None of the Above: Protest Voting in Latin American Democracies

Cohen, Mollie J.

doi:doi:10.3998/mpub.12738341

Mollie J. Cohen

Open Access

Around the world each year, millions of citizens turn out to vote but leave their ballots empty or spoil them. Increasingly, campaigns have emerged that promote "invalid" votes like these. Why do citizens choose to cast blank and spoiled votes? And how do campaigns mobilizing the invalid vote influence this decision? None of the Above answers these questions using evidence from presidential and gubernatorial elections in eighteen Latin American democracies. Author Mollie J. Cohen draws on a broad range of methods and sources, incorporating data from electoral management bodies, nationally representative surveys, survey experiments, focus groups, semi-structured interviews, and news sources.

Contrary to received wisdom, this book shows that most citizens cast blank or spoiled votes in presidential elections on purpose. By participating in invalid vote campaigns, citizens can voice their concerns about low-quality candidates while also expressing a preference for high-quality democracy. Campaigns promoting blank and spoiled votes come about more often, and succeed at higher rates, when incumbent politicians undermine the quality of elections. Surprisingly, invalid vote campaigns can shore up the quality of democracy in the short term. None of the Above shows that swings in blank and spoiled vote rates can serve as a warning about the trajectory of a country's democracy.

Resources

Filtering by: Keyword campaign characteristics

Start Over

Not finding what you are looking for? Help improve Fulcrum's search and share your feedback.

The figure plots average invalid vote campaign approval by condition, estimated using an OLS regression model including treatment dummies, estimated without controls.

Average Effect of Campaign Attributes on Approval

From Chapter 4

Figure 4.4. Average Effect of Campaign Attributes on Approval Source: Peru Study I. Note: The figure presents results from an OLS regression model including treatment dummies, estimated without controls. Whiskers represent 95% confidence intervals around point estimates. Differences between coefficients are significant with * p ≤ 0.05, two-tailed. For complete results, see table A4.7.

The figure plots average values of each mechanism by leadership condition, estimated using an OLS regression model including treatment dummies, estimated without controls.

Mechanisms by Campaign Promoter

From Chapter 4

Figure 4.5. Mechanisms by Campaign Promoter Source: Peru Study I. Note: The figure presents results from an OLS regression model estimated without demographic controls but controlling for a respondent’s assigned democracy condition. Whiskers represent 95% confidence intervals around point estimates. Differences between coefficients are significant with * p ≤ 0.05. For complete results, see table A4.7.

The figure plots average values of each mechanism by democratic orientation condition, estimated using an OLS regression model including treatment dummies, estimated without controls.

Mechanisms by Democratic Orientation

From Chapter 4

Figure 4.6. Mechanisms by Democratic Orientation Source: Peru Study I. Note: The figure presents results from an OLS regression model estimated without demographic controls but controlling for a respondent’s assigned leadership condition. Whiskers represent 95% confidence intervals around point estimates. Differences between coefficients do not reach standard thresholds of statistical significance. For complete results, see table A4.7.

The figure plots values for three mechanisms by treatment condition, support for democracy, and trust in parties. The top row plots estimated values, and the bottom row plots differences across treatment conditions, with 95% confidence intervals.

Campaign Perceptions Conditional on Treatment, Support for Democracy, and Trust in Parties

From Chapter 4

Figure 4.7. Campaign Perceptions Conditional on Treatment, Support for Democracy, and Trust in Parties Source: Peru Study I. Note: The figure presents results from OLS regression models estimated without demographic controls but controlling for assigned leadership and democracy conditions. The top panel plots point estimates, and the bottom panel reports cross-group differences, with 95% confidence intervals. For complete results, see tables A4.8 and A4.9.

The figure plots campaign approval by treatment condition, support for democracy, and trust in parties. The top row plots estimated values, and the bottom row plots differences across treatment conditions, with 95% confidence intervals.

Campaign Approval, Conditional on Trust in Parties, Democratic Orientation

From Chapter 4

Figure 4.8. Campaign Approval Conditional on Trust in Parties, Democratic Orientation Source: Peru Study I. Note: The figure presents results from OLS regression models estimated without demographic controls but controlling for assigned leadership and democracy conditions. The top panel plots point estimates, and the bottom panel reports cross-group differences, with a 95% confidence interval. For complete results, see tables A4.8 and A4.9.

University of Michigan Press
Ebook Collection

None of the Above: Protest Voting in Latin American Democracies

Overview

Contents

Stats

Resources

Search and Filter Resources

Limit your search

Average Effect of Campaign Attributes on Approval

Mechanisms by Campaign Promoter

Mechanisms by Democratic Orientation

Campaign Perceptions Conditional on Treatment, Support for Democracy, and Trust in Parties

Campaign Approval, Conditional on Trust in Parties, Democratic Orientation

None of the Above: Protest Voting in Latin American Democracies

Resources

Search and Filter Resources

Limit your search

Section

Keyword

Creator

Format

This is the en.blacklight.search.search_results_header

Search Results