Correction thread. tl;dr violent partisan views are rising a bit, but not by double-digits. This is a 1st effort to clarify last week’s Politico article w/ new, better data, plus an editor’s note on the article. I won’t post the link to avoid giving those #s more attention. 1/

Last week, @LilyMasonPhD & I co-wrote a Politico article on the topic w/ another research team. We combined our 2017-18 YouGov data (YG), w/ Dec ‘19 Voter Study Group (YG) data, plus Nationscape (Lucid) data from June & Sept ’20. 2/

We saw a slowly rising trend in the YG data, 2017-19. The new Lucid data showed a summer ’20 rise. But there was a huge jump from Dec ’19 YG to June ’20 Lucid w/ eye-popping violent views—a big red flag. Lily & I were skeptical, & we trust YG to be more representative. 3/

But past VSG & Nationscape data aligned, & Lily & I had Jan ’20 Lucid data only 5 pts more violent than Dec ‘19 YG. W/ spring turmoil & past benchmarks, a big rise was possible. Even so, we steered edits to focus on w/in sample trends. Part/all of a jump could be an artifact. /4

We thought a w/in Lucid trend was more reliable than its baselines. But a Politico figure most of us didn’t see before publication implied a continuous trend w/ a huge 2020 leap, & news/Twitter focused on the eye-popping Lucid baselines, not the trends. /5

Then, we unexpectedly got Sept ’20 VSG data the next day that showed signs of a small rise, but nothing like Lucid baselines. IOW, the most comparable & representative data shows a slow rise continuing the trends across years. That’s still concerning, but not sensational. /6

We don’t yet know why the numbers are different, but, speaking for myself, I now have no confidence in the Lucid baselines. I'm sorry for this. We had hoped to update the article, but Politico will only grant an editor’s note. I hope this begins to set the record straight. /7

Thanks to @rp_griffin & our co-authors working together on updates w/ our new data, incl. how it causes us to reevaluate the Nationscape/Lucid baselines. We stand by the article concerns & proposals for avoiding violence on the front & back ends of the piece. /8

Here are some comparable updates from the new YouGov analysis. In Sept, 14 percent of Republicans and 18 percent of Democrats said that violence could be justified to advance their parties’ political goals—an increase since 2017 when 8 percent of both groups said the same. /9

In Sept, 23% of Reps & 28% of Dems said there is at least “a little” justification for violence if the other party wins the election. Those figures are both up from 2017 when 12% of Reps & 18% of Dems said the same. 9% of strong partisans see “a great deal” of justification. /10

June & Sept 2020 Nationscape data partly corroborates a rise in the number of partisan respondents who said there was some justification of violence in the event of a 2020 presidential loss (+9%) or to advance political goals (+5%). /end

Good grief--reversed question labels. 2020 loss scenario is always higher. The weekend can't come soon enough.