Republican presidential nominee Donald Trump and Democratic presidential nominee Hillary Clinton speak in 2016 during the second presidential debate at Washington University in St. Louis. (John Locher,/Associated Press)
Opinion writer

The latest Monmouth University poll reports, “In considering who should be their party’s standard bearer, a majority of 56% prefer someone who would be a strong candidate against Trump even if they disagree with that candidate on most issues. Just 33% say they would prefer a nominee who they are aligned with on the issues even if that person would have a hard time beating Trump. Democratic women (61%) are more likely than men (45%) to say they would put their policy positions aside in order to get a nominee who could beat Trump.” The pollsters point out that in the past, “both parties consistently prioritized shared values over electability when selecting a nominee.”

If Democrats “may be willing to flip that equation in 2020 because of their desire to defeat Trump” what does that mean for the primary?

To start with, we should be wary about expecting voters to distinguish between candidates they like and ones that are “electable.” Very often, voters become convinced that the persons they like will be liked by others. Presto -- instant electability!

Moreover, voters aren’t good at guessing who’s electable and who’s not. Pundits aren’t much better. One need only look at GOP opinion polling during the 2016 primary (and into the general election) to find loads of voters and commentators who didn’t think President Trump was electable. Likewise, in 2008 it was widely assumed during the early part of the Democratic race that Hillary Clinton was the more electable of the two top contenders. Presidents Trump and Obama showed that initial assessments were wrong.

So how might Democrats desperate to win back the White House go about maximizing their party’s chances? Perhaps voters should work backwards. What should they be wary of?

First is the lesson of the 2018 midterms: Be wary of Sen. Bernie Sander-type ideological outliers. The Third Way, a group advocating a center-left party found: “Candidates backed by the mainstream New Democrat Coalition flipped at least 25 House seats from the red column to blue last night—that is the entire majority. By contrast, candidates endorsed by the Sanders-style Our Revolution and Justice Democrats failed to flip a single seat for Democrats.” Moreover:

Our election-eve survey with David Binder Research asked Democrats and Independents to think about the Democratic candidate in their district and asked if they would have preferred a more liberal, moderate, or conservative option. A plurality (37%) would not have preferred anyone different. Thirty-five percent would have preferred a more moderate (21%) or conservative option (14%). Just 16% said they would have preferred a more liberal candidate.

And these mainstream candidates delivered in battleground races last night, while the Sanders-style standard bearers faltered.

Now, that doesn’t mean Democrats shouldn’t pick a progressive; but they should be wary of tapping the most progressive candidate in the pack.

Second, avoid a contender who seems destined to turn off as many or more people than he or she’s going to win over from the pool of independents, ex-Republicans and non-voters. Think of the people (women, college-educated whites) and the regions (the Upper Midwest, the Southwest) where Democrats didn’t do as well as they needed to in 2016 and where they came back in 2018. (It should have been a flashing red light to Democrats when Clinton lost Michigan well into the primary calendar.) The gap between the popular and electoral vote may be here to stay; that means broadening the available pool of voters in the electoral-rich states will be key.

And finally Democrats would do well to find the person -- not the person with the best rhetoric -- that matches up against Trump. Consider if a candidate is young or youngish, knowledgeable, idealistic, empathetic, high-energy and appropriate for a change election. Trump is none of those and therefore may be at a disadvantage in a side-by-side comparison. Democrats don’t want to leave possible advantages on the table, thereby making it harder to dislodge Trump.

There could be more than a dozen candidates in the race when the first Democratic primary votes are cast. None will perfectly match a particular voter’s policy stances (unless the voter is a blood relative or employee, and maybe not even then). However, Democrats would be smart to steer toward a candidate who isn’t on the ideological fringe and is less likely to provoke an intense backlash among groups (e.g. non-college-educated voters, who are a majority of Americans) who must vote in greater numbers for Democrats than they did in 2016. A candidate with wide geographic appeal who can best exploit Trump’s vulnerabilities (age, cluelessness, meanness, incompetence) may have a greater chance than candidates who are at a disadvantage in critical states and whose deficits overlap with Trump’s.

Like we said, figuring out electability isn’t easy, even if you want to make that the basis of your pick.