Donating Time for Democracy: Encouraging Citizen Participation in Electoral Work in Mexico

Researchers:
Enrique Seira Bejarano
Andrei Gomberg
Fieldwork by:
Location:
Mexico
Timeline:
2016 - 2017
Target group:
  • Civil servants
Outcome of interest:
  • Electoral participation
  • Enrollment and attendance
  • Transparency and accountability
Intervention type:
  • Community participation
  • Digital and mobile
  • Information
  • Nudges and reminders
  • Recruitment and hiring
  • Training
Research papers:
Partners:
  • Que Funciona para el Desarrollo
  • Government of Mexico

In Mexico, more than one million citizens participate in running polling booths and counting votes in state and national elections. In 2015, however, as many as 100,000 citizens who had agreed to function as poll workers and received training for it did not show up on the day of the election. In this study, researchers are partnering with the national electoral authority to test a range of policy interventions designed to ultimately improve poll workers’ attendance during the 2017 state elections. 

Policy issue

Citizen participation is the foundation of democratic institutions. In situations in which citizens perceive their elections to be at high risk of fraud, inviting citizens themselves to manage polling booths and lead vote counts can increase the credibility and validity of election results.

However, this may bring logistical and organizational challenges. Citizens who participate must commit to many hours of training and day-of labor, and a high percentage of citizens tend to drop out during the training period or are no-shows on election day itself. Dropouts and absenteeism of citizen poll workers result in voting precincts being staffed by untrained citizens, which may increase the likelihood of vote-counting mistakes and thus risk annulment of precincts, calls for recounts, and challenges to the legitimacy of democratic institutions.

Past research has shown that a number of factors drive poll worker dropout. First, citizens may be poor forecasters of their future willingness to participate: They may sincerely wish to donate time in the future, but avoid it when the future arrives.1 Second, citizens may not see value in donating their time to democracy.2 Third, citizens who would donate their time may refrain from doing so out of dislike for certain teammates’ traits, including gender and socioeconomic status.3 What types of messages and forms of communication can governments employ to increase the reliability of election workers?

Context of the evaluation

Prior to 2000, Mexico experienced decades of one-party rule sustained by institutional election fraud. Broad distrust of election processes and allegations of election tampering continue to mar election results. However, full participation of trained citizen election workers can build trust and accuracy in the electoral process.

Citizen poll workers are selected randomly, with literacy being the primary prerequisite. The Mexican Constitution requires that votes be counted by citizens, but it has become increasingly difficult to recruit enough citizens to do this. In the 2015 elections, for example, 100,000 citizens who had agreed to function as poll workers and had been trained accordingly did not show up to serve on election day. Moreover, about half of recruited poll workers drop out during training.

Recruiting and training citizen poll workers involves three stages. First, over the course of two home visits, recruiters assess citizen eligibility and offer citizens a specific poll worker roll. Roughly two months later, poll workers meet for operational training. On election day in the following month, poll workers are tasked with running the balloting, counting the votes, and filling out totals in official documents. If one of the four poll workers assigned to a precinct does not show up, a trained substitute takes their place. When trained substitutes do not show up, the precinct accepts volunteers from the queue of voters, who fill in for the missing worker at the last minute and without training.

In this study, researchers are partnering with Mexico’s Instituto Nacional Electoral (INE), the national election commission, to test different approaches to poll worker recruitment and retention for the 2017 state-level elections. INE plans to implement the lessons learned from this study in the 2018 nationwide presidential election, which will require an unprecedented 1.4 million citizen poll workers to staff more than 150,000 polling stations across Mexico.

Details of the intervention

Researchers are conducting a randomized evaluation to test the impact of four different interventions at specific points in the recruiting and training cycle on poll worker dropout and absenteeism: information about the composition of their poll booth team, civic education, a public pledge of attendance, and social media engagement.

  1. Information about demographic composition of the poll-booth team: While being informed of their own role on the poll-booth team during the recruiter’s second visit to their home, a random set of poll workers will be provided with details like gender and education levels of their teammates. By comparing dropout, attendance, and team effectiveness of each group, researchers can examine issues related to discrimination.

  2. Civic education: In addition to the standard operational training, a random subset of poll workers will participate in a separate training module designed to strengthen their intentions to fulfill their poll worker duties. The module includes a pre-scripted 30-minute civic education talk by a recruiter, accompanied by a video and printed material with key takeaways.

  3. Pledge of attendance: During training, a random subset of citizen poll workers will have the opportunity to publicly pledge to their polling booth team that they will show up on election day.

  4. Social media messaging: After recruitment and continuing until election day, a subset of citizen poll workers who did not have the option of making public attendance pledges during training will be encouraged to communicate with each other through a popular social messaging platform with varying formats for group chats. This will provide insight into the dynamics of group communication and how social media in practice can affect delivery of public goods like civic action.

Researchers will conduct a baseline survey of all 200,000 citizen poll workers participating in the study. The survey will measure attitudes towards democracy, voting history, knowledge of politics, and demographics, among other topics. Researchers will then conduct a follow-up survey on the day of the training to measure expectations about each team member’s likelihood of participating on election day. In addition to surveys, researchers will use administrative data from the national election commission to assess poll worker attendance on election day, and procedural and vote-counting mistakes at poll booths.

Results of this impact evaluation are expected to inform scale-ups of citizen poll worker engagement programs in future elections.

Results and policy lessons

Results forthcoming.

1.
Laibson, David. 1997. “Golden Eggs and Hyperbolic Discounting.” The Quarterly Journal of Economics:443-477; O’Donoghe, Ted and Matthew Rabin. 1999. “Doing It Now or Later.” American Economic Review: 103-124.
2.
Gatcher, Simon, Daniele Nosenzo, and Martin Sefton. 2013.“Peer Effects in Pro-Social Behavior: Social Norms or Social Preferences?” Journal of the European Economic Association: 548-573.
3.
Alesina, Alberto and Eliana La Ferrara. 2000. “Participation in Heterogeneous Communities,” The Quarterly Journal of Economics: 847-904; La Ferrara, Eliana. 2002. “Inequality and group participation: Theory and evidence from rural Tanzania.” Journal of Public Economics: 235-273; Vigdor, Jacob. 2004. “Community composition and collective action: Analyzing initial mail response to the 2000 census.” The Review of Economics and Statistics: 303-312; Habyarimana, James et al. 2007. “Why Does Ethnic Diversity Undermine Public Goods Provision?” American Political Science Review: 709-725.