Partisan Ambivalence,Partisan Intensity,and Racial Attitudes: The Impact of Shifting Policy Positions on Partisan Evaluations in the 1960s

Abstract

The concept of ambivalence is important to the study of political psychology and behavior. We examine the causes of partisan ambivalence following the passage of major civil rights legislation to test our argument that the correlates of ambivalence will change following the alteration of long-standing party positions on a highly salient issue. We find support for this hypothesis for White Southerners during this time period—Indeed, our results demonstrate that the strength of one’s partisan attachment is unrelated to partisan evaluations for such individuals immediately following the passage of major civil rights legislation, but is again a predictor by the early 1970s.

Keywords

partisan ambivalence Southern politics party realignment

The concept of ambivalence has become increasingly important to the study of political psychology and public opinion. The considerable impact of ambivalence on subsequent attitude formation and behavior is well documented (e.g., Lavine, 2001; Lavine, Johnston, & Steenbergen, 2012; Meffert, Guge, & Lodge, 2004). Likewise, we now have a growing understanding of the causes of ambivalence (e.g., Keele & Wolak, 2008; Rudolph & Popp, 2007). Yet because nearly all of this evidence is from surveys conducted roughly in the last 25 years, the broad social and political context of these studies is similar. Here, we examine the causes of ambivalence toward the parties in a different era: the presidential elections immediately following the passage of major civil rights legislation in the 1960s. We argue that following the alteration of long-standing positions on a highly salient issue the influence of partisanship on ambivalence will be attenuated while that of the attitudes relevant to the policy will be heightened. Specifically, we believe that among individuals for whom the elite change is most salient—in this case, White Southerners—the degree to which individuals are one-sided or conflicted in their evaluations of the parties according to stated likes and dislikes of their own as well as the opposite party will be unaffected by partisan attachments during this period, and will instead be influenced by racial attitudes. The context of the mid-to-late 1960s offers a clear test of these hypotheses given the lack of visible major party positions on civil rights legislation until this point. Comparing the correlates of partisan ambivalence in the periods before, during, and after the major parties polarize on a salient issue enhances our knowledge of the contextual influences on an important psychological phenomenon.

The party system, especially in the South, was in flux in the 1960s following the passage of two important pieces of civil rights legislation. As a result of this abrupt change in policy stances on the issue of race, we argue that evaluations of the parties among Southerners will be less dependent on the strength of individuals’ party identification and more dependent on their racial attitudes compared with citizens elsewhere. We find support for the hypothesis that the correlates of ambivalence were different for White Southerners during this period—In both 1964 and 1968, the impact of partisan intensity is weaker among Southerners, and racial attitudes predict ambivalence in our 1968 Southern sample. We interpret this result as being consistent with evidence demonstrating that although partisans did not immediately leave their party during this period (Green & Palmquist, 1994; Green, Palmquist, & Shickler, 1998, 2002), the electorate did respond to elite polarization on the issue of race (Carmines & Stimson, 1989; Valentino & Sears, 2005). Even if White Southerners did not abandon their party identification straightaway, we argue that their partisan identity became “decoupled” from their evaluations of the parties and as a result their partisanship played a smaller role in shaping attitudes toward the parties. That is, Southerners’ expressed likes and dislikes of the parties, and thus their likelihood of experiencing ambivalence became divorced from their self-identified partisanship due to elites’ changing position on the salient issue of race. Moreover, we demonstrate that the relationship between partisan intensity and ambivalence—two distinct dimensions of attitude strength (J. M. Miller & Peterson, 2004)—among Whites was evident in the election immediately prior to the disruption, 1960, and it reappears in each of the four subsequent elections, 1972-1984, suggesting that once the new political reality firmly supplanted the previous status quo, partisanship’s influence on evaluations was again substantial. In that regard, our results also comport with recent evidence demonstrating the importance of “group-party” ambivalence (Shufeldt, 2017). Furthermore, the results are consistent with theories of issue-based partisan change in which elites’ polarization on a novel political issue produces gradual shifts in citizens’ partisan identities (Carsey & Layman, 2006; Carmines & Stimson, 1986, 1989; Layman & Carsey, 2002b,) and attitudes (Zaller, 1992).

The purpose of this article is to shed light on the way in which changing party positions on a salient issue affects the attitudes of the electorate—namely, polarization on an issue alters the criteria by which some in the electorate evaluate the parties. Importantly, one’s level of ambivalence toward the political parties is potentially unrelated to his or her own partisanship if the disruption is sufficiently great. Periodic disruption of the ties that hold together party coalitions is a common feature of American politics (Key, 1955; Schofield, Miller, & Martin, 2003; Sundquist, 1983), and such an occurrence may be on the horizon (G. Miller & Schofield, 2008). This article demonstrates how the electorate responds to such an occurrence—Partisan identities are resistant to change, but their importance for subsequent attitudes can be diminished. Investigating the correlates of ambivalence in the years both prior and subsequent to major party polarization on the issue of race enables us to assess over time changes in the relationship between shifting party positions on a salient issue and voters’ evaluations of the parties.

Background

An individual is ambivalent when he or she has internalized both positive and negative attitudes toward some stimulus object (e.g., Newby-Clark, McGregor, & Zanna, 2002; Thompson, Zanna, & Griffin, 1995). Considerable evidence documents the considerable influence of ambivalence on attitudes and behavior. Evaluative conflict can diminish the accessibility (Huckfeldt & Sprague, 2000), certainty (McGraw, Hasecke, & Conger, 2003; Meffert et al., 2004; Tetlock, 1986), extremity (Meffert et al., 2004), and stability (Alvarez & Brehm, 2002; Rudolph, 2005) with which individuals hold their views. Ambivalence toward the political parties and candidates influences how individuals seek information (Basinger & Lavine, 2005) and the timing of electoral decisions (Mutz, 2002; Nir & Druckman, 2008), as well as patterns of political involvement (Mutz, 2002) and voting behavior (Lavine et al., 2012; Mulligan, 2011).

Known sources of partisan ambivalence include one’s context, including the campaign environment (Keele & Wolak, 2008) and social networks (Huckfeldt, Mendez, & Osborn, 2004), as well as personal characteristics, such as knowledge and one’s need for cognition (Rudolph & Popp, 2007). One individual-level variable consistently shown to affect ambivalence is the intensity of one’s partisan attachment (Huckfeldt et al., 2004; Keele & Wolak, 2008; Rudolph, 2011; Rudolph & Popp, 2007). However, nearly the entirety of the evidence regarding both the causes and consequences of ambivalence consists of studies of survey responses conducted no earlier than the 1990s. We examine the causes of ambivalence in an era of political upheaval, particularly in a region where partisan loyalties might be most tried—the American South during the civil rights movement.

For several decades prior to the passage of major civil rights legislation the parties were polarizing over racial issues. During this time, Northern Democrats in Congress became increasingly likely to take stances on these issues and support attempts to bring legislation to the floor (Noel, 2013; Schickler, Pearson, & Feinstein, 2010). Likewise, the thinking of public intellectuals on the matter also began to change in the 1940s (Noel, 2012). Yet, many—probably most—in the electorate were likely unaware of shifting legislative coalitions or the changing racial attitudes among intellectuals expressed in the pages of outlets such as the National Review. Certainly, large segments of the electorate had not yet fully processed the change. For example, Carmines and Stimson (1986) find that the public was just noticing a difference in the two parties’ stances on racial issues in the mid-1960s. For these individuals, the passage of major civil rights legislation—the Civil Rights Act of 1964 and the Voting Rights Act of 1965—signaled that party coalitions were significantly changing along racial lines. Opposition to integration and extending equal rights to Blacks was rooted in resistance to societal change (Kinder & Sears, 1981; Sears, Lau, Tyler, & Allen, 1980; Valentino & Sears, 2005), prompting Southern Whites to reassess their partisan loyalties in response to these landmark laws and the parties’ stances on racial issues (E. Black & Black, 2002; M. Black, 2004; Carmines & Stimson, 1989; Hayes & McKee, 2007; Sundquist, 1983).

Still, although the public recognized the parties’ polarizing stands on racial issues (Carmines & Stimson, 1989; Valentino & Sears, 2005), individuals did not defect from their party identification in droves in the 1960s (e.g., Green et al., 2002; Lublin, 2004).¹ Consistent with theories of partisanship as a psychological attachment (Campbell, Converse, Miller, & Stokes, 1960), older individuals were particularly more resistant to abandoning their partisanship (Beck, 1977). Despite partisan identification remaining reasonably stable, there were other indications that White Southerners viewed the parties differently, including defections in presidential elections acting as “a leading indicator” (Osborne, Sears, & Valentino, 2011, p. 83) as well as an increase in split-ticket voting (Burden & Kimball, 2002).² A general characterization of the electorate following the disruption of the status quo in the 1960s was that partisanship mattered less to individuals’ subsequent attitudes and behavior in the South compared with the rest of the country (Stanley, 1988), as well as that the Democratic “Solid South” began to erode (E. Black & Black, 1992; Carmines & Stimson, 1989; Sundquist, 1983). The Southern electorate’s response to major party polarization on the dimension of race evidences the phenomenon of issue evolution in which shifts in elites’ issue positions produce partisan change (Carmines & Stimson, 1986, 1989), a pattern also observed for issues ranging from social welfare concerns (Abramowitz & Saunders, 1998) to cultural issues such as abortion (Adams, 1997; Carmines & Woods, 2002) and the environment (Lindaman & Haider-Markel, 2002).³ We argue that another way in which the reaction to a change in elite positions will manifest itself among the mass public is that the criteria used to evaluate the parties themselves will be different. As the parties staked out new and visible stances on racial issues during this period, White Southerners felt abandoned by their party and temporarily stopped using their partisan identities as a guide for evaluating the major parties. In other words, as a consequence of the emergence of party change on this highly visible and salient issue, we argue that the correlates of partisan ambivalence will be altered for White Southerners.⁴

Theory

We argue that the parties’ shifting signals on a salient issue weaken the influence of partisan attachments and increase those of policy-relevant attitudes on individuals’ evaluation of the two parties. To show why this is the case, we begin by conceptualizing partisanship as a psychological attachment (Campbell et al., 1960) and social identity (Green et al., 2002; S. Greene, 1999, 2004; Mason, 2015). Campbell et al.’s (1960) conceptualization of partisanship as a psychological attachment has been expanded by explicitly incorporating Tajfel’s (1978) definition of social identity, which emphasizes the importance that group membership can have on one’s self-identity and subsequent behavior. Key to this conceptualization is that an individual’s sense of belonging to a group consequently affects both one’s self- and group-image. Thus, under usual circumstances, one’s partisan identification should impact how one interprets and perceives parties, candidates, and issues, as well as how one behaves (e.g., Huddy, 2001; Huddy & Bankert, 2017; Huddy, Mason, & Aarøe, 2015; Lewis-Beck, Jacoby, Norpoth, & Weisberg, 2008). As a result, stronger partisans are, on average, less ambivalent than weaker partisans and independents (e.g., Keele & Wolak, 2008; Rudolph, 2011; Rudolph & Popp, 2007). However, some individuals do not have an exclusively or even predominantly one-sided evaluation of the two political parties. We argue that studying the concept of ambivalence in the periods before, during, and after major party polarization on the issue of race allows us to gain leverage on the question of how a sudden and radical change in party positions on a salient issue affects individuals’ evaluations of the parties.

When the parties diverge and then differentiate on a prominent issue, some individuals may feel “abandoned” by their party and the normal processes that influence ambivalence may be altered. Stated differently, identity-conflicting evaluations can occur when “party behavior and performance impinge on an area of personal importance to a voter” (Lavine et al., 2012, p. 8). In particular, although one’s evaluation of the parties is usually a function of what Rudolph and Popp (2007) term “directional” motivations, including partisan intensity, we argue that following large policy changes on a salient issue, this process weakens. A person who identifies with a party based on habit or tradition, but who feels he or she is no longer “part of the party” or that “the party changed, I stayed the same” is more likely to be ambivalent. Indeed, a person may be a strong partisan but will not necessarily strongly identify with the national party—He or she may strongly identify with the partisan label, but not with the party or its policies.⁵ In other words, partisans’ views of the parties and their traditional supporters and policies continue to exist, and thus conflict, with the new reality during periods of party upheaval. As a result, partisan identifiers view themselves as the true supporters of their party while viewing the national party’s priorities with skepticism and perhaps hostility.

We agree with Green et al. (2002, p. 162) that self-conceptions change gradually, and thus we do not expect individuals who become uncomfortable with the parties’ altered group images—what it means to be a Democrat and Republican fundamentally changes during tumultuous periods when parties shift positions and align with new social groups—to switch party loyalties immediately in the face of such a disruption. Given party identification’s centrality to one’s self-image, we instead anticipate that, as Green et al. (2002) write, “A person may identify with a party despite policy disagreement in the hopes that these tensions will subside” (p. 159). Consequently, we argue that the initial response for most partisans to parties’ changing policy positions and group images will be a “decoupling” of partisanship from partisan evaluations that alters the relationship between partisan intensity and partisan ambivalence. Thus, the act of disconnecting one’s partisan attachments from one’s party evaluations may be a first step in the process of partisan change and polarization as citizens ultimately decide whether to adopt their party’s new issue stance and image or switch parties (e.g., Carsey & Layman, 2006; Green et al., 2002).

We focus on evaluations of the parties following the passage of civil rights legislation in the 1960s. As mentioned, until this point the parties were hesitant to stake clear stands on the issue. For example, in 1960, 56% of respondents believed there was no difference between the parties when it came to protecting Black Americans from discrimination—while the number declined to 32% in 1964 and 38% in 1968. Furthermore, this decline coincided with an increase in the belief that Democrats were the party more likely to protect African Africans from discrimination.⁶

Given the parties’ relatively sudden split on civil rights policies that were largely aimed to improve social and political conditions for African Americans residing in the South, our expectation is that White Southerners are those who will feel abandoned, and thus partisan intensity should have a smaller impact on partisan ambivalence in the American South in this time period. We wish to note that our theory does not apply exclusively to Southern Democrats; Republican identifiers with more liberal racial attitudes might also feel abandoned by their party or perhaps will now have positive views of the national Democratic Party. Our first hypothesis is as follows:

Hypothesis 1: Partisan intensity will have a smaller impact on ambivalence in the South than elsewhere in the United States following the passage of major civil rights legislation.

We emphasize importantly here that partisan strength and ambivalence are not identical concepts. For example, using panel data, Rudolph (2011) writes, “The results of this analysis demonstrate that ambivalence and partisan strength are not mirror equivalents of each other. The results suggest that causality flows from partisan strength to ambivalence but not the reverse” (p. 568). This finding is consistent with evidence from Steenbergen, Lavine, and Goolsby (2006) and Lavine et al. (2012).

A related consequence of a significant party change on a major issue is that attitudes relating to the policy area should have a larger influence on how individuals evaluate the parties, as these attitudes are now more salient. Our theory here suggests that the importance of attitudes toward African Americans and the government’s role in promoting racial equality will increase. Given that the legislation largely targeted the South, we expect racial attitudes will matter more there than elsewhere. Thus, our second hypothesis is as follows:

Hypothesis 2: Racial attitudes will have a larger impact on partisan ambivalence in the South than elsewhere in the United States following the passage of major civil rights legislation.

The passage of highly visible and salient legislation on a controversial topic potentially disrupts how individuals evaluate the parties. Following an examination of two elections, 1964 and 1968, in detail, we examine all presidential elections from 1960 to 1984 to test our third hypothesis:

Hypothesis 3: The relationship between partisan intensity and ambivalence should be observable both before and well after major party polarization on the issue of race for White Southerners.

The logic of this hypothesis is that following the establishment of a new party system status quo, members of the electorate will recognize and accept the new party coalitions and as a consequence one’s partisan attachment will behave as it had previously. Our argument here is similar to that of Lavine et al. (2012), who argue that because partisanship is a social identity, once the system returns to a stable status quo, “the partisan perceptual screen is reinstated” (p. 198).

Data and Methods

To test our first two hypotheses, we examine data from the 1964 and 1968 CPS American National Election Studies (ANES).⁷ We select these two years to explore in detail because each follows significant civil rights legislation and because each survey has the full battery of questions needed to test our theory. Following this analysis, we examine all presidential election years 1960-1984 to test more completely the dynamic aspect of our theory. To examine the impact of partisan intensity and racial attitudes on partisan evaluations, we use Basinger and Lavine’s (2005) index of comparative ambivalence. The measure, which is a modification of a common objective measure (Thompson, Zanna, & Griffin, 1995), combines responses about both major parties from open-ended questions asking respondents what they like and dislike about the political parties to gauge the conflict regarding evaluations of the parties.⁸

Because ambivalence measures mishandle indifference (Rudolph, 2005; Thornton, 2011)—here, those respondents who offer no responses to the open-ended questions about the parties—we account for indifferent individuals by estimating a selection model (Heckman, 1979), as suggested by Kimball and Summary (2012) and Thornton (2013). In this scenario, the selection stage models if the respondent lists any statements and the outcome stage models the score from the Basinger and Lavine (2005) measure for those who offer at least one response. Given our theory’s emphasis on ambivalence, we are primarily interested in the coefficients in the outcome stage.

We define the South using census bureau coding; although this includes “border” states, we choose this coding to increase our Southern sample—and, in a sense this may make our tests more conservative as deep South states tend to be more different (e.g., Glaser & Gilens, 1997; McKee & Springer, 2015).⁹ In addition, we limit our analysis to White individuals given our focus on attitudes toward African Americans. Because evidence suggests a substantial portion of the variation toward the higher end of the ambivalence measure may not be meaningful (Thornton, 2011), we test the robustness of the selection model by coding individuals into three nominal categories of ambivalent, indifferent, and univalent. These results are presented in Supplementary Appendix B. We measure partisan intensity by “folding” the traditional partisan scale around its midpoint. It therefore ranges from zero to three, from pure independent to strong partisan. Our theory suggests that Southerners will be less influenced by partisan strength when evaluating the parties than will others.

Both the 1964 and 1968 ANES feature multiple questions regarding respondents’ attitudes toward African Americans and the government’s role in ensuring racial equality. We use six questions to create a scale of racial attitudes. We include a 3-point scale of support for school integration, a binary measure of attitudes toward Whites’ ability to block African Americans from moving into their neighborhood, a 3-point measure gauging support for the government’s involvement in making sure African Americans are treated fairly when looking for jobs, a 3-point scale of if the respondent thinks the civil rights movement is pushing too fast, a 3-point scale of one’s attitudes toward segregation, and finally, a feeling thermometer question measuring affect toward Blacks, which ranges from 0 to 100. We recode all measures to range from zero to one and so that higher values are associated with negative attitudes toward African Americans. The resulting scale is quite reliable in both years: It has a Cronbach’s α of .762 in 1964 and .742 in 1968. We include all respondents who answer at least four of the six questions (which is over 95% of the respondents in each year). Our theory suggests that the impact of this variable on ambivalence will be stronger in the South.

Given that our theory suggests that the effects of partisan intensity and racial attitudes on partisan ambivalence will vary among Southerners and non-Southerners, we interact each of our primary explanatory variables with a South dummy variable—coded one for Southerners and zero for all other respondents. Recall our first hypothesis that partisan intensity will have a larger impact on ambivalence for non-Southerners. Next, we hypothesize that racial attitudes will have a larger impact on ambivalence for Southerners.¹⁰

We include several control variables previously demonstrated to influence partisan ambivalence (e.g., Keele & Wolak, 2008; Rudolph & Popp, 2007).¹¹ Education is measured using a 6-point scale ranging from no high school to a college degree or higher. We control for knowledge in the form of whether or not the individual correctly identifies the party controlling the House of Representatives, coded one for answering correctly. Because those who care about the election may have more one-sided opinions, we include a binary measure of whether or not the individual cares about the election outcome. Finally, we include age, as those with more experience with the political system may be more likely to have opinions about the parties (e.g., Converse, 1969). Though not always necessary for the identification of a set of equations (Vella, 1998), we exclude age from the outcome equation. This decision does not drive our substantive results.

Results

The results from the selection models are presented in Tables 1 and 2 for 1964 and 1968, respectively. Our first hypothesis is that partisan intensity will have a smaller impact on how the parties are evaluated in the South than elsewhere as a result of the parties’ shifting policy priorities. We find support for this hypothesis in both years: The coefficient for the interaction term is positive and statistically significant, indicating that partisanship matters less in the South. Moreover, partisan intensity is a significant predictor of ambivalence—those who identify more strongly are less ambivalent—in our non-South sample. Among White Southerners, not only is the impact of partisanship smaller, but the marginal effect also cannot be distinguished from zero in either year.¹² Consistent with our theory, after the disruption to the status quo following the passage of major civil rights legislation, the strength of one’s partisan attachment did not systematically reduce ambivalence for White Southerners.

Table 1.

Partisan Ambivalence, 1964.

	Coefficient	SE	t
Outcome equation
South	−0.439	0.203	−2.16
Partisan intensity	−0.252	0.035	−7.12
Partisan Intensity $\times$ South	0.155	0.064	2.43
Racial attitudes	−0.066	0.123	−0.54
Racial Attitudes $\times$ South	0.128	0.226	0.57
Education	0.048	0.015	3.14
Knowledge	0.021	0.057	0.38
Care who wins	−0.323	0.058	−5.55
Constant	0.530	0.131	4.03
Selection equation
Partisan intensity	0.399	0.052	7.63
Racial attitudes	−0.004	0.191	−0.02
Education	0.131	0.033	3.96
Knowledge	0.404	0.097	4.16
Care who wins	0.560	0.095	5.92
Age	0.012	0.003	3.76
Constant	−1.210	0.237	−5.11
N	1,266
Pr(ρ=0)	0.004
Log-likelihood	–1685.178
χ²	114.40
p	.000

Note. Estimated using selection model.

Table 2.

Partisan Ambivalence, 1968.

	Coefficient	SE	t
Outcome equation
South	−0.265	0.206	−1.29
Partisan intensity	−0.233	0.041	−5.7
Partisan Intensity $\times$ South	0.160	0.067	2.37
Racial attitudes	−0.251	0.155	−1.62
Racial Attitudes $\times$ South	−0.154	0.262	−0.59
Education	0.056	0.018	3.12
Knowledge	0.158	0.076	2.07
Care who wins	−0.203	0.070	−2.90
Constant	0.373	0.175	2.13
Selection equation
Partisan intensity	0.275	0.055	4.97
Racial attitudes	−0.325	0.226	−1.44
Education	0.120	0.036	3.34
Knowledge	0.502	0.110	4.58
Care who wins	0.467	0.104	4.49
Age	0.016	0.004	4.40
Constant	−0.839	0.267	−3.15
N	1,148
Pr(ρ=0)	0.382
Log-likelihood	–1688.521
χ²	90.39
p	.000

Note. Estimated using selection model.

Our second hypothesis is that due to the increased salience of racial attitudes following the passage of civil rights legislation, racial attitudes should be a larger predictor of ambivalence in the South, where most of the legislation was directed. Here we find mixed results. We find some support for our hypothesis in 1968 but none in 1964. The coefficient for the interaction term is not statistically significant in either year. However, examining the marginal effect suggests that the impact is statistically significant for the Southern subsample in 1968 (p < .05), but not in 1964, when racial attitudes are unrelated to ambivalence in both samples.

There are a few plausible explanations for this result. First, perhaps the presence of George Wallace, a candidate for the American Independent Party in 1968 who made race an explicit part of his candidacy, increased the salience of these concerns—Although Barry Goldwater was an opponent of the legislation in 1964, he appealed to the issue much more in terms of states’ rights than race.¹³ Boyd (1972), for example, suggests that 1968 was an unusual election in terms of issue salience and issue voting. In addition, Green et al. (2002, Chapter 6) convincingly show using panel data that the passage of the Voting Rights Act (1965) truly was the pivotal moment that precipitated Southern party change as Whites responded to dramatic increases in Black voter registration rates, which remained anemic even after 1964. Finally, Beck (1977) argues that younger individuals were more receptive to change. Thus, perhaps the impact of racial attitudes is contingent not only on region, but also on age. To test this argument, we estimated a model in which we interacted racial attitudes with both age and region. We find support for this argument in 1968 but not in 1964. These results are available in Supplementary Appendix D. Green et al. (2002, Chapter 6) also help us reconcile the finding that age moderates the relationship between racial attitudes and party evaluations because they show that older Southerners are both more conservative and more likely to remain Democrats over time relative to their younger cohorts, who increasingly entered the electorate as Republicans. We are therefore unsurprised to learn that older Southerners are more likely to decouple their long-standing partisan ties from their party evaluations in response to party change on the issue of race. These racially conservative Whites instead began to evaluate the parties on the basis of their racial attitudes, and some of them gradually changed their party affiliation—Green et al. (2002, Chapter 6) estimate that approximately half of the Southern realignment is owed to conversion, as opposed to generational replacement. We also wish to note that we do not interpret this null finding as definitive evidence that racial attitudes were unimportant in 1964 but instead believe that it may be at least partially a function of the difficultly in gauging racial attitudes accurately.¹⁴

The impact of partisan attitudes might be conditional on both region and partisanship. To examine this possibility, we estimated a model with a three-way interaction between racial attitudes, region, and partisanship. The results, presented in Supplementary Appendix D, do not lend support for such an argument.¹⁵ Finally, we examined the possibility that knowledge moderates the impact of racial attitudes on partisan ambivalence—In particular, we estimated a model with a three-way interaction between region, racial attitudes, and knowledge about the parties’ positions on integrating public schools and preventing employment discrimination against Blacks. The specific coding, specification, and results of these models, presented in Supplementary Appendix D, indicate that knowledge does not moderate the impact of racial attitudes. This finding indicates that the party disruption was ubiquitous among Southerners in the first election following the passage of the Voting Acts Act (e.g., Green et al., 2002 Chapter 6), independent of these individuals’ specific knowledge of the parties’ postures toward public school integration and protecting Blacks from employment discrimination.

We argued that when citizens feel abandoned on a salient issue, and perhaps see the disintegration of long-standing party coalitions, their own partisanship is no longer as significant of a predictor in how they evaluate the parties. Although most Southerners initially held on to their partisan identity (Norpoth & Rusk, 1982), clearly many had mixed feelings about their party as the result of changing coalitions, especially those related to civil rights. When a party stakes a controversial position, it of course risks alienating its supporters. One possible consequence of this action is that some party members in the electorate may continue to identify with the party label, but this identity then has a smaller influence on how they evaluate the parties. Even if many Southerners were initially hesitant to abandon their party identification, their relationship with the parties and the party system fundamentally changed.

The Impact of Partisan Intensity on Ambivalence in the South, 1960-1984

Our theory has dynamic implications suggested by our third hypothesis that are not fully tested in the previous section. If our argument that a shift in policy positions led to a change in the way the parties are evaluated is correct, then the two years already examined should be different from other years. Namely, in the years following the disruption to the status quo, we should see a return to normalcy, and thus partisan intensity should again influence how individuals respond to the parties in the South, as the electorate will have had time to respond to the change in party positions. In this section, we examine the changing impact of partisan intensity on partisan ambivalence among White Southerners both before major party polarization on the issue of race and well after the changes had become established in a new party system status quo. We are limited in this analysis because the ANES does not contain an identical set of questions used to gauge racial attitudes over the period of interest. We can, however, examine a straightforward test of if the impact of partisan intensity on partisan evaluations was influenced by the behavior of political elites. To do so, we estimate a model identical to that presented in Table 1—with the exception that the racial attitude variable is omitted—for each presidential election year from 1960 to 1984.¹⁶ If our theory is correct, then once individuals adjust to the new reality of the party platforms, an increase in partisan intensity should again be associated with a decline in ambivalence and there should be no systematic difference in how Southerners and non-Southerners evaluate the parties. The results of interest are presented in Figure 1, which displays the coefficient of the interaction term between partisan intensity and region from the outcome stage of a selection model as well as the marginal effect of partisan intensity on ambivalence for White Southerners.¹⁷ The full results for each year are available in Supplementary Appendix E.

Figure 1.

The impact of partisan intensity on ambivalence among White Southerners, 1960-1984.

The results are consistent with our expectations. First, note that in 1960, before the passage of landmark civil rights legislation signaling shifting elite policy positions on the salient issue of race, we observe a significant relationship between partisan intensity and ambivalence in the South. Second, by the 1970s and 1980s, partisan intensity once again becomes associated with less ambivalence among White Southerners. Moreover, with the exception of 1964 and 1968, there is no discernible difference in the impact of partisanship on ambivalence across the two regions. In the 1970s, the South was no longer solidly Democratic in presidential elections and in 1972 Richard Nixon was the first Republican presidential candidate to capture the entirety of the former Confederacy, winning the Deep South states with close to 70% of the popular vote in the region. Moreover, by the 1970s, “Southern whites were only marginally different from the total electorate” in terms of the distribution of partisanship (Petrocik, 1987, p. 357), and the relationship between partisanship and other variables—for example, class (Nadeau & Stanley, 1993)—began to resemble that of the rest of the country. Similarly, among a survey of campaign activists, 50% of eventual partisan converts made the switch in the early 1970s (Clark, Bruce, Kessel, & Jacoby, 1991). In other words, the new political reality was firmly established by the 1970s and individuals responded to the parties in the way that we would expect: Their partisanship was a significant predictor of how they evaluated the parties. What we have shown is that as members of the electorate became accustomed to the new political coalitions, their partisanship once again became a major component of how they evaluated the parties.

The evidence presented here and in the previous section is consistent with those who argue that although the disruption to the status quo took a long time to influence the makeup of the Southern electorate (Green et al., 2002) and some citizens were initially reluctant to abandon their partisanship (Norpoth & Rusk, 1982), there were other indicators that the party coalitions had changed. For example, partisan defections in presidential elections increased, which is at least partly attributable to attitudes toward civil rights (Boyd, 1972). Our results demonstrate that the evaluative criteria among White Southerners were different following the change, consistent with Sears and Valentino (1997). Our interpretation of this evidence is that Southerners did not abandon their party wholesale, but instead held mixed feelings that were not assuaged by party loyalty, as usually occurs. The evidence presented in this article is consistent with the argument that partisanship is a social identity that is resistant to change, but we argue that its influence on partisan evaluations can be diminished if a party takes actions that drive a wedge between it and its members on a salient issue.

Conclusion

The importance of ambivalence to the study of political psychology and public opinion is now well established. But given that political scientists have devoted significant attention to the concept for only roughly the last 25 years, almost the entirety of the evidence documenting the causes and consequences of the phenomenon is gleaned from relatively recent data. From a practical standpoint, we provide evidence of its correlates using a different dataset in a different time, which is crucial for testing the generalizability of theories (Berinsky, Powell, Schickler, & Yohai, 2011; Bianco, Lynch, Miller, & Sened, 2006; Caughey, Berinsky, Chatfield, Hartman, Schickler, & Sekhon, n.d.). In doing so, we believe we have furthered the understanding of the relationship between partisan attachments and partisan ambivalence, as well as the reaction of White Southerners to the parties’ polarization on the issue of race during a turbulent and transformative period in American politics.

The results support our argument that the parties’ highly salient actions signifying new national coalitions can alter the relationship between ambivalence and its key predictors. In one of the two years under consideration, racial attitudes were a significant predictor of ambivalence among White Southerners but not of others in the electorate. Moreover, the impact of partisan intensity was indistinguishable from zero in our Southern sample in each year. Furthermore, an analysis of the impact of partisan intensity on ambivalence from 1960 to 1984 demonstrates that in the elections following the upheaval of the status quo, as the electorate became increasingly aware of the new political reality, partisan intensity once again was associated with a decline in ambivalence. In other words, the results demonstrate that when a party takes an action that is out-of-step with many of its members, some in the electorate will “decouple” their partisan identification from their evaluations of the parties. These results are consistent with evidence that although White Southerners did not immediately abandon their partisanship, their relationship with the parties fundamentally changed in the 1960s in response to a shift in the parties’ positions on the issue of race. The reshuffling of coalitions is a periodic feature of political systems and may happen in the near future (G. Miller & Schofield, 2008). The findings in this article suggest that when members of the electorate feel alienated, they may not abandon their partisan identity immediately, but the salience of that identity in formulating their evaluations of the parties—ambivalent or univalent—can be diminished. In other words, the electorate responds to changes in the makeup of party coalitions, even if that change is more gradual than a sudden realignment.

We also consider what our findings’ implications may be for understanding contemporary party competition and intraparty divisions. Namely, might existing issue and group cross-pressures portend a political realignment? We cautiously speculate that socially conservative Democrats—minorities and labor union members, for example—may shift away from their party due to the Democrats’ increasingly liberal social policy platform. We witnessed some of this defection in the 2016 presidential election when less educated Whites—a traditional Democratic constituency—turned sharply toward Donald Trump’s economic populism and racial conservatism (rhetoric that, if continued, likely would forestall a crack in the Democratic Party’s minority coalition). Interestingly, the Democratic Party identification advantage among less educated Whites remains, despite this defection at the polls. Time will tell whether the most recent presidential election was a “leading indicator” (Osborne et al., 2011) or an aberration due to the unique candidate combination, but divergence between partisan loyalty and presidential voting among this electoral subgroup is nonetheless a phenomenon that psephologists will want to monitor moving forward. On the flip side, we also can envision educated, wealthier Republicans defecting from their party if the GOP continues to advocate a nativist, racially hostile platform that is anathema to more cosmopolitan, economic conservatives. These twin developments could produce a wholesale electoral realignment, but we still rate this possibility as unlikely given the American two-party system’s extraordinary resilience, a fact that reinforces the Southern realignment’s exceptionalism.

Supplemental Material

Supplemental_Online_Appendix_APR – Supplemental material for Partisan Ambivalence, Partisan Intensity, and Racial Attitudes: The Impact of Shifting Policy Positions on Partisan Evaluations in the 1960s

Supplemental material, Supplemental_Online_Appendix_APR for Partisan Ambivalence, Partisan Intensity, and Racial Attitudes: The Impact of Shifting Policy Positions on Partisan Evaluations in the 1960s by Robert N. Lupton and Judd R. Thornton in American Politics Research

Footnotes

Acknowledgements

The authors would like to thank Jason Reifler for helpful comments. An earlier version of this paper was presented at the 2012 Annual Meeting of the Midwest Political Science Association where we received helpful feedback from Steven Greene and Eric Groenendyk.

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.

Notes

Supplemental Material

Supplemental material for this article is available online.

Author Biographies

Robert N. Lupton is an assistant professor in the Department of Political Science at the University of Connecticut. He studies the role of individual-level predispositions, primarily core values, ideology and partisanship, in public opinion and voting behavior.

Judd R. Thornton is an assistant professor in the Department of Political Science at Georgia State University. His primary focus is on mass political behavior. In particular, his interests include partisanship, beliefs systems and ideology, the interplay between elite behavior and mass opinion, and issues of measurement.

References

Abramowitz

A. I.

Saunders

K. L.

(1998). Ideological realignment in the American electorate. Journal of Politics, 60, 634-652.

Adams

G. D.

(1997). Abortion: Evidence of an issue evolution. American Journal of Political Science, 41, 718-737.

Alvarez

R. M.

Brehm

(2002). Hard choices, easy answers: Values, information, and American public opinion. Princeton, NJ: Princeton University Press.

Basinger

S. J.

Lavine

(2005). Ambivalence, information, and electoral choice. American Political Science Review, 99, 169-184.

Beck

P. A.

(1977). Partisan development in the postwar south. American Political Science Review, 71, 477-496.

Berinsky

Powell

E. N.

Schickler

Yohai

I. B.

(2011). Revisiting public opinion in the 1930s and 1940s. PS: Political Science & Politics, 44, 515-520.

Bianco

W. T.

Lynch

M. S.

Miller

G. J.

Sened

(2006). “A theory waiting to be discovered and used”: A reanalysis of canonical experiments on majority rule decision-making. Journal of Politics, 68, 837-850.

Black

(1992). The vital south: How presidents are elected. Cambridge, MA: Harvard University Press.

Black

(2002). The rise of the southern republicans. Cambridge, MA: Harvard University Press.

10.

Black

(2004). The transformation of the southern democratic party. Journal of Politics, 66, 1001-1017.

11.

Boyd

R. W.

(1972). Popular control of public policy: A normal vote analysis of the 1968 election. American Political Science Review, 66, 429-449.

12.

Breckler

S. J.

(1994). A comparison of numerical indexes for measuring attitude ambivalence. Educational and Psychological Measurement, 54, 350-365.

13.

Brown

(2003). Democratic senator Zell Miller endorses President Bush. Retrieved from http://transcripts.cnn.com/TRANSCRIPTS/0311/06/asb.00.html

14.

Burden

B. C.

Kimball

D. C.

(2002). Why Americans split their tickets. Ann Arbor: University of Michigan Press.

15.

Campbell

Converse

P. E.

Miller

W. E.

Stokes

D. E.

(1960). The American voter. New York, NY: John Wiley.

16.

Carmines

E. G.

Stimson

J. A.

(1986). On the structure and sequence of issue evolution. American Political Science Review, 80, 901-920.

17.

Carmines

E. G.

Stimson

J. A.

(1989). Issue evolution: Race and the transformation of American politics. Princeton, NJ: Princeton University Press.

18.

Carmines

E. G.

Woods

(2002). The role of party activists in the evolution of the abortion issue. Political Behavior, 24, 361-377.

19.

Carsey

T. M.

Layman

G. C.

(2006). Changing sides or changing minds? Party identification and policy preferences in the American electorate. American Journal of Political Science, 50, 464-477.

20.

Caughey

Berinsky

A. J.

Chatfield

Hartman

Schickler

Sekhon

J. S.

(n.d.) Population estimation and calibration weighting for nonresponse and sampling bias: Application to quota-sampled opinion polls, 1936-1953. Retrieved from https://www.dropbox.com/s/qnol9sffrb2bci1/CaugheyEtAl-Weights.pdf?dl=0

21.

Clark

J. A.

Bruce

J. M.

Kessel

J. H.

Jacoby

W. G.

(1991). I’d rather switch than fight: Lifelong democrats and converts to republicanism among campaign activists. American Journal of Political Science, 35, 577-597.

22.

Cohen

Karol

Noel

Zaller

(2008). The party decides: Presidential nominations before and after reform. Chicago, IL: The University of Chicago Press.

23.

Converse

P. E.

(1969). Of time and partisan stability. Comparative Political Studies, 2, 139-171.

24.

Glaser

J. M.

Gilens

(1997). Inter-regional migration and political resocialization: A study of racial attitudes under pressure. Public Opinion Quarterly, 61, 72-86.

25.

Green

D. P.

Palmquist

(1994). How stable is party identification? Political Behavior, 16, 437-466.

26.

Green

D. P

Palmquist

Schickler

(1998). Macropartisanship: A replication and critique. American Political Science Review, 92, 883-899.

27.

Green

D. P.

Palmquist

Schickler

(2002). Partisan hearts & minds: Political parties and the social identities of voters. New Haven, CT: Yale University Press.

28.

Greene

(1999). Understanding party identification: A social identity approach. Political Psychology, 20, 393-403.

29.

Greene

(2004). Social identity theory and party identification. Social Science Quarterly, 85, 136-153.

30.

Greene

W. H.

(2008). Econometric analysis (6th ed.). Upper Saddle River, NJ: Prentice Hall.

31.

Hanmer

M. J.

Kalkan

K. O.

(2013). Behind the curve: Clarifying the best approach to calculating predicted probabilities from limited dependent variable models. American Journal of Political Science, 57, 263-277.

32.

Hayes

McKee

S. E.

(2007). Toward a one-party south? American Politics Research, 36(3), 3-32.

33.

Heckman

J. J.

(1979). Sample selection bias as a specification error. Econometrica, 47, 153-161.

34.

Huckfeldt

Mendez

J. M.

Osborn

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

35.

Huckfeldt

Sprague

(2000). Political consequences of inconsistency: The accessibility and stability of abortion attitudes. Political Psychology, 21, 57-79.

36.

Huddy

(2001). From social to political identity: A critical examination of social identity theory. Political Psychology, 22, 127-156.

37.

Huddy

Bankert

(2017). Political partisanship as a social identity. In Thompson

(Ed.), Oxford research encyclopedia of politics. Oxford, UK: Oxford University Press. doi:10.1093/acrefore/9780190228637.013.250

38.

Huddy

Mason

Aarøe

(2015). Expressive partisanship: Campaign involvement, political emotion, and partisan identity. American Political Science Review, 109(1), 1-17.

39.

Jewell

R. D.

(2003). The effects of deadline pressure on attitudinal ambivalence. Marketing Letters, 14, 83-95.

40.

Karol

(2009). Party position change in American politics: Coalition management. New York, NY: Oxford University Press.

41.

Keele

Wolak

(2008). Contextual sources of ambivalence. Political Psychology, 29, 653-673.

42.

Key

V. O.

(1955). A theory of critical elections. Journal of Politics, 17(1), 3-18.

43.

Kimball

D. C.

Summary

(2012). Reconsidering the measure of partisan ambivalence. Paper prepared for presentation at the Annual Meeting of the American Political Science Association. Retrieved from https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2108218

44.

Kinder

D. R.

Sears

D. O.

(1981). Prejudice and politics: Symbolic racism versus threats to the good life. Journal of Personality and Social Psychology, 40, 414-431.

45.

Lavine

(2001). The electoral consequences of ambivalence toward presidential candidates. American Journal of Political Science, 45, 915-929.

46.

Lavine

Johnston

C. D.

Steenbergen

M. R.

(2012). The ambivalent partisan: How critical loyalty promotes democracy. New York, NY: Oxford University Press.

47.

Layman

G. C.

Carsey

T. M.

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

48.

Layman

G. C.

Carsey

T. M.

(2002b). Party polarization and party structuring of policy attitudes: A comparison of three NES studies. Political Behavior, 24, 199-236.

49.

Layman

G. C.

Carsey

T. M.

Green

J. C.

Herrera

Cooperman

(2010). Activists and conflict extension in American party politics. American Political Science Review, 104, 324-346.

50.

Lewis-Beck

M. S.

Jacoby

W. G.

Norpoth

Weisberg

H. F.

(2008). The American voter revisited. Ann Arbor: University of Michigan Press.

51.

Lindaman

Haider-Markel

D. P.

(2002). Issue evolution, political parties, and the culture war. Political Research Quarterly, 55, 91-110.

52.

Lublin

(2004). The republican south: Democratization and partisan change. Princeton, NJ: Princeton University Press.

53.

Mason

(2015). “I disrespectfully disagree”: The differential effects of partisan sorting on social and issue polarization. American Journal of Political Science, 59, 128-145.

54.

McGraw

K. M.

Hasecke

Conger

(2003). Ambivalence, uncertainty, and processes of candidate evaluation. Political Psychology, 24, 421-448.

55.

McKee

S. C.

Springer

M. J.

(2015). A tale of “two-souths”: White voting behavior in contemporary southern elections. Social Science Quarterly, 96, 588-607.

56.

Meffert

M. F.

Guge

Lodge

(2004). Good, bad, indifferent and ambivalent: The consequence of multidimensional political attitudes. In Saris

W. E.

Sniderman

P. M.

(Eds.), The issue of belief: Essays in the intersection of non-attitudes and attitude change (pp. 60-100). Princeton, NJ: Princeton University Press.

57.

Miller

Schofield

(2008). The transformation of the republican and democratic party coalitions in the U.S. Perspectives on Politics, 6, 433-450.

58.

Miller

J. M.

Peterson

D. A. M.

(2004). Theoretical and empirical implications of attitude strength. Journal of Politics, 66, 847-867.

59.

Mulligan

(2011). Partisan ambivalence, split-ticket voting, and divided government. Political Psychology, 32, 505-530.

60.

Mutz

D. C.

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

61.

Nadeau

Stanley

H. W.

(1993). Class polarization in partisanship among native southern whites, 1952-90. American Journal of Political Science, 37, 900-919.

62.

Newby-Clark

I. R.

McGregor

Zanna

M. P.

(2002). Thinking and caring about cognitive inconsistency: When and for whom does attitudinal ambivalence feel uncomfortable? Journal of Personality and Social Psychology, 82, 157-166.

63.

Nir

Druckman

J. N.

(2008). Campaign mixed-message flows and timing of vote decision. International Journal of Public Opinion Research, 20, 326-346.

64.

Noel

(2012). The coalition merchants: The ideological roots of the civil rights realignment. Journal of Politics, 74, 156-173.

65.

Noel

(2013). Political ideologies and political parties in America. New York, NY: Cambridge University Press.

66.

Norpoth

Rusk

J. G.

(1982). Partisan dealignment in the American electorate: Itemizing the deductions since 1964. American Journal of Political Science, 76, 522-537.

67.

Osborne

Sears

D. O.

Valentino

N. A.

(2011). The end of the democratic south: The impressionable years hypothesis. Political Psychology, 32, 81-108.

68.

Petrocik

J. R.

(1987). Realignment: New party coalitions and the nationalization of the south. Journal of Politics, 49, 347-375.

69.

Povey

Wellens

Conner

(2001). Attitudes toward following meat, vegetarian and vegan diets: An examination of the role of ambivalence. Appetite, 37(1), 15-26.

70.

Rainey

(2014). Arguing for a negligible effect. American Journal of Political Science, 58, 1083-1091.

71.

Rudolph

T. J.

(2005). Group attachment and the reduction of value-driven ambivalence. Political Psychology, 26, 905-928.

72.

Rudolph

T. J.

(2011). The dynamics of ambivalence. American Journal of Political Science, 55, 561-573.

73.

Rudolph

T. J.

Popp

(2007). An information processing theory of ambivalence. Political Psychology, 28, 563-586.

74.

Schickler

Pearson

Feinstein

B. D.

(2010). Congressional parties and civil rights politics from 1933 to 1972. Journal of Politics, 72, 672-689.

75.

Schofield

Miller

Martin

(2003). Critical elections and political realignments in the USA: 1860-2000. Political Studies, 51, 217-240.

76.

Sears

D. O.

Lau

R. R.

Tyler

T. R.

Allen

H. M.

Jr. (1980). Self-interest versus symbolic politics in policy attitudes and presidential voting. American Political Science Review, 74, 670-684.

77.

Sears

D. O.

Valentino

N. A.

(1997). Politics matters: Political events as catalysts for preadult socialization. American Political Science Review, 91(1), 45-65.

78.

Shufeldt

(2017). Party-group ambivalence and voter loyalty: Results from three experiments. American Politics Research, 46, 132-168. doi:10.1177/1532673X17705854

79.

Stanley

H. W.

(1988). Southern partisan changes: Realignment, dealignment, or both? Journal of Politics, 50, 64-88.

80.

Steenbergen

M. R.

Lavine

Goolsby

(2006). Reducing conceptual chaos: Distinguishing partisan ambivalence from partisan strength. Paper presented at the Annual Meeting of Midwest Political Science Association, April, Chicago, IL.

81.

Sundquist

J. L.

(1983). Dynamics of the party system (Rev. ed.). Washington, DC: Brookings Institution.

82.

Tajfel

(1978). Social categorization, social identity, and social comparisons. In Tajfel

(Ed.), Differentiation between social groups: Studies in the social psychology of intergroup relations (pp. 27-60). London, England: Academic Press.

83.

Tetlock

P. E.

(1986). A value pluralism model of ideological reasoning. Journal of Personality and Social Psychology, 50, 819-827.

84.

Thompson

M. M.

Zanna

M. P.

Griffin

D. W.

(1995). Let’s not be indifferent about (attitudinal) ambivalence. In Krosnick

(Ed.), Attitude strength: Antecedents and consequences (pp. 361-386). New York, NY: Lawrence Erlbaum.

85.

Thornton

J. R.

(2011). Ambivalent or indifferent: Examining the properties of measures of partisan ambivalence. Political Psychology, 32, 863-884.

86.

Thornton

J. R.

(2013). The impact of elite polarization on partisan ambivalence and indifference. Political Behavior, 35, 409-428.

87.

Valentino

N. A.

Sears

D. O.

(2005). Old times they are not forgotten: Race and partisan realignment in the contemporary south. American Journal of Political Science, 49, 672-688.

88.

Vella

(1998). Estimating models with sample selection bias: A survey. The Journal of Human Resources, 33, 127-169.

89.

Voting Rights Act. (1965). Public Law. 89-110; 79 Stat. 437.

90.

Wohlbrecht

(2000). The politics of women’s rights: Parties, positions, and change. Princeton, NJ: Princeton University Press.

91.

Zaller

(1992). The nature and origins of mass opinion. New York, NY: Cambridge University Press.

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

1.20 MB