Born To Run The Numbers: May 2020

Tom with a BTRTN update of the 2020 presidential race...not a prediction, just a “snapshot” of where the race stands right now, and how it might turn out if Election Day were held today.

THE LEAD

· It is early, to be sure, a lifetime until November in a race that will be far more unpredictable than ever, given the unprecedented impact of the coronavirus on our nation’s health and economy, and almost certainly the very election process itself.

· Joe Biden can almost certainly count on the electoral votes from 18 voting entities (16 states, DC and one Maine district) totaling 210 blue votes on the path to 270; Donald Trump can count on 20 states (plus two Nebraska districts) for a total of 125 solid red electoral votes.

· Thus, the outcome of the race will almost surely depend on 14 swing states (plus one district in Maine and one in Nebraska), holding 203 electoral votes, where either candidate has at least a shot of winning

· There have been a total of 46 polls conducted in these 14 swing states over the past three months; Biden has led in 31, Trump in 13, with two ties. Biden leads the early polling in 11 of the 14 states, while Trump is ahead in three.

· According to our BTRTN model, if the elections were held today, Biden would have a 73% chance of winning the presidency.

You are probably thinking one of two things right now, or both: 1) we have heard this before, in 2016, and look how that turned out, and/or 2) yikes, even if I believe this, reading it will only have the effect of making Democrats overconfident or complacent.

Yes, in 2016, Donald Trump pulled an inside straight on Election Day, overcoming 1 in 3 odds, helped by James Comey, GOP voter suppression, protest votes for third party candidates and the Russians. Yes, some of this could happen again. But if complacency was a factor in 2016, well, Dems should not be fooled again, regardless of the shape of the race in the run-up to Election Day.

But there is one undeniable takeaway from these early results: Joe Biden is a very viable candidate. He has been forced to endure the pandemic in his house; has been hit with the Tara Reade charges; and, yes, continued to be a gaffe-machine (such as the recent “then you ain’t black” statement). But nonetheless, he is faring extremely well in the polls in the early going. Perhaps his supporters are as resilient as Trump’s in the face of his flaws – and he has fewer flaws, and more supporters.

But, just in case the following message somehow gets lost with all the positive numbers, we offer this warning label, to be repeated in this document.

WARNING: No matter how good the numbers look at any given time, the Democrats will not win any election, and especially the presidential election, unless they work hard to earn it – registering voters, calling, texting, donating – all through the summer and fall, up to and including Election Day.

THE STATE OF PREDICTING PRESIDENTIAL RACES

You have probably noticed that forecasting elections has become more difficult.

Back in the good old days, Americans could more than occasionally so completely align on a presidential candidate that an overmatched one – say a Barry Goldwater, a George McGovern, or a Walter Mondale – could barely pick-off a state. Yes, once a decade there used to a “landslide” presidential election, which made it easy on the pundits.

Those days are gone. Polarization has set in with the totality of a glacier, wiping out virtually any notion of dominance in a presidential race. Three factors rise above all as catalysts in creating our current state – the 1964 Civil Rights Act, 1965 Voting Rights Act and the entitlement legislation, which ultimately aligned Democrats and liberals, and Republicans and conservatives along geographic lines; the “culture wars” of the 1980’s which divided us into warring camps not just over economic and defense policies, but how we lived our lives; and the rise of personal media, which ensured we could source any information we desired, including, sadly, our own set of facts, that were consistent with our own thought bubbles.

This polarization has, of course, made every presidential contest much closer, as the influence of our ideology can be more of a factor than the candidates themselves. We have not had a “landslide” since Ronald Reagan pummeled Mondale in 1984 (59% to 41%, 525 to 12 electoral votes, as Mondale took only his home state of Minnesota). George H.W. Bush did give Michael Dukakis a pretty good thumping in 1988 (53/46, 426/111), but since then every “loser” has managed to keep the margin within 8 points, and garner at least 150 electoral votes.

Every election is now a nail-biter. And since we continue to cling to the anachronistic Electoral College to elect our presidents, these races are contested, and won and lost, in a handful of “swing states.”

While races have become closer, prediction techniques have improved radically. The new toy, circa 2008, is to aggregate polls, thereby increasing sample sizes and reducing margins of error. Ever improving polling techniques have helped, as has the sheer volume of polls themselves. So in 2008, Nate Silver (and BTRTN), through the newfound wonders of “aggregated” polls and regression-based models, forecasted the entire race accurately in advance, state by state. Our followers watched, in awe, as the returns came in on election night exactly as predicted (Nate was perfect, we missed Florida) on a magic November night, and Barack Obama became president. And all that was dutifully repeated in 2012, albeit with a host of other “aggregators” now crowding the scene.

But the illusion of certainty was shattered in 2016, when Hillary Clinton, a 2-to-1 favorite on Election Day according to the aggregators, was defeated by Donald Trump. She was done in by the freakish alignment of a poor campaign strategy that nearly forsook the industrial heartland; James Comey’s ill-timed “more emails” and “nothing wrong with them” announcements; lower than expected turnout; some of it doubtlessly illegally suppressed; at times questionable and infrequent swing state polling; third party candidacies getting “protest vote” support, and on and on.

Nate Silver may have done a decent job at explaining probabilities (much better than we at BTRTN), but still, no one was prepared for the shocking Trump upset win. No one could quite believe it had happened, even though it was subsequently pointed out that, that same year, 2016, both the Cleveland Cavaliers and the Chicago Cubs had overcome similar odds by winning the final two games of their championship series after trailing three games to two. .

So now we come to 2020. Can we be any more confident of our predictions as we move toward Election Day? Has anything changed? Well – the world is even more polarized. The Electoral College is still in place. The prediction models are no better. And, oh yes, we are in the middle of a killer global pandemic, that has shut down the U.S. economy, and has no clear end in sight. Talk about uncertainty – we don’t even know in what form the November elections will take place, the usual mix of mail-in plus the ballot box, or simply the former.

(For what it is worth, an in-depth article in The New York Times concluded that there is little conclusive evidence that an all or mostly mail-in election would favor one party over the other. Essentially, mail-in results in slightly higher turnout, which would favor the Dems, but none of the studies have taken place in the context of coronavirus, and finding Democrats, who tend to be younger and more mobile, may be harder in the pandemic.)

And so we carry on. There is no doubt that the “snapshots” we (and others) will give in the coming months will be reasonably accurate representations of the state of the race. But there are no guarantees that where we land on November 2^nd will actually come to pass on November 3^rd. There will only be probabilities.

So why read on? Well, if you are the type of person who believes it is only worth watching an NBA game with less than five minutes to go, then by all means, stop now.

But if you want some sense of how the race stands in these remarkable times, read on.

WHAT ARE THE SWING STATES?

Any presidential campaign starts with a game plan – a road map to get to 270 votes. First, one sorts out which states are reliably blue or red, and thus not worthy of scarce campaign resources. And that leaves the battleground – the so-called swing states that are “in play.”

We believe Joe Biden can already count on 210 out of the 270 electoral votes he needs for the nomination. There are 13 states plus the District of Columbia, comprising 182 electoral votes, that have supported the Democratic candidate in each of the five 21^st century elections (including Hillary Clinton), usually by wide margins. Those 14 voting entities represent the “Blue Wall” and they are: California, Connecticut, Delaware, DC, Hawaii, Illinois, Maryland, Massachusetts, New Jersey, New York, Oregon, Rhode Island, Vermont and Washington.

We believe Biden can safely add in New Mexico, which went for Bush twice but then flipped to Obama for two double digit wins, and Clinton won by +8. And, our view is that Virginia and Colorado are also solid blue at this point. Clinton beat Trump by only +5 points in both states in 2016, but Biden is clobbering Trump in early polling in each, solidly in the double digits. Throw in Maine’s 1^st District (Maine is one of two states, along with Nebraska, that does not award delegates on a winner-take-all basis), which Clinton won by +15 points, and you have 18 voting entities totaling 210 electoral votes, that appear to be solid blue for Biden.

As for Trump, he has a ton of southern and western states that are deep red, 20 states plus two Nebraska districts that add up to 125 delegates that form the Trump red brick wall.

That leaves 14 swings states and two swing districts, 203 Electoral College votes “in play.” These include the famous six that Donald Trump flipped in 2016 versus Barack Obama’s 2012 winning reelection map: Florida, Iowa, Michigan, Ohio, Pennsylvania and Wisconsin. You then throw in the races that Hillary Clinton won by three points or less: Maine, Minnesota, New Hampshire and Nevada. Also include red states that were relatively narrow wins for Trump, or bigger wins but shifting demographics that will benefit the Democrats: Arizona, Georgia, North Carolina and Texas.

Could these categories change as the election progresses? Of course. Strategies change, polling changes, and states drift in and out of contention. But this is the playing field as of today and it will likely not change too much.

HOW THE SWING STATES ARE SHAPING UP?

You can be sure of a few things. The first is that Donald Trump’s campaign team believes in polling; they conduct internal polls and they base strategy on them. There is no question of #FakePolls within that campaign. And second, the polls they are staring at look a lot like the public polls. And they are extremely concerned – and have shared those concerns with Trump recently in the White House.

There have been 46 public polls conducted in the 14 swing states since March 1, and Biden has led Trump in 31 of them, with Trump ahead in 13 and two ties. It is not as if the 46 polls have been skewed toward the Democratic states; if anything, it has gone the other way. Of the 46 polls, 28 have been conducted in what we would view as “toss up” states, 14 in Trump-leaning states, and only four in Biden-leaning states.

Let’s look at each in turn, first focusing on the four swing states that went for Clinton in 2016, then the six that flipped from Obama to Trump that gave him his win in 2016, and finally the four that remained red for Trump but could be in-play in 2020.

Clinton Swing States (4): Maine, Minnesota, New Hampshire and Nevada

There has been little polling in these four states to date, just a single poll in each. Each of them show Biden ahead, even comfortably ahead. Maine’s poll was in March, and Biden was +10. Minnesota, in a brand new poll last week, had Biden +5. New Hampshire, in April, had him +8, and Nevada, also in April, was +4. Since these are Clinton states to begin with, and Biden leads in the scant polling to date, we have these states in the “Leaning to Biden” bucket.

The Flips (6): Arizona, Florida, Michigan, North Carolina, Pennsylvania and Wisconsin

Each of these states flipped from Obama to Trump in 2016, and the 102 electoral votes that came with them put Trump over the top. Trump won four of them by 1.2 points or less.

There has been extensive polling in these states, 28 polls in all over the past three months. Remarkably, Biden is ahead in 23 of them, and there is one tie. Trump is ahead in only 4 of the 28 polls in the six states he flipped in 2016.

On average, the polls have favored Biden in five of the six states, and particularly in Michigan and Pennsylvania.

Swing State	# Polls M/A/M	Avg. Margin
MICH	4	D + 6
PA	4	D + 7
WIS	4	D + 3
ARIZ	5	D + 4
FLA	7	D + 2
NC	4	R + 1

Despite the polls, we still consider these states to be toss-ups, but we see the first four, as of now, as “Leaning to Biden,” with Florida and North Carolina “Leaning to Trump.”

Trump Swing States (4): Georgia, Iowa, Ohio, Texas

Trump won these four states handily in 2016. The closest was Georgia by +5 points, then Iowa and Texas by +9 and Ohio by +11. Why do we think they may be in play for Biden in 2016?

For Georgia and Texas, it’s a matter of demographics. Georgia has added about 1.1 million new voters to its rolls since 2016, driven by automatic registration at motor vehicle bureaus and a rising population. The new demographics are younger, more ethnic (Asian and Hispanic); whites are down from 62% of the voter rolls to 59%. In addition, the GOP has been having trouble with Atlanta’s suburban white voters, so much so that Governor Brian Kemp named Kelly Loeffler to fill the vacated seat of the retired Johnny Isakson, in a blatant bid to appeal to those disaffected suburbanites, instead of Trump favorite Doug Collins. Erasing that 5-point Clinton loss is an opportunity for Biden, and one reason that Stacey Abrams could find herself on the ticket.

The polls are very tight – one poll back in early March had Trump up by +8, but since then there have been five polls that show one or the other of Trump or Biden up by two points or less. We rate Georgia as “Leaning to Trump,” but it could become a toss-up soon.

Texas was one of the few states that Trump did worse in than Mitt Romney, winning by +9 over Clinton versus Romney’s +16 over Obama. Clearly demographics are a factor here, too. Texas may be an election cycle away from being a toss-up state, but the threat is there. Two Texas polls had Trump up by roughly +5, and two others show a dead-heat. We have Texas “Leaning (rather heavily) to Trump” -- in part because it is highly unlikely at this point that Biden will commit significant resources to flip this large state when he has other viable paths to 270.

Ohio and Iowa are not moving blue demographically, but the polls are indicating that Biden is going to be very competitive in these Midwestern states. The strength he is showing in those heartland flipped states (Michigan, Iowa, Pennsylvania, Wisconsin) is apparent in these two as well. In Iowa, Biden was down only -2 points in the most recent poll in the beginning of May, and in Ohio, he was -3. We have both of these “Leaning to Trump” as well.

In addition, we have thrown Maine’s 2^nd district in the “Leaning to Trump” category, as he won it by +10 in 2016. It may actually be closer to a toss-up, given the overall polling, but we will leave it for now. We have Nebraska’s second district as a “Toss-Up Leaning to Trump”; he won it by only +2 in 2016.

This chart summarizes the breakdown of our BTRTN snapshot as of now:

BTRTN SNAPSHOT: 5/28/2020
Categories	Voting Entities	Electoral Votes
Total	56	538
DEM TOTAL	26	289
Dem Solid	18	210
Dem Lean	4	22
Dem Toss-up	4	57
GOP Toss-up	3	45
GOP Lean	5	79
GOP Solid	22	125
GOP TOTAL	30	249

There is a chart below that summarizes all 57 races in detail.

THE ODDS

At this early juncture, if the election were held today, we peg the odds of Biden winning the presidency at 73%. The swing state polling paints a clear picture, but is just one of many data points that are problematic for Trump, including:

· His relatively low approval rating of 44%; only George W. Bush won re-election with an approval rating of less than 50%, at 48%. And Bush, of course, barely won.

· The low (and partisan-driven) assessment of his handling of the coronavirus, with 50% disapproving to 42% approving

· The shattering of Trump’s “trump card,” the healthy state of the economy, by the pandemic

· The generic ballot, which consistently shows the Democrats ahead of the GOP by 6-10 points, on average about +8 of late.

· Apart from the 14 swing states, the early polls hold troubling news for Trump even in his strongholds. Consider five states: Mississippi, Missouri, Kansas, Indiana and Utah. Trump won each of them by roughly +20 points in 2016, but in recent polls in each states, he is up by only +10-12 points -- except for Utah, where he leads Biden in a brand new poll by only 3 points.

· And even in the deepest red states, we see the same phenomenon. Trump won Kentucky by +30 points, but recent polling has him up by half that amount. The same is true in North Dakota, which he won by +36 in 2016, but now leads by only half that.

% Odds of Winning
Biden	73%

Here is a chart that provides detail on each of the 56 voting entities, the 50 states, District of Columbia, three Nebraska districts and two Maine districts. The “swing entities” are between the two thick blank lines.

States	2020 Electoral Votes	Past Election Results (Margin Dem Minus GOP)					Swing State Recent Poll Avg	BTRTN Rating
		Past Election Results (Margin Dem Minus GOP)
		2000	2004	2008	2012	2016
DC	3	76	80	86	84	89		D Solid
HAW	4	18	9	45	43	32		D Solid
CAL	55	12	10	24	21	30		D Solid
VT	3	10	20	37	36	29		D Solid
MASS	11	27	25	26	23	27		D Solid
MD	10	16	13	25	25	25		D Solid
NY	29	25	18	27	27	21		D Solid
WASH	12	6	7	17	14	18		D Solid
ILL	20	12	10	25	16	16		D Solid
RI	4	29	21	28	27	16		D Solid
ME 1	1	8	12	23	22	15		D Solid
CT	7	18	10	22	18	13		D Solid
NJ	14	16	7	16	17	13		D Solid
DEL	3	13	8	25	19	12		D Solid
ORE	7	0.4	4	16	12	11		D Solid
NM	5	0.1	-1	15	10	8		D Solid
VA	13	-8	-8	6	3	5		D Solid
COL	9	-8	-5	9	5	5		D Solid

MAINE	2	5	9	17	15	3	Biden +10	D Lean
NH	4	-1	1	10	6	0.3	Biden +8	D Lean
NEV	6	-4	-3	13	7	2	Biden +4	D Lean
MINN	10	2	3	10	8	2	Biden +5	D Lean
MICH	16	5	3	16	10	-0.2	Biden +6	D TU
PA	20	4	3	10	5	-0.7	Biden +7	D TU
WIS	10	0.2	0.4	14	7	-0.8	Biden +3	D TU
ARIZ	11	-6	-10	-9	-11	-4	Biden +4	D TU
FLA	29	0.0	-5	3	1	-1.2	Biden +1	R TU
NC	15	-13	-12	0.3	-2	-4	Trump +1	R TU
NEB 2	1	-18	-22	1	-7	-2	n/a	R TU
GA	16	-12	-17	-5	-8	-5	Trump +4	R Lean
ME 2	1	1	6	12	9	-10	n/a	R Lean
IOWA	6	0.3	-1	10	6	-9	Trump +2	R Lean
OHIO	18	-4	-2	5	2	-11	Trump +3	R Lean
TX	38	-21	-23	-12	-16	-9	Trump +6	R Lean

SC	9	-16	-17	-9	-11	-14		R Solid
ALASK	3	-31	-26	-22	-13	-15		R Solid
MISSP	6	-17	-20	-13	-12	-19		R Solid
UTAH	6	-41	-46	-28	-48	-19		R Solid
IND	11	-16	-21	1	-11	-19		R Solid
MO	10	-3	-7	-0.1	-10	-19		R Solid
LA	8	-8	-15	-19	-17	-20		R Solid
MON	3	-25	-21	-2	-14	-21		R Solid
KAN	6	-21	-25	-15	-22	-21		R Solid
TENN	11	-4	-14	-15	-21	-26		R Solid
NEB 1	1	-23	-27	-10	-16	-22		R Solid
NEB 3	1	-46	-51	-39	-42	-55		R Solid
NEB	2	-29	-33	-15	-23	-26		R Solid
ARK	6	-5	-10	-20	-24	-27		R Solid
ALAB	9	-15	-26	-22	-22	-28		R Solid
KY	8	-15	-20	-16	-23	-30		R Solid
SD	3	-23	-22	-8	-18	-30		R Solid
IDAHO	4	-41	-38	-25	-32	-32		R Solid
ND	3	-28	-27	-10	-20	-36		R Solid
OKL	7	-22	-31	-31	-34	-36		R Solid
WV	5	-6	-13	-13	-27	-42		R Solid
WYO	3	-41	-40	-32	-41	-48		R Solid

Thursday, May 28, 2020

BTRTN 2020 Presidential Race State-by-State Snapshot: Extremely Early, But Biden Clearly Ahead