Usually ESY is focused only on the House of Representative. This post will be a bit different in that it will cover the Electoral College, Senate, and House.
Sign up for ESY below! We won’t bother you except to let you know about fresh, new content. We promise!
President Trump’s approval rating hovers in the low 40s, a bad number for any president seeking reelection. He’s doing worse than any president since Jimmy Carter has at this point in their presidency. But Trump’s approval differs wildly depending on the demographic group. This post will take a look at the president’s approval among non-college educated whites, college educated whites, blacks, and Latinos. And we’ll go deeper, looking at regional variations within these groups, a crucial dynamic that is frequently ignored or overlooked.
First, we’ll look at President Trump’s approval in key Electoral College states, then we’ll look at important Senate states, then the House battleground districts.
Trump’s Approval in Electoral College Swing States
The swing states can be separated into two broad categories of the Rust Belt and Sun Belt. The Rust Belt comprises much of the Midwest and Northern U.S. and the swing states of Michigan, Pennsylvania, Wisconsin, Minnesota, Iowa, and Ohio. I’m also including Maine and New Hampshire in the category due to their similar geographic and demographic characteristics.
The Sun Belt, sweeping from the southwest to southeast, includes the swing states of Arizona, Nevada, Texas, Georgia, and North Carolina. I’ll also stretch the Sun Belt definition and add in Colorado and Virginia.
Data aggregated from Gallup’s 2018 polling helps sketch the president’s approval among different demographics in these swing states. Unfortunately, the granular data is not publicly available, but this article in the Atlantic did provide some of the polling numbers, which are in the chart below. Clearly there are some holes, but the data available together with piece’s additional commentary fill in much of the big picture gaps. Sources: Daily Kos, Morning Consult, Atlantic/Gallup
Additionally, if we fill in the data holes with numbers from Civiqs, we get a similar picture. The graph below has this supplemental data color coated in red. To avoid mixing data sources, we will use Gallup’s numbers from the the first chart in the following analysis. This graph is just meant to confirm that the big picture painted by Gallup’s data is on mostly on target.
Sources: Daily Kos, Morning Consult, Atlantic/Gallup, Civiqs
Trump’s overall approval is higher in the Sun Belt versus the Rust Belt — averaging 46% versus 43%. This is a good sign for Democrats who think that the way to an Electoral College victory is through Michigan, Pennsylvania, and Wisconsin.
Looking at demographics, it’s not surprising that Trump does best with non-college whites, then college whites, then non-whites. What’s interesting is how much these groups differ from the Sun Belt to the Rust Belt. Trump’s approval in the Rust Belt among non-college whites is 11% higher (63% to 52%) than in the Sun Belt. This Demographic makes up 60% of the population in the Rust Belt and 40% in the Sun Belt, meaning that Trumps struggle in the Rust Belt is especially significant given how dominant these voters are in the region.
Trump’s approval among college-whites is 7% higher (44% to 37%) in the Sun Belt than the Rust Belt. And his approval among nonwhites is 4% lower. This coalition of college-educated whites and minorities has swung away from Trump and the Republicans, largely responsible for delivering Democrats the House in 2018. Given that the college-educated white and non-white coalition is more significant in the Sun Belt — where it makes up 58% of the population to the Rust Belt’s 40% — Democrats will need to win over and turn out these voters if they want to win any of Sun Belt states. The problem for Democrats, however, is that Trump does better among the coalition (7% better among college educated whites and 4% among nonwhites) in the Sun Belt than the Rust Belt.
Trump’s Approval in Senate Swing States
The story is largely the same in the Senate races, many of which overlap with the Electoral College swing states.
Sources: Daily Kos, Morning Consult, Atlantic/Gallup
The big picture: Trump does worse in the Rust Belt among non-college whites and better in the Sun Belt with college-educated whites and minorities.
The big outlier among these states is Alabama. This is the only state that is competitive for Senate but not for the Electoral College. Here, incumbent Democrat Doug Jones won a 2017 special election due to a weak and scandal plagued opponent, Roy Moore. Moore is running again, but if he loses the Republican nomination, Jones faces a tough road in a state that voted for Trump by 28% and in which the president has a 60% approval rating and his highest net approval of any state.
Trump’s Approval in House Swing Districts
There’s no way to get Trump’s approval ratings in each of the 51 swing House districts, so instead we’ll look at the demographics and regional characteristics. First up, the 17 battlegrounds in the Rust Belt.Sources: Daily Kos, CityLab
The districts are heavily suburban (14 of the 17 are predominantly suburban according to CityLab’s Density Index), overwhelmingly white (89%), and largely non-college white (59%). These demographics line up almost perfectly with the Rust Belt states overall, which are 88% white and 60% non-college white, meaning that Trumps average approval in these House seats is probably around the 43% average of the Rust Belt states overall.
Again, these districts are also heavily suburban, with 11 of the 12 having predominant suburban characteristics. These districts, unlike Rust Belt battlegrounds, do not match as neatly with the Sun Belt’s overall demographics. The districts have a slightly lower white percentage relative to the Sun Belt overall (62% to 65%), but the big difference is in their percentage of white non-college voters, which is 32% versus the Sun Belt’s average of 42%. This lower percentage of Trump’s biggest supporters would probably drag his approval rating at least a few points down from the Sun Belt’s average of 46%.
These districts, scattered across the U.S. are also heavily suburban (19 of 22), and more in line with national demographics than either the Sun or Rust Belt districts. They are 74% white and 47% non-college white, compared to the national averages of 69% and 47%. And, given that Trump’s national approval hovers around 41% to 43%, that’s likely where the average of these districts stands as well, although perhaps a point or two lower given the 5% difference in white voters.
Trump’s approval rating will be one of the most important factors in these House races, as well as in the Senate and Electoral College. We often hear about Trump’s approval rating among different demographic groups, but that analysis often lacks important regional nuance. Regardless of their top-line demographic identifiers, voters differ heavily on their views of the president depending on where they live.