It’s true that it sometimes seems like many years since Donald Trump was elected president. But his 2016 campaign shouldn’t be that hard to remember, as I observed this week at New York:

The conventional wisdom about Donald Trump is that he took a steadily growing extremist strain in the Republican Party, put it into hyperdrive, and won the presidency by polarizing the country to an unprecedented extent, boosting right-wing turnout and by various means suppressing Democratic turnout. In a classic contrarian take, Vox’s Matt Yglesias suggests that CW is based on selective memory and post-ex-facto reinterpretations based on the kind of president Trump has become, because many people who voted for him in 2016 perceived him as “moderate.”

“[I]t’s not true that Trump ran and won as an ideological extremist. He paired extremely offensive rhetoric on racial issues with positioning on key economic policy topics that led him to be perceived by the electorate as a whole as the most moderate GOP nominee in generations. His campaign was almost paint-by-numbers pragmatic moderation. He ditched a couple of unpopular GOP positions that were much cherished by party elites, like cutting Medicare benefits, delivered victory, and is beloved by the rank and file for it.”

Yglesias cites Pew data from the summer of 2016 to make the point that Trump was perceived by voters as more moderate than Hillary Clinton, and Gallup data to show that Trump was perceived as more moderate than past Republican nominees Mitt Romney, John McCain, or either Bush. Yes, Trump also ran on crude racial and cultural appeals that more conventionally conservative Republicans tend to deploy more subtly. But that enables him to appeal to a relatively small but crucial subset of Democrats who are moderate to liberal on economic and fiscal issues while being conservative on matters of culture and racial resentment:

“Had Trump ran on a conventional Republican platform of cutting Social Security and Medicare, Democrats would have hammered him for it — just as they hammered Bush and McCain and Romney — and won the votes of many older non-college whites who are racist enough to like Trump but sufficiently non-racist to have voted for Democrats in the past.”

Yglesias goes on to use this 2016 reminder to push back against the common assumption that Trump’s example shows Democrats that the old moderation-wins tenets of political science are outdated. Indeed, he revisits the academic support for the proposition that all other things being equal, being perceived as a moderate significantly enhances the prospects for victory. And then he makes a familiar if pointed observation about the left-leaning direction of the 2020 Democratic primaries:

“Win or lose the election, there’s just no way Democrats are going to be able to implement the Section 1325 repeal many of them promised at last week’s debate. But if they lose the election over charges of being soft on border enforcement, then they can’t do anything at all for Dreamers, for humane treatment of asylum seekers, or for a path to citizenship for the long-settled undocumented population. Taking an unpopular stand or two in pursuit of progress is fine.

“But extremism, like anything else, is best in moderation and ought to be saved for moments where the stakes are really high. Trump’s success in politics, meanwhile, confirms rather than debunks the basic political value of trying to take popular positions on the issues.”

The 2020 Democratic nominee, however, is not going to be running against the 2016 version of Donald Trump. Yglesias acknowledges that Trump the president has been far more satisfying to the right wing of his party than Trump the candidate, but doesn’t quite factor that into his prescription for Democrats. And it’s hardly incidental to any sound analysis of 2020, if only because Trump’s strategy for reelection is focused on base mobilization to an extraordinary extent. As my colleague Jonathan Chait pointed out last year, Trump has consolidated his Republican support in no small part by abandoning many of the policy positions that made him look moderate to voters in 2016:

“In office, he has instead governed as an orthodox right-winger. This explains why Trump has lost so much of his nonconservative support. But it also helps explain the Republican Party’s willingness to defend him. Instead of keeping his popular promises that helped get him elected, Trump instead adopted the unpopular stances of the conservative movement, which has in turn embraced him.”

The odds are very high that Trump will not be perceived as a “moderate” by voters after the kind of savage campaign he is almost certainly going to run, which should help the Democratic nominee appear more “moderate” by comparison. Alternatively, after months and months of Republican efforts to brand the Democratic candidate as a “socialist,” there may be no candidate “moderate” enough to overcome the imputed red hue.

All in all, Democrats would be well advised to stop worrying about ideological labels and focus on selecting a candidate who (a) looks strong against Trump in actual, empirical terms, and (b) would likely give them the greatest policy bang for the buck as president.