Trump’s coalition won the demographic battle. It’ll still lose the war.

Here’s one way to think about the 2016 election. We are witnessing a great race in this country between demographic and economic change that’s driving a new America, and reaction to those changes. On November 8, with a tremendous burst of speed, reaction to change caught up with change and surpassed it.

But is that advantage sustainable over the long haul, as change continues and reaction has to run ever faster simply to keep pace? Probably not. Those old legs will give out eventually, though we do not know exactly when. In the end, the race will be won by change — as it always is.

Looking back from 2032, we are far more likely to view the 2016 election as the last stand of America’s white working class, dreaming of a past that no longer exists, than as a fundamental transformation of the political system.

Consider the following. Democrats in 2016 benefited from a substantial shift in their direction from the white college-educated vote, which more than canceled out some decline in the strength of the minority vote. If these trends had been all there was to the election, the Democrats not only would have won, they would have increased their margin overall and in most states relative to 2012.

But that was not to be. Instead, their support crashed to historically low levels among white working-class (non-college) voters — a staggering 39-point deficit nationally, compared with 25 points in 2012 — pushing Democratic margins down in most states and allowing Trump to eke out an Electoral College victory, despite losing the popular vote. His narrowest victories took place in Florida and three Midwestern/Rust Belt states: Florida by 1.3 percentage points, Pennsylvania by 1 percentage point, Wisconsin by 0.9 points, and Michigan by 0.2 points.

Let’s take a closer look at these states and see how they might shape up going forward in the great race. In these states, Trump’s white working-class support ranged from 62 to 66 percent, which represented very sharp shifts toward the GOP, particularly in the Rust Belt states. Among minorities in these states — black, Latino, Asian, and those of “other” race, considered as a group — Clinton’s support ranged from 72 to 81 percent.

Over the next four cycles, to 2032, the share of white working-class eligible voters in each of these states is projected to drop by 8 to 9 percentage points, while minority voters continue their steady increase. (Depending on the state, there should also be small increases in white college voters.)

Here’s what all this means concretely, applied just one election ahead. If we assume that the support patterns from 2016, with their astronomically high white-working class support rates for Trump and relatively weak minority support rates for the Democratic candidate, hold in 2020, projected demographic shifts in the electorate would still, by themselves, produce a very different outcome.

The Democrats’ advantage in the national popular vote would bump up from a little more than 1 point to 3 points. Critically, this change would flip the Rust Belt trio of Pennsylvania, Michigan, and Wisconsin — plus Florida — back to the Democrats, producing a 303-235 victory for the Democratic candidate, even with the white working-class surge toward Trump replicated in 2020. In addition, Arizona, Georgia, and North Carolina, already very competitive in the 2016 election, would become even more contestable under this scenario.

And this is just one election ahead. Naturally, the effects of demographic change will be magnified the further away we get from 2016.

Therefore, all else equal, Trump or another Republican candidate will have to continue to increase white working-class margins or white working-class turnout (or both) to be successful in future cycles. But that will be a difficult task, to say the least, given just how high support from that population was for Trump in 2016.

Indeed, given that Trump has essentially promised to solve all the economic and social problems of his white working-class supporters by making America “great again,” he is liable to be judged harshly when, as seems likely, this does not come to pass. That will create an opening for Democrats to reach out to these voters with programs and ideas that might actually help their communities. It is vital that Democrats take advantage of that opening. Combined with the ongoing advantages Democrats have from demographic change — which is bringing states like Arizona, Georgia, and even Texas within their reach — that could create quite a formidable coalition.

This suggests that the rampant fear among Democrats that Trump and his Breitbart strategy are throwing up a permanent barrier to left advance — perhaps even threatening democracy itself — is overblown. He and his movement are clearly riding on demographic borrowed time. His greatest strength comes from the votes of less educated aging whites, who are declining. This is not to say that Trump’s populism will not continue to be a problem for some time, but rather that over the medium to long term, his movement has intrinsically limited growth potential.

In the end, the Trumpian populism of the 2010s will likely have no more staying power than the agrarian populism of the 1880s and ’90s, which was similarly driven by demographic groups on the decline and similarly undercut by ongoing structural change.

Today, as well, the great race, with some twists and turns, will be won by change, not the reaction to it.

Ruy Teixeira is a senior fellow at the Center for American Progress. His next book is The Optimistic Leftist: Why the 21st Century Will Be Better Than You Think, forthcoming from St. Martin’s Press in March.