Back in July, Gallup released the best barometer of President Trump’s support in all 50 states, with reference to his national numbers. That poll was useful and informative, but there were serious flaws in its methodology, as I noted in a write up published at the time. Today, Morning Consult, a pollster notable for its approval rating numbers of senators and governors and also one of the more favorable polls for Trump, rolled out a 50 state poll of its own. Besides being Christmas for polling geeks like myself, it was also a huge boon of information about President Trump’s strength. It also solves many of the Gallup poll’s issues with methodology, as it actually isolates when responses came in, so you can compare approval rating in a given state by month. This piece contains some of my observations and notes about this poll:
- It is important to note firstly that the poll has Trump with a September approval rating of 43/52, which is one of the best polls I’ve seen for Trump in a while. However, that isn’t an unreasonable number given that this is a poll of voters, who consistently give Trump better ratings than all adults, and that September was perhaps the only “good” month for Trump as his approval ratings generally went up, as voters liked his debt ceiling deal and his initial responses to Hurricanes Harvey and Irma. As we are now in October, his numbers have slid back to where they were at the start of September and are close to pre-Hurricane, so like with the Gallup poll, assume his numbers today are a tad lower than they actually are.
- Unlike the Gallup poll, no result jumps out as super wonky, which is a good thing, generally speaking. However, the most surprising result is probably Trump’s very tight 49/45 approval rating in Nebraska. Nebraska is one of the reddest states of all, having last gone blue in 1964, during Lyndon Johnson’s landslide for the ages, so this is pretty shocking. For what it’s worth, incumbent Senator Deb Fischer (R) is up in 2018, but Democrats currently don’t have a great attack plan given that they’re defending so many seats and it’s freaking Nebraska. But this poll suggests it might not be thoughtless folly to give it a shot. If there is a place where this result could be most useful to Democrats, it’s probably in NE-02, the Omaha district that is Nebraska’s only swing house district. Ex-Rep. Brad Ashford is back after losing in 2016 to Don Bacon in a tight race. If Trump’s unpopularity in the state is centered in Omaha (and I suspect it is), that could be good news for Ashford and put Democrats one tiny step closer to a house majority
- Taking a look at upcoming elections in 2017, which are often a referendum on the president, we first look at Alabama where the senate special election to replace Jeff Sessions is getting interesting between former attorney Doug Jones (D) and the ever controversial Roy Moore (R). Jones is trying to become the first Democrat to win an Alabama senate race since 1992, when Richard Shelby won, and he was a classic Dixiecrat (he would switch parties in 1994). Simply put, it’s a major uphill climb for Democrats. And this poll confirms that. The third Trumpiest state in the union is Alabama, where voters give Trump a 59/35 rating. Also worth saying, New Jersey is still very blue, giving Trump a 38/57 rating which is honestly generous for that state, meaning that Phil Murphy (D), who has the governor race to replace Chris Christie on ice, will probably cruise to victory. In Virginia, Trump gets a 42/53 rating, which is also probably a little generous, but reaffirms the fact that Virginia Democrats have no excuse to lose any of the statewide races and should be tying their Republican opponents to Trump at any chance they get. There will be a lot more Virginia coverage coming in the final few weeks, including from me, but this is worth pointing out
- Looking at the Upper Midwest, sometimes called the Rust Belt, the numbers are a tad interesting. Trump won the presidency by narrowly squeaking out wins in MI/PA/WI. This began a firestorm of hot takes about the Midwest’s fall to conservatism, something that I have questioned the validity of and in many ways, 2018 will be a test of that theory, thanks to an abundance of crucial house, state legislature, gubernatorial, and senate races in those states. What about Trump? Well, as many polls have indicated and I speculated, Trump isn’t doing that great in the Midwest. His numbers are 42/53 in Iowa, 39/56 in Minnesota, 40/55 in Michigan, 41/53 in Wisconsin, 45/51 in Pennsylvania. Now of course those numbers have to be compared to the rest of the nation, which I do in the PVI section below, but it’s another data point in the wrench that is the “Midwest is turning red” theory.
- A quick check on 2018 races: Tennessee, a very red state that is now kind of important since Senator Bob Corker is retiring and trying to take Trump down with him, is the second weakest of the modern day Solid South states (AR, LA, MS, AL, TN, SC), behind South Carolina, for Trump. Trump’s number in Tennessee is 53/42 suggesting that there is a tiny opening for Iraq War veteran and outsider James Mackler to make something happen. That’s a Hail Mary for sure but it’s intriguing. Trump is at 51/44 in Texas, a state where Ted Cruz is being challenged by new progressive favorite, Beto O’Rourke. Trump’s support in the Upper Plains states of Montana and North Dakota is pretty weak relative to the state (50/45 and 51/44, respectively), good news for potentially vulnerable Democratic Senators Jon Tester and Heidi Heitkamp who are both up in 2018. The same goes for Missouri and Indiana, two states that went hard for Trump in 2016. There, Trump’s support is just 49/45 and 50/45, respectively. Both states are home to the two most endangered Democrats in the senate, Claire McCaskill and Joe Donnelly. West Virginia still loves themselves some Donald Trump, giving him a 60/36 approval rating, which is bit higher than what West Virginia Metro Times found, but the topline is still the same. The good news for Democrats is that Joe Manchin (D) has built a very tough to beat image in that state as a favorite son. Finally and perhaps most interestingly, Trump sports a 44/51 approval rating in Arizona, where embattled Senator Jeff Flake (R) is in a lot of trouble, facing real opponents from the right and the left, suggesting a huge opportunity for Democrats.
- Lastly, I’ve compiled a PVI table for this poll. As many of you know, PVI is a metric developed by Cook Political Report to measure a state’s partisan lean in elections. You simply take the amount the person got nationwide and compare the amount in that state and work out the difference:
PVI Chart
State |
poll pvi |
2012 pvi |
2016 pvi |
ak |
R+9 |
R+9 |
R+11 |
AL |
R+16 |
R+13 |
R+14 |
AR |
R+11 |
R+14 |
R+14 |
AZ |
R+1 |
R+7 |
R+3 |
CA |
D+7 |
D+10 |
D+14 |
CO |
D+2 |
D+1 |
EVEN |
CT |
D+7 |
D+7 |
D+7 |
DC |
D+26 |
D+40 |
D+43 |
DE |
D+5 |
D+7 |
D+5 |
FL |
R+4 |
R+2 |
EVEN |
GA |
R+4 |
R+6 |
R+3 |
HI |
D+12 |
D+20 |
D+14 |
IA |
D+1 |
D+1 |
R+6 |
ID |
R+12 |
R+18 |
R+20 |
IL |
D+6 |
D+7 |
D+8 |
IN |
R+7 |
R+7 |
R+10 |
KS |
R+7 |
R+13 |
R+12 |
KY |
R+11 |
R+14 |
R+15 |
LA |
R+14 |
R+11 |
R+10 |
MA |
D+11 |
D+10 |
D+14 |
MD |
D+10 |
D+11 |
D+14 |
ME |
D+1 |
D+6 |
EVEN |
MI |
D+3 |
D+3 |
R+1 |
MN |
D+4 |
D+2 |
R+2 |
MO |
R+7 |
R+7 |
R+10 |
MS |
R+11 |
R+8 |
R+8 |
MT |
R+7 |
R+9 |
R+12 |
NC |
D+2 |
R+3 |
R+2 |
ND |
R+8 |
R+12 |
R+21 |
NE |
R+7 |
R+13 |
R+14 |
NH |
D+5 |
D+1 |
R+1 |
NJ |
D+5 |
D+7 |
D+7 |
NM |
EVEN |
D+3 |
EVEN |
NV |
R+1 |
D+1 |
EVEN |
NY |
D+6 |
D+12 |
D+11 |
OH |
R+3 |
EVEN |
R+4 |
OK |
R+13 |
R+19 |
R+19 |
OR |
D+4 |
D+4 |
D+2 |
PA |
R+2 |
D+1 |
EVEN |
RI |
D+8 |
D+12 |
D+6 |
SC |
R+8 |
R+7 |
R+7 |
SD |
R+8 |
R+11 |
R+16 |
TN |
R+10 |
R+12 |
R+13 |
TX |
R+8 |
R+10 |
R+5 |
UT |
R+9 |
R+27 |
R+21 |
VA |
D+1 |
EVEN |
D+2 |
VT |
D+10 |
D+16 |
D+9 |
WA |
D+6 |
D+6 |
D+5 |
WI |
D+2 |
D+2 |
R+1 |
WV |
R+17 |
R+16 |
R+22 |
WY |
R+18 |
R+23 |
R+26 |
This chart is not perfect, and there are downsides to the methodology of PVI (for instance, Trump got less percent of the vote of Minnesota when compared to Romney, but did better in PVI because PVI is calculated based on the victor (of the popular vote)’s margin). But it’s interesting to look at. For the most part, it is indicative of a map that looks a lot more like 2012 than 2016, especially in Iowa, Michigan, Minnesota, and Wisconsin (not sure what’s going on in Pennsylvania). The poll as a whole probably underestimates distaste for Trump in blue states like California, but it’s much more useful and instructive than the Gallup one and suggests big openings for Democrats in 2018 and beyond.