As the New Year began, I mulled a perennial topic of Democratic strategy in a piece for New York:

As the voting phase of the 2020 Democratic presidential primary gets underway next month, fears about conflicts between candidates (and their supporters) weakening the party will begin to multiply. So aside from perpetual (and, in some respects, futile) efforts to assess each candidate’s “electability,” it makes sense to begin examining the most viable candidates in terms of their ability to unite Democratic voters for the mega-intense drive to November. Here are some strategies to consider.

Determine Which Groups Need to Come Together

The first step for any candidate hoping to unify the party is determining which factions are most essential for a 2020 win. Is it average voters? Certain demographic groups? Some combination of activists, opinion leaders, donors, and elected officials?

At this point, most Democratic primary voters don’t seem terribly inclined toward internecine conflict. An Economist-YouGov survey taken at the end of 2019 asks registered voters who say they plan to participate in Democratic primaries or caucuses: “Are there any presidential candidates that you would be disappointed if they became the Democratic nominee?” Multiple answers are allowed. Ex-Republican Mike Bloomberg attracted the most “disappointment,” at 34 percent, with Joe Biden coming in at 23 percent, Bernie Sanders at 20 percent, Andrew Yang at 18 percent, Pete Buttigieg and Amy Klobuchar at 17 percent, and Elizabeth Warren at 14 percent. Some findings don’t exactly match what you’d expect. Buttigieg is thought by some to have fatally alienated African-American voters. Yet in this survey, only 14 percent of black voters say they’d be disappointed if Mayor Pete is the nominee, less than the 17 percent who said the same of Sanders, whose African-American support levels (mostly among young black voters) are notably stronger this year than in 2016.

Among party elites, you’d get a different picture. Big donors are famously hostile to both Sanders and, perhaps even more particularly, to the Wall Street–focused Warren, despite her generally high ratings among rank-and-file Democrats. Elected officials are more prone to support Biden, and to fear Sanders, so far. The “Twitter Left” of typically young, typically ideological activists is hostile to Biden, as you might expect, but also arguably even more hostile to Buttigieg, whose combination of youth and centrism seems to annoy millennial progressives in a particularly intense way.

In a piece presenting the argument for Sanders’s nomination, Matthew Yglesias suggests that lefty elites are potentially more troublesome in a general election than their less ideological counterparts, and thus need placating:

“Lots of moderate Democrats … find it annoying that Sanders and some of his followers are so committed to painting mainstream Democrats in such dark hues. And it is annoying! But annoying people won’t stop being annoying if he loses the nomination. If anything, they will be more annoying than ever as some refuse to get enthusiastic about the prospect of beating Trump. But if Sanders wins, partisan Democrats who just want to beat Trump will magically stop finding Bernie superfans annoying — the causes will be aligned, and the vast majority of people who want Trump out of the White House can collaborate in peace.”

If you are focused on the small factors that might make a difference in Democratic esprit de corps in a general election, eliminating the fights between Bernie-or-Bust folk and regular Democrats on social media that persisted in 2016 would make sense. But there is also the possibility that a Sanders (or Warren) nomination could give impetus to some sort of “centrist” third-party campaign like the one Howard Schultz threatened to wage earlier in the cycle, which for all its lack of natural voter appeal could serve as a lightning rod for Democratic-donor unhappiness. In other words, the target of each candidate’s unity effort will depend on identifying which voters are most put off by them.

Find Candidates With the Most “Unity” Potential

There are at least two traditional ways of assessing a candidate’s capacity to unify a potentially divided party that we can apply to the 2020 field: ideological positioning and second-choice support.

On the former point, being perceived as at the center of the party with respect to ideology can be a real asset. A new CBS-YouGov poll of Iowa asks if the various candidates are “too progressive,” “not progressive enough,” or “about right.” The rankings for “about right” are as follows: Buttigieg (65 percent); Biden (63 percent); Klobuchar (54 percent); Warren (44 percent); and Sanders (42 percent). Two candidates are far above average in the “too progressive” column: Sanders (50 percent) and Warren (41 percent). Two others stand out as “not progressive enough”: Klobuchar (35 percent) and Biden (31 percent).

But there are other respects in which candidates may or may not have the kind of broad appeal you’d want heading into a general election. Per the Morning Consult tracking poll, for example, Sanders is vastly more popular among younger voters and struggles toward the higher end of the age spectrum; to a less extreme extent, that is true of Yang as well. Biden and Buttigieg have the opposite problem. Warren displays fairly even levels of support by age. Biden, Sanders, Warren, and Yang have support spread reasonably well among white as well as nonwhite voters, while Buttigieg famously struggles with nonwhite voters.

Morning Consult regularly asks its large tracking-poll sample about second-choice candidates. In the latest poll, Sanders is the top second choice of Biden supporters (at 29 percent), followed by Warren (19 percent), Buttigieg (12 percent), and Bloomberg (11 percent). For Sanders supporters, the second choice is Warren (at 32 percent), followed by Biden (28 percent). Among Warren voters, the second preference is Sanders (33 percent) and Biden (24 percent). The runners-up for Buttigieg backers are Biden (27 percent) and Warren (20 percent).

More than anything else, these findings suggest that Biden and Sanders are both stronger unity figures than you might think. But candidates like Buttigieg, who are perceived as ideologically near the party’s center, have some strengths to build on as well.

Endorsements and Running Mates Can Help

As the nomination contest winds down, candidates will have other opportunities to broaden their appeal. First, they can seek endorsements from defeated candidates. Julián Castro’s endorsement of Warren shortly after he withdrew from the race may not bring her that much tangible help, given his own poor standing in the polls, but if others join him, that could burnish her reputation as a unity candidate. Conversely, in 2016, Sanders’s reluctance to endorse Hillary Clinton as soon as he might have (he waited until mid-July) contributed to a persistent sense of a divided party, even though Sanders energetically campaigned for the nominee later on.

Once a candidate has secured the nomination, they can reach out to disappointed party factions of identity groups via their choice of running mate. It is generally expected that any male nominee in 2020 will seek a woman as a running mate, and vice versa; the days of two white men forming a ticket are probably over for the time being among Democrats. But trans-ideological and multigenerational ticket-making could be a factor as well, if the nomination contest is close and tensions have developed.

The late-septuagenarians Biden and Sanders could have a particularly rich opportunity to promote unity in their veep selection, as either might be limited to a single term. Running with a younger, perhaps ideologically distinctive veep could essentially signal to Democrats jittery about them that their nomination wouldn’t necessarily represent some sort of definitive settlement on the party’s direction.

One unity tool that has never been tried by a Democrat (though Ronald Reagan briefly and unsuccessfully attempted it in 1976) is the choice of a coalition-building running mate before the nomination is secured. Hints were dropped by Biden’s campaign that he might try that gambit in 2020, with former Georgia gubernatorial nominee Stacey Abrams being the putative veep whose name was most mentioned. Abrams ruled it out for the time being, but not for good, and Biden (or some other candidate) could come back around to the idea between now and the convention.

Surviving Contested Conventions and Other Late Threats to Solidarity

The need for party unity could skyrocket if certain events threaten to further divide Democrats. A contested convention is unlikely but possible, and would certainly make unity gestures very important in restoring normalcy before the sprint to November began. Or if the primaries produce a nominee (Sanders? Biden? Buttigieg?) who provokes threats of defections to a third-party candidate or to the living-room couch on Election Day, then deploying opinion leaders with credibility among disaffected voters would be important. In general, the closer the ultimate contest looks, the more essential it will be to convey the impression that all hands must be on deck.

There Could Be Several “Unity Candidates”

A unified Democratic Party is within the grasp of any nominee, mostly thanks to the Great Unifier in the Oval Office, whose sins make those of any competitor pale by comparison, and whose second term could represent a veritable hellscape for the party and the country. But the need for unity should cross the minds of eager Democrats every time they speak to a reporter or post on Twitter about the unacceptable characteristics of candidates other than their own. It’s not just that the infernal forces of MAGA will repeat, amplify, and exaggerate any such negative talk no matter how hypocritical it may be to do so; it’s also that it may not take much to convince low-information voters that “they’re all just lying politicians” — so why not stay home, or vote for the devil that you know?

Trump’s sudden announcement that the U.S. has “taken out’ a revered Iranian military commander led me to this observation for New York:

The president who has endlessly and redundantly attacked open-ended military commitments and expensive and extended conventional warfare has taken an action that many experts think will push Iran to launch a full-on regional war, perhaps even involving interested bystanders like Israel and Saudi Arabia. It could be the mother of all quagmires. So how did this happen and what will Trump do if the sudden strike he authorized sparks significant retaliatory measures from Iran?

We may not know for years, if ever, exactly what transpired in the White House, but the two relevant things to remember about the commander-in-chief are (a) he is by nature a bully, whose creed has always been winning by intimidation with superior force (or in business, via ruinous lawsuits) and (b) to the extent he thinks about war and peace he’s a total throwback to Andrew Jackson and represents Old Hickory’s peculiar combination of non-interventionism and violent militarism. I explained this latter characteristic when he launched an attack on Syria back in 2017:

“Trump [is] a self-consciously ‘Jacksonian’ president who simultaneously reveres military force while despising the ‘globalist’ ideologies that have both justified and restrained its use so often in U.S. history. ‘Jacksonians’ typically oppose entangling alliances and international nation-building exercises, but not only accept but welcome massive violence when America is ‘crossed.’ For Trump in particular, intimidation of enemies is as important to international affairs as it is to business life. That is why Trump constantly attacked Barack Obama for failing to back up his ‘red line’ threats against Syria’s use of chemical weapons (an attack he repeated before launching the cruise missiles last night) even as Trump himself denied any interest in ‘taking sides’ in that country’s messy civil war …

“Trump’s decision to act without consulting, much less asking authorization from, Congress makes perfect sense. For one thing, the ‘strategy’ or ‘comprehensive plan’ that so many senators asked for in their own initial reactions to the attacks on Syria last night may very well not exist. If the missiles were simply intended to put things right after Obama’s ‘cowardice’ and send a message to the world, then there is not and will never be a ‘strategy’ or ‘plan’ with respect to Syria; the violence was an end in itself.”

It’s very likely that is the case today, too. And projecting his own taste for high-stakes gamesmanship onto the Iranians, he may well believe one show of righteous and lethal American power will make them back down.

You don’t reason with these people, he seems to believe: You just have to show ‘em Uncle Sam is the boss.

But what if Trump has miscalculated by assassinating a figure so central to Iran’s national and religious sensibilities that their own sense of honor demands they do not back down? Or what if they believe they can bully the bully thanks to his well-known distaste for getting tied down into an extended military conflict?

Clearly Trump hopes this is a one-off game that he’s already won. In his brief remarks today from Mar-a-Lago, he treated the assassination of Soleimani as a preventive act that interrupted “sinister attacks” on U.S. personnel. And he pointedly said: “We took action last night to stop a war. We did not take action to start a war.” But it might not be entirely up to him.

It’s at this juncture that things could get really scary. Even if Trump somehow erroneously thinks a conventional war with Iran would ensure his reelection, he is unlikely to want to burden his second term with the kind of endless no-win commitment that sank the despised George W. Bush, in Trump’s own contemptuous estimation. So in the Jacksonian tradition, the most appropriate response to any serious Iranian escalation toward all-out war would be a counter-escalation of great ferocity, which would deliberately ignore any sense of proportion or calibration and display America’s might at its fullest. And if that is Trump’s inclination, it’s unclear whether any of the people around him — or his Republican allies in Congress — who have mostly been frothing for war with Iran for years, would restrain him. It’s certainly doubtful that two of Trump’s favorite allies this side of Russia, Saudi Arabia and Israel, would object to the U.S. getting medieval on their ancient enemy.

Given Trump’s combined taste for violence and distaste for messy conventional warfare, and his limited concern for the lives of non-Americans, you have to assume the worst is possible. Remember this comment (per Alex Ward) he made last summer about Afghanistan?

“Ahead of a Monday meeting with Pakistani Prime Minister Imran Khan, Trump told reporters at the White House that he could win the war in Afghanistan in just one week if he really, really wanted to. But Trump says he won’t do that because he doesn’t want millions to die.

“’I don’t want to kill 10 million people,’ he said. ‘I have plans on Afghanistan that if I wanted to win that war, Afghanistan would be wiped off the face of the earth, it would be gone, it would be over in literally 10 days.’”

If Trump has plans to nuke Afghanistan, he most definitely has plans to nuke Iran, particularly given global worries about Tehran’s own nuclear program. I’m not predicting he’d do anything quite that insane, but the Jacksonian logic of sudden and terrifying force as a first and last resort means nothing lethal is going to be off the table. Lord have mercy on us all.

Amidst all the retrospectives of the last decade, I had to note at New York the upcoming decennial anniversary of a bad moment for Democrats:

I’ve only seen one take on the 2010s, from USA Today’s Jill Lawrence, that gives proper weight to the shocking event that showed in the world of politics, the “teens” would not reflect a continuation of the strong Democratic trends of 2006 (which made Nancy Pelosi Speaker of the House) and 2008 (when Obama won the presidency in a near-landslide).

That would be Republican Scott Brown’s January 2010 conquest of the Massachusetts Senate seat held since 1962 by the “liberal lion” of that chamber, Ted Kennedy. At the time there were some Democratic hopes that the special election was an aberration based on irregular turnout and a feckless campaign by Democrat Martha Coakley. But as Lawrence pointed out, it was the ultimate omen:

“To call Brown’s win a trauma for Democrats would be an understatement. And yet, although it was a DefCon 1 warning about the midterm to come, they went into those 2010 November elections unprepared for the debacle. Wipeouts in the state legislatures that would draw up new and in some cases egregiously gerrymandered election maps based on the 2010 Census. Wipeouts in races for the governors who would have been able to temper or veto those maps. Wipeouts in the House, installing a GOP majority hellbent on thwarting Obama.

“The trend continued through 2014, from state legislatures, governors and the House to a 2014 Republican Senate takeover that offered a glide path for conservative judges and justices when the next jolt arrived two years later in the form of Trump.”

Yes, Obama was reelected in between those two strong Republican years, though by a significantly reduced margin (dropping from 7.2 percent to 3.9 percent in the popular vote). And of course, Brown lost his Senate seat to Elizabeth Warren in 2012. At the time, many observers (myself included) deduced that the demographic disparity between midterm and presidential electorates explained a lot of the apparent oscillation of results, which augured well for Democratic prospects in 2016. And then you-know-what happened.

Now, three weeks before the tenth anniversary of the Brown shocker, there is no easily discernible pattern in American party politics going forward. Democrats did better in 2018 than Republicans did in 2014, by any standard other than net Senate seat gained (Republicans picked up nine net seats in 2014 and actually gained two more in 2018 thanks to a heavily skewed landscape). But that wasn’t unusual for a president’s first midterm, particularly a president as unpopular as Donald Trump. The 2020 election is widely expected to be a barn-burner, and while Democrats should continue to benefit gradually but steadily from demographic changes in their favor, Republicans have proven quite good at maximizing their power via a combination of voter suppression, gerrymandering, the unrepresentative nature of the Senate and the Electoral College, and ruthless demagoguery. If Trump hangs onto the White House next year and his party hangs onto the Senate, the GOP could establish a hold on the federal judiciary lasting for decades, while continuing to punch above their popular weight in other arenas thanks to the structural advantages they maintain.

Brown’s victory and the ensuing struggle to enact Obamacare (and other, less successful, elements of the president’s agenda) without a Democratic Senate supermajority also offered a tutorial on the obstruction a disciplined Republican majority could mount, and the price Democrats would have to pay in policy compromises to govern even with a strong position in Congress (which they were soon to lose).

Without question, Democrats will be ebullient if they manage to defenestrate Trump, particularly if that win is accompanied by the first Democratic trifecta (control of the White House and both Houses of Congress) since 2008. But as 2010 quickly showed, political fortune can change almost instantly, and the work of building a governing majority never ends.

With all the justifiable focus on the presidential contest, it’s easy to forget how important the battles for control of Congress in 2020 will be. I wrote about the Senate landscape this week for New York:

The general feeling going into this election cycle has been that flipping the Senate will be difficult for Democrats, even if they depose Trump. They have a three-seat majority, and in Alabama’s incongruous Democrat Doug Jones, a pretty easy mark. Only two Republican incumbents (Cory Gardner of Colorado and Susan Collins of Maine) up for reelection represent states carried by Hillary Clinton in 2016, and aside from Jones, Michigan Democrat Gary Peters is in a Trump ’16 state.

But as the Cook Political Report’s veteran Senate guru Jennifer Duffy notes, Democrats are off to a very good start, particularly on the fundraising front, in their effort to place as many GOP Senate seats into play as possible:

“If there is anything that sticks out in Senate races this cycle, it’s the early spending on television advertising in the most competitive races. As of December 19, just over $32.5 million has been spent in eight key races. Democrats have outspent Republicans, $21.9 million to $10.6 million, according to data provided by Advertising Analytics. The Senate race in Maine has seen the highest level of spending at $8.2 million. To put this in some perspective, Collins spent $5.6 million on her 2014 reelection bid, and independent expenditures amounted to less than $2 million. Advertising Analytics estimates that $55 million will be spent on television advertising in Maine this cycle, an astonishing amount for a state with three relatively inexpensive media markets. Democrats have outspent Republicans almost two to one and nearly all that money has been on ads criticizing Collins.

“Democrats have also outspent Republicans in Arizona, Colorado, Iowa, and Kentucky. Republicans have spent considerably more than Democrats in Alabama and North Carolina, but most of the money is being spent in GOP primaries.”

Duffy notes that a Senate impeachment trial could create some general-election problems for GOP incumbents Collins, Gardner, and Martha McSally — who dare not offend the Trumpian base with a vote to remove him from office. But there are signs Democrats also have a shot against North Carolina’s Thom Tillis and Iowa’s Joni Ernst — and perhaps the occupants (David Perdue and the top Republican in the 2020 special “jungle primary”) of the two Georgia Senate seats at risk. Even Kansas is a possible pickup so long as Kris Kobach is in the race.

“It appears that there will be at least five GOP-held seats in play, with a chance that Democrats could add one or two more. This puts Democrats in a position to win the majority, even if they lose Alabama and/or Michigan. This is not to suggest that Democrats will win the majority, only that their prospects are considerably better today than they were five months ago.”

If Mike Pence is reelected as vice president, of course, Democrats would need to post a net gain of four Senate seats to be in a position to block — Joe Manchin willing — Trump executive- and judicial-branch nominees, not to mention whatever nasty legislative treats he has in mind. Given the strong likelihood that we’ll see remarkable levels of straight-ticket voting in 2020, prospects for a Democratic Congress confronting a raging, reelected Trump are limited. More likely, Senate Democrats will be focused on giving a president of their own, if they get one, a fighting chance for success.

On the occasion of Donald Trump’s impeachment by the U.S. House of Representatives, I offered a quick history lesson at New York to counter all the angst about the partisanship of the process:

[Y]ou may have already heard handwringing comments about the exceptional partisanship that House members exhibited in the vote to impeach Donald Trump. It wouldn’t have happened in the days when members of Congress socialized with each other and worked on legislation in a spirit of comity, etc., etc.

Perhaps there is some merit in that much-rehearsed paean to lost bipartisanship, so full of manufactured nostalgia that it ought to be made into a Hallmark Channel movie. But actually, the two previous examples of House votes on presidential impeachment were arguably just as partisan, differing only in degree.

The principal vote to impeach Andrew Johnson (in those days the House voted on a general resolution of impeachment before drafting individual articles), on February 24, 1868, was carried by a 126-47 margin, with 17 members not voting. Of those who did vote, all but two Republicans voted “aye” and all the Democrats voted “nay,” according to the official House history. The later votes on the 11 individual articles eventually sent over to the Senate are hard to find online, but the House history indicates that the margins were “similar.”

When the House voted on the impeachment of Bill Clinton on December 19, 1998 (Thursday is the 21st anniversary of that event), the two articles on which the president was later tried by the Senate passed, per the New York Times, on near-party-line votes. The first, impeaching him for perjury before a federal grand jury, passed by a 228-206 margin in the Republican-controlled chamber, with five Republicans and five Democrats defecting. The second, alleging obstruction of justice, passed by a narrower 221-212 vote. This time 12 Republicans and five Democrats broke ranks.

Yes, there was theoretically more bipartisanship in votes on two articles of impeachment the House rejected, mostly because some Republicans considered them redundant or too easily mockable. An article alleging perjury in the Paula Jones civil case lost 229-205, with 28 Republicans defecting and no Democrats breaking ranks. And a final article alleging “abuse of power” because Clinton apparently offended the dignity of Judiciary Committee Republicans by answering their written interrogatories evasively, didn’t resonate much outside that committee; it was defeated by a wide margin with 81 Republicans (and one Democrat) breaking ranks.

There were, of course, many differences between the three impeachments. Andrew Johnson was pretty much a lame duck, rejected by both major parties as a reelection prospect (though he did get some votes at the later Democratic Convention; he was later elected to the Senate as a Democrat, rejoining his pre-Civil War party), by the time he was impeached. Clinton was a popular second-term president. Johnson was very nearly convicted and removed from office by the Senate; there was never any real chance that would happen to Clinton.

But the most conspicuous thing distinguishing Trump’s impeachment from those of his predecessors has been the full-throated defense of his conduct by his own party. By the time he was impeached, Johnson was virtually a man without a party; southern Democrats, while voting against impeachment and removal, couldn’t be that thrilled about the man who demonized them during the Civil War as traitors who deserved death in battle or by hanging. And he was acquitted mostly because of the concerns some Republicans had that the Tenure of Office Act (banning removal of Cabinet members without congressional concurrence), the backbone of the impeachment articles after Johnson defied it, was unconstitutional (as it was eventually held to be by the Supreme Court). As for Clinton, many, many Democrats condemned his behavior during the Lewinsky scandal, and there was robust support in their ranks in Congress for a measuring censuring him.

So you could say that this third presidential impeachment was more partisan than the first two in that the two parties were not simply arguing over what should be done about a president’s misconduct, but whether it existed at all. And I suspect that his party will be found guilty for that misdemeanor at the bar of history, if not sooner in the 2020 elections.

When U.S. Rep. Jeff Van Drew decided to switch parties after Democrats in his district did not take kindly to his refusal to back an impeachment inquiry, I noticed something about his record that will give his new friends heartburn, and wrote about it for New York:

New Jersey congressman Jeff Van Drew’s impending Democrat-to-Republican Party switch has gotten a lot of attention as another sign of partisan polarization over impeachment (his was one of just two Democratic votes against the formal launch of impeachment proceedings in October, and he had signaled his opposition to the articles of impeachment that will hit the House floor this week). This was punctuated over the weekend by a mass resignation of most of his staff.

The move was probably made inevitable when Democrats back home began to abandon Van Drew much as his staff did later. But Chris Cillizza thinks it’s a great trophy for Trump:

“The image of Van Drew, then, being driven from his party because of impeachment plays directly into Trump’s hands. The President has long argued that Democrats are blinded by their hatred for him and that this latest congressional reaction to his behavior with Ukrainian President Volodymyr Zelensky is nothing more than a partisan hoax.”

I don’t know about that; I’d say having actual Democrats instead of former Democrats opposing impeachment is more valuable to his cause and to the “Democrats in disarray” narrative political media love so much. But there’s an aspect of Van Drew’s apostasy that could cause Trump and the GOP some heartburn, too. After the 2018 elections (when Rodney Frelinghuysen and Charlie Dent retired), the anti-abortion movement could boast it had finally hunted to extinction the once-vibrant herd of pro-choice House Republicans. Whatever else he is now, Van Drew is pro-choice, having received a 100 percent rating from Planned Parenthood in the New Jersey legislature, and then saying this during his 2018 congressional campaign:

I’m strongly and unequivocally pro-choice: I support Roe v. Wade and a woman’s right to choose. Any efforts to weaken or undermine that right will face my fierce opposition.

It is most definitely going to annoy Trump’s extremely important conservative Evangelical and traditionalist Catholic allies that a friend of the baby-killers has snuck back into the tent. And they may not be happy that POTUS has publicly offered Van Drew fundraising help in his 2020 reelection bid.  want to know whether the president has promised Van Drew support in his 2020 reelection bid.

Before the party switch, Van Drew had three Republicans already in the field running against him, one of whom is a self-funding business executive who called the incumbent a “weasel.” I’d be shocked if his primary opponents don’t make a big deal out of his position on abortion, which in turn could cause problems for Trump if he defends the apostate, when he really needs anti-abortion activists whipping the faithful into a frenzy before November.

If, like me, you’ve been watching as much of the House Judiciary Committee deliberations on impeachment as you can stand, you have endlessly and interminably and redundantly heard Republicans argue that Democrats are engaged in a “coup” to overturn Trump’s election. This is how I responded at New York:

As the Kabuki Theater of impeachment rolls along in the U.S. House, there are legitimate fears about how it will all play out for Democrats, as I explained this week at New York:

 It’s important to internalize a couple of basic facts about the big picture with respect to the impeachment process: (1) Trump is not going to be removed from office by the Senate, and (2) whatever impact impeachment has on future presidents or on historical judgments of this Congress — both factors often cited by Democrats favoring impeachment — we won’t know it for a good while. So the only thing relevant to analyze is the effect that this process and its trajectory may have on the 2020 elections.

We just don’t know at this point how the voting public will adjudge the impeachment inquiry, the House impeachment, or the Senate acquittal. Public support for impeachment spiked a bit just before and just after the formal process was initiated, but it has stabilized amid evidence that it’s not very popular in the Rust Belt battleground states where Trump pulled his upset in 2016. Attitudes toward impeachment, not surprisingly, are beginning to pretty closely match attitudes toward Trump and his reelection bid. So the best early evidence is that the impact of impeachment may be on the margins of the election, where enthusiasm and turnout patterns are legitimately important and not just the subject of spin.

That should concern Democrats. The decision by Nancy Pelosi to quickly enact narrow articles of impeachment before the 2020 election year formally begins may well indicate that she views it as a distraction at best and as a potential problem for Democrats at worst. These attitudes should be unsurprising given her steady resistance to going down that road until the Ukraine scandal broke and the Democratic rank and file shifted massively into the pro-impeachment camp. As I noted when the articles were announced, pro-impeachment Democrats — including those who favored taking this step before the Ukraine scandal appeared and Pelosi climbed aboard — may soon feel cheated out of a deep, broad investigation that would accompany the 2020 campaign:

“It’s certainly not hard to suspect that Pelosi is really just cutting losses by focusing on one incident of Trump’s misgovernment and racing to impeach him by year’s end so that House Democrats can move on to their previously scheduled election-year agenda. And the impression that she’s ready to ‘move on’ is certainly reinforced by the fact that she is holding another presser today to announce support for the administration’s renegotiated U.S.-Mexico-Canada trade agreement.”

At the same time, though, impeachment has gone far enough to unite Republicans behind Trump and give them the relentless talking point they have already agreed upon and will repeat like a meditative chant at any moment Republican partisans need a fresh burst of energy: Impeachment is a coup designed to rob American voters of their right to elect — or reelect — the president of their choice. A variation on this theme that we will briefly hear before the Senate trial concludes is that panic-stricken Democrats know the only way they can keep Trump from a second term is to secure his removal from office. But he and the GOP will thwart these traitors, and the preordained acquittal will be celebrated as total exoneration wherever MAGA folk gather.

So in one party, you will have excited, triumphant fans of the president, who once again eluded and outsmarted his elitist enemies, snake-dancing to the polls to secure another four years for their hero. In the other, you will have some people who want to forget about impeachment altogether and talk about health care whenever the party’s presidential candidates aren’t bickering about it, plus some people who are in a state of simmering resentment that their congressional leaders just went through the motions and didn’t expose Trump’s broader crimes and misdemeanors.

This isn’t an equation that works out very well for Democrats in 2020. Yes, of course, they can still beat Trump, and unless his job-approval rating finally rises, they probably will if their presidential nominee is decent and competent and acceptable to all party factions. But it may well be that Pelosi’s effort to thread the needle on impeachment will instead show her to have fallen between two stools, disappointing Democrats yet giving Republicans the hate-rage jet fuel on which they thrive.

I don’t routinely post items here on the demise of presidential candidacies, but Kamala Harris’ involved strategic issues to an unusual degree, so I’m sharing what I wrote up for New York.

When Steve Bullock and Joe Sestak withdrew from the 2020 presidential race at the beginning of this week, it represented the inevitable, arguably overdue winnowing of a might-have-been and a never-was contender. Kamala Harris’ surprise withdrawal today was more significant, representing the demise of a candidacy that made a lot of strategic sense and that for a brief moment last summer looked very formidable.

It’s unclear at this early juncture whether the withdrawal was the product of the “disarray in the campaign” that has been written about abundantly in recent weeks, most pungently at Politico:

“Kamala Harris’ campaign is careening toward a crackup.

“As the California senator crisscrosses the country trying to revive her sputtering presidential bid, aides at her fast-shrinking headquarters are deep into the finger-pointing stages …

“[One] person described the current state of the campaign in blunt terms: ‘No discipline. No plan. No strategy.’”

Most accounts of Harris campaign troubles focused on overspending and on confusion at the top of the organization where campaign manager Juan Rodriguez and campaign chair (and the candidate’s sister) Maya Harris were twin authorities.

But beneath the day-to-day problems was a campaign whose plausible strategic objectives simply weren’t being met. The original Harris plan was modeled to a considerable extent on Barack Obama’s in 2008, as I observed last fall:

“[I]t’s the strategy successfully pursued by another freshman senator with a multiracial background in 2008: establish your political chops by winning in nearly-all-white Iowa and then consolidate minority support in the South and in urban states with large African-American populations. Indeed, Harris has an advantage that Barack Obama did not enjoy: her own home state of California has moved its primary up until March 3, just after the initial quartet of events in Iowa, New Hampshire, Nevada, and South Carolina.

“In the sports language often used (along with combat and gambling lingo) by political operatives, one of Harris’s people called this strategy: ‘the SEC primary meets the West Coast offense.’ And it makes sense, on paper, particularly if Harris can go into South Carolina with a head of steam and win there.”

Aside from the challenge of trying to get traction in a crowded field in Iowa, Harris had to do to Joe Biden what Obama did to Hillary Clinton in 2008: shake loose a strong attachment to a white front-runner among African-American voters, particularly in South Carolina. For a moment, after she seized the spotlight in the June candidate debate with a strong criticism of Joe Biden’s understanding of racial issues, it looked like she was well on her way to doing just that, as I noted at the time:

“Totally aside from the substantive impact of Harris’s challenge to Biden’s record on school busing and racial justice generally, it’s not good for the former veep that two of the strongest performers in the first round of debates have been Harris and Cory Booker, who represent a generational and a racial contrast to him. They are both gunning for Biden in South Carolina, and if one or both begins to carve into his African-American support, he’s in serious trouble.”

And in part because she was the clear star of this debate, it was Harris, not Booker, who caught fire. A Quinnipiac poll immediately afterwards showed her leaping into second place, just two points behind Biden, and well ahead of Bernie Sanders and Elizabeth Warren. Perhaps even more importantly, a CNN survey showed Harris cutting Biden’s advantage over her among nonwhite voters from 26 percent to six percent.

But even before her underwhelming performance in the second round of debates in July, there were signs Harris’ boom was subsiding. By August her national polling numbers were back down into the single digits, and it became obvious she was significantly trailing Obama’s trajectory at the same stage of the 2008 campaign, particularly in terms of African-American support and positioning in Iowa. The rise of Elizabeth Warren during this same period took a lot of the spotlight away from the Californian as well.

By September Harris recognized that without a better showing in Iowa she was unlikely to enjoy a South Carolina breakthrough. So she emulated another successful campaign of the past, John Kerry’s in 2004, in putting all her resources into the first-in-the-nation-caucus state, famously telling Senate colleagues she was “f___ing moving to Iowa.” But she never got traction there. In a September Iowa Poll from Ann Selzer, she was at six percent, and then in Selzer’s November poll, she had dropped to three percent. And that’s about the time when the “disarray” that got so much attention became impossible to ignore.

Contributing to Harris’ strategic failures were some messaging missteps. In the crucial July debate when she began to lose steam, she got mired in a confusing explanation of her differences with Bernie Sanders and Elizabeth Warren on Medicare For All, and walked right into a savage attack on her criminal justice reform record by Tulsi Gabbard. And while she issued well-regarded proposals on teachers’ pay and tax reforms, she never really achieved a signature policy proposal, which enhanced the impression that she was mostly focused on her positioning in the field rather than making a compelling case for her nomination or her ability to beat Trump. And her debate stumbles quickly diminished thoughts of this ex-prosecutor dismantling Trump in a general election tilt.

Perhaps the coup de grace in terms of Harris’ trajectory in the race occurred earlier this week when two national polls showed her even or actually behind Michael Bloomberg, who very recently entered the race. She was continuing to go nowhere fast, and there was even talk in California that if she didn’t get out of the race and mend fences back home, she might court a 2022 Senate primary opponent.

It’s unclear exactly how Harris’ withdrawal will affect the race. The potential beneficiary who needs help the most is the one remaining African-American candidate, Cory Booker, who has been laboring in Harris’ shadow in both Iowa and South Carolina, the two key states for him as well. And it creates a real battle royal in California, where Harris retained some significant support despite slipping behind the leading national candidates there. Arguably Joe Biden, who was for a while vulnerable to Harris’ strategy, will be relieved to see the back of her. And to the extent that Harris at her best was a potential unity candidate for the party, her absence could create a fresh competition for that mantle.

All in all, Harris had a lot of potential but failed to capitalize on it, which has led some observers to compare her to 2016 Republican presidential candidate Marco Rubio.

At 55, Harris isn’t quite as young as the 48-year-old Rubio, but like him, she is young enough to contemplate a national political comeback–perhaps even by becoming someone’s running-mate next year.

Given how close the 2020 election is likely to be, Democrats need to pay attention to some relatively small swing-voter groups, and I wrote about one of them at New York:

[A] lot of self-identified political independents are quite proud of themselves for, well, their “independence.” A majority of them, as many political scientists have explained, are functionally partisan in their voting habits, but since they are theoretically open to going the other way, they aren’t like those knee-jerk D’s and R’s, or so they imagine. It is true that some indies have a mix of positions on issues that don’t nicely comport with either major party’s — or any minor party’s — views (notably economic liberals who are also social conservatives; those with the opposite configuration can always vote Libertarian). It’s a free country, and it’s fine with me if they want to let their freak flag fly.

There is one species of nonpartisan, however, who might be considered different from others and even pernicious: those who oscillate from party to party not based on issue adherence, or even the attractiveness or repulsiveness of individual candidates, but because they want to keep all parties and all factions in some sort of equipoise where they don’t get to have their way. These “checks and balances” voters are often very proud of themselves for the civic virtue they display in limiting the power of the overwhelming majority of citizens who are partisan. And when the two major parties are equally strong, they can even determine outcomes, as Nate Cohn and Claire Cain Miller explain in examining some Siena College polling data from battleground states concerning voters who supported Trump in 2016 and 2018 but voted Democratic in the 2018 midterms:

“Many of the voters who said they voted Democratic but now intended to vote for Mr. Trump offered explanations that reflect longstanding theories about why the party out of power tends to excel in midterms.

“Michelle Bassaro, 61, is a Trump supporter, but in the midterm election, she voted for the Democrat in her district to balance the administration’s power. She said she had voted for Republicans when Democrats were in the White House for the same reason, consistent with research that shows that some people intentionally vote for divided government.”

The research these writers referred to is a bit dated but still relevant; one study showed that an estimated 16 percent of voters, as of 2008, preferred divided government. A lot of them don’t act on this sentiment — or don’t really have the practical option to do so in a particular election — but there are enough to be dangerous in a close contest. And dangerous they are, I believe.

Perpetually divided government (which we have had more often than not at the federal level in the post–World War II era) is an invitation to gridlock, dysfunction, and citizen dissatisfaction. It’s even more damaging now that the ideological polarization of the two major parties has made bipartisan coalitions vastly less likely than in the days when liberal Republicans and conservative Democrats walked the Earth. Yet some of the same voters who consider themselves shrewd and civic-minded for keeping the two parties in balance tend to complain about stuff not getting done:

“Danny Destival, 56, who runs a greenhouse supply business in Panama City, Fla., said he’s ‘been a Southern Democrat all my life.’ But in 2016, he cast his first Republican vote because he liked that Mr. Trump was a businessman, not a politician — and he disliked Hillary Clinton.

“His main priority is voting for ‘the person who’s going to get more done’ — that’s why he stuck with the Democrats in the midterms — but at the national level, he said, the Democrats have disappointed him on that front.

“’If you’re going to Washington, you need to do something,’ he said. ‘If the only thing you’re going to do the whole time you’re there is try to get rid of the president, that’s a problem. I mean, Trump is not a great person, but you’ve got to get some work done.'”

This “swing voter” does not seem to be aware that he is part of the problem he is complaining about. And his preferred candidate for president is a symptom of how haywire things can go if the normal processes of policymaking and legislation are frustrated by divided government and the consequent gridlock. You get voters throwing up their hands and then supporting a demagogue who claims he will “drain the swamp,” only to run one of the most corrupt administrations in history with legislative accomplishments — even when his party did have unified control of the government — that would fit in a thimble. Trump is also emblematic of the recklessness a president can exhibit when thinking of himself as empowered to overpower and dominate other institutions or the rule of law itself.

Complain as we might about the folly of using one’s vote to alternate perpetually between the parties, it’s enough of a reality to sober any Democrats who believe their midterm victory in 2018 gives them an automatic upper hand in 2020. You might imagine logically that the phenomenon of a party winning a second consecutive presidential election while losing the intervening midterm would be relatively rare. It has actually happened (just going back to the end of World War II) in 1956, 1972, 1984, 1988, 1996, and 2012. Some of it has to do with differential turnout patterns in presidential and midterm elections, but the regularity with which the president’s party loses ground in the midterms suggests that voter oscillation, whether or not it consciously reflects a desire for divided government, is likely a factor. Democrats need to include in their 2020 messaging some recognition of this fact, and they to make it clear that any of their voters from the 2018 midterm who think voting for Trump will keep things under control in Washington will risk ushering in the most uncontrolled presidential term since Andrew Johnson decided to try to veto the results of the Civil War.