(iPoster) Bias or Bandwidth? Mission Timing and Recommendation Politics in OAS Monitoring

• In Event: Poster Session: International Collaboration Section

Abstract

International election observation is often treated as a credibility enhancing constraint on incumbents. Yet it also depends on negotiated cooperation between monitors and host governments, and it culminates in standardized organizational outputs, especially post-election recommendations produced through routinized templates and institutionalized topic categories (Hyde 2011; Kelley 2012; Barnett and Finnemore 2004). This paper asks when and why the Organization of American States issues more or fewer recommendations, and whether earlier operational access changes what the OAS recommends, not only how much it recommends. I develop a bias or bandwidth framework that treats mission timing as a collaboration problem with two mechanisms. Earlier acceptance, measured as the lead time between government acceptance and election day, can expand diagnostic bandwidth by increasing the time available to observe pre-election administration, verify irregularities, and draft more granular guidance. But earlier acceptance can also reflect deeper cooperation that facilitates pre-election remediation and relational problem solving, reducing what remains publicly recommendable when the final report is written. I test these expectations using an original election episode dataset of OAS observation missions merged with Varieties of Democracy measures of electoral integrity and democracy. The analysis covers 179 election episodes, with 146 episodes in the timing models. I estimate country and year fixed effects specifications using count data estimators appropriate for overdispersed outcomes. Across specifications, larger missions are consistently associated with a higher volume of recommendations, and earlier acceptance is strongly associated with the volume of recommendations net of election quality. To probe mechanisms, I estimate topic-level models in levels and in composition by modeling topic counts with an offset for episode totals. The composition results indicate a largely stable portfolio of issue areas: additional time and personnel primarily scale up dominant categories rather than systematically reallocating attention away from politically sensitive domains. These findings contribute to debates on international collaboration by showing how access bargaining, bureaucratic routines, and classification systems jointly structure the information and guidance that international organizations publicly produce in cooperative oversight arrangements.

Author

• Dayra Lascano, Cornell University