President Trump leaves the White House to board Marine One before a trip to France on July 12 in Washington. (Matt McClain/The Washington Post)

President Trump drew an interesting analogy on Twitter over the weekend.

The comparison is … a little strained. Clinton didn’t “illegally” get any debate questions; CNN contributor Donna Brazile sent a question to her campaign and lost her role with the network as a result. The 33,000 (or so) emails that were deleted from her personal server were determined by Clinton’s lawyers to not be work-related and therefore didn’t need to be turned over to the State Department.

But this isn’t really about drawing a one-to-one analogy between what Clinton did and what Trump’s son, Donald Trump Jr., did. (In case you were on vacation last week, Trump Jr. eagerly agreed to receive negative information about Clinton in a meeting in June 2016, despite being told that the information came from the Russian government.) Instead, it’s about continuing to position Hillary Clinton as his foil, well past the Nov. 8, 2016, sell-by date.

Why? A new survey from NBC News and the Wall Street Journal offers some insight.

The pollsters polled a number of counties that were significant in Trump’s 2016 victory. They fell into three categories: Counties Trump won, counties that Trump won despite Barack Obama winning them in 2012, and counties that Trump won by a much wider margin than Mitt Romney did four years prior.

In all Trump counties, half of respondents approve of the job he’s doing in office — a much better evaluation than Trump gets in the country on the whole.

In those counties that flipped, he is a bit underwater, with a net minus-7 approval rating — still well above the minus-22 net rating he received nationally in the new Post-ABC poll. In those counties where Trump surged, he’s very popular.

But.

As Nate Silver pointed out on Twitter on Sunday, it’s worth putting those figures into broader context. If you look at how Trump did on average during the 2016 election in each of those groupings of counties, there’s a distinct difference. On average, Trump won the counties he won by 42 percentage points last year; now, his approval is only four points above water. In those counties where he surged, he won by an average of 45 points, making his plus-16 net approval seem a bit less impressive.



That’s a misleading metric, though, since a lot of very small counties voted overwhelmingly for Trump. If we look at the total vote tally in each type — all the Trump votes in counties he won versus the overall total votes cast, for example — the differences are a bit smaller. But there are still differences.

We’re comparing apples and oranges a bit here, certainly. Job approval isn’t the same as a vote in an election (though there’s obviously some correlation). But that Trump’s job performance isn’t a (mostly) binary choice between him and Clinton is also disadvantageous to him. How you feel about Clinton is largely irrelevant to how you feel Trump is doing in office.

Hence tweets like the one from this weekend. Trump is fond of reminding his supporters what the alternative might have been as a means of reminding them why they voted for him in the first place. Sure, you may not love what Trump’s up to. But, he hopes the argument goes, at least he ain’t Clinton.

Is this a sound strategy? It depends on what the desired outcome is. The approval numbers, though, suggest that trouble looms in the distant future, unless Trump can convince voters in those swing counties that he’s doing more for them than simply not being his onetime Democratic opponent.