–.055	–.007
	Mean	–.032	.014	–.067	–.011
	High	–.042	.019	–.088	–.014
Mean	Low	.001	.005	–.009	.013
	Mean	.001	.008	–.014	.017
	High	.001	.010	–.019	.022
High	Low	.023	.013	.005	.060
	Mean	.034	.016	.009	.073
	High	.045	.021	.012	.096

Note. Entries (INDs) are standardized coefficients capturing specific indirect relationships at different values of internal efficacy (V) and strong tie homogeneity (W); LL95 and UL95 are lower and upper bounds of 95% bias-corrected confidence intervals based on the bootstrapping of 5,000 samples; Low signifies values one standard deviation below the mean, whereas High signifies values one standard deviation above the mean.

Discussion

We aimed to shed light on the highly publicized democracy dilemma by assessing the processes wherein the interplay of cross-cutting discussion and strong tie homogeneity is simultaneously associated with the outcomes of deliberative and participatory democracy. The findings presented here provide manifold implications for the literature that has examined the effects of cross-cutting exposure on the health of a democratic society.

First, our results significantly add to understanding of the debate surrounding the influence of encountering disagreement on political participation by demonstrating that strong tie homogeneity and cross-cutting discussion work in concert to predict greater mobilization. This relationship may transpire through the indirect mechanism of attitude polarization.⁴ The interactive effect of these two discussion characteristics resonates well with earlier findings that the combination of agreement and dissent can foster a citizen’s stronger conviction and motivation to take an action. For example, Wojcieszak observed that engagement in a highly likeminded online group in conjunction with disagreement in a close offline network of family, friends, and acquaintances predicted increased polarization (2010) and participation (2009). On a somewhat different note, McClurg (2006b) found that the inverse relationship between disagreement within a dyad and political participation did not manifest among those who were entrenched in a favorable neighborhood context. These studies are important scholarly endeavors to examine the interactive role of homogeneity and heterogeneity in promotion of participatory democracy. However, our research differs from these previous attempts with regard to the locus of disagreement. Though of differing foci, these studies examined disagreement that was mainly inherent in an immediate circle.⁵ Yet we believe that exposure to dissimilar views frequently takes place outside of a closely knit group (Mutz, 2006; Mutz & Mondak, 2006), and our measure of cross-cutting experiences reflects a more realistic communication environment wherein people encounter disagreeing others either internally or externally.

Second, the present study offers an important direction for future scholarship on deliberative democracy. In particular, we employed an often ignored component of deliberative democracy: dialogic openness (Campbell & Kwak, 2012). By and large, our results are consistent with prior work highlighting deliberative benefits of cross-cutting discussion such as awareness of legitimate rationales for opposing viewpoints and political tolerance (Mutz, 2002b, 2006; Nir, 2005; Price et al., 2002). Yet our study extends this line of research and suggests that strong tie homogeneity could be an important catalyst of ongoing deliberative orientation to converse with others who have unfamiliar or contrasting views. Similar to hard core individuals with strong convictions who would not fear expressing their dissent when they face the opinion climate countering their views, those who belong to a strong supportive network are more likely to continue their deliberative tendencies (see also Milgram, 1974). Thus strong tie homogeneity seems to provide resources necessary for repeated interaction with incongruent others.

Some may dismiss it as a discomforting situation that people who have more ardent views owing to their homogeneous strong ties merely purport to uphold their own beliefs by talking with less similar others. Nonetheless, just as selective search for opinion-reinforcing information may coincide with greater contact with dissimilar views (Garrett, 2009; Knobloch-Westerwick & Meng, 2009), those who are determined to carry through their conviction during deliberation are in due course more likely to hear alternative perspectives and can ultimately understand nonlikeminded ideas without compromising their own stance. Supporting this line of reasoning, Thorson (2012) recently demonstrated that engagement in persuasion attempts to proselytize others were linked to some tangible deliberative benefits including articulation of own as well as oppositional rationales and campaign learning. We believe that these individuals who are interested in advancing their opinions during conversations would fulfill the goal of deliberative democracy much more effectively than those who are predisposed to remain silent and refrain from controversies perhaps due to the lack of firm views shared with close ties.

The third implication of this study directly pertains to the democracy dilemma we started with. Although the issue became widely known since Mutz (2002a, 2002b, 2006) empirically showed puzzling consequences of cross-cutting discussion, the overarching quandary can trace back to early theorists who contemplated an ideal balance between unity and diversity. For instance, Berelson (1952) argued that a healthy democratic system requires “enough cleavage to stimulate debate and action, [and] enough consensus to hold the society together even under strain.” (p. 328). Furthermore, Berelson et al. (1954) keenly noted that extreme consensus may refer to a totalitarian system that is highly restrictive of citizen liberty, whereas excessive incongruity could destroy the cohesiveness of the society, ultimately yielding political apathy. Echoing these early remarks, Nir (2005) concluded, “[f]ar from being detrimental to a healthy democratic order, however, a balance [emphasis in original] between political cleavage and consensus may be the precondition for it” (p. 439). In accordance with these considerations, our findings highlight that the mix of internal consensus and disagreement experienced in the form of cross-cutting discussion can indeed set the ideal stage for deliberative and participatory democracy. That is, the existence of a highly sympathetic strong tie network could empower citizens to glean benefits from communication experiences that cut across lines of political difference. On the practical front, this study indicates that for those who are equipped with the essential resources, talking about politics with others—indeed, arguing about politics—can actually enhance one’s level of engagement in the democratic process. The key contribution here is better understanding of the conditions under which this constructive outcome is facilitated.

Building on the results of the current research, we envision a dual process of democracy in which two types of democracy, with the proper recipe of homogeneous and heterogeneous ingredients, could develop in a parallel direction. First, individuals, through workings of inoculation, collective identity, or biased processing, may be induced to reinforce their common attitudes within their close community and act upon them, when they are confronted with counterattitudinal perspectives. Second, continuous deliberation with lesser known others may in the end form an affiliation among people of dissimilar beliefs and norms, which can eventually extend the scope of a closely held communication network while propelling mutual understanding and tolerance—that is, affective mechanism (Mutz, 2002b, 2006). From a normative standpoint, the first process may raise alarm, given that it may plainly refer to increased participation from the polarized public (see, for example, Abramowitz & Saunders, 2005). However, we believe that this concern can be alleviated by the second process. Our findings clearly indicate that the same pairing of strong tie homogeneity and cross-cutting discussion that could encourage attitude polarization and political participation can also facilitate a citizen’s tendency to converse with others who may hold alternative opinions. Although the interplay of internal homogeneity and counterattitudinal exposure may reinforce previously held attitudes, it can also promote more thoughtful consideration of dissimilar views. In short, we contend that the aforementioned dual process of democracy can facilitate the emergence of ideal citizens in a pluralistic society as defined by Huckfeldt, Mendez, et al. (2004)—that is, “tolerant gladiators—combatants with the capacity to recognize and respect the rights and responsibilities of their political adversaries” (p. 91).

However, it is worth noting that benefits from this dual process of democracy may be unequally distributed across different segments of population. While acknowledging this possible disparity in gains, we aimed to ascertain a key condition under which the deliberative outcome can translate to meaningful political action. To that end, we examined how internal efficacy moderates the overall indirect process in which the joint workings of cross-cutting discussion and strong tie homogeneity are tied to political participation via the conduit of dialogic openness. The results showed that the indirect association was positive among those high in efficacy. This indicates that, at least among efficacious citizens, cross-cutting discussion worked in coordination with strong tie homogeneity to predict boosted political participation indirectly through enhanced dialogic openness. Furthermore, as the level of inner agreement among these efficacious individuals increased, the positive indirect path from cross-cutting discussion to participation via dialogic openness became even more apparent. These findings signify that the two democratic outcomes that the combination of counterattitudinal exposure and internal consensus can produce are likely to coexist among efficacious individuals.

Conversely, the corresponding indirect link for those low in efficacy was negative. This negative indirect link denotes that dialogic openness encouraged by the collaborative workings of cross-cutting discussion and strong tie homogeneity may result in depressed participation. That is, inefficacious people’s cross-cutting discussion in conjunction with homogeneity within their immediate circle can entail more open tendencies to converse with others of less familiar views, but such an ongoing deliberative orientation may ultimately coincide with less active citizenry. In this vein, it can be inferred that citizens low in efficacy likely lack civic competence to convert the outcomes of open deliberative tendencies into meaningful political behaviors. Moreover, strong tie homogeneity fueled the negative association, with the most adverse relationship observed among individuals whose strong tie network is highly homogeneous. This implies that sustained exposure to disagreement coupled with recurring confirmation of in-group beliefs would merely create additional confusion among less efficacious citizens. Being unable to digest a broader scope of dissimilar viewpoints that are available through their likeminded circle of contact and prolonged deliberative practices, they may be more susceptible to some of the well-recognized hindering mechanisms of deliberation, such as conflict avoidance and ambivalence, which can consequently lead to political inertia. In sum, the current research illustrates that for people with low internal efficacy, deliberative benefits can outweigh participatory outcomes as they seem incapable of translating more open deliberative orientation into greater participation and that this gap tends to widen when they are continuously under cross-pressures of contrasting perspectives.

The present study is not without limitations. First, because the survey data are cross-sectional in nature, patterns observed here should be qualified as correlation and they are not sufficient for causal inferences. Nonetheless, treating deliberation as the antecedent of participation is consistent with a large body of work on network heterogeneity (e.g., Kwak et al., 2005; Mutz, 2006; Wojcieszak et al., 2010). Admittedly, the most troubling causal inferences assumed in this study involve individual links from cross-cutting discussion and strong tie homogeneity to dialogic openness. As discussed earlier, there are also reasons to expect that propensity for deliberation can affect the frequency of counterattitudinal exposure or the degree of core tie similarity. However, the causal relationship of dialogic openness with cross-cutting discussion or strong tie homogeneity is less likely to be one-sided, instead creating a spiral relationship (Slater, 2007). In this vein, future studies can assess, with longitudinal data, the reciprocal relationship that dialogic openness may hold respectively with cross-cutting discussion and strong tie homogeneity.

Another limitation of this study stems from the single-item index of dialogic openness. With only one item, we were less likely to fully capture an individual’s ongoing deliberative orientation. Moreover, although we employed this measure motivated by the lack of attention to the sustained engagement in deliberation, our analyses did not incorporate more conventional indicators of deliberative democracy such as awareness of legitimate rationales for opposing viewpoints and political tolerance. However, these more frequently explored outcomes do not fully reflect the prospect of deliberative democracy. After all, deliberative democracy is in peril to the extent that people avoid future deliberation with nonlikeminded others after they become cognizant and tolerant of oppositional ideas. Accordingly, we urge the field to develop a more comprehensive scale of deliberative benefits that can effectively encapsulate a citizen’s ongoing openness to alternative views and actual cognitive outcomes of deliberation. Along this line, future studies may employ multidimensional scaling that can exhibit multifaceted nature of deliberation (Burkhalter et al., 2002).

Finally, it is noteworthy that our findings never showed an inverse relationship between cross-cutting discussion and participation, seemingly contracting the Mutz’s (2002a, 2006) well-recognized conclusion. We first emphasize that the inconsistency does not result from differing definitions of heterogeneity, given that the measures adopted in both studies represent the nonlikeminded element (i.e., dangerous discussion) as opposed to the diversity aspect (Eveland & Hively, 2009). Rather, the difference between Mutz’s and our study seems to denote further subattributes of nonlikemindedness: the extent of dissimilarity and the frequency of exposure to oppositional views. Admittedly, our measure of cross-cutting discussion did not express the degree of dissimilarity (e.g., how severe disagreement was). Perhaps for this reason, studies treating the frequency of discussion as an indicator of network heterogeneity yielded positive relationships with participation (Kwak et al., 2005; Scheufele et al., 2004, 2006). On the other hand, Mutz’s measure tapped closely the degree of nonlikemindeness. However, there was no independent frequency measure that could encapsulate disagreement arising beyond those with whom the respondents regularly discuss politics. Consequently, her measure was suitable to capture more chronic exposure to disagreement, yet it largely ignored inadvertent encounters, which many including Mutz herself deemed as more common forms of cross-cutting experiences (Brundidge, 2010; Wojcieszak & Mutz, 2009). Given the strengths and weaknesses of each approach, future studies can incorporate separate measures of the degree of dissimilarity and the frequency of cross-cutting conversations, and examine the respective effects of persistent heterogeneity and incidental exposure.

In closing, we provide an important suggestion for future scholarly endeavors to assess the influence of cross-cutting communication experiences in the context of the emerging information environment. We have thus far argued that the combination of solid inner consensus and frequent contacts with dissimilar views best contributes to deliberative and participatory democracy. Then, the new information environment seems to paint a promising picture for the health of a democratic society since it can accelerate the pairing of internal homogeneity and exposure to divergent perspectives. Indeed, some communication technologies that afford a high degree of control (e.g., voice calling, texting, and instant messaging) can reinforce coherence with core ties, whereas mass media types of behaviors that are less conditioned by social selectivity (e.g., informational use of news sites, portals, and search engines) can increase the likelihood of coming across heterogeneous viewpoints. Therefore, efforts to further develop this line of inquiry will benefit by exploring how various communication technologies, some of which can buttress strong tie cohesiveness and others offering more opportunities to encounter less familiar stances, interact with one another in promotion (or discouragement) of deliberative and participatory democracy.

Footnotes

Acknowledgements

The authors wish to thank the editor, Professor Michael E. Roloff, and the three anonymous reviewers for their insightful feedback and comments on this work.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the authorship and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by The Howard R. Marsh Endowment and The Constance F. and Arnold C. Pohs Research and Technology Endowment at the University of Michigan.

Notes

Author Biographies

Hoon Lee (PhD, Communication, University of Michigan) is a postdoctoral research fellow at the Nam Center for Korean Studies at the University of Michigan. His current research investigates social impacts of entertainment media and new communication technologies in cross-national and cross-cultural contexts.

Nojin Kwak (PhD, Communication, University of Wisconsin-Madison) is an associate professor in the Department of Communication Studies at the University of Michigan and the director of the Nam Center for Korean Studies at the University of Michigan. His research centers on the role of communication media in civic and political engagement.

Scott W. Campbell (PhD, Communication, University of Kansas) is associate professor of Communication Studies and Pohs Professor of Telecommunications at the University of Michigan. His research examines the social implications of new media, with an emphasis on mobile telephony.

References

Abramowitz

Saunders

(2005). Why can’t we all just get along? The reality of polarized America. Forum, 3(2), 1-22.

Bandura

(1997). Self efficacy: The exercise of control. New York, NY: W. H. Freeman.

Berelson

B. R.

(1952). Democratic theory and public opinion. Public Opinion Quarterly, 16, 313-330.

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.

Binder

A. R.

Dalrymple

K. E.

Brossard

Scheufele

D. A.

(2009). The soul of a polarized democracy. Communication Research, 36, 315-340.

Brader

(2006). Campaigning for hearts and minds. Chicago, IL: University of Chicago Press.

Brundidge

(2010). Encountering “difference” in the contemporary public sphere: The contribution of the Internet to the heterogeneity of political discussion networks. Journal of Communication, 60, 680-700.

Burkhalter

Gastil

Kelshaw

(2002). A conceptual definition and theoretical model of public deliberation in small face-to-face groups. Communication Theory, 12, 398-422.

Calhoun

(1988). Populist politics, communications media and large scale societal integration. Sociological Theory, 6, 219-241.

10.

Campbell

S. W.

Kwak

(2011). Political involvement in “mobilized” society: The interactive relationships among mobile communication, network characteristics, and political participation. Journal of Communication, 61, 1005-1024.

11.

Campbell

S. W.

Kwak

(2012). Mobile communication and strong network ties: Shrinking or expanding spheres of public discourse? New Media & Society, 14, 262-280.

12.

Chen

Casper

W. J.

Cortina

J. M.

(2001). The roles of self-efficacy and task complexity in the relationships among cognitive ability, conscientiousness, and work-related performance: A meta-analytic examination. Human Performance, 14, 209-230.

13.

Cronbach

L. J.

(1987). Statistical tests for moderator variables: Flaws in analyses recently proposed. Psychological Bulletin, 102, 414-417.

14.

Dodson

(2010). The return of the American voter? Party polarization and voting behavior, 1988 to 2004. Sociological Perspectives, 53, 443-449.

15.

Eveland

W. P.

Jr. Hively

M. H.

(2009). Political discussion frequency, network size, and “heterogeneity” of discussion as predictors of political knowledge and participation. Journal of Communication, 59, 205-224.

16.

Garrett

R. K.

(2009). Politically motivated reinforcement seeking: Reframing the selective exposure debate. Journal of Communication, 59, 676-699.

17.

Gastil

Black

Moscovitz

(2008). Ideology, attitude change, and deliberation in small face-to-face groups. Political Communication, 25, 23-46.

18.

Gergen

K. J.

(2008). Mobile communication and the transformation of the democratic process. In Katz

J. E.

(Ed.), Handbook of mobile communication studies (pp. 297-310). Cambridge, MA: MIT Press.

19.

Glynn

C. J.

McLeod

J. M.

(1984). Public opinion du jour: An examination of the spiral of silence. Public Opinion Quarterly, 48, 731-740.

20.

Granovetter

M. S.

(1973). The strength of weak ties. American Journal of Sociology, 78, 1360-1380.

21.

Hayes

A. F.

(2012). PROCESS: A versatile computational tool for observed variable moderation, mediation, and conditional process modeling [White paper]. Retrieved from http://www.afhayes.com/public/process2012.pdf

22.

Hayes

A. F.

Preacher

K. J.

Myers

T. A.

(2011). Mediation and the estimation of indirect effects in political communication research. In Bucy

E. P.

Holbert

R. L.

(Eds.), The sourcebook for political communication research: Methods, measures, and analytical techniques (pp. 434-465). New York, NY: Routledge.

23.

Haythornthwaite

(2002). Strong, weak, and latent ties and the impact of new media. Information Society, 18, 385-401.

24.

Huckfeldt

Johnson

Sprague

(2004). Political disagreement: The survival of diverse opinions within communication networks. New York, NY: Cambridge University Press.

25.

Huckfeldt

Mendez

J. M.

Osborn

(2004). Disagreement, ambivalence, and engagement: The political consequences of heterogeneous networks. Political Psychology, 25, 65-95.

26.

Ikeda

Boase

(2011). Multiple discussion networks and their consequence for political participation. Communication Research, 38, 634-659.

27.

Jaccard

Turrisi

Wan

C. K.

(1990). Interaction effects in multiple regression. Newbury Park, CA: SAGE.

28.

Jang

S.-J.

(2009). Are diverse political networks always bad for participatory democracy? American Politics Research, 37, 879-898.

29.

Kim

Scheufele

D. A.

Han

J. Y.

(2011). Structure or predisposition? Exploring the interaction effect of discussion orientation and discussion heterogeneity on political participation. Mass Communication and Society, 14, 502-526.

30.

Knobloch-Westerwick

Meng

(2009). Looking the other way. Communication Research, 36, 426-448.

31.

Kunda

(1990). The case for motivated reasoning. Psychological Bulletin, 108, 480-498.

32.

Kwak

(1999). Revisiting the knowledge gap hypothesis: Education, motivation, and media use. Communication Research, 26, 385-413.

33.

Kwak

Williams

Wang

Lee

(2005). Talking politics and engaging politics: An examination of the interactive relationships between structural features of political talk and discussion engagement. Communication Research, 32, 87-111.

34.

Lasorsa

D. L.

(1991). Effects of newspaper competition on public opinion diversity. Journalism Quarterly, 68, 38-47.

35.

Lee

F. L. F.

(2012). Does discussion with disagreement discourage all types of political participation? Survey evidence from Hong Kong. Communication Research, 39, 543-562.

36.

Lee

F. L. F.

Chan

J. M.

(2009). The political consequences of ambivalence: The case of democratic reform in Hong Kong. International Journal of Public Opinion Research, 21, 47-64.

37.

Lee

(2012). Communication mediation model of late-night comedy: The mediating role of structural features of interpersonal talk between comedy viewing and political participation. Mass Communication and Society, 15, 647-671.

38.

MacKinnon

D. P.

Fairchild

A. J.

Fritz

M. S.

(2007). Mediation analysis. Annual Review of Psychology, 58, 593-614.

39.

Mansbridge

(1983). Beyond adversary democarcy. Chicago, IL: University of Chicago Press.

40.

Mansbridge

Morris

(2001). Oppositional consciousness: The subjective roots of social protest. Chicago, IL: University of Chicago Press.

41.

Marcus

G. E.

Sullivan

J. L.

Theiss-Morse

Wood

S. L.

(1995). With malice toward some: How people make civil liberties judgments. New York, NY: Cambridge University Press.

42.

Matthes

(2012). Exposure to counterattitudinal news coverage and the timing of voting decisions. Communication Research, 39, 147-169.

43.

Matthes

Rios Morrison

Schemer

(2010). A spiral of silence for some: Attitude certainty and the expression of political minority opinions. Communication Research, 37, 774-800.

44.

McClurg

S. D.

(2006a). The electoral relevance of political talk: Examining disagreement and expertise effects in social networks on political participation. American Journal of Political Science, 50, 737-754.

45.

McClurg

S. D.

(2006b). Political disagreement in context: The conditional effect of neighborhood context, disagreement and political talk on electoral participation. Political Behavior, 28, 349-366.

46.

McGuire

W. J.

(1961). The effectiveness of supportive and refutational defenses in immunizing and restoring beliefs against persuasion. Sociometry, 24, 184-197.

47.

McLeod

J. M.

Scheufele

D. A.

Moy

Horowitz

E. D.

Holbert

R. L.

Zhang

Zubric

(1999). Understanding deliberation: The effects of discussion networks on participation in a public forum. Communication Research, 26, 743-774.

48.

McPherson

Smith-Lovin

Cook

J. M.

(2001). Birds of a feather: Homophily in social networks. Annual Review of Sociology, 27, 415-444.

49.

Milgram

(1974). Obedience to authority: An experimental view. New York, NY: Harper Colophon Books.

50.

Morey

A. C.

Eveland

W. P.

Jr. Hutchens

M. J.

(2012). The “who” matters: Types of interpersonal relationships and avoidance of political disagreement. Political Communication, 29, 86-103.

51.

Mutz

D. C.

(2002a). The consequences of cross-cutting networks for political participation. American Journal of Political Science, 46, 838-855.

52.

Mutz

D. C.

(2002b). Cross-cutting social networks: Testing democratic theory in practice. American Political Science Review, 96, 111-126.

53.

Mutz

D. C.

(2006). Hearing the other side: Deliberative versus participatory democracy. Cambridge, MA: Cambridge University Press.

54.

Mutz

D. C.

Mondak

J. J.

(2006). The workplace as a context for cross-cutting political discourse. Journal of Politics, 68, 140-155.

55.

Neuwirth

Frederick

Mayo

(2007). The spiral of silence and fear of isolation. Journal of Communication, 57, 450-468.

56.

Nickerson

R. S.

(1998). Confirmation bias: A ubiquitous phenomenon in many guises. Review of General Psychology, 2, 175-220.

57.

Niemi

R. G.

Craig

S. C.

Mattei

(1991). Measuring internal political efficacy in the 1988 national election study. American Political Science Review, 85, 1407-1413.

58.

Nir

(2005). Ambivalent social networks and their consequences for participation. International Journal of Public Opinion Research, 17, 422-442.

59.

Noelle-Neumann

(1993). The spiral of silence: Public opinion, our social skin. Chicago, IL: University of Chicago Press.

60.

Norris

(2002). The bridging and bonding role of online communities. Harvard International Journal of Press/Politics, 7(3), 3-13.

61.

Park

S. G.

(2000). The significance of civility in deliberative democracy. Korean Journal of Journalism & Communication Studies, 45, 162-194.

62.

Petty

R. E.

Cacioppo

J. T.

(1990). Involvement and persuasion: Tradition versus integration. Psychological Bulletin, 107, 367-374.

63.

Pfau

Burgoon

(1988). Inoculation in political campaign communication. Human Communication Research, 15, 91-111.

64.

Pfau

Tusing

K. J.

Koerner

A. F.

Lee

Godbold

L. C.

Penaloza

L. J.

Hong

Y. Y.-H.

(1997). Enriching the inoculation construct: The role of critical components in the process of resistance. Human Communication Research, 24, 187-215.

65.

Price

Cappella

J. N.

Nir

(2002). Does more disagreement contribute to more deliberative opinion? Political Communication, 19, 95-112.

66.

Putnam

R. D.

(2000). Bowling alone: The collapse and revival of American community. New York, NY: Simon and Shuster.

67.

Putnam

R. D.

Yonish

(1999, May). How important is response rate? An evaluation of a “mail panel” survey archive. Paper presented at the the annual meeting of the American Association for Public Opinion Research, Saint Pete Beach, FL.

68.

Rimal

R. N.

Real

(2003). Perceived risk and efficacy beliefs as motivators of change: Use of the risk perception attitude (RPA) framework to understand health behaviors. Human Communication Research, 29, 370-399.

69.

Rudolph

T. J.

Gangl

Stevens

(2000). The effects of efficacy and emotions on campaign involvement. Journal of Politics, 62, 1189-1197.

70.

Scheufele

D. A.

(2002). Examining differential gains from mass media and their implications for participatory behavior. Communication Research, 29, 46-65.

71.

Scheufele

D. A.

Hardy

B. W.

Brossard

Waismel-Manor

I. S.

Nisbet

E. C.

(2006). Democracy based on difference: Examining the links between structural heterogeneity, heterogeneity of discussion networks, and democratic citizenship. Journal of Communication, 56, 728-753.

72.

Scheufele

D. A.

Nisbet

M. C.

Brossard

(2003). Pathways to political participation? Religion, communication contexts, and mass media. International Journal of Public Opinion Research, 15, 300-324.

73.

Scheufele

D. A.

Nisbet

M. C.

Brossard

Nisbet

E. C.

(2004). Social structure and citizenship: Examining the impacts of social setting, network heterogeneity, and informational variables on political participation. Political Communication, 21, 315-338.

74.

Schuck

A. R. T.

de Vreese

C. H.

(2012). When good news is bad news: Explicating the moderated mediation dynamic behind the reversed mobilization effect. Journal of Communication, 62, 57-77.

75.

Schudson

(1984). Embarrassment and Erving Goffman’s idea of human nature. Theory and Society, 13, 633-648.

76.

Slater

M. D.

(2007). Reinforcing spirals: The mutual influence of media selectivity and media effects and their impact on individual behavior and social identity. Communication Theory, 17, 281-303.

77.

Stroud

N. J.

(2011). Niche news: The politics of news choice. New York, NY: Oxford University Press.

78.

Tajfel

Turner

(1986). The social identity of inter-group behavior. In Worchel

Austin

W. G.

(Eds.), Psychology of intergroup relations (pp. 7-24). Chicago, IL: Nelson-Hall.

79.

Thorson

(2012). Beyond opinion leaders: How attempts to persuade foster political awareness and campaign learning. Communication Research. Advance online publication. doi:10.1177/0093650212443824

80.

Ulbig

S. G.

Funk

C. L.

(1999). Conflict avoidance and political participation. Political Behavior, 21, 265-282.

81.

Valentino

N. A.

Gregorowicz

Groenendyk

(2009). Efficacy, emotions and the habit of participation. Political Behavior, 31, 307-330.

82.

Valenzuela

Kim

Gil

Zúñiga

(2012). Social networks that matter: Exploring the role of political discussion for online political participation. International Journal of Public Opinion Research, 24, 163-184.

83.

Witte

Allen

(2000). A meta-analysis of fear appeals: Implications for effective public health campaigns. Health Education & Behavior, 27, 591-615.

84.

Wojcieszak

(2009). “Carrying online participation offline”—Mobilization by radical online groups and politically dissimilar offline ties. Journal of Communication, 59, 564-586.

85.

Wojcieszak

(2010). “Don’t talk to me”: Effects of ideologically homogeneous online groups and politically dissimilar offline ties on extremism. New Media & Society, 12, 637-655.

86.

Wojcieszak

(2011a). Deliberation and attitude polarization. Journal of Communication, 61, 596-617.

87.

Wojcieszak

(2011b). Pulling toward or pulling away: Deliberation, disagreement, and opinion extremity in political participation. Social Science Quarterly, 92, 207-225.

88.

Wojcieszak

(2011c). When deliberation divides: Processes underlying mobilization to collective action. Communication Monographs, 78, 324-346.

89.

Wojcieszak

Baek

Y. M.

Delli Carpini

M. X.

(2010). Deliberative and participatory democracy? Ideological strength and the processes leading from deliberation to political engagement. International Journal of Public Opinion Research, 22, 154-180.

90.

Wojcieszak

Mutz

D. C.

(2009). Online groups and political discourse: Do online discussion spaces facilitate exposure to political disagreement? Journal of Communication, 59, 40-56.

91.

Wojcieszak

Price

(2010). Bridging the divide or intensifying the conflict? How disagreement affects strong predilections about sexual minorities. Political Psychology, 31, 315-339.

92.

Wojcieszak

Price

(2012). Perceived versus actual disagreement: Which influences deliberative experiences? Journal of Communication, 62, 418-436.