The data are from a stratified random national sample of 2,107 voters, fielded from April 28 to May 15, 2026. Respondents rated each message on the familiar 0 to 100 scale of the extent to which they agreed or disagreed with it, with 100 being the highest and zero the lowest. For methodological details, check out our website, americanmessage.org. The study reports voter ratings on the major perennial issues two or three others in the news, from a stratified random national sample of 2,107 voters selected to resemble the voting population. Respondents rated subsets of 108 statements, randomized per subject, of which 96 were affirmative messages and 12 were opposition messages on each of the same issues taken from actual statements or written to be difficult to beat. This data were drawn from a broader project including a companion study on “toxic terms,” aimed at addressing the extent to which our language is alienating voters and identifying common-language alternatives the public better understood and preferred.

 

Voters rated value statements across the major perennial and most contested current issues on a 0-100 scale, from strongly disagree to strongly agree. The statements were written in everyday language and designed to be emotionally compelling. They were not designed, and empirically did not, privilege any one point on the political or progressive spectrum. On every issue, we tested variations on a theme that would allow you to find highly rated messages that match or augment your voice, using language you would likely use and that you can use with confidence because it beats the most difficult, most disinforming language from the other side. The number next to each message tell you its average rating.

 

For perspective, average ratings across the voting population above 65 or strong. Numbers above 70, which we used as a cutoƯ, even though many of the messages in the high 60s beat corresponding opposition messages by 15-20 points, are extremely high. The presence of any in the 80s is unusual. Any message reported here is one we recommend using. If you want to see how they might play with your constituents, you can read the crosstabs here, which array the messages in descending order for each major demographic, although even with samples of this size, the numbers of participants, except for males vs. females, regions of the country, or college vs. non-college, are to be interpreted with caution.

 

We retained 87 items reflecting core American values which leaned strongly progressive but drew support from voters across the political spectrum, with a mean rating of 81.5 and a range of 71.1 ("The best way to protect our border is to hire as many immigration judges as new immigration agents, so they can decide in weeks, not years, whether someone is fleeing terror, torture, or tyranny in their native land before allowing them in") to 89.9 (“If you have to go to the ER, your biggest worry should be the reason you came in, not the bill you'll come out with”). As intended, opposition messages were strong as well, with a mean just under 60 (59.3) and a range of 47.9 to 72.9.

 

We have arrayed the statements in tables organized by issue, both to present the scientific results (statements and their corresponding ratings) and to provide messages that can be taken directly from the tables and used by political leaders and advocates.