Brad Parscale, manager of President Trump’s 2020 reelection campaign, told CBS News’s “Face the Nation” Sunday which states that the president lost in 2016 were pickup possibilities.

“Obviously, we have to go back and win Michigan again, Pennsylvania, Wisconsin,” Parscale said, referring to three states that Trump won by a combined 78,000 votes. “We plan on also being in Minnesota very soon. I think New Mexico is in play in 2020.

"I think New Hampshire, I think we continue to grow the map. I think Nevada, you know even Colorado. And so those are, those are states we did not win in, in 2016 that I think are open for 2020.”

As a general rule, states don’t suddenly jump from strongly supporting one major party to supporting the other. There’s often a longer-term trend that can suddenly benefit a candidate, as it did Trump in 2016.

Consider the three states Parscale mentioned at first. In 2008, all three states were from three to nine percentage points more Democratic than the country on the whole. In 2012, the states were from two to six points more Democratic. In 2016, they were each two or three points more Republican than the country overall — allowing Trump to claim victory in a contest where he trailed in the national popular vote by a little over two points. The shift from 2012 to 2016 was more dramatic than the shift from 2008 to 2012, but the shift was already underway.

Three of the states Parscale mentioned — Minnesota, Nevada and New Hampshire — look somewhat different. All three shifted to the GOP since 2012, but none of the three shifted enough to hand Trump a victory three years ago. In Nevada and New Hampshire, that shift followed a similar shift to voting more Republican from 2008 to 2012. (Again these are relative changes, so the fact that Barack Obama won by a narrower margin in 2012 than 2008 doesn’t affect these numbers.)

In the abstract, then, these states seem like possibilities for the Trump campaign. That ambition is complicated a bit by the results of the 2018 midterms, where Democrats easily held a Senate seat in Minnesota and picked up a Senate seat in Nevada.

Then there are New Mexico and Colorado. Trump’s team has had its eye on New Mexico for a while. Shortly before the 2016 election, then-communications staffer Jason Miller proclaimed that the campaign saw both Michigan and New Mexico in play. They were right about Michigan, barely. They were quite wrong about New Mexico.

New Mexico has consistently voted Democratic by a healthy margin although, relative to the national popular vote, that margin fell slightly in 2016. Colorado shifted more to the Democrats.

Democrats picked up House seats in both states in 2018.

What Parscale’s prediction ignores, too, is Republican states about which the campaign might want to be worried. According to the Washington Examiner, Trump has already moved to lock down Arizona, a traditionally red state that elected a Democrat to the Senate last year. Then there are Georgia and Texas, each of which (like Arizona) voted much more heavily Democratic in 2016 than in 2012.

In both Georgia and Texas, Democrats came close to picking up statewide seats in 2018 — but in each case, the Democrats lost by single-digit margins. None of these states is a gimme for Democrats. But Arizona voted Republican in 2016 relative to the country by a narrower margin than New Mexico voted relatively Democratic.

If the Democrats manage to pick up Texas in 2020, it’s not as though Trump winning Minnesota could offset the loss. Texas is worth 38 electoral votes to Minnesota’s 10. If Trump wins the five states Parscale floated on “Face the Nation,” Trump gains 34 electoral votes in total. Arizona and Georgia are another 27 electoral votes.

We noted in February that Trump should keep an eye on Texas, given that it was the state he won in 2016 where his approval rating was the lowest. In most states, Trump’s approval improved from 2017 to 2018 according to Gallup, but in most states Trump’s approval rating is underwater (meaning that more people disapprove than approve of the job he’s doing).

In three states mentioned by Parscale — Nevada, Pennsylvania and New Hampshire — Trump’s net approval rating dropped from 2017 to 2018. (Those are the dark blue arrows.) In the rest of those targeted states and in the three danger zones for Trump, his net approval rating improved.

In every one of the states mentioned above, though, Trump’s approval rating is at least eight points underwater. On average, it’s 14 points more negative than positive.

Notice that we didn’t talk about other states that Trump won narrowly in 2016, such as North Carolina or Florida. Those two states voted fairly consistently over the past few election cycles. In both states, Trump’s approval rating is underwater.

Trump’s team obviously wants to win more than the 30 states it picked up in 2016. There’s good reason to believe, though, that they’re going to be in a tough fight just to hold that many.