Data: Ad campaign significantly
improves non-partisan behavior.

TRIAL FINDS MESSAGING CAMPAIGN SIGNIFICANTLY BOOSTS COLLABORATIVE BEHAVIOR

A new e.pluribus.US ad campaign meaningfully increases public action to reduce partisan dysfunction.

In multiple, large-scale, randomized, controlled field trials, Americans exposed to the campaign exhibited significantly greater behavior supporting collaborative politicians than did those in a control group.

This finding offers a promising tool for reducing dysfunctional partisanship and galvanizing support for pro-collaboration politicians.

FINDINGS

106,323 citizens participated across two separate trials, from seven congressional districts in Texas, Oklahoma, Arkansas and Florida (this first phase tested right-leaning audiences). A Test Group was shown a series of social media videos comprising the campaign, while a separate Control Group saw videos unrelated to the campaign. Both groups were then invited to see actions they could take to reduce partisan division.

The Outcome:

• (1) The Test Group clicked the Invitation to see the actions at a 2.5 – 5.0% greater rate than Control.
  
• (2) The Test group was additionally 23% more likely than Control to actually take one of the actions to reduce partisanship, once it viewed the list.

These results validate that the messaging campaign meaningfully improves an audience’s propensity to act to reduce partisan dysfunction. Further information is provided in a full Technical Report.

BACKGROUND

e.pluribus.US conceives, builds and tests interventions to scalably improve public attitudes toward working with political opponents. In this initiative we test methods of influencing public opinion via advertising. The campaign cites familiar examples of ordinary citizens resolving disagreements in their daily lives, then questions whether politicians are as effective at overcoming disagreement with each other. Video 1 shows one of these messages.

There exists a wealth of initiatives to help citizens address partisanship, should they choose to. There are fewer initiatives aimed at motivating them to actually do so. This project targets that goal.

We believe most citizens are familiar with disagreement and how to work through it to accomplish things with others, however they do not, in sufficient numbers, demand that same behavior from their politicians. In order to motivate them to do so, the campaign messaging grounds itself in issues citizens already prioritize — eg. abortion, gun violence, inflation, border security — leverages frustration with politicians’ lack of progress on those issues, and draws a contrast between (a) how citizens approach resolving day-to-day disagreements and (b) how politicians approach resolving political disagreements.

Figures 1 and 2 show that after being treated with that rationale, citizens were 2.5% more likely in Trial 1 and 5% moreso in Trial 2 to consider actions they might take to reduce partisanship and then 23% more likely to actually take one of those actions.

Note: due to peculiarities of testing on the Meta platform, these raw results understate the impact of the campaign. On an accurately adjusted basis, Test outperformed Control by 5.1% and 11.4% margins, respectively, in clicking to see the actions, and by a combined 46% in actually taking action, as explained in the full Technical Report.

Next steps for this initiative include testing left-leaning audiences, increasing sample sizes, broadening demographic participation via other social media platforms and continually expanding the list of options for taking action. Our ultimate goal is to mass-apply the campaign in areas where it can influence voters to support candidates committed to work with opponents to get things done.

ACKNOWLEDGEMENTS

We thank:

• ● Stanford University’s Polarization and Social Change Lab
• ● The Polarization Lab at Duke University
• ● Civic Health Project and
• ● the Strengthening Democracy Challenge

for their inspiration, extensive prior work in this area, guidance on planning the trials and feedback on its results. A full list of acknowledgements is in the Technical Report.

RESULTS TABLES

Tables 1 & 2 set forth metrics on trial participation and engagement rates. The terms used are defined in the full Technical Report. Trial 1 was conducted in October of 2023 in five congressional districts across Texas, Oklahoma and Arkansas, using quasi-random test/control group assignments based upon alternating birth months. Trial 2 was conducted in January of 2024 in two congressional districts in Florida using fully random test/control group assignment.

Table 1 shows that the Test group subjects were 2.48% more likely in Trial 1 and 5.03% more likely in Trial 2 to accept an invite to take action to reduce partisan dysfunction. (After adjustment for peculiarities of testing on Meta, these numbers are 5.1% and 11.4%, respectively.)

A secondary question was, among subjects that clicked to view the list of ways to reduce partisanship, were those that had seen the campaign more likely to actually take one of those actions than those that had not? In both trials subjects were significantly more likely to take an action if they had seen the campaign. However we caveat that at this secondary level of participation, sample numbers become small. We can evaluate a larger sample size by combining the results across all the trials, as shown in Table 2. A follow-up test could confirm this outcome with larger sample size.

Full details on methodology, results, limitations/caveats and acknowledgments are available in the full Technical Report.

Or, if you want some humor, see the “post-credits” review of our screwup incidents, wherein we learned the hard way.