Everyone knows that President Trump is historically unpopular and his low approval rating is putting Republicans in real danger of losing at least the House in 2018. At this point, that’s old news.
But not everyone has a good grasp on the granular parts of presidential approval—how intensely people feel about the president, which groups do and don’t like him and how much that matters for the midterms and our understanding of public opinion in the Trump era more generally.
That’s why I spent some time with a new, centralized dataset from SurveyMonkey that deals with these questions. I looked through their data, did some math, and came up with four different charts that demonstrate some of the most important facts about Trump’s approval rating.
Before getting started, I have one quick note for readers who might be confused because they have personal experience either making or filling out very non-political polls on SurveyMonkey. This is that same SurveyMonkey – the company just also usse its platform to conduct scientific political surveys. Mark Blumenthal, head of election polling, explained it to me: “We draw random samples of those who have just completed a survey on our platform and invite them to answer a few more questions about important issues and current events, including Trump approval” and then they “statistically adjust the data so it is representative of the American public.”
So here are the graphics and the explanations:
There Aren’t (And Never Have Been) Many Lukewarm Trump Disapprovers
The conventional wisdom on politics is often wrong. But these polls show that it’s right about one thing – Americans who don’t like Trump really don’t like him.
This graphic shows the percentage of Trump disapprovers who “strongly” or “somewhat” disapprove of Trump’s job performance in a series of SurveyMonkey polls stretching from January 2017 to now.
The top line tells most of the story. If you put every American who disapproves of Trump in a room (usually more than half of the people that SurveyMonkey polls) and picked someone randomly, the odds of picking a “strong disapprover” would be greater than 3-to-1.
This is a problem for down-ballot Republicans. Democrats have won some high profile races (like the special election in Pennsylvania’s 18th District) partially by riding strong anti-Trump sentiment. The president has motived Democrats to turn out at a high rate while driving some more typical GOP voters to cast their ballot for the blue team. That anti-Trump enthusiasm is part of the reason Democrats are currently the favorites to take the House in November.
Trump’s Base is More Divided Than Many Think
The conventional wisdom is a little less helpful when it comes to Trump supporters. Diehard Trump supporters have garnered a lot of media attention, but not every Trump approver is enthusiastic.
This graphic is nearly identical to the previous one—the difference is that we’re comparing strong Trump approvers and those who only somewhat approve of the president’s job performance.
In most of SurveyMonkey’s polls, strong Trump approvers outnumbered those who approved somewhat, but the split isn’t as lopsided as it is Trump disapprovers (this is similar to what we’ve seen in YouGov polls of Trump voters). Maybe more interestingly, the mix of somewhat and strong Trump approvers seems to have changed at least a little bit over time. Every poll is noisy, so these shifts shouldn’t be overinterpreted. But the lack of stability in this graphic compared to the last one (plus real changes we’ve seen in topline averages) suggests that Trump supporters aren’t immovable. Sometimes they go from supporting Trump enthusiastically to being more lukewarm. And sometimes they go from approving of Trump to not approving (or back to approval).
This is another problem for the GOP in the midterm elections. A drop in Republican turnout (which we saw in numerous races in both 2017 and 2018) combined with strong Democratic turnout and crossover voting would add up to big problems for the House and Senate GOP. Figuring out the exact odds of a Democratic House or Senate takeover is a trickier task, but lower GOP turnout is one possible ingredient of a bad midterm performance.
The Demographic Patterns Are Pretty Familiar
SurveyMonkey also broke its data down by demographic groups. We can look at four key groups simultaneously – college educated whites, non-college educated whites, non-white college graduates and non-white Americans without a college degree.
Trump’s net approval rating is consistently positive only with one group—blue collar whites. This group is an important part of the Trump coalition. They carried him across the finish line in states like Wisconsin, Pennsylvania, Michigan, Iowa, and Ohio in 2016. But the broader group isn’t completely in his corner. SurveyMonkey often showed that a solid minority (often more than 40 percent) of non-college-educated white adults (a broader group than blue collar white voters) disapproved of Trump’s job performance.
Nonwhite college graduates and nonwhites without a college degree aren’t the most commonly used demographic groups. So to get a sense of where Trump stands with different racial groups, I plotted his approval rating with African-American, white and Hispanic Americans.
These patterns should be familiar.
Black voters are a big part of the Democratic base, so strong disapproval of Trump (who has, to make an extreme understatement, not always been racially sensitive) makes sense. The same goes for Hispanics – it’s a young, Democratic group and Trump has done little to win them over. Whites, on the other hand, seem to bounce around with the overall approval of blue collar whites and the disapproval of college-educated whites canceling out to a net rating that’s closer to zero.
There are some problems for the GOP in these numbers. As Dave Wasserman has pointed out, college-educated voters often turn have higher midterm turnout than other groups. If those voters continue to disapprove of Trump, they might vote against Republican House members, endangering them in key districts. Nonwhite voters often don’t turn out in high numbers during midterm elections, but they did in Alabama’s special election—and it’s impossible to know ahead of time whether that’s the new normal or not. In any case, the GOP seems to be dealing with something that most parties in power go through —they overreach in their first two years, get punished by the public and then have to try to regroup before the next presidential election.
