On Tuesday, President Trump, in an extremely in-character move, sent out some weird poll-related tweets.
These missives are part of the president’s multi-day attempt to incite a feud with a group of four very progressive first-term congresswomen — Reps. Alexandria Ocasio-Cortez (N.Y.), Ilhan Omar (Minn.), Ayanna Pressley (Mass.) and Rashida Tlaib (Mich.), all of whom are women of color. Over the weekend, Trump tweeted that they should “go back and help fix the totally broken and crime infested places from which they came” and claimed that House Speaker Nancy Pelosi (D-Calif.) would likely help with the travel arrangements, inserting himself into a dispute the four, collectively known as “the Squad,” were having with Pelosi over border funding. But the racist attacks on these congresswomen, all of whom are U.S. citizens (including three who were born in the United States), quickly became the story instead. Since then, Trump has doubled down in his own defense.
There’s a lot to analyze in those two tweets. Trump’s emphasis on polling is particularly noteworthy, especially because of the way he has approached numbers such as these in the past. Trump clearly believes these progressive congresswomen are unpopular and wants to make sure the Democratic Party is strongly associated with them, because he thinks it will help him win reelection in 2020. The evidence Trump is citing appears to come from this Axios report, which talks through the results of a poll of likely voters who are white and have two years or less of college education.
As is often the case with Trump’s polling tweets, there are numerous red flags and analytical mistakes — with some kernels of truth sprinkled in.
First, Trump is using bad data. Axios broke some of the most important rules of poll reporting in this story: It failed to disclose basic methodological details or reveal who conducted the survey. If the pollster was a party operative who had a rooting interest in the conflict between Pelosi and Ocasio-Cortez, or a less-than-competent researcher who made simple methodological mistakes, we would have no way of knowing, because we don’t know who conducted it and we haven’t seen their more detailed data. But Axios pushed out this dubiously sourced data anyway, and it has now made it all the way to The White House.
Second, the poll doesn’t say what Trump claims it said. The president’s wording implies that Omar and Ocasio-Cortez have low national favorability ratings. But the poll only surveyed white voters who had two years of college or less — an already-Republican-leaning group. I would guess that in a national poll, Ocasio-Cortez and Omar would post better, though maybe still not great, favorability ratings.
Third, the poll doesn’t show what Trump claims that it shows. As HuffPost’s Ariel Edwards-Levy explains, the basic idea is that the poll — or at least, what we’ve seen of it — doesn’t actually ask voters whether they see Ocasio-Cortez, Omar and other progressive Democrats as the faces of the party. And even if voters did think of the Democrats as Ocasio-Cortez’s party, that perception would likely change by 2020. Democrats are in the middle of a presidential primary. The nominee ought to, at least temporarily, supplant Ocasio-Cortez as one of the most visible Democrats in the country, thanks to a wave of media attention and, likely, attacks from Trump himself. Ocasio-Cortez’s poll numbers may be interesting on an academic level, but Trump will ultimately end up running against someone else in 2020.
That being said, there are some real insights in Trump’s tweets. Most Americans don’t like socialism and most don’t want to vote for a socialist presidential candidate. Moreover, there is some evidence that candidates who are on the ideological extremes alienate swing voters and pay a price at the ballot box, though that evidence seems stronger in congressional races than at the presidential level. So trying to use Ocasio-Cortez and other highly progressive members to cast the party as too far to the left is not the worst political strategy.
But Trump’s overall analysis is sloppy. Americans generally did agree that his initial tweets were racist, and statements like these are part of the reason he has struggled to lift his approval rating out of the low to mid-40s. Trump really might draw a Democratic opponent who is too far to the left for swing voters, capitalize on the strong economy and lack of costly foreign wars and win again. But sending out racist tweets and defending them with bad elections analysis probably won’t help him get there.