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.
Prosecutor Responsiveness to the Public and the Electoral Mechanism
Last registered on February 07, 2020


Trial Information
General Information
Prosecutor Responsiveness to the Public and the Electoral Mechanism
Initial registration date
February 04, 2020
Last updated
February 07, 2020 3:42 PM EST
Primary Investigator
West Virginia University
Other Primary Investigator(s)
Additional Trial Information
In development
Start date
End date
Secondary IDs
The United States is unique in that important legal actors are selected by popular elections. As a consequence, elections act as the primary accountability mechanism for the substantial discretion prosecutors exercise. Numerous problems can be expected - - voters do not have the human capital to properly evaluate prosecutor decision making and they are not incentivized to become informed (i.e., free riding). Nevertheless, recent empirical evidence suggests that prosecutors respond to the election cycle. Here, I investigate further whether elections encourage prosecutors to be responsive to their community's interest. To do so, I will conduct a randomized control trial where prosecutor offices will be contacted to provide basic information about their office. The outcome variable of interest is whether or not any employee of the office responds to the information request. Response rates of prosecutors with a re-election in the upcoming election cycle (in 2020) will be compared to those without elections in the year will be compared. In addition, offices run by appointed prosecutors will be considered.
External Link(s)
Registration Citation
McCannon, Bryan. 2020. "Prosecutor Responsiveness to the Public and the Electoral Mechanism." AEA RCT Registry. February 07. https://doi.org/10.1257/rct.5395-1.0.
Experimental Details
A total of 400 counties in the U.S. are selected at random (from all counties in the 45 states that have popular elections to select prosecutors). The prosecutor office that serves each county is identified. For these 400 offices, contact email information is collected, along with the dates of the head prosecutor's next/last election. A short survey is created. The questions on the survey are from the last Department of Justice Census of State Prosecutors survey from 2007. Thus, basic factual information is asked. The survey is emailed to the 400 randomly-selected offices (100 per month for 4 months). The data provided can be compared to that same offices responses in 2007 and those offices that do respond to the survey can be compared to those who did not respond.
Intervention Start Date
Intervention End Date
Primary Outcomes
Primary Outcomes (end points)
The primary outcome variable of interest is whether any employee of the office complete the survey.
Primary Outcomes (explanation)
This will simply be measured as an indicator variable for whether a response for that office is recorded in the 30 days/90 days after the email is sent.
Secondary Outcomes
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
From a list of every county in the United States, those from the five states that do not elect prosecutors will be removed. From the remaining counties, 400 will be selected at random to be a part of the research trial. Some of these will be from offices where the head prosecutor is running for re-election in 2020. The rest will be from offices where the head prosecutor is not up for re-election in 2020.
Experimental Design Details
Randomization Method
To select the counties to be included in the study, the random number generator function in Excel will be used. The 400 counties with the lowest drawn numbers will be the subject pool. In some areas of the country, multiple counties are merged together into a wider prosecutorial district. If two counties are selected to be in the study that are in the same prosecutorial district, then the one with the smaller value to the random number draw will be kept and a replacement county will be selected (following the same random selection mechanism as before).
Randomization Unit
I randomize at the county level.
Was the treatment clustered?
Experiment Characteristics
Sample size: planned number of clusters
400 prosecutor offices
Sample size: planned number of observations
Sample size (or number of clusters) by treatment arms
100 offices contacted each month for 4 months (February, March, April, and May 2020).
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
IRB Name
West Virginia University
IRB Approval Date
IRB Approval Number
# 1912821070
Post Trial Information
Study Withdrawal
Is the intervention completed?
Is data collection complete?
Data Publication
Data Publication
Is public data available?
Program Files
Program Files
Reports, Papers & Other Materials
Relevant Paper(s)