Happy Winners, Sore Partisans? Political Trust, Partisanship, and the Populist Assault on Electoral Integrity in Mexico

Abstract

Winning elections usually make partisan voters more politically satisfied and confident. However, if they voted for a president that actively undermines the legitimacy of democratic institutions, they will be compelled to accommodate their views and update their judgment on a selective basis. They will support the regime's performance and yet distrust the institutions denounced by the government. This claim is tested using data from a representative survey conducted in Mexico. In this country, the president is a populist leader who consistently denounces all constraints on the executive. In particular, the president frequently undermines the institutions safeguarding free and fair elections. The analysis reveals that the gap in political trust reflects the opposite reactions from partisan winners and losers to the executive's antagonizing behavior. Voters supporting the winning party are more satisfied with democracy. However, they are less likely to trust the integrity of elections than the partisan losers.

Assuming that elected authorities would honor democratic values and comply with the rule of law is increasingly unwarranted. All over the world, authoritarian and illiberal candidates are rising to the executive office through competitive elections (Hellmeier et al., 2021; Inglehart and Norris, 2017; Levitsky and Ziblatt, 2018). Once in power, they are “walking the talk”—proceeding to dismantle democracy as swiftly as possible (Haggard and Kaufman, 2021; Lührmann et al., 2021). In doing so, they frequently attack those who oppose their attempts to concentrate power. In these circumstances, how do voters respond to illiberal rulers? More precisely, what are the consequences for political trust when the incumbent undermines the legitimacy of democratic institutions?

An increasing body of research shows that citizens’ commitment to democratic values and institutions depends on their partisan and political allegiances. Despite their alleged commitment to democracy in principle, some voters are willing to tolerate infringements to essential freedoms, the rule of law, or the integrity of elections when they feel represented by the executive. Provided their preferred politicians are in power, they might tolerate and even endorse reforms that could drive the regime into the path of autocratization (e.g., Cohen et al., 2022; Graham and Svolik, 2020; Mazepus and Toshkov, 2021; Moehler, 2009; Singer, 2018, 2021; Svolik, 2020).

This research further argues that political trust is decisively shaped by partisan identification with the illiberal executive. The central claim is that the winner-loser gap in political trust among partisan voters is contingent on the incumbent's commitment to the values and procedures underpinning the democratic regime. If the executive assails and undermines the legitimacy of democratic freedoms or institutions, the partisan voters will privilege their political allegiances when judging the regime's performance and will accommodate their views to reflect those of the government. As a consequence, political trust will exhibit a contradictory, counterintuitive pattern: the partisan winners will be more satisfied with the political status quo and yet distrust institutions that are crucial to democracy.

As a test of the argument, I provide an analysis of trust in Mexico's political institutions under Andrés Manuel López Obrador (AMLO), a populist president (Monsiváis Carrillo, 2020; Sarsfield, 2022). Drawing on data from a representative survey, I examine whether the winner-loser gap among partisan voters conforms to conventional expectations (Anderson et al., 2005). For once, it could be expected that voters identified with the winning party express trust and confidence in each one of the regime's institutions, from the presidency to the integrity of elections. They should, indeed, as López Obrador achieved a resounding victory in what is generally perceived as a fair and legitimate election. However, once in office, AMLO soon uncovered a hegemonic ambition by strategically attempting to override the rule of law and curtailing the checks and balances. Furthermore, deploying a populist worldview, Lopez Obrador assumes that any constraint on the executive authority is illegitimate while claiming that the people entrusted him with the mandate to deliver the regime's “fourth transformation (4T).”

In these circumstances, voters identified with National Regeneration Movement (MORENA), the incumbent party, need to accommodate the fact that the government actively incites polarization and attacks the democratic legitimacy of political pluralism and institutional constraints on the executive. For instance, the president constantly delivers scathing attacks against journalists and media outlets (Article 19, 2021). Indeed, one of the most consequential targets of the president is the National Electoral Institute (INE), an independent and highly professional election body management (EMB) that performed a decisive role in guaranteeing that the 2018 election was fair and clean. The president and his allies constantly claim that the EMB collaborated in committing “electoral frauds” in the past (e.g., López Obrador, 2022). They also contend that the INE is an expensive and “anti-democratic” body, serving a “conservative” agenda. ¹ The question is, then, if MORENA identifiers are satisfied with the political regime, will they trust the executive enough to mirror the official stance toward the INE and the integrity of elections?

This study contributes to current research on the relationship between partisanship and public support for the gradual assault on democracy launched by personalistic-populist executives. Partisanship is increasingly weak and volatile in Mexico—at best, half of the electorate identifies with political parties. However, the analysis suggests that partisanship plays a crucial role in the relationship between voters and political institutions in this country. The main finding is that the MORENA partisans and voters identified with opposition parties differ in their assessments of the functioning of democracy and the trustworthiness of political institutions, but not in a conventional way. Instead, the partisan gap in political trust matches the president's polarizing rhetoric. MORENA identifiers are the most satisfied with democracy and decidedly trustful in the presidency. At the same time, they distrust the INE and the judges and remain skeptical of the integrity of elections to a greater degree than the partisan losers. Even as nonpartisan citizens fall beyond the scope of the analysis, the evidence indicates that voters identified with the incumbent and opposition parties might respond to the executive in a manner that reflects their different stakes in the status quo. The partisan winners seem inclined to endorse the president's antagonizing speech, while the partisan losers are likely reaffirming their trust in institutions that constrain the executive's reach and guarantee fair elections.

The organization of this article is as follows: the first part provides the theoretical basis for the hypotheses. In this part, I will discuss the relationship between illiberal governments, partisanship, and political trust. The second part offers a more detailed account of the case study and the political context. Then, the third section describes the data, variables, and method. The fourth section presents and discusses the results. The main findings are summarized in the conclusions.

Political Trust and the Illiberal Assault on Political Institutions

Political trust is an essential element of political legitimacy (Booth and Seligson, 2009). As an attitude, it is “concerned with the political institutions that link overarching democratic principles to everyday political actors and policies” (van der Meer and Zmerli, 2017: 4). It is different from people's commitment to democratic principles but also from the approval of specific actors, such as elected authorities or political leaders (Norris, 2017). Typically, political trust reflects attitudes toward the core institutions of contemporary democracies, such as parliaments, the judiciary, governments, political parties, or elections. In addition, political trust is akin to satisfaction with democracy, as both types of attitudes reflect evaluations of regime performance (Martini and Quaranta, 2020; van der Meer and Hakhverdian, 2017).

Studies of political trust conventionally assume that the incumbent government is committed to democracy. However, as the third wave of autocratization “turns viral” (Hellmeier et al., 2021; Lührmann and Lindberg, 2019), the democratic credentials of elected governments should not be taken for granted. A distinctive feature of illiberal rulers is their deliberate animosity toward all forms of constraints on their power. At the core of illiberal politics lies the idea that popular sovereignty should know no boundaries and remain unchecked. Therefore, its implications are both anti-pluralist and anti-institutional—there is no place for disagreement or diverging political views, and the separation of powers and the rule of law as seen as illegitimate (Wagrandl, 2021: 110–113). Illiberal incumbents typically denounce the free press, opposition parties, independent political institutions, or electoral fairness in their quest to cement their rule.

The question is, how do citizens assess the performance of political institutions in these circumstances? What are the consequences for political trust when the sitting president constantly deprecates actors, procedures, or institutions? The literature suggests that citizens evaluate the regime's functioning on the basis of the quality of procedures and outcomes. Recent studies indicate that both actual and perceived economic performance increases satisfaction with democracy (e.g., Christmann, 2018; Claassen and Magalhães, 2021). The quality of political institutions, low corruption, and effective government, among others, equally strengthen political trust and democratic satisfaction (Christmann, 2018; Claassen and Magalhães, 2021; van der Meer and Hakhverdian, 2017). Moreover, the positive influence of favorable outcomes and institutional effectiveness is reinforced by perceptions of procedural fairness, education, and a proper democratic understanding (Magalhães, 2016; Mauk, 2021).

Nevertheless, while citizens could sincerely be motivated to express accurate judgments about the regime's performance, they are equally motivated by “directional” or partisan goals. According to research on partisan motivated reasoning, citizens will acquire and interpret new information guided by their partisan preferences and loyalties (Druckman, 2012: 200–202; Leeper and Slothuus, 2014; Taber and Lodge, 2012: 159–162). Partisan motivated reasoning is stimulated by political polarization, with voters forming their opinions and decision more on the basis of partisan endorsements and less on accurate information (Druckman et al., 2013). Citizens will follow partisan cues and express opinions that align with those of their preferred parties—for instance, regarding the economy's performance (Bisgaard and Slothuus, 2018: 458–459). Likewise, when party elites strongly disagree over politically relevant facts, voters adopt beliefs that do not simply reflect insincere cheerleading but their actual reasoning based on their partisan identification (Peterson and Iyengar, 2021).

Thus, in a situation in which the sitting executive denounces independent institutions or delegitimizes lawful dissent and opposition, it is highly plausible that the executive's polarizing claims would successfully resonate with certain groups of voters—those who share the president's partisan identity. In this study, my contention is that partisan voters will respond to the incumbent's delegitimization efforts depending on their status as winners or losers of elections. The contingent status of voters as electoral winners or losers performs a decisive role in shaping political support, policy preferences, and even attitudes toward institutional reform (Anderson et al., 2005; Bowler and Donovan, 2007; Curini et al., 2012; Martini and Quaranta, 2019).

In ordinary circumstances, winning elections turn voters into happy, trustful citizens. Losing produces the opposite effect—distrust and dissatisfaction (Anderson et al., 2005). However, when the incumbent government assails democratic institutions, the partisan winners are put in a difficult situation: they are supposed to be happy with the status quo but also embittered by those political actors or institutions reviled by the government. How are they supposed to react? Gathering independent, non-partisan information and then deciding whether the executive is right or wrong? Or just believe what their leaders are claiming? In the interest of democratic resilience, it certainly would be better if citizens were to assess the facts and then decide whether to trust institutions to a greater or lesser extent. Motivated by their partisan allegiances, it is more likely that they will accommodate their beliefs to reflect the executive's stance on the issue.

Current research shows that voters tend to prioritize their political allegiances and partisanship over their commitment to democracy. In Latin America, for instance, under certain conditions, both citizens and legislative elites in the winning camp agree to the overarching concentration of power in the executive or ignore declines in the quality of democracy (Singer, 2018, 2021). Similarly, after Jair Bolsonaro's election in Brazil, the winning voters displayed a renewed enthusiasm for democracy while also agreeing to bolster his clout by dismantling institutional checks on the executive (Cohen et al., 2022). The literature indicates, furthermore, that partisanship is crucial to explaining why voters are willing to trade off democratic transgressions for political representation. Svolik (2020: 5–6) demonstrates that the incumbents strategically induce their partisan supporters to tolerate undemocratic behavior when political polarization is high. Out of partisan loyalty, voters will condone the executive's transgressions to prevent the challenger from reaching power. Likewise, using experimental data, Graham and Svolik (2020) find that Americans are ready to trade off democratic principles when these principles conflict with their partisan loyalty, political ideology, or policy preferences. Along the same line, Mazepus and Toshkov (2021) gauge survey data from 26 European countries to show that partisan losers are more supportive of checks and balances. Conducting an experimental study in Ukraine, they also offer evidence that the partisan losers are more likely to oppose institutional reforms that might infringe on the independence of the judiciary.

In this study, I further argue the winner-loser gap in political trust among partisan voters will be conditioned by the executive's illiberal assailment of political institutions. When the executive seeks to subvert democracy by denouncing legitimate institutions, the winners will be less interested in finding out if those particular political institutions are indeed trustworthy or not. Instead, their partisan loyalties will inform their political judgment. Therefore, the winner-loser gap will display a counterintuitive pattern, driven by the winners’ contorting endeavors to keep up with the executive’ denunciation of political actors or institutions. The empirical hypotheses are, thus, the following:

H1:

The partisan winners will be more politically satisfied and trustful than losers on the condition that the executive remains neutral if not openly favorable towards the regime or political institutions.

At the same time, taking the government's side, the voters who support the incumbent party will distrust those institutions deemed reprehensible by the president. Therefore:

H2:

The partisan winners will be less politically satisfied and trustful than losers in those particular cases in which a concrete set of political processes, actors, or institutions are considered illegitimate or reprehensible by the executive and its political allies.

The Case Study: Mexico

The Mexican case is particularly well-suited to test the hypotheses. This is a case in which the partisan winners should express confidence in institutions and satisfaction with democracy after AMLO's resounding victory in the 2018 presidential election. At the same time, this is also a case in which the incumbent government is presided by a populist and personalistic leader, striving for the “aggrandizement of the executive” (Bermeo, 2016). Thus, such a populist government is expected to condition the partisan winners by undermining their confidence in the integrity of elections—one of the major foundations of the Mexican democracy.

Electoral Integrity and Public Opinion

It is worth underscoring that electoral integrity and the independence of the election management body (EMB) play a crucial role in the Mexican regime. Electoral integrity is the “agreed-upon international conventions and universal standards about elections reflecting global norms applying to all countries worldwide throughout the electoral cycle” (Norris, 2014: 21). Electoral management bodies are essential for electoral integrity. EMBs are legally responsible for “some or all of the elements that are essential for the conduct of elections and direct democracy instruments” (Catt et al., 2014: 5). These elements include voter registration, boundary delimitation, media monitoring, candidate validation, counting the votes, and many others.

Building an institutional framework for safeguarding free and fair elections was essential to Mexico's democratization (Eisenstadt, 2004; Schedler, 2005). The country's EMB, the Federal Electoral Institute (IFE), was created in the early 1990s but did not achieve complete independence until 1996. Almost two decades later, in 2014, a major constitutional reform turned the IFE into the National Electoral Institute (INE). The INE acquired new faculties and functions, additional to those already performed by the IFE. As an institution, the INE is generally regarded as a highly effective organization, capable of conducting elections in increasingly challenging circumstances.

Elections and the independence of electoral authorities have been a matter of bitter partisan contention ever since the country began its protracted transition to democracy and, more recently, since Felipe Calderon Hinojosa was declared the winner of the 2006 presidential election by a thin margin of 0.5 per cent over the runner-up, López Obrador (Eisenstadt, 2007; Estrada and Poiré, 2007). Analyzing legal reforms since 1996, Langston (2020) argues that the rules of electoral governance in Mexico are immersed in a non-cooperative game where party leaders constantly cheat the rules but demand fairness and a level playing field at the same time. This, in turn, forces party elites to negotiate new regulations to appease the loser's grievances in the short term. In the long term, all players will cheat again and cry foul when the outcome upsets their electoral expectations. Then, party elites will push for negotiated legal change, and the cycle of reforms and elections enters a new phase.

Mexican voters, on the other hand, react to electoral outcomes by staying congruent with their choices on the ballot. In each presidential election from 2000 to 2012, voters who endorsed the winning candidates were more satisfied and trustful in the IFE, electoral tribunal, or the quality of elections than those who supported the losers (Cantú and García-Ponce, 2015; Ugues Jr., 2010, 2018; Ugues Jr. and Medina Vidal, 2015). Remarkably, Lopez Obrador supporters were the most aggrieved after the 2006 and 2012 defeats and even arrived already skeptical of electoral integrity on election day in 2012 (Cantú and García-Ponce, 2015).

In addition to candidate preferences, party identification is typically correlated with voters’ attitudes toward the integrity of elections or the electoral authorities. In Mexico, partisanship is still unsettled and volatile, presidential elections are mostly candidate-centered, and campaigns are decisive in persuading even weak partisans (Greene, 2011). In 2018, for instance, López Obrador was elected with 53 per cent of the vote. However, his victory was largely driven by independent voters (Greene and Sánchez-Talanquer, 2018: 37). Against this background, some studies indicate that the partisan winners are consistently more trustful than the losers. The analysis in Ugues Jr. (2018) shows that voters identified with the winning party are more likely to perceive that the election was clean in either 2000, 2006, or 2012. At the same time, in the aftermath of the 2006 presidential contest, voters close to the Party of Democratic Revolution (PRD)—López Obrador's party back then—were distinctively less confident in the election and also more distrustful of the IFE and the electoral tribunal (Ugues Jr., 2018; Ugues Jr. and Medina Vidal, 2015: 236–238).

Thus, according to recent scholarship on the Mexican voter, partisanship is a crucial mechanism linking voters, electoral choices, and trust in the integrity of elections and electoral authorities. If anything, prior research supports the expectation that voters identified with the winning party are more satisfied and confident than the losing partisans. Unless, of course, the newly elected executive renovates his campaign against the established rules of electoral governance.

Populism and the Illiberal Assault on Political Institutions

The 2018 presidential election is generally regarded as an event of major political significance (Aparicio and Castro Cornejo, 2020). AMLO's triumph produced the third alternation in the executive office since the year 2000, and for the first time, a so-called leftist candidate would become president. Moreover, López Obrador won by a landslide with 53.2 per cent of the effective vote. No other presidential candidate had reached an absolute majority since the electoral democracy was established in the late 1990s. Furthermore, the president would command a robust and disciplined majority in both Chambers of Congress. The last time a president would enjoy a unified government was in 1997, when the Revolutionary Institutional Party (PRI) lost the Chamber of Deputies to the opposition.

In addition, the 2018 election was largely undisputed. The electoral authorities delivered a professional and effective performance, and the losing candidates conceded gracefully. Conversely, in two previous attempts, López Obrador lost the elections claiming to be the victim of electoral fraud. In 2006, he alleged that the “mafia in power” stole the election, refused to concede, and emphatically said “go to hell” to the electoral institutions. Running again in 2012, he lost to Enrique Peña Nieto by six points (39.17 vs. 32.41 per cent) and, for the second time, López Obrador alleged fraud and blamed the electoral authorities for his defeat.

In ordinary circumstances, after such a decisive victory, the partisan winners would have every reason to be happy and trustful of political institutions. However, under the new government, the voters will have to accommodate the assailing of political institutions by a charismatic and personalistic president. If anything, AMLO's inauguration in 2018 marks the beginning of an administration commanded by a personalistic-populist leader driven by an illiberal view of democracy. According to the ideational approach, the distinctive feature of populism is a Manichean discourse portraying politics as the permanent struggle between good and evil—the former incarnated by an idealized and homogenous notion of the people, the latter by an equally undifferentiated corrupt elite (Hawkins and Rovira Kaltwasser, 2017). López Obrador discourse is prototypically populist, embracing a worldview in which the conspiring elite—typically, the “Mafia in power”—incarnates the political evil while the “wise” and “honest” people represent all desirable virtues (Monsiváis Carrillo, 2020; Sarsfield, 2022).

In addition to his populism, AMLO is a personalistic leader who believes that his personal charisma and individual qualities are sufficient to legitimize his political initiatives and decisions (Tannenberg et al., 2020: 79)—thus living up to his early portrayal as the “Tropical Messiah” (Krauze, 2006). In Latin America, populist and personalistic presidents are often precursors of episodes of democratic backsliding (Balderacchi, 2018; Levitsky and Loxton, 2013). Indeed, personalistic presidents are more likely to erode the horizontal mechanisms of accountability or the rule of law when they command a large legislative majority by a weakly institutionalized political party (Rhodes-Purdy and Madrid, 2020). When AMLO took office, his government commanded a supersized majority in Congress and the loyalty of MORENA. Even after becoming Mexico's dominant party in 2018, MORENA remains weakly institutionalized, disconnected from social groups, and incapable of balancing the President (Bruhn, 2021).

Once in power, AMLO launched a political transformation wrapped in populist discourse that essentially aims to the illiberal concentration of authority in the presidency. One of the core ideas of illiberalism is that popular sovereignty should remain unlimited (Wagrandl, 2021). On account of his electoral legitimacy, López Obrador claims to embody the sole source of popular representation and democratic authority. Allegedly, his government would restore the sovereignty of the people after more than three decades of “neoliberalism” and the “evil of corruption” through the “fourth transformation.” Consequently, political pluralism is considered illegitimate and a threat to the “nation” and the president's “transformative project. ² ”

On the other hand, illiberal rulers reject all sorts of institutional constraints. The Mexican president's illiberalism is manifest in his stern opposition to public accountability and checks on the executive power. From the onset of his administration, AMLO strived to centralize power by curtailing most constraints on the executive while fostering polarization and displaying an open disregard for the rule of law. ³ In this regard, Dussauge-Laguna (2021) shows that AMLO's populist “doublespeak” has increased his control over the public bureaucracy and weakened the administrative capacities of the state at the same time. On behalf of “republican austerity” and fighting “corruption,” he has strived to subordinate the Supreme Court and undermine the constitutional autonomy of several state agencies (Reveles Vázquez, 2021). The president has also undermined political pluralism and tolerance by portraying all sorts of opposition and public criticism as illegitimate threats to the alleged “transformation” he claims to represent. Sánchez-Talanquer and Greene (2021: 68–69) argue that the government's ability to fulfill citizens’ demands is paradoxically hampered by the enduring weakness in state capacity and highlight the risk involved in the president granting increasing concessions to the military. Moreover, the government consistently disseminates misleading information, launches deliberate attacks against the media and the free press, and remains impassive when journalists become the victims of criminal violence (Article 19, 2021).

In particular, the populist undermining of democratic institutions is directed at the integrity of elections. Ever since his defeat in the 2006 presidential election, López Obrador has accused the electoral authorities of being complicit in fraud and corruption. However, the president's animosity toward the INE and the rules safeguarding free and fair elections has increased during his term. On every occasion in which compliance with electoral regulations is at stake, López Obrador scathingly attacks the electoral authorities and threatens to impose overarching reforms (e.g., Reina, 2022). Each time, the president's plan involves curtailing the independence of electoral authorities and handing over control of election management to the executive—as was customary under the competitive-authoritarian regime.

Indeed, AMLO's most recent proposal of constitutional reform substantially undermines the current system of electoral governance by, among other things, turning the INE into a different body and having the electoral councilors and magistrates elected by the public. ⁴ Furthermore, the president and his allies are willing to “walk the talk” by blatantly breaching electoral regulations and defying the electoral authorities (Manetto, 2022). At every opportunity, the incumbent government and the partisan elite in Congress, governorships or mass media accuse the electoral authorities of illegitimate interests, demand their removal, and pledge loyalty to the president.

Data, Measurement, and Method

The empirical analysis draws on the National Survey of Civic Culture 2020 (ENCUCI 2020) (INEGI, 2021). ENCUCI 2020 is a representative survey of the national population in México older than 15 years. It was produced by a collaboration between the INE and the National Institute of Statistics and Geography (INEGI). The sampling method uses a probabilistic, three-stage, stratified, and clustered sample design. The survey was conducted through face-to-face interviews between August 17 and September 18, 2020, during the Covid-19 pandemic. The database comprises 21,519 valid observations.

Dependent Variables

The analysis focuses on three sets of dependent variables. The first variable is an indicator of intermediate political support—Satisfaction with democracy, which provides a general assessment of the public's evaluation of the regime's performance. The questionnaire replicates previous studies asking: “How satisfied are you with the democracy we have in Mexico today?” The question is presented to those interviewees who had previously declared knowing or “having heard what democracy is.” Satisfaction with democracy is measured as a four-point ordinal variable as are the other dependent variables.

The second set of variables examines trust in the core institutions of contemporary liberal democracies plus the electoral management body: Trust in the Judges, Trust in the President, Trust in Senators and Deputies, and Trust in the INE. ENCUCI 2020 is one of the few publicly available surveys that explicitly inquire about confidence in the National Electoral Institute under Lopez Obrador's administration.

The third set gauges perceptions of electoral integrity and malpractice (Norris, 2019). Perceptions of electoral integrity measure citizens’ beliefs that the votes are counted impartially, or that election officials are fair. Perceptions of electoral malpractice tap on voters’ beliefs about fraud or electoral manipulation: Opposition candidates are prevented from running; voters are bribed; rich people buy election; voters receive threats at the polls; and the government channels resources to the parties it favors. All the dependent variables were reversed. The highest value (4) indicates the maximum presence of the attribute and the lowest (1) signals the minimum.

Independent and Control Variables

The independent variable, party identification, distinguishes between partisan winners, partisan losers, and independent voters. First, I select the subset of individuals who declare having voted in 2018 (83.6 per cent). Then, voters are classified according to their partisan identification. Those who “feel close” to MORENA are coded as the “partisan winners” (26.8 per cent). MORENA is AMLO's political offspring—a “social movement” established as a political party in 2014 that relies on his founder's charisma and informal leadership. Before MORENA's formal recognition as a political party in 2014, AMLO's partisan supporters were primarily congregated in the Party of the Democratic Revolution (PRD). Thereafter, MORENA became the main source of party identification among AMLO's followers. ⁵

The “partisan losers” or “opposition partisans,” on the other hand, are those who self-identify with any of the three main political parties in the opposition (26.9 per cent): The National Action Party (PAN), the Revolutionary Institutional Party (PRI), and the Party of the Democratic Revolution (PRD). In 2018, PAN and PRD formed an alliance to compete in the presidential race. It should be noted that a substantial proportion of the electorate does not identify with political parties at all (46.2 per cent). These voters are classified as “independents.”

In addition, the analysis controls for those variables that are important correlates of political trust: interest in politics, interpersonal trust, and socio-demographic indicators such as sex, age group, level of education, employment, speakers of an indigenous language, monetary resources in the household, and the macro-region where the interview was conducted. See the Appendix for more information regarding the question wording and descriptive statistics.

Method

The analysis is performed using ordinal-logistic regression models for survey data. The analysis stratifies by urban, urban-complement, or rural attributes of the locality (“dominio”) and the subnational state (“entidad”) where the survey was conducted. The survey weight is also applied (“fac_sel”). If clustered robust standard errors or ordinary standard errors are used instead, the results are almost equivalent. Under these specifications, the independent variable is statistically significant in two additional tests (specifically, models 3 and 6 in Table 1 below). These results are in line with the theoretical expectations, but regression models for survey data provide a more stringent test of the hypotheses by estimating standard errors that produce wider confidence intervals. Thus, I adopt a more conservative stance and discuss the results from the regression models for survey data.

Table 1.

The Partisan Gap in Political Trust in Mexico (2020).

[Figure omitted. See PDF]

Variables	Models
	(1)	(2)	(3)	(4)
	Satisfaction with Democracy	Trust in the Presidency	Trust in the Senators and Deputies	Trust in the Judges
Party identification
MORENA identifiers	0.998***	1.431***	−0.064	−.441***
	(0.078)	(0.069)	(0.070)	(0.064)
Independents	−1.135*	−0.0004	−0.600***	−0.557***
	(0.068)	(0.058)	(0.060)	(0.060)
Control variables	Yes	Yes	Yes	Yes
Observations	10,144	13,915	13,617	13,341
	Models
	(5)	(6)	(7)	(8)
	Trust in the INE	Votes are Counted Fairly	Election Officials are Fair	Opposition Candidates are Banned
Party identification
MORENA identifiers	−0.461***	−0.109	−0.160*	−0.061
	(0.070)	(0.062)	(0.063)	(0.067)
Independents	−0.609***	−0.395***	−0.321**	−0.118
	(0.062)	(0.057)	(0.056)	(0.062)
Control variables	Yes	Yes	Yes	Yes
Observations	13,906	13,585	13,437	12,873
	Models
	(9)	(10)	(11)	(12)
	Voters are Bribed	The Rich Buy Elections	Voters Receive Threats	The Government Sponsors Political Parties
Party identification
MORENA identifiers	0.179**	0.459***	0.009	−0.050
	(0.067)	(0.069)	(0.065)	(0.066)
Independents	0.174**	0.126	0.021	0.153*
	(0.063)	(0.064)	(0.059)	(0.062)
Control variables	Yes	Yes	Yes	Yes
Observations	0.179**	0.459***	0.009	−0.050

Note: Ordinal-logistic regression models for survey data. Linearized standard errors in parentheses. Regression coefficients are estimated by taking the partisan losers as the reference category. Control variables are included in the analysis but not shown. The complete results are available in the Appendix. *p < 0.05, **p < 0.01, ***p < 0.001.

Results and Discussion

According to conventional expectations, voters identified with the winning party should be more politically satisfied than the partisan losers across all dimensions under scrutiny (e.g., Anderson et al., 2005). The results displayed in Table 1 suggest, however, that such expectations do not hold among voters identified with MORENA—the incumbent party since 2018. The partisan winners conform to the expectations in just two out of twelve instances—they are more satisfied with democracy (model 1) and express more confidence in the presidency (model 2). However, Morenistas and voters identified with the opposition parties are equally trustful in their representatives in Congress (model 3). Likewise, partisan voters share the same views about the vote count, the opposition candidates’ right to compete, violence at the polls, or the unlawful biases sponsored by the government, regardless of their party identification. Nevertheless, conventional expectations fail to explain the partisan gap in crucial variables—trust in the judges (model 4), trust in the INE (model 5), perceptions about election officials (model 7), vote-buying (model 9), and the rich people buying elections (model 10). In all these cases, the expected relationship is inverted—those who are more likely to trust the EMB, the judiciary, and the quality of elections are voters identified with the opposition parties.

At first sight, such a pattern of political trust makes little sense. Why would Morenistas be satisfied with democracy and yet distrust the EMB or the integrity of elections simultaneously? These results are difficult to explain unless the populist assailing of political institutions is accounted for. The first hypothesis predicts that partisan supporters of the sitting president will be more satisfied with democracy and political institutions as long as the executive avoids spreading accusations against the political regime or specific political institutions. The data reveals strong support for the president and the belief that democracy is performing satisfactorily among MORENA identifiers—Figure 1 illustrates this point. On the other hand, the second hypothesis predicts that the executive's animosity toward independent and countervailing democratic institutions will motivate his partisan supporters to modify their political attitudes. The partisan winners will be no longer unconditionally supportive of political institutions and instead reproduce the incumbent's hostility. Accordingly, as depicted in Figure 1, voters identified with MORENA are less enthusiastic about the judges and the EMB than voters identified with the opposition parties.

Figure 1.

Party Identification and Political Trust in Mexico. Note: Predicted Probabilities and 95% Confidence Intervals Based on Estimations from Table 1. All Covariates at Their Means.

[Figure omitted. See PDF]

The second hypothesis is also helpful for making sense of the partisan gap in perceptions of electoral integrity. Despite being the winners in 2018, MORENA identifiers seem persuaded that electoral malpractice persists. Their attitudes, however, largely coincide with the president's point of view. For instance, the incumbent's supporters are more likely to believe that “the rich buy elections” than voters identified with the opposition parties (model 10). This is a finding that coincides with the AMLO's anti-elitist rhetoric and recurrent praise of the ordinary people (Sarsfield, 2022). Morenistas are also more inclined to believe that “election officials are unfair” (model 7), which is also consistent with the president's stance on the electoral authorities. Conversely, MORENA partisans express more ambiguous opinions when the official position is unclear or undefined. López Obrador rarely denounces the increasing violence in elections or alludes to the persisting shortcomings in monitoring campaign finances. Moreover, the President hardly complains about the vote count anymore. During the post-election conflict in 2006, AMLO's followers demanded a recount “vote by vote”—“voto por voto, cassilla por casilla.” In 2018, the tally favored the President-elect by over 17 million votes over the runner-up. Interestingly, Morenistas do not seem particularly disturbed by the vote count, threats of violence at the polls, or the government channeling funds to the ruling party.

In synthesis, the results suggest that MORENA partisans are satisfied with the current state of affairs, except that their attitudes coincide with the presidential reluctance to trust the electoral authority or the integrity of elections. Indeed, the partisan winners are satisfied with democracy and trust the president but distrust the honesty of the administrative body that played an essential role in safeguarding the integrity of the vote when AMLO was elected.

The main issue with the results discussed so far is that they are also compatible with alternative models of attitude formation. That is, alternative explanations need to be addressed and discarded before assuming the hypotheses are supported by the evidence. For instance, the results might be driven by an enduring stream of political discontent preceding AMLO's polarizing discourse in the presidency. The sources of political disaffection in Mexico run deep and remain unchallenged. López Obrador has instigated political disbelief and disparaged electoral institutions since his first defeat in 2006. Cantú and García-Ponce (2015) demonstrate that AMLO's supporters in 2012 were already skeptical of the honesty of the 2012 election weeks ahead of the voting day. Moreover, in 2018, AMLO capitalized on the electorate's widespread discontent and channeled the antisystem urge of many voters in his third bid for the presidency (Castro Cornejo et al., 2020; Greene and Sánchez-Talanquer, 2018). Therefore, it could be the case that Morenistas have always been skeptical of the INE and other institutions and so they remain.

On the other hand, the outcomes are also compatible with an attitudinal swing among opposition identifiers. The results in model 4 align with this alternative explanation, as PAN, PRI, and PRD partisans are more likely to trust the judges than Morenistas. Voters identified with the political opposition may expect the judiciary to reinforce the limits on the executive power. This finding coincides with recent research showing that voters identified with the losing parties or candidates are willing to oppose or reject the executive's effort at dismantling the separation of power or crucial checks on his authority (Mazepus and Toshkov, 2021; Singer, 2021). If that is the case, it is important to disentangle the presidential effects on the partisan winners from his influence on the opposition identifiers. Otherwise, the impact of the presidential populism on the partisan winners could be overestimated or even wrongly assumed to be true when, in fact, the outcome is produced by the partisan losers’ attitudinal shift.

Now, it is difficult to test the alternative explanations with the available data. In order to estimate the attitudinal swing among Morenistas, it would be necessary to assess their political attitudes against a baseline. Ideally, the analysis should be performed using panel data, which are unavailable. As an alternative, I performed a set of additional tests leveraging data from several cross-sectional surveys. ⁶ Although the surveys’ sampling methods, item availability, and question-wording substantially diverge from one project to the other, they provide valuable information to assess the likelihood that the results follow from the main hypotheses instead of the alternative explanations.

In the first place, I evaluate the possibility that the results overestimate the effect of the government's delegitimatizing efforts directed at the electoral authorities on MORENA identifiers. The alternative hypothesis is that Morenistas’ attitudes toward elections or the INE remained the same before and after López Obrador was elected president. In such a case, they should be unmoved by their newly acquired status as partisan winners. I begin the analysis by examining the partisan gap in trust in elections using AmericasBarometer's data. According to the results, from 2008 to 2017, PRD and MORENA identifiers reacted as partisan losers, distrusting elections to a more significant degree than other partisans. ⁷ Yet, early in 2018, the World Values Survey 7 (WVS-7) uncovers a different pattern. ⁸ The analysis reveals that MORENA identifiers did not differ from other voters’ attitudes toward elections, the INE, or several indicators of electoral integrity and malpractice. As the election date approached, MORENA partisans were as confident in electoral fairness as anybody else.

Furthermore, a few weeks after the 2018 election, Morenistas turned into happy winners. ⁹ They believed in the trustworthiness of the outcome and were also a bit more likely to think that the election was free and fair than the partisan losers. The AmericasBarometer's data also indicates that the most remarkable shift in trust in elections between 2017 and 2019 took place on the side of voters identified with MORENA. Early in 2019, as López Obrador's government was recently inaugurated, Morenistas were significantly more trustful in elections than in 2017. The ENCUCI 2020 data also indicates that Morenistas’ cheerful mood toward elections began to recede as AMLO's term advanced. By mid-2020, the partisan gap was already reversed—see Table 1 above. Opposition partisans already expressed more trust in INE and electoral fairness than MORENA voters.

The evidence suggests that attitudes of trust among MORENA identifiers did vary before and after AMLO was elected. Still, MORENA partisans seem somewhat skeptical of elections, even after AMLO's inauguration. Surely, their attitudes improved in 2019 relative to 2017 or the PRD identifiers in earlier elections, but only as much to reach the level of trust among the opposition identifiers. ¹⁰ Therefore, if Morenistas became mistrustful of elections again to stay congruent with the presidential stance, it is likely that their move toward election distrust was relatively modest since they were already less confident.

Likewise, it could be the case that executive's influence on his partisan adepts might be overestimated. The president might be driving the opposition voters to modify their attitudes as well. Therefore, I examine the alternative hypothesis that trust in elections among opposition identifiers is not affected by López Obrador's populism. The expectation is that the partisan losers display similar levels of trust in elections at different moments during AMLO's administration. To perform the analysis, I rely again on data from the AmericasBarometer. However, the 2021 AmericasBarometer survey in Mexico did not field questions gauging party identification, making a strict contrast between partisan voters unfeasible. Consequently, I focus on citizens’ evaluation of AMLO's performance and contrast those who approve with those who disapprove of his government.

The analysis reveals that AMLO's supporters are as confident in elections in 2019 as in 2021. ¹¹ As would be expected, they are also more trustful in elections than citizens who disapprove of the executive's performance. However, there is something unusual among citizens dissatisfied with the government. Not only are they more confident in elections in 2021 than in 2019, but they are also more trustful in 2021 than in any previous year in the series. By the year 2021, the presidential animosity was already explicit and persistent. The findings thus suggest that the executive's opponents were sensitive to his government's antagonizing of electoral authorities as they were more inclined to trust the elections. ¹²

The results from the additional tests provide evidence aligned with the theoretical expectations but also suggest some caution in the interpretation of the data. The analysis reveals that the partisan winners did modify their political attitudes in order to reproduce the polarizing presidential speech. However, the evidence also indicates that trust in elections is not simply driven by motivated reasoning on the part of MORENA identifiers. On the one hand, it is likely that attitudinal change among Morenistas is somewhat moderate because they never were particularly enthusiastic about electoral integrity even after AMLO was elected. On the other hand, attitudinal updates among the opposition identifiers seemingly perform a crucial role as well. Therefore, the picture portraying partisan winners as the only ones accommodating their judgment is likely incomplete. Provided that current data limitations prevent estimating the exact magnitude of attitudinal change in both Morenistas and opposition identifiers, future research should collect the appropriate data to assess partisan-motivated reasoning among Mexican voters.

What are the substantive implications of these findings? The most remarkable fact uncovered by the analysis is the partisan gap in trust in the INE, integrity of elections, and judges that mirrors the presidential animosity toward countervailing and independent political institutions. AMLO's illiberal assault on electoral authorities or the checks and balances is arguably reflected in the relationship between partisanship and trust in political institutions. However, it is important to note that the size of the observed gap in trust in the INE is moderate relative to the divergence in satisfaction with democracy or trust in the presidency. ¹³ That is, Morenistas are decidedly more satisfied with democracy or trustful in the executive than opposition identifiers, and although they distrust the INE or judges more than their opponents, the difference is smaller. Moreover, it should be underscored that public trust in Mexico's EMB is fairly robust. According to ENCUCI 2020 data, 61 per cent of the electorate trust the INE either somewhat or a lot. Confidence in the electoral authority is resilient even among MORENA identifiers (58%), and AMLO's unprecedented 53 per cent of the vote in 2018 is still behind the proportion of the electorate feeling confident in the Mexican EMB. ¹⁴ Likewise, public support for the INE is substantially higher than the vote collected by MORENA in 2018 (37%) or the 2021 midterms (34%).

Finally, as the findings are circumscribed to party identifiers, we do not know if nonpartisan AMLO supporters are reacting in a similar manner to partisan winners. Still, the independent voters deserve more systematic attention. The independents hold negative views of the judges, the INE, or the elections, much like MORENA identifiers. However, they do not share Morenistas’ enthusiasm for the status quo. Nonpartisans are as dissatisfied with democracy or the president as the opposition identifiers. Data limitations prevent testing if their attitudes of distrust and disaffection respond to López Obrador's populist illiberalism or simply remain the same as before. In any case, political disaffection among the independents could also represent a significant threat to the stability of democracy. Nonpartisans might support the incumbent's efforts to override the rule of law and the separation of powers out of anti-system attitudes or the urge to expedite political change.

Conclusions

The partisan gap in political trust usually follows a consistent pattern: the partisan winners are more satisfied than the losers. This research contends that such a pattern is contingent on the government's commitment to democracy and respect for political institutions. When the incumbent government actively undermines the legitimacy of specific political actors or institutions, voters rely on their partisan identification to decide whether to trust those institutions or not. Consequently, the partisan gap might display an unusual pattern, as the gap in trust in political institutions and electoral integrity in Mexico illustrates.

After AMLO's victory in the 2018 presidential election, partisan winners had every reason to be satisfied with democracy and its core institutions. Capitalizing on widespread political discontent, their candidate had achieved unqualified electoral success. The winning coalition had even attained a supersized majority in both Chambers of Congress. However, the electorate was soon exposed to AMLO's populist discourse. In particular, Lopez Obrador's polarizing speech addressed the INE, the country's electoral management body—an institution that performed a key role in guaranteeing that the 2018 national election was free and fair. The president consistently denounced the INE for allegedly being complicit in electoral frauds in the past and resisting the political transformation pursued by his government.

The argument in this study asserted that AMLO's populist illiberalism—the attempt to override limits to executive power while claiming to act on behalf of the people—induced partisan-motivated reasoning among MORENA identifiers, the partisan winners. Morenistas were expected to update their attitudes in order to match the presidential view. The evidence reveals that Morenistas and opposition partisans diverge in their assessments of the regime performance and trustworthiness of political institutions. The observed partisan gap, furthermore, is congruent with the executive's animosity. MORENA identifiers are clearly more satisfied with the regime's performance and trustful in the president than the partisan losers. Conversely, opposition identifiers are more confident in the INE, the integrity of elections, or the judges than Morenistas. However, the analysis suggests that the results could not be explained by attitudinal shifts among MORENA voters alone. AMLO's efforts at corroding public trust in the INE or electoral integrity is likely pushing opposition partisans in the opposite direction as well, as they are presumably reaffirming their confidence in elections or the institutions constraining executive power.

At the same time, it is worth noting that trust in the INE remains fairly robust despite the partisan divergences uncovered in the analysis. In addition, the analysis is circumscribed to partisan voters in a country where partisanship is weak and volatile. The independent voters fall beyond the scope of the findings, but deserve more systematic scrutiny as they are mostly dissatisfied and disaffected. In any case, the available data leaves some crucial questions unanswered and thus invites further research on partisanship and political trust in order to provide a more accurate assessment of the attitudinal shift in Mexico.

All things considered, this research offers evidence that partisan assessments of political institutions in Mexico match the president's illiberal view of democracy and populist animosity. As such, it contributes to a growing and important literature studying the relationship between partisanship, political support, and democratic decline instigated by populist or illiberal rulers. Recent studies demonstrate that voters tolerate the incumbent government's gradual attempts at the subversion of democracy as a consequence of their partisan allegiances or political identification (Cohen et al., 2022; Graham and Svolik, 2020; Singer, 2018). Similarly, other studies reveal that the partisan losers are aware of the stakes involved in the incumbents’ attempt to aggrandize executive power and are willing to stand up for democracy (Mazepus and Toshkov, 2021; Singer, 2021). In turn, this study provides evidence that MORENA and the opposition identifiers are sensitive to the incumbent government's antagonizing relationship with electoral authorities and checks and balances. Indeed, as the government's efforts to overhaul Mexico's electoral governance and other constraints on the executive have only increased over time, the partisan gap should be expected to widen as well. Therefore, further studies addressing partisan and nonpartisan reactions to AMLO's polarizing discourse and behavior could improve our current understanding of attitudinal change and political support in Mexico and elsewhere.

Supplemental Material

sj-docx-1-pla-10.1177_1866802X221136147 - Supplemental material for Happy Winners, Sore Partisans? Political Trust, Partisanship, and the Populist Assault on Electoral Integrity in Mexico

Supplemental material, sj-docx-1-pla-10.1177_1866802X221136147 for Happy Winners, Sore Partisans? Political Trust, Partisanship, and the Populist Assault on Electoral Integrity in Mexico by Alejandro Monsiváis-Carrillo in Journal of Politics in Latin America