Do Robotic Calls from Credible Sources Influence Voter Turnout or Vote Choice?
Last registered on July 20, 2016

Pre-Trial

Trial Information
General Information
Title
Do Robotic Calls from Credible Sources Influence Voter Turnout or Vote Choice?
RCT ID
AEARCTR-0001430
Initial registration date
July 20, 2016
Last updated
July 20, 2016 12:36 PM EDT
Location(s)
Primary Investigator
Affiliation
Columbia University
Other Primary Investigator(s)
PI Affiliation
University of Texas at Austin
PI Affiliation
Yale University
Additional Trial Information
Status
Completed
Start date
2003-03-06
End date
2003-03-08
Secondary IDs
Abstract
The effectiveness of prerecorded phone calls was assessed in the context of a Texas Republican primary election that featured a contest for state Supreme Court. Automated calls endorsing one of the judicial candidates were recorded by the sitting Republican governor and directed at more than a quarter million people identified as likely voters and probably supporters of the governor. Two experimental designs were used to evaluate the calls' effectiveness. The first design randomly assigned households to treatment and control conditions in order to gauge the calls' effects on individuals' voter turnout, as measured by public records. The second design randomly assigned precincts to treatment and control conditions in order to assess whether the calls increased the precinct-level vote margin of the endorsed candidate. Results suggest that the automated calls had weak and statistically insignificant effects on turnout and vote margins.
External Link(s)
Registration Citation
Citation
Gerber, Alan, Donald Green and Daron Shaw. 2016. "Do Robotic Calls from Credible Sources Influence Voter Turnout or Vote Choice?." AEA RCT Registry. July 20. https://doi.org/10.1257/rct.1430-1.0.
Former Citation
Gerber, Alan et al. 2016. "Do Robotic Calls from Credible Sources Influence Voter Turnout or Vote Choice?." AEA RCT Registry. July 20. http://www.socialscienceregistry.org/trials/1430/history/9492.
Experimental Details
Interventions
Intervention(s)
Automated phone calls with the recorded voice of Governor Rick Perry urging Republicans to support a candidate for Texas Supreme Court.
Intervention Start Date
2003-03-06
Intervention End Date
2003-03-08
Primary Outcomes
Primary Outcomes (end points)
Effects of Robotic Calls on Voter Turnout Based on Precinct Level Random Assignment

- Percentage Voting


Effect of Robotic Calls on Voter Turnout Based on Household Level Random Assignment, by Household Size:

- One-Voter Households Percentage Voting

- Two-Voter Households Percentage Voting


Estimated Effects of Robotic Calls on Precinct-Level Outcomes:

- GOP Primary Votes Cast

- GOP Turnout

- Votes Cast for Willett or Smith

- Willett Votes

- Net Willett-Smith Difference
Primary Outcomes (explanation)
Secondary Outcomes
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
This study consisted of a two-tiered experiment aimed at evaluating how automated phone calls affect both voter mobilization and persuasion. This evaluation was conducted in the context of a Texas Republican primary election that featured a contest for state Supreme Court. Automated calls endorsing one of the judicial candidates, Don Willett, were recorded by the sitting Republican governor, Rick Perry, and directed at more than a quarter million people identified as likely voters and probable supporters of the governor. The working hypothesis was that calls conveying a message from a prominent and credible source, such as the sitting Republican Governor, should have a persuasive effect.

The first experimental analysis randomly assigned precincts containing between 100 and 700 (likely) Republican and pro-Perry households to treatment and control groups. The turnout and persuasion hypotheses were assessed by comparing average outcomes in the treatment and control precincts.

The second analysis occurred in precincts containing fewer than 100 or more than 700 Republican households. Within these precincts, households, rather than precincts, were randomly assigned to treatment or control groups.
Experimental Design Details
Randomization Method
Experiment 1:

Turnout was measured at the level of the individual voter. Random assignment of targeted households was conducted using a random generator in SPSS.

Experiment 2:

Within the precincts selected for experiment 2, households, rather than precincts, were randomly assigned to treatment or control groups. Robotic calls were directed only to phone numbers belonging to individuals in the treatment group.
Randomization Unit
Experiment 1:

- Turnout measured at the level of the individual voter and assigned at the household level (where households were defined as shared phone numbers)

- Vote choice observed at the precinct level


Experiment 2:

Precincts were assigned to treatment or control using the same targeting criteria as the household-level random assignment. Turnout and vote choice observed at the household level

Was the treatment clustered?
Yes
Experiment Characteristics
Sample size: planned number of clusters
1,597 precincts
Sample size: planned number of observations
292,602 voters
Sample size (or number of clusters) by treatment arms
Experiment 1:

- The 815 treatment precincts contained a total of 147,660 voters, as compared to 782 control precincts, which contained 144,942 voters. Vote choice was observed only at the precinct level.

Experiment 2:

- A total of 105,270 voters lived in households whose phone number was assigned to the treatment group. The control group comprised 65,657 voters.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
IRB
INSTITUTIONAL REVIEW BOARDS (IRBs)
IRB Name
IRB Approval Date
IRB Approval Number
Post-Trial
Post Trial Information
Study Withdrawal
Intervention
Is the intervention completed?
Yes
Intervention Completion Date
March 08, 2003, 12:00 AM +00:00
Is data collection complete?
Yes
Data Collection Completion Date
March 08, 2003, 12:00 AM +00:00
Final Sample Size: Number of Clusters (Unit of Randomization)
1,597 precincts
Was attrition correlated with treatment status?
No
Final Sample Size: Total Number of Observations
292,602 voters
Final Sample Size (or Number of Clusters) by Treatment Arms
Experiment 1: - The 815 treatment precincts contained a total of 147,660 voters, as compared to 782 control precincts, which contained 144,942 voters. Vote choice was observed only at the precinct level. Experiment 2: - A total of 105,270 voters lived in households whose phone number was assigned to the treatment group. The control group comprised 65,657 voters.
Data Publication
Data Publication
Is public data available?
Yes
Program Files
Program Files
No
Reports and Papers
Preliminary Reports
Relevant Papers