Research description
Describe the Christmas sample exclusion
- Review negative comments after Thanksgiving
- Identify fully rural counties in top third of Trump vote share
- Remove 60 counties before the Christmas campaign
Illustrative summary of the source’s sample change, covering 362 zip code areas in 13 states. This was not a randomised test of exclusion or proof that exclusion prevented backlash.
See source →How
When to use
In politically polarized environments, public health messaging campaigns risk triggering political reactance among resistant populations. Health agencies face the challenge that aggressive outreach in highly hostile areas could provoke counter-behavioral responses. Exclude specific high-risk, politically resistant geographic clusters to safeguard campaign safety and preserve overall intervention effectiveness.
Exclude highly politically resistant geographic clusters from public health advertising campaigns using demographic and election metrics.
- Research teams collect county-level election results and urban-rural demographic indicators across target jurisdictions.
- Analysts identify counties that fall into the top tercile of opposition party vote share and meet fully rural criteria.
- Campaign managers remove identified high-risk counties and their constituent zip codes from social media advertising target lists.
- Marketing partners configure platform geographic targeting parameters to ensure zero ad impression delivery in excluded counties.
- Evaluation teams monitor comment sections during active campaigns to verify backlash reduction across remaining target areas.
Example
Researchers excluded 60 fully rural counties in the top tercile of votes for Donald Trump in the 2020 election before the Christmas campaign after receiving negative comments during Thanksgiving.
About the study
Participants
60 fully rural counties (362 zip code tabulation areas) in the top tercile of Trump vote share in 13 US states.
Study Design
Targeted sample exclusion prior to Christmas ad deployment based on political vote share and rural classification following qualitative comment feedback from Thanksgiving.
Results
Post-exclusion Christmas campaign generated a 4.2% drop in zip-code COVID-19 cases (P = 0.007) and a 1.041 percentage point drop in travel distance (P = 0.011), with no evidence of harmful travel backlash.