English (pdf)
Article in xml format
Article references
Automatic translation
Send this article by e-mail
Cited by SciELO 
Cited by Google 
Similars in
SciELO 
Similars in Google 
Permalink
Introduction
Trust in elections is an important expression of faith in contemporary democratic institutions. Voters usually follow partisan cues or react to their status as winners or losers in electoral contests to decide whether elections are trustworthy or not (Maldonado and Seligson 2014; Cantú and García-Ponce 2015; Ugues Jr. 2018). Voters are also more inclined to trust elections when the electoral processes are conducted with honesty and fairness. If elections are marred with fraud or manipulation, citizens will grow detached and skeptical (Birch 2008; McAllister and White 2015; Fortin-Rittberger et al. 2017; Wellman et al. 2018; Norris 2019; Mauk 2020).
This research argues that an informed understanding of democracy matters for trust in elections, albeit in a conditional manner. An informed understanding of democracy entails that democracy is defined as a regime in which citizens have equal political rights and civil freedoms, elections are free and fair, and the government is bound by the rule of law (Welzel 2013; Cho 2014; Moreno and Welzel 2014; Shin and Kim 2018; Wegscheider and Stark 2020). An increasing body of research shows that an accurate notion of democracy is correlated with important political outcomes, from support for democracy to political trust (Cho 2014; Flesken and Hartl 2018; Mauk 2021). Along this line, the main contention in this study is that citizens’ belief in the trustworthiness of elections will not only depend on the quality of the electoral cycle, but also on how they understand democracy. An informed notion of democracy should strengthen the positive relationship between perception of electoral integrity and trust in elections. Symmetrically, being informed about democracy should enhance the negative correlation between perceptions of electoral malpractice and confidence in elections.
An informed notion of democracy could be particularly relevant for trust in elections in regimes where manipulation or malpractice has a limited reach, but the integrity of the electoral cycle is still less than complete. As this study will argue, this is the case of trust in elections in Mexico. For several decades, the Mexican regime was a competitive authoritarian regime in which elections were systematically manipulated by the ruling party (Levitsky and Way 2010). Authoritarian elections receded after a protracted and often contentious dynamic of political transition. In this case, establishing a network of independent and professional institutions of electoral governance was critical to advancing democratization (Eisenstadt 2004; Schedler 2005; Magaloni 2006).1
Ever since competitive elections became “the only game in town,” the organization of national elections in Mexico abide by international standards (OAS 2018; V-Dem 2019, 58). As an illustration, the 2021 midterm elections were clean and inclusive despite considerable logistical challenges and significant political shortcomings (OEA 2021). Over the years, nonetheless, voter confidence in elections and electoral authorities have declined steadily, puzzling scholars and observers alike (Ortega and Somuano 2015, 60). Ortega and Somuano (2015, 4) note that mistrust in the electoral management body contrasts with frequent technical and logistic adaptations safeguarding elections. Similarly, Sonnleitner (2016, 52) aptly describes what he dubs the “Mexican paradox”: “Never before in this country were elections as free and competitive, plural and trustful, delivering so many peaceful alternations at every level of government. And yet, […] elections are now facing a deep crisis of legitimacy.” For many voters, it is as if elections were still plagued with fraud and manipulation.
Focusing primarily on the study of a single country, this article aims to contribute to the growing scholarship on citizens’ understanding of democracy and its behavioral and attitudinal consequences in Latin America. Specifically, it attempts to bridge studies on the public’s conceptions of democracy with research on electoral integrity and election legitimacy in the countries of the region. It also expects to cast a new light on the “Mexican paradox.” The main findings suggest that an informed notion of democracy help voters become more sensitive to the integrity of the electoral cycle and thus more likely to trust elections.
This study is organized as follows: The first section provides a review of the Mexican case to frame the analysis. The second advances the theoretical argument and formulate the hypotheses. In the third part, the study’s data, variables, and method are described. The following section examines and discusses the results, replicating the analysis in a group of Latin American countries. All arguments and findings are summed up in the conclusions.
Trust in Elections: The Mexican Paradox
In Mexico, elections are conducted under an increasingly sophisticated framework of electoral governance. Yet, many voters seem to believe that authoritarian elections are still in place.
It could be argued that partisan divisions frequently shadow major organizational and logistic advances safeguarding the integrity of elections. In this case, a bitter partisan conflict following the 2006 presidential election eroded trust in the electoral outcomes-denouncing fraud, the losing candidate infamously claimed, “to hell with your institutions” (Bruhn and Greene 2007; Eisenstadt 2007). Moreover, several studies confirm that voters’ electoral choices in the 2006 and 2012 presidential contests influenced their attitudes towards the IFE, the TEPJF, and the electoral outcome (Cantú and García-Ponce 2015; Ugues Jr. and Medina Vidal 2015; Sonnleitner 2016; Ugues Jr. 2018).
On the other hand, even when substantial progress in safeguarding democratic elections has been made over the last two decades, crucial challenges persist (Montes de Oca 2018; OAS 2018; Méndez de Hoyos et al. 2019, 9). To begin with, the de facto independence of electoral authorities is frequently contested. More often than not, the selection of the IFE-INE’s board of experts is a highly controversial process, whereby political parties seek to designate their own gatekeepers while also retaining in practice the power to dismiss councilors (Estévez et al. 2008; Langston 2020).2 Likewise, subnational electoral authorities diverge in their level of formal independence and professionalism (Méndez de Hoyos 2013; Salazar Elena and Flores-Ivich 2013). More commonly, they function as arenas of partisanship, allowing the persistence of local electoral-authoritarian enclaves (Cantú 2014).
Another significant challenge to the integrity of elections comes from the weak enforcement of the rule of law. According to Casar and Ugalde (2018), the average cost of a campaign in gubernatorial races is ten times higher than the legal limit. In consequence, large sums of illicit money flow into national and local elections as parties and candidates divert substantial amounts of public funds, reach for illegal private contributions, or accept funding from organized crime. At the same time, drug cartels and criminal organizations increasingly participate in the electoral arena. Ponce (2016) finds that electoral competition at the local level is eroded by violence and crime related to drug trafficking, as well as repressive measures adopted by Mexican authorities. Similarly, Ley (2017) shows that the strategic use of violence by organized crime groups during electoral campaigns reduces electoral participation-voters who live in regions where organized crime was engaged in high-profile violence were more reluctant to participate. While mostly considered a success, the 2018 national and local elections were also deemed “exceptionally violent” (OAS 2018, 87).
In any case, electoral malpractices are not as pervasive or systematic as to override the democratic quality of the electoral cycle. Why then are Mexican voters largely disinclined to trust elections? This study will argue that public understanding of democracy deserves consideration when studying the relationship between perception of electoral integrity and trust in elections.
Theory and Hypotheses
2.1. Electoral integrity and trust in elections
The idea of electoral integrity captures the notion that there is a set of standards that elections must fulfill to be recognized as democratic. While there is little doubt that the democratic quality of elections varies worldwide, many disagreements persist regarding how to precisely define and measure electoral integrity (van Ham 2015). Following Norris (2014, 21), electoral integrity is defined as “agreed-upon international conventions and universal standards about elections reflecting global norm applying to all countries worldwide throughout the electoral cycle.” In this definition, the electoral cycle encompasses all rules and activities advanced during the pre-electoral period, the campaign, on polling day, and its aftermath. Thus, electoral integrity is achieved when all the activities during the electoral cycle are professionally designed and performed, and all the rules are clearly defined and enforced. In many circumstances, the efforts aimed at improving the quality of elections are hindered by different types of malpractice. From minor administrative failures or inadequate regulations to major attempts to distort or commit fraud, the menu of electoral manipulation is assorted with a wide range of options (Schedler 2002).
Current research suggests that voters are more likely to believe in the trustworthiness of political institutions, including elections, when the electoral cycle safeguards inclusion, freedom, and fairness. For instance, in a group of post-communist societies, McAllister and White (2015) find that citizens’ views about electoral integrity significantly impact their satisfaction with democracy. Similarly, Norris (2019) finds that perceptions of integrity and malpractice are correlated with the level of satisfaction with democracy across the globe and explain distrust in the outcome of the 2016 election in the United States. Birch (2008) equally reveals that proportional electoral systems and the public funding of parties have positive impacts on confidence in elections.
Along the same line, Maldonado and Seligson (2014) show that the quality of democracy reduces the winner-loser gap in trust in elections in Latin America. Likewise, Fortin-Rittberger et al. (2017) demonstrate that high levels of electoral fraud at the country level reduce satisfaction with democracy, while increasing electoral manipulation and malpractice reduces the winner-loser gap in voter satisfaction with democracy; at the same time, electoral malpractice produces a lesser effect on satisfaction in third-wave democracies. Furthermore, using a harmonized array of surveys from Asia, Europe, and Latin America, Mauk (2020) finds that perceptions of electoral fairness mediate in the relationship between electoral losers and political trust, producing an indirect negative effect on trust in political institutions. Such an effect, nonetheless, is contingent on the country’s level of electoral integrity.
2.2. Understanding of democracy
The relationship between democratic attitudes and the evaluation of the quality of elections is far less studied. Some scholars argue that people who express favorable views of democracy are more likely to manifest positive opinions of the electoral management bodies (EMBs). For instance, Ugues Jr. and Medina Vidal (2015, 234) assert that “an acculturation toward liberal democratic values” should be correlated with more positive assessments of electoral authorities. Yet, after assessing public opinion in the aftermath of the 2006 and 2012 contests, Ugues Jr. and Medina Vidal (2015) report mixed results. People who believe that Mexico is a democracy produce positive evaluations of electoral authorities, but only if the analysis excludes the perception that elections are clean.
However, in the study of citizen attitudes towards democracy, it is often overlooked that people define and understand “democracy” in different ways. Comparative research indicates that citizens differ in the normative and institutional criteria they use to give meaning to democracy (Canache 2012; Landwehr and Steiner 2017; Quaranta 2018; Osterberg-Kaufmann et al. 2020). At the same time, many citizens are unaware of the attributes that adequately define the regime, and not everyone is able to distinguish democracy from authoritarian rule (Shin and Kim 2018). In Latin America, for instance, incoherent conceptions of democracy are widespread (Schedler and Sarsfield 2009; Carlin and Singer 2011).
People may associate democracy with many valuable ideals and policy outcomes, but the hallmark of a proper conceptualization is connected with the normative and institutional underpinnings of liberal democracy (Moreno and Welzel 2014, 62-63). Following the established scholarship in political science, a liberal-democratic regime involves equal freedoms of expression and association, universal suffrage, free and fair elections, and the rule of law to protect individuals against encroachments from the state (e.g., Dahl 1971; Coppedge 2012; Munck 2016; Lührmann et al. 2018). Likewise, an informed understanding of democracy implies that people associate the notion of “democracy” with civil and political freedoms, free and fair elections, and the checks and balances that subject state authorities to the rule of law (Cho 2014; Moreno and Welzel 2014; Shin and Kim 2018; Mauk 2021). An informed understanding of democracy also requires not to confuse authoritarianism and democracy (Shin and Kim, 2018). Citizens should be able to discard from their notion of democracy all elements contradicting democratic values that usually belong in authoritarian systems.
2.3. The argument
The central claim in this research is that the understanding of democracy reinforces the relationship between people’s assessment of the quality of the electoral cycle and their confidence in elections. It should be noted that the expected relationship between the understanding of democracy and trust in elections is not direct but conditional. People could distrust either the electoral process or its outcomes regardless of whether they properly understand democracy or not. Nevertheless, people should trust elections to a greater or lesser degree depending on how their understanding of democracy conditions their evaluation of the electoral process.
To assess the influence of the understanding of democracy, it is necessary to examine, first, the relationship between perception of electoral integrity and trust in elections. This study postulates that perceptions of electoral integrity and electoral malpractice are correlated with trust in elections. When voters perceive, for instance, that election officials are fair and professional, media coverage of the campaigns is balanced, or all parties compete in a level playing field, they will be more likely to express trust in elections. However, if they experience violence at the polls, learn that some voters are being coerced or bribed, or believe that money is biasing the electoral campaigns, their attitudes toward elections as a political institution will be unfavorable. Therefore, trust in elections will result from voters’ assessment of integrity and malpractice during the electoral cycle. The first set of hypotheses are the following:
H1: Perception of integrity in the electoral cycle increases trust in elections.
H2: Perception of malpractice in the electoral cycle reduces trust in elections.
On the other hand, following the analysis of Flesken and Hartl (2018) of perceptions of electoral integrity on the basis of motivated reasoning, an informed understanding of democracy should motivate citizens to consider the integrity of elections more carefully. Motivated reasoning describes how individuals process information in different ways, depending on their goals or motivations. Particularly, accuracy goals motivate people to reach correct conclusions based on the available information (Flesken and Hartl, 2018: 709). An informed understanding of democracy should, therefore, enhance citizens’ motivated reasoning about the integrity of elections.
This study argues that people who acknowledge the democratic significance of citizens’ rights and freedoms, the rule of law, and clean elections might also be more appreciative of the quality of elections. On the basis of motivated reasoning, citizens who are able to identify the defining features of democracy should also be more motivated to correctly identify evidence of integrity or malpractice during the electoral cycle. However, it should be noted that citizens who are properly informed about democracy are not necessarily expected to trust elections. The relationship between an informed understanding of democracy and trust in elections is contingent on how citizens judge the electoral process.
Voters who accurately understand democracy might express trust in elections to a greater or lesser degree depending on the actual integrity of the electoral cycle. If voters believe that elections are conducted with a certain level of integrity, their trust in elections will be conditioned by their understanding of democracy. If they do not correctly understand democracy as a regime of rights and freedoms, their trust in elections will depend solely on their evaluation of the integrity of the electoral process-all else being equal. Instead, if they accurately understand democracy, their trust in elections will increase. Thus, an accurate understanding of democracy shifts in a positive direction the relationship between perception of electoral integrity and trust in elections. A similar reasoning applies to the perception of electoral malpractice, albeit in a negative direction. Voters are more likely to distrust elections if they are aware of pervasive malpractice. However, they will distrust elections even more if they understand democracy-if they do not, their distrust will only depend on their assessment of electoral manipulations.
The main empirical expectations of the study are thus expressed in this set of hypotheses:
H3: An informed understanding of democracy positively conditions the relationship between perception of electoral integrity and trust in elections.
H4: An informed understanding of democracy negatively conditions the relationship between perception of electoral malpractice and trust in elections.
Data and Method
The hypotheses outlined above are tested using the 7th wave of the World Value Survey (WVS-7) (Haerpfer et al. 2020). Mexico’s WVS-7 is a representative survey of the national voting-age population (over 18 years) collected from January to May 2018, using a multi-stage probability sampling design. The dataset comprises 1,741 face-to-face interviews conducted in 496 polling points in the country’s 32 federal states.
3.1. Trust in elections
The dependent variable, Trust in elections, is a four-point ordinal variable measuring “(h)ow much confidence do you have in the elections?” The coding of the original variable was reversed. It now runs from “Not at all” (=1) to “A great deal” of confidence (=4). As illustrated in Figure 1, trust in elections was hardly impressive at the time of the survey. People who expressed “Quite a lot” or “A great deal” of trust cover barely a fifth of the Mexican electorate (20%), while those who did not trust the elections “at all” represent half of the voting-age population (50%). In a group of nine Latin American countries included in the WVS-7, only voters from Guatemala, Colombia, and Peru express less trust in elections than Mexico (see Table A3 in the appendix).
3.2. Independent variables
The first set of independent variables are Perception of electoral integrity and Perception of electoral malpractice. Following Norris (2019)’s, perception of electoral integrity taps on “how often” citizens believe “Votes are counted fairly,” “Journalists provide fair coverage of elections,” “Election officials are fair,” “Voters are offered a genuine choice in the elections,” and “Women have equal opportunities to run for office.” All items are summed up into an additive index (alpha= 0.6029) and rescaled from 0 to 1. The same procedure was applied to produce an indicator of perception of electoral malpractice (alpha= 0.6495). The upper value signals that elections are very often tainted by misconduct and manipulation: “Opposition candidates are prevented from running,” “TV news favors the governing party,” “Voters are bribed,” “Rich people buy the elections,” and “Voters are threatened with violence at the polls.”
On the other hand, understanding of democracy assesses whether an informed understanding of democracy modifies the relationship between perception of election integrity and trust in elections. In this regard, the study followed Cho (2014)’s procedure-which is similar to Welzel’s (2013). Leveraging a set of items asking how much a series of attributes are essential for democracy, first, it summed up three items that form a liberal notion of democracy: “people choose their leaders in free elections,” “civil rights protect people from state oppression,” and “women have the same rights as men.” Additionally, to assess whether citizens are able to discard authoritarian features from their idea of democracy (Shin and Kim 2018), it reversed the following three items: “religious authorities ultimately interpret the laws,” “the army takes over when the government is incompetent,” and “people obey their rulers.” Finally, an additive indicator was formed comprising both liberal and non-authoritarian views of democracy, transforming it into a 0 to 1 scale (alpha=0.6273). The resulting variable measures the prevalence of the liberal-electoral notion of democracy in the Mexican population.
The mean value of the independent variables in the Mexican case reveals that perceptions of electoral malpractice (0.61) surpass perceptions of electoral integrity (0.49) and indicates that the average understanding of democracy barely reaches the middle point of the scale (0.49).
3.3 Control variables
The analysis controls for several alternative explanations, starting with voters’ partisan alignments as a proxy for voters’ identification with the incumbent or the opposition parties (e.g., Cantú and García-Ponce 2015; Ugues Jr. and Medina Vidal 2015). First, the analysis controls for voters’ identification (or lack thereof) with a political party. Then, it includes life satisfaction, an important predictor of political support (Esaiasson et al. 2020), as well as relevant correlates of political trust, such as interest in politics or ideology. The analysis also contemplates a set of socio-demographic indicators, including sex, age in years, educational attainment, income, and urban-rural residence.
The descriptive statistics of all independent and control variables are shown in Table A1 in the appendix. Figures A1 to A3 describe the distribution of the independent variables.
3.4. Method
The analysis models the dependent variable-a four-point ordinal variable-using ordinal logit regressions. The survey design comprises 31 strata-one for each of the 31 federal states included in the sample-, 496 primary sampling units, and the sample’s weight.3 Linearized standard errors are used to account for the complex survey design. If robust standard errors are used instead, the results are equivalent.
Results and Discussion
4.1. Results
The main empirical expectation in this study is that voters will be to a greater or lesser degree inclined to trust elections as a result of their perceptions of both electoral integrity and malpractice, and the conditional role of the understanding of democracy. First, the main effects of the independent variables are tested. Table A2 in the appendix displays the correlation coefficients between the dependent and independent variables. Further details concerning the robustness of the analysis are addressed after discussing the regression’s results.
The first model (M1) confirms that perception of electoral integrity increases the likelihood of trusting elections (H1). However, the second model (M2) reveals that perception of electoral malpractice produces no effect whatsoever on the dependent variable (H2). Furthermore, as expected, the direct relationship between the understanding of democracy and trust in elections is not systematic. In M1 and M2, the sign of the coefficient is positive but statistically non-significant. If voters do understand what democracy means, it does not make in itself any difference in their willingness to trust elections. (see table 1).
Table 1. The conditional role of the understanding of democracy on trust in elections
Source: Own elaboration. Note: Ordered logistic models. Linearized standard errors in brackets; * p<0.05, ** p<0.01, *** p<0.001.
The second step in the analysis addresses the conditional relationship between understanding of democracy and the independent variables. The results are displayed in models M3 and M4. The results in M3 indicate that the interaction between understanding of democracy and perception of electoral integrity is statistically significant (H3). In turn, M4 reveals that the impact of perception of electoral malpractice on the outcome is not systematically modified by an accurate understanding of democracy (H4). The results in M5 uphold these findings: the understanding of democracy systematically interacts with perception of electoral integrity, but not with perception of electoral malpractice. An informed understanding of democracy was expected to boost the negative influence of electoral malpractice, but the study evidence suggests that this is not the case.
In models M3 and M5, the coefficient of understanding of democracy is negative and statistically significant. It must be noted that the negative coefficient does not indicate an inverse relationship between understanding of democracy and trust in elections-such a relationship is already conditioned by the interaction between perception of electoral integrity and understanding of democracy. Instead, the negative coefficient indicates that understanding of democracy is negatively correlated with trust in elections when perception of electoral integrity is at its lowest value. When people believe that electoral integrity is absent, they will distrust the elections to a greater extent if they understand democracy. I will further discuss this finding below.
4.2. Discussion
The empirical analysis suggests that the conditioning role of the understanding of democracy matters for trust in elections. However, as it turns out, such a conditional role specifically matters when people assess the integrity of elections.
In line with the expectations of H3, the understanding of democracy reinforces the positive relationship between perception of electoral integrity and trust in elections. As people’s understanding of democracy becomes more accurate, their perception of electoral integrity improves the probability that they will trust elections. As an illustration, see Figure 2 which shows predicted probabilities that citizens will trust elections when perception of electoral integrity interacts with an informed understanding of democracy, keeping all covariates at their means.
Source: Own elaboration
Figure 2. Trust in elections: Predicted probabilities Note: The conditional effects on probability and their 95% confidence intervals (CI) are based on estimations from M5 in Table 1, keeping all covariates at their means.
In the hypothetical case that voters perceive no sign of electoral integrity, the probability of feeling “Not at all” trust in elections increases from 0.47 (CI 95%: 0.28-0.66) to 0.85 (CI 95%: 0.74-0.96), as understanding of democracy changes from 0 to 1 (see Panel A in Figure 2). Likewise, if the electoral cycle is conducted with complete honesty and fairness, voters who understand democracy will be more likely to express “Quite a lot” trust in elections (0.32, CI 95%: 0.32-0.42), or even “A lot of trust” (0.28, CI 95%: 0.11-0.45), than those who do not properly understand democracy (0.17 and 0.05, respectively).
Furthermore, it seems that even a somewhat confused understanding of democracy is necessary to reinforce the positive influence of perception of electoral integrity on trust in elections. For example, at a score as low as 0.37 in understanding of democracy, which marks the 25th percentile of the electorate, the interaction between understanding of democracy and perception of electoral integrity already produces a statistical difference in the “Not at all” and “Quite a lot” categories. Moreover, when the score for understanding of democracy is 0.58 (the 75th percentile), those who perceive integrity will already be more likely to express “A lot of trust” in elections.
The conditioning role performed by the understanding of democracy is also statistically significant if the analysis is performed on the items that form the index of perception of electoral integrity (see Table A3 in the appendix). The likelihood that citizens believe in the trustworthiness of elections increases when they find that votes are counted fairly, journalists provide fair coverage of elections, election officials are fair, or women have equal opportunities to run for office, on the condition that they are fully aware of the definition of democracy. There is only one item in which the interaction coefficient is not statistically significant- “Voters are offered a genuine choice in the elections.”
Additionally, when the analysis evaluates the interaction between understanding of democracy and perception of electoral integrity, the coefficient of the understanding of democracy term becomes negative and statistically significant. It should be emphasized that this does not mean that an increase in the understanding of democracy produces a decrease in the probability of observing trust in elections. It means that the relationship between understanding of democracy and the dependent variable becomes negative when citizens might perceive no-electoral integrity whatsoever. Those who understand democracy are less likely to express confidence in elections if they do not find a single trace of honesty or fairness in the electoral process (see Panel B in Figure 2).
On the other hand, the analysis disconfirms the relationship predicted by H2 and H4. The results indicate that perception of electoral malpractice is uncorrelated with trust in elections. When further examining such a relationship, the data reveal that some of the items that constitute the index of perception of electoral malpractice are, indeed, negatively associated with trust in elections (Table A4). People who believe that voters are bribed, rich people buy the elections, or feel threatened by violence at the polls, are less likely to trust elections-as would be expected. In one case, there is no relationship (“TV news favors the governing party”). In another, the relationship is positive: people who believe that “Opposition candidates are prevented from running” will trust the elections. In any case, none of the separate attributes of electoral malpractice is statistically conditioned by understanding of democracy.
These results suggest that voters do not need to be properly informed about democracy to become disappointed by concrete instances of electoral manipulation, such as violence at the polls, vote-buying, or the rich influencing the outcomes. Voters are equally appalled by specific instances of malpractice, whether they understand democracy or not. At the same time, the analysis reveals that trust in elections is primarily fueled by perceptions of electoral integrity, especially when people are aware of the definition of democracy.
These findings help cast the “Mexican paradox” in a new light. As described earlier, such a paradox arises from enduring attitudes of public distrust toward elections, despite the country’s independent and highly professional EMBs. The results suggest that the legacy of distrust in elections persists and is difficult to dissipate. Mexican citizens still believe that vote-buying or biased outcomes are pervasive. Partisan conflicts and elite polarization might be responsible to a certain extent for the skepticism of citizens (e.g., Eisenstadt 2007; Ugues Jr. 2018; Langston 2020). Nevertheless, the analysis reveals that citizens’ concrete experiences with electoral dynamics matter substantially, especially when they are better informed about democracy.
The evidence indicates that voters who perceive fairness and honesty during the electoral cycle are more likely to express confidence in elections. However, the crucial finding highlights that it takes some democratic sophistication to appreciate the political significance of electoral integrity. When voters detect electoral integrity, they are more likely to trust elections on the condition that they are also aware of the definition and meaning of democracy. An accurate notion of democracy seemingly enables people to recognize more clearly the link between specific instances of integrity and the trustworthiness of elections as a political institution.
In Mexico, an informed notion of democracy acts as a crucial yardstick to assess institutional performance. Even when some political actors frequently seek to undermine or capture electoral authorities, the relationships uncovered in this research are robust to the influence of voters’ political affinities and partisanship. Therefore, trust in elections depends not only on citizens’ belief in the integrity of elections, but also on their understanding of democracy. The underside of this is not simply that support for democracy in this country is usually moderate and increasingly weakened by violence related to drug trafficking (Crow 2010; Moreno and O’Neil 2014; Hiskey et al. 2020). There is also evidence that only a minority in the electorate share a coherent notion of liberal democracy (Schedler and Sarsfield 2009; Monsiváis-Carrillo 2018).
4.3. Is Mexico a special case?
This research primarily focuses on the “Mexican paradox.” Nevertheless, it is crucial to take a step forward and ask how far these findings are able to travel to other Latin American countries. The World Value Survey does not sample all countries in the region, but it includes data from Argentina, Brazil, Chile, Colombia, Ecuador, Guatemala, Nicaragua, and Peru (see Table A4). The analysis was replicated using a pooled database with observations from all these countries, excluding Mexico. The results, as shown in Table A6 in the appendix, confirm that the conditional importance of the understanding of democracy for trust in elections is not exclusive to Mexico. It is a relationship that is also present in other Latin American countries as well.
Indeed, the data confirm that perceptions of electoral integrity increase the likelihood of trust in elections, while perceptions of electoral malpractice produce the opposite effect. Moreover, in the Latin American countries included in the analysis, the data support H3 and H4. On the one hand, when citizens detect integrity during the electoral cycle, they will trust the elections to a greater degree if they are adequately informed about democracy. On the other hand, the understanding of democracy augments the negative influence of electoral malpractice on the outcome. As expected, citizens who are informed about democracy react more negatively to signs of electoral distortion or manipulation.
4.4. Robustness analysis
To assess the robustness of the results, additional tests were performed. In the first place, the study addresses potential endogeneity between independent and dependent variables. Table A2 in the appendix suggests that the correlation between dependent and independent variables is not a cause of concern. Another potential source of endogeneity is the omitted variable bias; that is, the possibility that both perceptions of electoral integrity or malpractice and trust in elections are effects of a variable excluded from the analysis. Nonetheless, within the limits of observational data, the analysis is replicated using a wide array of control variables, aside from those reported above, with similar results. No-reason is found to believe that crucial variables were omitted.
Then, it could be argued that the relationship between perceptions of electoral integrity and malpractice and trust in elections is affected by the risk of reverse causality. While such a risk is difficult to dispel using survey data, it is worth noting that both perceptions of electoral integrity and malpractice summarize factual beliefs with regard to concrete features of the electoral cycle. In contrast, trust in elections gauges more diffuse attitudes towards the political system (e.g., Norris 2017; van der Meer and Zmerli 2017). Conventional assumptions in the study of political trust suggest that attitudes of intermediate political support-such as trust in elections-are inductive generalizations shaped after concrete experiences with the performance of the regime. The opposite line of reasoning is more difficult to justify-namely, that voters conclude that “Votes are counted fairly” or “Voters are bribed,” among others, because they find the elections trustworthy in the first place.
Finally, as shown in M3 and M4 in Table 1 above, the relationship between perception of electoral integrity and the dependent variable is tested separately from the relationship between perception of electoral malpractice and the outcome. Even though perception of electoral integrity and perception of electoral malpractice tap into different conceptual domains, it could be argued that including both variables in the same model inflates the coefficients and artificially produces statistically significant results. Models M3 to M5 confirm that this is not the case. M28 to M30 in Table A6 further corroborate these findings, this time using data from eight Latin American countries. A supplementary test replicates the analysis using a variable that subtracts the index of perception of electoral malpractice from the index of perception of electoral integrity. Table A7 displays the results, showing the expected outcome in Latin America. In Mexico, however, it is likely that subtracting perception of electoral malpractice is canceling the relationship between perception of electoral integrity and trust in elections. If anything, this highlights the specific relevance of perception of electoral integrity in this country.
Conclusions
Trust in elections is a valuable asset in developing democracies. This research provides evidence that the understanding of democracy as a liberal-electoral regime plays a decisive role in the likelihood that people will trust elections. An informed understanding of democracy reinforces the relationship between perceptions of electoral integrity and trust in elections. Citizens are more likely to express confidence in elections when they believe that the electoral cycle is conducted with honesty. Furthermore, once citizens collect evidence of the integrity of elections, they will express more trust if they understand democracy as a liberal-electoral regime than if they are less informed about the meaning of democracy.
The empirical analysis focuses on Mexico, but the relationships uncovered in this case are present in other Latin American countries as well. The results offer a new perspective on the “Mexican paradox”-many citizens distrust elections despite the highly sophisticated system of electoral administration developed in the country during the last two decades. Evidence suggests, on the one hand, that the legacy of distrust in authoritarian elections persists. In Mexico, people who perceive that violence, bribes, or money shape the electoral contests usually distrust the elections. However, the understanding of democracy does not modify the relationship between perceiving electoral malpractice and trust in elections, in contrast with the results from several Latin American countries.
The crucial finding is that the relationship between perception of electoral integrity and trust in elections is strengthened when people understand democracy. This finding is present in both Mexico and the Latin American countries included in the analysis. According to the results, an informed understanding of democracy enhances the political significance of electoral integrity. If voters believe that elections are regularly conducted with honesty and fairness, they are more likely to trust elections on the condition of being well informed about the definition and meaning of democracy. In Mexico, for instance, an informed understanding of democracy performs a crucial role in enabling citizens to appreciate the political relevance of electoral integrity, even when some political actors routinely discredit the electoral authorities and the election outcomes.
The findings in this study suggest that more research is needed to examine the conditional influence of the understanding of democracy on trust in elections and political institutions in Latin America. Following recent contributions, it is important to assess how the quality of democracy and the integrity of elections at the country level interact with citizens’ notions of democracy (e.g., Mauk 2020; Wegscheider and Stark 2020; Mauk 2021). However, it is equally important to evaluate the consequences of potential policy interventions aimed at civic education. A relevant lesson for political authorities, civic organizations, or even political parties across the region is that a liberal notion of democracy matters. Therefore, public policies could promote voter education on the foundations and functioning of liberal democracy. In a region where weak institutions and personalistic leaders prevail, efforts at voter education on democracy might seem unwarranted. Still, as this research shows, these efforts could make a difference. When concrete evidence of election integrity is at hand, an accurate understanding of democracy might offer valuable public support to the usually contested development of democratic institutions.
