Partisan dealignment in Mexico

Several studies show that partisanship is of central importance to understanding patterns of political engagement of the Mexican electorate. Similar to their counterparts in many old and new democracies (^{Gunther et al., 2016}; ^{Nadeau et al., 2017}), Mexican partisans tend to vote for the party with which they identify (^{Moreno, 2009}; ^{Somuano Ventura, 2014}; ^{Somuano Ventura & Ortega Ruiz, 2003}). They are also more likely than independents to vote on election day (^{Buendía & Somuano, 2003}; ^{Moreno, 2009}; ^{Temkin Tedwab & Flores-Ivich, 2014}), to be interested in public affairs and share political information, and to become involved in campaign and protest activities (^{Somuano Ventura, 2014}). However, like many other democracies, Mexico has experienced a profound process of partisan dealignment in recent decades (^{Díaz Jiménez, 2019}; ^{Moreno, 2018}). The proportion of partisan citizens decreased from approximately three-quarters of the electorate in 1997 to only half in 2018 (Figure 1).

Sources: 1997: ITAM-Arcop survey reported in ^{Moreno (2009}, ²⁰¹²⁾; 1998-2009: annual averages obtained from Reforma Newspaper surveys reported in ^{Moreno (2009}, ²⁰¹²⁾. For 2010 onward, the annual averages of the electoral surveys of Buendia & Laredo, reported in ^{Díaz Jiménez (2019)}, are used.

Figure 1 Partisanship and partisan independence in Mexico, 1997-2018 

The significant increase in the share of independent voters has had various consequences for elections and electoral behavior, largely because independents seem to be more likely than partisans to delay their voting decision (^{Díaz Jiménez, 2019}) and to be more susceptible to media and campaign effects (^{Greene, 2015}). Therefore, it is important to understand the relevant characteristics of this segment of the electorate beyond its electoral behavior. Among these characteristics are the levels of political engagement or disaffection.

Partisanship and cognitive mobilization in Mexico

The relationship between partisanship and CM and the consequences of this relationship for political culture and citizen participation in the Mexican case constitute a relevant but developing research agenda (^{Meixueiro, 2014}; ^{Ortega et al., 2011}; ^{Somuano Ventura, 2014}; ^{Temkin Yedwab et al., 2008}). Some studies, for example, suggest that the proportion of cognitive partisans and apartisans increased from 2006 to 2012 while the percentage of apoliticals and ritual partisans decreased during the same period (^{Meixueiro, 2014}; ^{Somuano Ventura, 2014}). However, such studies do not provide evidence of significant contrasts in the level of political engagement between groups with high and low degrees of CM. And although some previous analyses have documented relevant differences in some civic attitudes and modes of political participation among the cog-partisan groups (^{Temkin Yedwab et al., 2008}), there are other possible significant between-group differences in political orientations that have not yet been sufficiently explored.

This research seeks to contribute to the study of the consequences of the loosening of partisan attachments and growing cognitive mobilization that has occurred throughout the country in recent decades for patterns of political engagement and disaffection among Mexicans. On the one hand, it is of particular interest to examine whether this dealignment (and the consequent decline in parties’ mobilization capacity) may foster a significant decrease in the attitudinal components of political engagement and a possible increase in the levels of political disaffection among the population. On the other hand, it is also important to examine whether CM is able to gradually take the place of partisanship as a mobilizing factor of civic attitudes among the population.

Political engagement and political disaffection as dependent variables

Political engagement is a complex and multidimensional concept that refers to the active involvement of citizens in politics and public affairs (^{Delli Carpini, 2004}; ^{Putnam, 1993}). Certainly, the most visible aspects of political engagement are the actions of citizens “intended directly or indirectly to affect the selection of elected representatives and/or the development, implementation, or enforcement of public policy through government” (^{Delli Carpini, 2004, p. 397}) (behavioral dimensions). However, it is also true that the set of attitudes, values, resources and skills that encourage and enable the political participation of individuals, including their levels of political interest, knowledge, efficacy and trust (attitudinal dimensions), are also relevant components of the concept (^{Delli Carpini, 2004}; ^{Moy & Hussain, 2011}; ^{Verba et al., 1997}).

Among the various political attitudes that favor citizen participation in contemporary democracies are political knowledge and efficacy. The first is understood as the information that citizens acquire and remember about the various actors, processes and institutions of the political system, and the second is understood as “the feeling that individual political action does have, or can have, an impact upon the political process, i.e., that it is worthwhile to perform one’s civic duties” (^{Campbell et al., 1954, p. 187}). Political efficacy is also a significant predictor of various forms of political participation (^{Dalton, 2020}). The specialized literature distinguishes between internal efficacy, understood as the degree of confidence that citizens have in their abilities to understand politics and participate effectively in it (^{Craig et al., 1990}), and external efficacy, which refers to citizens’ perception of the capacity and willingness of government and political institutions to respond to their demands and needs (responsiveness) (^{Balch, 1974}; ^{Campbell et al., 1960}; ^{Converse, 1972}). Another relevant civic attitude is political trust, which “refers to citizens’ assessments of the core institutions of the polity” (^{Zmerli et al., 2007, p. 41}). More specifically, this attitude “entails a positive evaluation of the most relevant attributes that makes each political institution trustworthy, such as credibility, fairness, competence, transparency in its policy-making, and openness to competing views” (Levi & Stoker, in ^{Zmerli et al., 2007, p. 41}).

Political disaffection is also a difficult concept to define and operationalize (^{de Vreese, 2005}; ^{de Vreese & Semetko, 2002}). It has often been defined in opposition to attitudes of political interest, efficacy and trust. In general, it may be defined as “a certain estrangement of members of the polity from both its core political institutions and, more generally, from politics” (^{Gunther & Montero, 2006, p. 49}; ^{Gunther et al., 2007, p. 33}) and as the “subjective feeling of powerlessness, cynicism, and lack of confidence in the political process, politicians, and democratic institutions” (^{Torcal & Montero, 2006, p. 6}).

Based on previous research (^{Maldonado Hernández, 2013}; ^{Torcal, 2003}; ^{Torcal & Montero, 2006}; ^{Yamamoto & Kushin, 2014}), this study identifies five different (though interrelated) dimensions of political disaffection: political disengagement, institutional disaffection, political cynicism, political apathy and political skepticism. Political disengagement refers to a set of “attitudes related to a general distrust of politics and to the respondent’s lack of engagement with the political process” (^{Torcal & Montero, 2006, p. 6}). And institutional disaffection focuses on “beliefs about the lack of responsiveness of political authorities and institutions, and citizens’ lack of confidence in the political institutions” (^{Torcal & Montero, 2006, p. 7}). A related concept is political cynicism, which is also defined as a generalized distrust of the political system (^{Austin & Pinkleton, 1995}, ¹⁹⁹⁹⁾. Cynical citizens tend to view politics as personally irrelevant; therefore, they generally refrain from participating in the political process (^{Austin & Pinkleton, 1995}, ¹⁹⁹⁹; ^{Crotty & Jacobson, 1980}). Political apathy is also another relevant component of disaffection and is defined as indifference, lack of interest and/or lack of attention toward politics (^{Bennett, 1986}).

Both political cynicism and apathy are usually considered components of disaffection with negative implications for citizens’ democratic engagement, since they can reinforce each other in a spiral of political disaffection (^{Austin & Pinkleton, 1995}, ¹⁹⁹⁹; ^{Pinkleton & Austin, 2004}). However, as ^{Yamamoto and Kushin (2014)} note, not all dimensions of political disaffection have negative consequences for democracy; some of them, such as political skepticism, may even have some positive effects (^{Austin & Pinkleton, 1995}). Skepticism can be defined as a lack of trust in the political process but without the refusal to participate in it (^{Cappella & Jamieson, 1997}). In contrast to political apathy and cynicism, skepticism stimulates the search for, verification of and contrasting of information from various channels of political communication (such as traditional and social media); therefore, it may be considered as a more positive aspect of disaffection (^{Austin & Pinkleton, 1999}; ^{Pinkleton & Austin, 2001}).

Research questions

Drawing on the literature on cognitive mobilization and political disaffection, the following research questions are posed:

According to theories of CM, the erosion of partisan loyalties does not necessarily result in a decline in civic attitudes since the possible loss of the mobilizing effect of partisanship in the behavioral and attitudinal components of political engagement could be gradually compensated by an increase in the levels of CM, particularly among younger generations. Similarly, it is also expected that growing levels of CM could limit the increase in the negative dimensions of political disaffection, such as political cynicism and apathy, as proposed in the following hypotheses:

Design and sample

To test the above hypotheses, multivariate statistical analyses were carried out using the postelection wave of a panel study, a survey design widely used in studies of public opinion, political communication and political behavior (^{Eveland & Morey, 2011}; ^{de Vreese & Semetko, 2002}; ^{Iyengar & Vavreck, 2012}).³

Dependent variables

A number of indicator scales of various attitudinal dimensions of political engagement were constructed, drawing on the political engagement-disengagement continuum (based on political knowledge and internal political efficacy) and the political engagement-institutional disaffection continuum (based on external political efficacy and institutional trust). I also developed attitudinal scales of specific dimensions of political disaffection, such as political cynicism, apathy and skepticism.

Independent and control variables

Cognitive-Partisan Index.^{Dalton’s (1984}, ²⁰¹³, ²⁰²⁰⁾ procedure was followed by combining indicators of respondents’ levels of education and political interest to create an additive index of CM. After that, I classified panel participants into two groups according to their levels of CM (low and high) and then combined both groups with the presence or absence of a party identification declared by respondents in the first wave of the panel study. Participants in the study were finally classified into four groups: cognitive partisans (n = 245, 35%), ritual partisans (n = 125, 17.8%), apartisans (n = 173, 24.7%) and apolitical independents (n = 156, 22.3%). Additionally, indicator scales of political information consumption (in traditional and social media) and political conversation (offline and online) were also developed, as these are often relevant predictors of civic attitudes, according to previous studies on the mobilization effects of mass media in Mexico (^{Díaz Jiménez, 2017}; ^{Díaz Jiménez & Muñiz, 2017}). Age and income scales and a dummy variable for gender were also developed as control variables.

Political engagement or disengagement?

I first examine the political engagement-political disengagement continuum. It should be noted that positive standardized regression coefficients in Table 2 indicate that the variable in question is positively associated with that particular dimension of citizens’ political engagement while negative coefficients indicate that the variable is related to political disaffection. In general, the results of the models provide evidence of a more relevant role of CM than of partisan identification as a predictor of political orientations among the population, given that, unlike ritual partisans, cognitive partisans and apartisans showed higher levels of internal political efficacy than apolitical independents (Table 2).

In other words, in the case of this first dimension, specifically regarding the internal efficacy component, it can be concluded that cognitive partisans and apartisans are significantly more politically engaged (effective) than apolitical independents, even after controlling for a number of social and political information consumption variables. Therefore, CM clearly favors political engagement rather than disengagement among citizens. It is important to note that other variables in the model, such as attention to political information in traditional media and interpersonal and interactive political conversation, also contribute to increasing citizen engagement. However, it is also noteworthy that neither partisanship nor CM have any significant impact on political knowledge.

Political engagement or institutional disaffection?

On the other hand, regarding the political engagement-institutional disaffection continuum, partisanship proved to be a more relevant predictor of the observed variation in the levels of political trust than CM. The regression analysis results indicate that cognitive and ritual partisans have significantly higher levels of institutional trust than apartisans and apolitical independents, even after controlling by social background and political information consumption variables.

The findings are similar to those of previous comparative studies of Latin America (^{Cisneros, 2020}) showing that it is partisan identification and not CM that has a more relevant role in fostering citizens’ trust in political institutions. However, it is worth mentioning that consumption of political information in traditional and social media was also a strong predictor of political trust among citizens. This finding lends support to theories of CM that emphasize the mobilization effects of political communications on citizens’ political involvement (^{Norris, 2000a}, ^2000b). In the case of external political efficacy, partisanship and CM are both significant predictors of this civic attitude, but its impact is negative. Cognitive partisans show significantly lower levels of external efficacy than apolitical independents; interestingly, the latter group shows the highest degree of this variable.

Political cynicism, apathy and skepticism

Regarding the analysis of the three dimensions of political disaffection as dependent variables of the study, the results of regression models show that, with the exception of political apathy, neither partisanship nor CM seem to be relevant explanatory variables of such orientations. There is no statistically significant a ssociation b etween cog-partisan categories and political disaffection dimensions while controlling for the other variables in the model. Only age is positively and significantly associated with political cynicism, and the only significant predictor of political skepticism is offline political conversation (Table 3).

However, in the case of political apathy, CM is a relevant explanatory variable since both cognitive partisans and apartisans seem to be significantly less apathetic than apolitical independents. Therefore, the results suggest that along with age, variables such as high levels of CM and political information consumption in social media contributed to significantly reducing the levels of political apathy of the Mexican electorate in 2018 (Table 3). However, an interesting finding is that high levels of online political conversation seem to be significantly associated with greater political apathy among citizens.

It should be noted that, in general, the results of the analyses suggest that except for political apathy, the effects of partisanship and CM are rather limited as an explanation for political disaffection among Mexicans. Moreover, the R squared values of the models are very low in all cases, which suggests that even in the case of political apathy, partisanship and CM have a modest role in explaining the total observed variation in these three dimensions of disaffection.