NEW UPDATE: Completed trials may now upload and register supplementary documents (e.g. null results reports, populated pre-analysis plans, or post-trial results reports) in the Post Trial section under Reports, Papers, & Other Materials.
Do Citizens Respond to Better Information on Local Public Finance? Evidence from a Field Experiment During the 2019 Colombian Municipal Elections
Initial registration date
October 26, 2019
October 28, 2019 11:19 AM EDT
This section is unavailable to the public. Use the button below
to request access to this information.
Other Primary Investigator(s)
Additional Trial Information
We hypothesize that more transparency about the sources and use of public funds helps voters make better judgements about their elected officials, making them more likely to vote in order to effect political change. We examine the effect of information on voter behavior by randomizing whether we provide a municipality's journalists, community leaders, and the general public with simple and accessible information about local public finance collected by the central government.
Fonseca, Angela and Luis Reyes. 2019. "Do Citizens Respond to Better Information on Local Public Finance? Evidence from a Field Experiment During the 2019 Colombian Municipal Elections." AEA RCT Registry. October 28.
We evaluate a program that disseminates information collected by the central government about public finance in Colombia.
Intervention Start Date
Intervention End Date
Primary Outcomes (end points)
Voter turnout, change in incumbent party (for mayor and city council), blank/null voting.
Primary Outcomes (explanation)
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
We disseminate, in an easy-to-understand format, publicly available information about public finance to randomly chosen Colombian municipalities.
Experimental Design Details
We randomly chose municipalities using random numbers generated by Stata from a random seed obtained from random.org. We chose the municipalities from a list of all Colombian municipalities that excluded six major capital cities - Bogotá, Barranquilla, Cúcuta, Medellín, Cali, and Bucaramanga, as well as municipalities that as of 2017 (the latest available information) did not have their own radio stations.
We randomized at the municipality level. Outcomes are observed at the polling station level.
Was the treatment clustered?
Sample size: planned number of clusters
There is a total of 746 clusters (municipalities)
Sample size: planned number of observations
11826 polling stations.
Sample size (or number of clusters) by treatment arms
104 clusters outreach during municipal and national elections; 104 clusters outreach during national elections only; 208 clusters information through Facebook ads.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
For alpha=0.05 and power=0.8:
For an effect size of .2 standard deviations, a balanced experiment requires 199 clusters/municipalities. If the effect size is allowed to be 0.31 standard deviations (to be able to detect effects greater than 5 percentage points in turnout), each treatment group requires 85 municipalities.
For alpha=0.1 and power=0.8:
Assuming rho=.29, for an effect size of .2 standard deviations, a balanced experiment requires 158 clusters/municipalities. If the effect size is allowed to be 0.31 standard deviations (to be able to detect effects greater than 5 percentage points in turnout), each treatment group requires 67 municipalities.
INSTITUTIONAL REVIEW BOARDS (IRBs)