Trump approval falls more in KY than any other Southern state

Trump Net Approval Drop
From the Morning Consult tracking poll

Morning Consult has released their latest monthly poll on President Trump’s net popularity, and the Kentucky numbers are a little surprising.

The “net popularity” number is less familiar than the straight-up “approval rating.” The net number is the difference between those who approve and those who disapprove, and the “swing” is the difference in that net number across two or more polling cycles.

So, if an elected official has 60% who approve of their performance, and 40% who disapprove, then that official has a +20-point net approval rating. If, on the other hand, we poll for that official a year later, and their approval has dropped to 45% (from 60%), and their disapproval has risen to 55%, then they have a net approval of -10 points … AND, a change or swing of 30 points (+20 to -10).

A 20-point net approval rating is pretty strong. And, a 30-point swing is pretty dramatic. So how is President Trump doing after 1.5 years in office?

His approval numbers have dropped in every state, which is not in itself surprising. Almost every president sees their approval highest at the start of their term, and then sees it fall as the term goes on. For Trump, that translates into an average net approval of +14 points at the start of his term, and a -3 point approval rating now, if you average all the states. His net approval has swung from -6 points in Louisiana to -31 point (!) in New Mexico.

Here’s the part that is interesting for Kentucky. Trump’s net approval has fallen -19 points in Kentucky. That is a greater swing than in any other southern state. And in fact, that is a greater drop in approval margin than in 35 other states.

Now, don’t misunderstand. Trump still has a net approval of +15 points in Kentucky. There are still a majority of Kentuckians who approve of the job he is doing. BUT, that number has fallen further in Kentucky than in any other state in the Old South.

Here’s a map showing the net approval point swing in each state:

Trump Net Approval Drop

And for the data geeks like myself, here are the actual numbers:

StateJan 17May 18Points Change
New Mexico17-14-31
New York8-21-29
Rhode Island-4-24-20
New Jersey2-16-18
North Dakota236-17
North Carolina182-16
New Hampshire1-15-16
South Carolina2514-11
West Virgina3727-10
South Dakota2114-7


We have asked for the cross-tabs for the poll. If we get them, we’ll either update this story or post a separate analysis.

Why are these numbers like this? Why Kentucky? Post your thoughts in the comments.


  • Kentucky was not a deep red Republican state until recently. I think conservative religion and its parallel wedge issues like gay marriage have swung the state red in rural areas, but Trump’s vulgarity and cruelty have also turned religious voters off. Just wait until international retaliation of Trump tariffs kicks in – tariffs on bourbon, cars, washing machines, etc – Kentuckians will not be in thrall to Trump after that.

  • This kind of polling gets really tricky with the sampling error.
    First you have the error that comes with sampling in two polls.
    You are dealing with error from two samples not merely one
    that adds real volatility to the sampling error.
    It has been too long ago for me to remember the equation for calculating
    the random error in such polling, but the plus or minus percentage points error
    no matter the sample size gets significantly wider.

    Then we have the issue of how large the sample size is for each of the states.

    if you have sample sizes of say 1000 in each state at each point in time, you are fine.
    But if this is being done with samples of say 3000 for all 50 states,
    that gives you a sample size of average about 60 in each state.
    In that case, the estimates would be pretty much a crap shoot
    because you are estimating a population parameter for each state on small samples with impossibly large random error intervals..
    No way to know if the second point is of any concern here or not.

    • I THINK I remember reading that they were up to 20,000 poll responses. But, I’d have to go back and look.

  • I suspect that some Kentuckians have realized that not only are coal jobs never coming back, despite what Trump and McConman have told them but that also many, many Kentuckians will be losing one sort of benefit or another. I also believe that disapproval of Beavin is translating to loss of confidence in Trump and his similar policies. Yes, Kentuckians can be hard-headed, but they can also learn from their mistakes.

    • Let us hope your last sentence is doubly true over these next three election cycles.

  • I would like to think it’s because our collective political IQ has increased, but if that were the case, the numbers would have been even better. I think those who are beginning to change their minds realize that Kentucky has yet to see any real benefit from his policies. Quite the contrary, our medical costs have risen, the coal industry has not been revived, and Bevin’s policies have likewise not benefited the middle and working poor and impoverished. Surely, at least a few have noticed that Trump and the truth are simply not acquainted and have paid enough attention to realize he hasn’t the first idea what he is doing. Then again, it could simply be a polling error. Given the number of Kentuckians who voted for him in the first place, it’s hard to be optimistic.

    • I dunno — he still has a majority that approve, it’s just much smaller. And, various polling outfits have noted that the Trump approval rating doesn’t necessarily flow down the ballot. I have hope for this fall.

  • I suspect it may be related to an overall republican backlash due to the way Governor Matt Bevins and the Repiblican controlled house and senate stole pensions from teachers and KY state employees