This paper asks how well states implement recommendations from temporary national truth commissions, focusing on 11 Latin American cases and using existing coded data to measure compliance. It contributes by applying a multilevel ordered logistic model to recommendation-level implementation data and assessing which recommendation types and contextual factors are associated with higher or lower compliance.
Multilevel ordered logistic regression Recommendation-level implementation data 11 Latin American truth commissions
Data and methods
The authors use data compiled by Skaar et al. (2022) on recommendations issued by 11 Latin American truth commissions and analyze implementation outcomes at the recommendation level. They apply multilevel ordered logistic regression to model variation in implementation while accounting for the nested structure of recommendations within commissions.
Main findings
The analysis reports that commissions issuing a larger number of recommendations are associated with lower average compliance. Recommendation content matters: backward-looking recommendations are less likely to be implemented, while recommendations calling for legal reform are more likely to be implemented. Factors often thought to predict truth commission success—such as the nature of the political transition or measures of civil society strength—do not show significant associations in these models, according to the abstract.
- More recommendations from a commission → lower compliance, on average.
- Backward-looking recommendations implemented less often; legal-reform recommendations implemented more often.
- Transition type and civil society strength are not significant predictors in these models.
