RESEARCH INTERESTS
My research agenda is centered on public opinion formation, voting behavior, and political campaigns. I examine the influence of anger and negative partisanship on voting behavior in new democracies, particularly in the context of the breakdown of the Mexican party system. My work seeks to better understand how anger plays a role in motivating voters in biased assimilation of political information. In turn, campaigns are able to translate voters’ anger into electoral behavior.
- American and Comparative Politics.
- Survey/Experimental Methods, Public Opinion, and Voting Behavior.
- Political Campaigns, Partisanship, and Motivated Reasoning.
- Misinformation, Clientelism and Vote-Buying.
- Survey Research, Expert Opinion, and Climate Change.
My research agenda is centered on public opinion formation, voting behavior, and political campaigns. I examine the influence of anger and negative partisanship on voting behavior in new democracies, particularly in the context of the breakdown of the Mexican party system. My work seeks to better understand how anger plays a role in motivating voters in biased assimilation of political information. In turn, campaigns are able to translate voters’ anger into electoral behavior.
BOOK PROJECT
My book project--“Politics of Anger: Negative Partisanship, Campaigns, and the Breakdown of the Mexican Party System”— explains why voters stopped supporting traditional parties and triggered the breakdown of one of the most stable party systems in Latin America. My research identifies the powerful role of anger explaining the rise of the populist left in the 2018 presidential election in Mexico. In a context of perceived poor governance and widespread corruption, the Mexican left identified the source of grievance—the incumbent PRI and the PAN as part of the political establishment—and translated voters’ anger into electoral behavior. Based on an original two-wave panel survey, a postelectoral survey as well as nine survey experiments conducted during the presidential campaign, my book project answers two interrelated questions: 1) what conditions changed to lead voters to support the left after rejecting the same presidential candidate in the two previous presidential elections and, 2) what variables activated anger among Mexican voters?
My research identifies three variables that moderated voters’ anger and made possible to translate it into political behavior: negative partisanship, misinformation, and political campaigns. My research finds that the proportion of partisans in Mexico is higher than previous studies suggested. Unlike Latin American cases in which party identification collapsed, partisanship in Mexico is widespread and negative. Since Mexico’s transition to democracy, the proportion of voters who both like their party and greatly dislike the opposing parties has dramatically increase making them more likely to engage in motivated reasoning. For example, based on experimental evidence, my research finds that voters engage in motivated reasoning when justifying corruption among their own co-partisan candidates. When voters were informed that a candidate (without informing the name of the candidate) engaged in corruption, respondents unequivocally consider such actions as corruption. However, when the name of their co-partisan candidate was explicitly mentioned—and engage in the same activities—voters reject to qualify them as corruption. My research also studies additional cases in which anger and negative partisanship make voters engage in defensive mechanisms and process information with a bias toward their preexisting predispositions: the economy, normative attitudes towards democracy, and vote-buying.
Second, in young party systems like Mexico, voters also engage in politically motivated reasoning reinforcing inaccurate beliefs or facts. The Mexican left relied on misinformation to activate their partisan base and remain electorally viable through specific mechanisms: 1) a belief that their candidate lost the 2006 and 2012 presidential elections due to electoral fraud (even though the best available evidence contradicts such misperception), 2) conspiratorial thinking that a “political mafia” of politicians and businessmen control the country, and 3) a sense of grievance against the political establishment driven by the perception that the Mexican left was not going to be “allowed” to win the presidency. These political misperceptions, particularly among voters with high levels of anger, undermined voters’ ability to form accurate opinions while reinforcing their partisan attachments.
And third, political campaigns were essential for the Mexican left to activating anger. Unlike cases in which a leftist candidate moderated his programmatic and personalistic appeals (e.g. Brazil’s Lula), the Mexican left—specifically Andrés Manuel López Obrador—did not change significantly. Instead, it was the political environment that changed, allowing the Mexican left to be perceived as the electoral vehicle to punish the PRI and the broader party system. Moreover, voters, including many not planning on voting for the left at the beginning of the campaign, were responsive to campaign information and behaved strategically by supporting López Obrador. These three elements helped the left mobilize a broad coalition of loyal partisans and anti-PRI voters, allowing them to win the Presidency and Congress for the first time in Mexican history.
My research identifies three variables that moderated voters’ anger and made possible to translate it into political behavior: negative partisanship, misinformation, and political campaigns. My research finds that the proportion of partisans in Mexico is higher than previous studies suggested. Unlike Latin American cases in which party identification collapsed, partisanship in Mexico is widespread and negative. Since Mexico’s transition to democracy, the proportion of voters who both like their party and greatly dislike the opposing parties has dramatically increase making them more likely to engage in motivated reasoning. For example, based on experimental evidence, my research finds that voters engage in motivated reasoning when justifying corruption among their own co-partisan candidates. When voters were informed that a candidate (without informing the name of the candidate) engaged in corruption, respondents unequivocally consider such actions as corruption. However, when the name of their co-partisan candidate was explicitly mentioned—and engage in the same activities—voters reject to qualify them as corruption. My research also studies additional cases in which anger and negative partisanship make voters engage in defensive mechanisms and process information with a bias toward their preexisting predispositions: the economy, normative attitudes towards democracy, and vote-buying.
Second, in young party systems like Mexico, voters also engage in politically motivated reasoning reinforcing inaccurate beliefs or facts. The Mexican left relied on misinformation to activate their partisan base and remain electorally viable through specific mechanisms: 1) a belief that their candidate lost the 2006 and 2012 presidential elections due to electoral fraud (even though the best available evidence contradicts such misperception), 2) conspiratorial thinking that a “political mafia” of politicians and businessmen control the country, and 3) a sense of grievance against the political establishment driven by the perception that the Mexican left was not going to be “allowed” to win the presidency. These political misperceptions, particularly among voters with high levels of anger, undermined voters’ ability to form accurate opinions while reinforcing their partisan attachments.
And third, political campaigns were essential for the Mexican left to activating anger. Unlike cases in which a leftist candidate moderated his programmatic and personalistic appeals (e.g. Brazil’s Lula), the Mexican left—specifically Andrés Manuel López Obrador—did not change significantly. Instead, it was the political environment that changed, allowing the Mexican left to be perceived as the electoral vehicle to punish the PRI and the broader party system. Moreover, voters, including many not planning on voting for the left at the beginning of the campaign, were responsive to campaign information and behaved strategically by supporting López Obrador. These three elements helped the left mobilize a broad coalition of loyal partisans and anti-PRI voters, allowing them to win the Presidency and Congress for the first time in Mexican history.
COLLABORATIVE PROJECTS
I am also participating in a number of collaborative research projects. As investigator of the Mexico team that conducts the National Electoral Study—which is part of the Comparative Studies of Electoral Systems (CSES)—we study vote-buying in recent elections at the gubernatorial, legislative, and presidential level in Mexico. Methodologically, we find that sensitive survey techniques like list experiments underestimate vote buying. We identify conditions under which list experiments estimate results that contradict direct questioning or provide nonsensical results (e.g. negative prevalence of a sensitive behavior). Given the cognitive taxing process required by the technique, list experiments are difficult to use among less educated populations (who, in fact, are the most likely to be targeted by clientelism). In turn, direct questioning that help voters remember whether they receive a gift makes respondents more likely to report receiving such gifts. We also find that Mexican parties have adopted the GOTV model of voter mobilization. However, instead of only delivering messages to increase turnout, candidates deliver clientelistic gifts in households without the need of political brokers.
My second collaborative project seeks to understand why voters support corrupt politicians in new democracies, focusing on voters’ responsiveness to sources of information, which sheds light on an important link in the larger process through which citizens hold politicians accountable. Relying on original survey experiments conducted in Colombia, El Salvador, Mexico, and Argentina, we find that only accusations of corruption coming from perceived credible sources and from media outlets that are ideologically close to voters’ partisan preferences drive down levels of support for corrupt politicians. We also find that voters are more likely to forgive politicians’ wrongdoings when benefits are redistributed than when they are not (illicit enrichment). My research expands on this collaborative project by studying how partisanship makes some voters less likely to hold co-partisan politicians accountable even when they clearly misbehave.
I also participate in a project that studies a quintessential topic of misinformation among elites and the American public: climate change. Relying on an original elite survey of top environmental biologists, we found that experts agree on climatic change projections, but have less consensus about strategies of climate change adaptation, and varying levels of certainty about consequences of climate change.
My second collaborative project seeks to understand why voters support corrupt politicians in new democracies, focusing on voters’ responsiveness to sources of information, which sheds light on an important link in the larger process through which citizens hold politicians accountable. Relying on original survey experiments conducted in Colombia, El Salvador, Mexico, and Argentina, we find that only accusations of corruption coming from perceived credible sources and from media outlets that are ideologically close to voters’ partisan preferences drive down levels of support for corrupt politicians. We also find that voters are more likely to forgive politicians’ wrongdoings when benefits are redistributed than when they are not (illicit enrichment). My research expands on this collaborative project by studying how partisanship makes some voters less likely to hold co-partisan politicians accountable even when they clearly misbehave.
I also participate in a project that studies a quintessential topic of misinformation among elites and the American public: climate change. Relying on an original elite survey of top environmental biologists, we found that experts agree on climatic change projections, but have less consensus about strategies of climate change adaptation, and varying levels of certainty about consequences of climate change.
PEER-REVIEWED PUBLICATIONS
Castro Cornejo, Rodrigo. Forthcoming. "Who is Partisan? Partisanship and Question Wording Effects in Latin America." Public Opinion Quarterly.
(with Ulises Beltrán). Forthcoming. Clientelism as GOTV Campaigning. Modern Campaigns and Vote-Buying in New Democracies. Política y Gobierno.
Botero, Sandra, Rodrigo Castro Cornejo, Laura Gamboa, Nara Pavao and David Nickerson. 2015. “Says Who? The Importance of Contextually Credible Sources for Corruption Allegations.” Political Research Quarterly. 68 (3): 493-504. PDF
Javeline, Debra, Jessica J. Hellmann, Jason S. McLachlan, Dov F. Sax, and Mark W. Schwartz and Rodrigo Castro Cornejo. 2015. “Expert Opinion on Extinction Risk and Climate Change Adaptation.” Elementa. 3(57). PDF
Javeline, Debra, Jessica Hellman, Rodrigo Castro Cornejo and Gregory Schufeldt. 2014. "Expert Opinion on Climate Change and Threats to Biodiversity," BioScience 63(8): 666-673. PDF
(with Ulises Beltrán). Forthcoming. Clientelism as GOTV Campaigning. Modern Campaigns and Vote-Buying in New Democracies. Política y Gobierno.
Botero, Sandra, Rodrigo Castro Cornejo, Laura Gamboa, Nara Pavao and David Nickerson. 2015. “Says Who? The Importance of Contextually Credible Sources for Corruption Allegations.” Political Research Quarterly. 68 (3): 493-504. PDF
Javeline, Debra, Jessica J. Hellmann, Jason S. McLachlan, Dov F. Sax, and Mark W. Schwartz and Rodrigo Castro Cornejo. 2015. “Expert Opinion on Extinction Risk and Climate Change Adaptation.” Elementa. 3(57). PDF
Javeline, Debra, Jessica Hellman, Rodrigo Castro Cornejo and Gregory Schufeldt. 2014. "Expert Opinion on Climate Change and Threats to Biodiversity," BioScience 63(8): 666-673. PDF
UNDER REVIEW
Do Electoral Polls Affect Voters' Behavior? Experimental Evidence from Mexico. R&R. (PDF).
This study finds that voters who are aware of their candidates' standing in the race strategically interpret polling results and switch their support to alternative candidates. This research contributes to the campaigns literature in two important ways. This article highlights the major role that partisanship and voters’ levels of information play moderating voters’ responsiveness to polling information. This study accounts for the role that voters’ expectations about the outcome of the election play on their incentives for either continuing supporting their co-partisan candidate or defecting. Such expectations are moderated by a partisan bias encouraging voters to engage in motivated reasoning. Similarly, the findings of this study suggest that vote shifts in new democracies are not exclusively an outcome of campaign persuasion as most comparative studies assume.
Who Receive Electoral Gifts? Sensitive Behavior and Question Wording Effects (with Ulises Beltrán). R&R. (PDF)
This research focuses on survey measurement of sensitive behavior, particularly, estimating the proportion of voters that receive electoral gifts during campaigns. This research shows that prior studies have been based on a survey methodology that systematically underestimates vote buying. Survey questions that rely on filter questions and include the phrase “in exchange for your vote” make respondents less likely to self-report receiving gifts during political campaigns. In turn, direct questioning that help respondents remember whether they received an electoral gift makes them more likely to report it, even gifts that constitute a clientelistic exchange and are considered socially undesirable. The findings of this paper have important implications for the study of recalled events whose reporting may be reduced by social desirability bias.
How Do Campaigns Matter (in New Democracies)? Vote Choice and the Stability of Partisan Attachments R&R.
This research contributes to the campaigns literature by pointing out a different mechanism for which campaigns matter in new democracies. While most research has focused on the process in which partisanship develops in new democracies or the strength of voters’ partisan attachments in young party systems, this study focuses on a second dimension of partisanship overlooked by most comparative studies on campaign effects: individual level stability. Some voters in new democracies are able to build strong partisanship as the screen through which they observe the political world leading individuals to interpret new information so as to reinforce. However, some voters lack long-term partisan attachments enabling them to update their party identification as the campaign unfolds. These voters have a harder time to reinforce their pre-campaign predispositions and are more likely to change their vote choice throughout campaigns. The findings suggest that for these voters, partisanship and voting choice are empirically intertwined.
List Experiments and Vote Buying in Latin America (with Ulises Beltrán). (PDF)
The manuscript reports the results of experimental survey data conducted in Mexico. We conducted list experiments to estimate the percent of respondents that received electoral gifts during the legislative (2015) and subnational (2015 and 2017) campaigns in Mexico. Consistent with recent studies on sensitive survey techniques, our research finds that list experiments seem to methodologically work better among more sophisticated voters (e.g. higher levels of education). Such findings suggest that previous studies that rely on the list experiment tend to underestimate the percent of voters that receive electoral gifts since this technique tends to work better among respondents who are, in fact, the least likely to be targeted by clientelistic strategies. Given levels of education in the region, we suggest that research that relies solely on list experiments approach its empirical findings with caution.
Under Friendly Fire: Partisan Press, Fragmented Opposition and Voting Behavior (with Sandra Botero, Laura Gamboa, Nara Pavão, and David W. Nickerson).
Statements in which a one-sided partisan media source criticizes a politician aligned with it—friendly fire—are particularly persuasive. This literature assumes a bipartisan context, where partisan news outlets send cues that neatly map on to two opposing partisan options. What happens in fragmented party systems? We argue that when there is a dominant party on one side of the political spectrum with a strong link with a media outlet, voters treat attacks against a co-partisan candidate as friendly fire. But when there is a fragmented opposition, we expect that the strength of the signal conveyed by the friendly fire is diminished. We fielded a survey experiment in Argentina where, until 2015, there was one governing party openly associated with a media outlet and a fragmented opposition. Accusations made by a congenial newspaper increase the punishment of the co-partisan candidate. The fragmented nature of the opposition changes the dynamic of friendly fire. Only partisan and sophisticated opposition voters treat attacks on opposition candidates as friendly fire. These voters are better able to overcome the lack of clear partisan link with the opposition newspaper and punish their co-partisan candidate.
Are All Types of Corruption Created Equal in the Eyes of Voters? (with Sandra Botero, Laura Gamboa, Nara Pavão, and David W. Nickerson).
Do voters evaluate some forms of political wrongdoing more harshly than others? Do they punish private enrichment and clientelism equally? We argue that voters’ responses to political wrongdoing are a function of the expected benefits voters associate with specific types of malpractice. We conducted a survey experiment varying two common types of political wrongdoing and measuring citizens’ evaluations of political candidates in Argentina. The results show that respondents punish politicians engaged in private enrichment more severely than politicians engaged in clientelism. Respondents were less willing to evaluate a candidate favorably on a series of dimensions when the candidate was accused of private enrichment rather than clientelism. We test two arguments that could provide a mechanism for this phenomenon. While the strength of one’s partisan affiliation does not moderate the treatment effect, we find that respondents with low socio-economic status punish illicit enrichment more harshly than clientelism and that high socio-economic respondents punish both types of wrongdoing equally.
This study finds that voters who are aware of their candidates' standing in the race strategically interpret polling results and switch their support to alternative candidates. This research contributes to the campaigns literature in two important ways. This article highlights the major role that partisanship and voters’ levels of information play moderating voters’ responsiveness to polling information. This study accounts for the role that voters’ expectations about the outcome of the election play on their incentives for either continuing supporting their co-partisan candidate or defecting. Such expectations are moderated by a partisan bias encouraging voters to engage in motivated reasoning. Similarly, the findings of this study suggest that vote shifts in new democracies are not exclusively an outcome of campaign persuasion as most comparative studies assume.
Who Receive Electoral Gifts? Sensitive Behavior and Question Wording Effects (with Ulises Beltrán). R&R. (PDF)
This research focuses on survey measurement of sensitive behavior, particularly, estimating the proportion of voters that receive electoral gifts during campaigns. This research shows that prior studies have been based on a survey methodology that systematically underestimates vote buying. Survey questions that rely on filter questions and include the phrase “in exchange for your vote” make respondents less likely to self-report receiving gifts during political campaigns. In turn, direct questioning that help respondents remember whether they received an electoral gift makes them more likely to report it, even gifts that constitute a clientelistic exchange and are considered socially undesirable. The findings of this paper have important implications for the study of recalled events whose reporting may be reduced by social desirability bias.
How Do Campaigns Matter (in New Democracies)? Vote Choice and the Stability of Partisan Attachments R&R.
This research contributes to the campaigns literature by pointing out a different mechanism for which campaigns matter in new democracies. While most research has focused on the process in which partisanship develops in new democracies or the strength of voters’ partisan attachments in young party systems, this study focuses on a second dimension of partisanship overlooked by most comparative studies on campaign effects: individual level stability. Some voters in new democracies are able to build strong partisanship as the screen through which they observe the political world leading individuals to interpret new information so as to reinforce. However, some voters lack long-term partisan attachments enabling them to update their party identification as the campaign unfolds. These voters have a harder time to reinforce their pre-campaign predispositions and are more likely to change their vote choice throughout campaigns. The findings suggest that for these voters, partisanship and voting choice are empirically intertwined.
List Experiments and Vote Buying in Latin America (with Ulises Beltrán). (PDF)
The manuscript reports the results of experimental survey data conducted in Mexico. We conducted list experiments to estimate the percent of respondents that received electoral gifts during the legislative (2015) and subnational (2015 and 2017) campaigns in Mexico. Consistent with recent studies on sensitive survey techniques, our research finds that list experiments seem to methodologically work better among more sophisticated voters (e.g. higher levels of education). Such findings suggest that previous studies that rely on the list experiment tend to underestimate the percent of voters that receive electoral gifts since this technique tends to work better among respondents who are, in fact, the least likely to be targeted by clientelistic strategies. Given levels of education in the region, we suggest that research that relies solely on list experiments approach its empirical findings with caution.
Under Friendly Fire: Partisan Press, Fragmented Opposition and Voting Behavior (with Sandra Botero, Laura Gamboa, Nara Pavão, and David W. Nickerson).
Statements in which a one-sided partisan media source criticizes a politician aligned with it—friendly fire—are particularly persuasive. This literature assumes a bipartisan context, where partisan news outlets send cues that neatly map on to two opposing partisan options. What happens in fragmented party systems? We argue that when there is a dominant party on one side of the political spectrum with a strong link with a media outlet, voters treat attacks against a co-partisan candidate as friendly fire. But when there is a fragmented opposition, we expect that the strength of the signal conveyed by the friendly fire is diminished. We fielded a survey experiment in Argentina where, until 2015, there was one governing party openly associated with a media outlet and a fragmented opposition. Accusations made by a congenial newspaper increase the punishment of the co-partisan candidate. The fragmented nature of the opposition changes the dynamic of friendly fire. Only partisan and sophisticated opposition voters treat attacks on opposition candidates as friendly fire. These voters are better able to overcome the lack of clear partisan link with the opposition newspaper and punish their co-partisan candidate.
Are All Types of Corruption Created Equal in the Eyes of Voters? (with Sandra Botero, Laura Gamboa, Nara Pavão, and David W. Nickerson).
Do voters evaluate some forms of political wrongdoing more harshly than others? Do they punish private enrichment and clientelism equally? We argue that voters’ responses to political wrongdoing are a function of the expected benefits voters associate with specific types of malpractice. We conducted a survey experiment varying two common types of political wrongdoing and measuring citizens’ evaluations of political candidates in Argentina. The results show that respondents punish politicians engaged in private enrichment more severely than politicians engaged in clientelism. Respondents were less willing to evaluate a candidate favorably on a series of dimensions when the candidate was accused of private enrichment rather than clientelism. We test two arguments that could provide a mechanism for this phenomenon. While the strength of one’s partisan affiliation does not moderate the treatment effect, we find that respondents with low socio-economic status punish illicit enrichment more harshly than clientelism and that high socio-economic respondents punish both types of wrongdoing equally.
WORKING PAPERS
Is My Candidate Behind? The Poll is Not Credible. Experimental Evidence from Mexico.
Clientelism or Political Communication? Campaign Mobilization in Mexico (with Ulises Beltrán)
Why Do Voters Support Candidates Illegally Funded by Drug-Cartels? Experimental Evidence from Mexico.
Who's Right? Support for Conservative Parties in Latin America (with Peter Johannessen).
Religious Engagement, Civic Skills, and Political Participation in Latin America (With Andre Audette and Mark Brockway)
Does Candidate's Religious Affiliation Affect Vote Choice? Experimental Evidence from Mexico
Clientelism or Political Communication? Campaign Mobilization in Mexico (with Ulises Beltrán)
Why Do Voters Support Candidates Illegally Funded by Drug-Cartels? Experimental Evidence from Mexico.
Who's Right? Support for Conservative Parties in Latin America (with Peter Johannessen).
Religious Engagement, Civic Skills, and Political Participation in Latin America (With Andre Audette and Mark Brockway)
Does Candidate's Religious Affiliation Affect Vote Choice? Experimental Evidence from Mexico
