The Ron Swanson Effect: Disconnection in Attitudes towards Government & Sorting into Government

How to cite

Ham, Dasom; Sumner, Jane Lawrence. 2026. "The Ron Swanson Effect: Disconnection in Attitudes towards Government & Sorting into Government." AEA RCT Registry. AEARCTR-0018634. https://doi.org/10.1257/rct.18634-1.2

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Dasom Ham* , Jane Lawrence Sumner

* Primary investigator

Registry ID
AEARCTR-0018634
DOI
https://doi.org/10.1257/rct.18634-1.2
First registered
Last updated
Status
Completed
Trial dates
2026-05-20 – 2026-05-26
Countries
United States of America ()
JEL codes
J45, H83
Registry record
http://www.socialscienceregistry.org/trials/18634

Abstract

This study investigates whether living in US deindustrialized communities could lead to a disconnect in attitudes towards government and likelihood of applying to government. We hypothesize that individuals living in deindustrialized communities have negative attitudes towards government due to perceived first-hand experience interactions with government services. However, since government jobs are the last "off-shoreable" job in the labor market, we hypothesize, despite negative attitudes toward government, these individuals are more likely to apply to government jobs.

We set out to test the following RQs:

1. RQ1: Do people located in deindustrialized areas report a large gap between negative attitudes toward government and interest in government employment, all else equal?
2. RQ2: What are the effects of offering information on benefits of working in government on likelihood of applying for government job? Does this differ based on living in a deindustrialized area?

Our main outcomes are attitude towards local, state, and federal government itself and government jobs, whether would like to work at local, state, and federal government, and likelihood of applying to a local, state, and federal government job.

We will recruit 1500 US adults from Bovitz-Forthright, an online survey vendor. Our study will ask individuals about demographic characteristics, preference for non-wage amenities, attitudes towards local, state, and federal government itself and associated jobs, and whether would like to work for local, state, and federal government.

We will then take a subsample of respondents who have not worked at the government and randomly assign them to a treatment and control group. The treated group will receive an information intervention where they be given a list of selected benefits (or non-wage amenities) associated with government work. Then, we will compare the treatment and control group and conduct a heterogeneity analysis by individuals who lived in deindustrialized US counties vs. not exposed US counties.

Intervention

Individuals who have not worked in local, state, or federal government will be randomly assigned to treatment and control group. The treatment group will receive information about selected available non-wage amenities in the government jobs while the control group will receive a placebo information.

Experimental design

We will recruit adults who live in the USA through an online survey, Qualtrics via data vendor, Bovitz-Forthright. The targeted sample size is 1500.

For our information intervention, we will restrict our sample to only individuals who are not current or former local, state, and federal government employees. We will randomly assign these individuals to (1) treatment group or (2) control group. The treatment arm will receive an information intervention (see details above).

Design details

Randomization method
Conditional on not working or have hard worked in government, randomization done via Qualtrics survey flow
Randomization unit
Individual
Clusters
See attached document for details on clustering standard errors.
Observations
We expect about 1500 respondents, so 1500 observations (or each respondent = an observation).
Arms
Since we are conditioning on a sample of 1500, assuming 100 individuals reported currently or formerly working in the government, we expect the following: For treatment arm, there will be approximately 700 respondents. For control arm, there will be approximately 700 respondents. However, we expect our sample size for each arm to be slightly smaller due to sample restriction after conditioning.
Primary outcomes
Likelihood of applying for local, state, and federal jobs Attitude towards local, state, and federal government Attitude/feeling towards local, state, and federal government work Gap in whether would like to work and attitude/feeling towards local, state, and federal government We will do heterogeneity by whether live in a de-industrialized community
Secondary outcomes
Selecting “Stability” and “health insurance” as one of three important non-wage amenities Availability in job opportunities for past 12 months

Cited by

No citation data available for this registration yet.

Keywords