September 2019 update: 2020 Democratic presidential nomination scenarios and general election polling

With the third Democratic presidential nomination debate set for September 12, 2019 in Houston, Texas, it is time for an updated assessment of the relative position of the now-20 declared candidates. The more stringent criteria to qualify for this debate—the first to be held on only one night—presaged the end for three more campaigns. Washington Governor Jay Inslee announced on August 21 he would attempt instead to win a third term as governor. United States House of Representatives (“Representative”) member Seth Moulton of Massachusetts dropped out on August 23 (to seek reelection to his House seat), followed by New York Senator Kirsten Gillibrand on August 28. The six candidates who have thus far abandoned their quest to be the 2020 Democratic presidential nominee all exited the race with grace, class and dignity, and I commend them for it.

To learn how I calculate the value I assign to each candidate, NSW-WAPA (national-and-state-weighted weighted-adjusted polling average), please see here[1]. I recently made two other methodological changes. One, I now treat as distinct polling entities two sets of ABC News/Washington Post national-level polls of the 2020 Democratic nomination contest: 1) Two polls, conducted January 21-24 and April 22-25, of adults only which simply asked respondents to name their first choice (as opposed to being read a list of names and being asked to choose one) and 2) two more recent polls (June 28-July 1, September 2-5) which used a list of names. Two, as of their poll conducted August 24-27, I am now using the “likely voters” version of the Harris X tracking poll; I treat these as coming from a distinct “pollster” than the “registered voters” version.

And, of course, here is the September 2019 lighthouse photograph in my Down East 2019 Maine Lighthouses wall calendar.

**********

I begin with a heartfelt methodological apologia.

While conducting the “post-first-debate-polls-only” analysis I discuss below, I noticed a significant glitch in how I weighted polling within distinct polling firms. I was also not incorporating more recent national polls correctly. The combined effect was to lower every candidate’s final NSW-WAPA (thus) at most 1.7 percentage points (“points”), while increasing the “Don’t Know/Other” value 4.6 points. However, the relative ordering of the candidates and the spacing between them was unaffected.

Even so, it is an embarrassing error on my part, and I apologize.

Table 1 below aggregates data from all national and state-level polls publicly released since January 1, 2019 (as of 12 am EST September 12, 2019), including:

• 178 national polls (including 36 weekly Morning Consult tracking polls)
• 21 Iowa caucuses polls
• 25 New Hampshire primary polls
• 6 Nevada caucuses polls
• 19 South Carolina primary polls
• 43 Super Tuesday polls[2]
• 36 polls from 15 other states.[3]

This makes a total of 328 polls, up from 293 in the last update.

Table 1: National-and-state-weighted WAPA for declared 2020 Democratic presidential nomination candidates

Candidate	National	IA	NH	NV	SC	Post-SC	NSW-WAPA
Biden	29.0	24.9	23.8	28.6	40.3	27.8	28.7
Sanders	16.2	16.6	19.5	19.9	14.3	15.8	17.4
Warren	13.0	14.9	15.5	17.5	10.5	14.2	14.5
Harris	8.8	10.1	8.6	8.0	10.8	9.1	9.3
Buttigieg	5.3	9.6	8.9	6.5	5.3	6.3	7.5
O’Rourke	3.3	2.8	2.6	2.6	1.9	6.5	2.7
Booker	2.3	2.6	1.6	1.6	3.7	1.6	2.3
Klobuchar	1.2	2.8	1.0	1.0	0.7	1.0	1.5
Yang	1.4	0.8	1.9	1.5	0.9	0.9	1.2
Gabbard	0.8	0.8	2.0	1.2	0.4	0.7	1.1
Castro	1.0	0.8	0.2	1.2	0.2	1.4	0.70
Steyer	0.2	0.1	0.5	2.1	0.5	0.2	0.65
Delaney	0.3	1.0	0.6	0.00	0.4	0.2	0.49
Bennet	0.3	0.3	0.2	0.4	0.2	0.4	0.28
Williamson	0.3	0.1	0.4	0.4	0.4	0.1	0.28
Ryan	0.3	0.1	0.2	0.4	0.3	0.2	0.23
de Blasio	0.3	0.1	0.00	0.6	0.1	0.1	0.19
Bullock	0.3	0.4	0.00	0.2	0.1	0.1	0.17
Sestak	0.02	0.1	0.00	0.2	0.00	0.1	0.07
Messam	0.00	0.00	0.00	0.00	0.1	0.04	0.03
DK/Other	13.9	9.6	11.4	4.7	7.6	12.3	9.5

The race has settled into a kind of stasis following the first two rounds of debates. Former Vice President Joe Biden remains the frontrunner (28.7), primarily because of his 26-point lead in South Carolina primary polls. However, he is less strong in Iowa and New Hampshire, where the candidates battling for second place overall, Vermont Senator Bernie Sanders (17.4) and Massachusetts Senator Elizabeth Warren (14.5), are even closer to first place. Rounding out a clear top five, both overall and in the four earliest states, are California Senator Kamala Harris (9.3) and South Bend, Indiana Mayor Pete Buttigieg (7.5). These five candidates account for over three-quarters (77.4%) of Democratic voter preferences at this point, and the conventional wisdom is the 2020 Democratic presidential nominee will be one of them.

In the next tier are candidates with NSW-WAPA between 1.0 and 3.0 who could yet rise in the polls with strong debate performances: former Texas Representative Beto O’Rourke, New Jersey Senator Cory Booker, Minnesota Senator Amy Klobuchar, entrepreneur Andrew Yang and Hawaii Representative Tulsi Gabbard. Of this group, only Gabbard did not qualify for the September 2019 debate, though may yet meet the criteria for the fourth round of Democratic presidential nomination debates (October 15 and possibly October 16). Just behind Gabbard, essentially tied for 11^th place, are former Secretary of Housing and Urban Development Julian Castro and billionaire activist Tom Steyer; Castro is the 10^th and final September 2019 debate qualifier, while Steyer has already qualified for the October 2019 debate(s).

These 12 candidates total 87.0% of Democratic voter preferences. With 9.5%[4] undecided or choosing an unlisted candidate, the remaining seven candidates are divvying up just 3.5% between them; as none of them appears close to making the October 2019 debate(s), I expect them to end their campaigns by the end of 2019.

**********

Because I have been playing them out in my head, I will sketch out some likely nomination-winning scenarios at this point; as these are purely thought experiments, they should not be taken as predictions. These scenarios incorporate additional information like endorsements, my own study of the presidential nomination process[5] and the data in Table 2, derived only from polls that began on June 28, 2019 or later (i.e., after the first round of debates; n=137).

Table 2: National-and-state-weighted WAPA for declared 2020 Democratic presidential nomination candidates—starting date of poll June 28, 2019 or later

Candidate	National	IA	NH	NV	SC	Post-SC	NSW-WAPA
Biden	27.8	23.8	21.9	26.1	38.5	26.2	27.0
Sanders	15.9	14.5	19.8	21.4	15.0	15.3	17.3
Warren	15.4	18.7	19.0	19.0	12.0	16.7	17.2
Harris	9.2	12.2	10.2	8.2	12.9	10.0	10.8
Buttigieg	5.0	9.3	9.1	5.3	4.7	5.4	7.0
Booker	2.1	2.0	1.6	1.2	3.2	1.3	1.9
O’Rourke	2.4	1.4	1.2	1.8	1.0	6.2	1.9
Yang	1.9	1.0	2.4	1.3	0.9	1.1	1.4
Gabbard	1.0	0.9	2.6	1.3	0.5	0.9	1.3
Klobuchar	1.1	2.5	1.2	0.9	0.7	0.6	1.3
Steyer	0.4	0.02	0.7	3.4	0.9	0.3	1.0
Castro	1.1	0.7	0.3	1.4	0.2	1.4	0.75
Delaney	0.3	0.8	0.6	0.00	0.7	0.2	0.52
Bennet	0.3	0.3	0.4	0.5	0.3	0.5	0.38
Williamson	0.5	0.1	0.6	0.3	0.5	0.2	0.36
de Blasio	0.4	0.3	0.00	0.9	0.1	0.1	0.28
Ryan	0.4	0.00	0.00	0.6	0.5	0.2	0.25
Bullock	0.3	0.5	0.00	0.3	0.2	0.2	0.24
Sestak	0.03	0.1	0.00	0.4	0.00	0.1	0.11
Messam	0.00	0.00	0.00	0.00	0.00	0.1	0.01
DK/Other	13.7	9.6	7.2	3.8	6.0	12.2	7.8

The top five remains the same (and combine for a slightly higher 79.3%), though Biden and Buttigieg are lower, while Sanders, Harris and, especially, Warren are higher.  The next tier of seven candidates is closely bunched together between 0.7 and 1.9. And a slightly smaller percentage (7.8) are undecided or prefer a different candidate.

But the real differences may be seen in Iowa, where Warren is now a close second to Biden, and New Hampshire, which is essentially a three-way tie between Biden, Sanders and Warren; Nevada is similar, with Biden slightly further ahead—and Steyer has his best showing (3.4) by far. Gabbard (2.6) and Yang (2.4) are similarly rising in New Hampshire.

In these scenarios I assume two things:

1. Results in the four early states will continue to have an outsized impact on all subsequent contests,
2. Candidates not finishing in the top five in any of Iowa, New Hampshire or South Carolina will end their campaign before March 3, 2020

Scenario 1: Biden wins easily.

This is the most obvious scenario: the front-runner wins. Biden has consistently led national and most state polls. He leads Harris in endorsement “points,” though relatively few Democratic party officials have endorsed a candidate. He is winning the “Hillary Clinton 2016 coalition”: older, moderate/conservative and black Democrats. And even after two shaky debate performances, Biden is still nearly 10 points ahead of Sanders and Warren overall in the most recent polling—and about five points ahead in the leadoff Iowa caucuses.

In this scenario, Biden wins the Iowa caucuses by closer to 10 points, with Sanders, Warren and Buttigieg splitting the votes of whiter, younger, more liberal Democrats, and Harris a distant fifth; no other candidate is close. Biden, who lately has been downplaying his chances in the early states, looks like the winner he was presumed to be.

Klobuchar, who had pinned everything on Iowa, drops out of the race and endorse Biden (instead of a Senate colleague or the untested Buttigieg). This, along with support from supporters of former candidates, breaks the logjam in New Hampshire. Biden wins there in the high single digits as Sanders, Warren and Buttigieg continue to split the not-Biden vote between them. Nevada becomes an afterthought (allowing Steyer to finish a surprising fourth), as all eyes turn to South Carolina—where Biden wins easily.

After the remaining candidates make their last stands on Super Tuesday (Warren in Massachusetts; Harris, Yang, Steyer and Gabbard in California; O’Rourke and Castro in Texas), Biden wins the majority of delegates awarded that day, after which only Sanders and Warren are serious challengers. Uncommitted officials, sensing a chance to focus on President Donald J. Trump, quickly coalesce behind Biden. He effectively wraps up the nomination on April 28, when he trounces Sanders and Warren in the Pennsylvania primary, the state where he was born, and whose southern neighbor (Delaware) he served in the Senate for 36 years.

Scenario 2: Warren’s early state strength vaults her to the nomination.

This is essentially the inverse of Scenario 1 and, to my mind, about equally as likely. Warren wins the Iowa caucuses because her far-more enthusiastic supporters show up to a caucus site on a Monday night in early February. Biden’s supporters, however, do not caucus in expected numbers, and he finishes a disappointing third, behind Sanders and barely ahead of Buttigieg and Harris. Booker finishes a surprisingly strong sixth, thanks to a core of enthusiastic supporters, though it is still disappointing. Nobody else is even close; Klobuchar again drops out of the race, though this time she endorses Warren. Harris, meanwhile, focuses on doing well in South Carolina and California.

Sensing a winner, supporters of other “liberal” candidates break for Warren, who edges out Sanders to win the New Hampshire primary; Biden actually finishes third behind Buttigieg. Yang and Gabbard finish higher than expected, but not enough to garner any momentum. Warren then makes it three-for-three in Nevada, as even some Sanders voters caucus for her; Biden finishes fourth, behind a surprisingly-strong Steyer.

In a matter of days, the very core of Biden’s appeal—his “electability”—is irreparably damaged. He still wins the South Carolina primary on the strength of moderate/conservative Democrats alarmed by the one-two punch of Warren and Sanders; pragmatic black Democrats split their votes between Biden, Harris and Warren (with a smattering for Booker), who finish in that order.

Just prior to Super Tuesday, though, Booker, Buttigieg, Castro and Yang—all eyeing the vice-presidential nomination—drop out of the race and endorse Warren. As a result, on Super Tuesday, besides her home states of Massachusetts, Warren wins Colorado, Minnesota and, in a narrow upset, Virginia. Biden wins the southern states of Alabama, North Carolina, Oklahoma and Tennessee, but by smaller margins than expected.

Texas is an inconclusive muddle, with Biden, Warren and O’Rourke battling for first place (roughly in that order)—but the game-changer is California. On the strength of white liberals (who have broken decisively from Sanders, especially women) and a slight plurality of LatinX voters, and with Biden and Harris splitting the black vote between them, Warren wins the California primary.

Harris, O’Rourke and Steyer see the writing on the wall and drop out, endorsing Warren. Gabbard also drops out, but chooses to endorse Sanders instead.

It is now effectively a two-person race between Biden and Warren, the clear front-runner. Sanders soldiers on, despite not having won a single contest, though his percentages languish around 10-15%, just enough to win some delegates here and there.

Biden continues to win primaries in the south (including the Florida primary on March 17) with Harris and Booker out of the race, but his only win outside the south is Ohio on March 10. Making his last stand in Pennsylvania, his high-single-digit win is deemed a disappointment. Warren then wins the Indiana primary on May 3, effectively wrapping up the nomination.

Scenarios 1a and 2a: Biden and Warren split the early states

It is highly plausible that Biden wins Iowa and South Carolina, while Warren wins New Hampshire and Nevada. Sanders, Harris and Buttigieg survive to battle on Super Tuesday, but Warren ends up winning California (as just enough Harris supporters instead vote for Warren), while Biden wins Texas (with O’Rourke a distant third). Sanders soldiers on, but it would once again be a Biden-Warren race. Given his base of support, Biden would probably be the slight favorite in this scenario.

In fact, nearly every scenario I game out ends up with either Biden or Warren as the nominee. There are, however, some entertaining (if less likely) exceptions:

Scenario 3: Sanders wins Iowa and New Hampshire

This is essentially 2016 all over again (Sanders barely lost Iowa before a landslide win in New Hampshire), except with three or four other viable candidates instead of one.

In this scenario, Biden and Warren begin to attack each other directly in the September and October debates, ultimately depressing turnout for both as voting begins in February 2020. The beneficiary is Sanders, who suddenly appears to be the sage elder statesman (and whose head-to-head polling numbers against Trump cause a second look at his candidacy): his loyal supporters push him over the top in Iowa—ahead of a surprisingly-strong Buttigieg, who picked up many disgruntled Biden and Warren voters. Harris also benefits, essentially tying the latter two for third place.

New Hampshire now becomes a battle between Sanders and Buttigieg, with Harris focusing on South Carolina and California. Once again, New Hampshire supports the familiar neighbor, though Buttigieg again makes it surprisingly close.

And national Democrats get very nervous, despite those head-to-head polls.

Ignoring Nevada (which Sanders wins easily), and taking solace in the prospect of a solid, historic ticket to go against Trump and Vice President Mike Pence[6], they rapidly and tactically endorse Harris and Buttigieg. Having finished no higher than sixth in any of the first three contests, every other candidate drops out and endorses either Harris or Buttigieg (except Williamson and Gabbard, who back Sanders).

Biden and Warren fight on, but with their candidacies fatally damaged, Harris wins South Carolina, with Buttigieg a solid second, well ahead of Sanders. Harris and Buttigieg then make a strategic decision to make every Super Tuesday contest a one-on-one battle against Sanders, who can no longer win with around 20-25% of the vote.

On March 3, Harris wins California and the southern states (Alabama, North Carolina, Oklahoma, Tennessee, Texas, Virginia) while Buttigieg wins Colorado, Massachusetts and Minnesota. Sanders finishes second or third in every state, collecting some delegates, while Biden and Warren see their campaigns end with a whimper.

Opening a commanding one-two lead in delegates, Harris and Buttigieg follow the same path as Kerry and North Carolina Senator John Edwards in 2004: a relatively calm trek through the remaining primaries and caucuses, with Harris steadily closing in on the nomination. Once she does, she surprises nobody by selecting Buttigieg as her running mate, just as Kerry chose Edwards in 2004.

Scenario 4: Someone other than Biden, Sanders or Warren wins Iowa and/or New Hampshire

This is the true wild card scenario, which is nearly impossible to game out at this point.

Still, let us suppose Booker continues to have very strong debate performances, and with them now on a single night (for the sake of argument, the October debate is one night), this finally resonates with voters, who take a long second look at him.

As a result, he surprises everyone by winning the Iowa caucuses. Rather than campaign in New Hampshire or Nevada, which he concedes to Sanders and Warren, he focuses exclusively on South Carolina. Sensing weakness in Biden, and excited by his Iowa win, black voters overwhelmingly support Booker, who just edges Biden, effectively ending that latter’s campaign.

Booker and Warren (who won New Hampshire and Nevada) then battle it out on Super Tuesday. With Harris making a last stand in California, and O’Rourke doing the same in Texas, Warren wins the former, Booker the latter. Warren and Booker now effectively replay the 2008 campaign between then-Senators Barack Obama and Clinton, with the most likely outcome either a Warren-Booker or a Booker-Warren ticket.

An alternative scenario sees Buttigieg finishing in the top two or three in Iowa (behind Biden and Warren in some order), then winning the New Hampshire primary. Nevada is again an afterthought, and Biden wins South Carolina. This feels somewhat like the 1988 Republican nomination battle, when then-Senator Bob Dole of Kansas won Iowa, but ultimately lost to the front runner, then-Vice-President George H. W. Bush; Biden thus eventually prevails.

What these scenarios suggests is that, based upon what we know now, Biden and Warren are far and away the most likely nominees, but there are universes in which Harris or Booker win the nomination. By contrast, it is very difficult to see Sanders or (probably) Buttigieg doing so…or any other candidate, for that matter.

**********

Because Democrats other than Biden and Sanders, who would beat Trump nationally by 9.0 and 5.7 points, respectively, are also winning (or barely losing) hypothetical head-to-head match-ups, “2020 Democratic nominee” (averages vs. Trump weighted by likelihood of being the nominee) now beats Trump by 4.1 points. This is higher than the median Democratic presidential margin (+3.0 points) in the previous six presidential elections, which include three elections with an incumbent seeking reelection and three elections with no incumbent. However, once you exclude Biden and Sanders, the margin over Trump decreases to 1.0 points; Warren would hypothetically win by 2.5 points and Harris by 1.9 points, while Buttigieg, O’Rourke and Booker would lose by less than one point.

Still, given that state-level results actually determine the winner of a presidential election (via the Electoral College), it is more informative to look to those polls, where they are publicly-available. Using my 3W-RDM, a measure of how much more or less Democratic a state’s voting is relative to the nation as a whole, this polling[7] implies Democrats would win the national popular vote by between 3.2 (excluding Biden and Sanders) and 6.5 (including Biden and Sanders) points on average. Most encouraging to Democrats should be the polls from North Carolina (R+6.0) and Texas (R+15.3), which show a very close race, implying a national Democratic lead of 5-7-and 12-15-points, respectively; these polls confirm strong opportunities for Democrats in the southeast and southwest. By contrast, however, a few polls from Democratic-leaning Maine (D+5.9) and Nevada (D+2.0) imply Democrats would lose nationwide by 1-6 points. Those remain the exceptions, however, to what continues to be encouraging news for Democrats in 2020.

Until next time…

[1] Essentially, polls are weighted within areal units (nation, state) by days to the nominating contest and pollster quality to form a unit-specific average, then a weighted average is taken across Iowa (weight=5), New Hampshire (5), Nevada (4), South Carolina (4), the time-weighted average of all subsequent contests (2) and nationwide (1). Within the subsequent contests, I now weight the 10 March 3, 2020 “Super Tuesday” states (Alabama, California, Colorado, Massachusetts, Minnesota, North Carolina, Oklahoma, Tennessee, Texas, Virginia) twice as much as the subsequent contests. As of this writing, I have at least one poll from (in chronological order) Maine, Michigan, Mississippi, Missouri, Ohio, Washington, Arizona, Florida, Illinois, Georgia, Wisconsin, Pennsylvania, Indiana, Oregon and New Jersey.

[2] Primarily Texas (15), California (14)

[3] Primarily Florida (9), Pennsylvania (5), Wisconsin (5)

[4] This does include polls that limit the number of candidates queried.

[5] As a doctoral student in government at Harvard in the early 1990s, I was a teaching assistant for an undergraduate course on this very topic.

[6] As of this writing, anyway.

[7] From Pennsylvania, New Hampshire, Wisconsin, Michigan, North Carolina, Texas, Iowa, Arizona, South Carolina, Minnesota, Nevada, Massachusetts, Florida, New York, Kentucky, Maine, Ohio, North Dakota, California, Alaska, Washington, Colorado, Missouri, Utah.