A terrible long-term strategy.

And it really might be long-term. One source quoted here thinks Republicans are likely to lose the popular vote in election after election for the foreseeable future thanks to the Democratic skew of the most populous states. Only one state in the top 10 by population is red and seems to be trending conspicuously redder in recent years — Ohio, ironically thought of not long ago as the ultimate swing state. All the rest are either blue leviathans like California and New York, deep purple like Pennsylvania and Florida, or red strongholds that look to be getting bluer, like Texas and Georgia. If you have a roughly 50/50 country *except* for CA, NY, and IL then guess what: You’re going to lose the popular vote more often than not. Whether it’s Trump or someone else at the top of the ballot.

Senior Republicans are resigned to President Trump losing the popular vote in 2020, conceding the limits of the flamboyant incumbent’s political appeal and revealing just how central the Electoral College has become to the party’s White House prospects…

“California, Illinois, and New York, make it very, very difficult for anybody on our side to ever again to win the popular vote,” said David Carney, a Republican strategist in New Hampshire. Asked if he expects Trump to defy the odds next year, Carney said flatly, “No,” but added, “the president shouldn’t worry about it. Two hundred seventy — that’s what people remember.”…

“The popular vote is irrelevant because it’s not how our system works,” said Dan Eberhart, an energy executive and Republican donor in Arizona who supports Trump. “The president is right to focus on voters in states with the biggest number of electoral votes.”

Trump’s own team is “privately bracing for a second popular vote defeat,” claims David Drucker.

Two things make this terrible strategy. First, even if it’s true (as some sources claimed to Drucker) that Trump is more prone to a PV/EV split than the average Republican nominee because he insists on playing only to his base, the odds of such a split happening remain low. In September 2016, with Trump running a similar game plan, FiveThirtyEight estimated the odds of a PV/EV split at just 6.1 percent. By the end of the campaign, those odds had increased but only by a little, to 10 percent or so. It takes a perfect storm win the electoral vote while losing the popular vote: You need to somehow pull out very narrow wins in one swing state after another while getting pasted in your opponent’s strongholds and maybe even ceding some ground to them in your own strongholds.

Which is exactly what happened with Trump. He walked a tightrope in 2016, suffering landslides in blue states, hanging on for closer-than-expected wins in red states like Arizona and Georgia, but coming through with razor-thin victories in key states like Florida, North Carolina, and the Rust Belt. You’ve heard the numbers many times about how close the vote was in Wisconsin, Michigan, and Pennsylvania and that was against an extremely unpopular Democratic nominee. To achieve another PV/EV split, Trump somehow has to duplicate all of that, especially the coin-flip results in the midwest. It’s hard under the best circumstances and will be harder if the new Democratic nominee is reasonably popular. (Whoever it ends up being is almost a lock to be better liked than Hillary.) I haven’t seen any recent estimates of the probability of a PV/EV split in 2020 but I’d guess it’s much more likely that Trump wins both the popular vote and the electoral vote on the strength of a great economy than that he somehow threads the needle for a second term while losing the popular vote.

So why not focus on trying to do that? Reach out a bit to the middle and look to expand his popularity instead of playing another long longshot via a PV/EV split. Or is it already too late to try to do that?

The second reason this is terrible is that the country won’t forever tolerate handing power to a party that consistently loses the popular vote. Republicans can (and do) celebrate the virtues of the electoral college all they want; the fact remains that many Americans, whose grasp our constitutional scheme isn’t real firm to begin with, will eventually find it impossible to reconcile the idea of democracy with a scheme in which one party semi-routinely takes power despite finishing second in overall votes. That sort of outcome has been tolerated historically because it’s been flukey. After Bush 2000 and Trump 2016 it’s not as flukey, and now per Drucker we have Team Trump and the GOP increasingly comfortable with the idea that PV/EV splits are fine and sustainable outcomes. They aren’t. The left is already clamoring to end the electoral college because they fear this outcome again in 2020. If it happens and Trump wins another term with fewer votes than his opponent, abolishing the EC will become a core part of the Democratic platform (at least until the country starts trending redder again). There will be a confrontation of some sort over our constitutional model, inevitably.

Already a collection of states worth 181 electoral votes has enacted legislation that will award their electoral votes to the popular-vote winner if/when enough states adopt the same bill to form a majority of 270. Not all of them are deep blue, either: One week ago today Colorado became the latest state to do it. The more this becomes a core issue for the left, the more they’ll seize election wins at the state level in purple states to pass bills like this. It’ll be awhile before they get anywhere near 270, but the more the legislation advances the more central this debate will become to America’s politics. And don’t forget: None other than Donald J. Trump is on record as saying he’d be fine with switching to a model in which the national popular vote winner becomes president.

In the meantime, though, he’s going to double-, triple-, and quadruple-down on dopey feuds and other goofy base panders that only the people who are already committed to voting for him enjoy.