Our Five Midterm Observations



1. Given the choice between extremism and the status quo, voters chose the latter.

Across not only swing states but solidly blue and red ones as well, our polling has consistently found voters feel that only the far right and the far left really get listened to in politics and that those in the middle get ignored.


The success of more moderate Democratic candidates against election deniers suggests that showing how the GOP would pose a risk to our country, rather than provide a check on the President, paid dividends.


In the end, voters cared more about keeping the far right out of office than punishing the administration. In fact, according to exit polls, Democrats won among voters who call themselves “moderate by double digits, 56-41. Democrats even won voters who “somewhat disapprove” of Joe Biden by 4 points, 49-45. And candidates who support QAnon conspiracy theories lost significant ground compared to 2020.



2. The winning case against GOP extremism was economic as well.

In early summer, we saw it was important to talk about how the GOP’s economic plans are “more extreme than ever before,” and their obsession with banning abortion would get in the way of addressing pocketbook issues. The threat Republicans pose to Social Security and Medicare appeared especially salient. While it appears that Republicans won just over half of voters over 50, Democrats likely reduced the GOP’s vote share among older voters by driving home this point in the final days.



3. GOP extremism on abortion can’t be denied as a major factor, in both turnout and persuasion.

On the persuasion side, we consistently saw in our research that highlighting Republican extremism on abortion was important - especially to address atttacks about “woke culture” coming from the right.


All 5 referendums on abortion (plus Nevada’s expanded Equal Rights Amendment) outpaced Democrats at the top of the ticket — obviously in red states, but even in blue ones too. This is just one reason we look at our data by different “assumptions” — while we guessed there would be increased enthusiasm to vote among Democrats infuriated with the Dobbs decision, we made sure our campaigns were ready for any turnout scenario (and that they knew just how much turnout would affect their odds).



4. (At least some) politics is (still) local.

Democrats outperformed expectations in red and purple states and at the same time underperformed in blue states like California and New York. The divergent trends suggest that while many campaigns focused on national dynamics (like inflation, abortion) and views of Biden during the campaigns, local dynamics (like homelessness) still played a major role in determining outcomes. As we saw in our research throughout the cycle, candidates and leaders need to think both about the national context and local dynamics to reach voters - not just one or the other.


5. Crime was a crucial issue for some, but Republican candidates may have underestimated concerns about gun crimes specifically.

Exit polls indicate as many voters said their top issue was “crime” in general as said “gun violence” specifically (11% each). But Democrats did better with voters worried about guns (+23 margin) than Republicans did with voters concerned about crime (+16). It’s possible that in elevating “crime” in their closing ads, Republicans drove backlash with voters concerned about gun crime. Dynamics like these are why we strongly believe message research needs to test for backlash effects.