There’s nothing I enjoy more than a challenge to a deeply entrenched bit of conventional wisdom, and I had the chance to write about one this week for New York:

It’s generally assumed that if Donald Trump could just talk about the economy (or let the numbers do the talking for him) and lay off the racism and misogyny and other unpleasantness he so enjoys, he’d have a much better chance of reelection.

But now comes Tom Edsall with some contrary analysis suggesting that good economic times may not be so great for Trump, at least in the pivotal Midwest region:

“John C. Austin, director of the Michigan Economic Center and a nonresident senior fellow at Brookings, has explored the politics of the Midwest from a different vantage point. He examined income and other economic trends in 15 Midwestern congressional districts, including Pennsylvania, that went from Republican to Democrat [in 2018]. In the July 27 issue of Politico Magazine, Austin made a point of saying that the conventional wisdom is wrong. Contrary to the perception that a rebounding economy will work to the president’s benefit, there is growing evidence in Michigan and throughout the Rust Belt that metro areas that are bouncing back — and there are a bunch — are turning blue again. Austin noted that 10 of the 15 districts that flipped from Republican to Democratic in Michigan, Illinois, Minnesota, Pennsylvania, and Iowa ‘have income growth rates that exceed their state averages.’

“Of the remaining five flipped districts, in which growth was below the state average, three were in Pennsylvania, where Democratic victories resulted from a state Supreme Court decision ordering the replacement of the Republican gerrymander of congressional districts, making those districts much more favorable to Democratic candidates.”

Why would people apparently benefiting from “Trump’s economy” vote Democratic more often? Perhaps they are simply less resentful:

It’s also possible that the demographic groups benefiting most from “Trump’s economy” are more likely to lean Democratic to begin with than those in low-growth or struggling small towns and rural areas. In any event, the phenomenon Edsall notes was not limited to the Midwest in 2018:

“The Economic Innovation Group, a Washington think tank that studies regional inequality, ranked all 435 congressional districts into five groups based on their economic condition: the prosperous, the comfortable, the mid-tier, the at risk, and the distressed.

“An examination by the group of all of the congressional districts across the nation that flipped in 2018 from red to blue produced intriguing results.

“Of the 43 congressional districts that shifted from Republican to Democratic control, 23, more than half, were ranked as prosperous, and seven, or 16.3 percent, were ranked as comfortable. Altogether, almost 70 percent of the districts that switched from Republican to Democratic were ranked in the top-two economic categories.”

This kind of data suggests that since relatively good economic times didn’t help Trump’s party last year, they may not next year, either. And it may help explain why Trump has lately been opening the spigots on a fresh effusion of filthy racist invective: Maybe he really does need it to build the resentment toward “outsiders” that stimulates his base, whether by distracting them from their economic misery or offsetting the pacific effects of economic growth.

Edsall also notes that a major noneconomic trend is working in Trump’s favor as he tries to replicate his 2016 Midwestern success: The region’s population continues to age.

“In five states — Michigan, Ohio, Pennsylvania, Minnesota, and Wisconsin — the number of 18-to-35-year-olds, the most liberal age group, grew by 56,448 between 2016 and 2018, according to Mark Muro, a senior fellow at Brookings.

“That growth pales in comparison with the rising number of people 65 and older, a core of Republican support, which grew by 685,005 — an advantage of better than 12 old people for each young person. ”

The underlying reality may be that Trump is far more dependent on grumpy old men living outside thriving metropolitan areas than on the diverse populations living in them. If so, we can expect him to keep the hatefest going and to ignore the advice of “experts” telling him that his theme song should be “Happy Days Are Here Again.”

I love political history, so I felt constrained to comment at New York on an effort to draw a lesson from a famous campaign from the past in dealing with Donald Trump:

One of the raging debates among Democrats and progressives going into the 2020 elections is whether the president’s white-nationalist tendencies should be central to the campaign to oust him or whether, instead, (a) the focus should be on his corruption and the broken promises to his voters, or (b) his opponent should largely ignore him and emphasize her or his own policy proposals. The discussion over this question is heavily influenced by fears that white working-class voters are amenable to Donald Trump’s racial appeals and are better persuaded by appeals on other issues.

But when Trump is so blatant in his racism, as he has been since his July 14 hate tweets against four nonwhite members of Congress, it feels morally as well as politically feeble to try to take race off the table for 2020. Now comes a voice speaking from the experience of a fight against a far more notorious racist demagogue — Louisiana’s David Duke, a former Imperial Wizard of the Ku Klux Klan. Anti-racism activist Tim Wise, a veteran of two anti-Duke campaigns (a Senate race in 1990 and the famous gubernatorial “Race From Hell” in 1991), argues in a Washington Post op-ed that you have to call out racism vocally and relentlessly:

“To win an election where the issue of race is front-and-center, anti-racists must make it clear to voters that when they cast their ballots, they are making a moral choice about the kind of people they want to be and the kind of nation in which they want to live.”

In opposing Duke’s 1990 challenge to Democratic senator Bennett Johnston, Wise remembers, he and his colleagues in an anti-racist organization were advised not to play Duke’s racial game:

“Our organization, which worked independently of Johnston’s campaign, saw Duke’s racism as the issue. But we had consultants telling us similarly not to focus on it too much: Point out his Klan past and affiliations with white supremacist groups, we were told, but don’t try to underscore or challenge his contemporary racial messaging. That would ‘play into his hands,’ they said. They encouraged us, instead, to talk about reports of his delinquent taxes and avoiding military service.

“So we played that game, and the results weren’t pretty. We ran an expensive TV ad in which we mixed the messages, mentioning Duke’s white supremacist ties alongside his tax history and failure to serve in Vietnam, as if those issues were of equal importance. Highly stylized, the ad seemed crafted more to win awards than to drive voters. The result? Duke got 44 percent of the vote, with about 60 percent of the white vote. He lost, but Duke-ism had proved itself potent.”

A year later, Duke ran for governor, and, to the horror of the nation, he led incumbent Republican governor Buddy Roemer in Louisiana’s nonpartisan “jungle primary” to win a runoff post against the ethically tarnished Democratic warhorse Edwin Edwards. At that point, says Wise, those counseling a nonconfrontational approach to Duke were ignored:

“Unlike 1990, the message in 1991 was all about the fundamental danger posed by hate to Louisiana and America. Even the message that businesses and tourists would boycott the state if Duke won was ultimately rooted in a moral imperative. After all, it was his extremism that would drive companies and tourists away, and rightly so. Our bumper stickers that read, ‘Vote for the crook, it’s important,’ operated on the premise that whatever one might think of Duke’s opponent, then-former Democratic governor Edwin Edwards and his ethically challenged past, Duke’s racism was worse.

“In that election, hope won, not hate, even though Duke still won the white vote in the eventual, decisive runoff. He got more total votes in 1991 than in 1990, but his share fell to 39 percent overall and about 55 percent among whites, in part because racially progressive whites showed up in larger numbers, inspired by the moral message. And black turnout surged.”

And so, argues Wise, the same strategy represents the best and perhaps the only way to take on Donald Trump:

“The lesson now for Democrats is that they must make this election about the threat of Trumpism, which is racist at its core. That doesn’t mean that policy ideas aren’t important, but first and foremost, it’s about making it clear to voters what the stakes are. No issue — climate, jobs, health coverage — overrides the importance of getting a bigot with authoritarian tendencies out of office. Focusing on look-how-much-I’ve-thought-about-this-stuff might make for good primary-debate theater, but it’s not going to move the needle in 2020.”

Personally, it strikes me as probable that this full-frontal attack on Trump’s racism is necessary but not sufficient to the project of ejecting him from the White House. Wise is right that ignoring Trump’s white-nationalist appeals normalizes his behavior and makes it acceptable as conventional politics when it needs to be made shameful in respectable society. David Duke famously made himself and his modern Klan the relatively respectable face of white racism — the “man in the gray flannel hood.” He campaigned on anti-big-government, welfare-reform, and crime-control dog whistles, and opponents who accepted the legitimacy of these appeals played his game.

But a 1991 gubernatorial campaign is hardly a template for a 2020 presidential election. Trump is president of the United States and the unquestioned leader of the Republican Party, not a state legislator who has been repudiated by most of his party’s elected officials (including, as Wise notes, then-President George H.W. Bush). The business community that largely united to repudiate Duke as deadly to Louisiana’s economy has more or less made its peace with Trump, particularly given his business-friendly mix of anti-regulation and anti-tax initiatives laced with goose-the-gas fiscal and monetary policies. And as shocking as Trump’s history of race-baiting is, it hardly holds a candle to Duke’s.

As veteran Louisiana journalist John Maginnis explains in his brilliant account of the “Race From Hell” (in his book Cross to Bear), Duke might have won had the Klan been the only skeleton in his closet. What seems to have turned the contest around was the circulation of images of Duke wearing Nazi garb to protest a Jewish leftist speaker at an event at LSU. In a televised debate near the end of the campaign, Edwards jolted the audience when he responded to a standard Duke rap on welfare reform by saying, “David, I was working on welfare reform back when you were still goose-stepping around Baton Rouge.” At a time when many World War II veterans were still alive and voting, Duke’s identification with this particular brand of racism was disastrous to him.

The moral of the story is that racism must be called out, particularly by the party that depends so heavily on nonwhite voter support and professes a commitment to equality and justice. But Trump isn’t going to make it easy by goose-stepping. It will take other appeals on the economy, on health care, on climate change, on this administration’s corruption, and, yes, on the policy thinking of the Democratic nominee, to make the values and interests of a majority of the electorate converge.

We all know how polarized the political climate has become immediately before and during the Trump Era. But it’s steadily getting more intense, as I noted this week at New York:

There’s been a lot of arguing back and forth about the net political impact of the president’s recent lurch into explicit racism (as imposed to incessantly implicit racism, which is his usual M.O.). For what it is worth, the first week of polling after he told nonwhite members of Congress to “go back where you came from” didn’t show much in the way of movement. At FiveThirtyEight, the 11 public polls (with results adjusted for quality and partisan bias) taken since July 14’s tweets show Trump’s approval rating going up by 0.4 percent, and very much within the narrow band of popularity he’s enjoyed throughout his presidency.

One thing that has become clear, however, is that Trump – or if you think the problem is more systemic, the Trump Era – has slowly and steadily polarized public opinion along partisan lines to an amazing extent. A new NPR/Marist poll released yesterday, which was in the field from July 15-17, showed 90 percent of self-identified Republicans approving of the job the president is doing, and a precisely equal 90 percent of Democrats disapproving. These are not casual judgments, moreover: 80 percent of Democrats strongly disapprove of Trump while 72 percent of Republicans strongly approve.

Yes, self-identified independents are less adamant, but even 39 percent of them strongly disapprove of Trump, while 22 percent strongly approve. So basically, nearly all Democrats and Republicans and nearly two-thirds of independents, either love or hate the man. No wonder he (like many Democrats and observers generally) think 2020 is going to be about base mobilization and turnout. Long before the white-hot frenzy of the general election campaign melts ambivalence while driving truly nonpartisan voters into aggressive plague-on-both-your-houses disengagement, it’s already hard to be neutral about the 45th president. There are going to be many millions of Americans very upset on November 4, 2020.

It’s pretty generally recognized that four 2020 Democratic presidential candidates have separated themselves from the rest of the field. But not many observers are teasing out the implications if it continues, as I tried to do this week at New York:

[L]ess than a year before Democrats meet in Milwaukee to nominate a presidential candidate, the field’s two old white men have lost a significant amount of support, while two women (Elizabeth Warren and Kamala Harris) have gained significant strength. While everyone understands that Warren and Harris represent a threat to Biden and Sanders, it’s underappreciated how much all four candidates are now separated from the rest of the field.

Using the RealClearPolitics national polling averages as a yardstick, as recently as two months ago, Biden was at 38 percent and Sanders at 19 percent, with Warren, Harris, and Buttigieg clustered in the high single digits; O’Rourke and Booker were a few points behind them; and everyone else was struggling for oxygen. Now Biden (28 percent), Sanders (15 percent), Warren (ditto), and Harris (13 percent) are more closely bunched, with Buttigieg the next in sight at 5 percent. After a boffo second quarter in fundraising, Mayor Pete may have the resources to lift himself from the madding crowd the way Warren and Harris have done. And some other current bottom-feeder could rocket into contention with a big-time performance in the July 30 and 31 debates in Detroit (after that, the heightened requirements for participation in two rounds of autumn debates are likely to serve as an abattoir for struggling candidacies).

And as noted, there are scenarios under which each of these four candidates could put it all away early. Most obviously, despite his recent loss of support nationally, Biden remains in the lead in all four of the early states that vote in February. If he wins them all, he’d be extremely hard to stop. Sanders has one of the more consistent bases of support in the key early states, is showing signs of addressing his minority-voting weakness from 2016, and has plenty of money, so he could plausibly win Iowa and New Hampshire and roar into Super Tuesday with a head of steam. Warren has what is generally conceded to be the best Iowa organization, is rising rapidly in New Hampshire, and can compete with both Sanders and Biden in harvesting good will among a national constituency of admirers. And Harris, building on her first-debate success, has a very plausible Obama-esque strategy of wrecking Biden in South Carolina among the black voters he heavily relies upon, and then forging into front-runner status with a big win in her native California just a few days later.

But if all of these candidates fall just a bit short of their most ambitious goals, and fail to land the knockout punch when they need it, then we could be looking at a truly protracted battle in which no one rival has an indomitable advantage. That does not necessarily mean a contested convention would occur: candidates could try to achieve some breakthrough via unconventional coalitions, alliances among themselves, or early running-mate announcements (tried unsuccessfully by Republicans Ronald Reagan in 1976 and Ted Cruz in 2016, but still worth considering). Democrats do not strictly require pledged delegates to stick with their original candidates (particularly if that candidate has withdrawn), so things could change before delegates arrive in Milwaukee. And then, yes, a gridlocked convention could turn to suddenly re-enfranchised superdelegates, all 764 of them, to help make a decision after the first ballot ends.

If nobody’s made a big move by early March, it’s time to start thinking seriously about these wild convention scenarios — and about which candidate might best unify the Donkey Party. By this time next year, the fear of letting Trump win again will be so palpable that you will be able to weigh it on scales and grill it up for an anxious meal.

Since the 2020 presidential election is looking to be wild and wooly and passionate, it’s important that Democrats are clear-eyed about the implications of abandoning swing voter persuasion altogether in the pursuit of base mobilization. I wrote about that this week at New York:

Of all the reasons Trump defeated Hillary Clinton in 2016, the most compelling (to me, at least) is that he sprang the upset because an awful lot of voters who didn’t really want the mogul to become president stayed home, voted for a minor-party candidate or even cast a “protest vote” for Trump on the theory that he could not possibly win. From that sound hypothesis has grown the somewhat more dubious postulate that in 2020, if turnout is high, the 45th president is toast. Part of the reason people instinctively believe this is that there are more Democratic than Republicans out there; in other words, an energized Democratic base is going to be larger than its GOP counterpart. But the more tangible rationale is probably a lot simpler: 2018 was a very high-turnout midterm, in which Democrats did well. So if turnout stays high, the donkeys will again romp, right?

Well, that’s plausible, but hardly a lead-pipe cinch, in part because 2018’s high turnout was in fact skewed positively toward Democrats, which may or may not happen again. Historically (and particularly in the previous two midterms in 2010 and 2014), midterm turnout patterns strongly favored Republicans because the older and whiter voters who leaned GOP were since time immemorial more likely to show up for non-presidential elections. In 2018, though, that pattern was turned on its head, as Louis Jacobson recently explained at the Cook Political Report:

“In four states (Idaho, Kansas, Oklahoma, and South Dakota) Democratic gubernatorial vote totals in 2018 were more than 10 percent higher than the party’s showing in the presidential year of 2016. In six more, vote totals increased, but by less than 10 percent (Alaska, Colorado, Georgia, Minnesota, Nebraska, and New Mexico). And in 11 other states, Democratic votes for governor dropped by less than 10 percent from presidential levels (Alabama, Hawaii, Iowa, Michigan, Oregon, Pennsylvania, South Carolina, Tennessee, Texas, Wisconsin, and Wyoming) …

“In three states, gubernatorial votes for the GOP nominee increased over Trump’s total by more than 10 percent (Maryland, Massachusetts, and Vermont). In four others, GOP votes increased but by less than 10 percent (Arizona, California, Hawaii and Oregon). And in five states, GOP candidates dropped from presidential-year vote totals by less than 10 percent (Connecticut, Georgia, New Mexico, Texas, and Wisconsin).

“All told, then, Republicans saw gains or modest losses in 2018 in 12 states. However, the strongest showings came in blue states where the party had broken with the Trump-era pattern by running a moderate candidate with crossover appeal. Subtracting those states leaves about nine states in which the GOP gubernatorial nominee overperformed Trump or didn’t see a big fall-off.”

The partisan gap in extremely high midterm turnout was even more notable, Jacobson found, in Senate races. In the most similar recent midterm, moreover, that in 2010, neither party beat presidential turnout levels in more than a few states.

So if Democrats can match those patterns in 2020, victory should be relatively easy. But another way to look at it is that Republicans may have a larger pool of presidential voters who stayed home in 2018 than do Democrats, which is unusual but hardly impossible. That is essentially what Nate Cohn found in a new analysis of what a high-turnout presidential election in 2020 might look like:

“The voters who turned out in 2016, but stayed home in 2018, were relatively favorable to Mr. Trump, and they’re presumably more likely to join the electorate than those who turned out in neither election. In a high-turnout election, these Trump supporters could turn out at a higher rate than the more Democratic group of voters who didn’t vote in either election, potentially shifting the electorate toward the president.”

Perhaps these midterm stay-at-homes are now alienated from Trump, and will either stay home again or flip to the Democratic candidate. But it’s not a sure thing, particularly if you look at the Electoral College map:

“The danger for Democrats is that higher turnout would do little to help them in the Electoral College if it did not improve their position in the crucial Midwestern battlegrounds. Higher turnout could even help the president there, where an outsize number of white working-class voters who back the president stayed home in 2018, potentially creating a larger split between the national vote and the Electoral College in 2020 than in 2016.”

The arguments over the identity of marginal voters and nonvoters could lead in various directions. But the one inescapable conclusion is that driving turnout upward by making the 2020 election a frenzied test of comparative “enthusiasm” is a dangerous game for Democrats, if it’s not supplemented by (1) a least some “swing voter” strategy, and (2) quieter turnout measures (e.g., voter registration drives, data-driven voter-targeting messages, or even traditional knock-and-drag GOTV efforts) that don’t run the risk of riling up both sides equally.

It’s certainly safe to say that Trump and his party are risking disaster by making every presidential utterance an outrage that at best excites his base as much as it infuriates Democrats. And since time immemorial, ideologues on both the left and the right have asserted as a matter of quasi-religious faith that some hidden majority favor their prescriptions, with tens of millions of citizens refusing to vote because the radicalism they crave has been withheld by Establishment centrists. In 2020, though, the stakes are higher than ever if Democrats are wrong about relying on the intensity of voter enthusiasm. It would be smart for them to have a backup plan.