The big gubernatorial victories in Kentucky and Louisiana have unleashed optimism in the Democratic Party. They underscored the story we’ve seen again and again in the Trump era: Democratic gains in the suburbs and among college-educated whites are helping Democrats assemble a potent anti-Trump coalition, even in some of the reddest parts of the country.

But behind the numbers, there is another reality lurking: It remains unclear how much this will matter in 2020 against President Trump — because of the peculiar nature of the advantage that Trump may hold in the electoral college.

A new analysis from David Wasserman for NBC News sheds fresh light on both sides of this story. It suggests that, perversely enough, the president’s advantage in the electoral college may actually be deepening, despite his party’s struggles in suburban and more educated parts of the country.

Which raises a question: Do Democrats have a plan for that? As it happens, there is an answer out there, and there are some signs they’re already gravitating toward it.

Wasserman looked at the Kentucky and Louisiana outcomes and concluded that they were plainly driven by big Democratic gains among suburban and educated voters.

Turnout skyrocketed, but on both sides, because Trump is energizing pro-Republican turnout, as well. The difference-maker was that Democrats engineered big swings in suburbs with a lot of college-educated voters, such as the areas outside New Orleans and in northern Kentucky, across the river from Cincinnati.

The crucial point here, though, is that Republicans also made gains — that is, they increased their margins, as well as turnout — in more rural parts of the country. Wasserman created this chart:

In the most educated areas in both states, the Democratic margin and turnout surged. But the same happened for Republicans in the remaining areas in the two states. On balance, in those states, Democrats still fared better from those competing dynamics.

But here’s the rub. Those same dynamics have real ramifications for Trump’s advantage in the electoral college. That’s because the states with the most-educated populations won’t be decisive in the electoral college, while the states with the less-educated populations will be crucial, particularly in the Rust Belt.

For instance, in Wisconsin, the percentage of eligible voters who are white and have college degrees is a low 27 percent. In Michigan and Pennsylvania, it’s 33 percent, according to Wasserman’s data. By contrast, in a dozen leading blue states, those voters average about 40 percent.

This means Democrats could run up huge margins in urban/suburban and cosmopolitan America, winning the popular vote by a greater margin than in 2016, while still losing in the electoral college. As Wasserman noted, the fact that Republicans are also improving in their areas suggests that Trump’s electoral college advantage — relative to the popular vote — may be deepening.

So what does this mean for Democrats in 2020?

It means they have to forget about debates over whether to prioritize the Rust Belt or the Sun Belt states — and instead treat it as a given that they must try just as hard in, say, Arizona, as in Wisconsin. Arizona, as it happens, has a higher population of eligible voters who are white and college-educated, at 39 percent.

Wisconsin will be incredibly hard-fought — Democrats there worry that Trump’s pool of support among non-college-educated whites could be even deeper this time — so Arizona might end up being essential to a Democratic electoral college victory, and must be treated as such. Florida should, as well (a win there for Democrats would essentially spell doom for Trump, pretty much no matter what happens in the Rust Belt).

A spokesman for Priorities USA, the biggest Democratic super PAC, tells me the group is already advertising in Arizona and Florida, along with Wisconsin, Pennsylvania and Michigan. This suggests Democrats grasp the need to keep open as many electoral college paths as possible.

This also means Democrats will have to forget about prioritizing winning back blue-collar whites over building a coalition of minorities, young voters and college-educated whites. It’s annoying to hear that Democrats will have to do both these things — it’s an obvious point, and the devil is in the details of how to get this balance right — but it’s just true.

That’s because we don’t know how deep Trump’s pool of support among non-college-educated whites will prove to be this time. So Democrats need to do all they can to win non-college-educated white voters, keeping Trump’s margin down among them to the greatest degree possible (and because it’s the right thing to do), while also maximizing turnout and vote share among those other constituencies and in the suburbs, to offset that Trumpian advantage.

As it happens, there are nonwhite, young, and college-educated voters in the Rust Belt states, as well, even if there are lower percentages of them, and that could help make the difference against Trump’s advantage with non-college-educated whites, too.

“The 2020 battlefield is not contained in one particular region,” Josh Schwerin, the Priorities USA spokesman, told me. “There are all types of voters in each of these states and oversimplifying the demographics would take votes off the table.”

It is often said that Democrats are in denial about Trump’s perfectly good shot at getting reelected, and about his lingering advantage in the electoral college. But much of what I’ve seen suggests they understand the electoral college challenges they face along these lines.

Read more: