Month: August 2020

The Republican advantage in the Electoral College is real

As I detail here, the Electoral College (“EC”), not direct popular vote, determines who wins American presidential elections. Since 1856, the first presidential election in which the two major candidates were a Democrat (James Buchanan) and Republican (John C. Fremont), there were four presidential elections in which one candidate won the EC while another candidate won the popular vote; in all four elections—1876, 1888, 2000, 2016—the Republican won the EC and, thus, the presidency.

Those elections—just four out of 41 (10%)—could be considered flukes, were it not for the fact Republicans maintain a clear, quantifiable advantage in the EC.

One way to think about this is to consider a presidential election in which the two major-party candidates receive exactly the same number of popular votes. Put another way, this is a situation where the difference between the Democratic percentage of the popular vote and the Republican percentage of the popular vote equals 0.0%.

If there was no partisan advantage in the EC, we would expect both candidates to receive 269 electoral votes (“EV”), exactly half of the 538 available to them. Or, at least, a number very close to 269, allowing for third-party candidacies and “faithless” electors who vote for someone other than the plurality winner of their state.

Table 1 lists the winner, political party, popular vote margin (Democratic % – Republican %) and number of EV won by the winning candidate for the 17 presidential elections from 1952 through 2016. I chose 1952 because it was the first presidential election to feature television commercials by the major candidates and televised nomination conventions. It is also the first election to show cracks in the previously solid Democratic south: Republican Dwight D. Eisenhower won Florida, Tennessee, Texas and Virginia in 1952, adding Kentucky and Louisiana in 1956. As usual, all elections data come from Dave Leip’s indispensable Atlas of U.S. Presidential Elections.

Table 1: Winning Presidential Party, Margin of Victory (Dem % – GOP %) and Electoral Votes Won: 1952-2016

Year Electoral College Winner Party Margin EV
1952 Eisenhower Republican -10.9% 442
1956 Eisenhower Republican -15.4% 457
1960 Kennedy Democratic 0.2% 303
1964 Johnson Democratic 22.6% 486
1968 Nixon Republican -0.7% 301
1972 Nixon Republican -23.1% 520
1976 Carter Democratic 2.1% 297
1980 Reagan Republican -9.7% 489
1984 Reagan Republican -18.2% 525
1988 GHW Bush Republican -7.7% 426
1992 B Clinton Democratic 5.6% 370
1996 B Clinton Democratic 8.5% 379
2000 GW Bush Republican -0.5% 271
2004 GW Bush Republican -2.5% 286
2008 Obama Democratic 7.3% 365
2012 Obama Democratic 3.9% 332
2016 Trump Republican -2.1% 304

Republicans won 10 of these elections, by an average margin of 8.6% in the popular vote and 393.9 EV; this includes 2000 and 2016, when Democrats Al Gore and Hillary Clinton, respectively, won the popular vote but lost the EC. Democrats, meanwhile, won seven of these elections by an average margin of 7.1% and 361.7 EV.

In three elections, the Democratic and Republican percentages of the popular vote differed by less than one percentage point (“point”): 1960, when Democrat John F. Kennedy won by 0.2% with 303 EV; 1968, when Republican Richard M. Nixon won by 0.7% with 301 EV, with 13.5% and 46 EV for American Independent nominee George Wallace; and 2000, when Republican George W. Bush lost by 0.5%, but still eked out 271 EV after a controversial recount in Florida.

It is difficult to discern any sort of pattern here, other than the higher the popular vote margin, the more EV you win. Figure 1 shows this clearly.

Figure 1: Popular Vote Win Margin and Electoral Votes Won, 1952-2016

Winning Electoral College

As expected, there is a strong linear association between popular vote margin and EV won—including 2000 and 2016; margin alone accounts for 86% of the variance in EV. The formula was calculated using ordinary least squares (“OLS”) regression, and it tells us the average number of EV one would expect a presidential candidate to win based upon their popular vote margin.

Thus, for every 1.0-point increase in popular vote margin (expressed as 0.01), that candidate wins an 11.1 additional EV, on average. Moreover, when the margin is 0.0 points—a popular vote tie—the winner should receive 292 EV, 23 more than the expected 269 or so. Also, to earn 270 EV, the winner would actually LOSE the popular vote by 2.0 points!

How is this possible?

Figures 2 and 3, which break down the popular vote margin-EV association by party, help to explain.

Figure 2: Popular Vote Margin and Electoral Votes Won: Democratic Presidential Candidates, 1952-2016

Dem Electoral College

Figure 3: Popular Vote Margin and Electoral Votes Won: Republican Presidential Candidates, 1952-2016

GOP Electoral College

For both major political parties, every 1.0-point in popular vote margin increases EV earned by an average of 12.3. And in both models, popular vote margin alone accounts for 92% of variance in EV; these two variables are VERY strongly linearly associated.

However, it is where the fitted line crosses the Y axis that makes all the difference—this is the expected EV won by each political party in the event of a tied popular vote. For Democrats, a tie equates to only 251.0 EV, on average, 19 fewer than needed to win the presidency. For Republicans, however, that same tie equates to 282.7 EV, 12.7 more than needed to win the presidency. The slopes and r-squared values are identical, the Republican line is just 31.7 points higher at every value of popular vote margin.

In other words, on average, a tied total popular vote translates to a 283-251 Republican win in the Electoral College, with 4 EV going to third-party candidates or otherwise up for grabs. That translates to a 32-EV Republican advantage in the Electoral College.

Another way to measure Republican advantage is to calculate what popular vote margin a presidential candidate needs, on average, to secure 270 EV. For the Democratic presidential nominee, the formula is:

Popular Vote Margin = (270 – 251.0) / 1233.1) = = +1.5%

And for the Republican presidential nominee, the formula is:

Popular Vote Margin = (270 – 282.7) / 1230.8) = = -1.0%

That is, a Democratic presidential nominee must win the national popular vote by at least 1.5 points to secure a minimum 270 votes, while a Republican presidential nominee can do so while losing the total popular vote by 1.0 points. Third-party EV and faithless electors keep the values from being identical. Still, that translates to a 2.5-point popular vote advantage for Republicans in the Electoral College!

This is a very robust finding. For example, while the 2016 election looks like an outlier—as does, to be fair, 1980, when Republican Ronald Reagan converted a 9.7-point popular vote margin of victory into 489 EV—removing it only improves the Democratic position slightly: a tied total popular vote still gives the Republican a 279.4 to 254.1 EC victory, while only reducing the popular vote advantage to 2.1 points.

Also, starting in 1964—the first election in which Alaska, Hawaii and the District of Columbia all contributed EV—actually increases the Republican advantage. The latter nominee would win 286.4 to 248.5 if the total popular vote were tied nationally, a 37.9 EV advantage equivalent to winning Texas’ 38 EV. Moreover, since 1964, a Democrat would need to win nationally by 1.8 points, on average, to win 270 EV, while a Republican could lose by 1.3 points—a gap of 3.1 points!

**********

My analysis of national- and state-level polling suggests 2020 Democratic presidential nominee Joseph R. Biden, Jr. currently leads Republican President Donald J. Trump by 7.5 points, up slightly from the last time I wrote about the state of the race.

Plugging 0.075 into the Democratic formula yields a projected EV total of 343.5. This is remarkably close to the 349.6 EV I estimate Biden will received based on in my model; the number increases to 352 if I simply count up the EV from states I calculate Biden has >50% chance to win.

I also assess Biden’s chances if all polls are systematically over-estimating Democratic strength by 3 points and if all polls are systematically under-estimating Democratic strength by 3 points. That is, I consider a universe in which Biden is actually ahead by 4.5 points or by 10.5 points.

Entering 0.045 into the Democratic formula yields an expected 306.5 EV—basically, the states won by Clinton in 2016 plus Michigan, Pennsylvania, Wisconsin and Florida. My estimate of Biden’s EV in this scenario is 301.7, or 308 using the states where Biden is better than even money.

Entering 0.105 into the Democratic formula yields an expected 380.5 EV—basically, the previous scenario plus Arizona, North Carolina, Ohio, Georgia and Iowa…but not Texas. The polling data, however, suggest Biden would do even better—389.6 EV, or 412 using the states where Biden is better than even money, including a 57.6% chance of winning Texas.

In other words, results from this simple one-variable model align almost exactly with the state of the race based on available polling data.

**********

Why this is the case, however, is a far more complicated question. The most direct answer is that Democratic votes are distributed less efficiently than Republican votes across the states. Democrats routinely win large states like California, New York and Illinois—104 total EV—by double-digit margins, padding their national vote total while adding 0 EV. Republicans, by contrast, win larger states by narrower margins, as they did in 2016, and smaller states like Wyoming, Idaho and Utah by huge margins which only minimally impact their national vote total.

Put differently, the Democratic strategy to run up the popular vote paradoxically hurts them in the Electoral College, while Republican strategy to eschew large national vote totals in favor of narrower wins in key states boosts them.

On a related note, Republicans have an advantage in swing states, as my 3W-RDMdemonstrates; this is a measure of how Democratic a state votes relative to the nation. Based solely on this measure, Table 2 lists how I would anticipate the following states to vote if Biden and Trump exactly tied in the total popular vote:

Table 2: Expected 2020 vote margins in 18 key states, based on tied popular vote

State EV 3W-RDM Current Biden polling
Oregon 7 8.7 n/a
New Mexico 5 6.5 11.3
Maine 4 5.9 10.4
Michigan 16 2.2 7.1
Colorado 9 2.2 12.4
Nevada 6 2.0 3.5
Minnesota 10 1.5 8.1
Virginia 13 1.5 9.6
Wisconsin 10 0.7 4.6
New Hampshire 4 0.1 5.2
Pennsylvania 20 -0.4 5.2
Florida 29 -3.4 3.9
Iowa 6 -4.7 -1.7
Ohio 18 -5.8 0.7
North Carolina 15 -6.0 2.0
Georgia 16 -9.6 -0.5
Arizona 11 -9.7 2.7
Texas 38 -15.3 -2.0

Assuming Biden starts with 175 EV[1] and Trump starts with 126 EV,[2] that leaves 237 EV up for grabs. In this popular-vote-tie scenario, Biden wins 84 of those EV, though Wisconsin and New Hampshire could be looking at a recount, for 259 EV. Trump wins the remaining 153 EV, with a recount possible in Pennsylvania, for 279 EV, very close to the 283-251 EV margin estimated earlier.

But while in this scenario Biden would narrowly win states like Michigan, Colorado, Nevada, Minnesota and Virginia—Trump would be looking at far easier wins in Florida, Iowa, Ohio, North Carolina, Georgia, Arizona and, especially, Texas. This is because of the enormous gap between Pennsylvania, at 0.4 points more Republican, and Florida, at 3.4 points more Republican. While a Democrat could theoretically win 279 EV—and the White House—by winning the total popular vote by 0.4 points, s/he would have to win by at least 3.4 points nationally to have a little breathing room.

That all said, Biden’s current estimated lead of 7.5 points gets him those 308 EV relatively easily, while making him slightly favored in Arizona and North Carolina, perched on the razor’s edge in Ohio and Georgia, and pounding on the door in Iowa and Texas. A slight polling error in his favor, strong Democratic turnout/depressed Republican turnout and a decisive win among late-deciding voters, and Biden could turn 343-352 EV into 412 EV.

When you know in advance how high the mountain you need to climb is, it is far easier to prepare to climb it.

Until next time…please stay safe and healthy…

[1] District of Columbia, Hawaii, Vermont, California, Maryland, Massachusetts, New York, Rhode Island, Illinois, Connecticut, Delaware, Washington, New Jersey

[2] South Carolina, Missouri, Indiana, Mississippi, Montana, Alaska, Louisiana, Kansas, Nebraska, South Dakota, Tennessee, Arkansas, Alabama, Kentucky, North Dakota, Utah, Idaho, West Virginia, Oklahoma, Wyoming

Matt Berger12 Comments

An update on Emerson College polling

In two essays I published in November 2019 (here and here), I addressed differences between polling results obtained by highly-respected Emerson College and those from other pollsters in the same race. Emerson College has an A- rating from FiveThirtyEight.com, and their analysts calculate that, on average, Emerson College polls are about 0.3 points more Democratic than other pollsters in the same elections.

Emerson College recently released new national and state-level polls, renewing interest in their work. I therefore decided to update my assessment by comparing Emerson College polls to those released by other pollsters in the same races, using all publicly-available polls released since January 1, 2019.

I Voted sticker

I begin with national polls of the 2020 presidential election between Democrat Joseph R. Biden, Jr., the former Vice President, and Republican incumbent Donald J. Trump. The 18 polls Emerson College has released of this race, weighted by time, average to a Biden lead of 3.6 percentage points (“points”) over Trump; this increases to 3.9 if I do not adjust for Emerson’s calculated Democratic lean. By comparison, the other 370 polls of this race—using my weighted-adjusted polling average (WAPA)—give Biden a 7.4-point lead, with essentially no partisan lean. That is, relative to other polls assessing Biden vs. Trump nationally, Emerson College polls are 3.5-3.8 points more Republican.

Table 1 repeats this comparison for every state in which Emerson College has polled the presidential election at least twice since January 1, 2019; the month of the most recent Emerson College poll is also listed. States are sorted by number of Emerson College polls and poll recency.

Table 1: Comparing state-level WAPA, Biden vs. Trump, Emerson College to all other pollsters

State # Emerson Polls Emerson College All Other Pollsters Emerson minus Other
Iowa 4

(12/2019)

 -1.8 -2.2 D+0.4
Texas 3

(5/2020)

 -1.4 -2.1 D+0.7
New Hampshire 3

(11/2019)

 6.7 4.7 D+2.0
Massachusetts 2

(5/2020)

 34.4 30.6 D+3.8
Ohio 2

(5/2020)

 -0.4 1.0 D-1.4
California 2

(5/2020)

 29.3 27.6 D+1.9
Michigan 2

(11/2019)

 11.0 7.0 D+4.0
Nevada 2

(11/2019)

 -0.4 +4.6 D-4.2
TOTAL/AVERAGE 20 9.7 8.9 D+0.8
Weighted by # polls 20 7.8 7.0 D+0.8

These eight states tell a very different story. On average, these polls show an average 0.8 points more support for Biden than all other polls in these states, irrespective of the number of polls Emerson College has conducted in that state. One major caveat is that Emerson College has yet to conduct a presidential horse race poll in calendar year 2020 in four of these states. If anything, though, state-level Emerson College polls of Biden vs. Trump have become more Democratic leaning: the four states with no calendar year 2020 polls show a Democratic lean of 0.45, broadly similar to the FiveThirtyEight estimate of 0.3, compared to 1.25 for the four states with a May 2020 Emerson College poll. Remove Massachusetts (home to Emerson College), however, and the difference vanishes.

I will post my assessment of polling in 2020 election for United States Senate (“Senate”) and governor around Labor Day; this assessment will only use polls released since January 1, 2020. At this point, though, Emerson College has released only two polls, in total, of any of this year’s 35 Senate and 11 gubernatorial elections. Both were conducted in Montana with 531 likely voters between July 31 and August 2, 2020. In the Senate race, Democratic Governor Steve Bullock trails incumbent Republican Steve Daines 44-50. The other six polls of this election give Bullock a narrow 0.9-point lead, for a pro-Republican lean of 6.9 points. Similarly, in the open governor’s race, Democratic Lieutenant Governor Mike Cooney trails Republican United States House of Representatives member Greg Gianforte 41 to 50; three other polls have Cooney down by 5.2 points, for a pro-Republican lean of 3.8 points. With all necessary caveats for small numbers, the average pro-Republican bias in these two polls is 5.4.

Curiously, while Emerson College presidential election polls force respondents to choose only between Biden and Trump—meaning their vote shares sum to 100%–in the two recent Montana polls, an average 7.5% of respondents chose an option besides the named Democrat and named Republican.

Still, this does not explain why, at the national level, Emerson College 2020 presidential election polls lean 3.5-3.8 points more Republican, while the sparse state-level polling leans 0.8 points more Democratic. Splitting the difference implies an Emerson College pro-Republican lean of about 1.4 points, but I would like to see far more state-level polling to have any confidence in that value.

Until next time…please stay safe and healthy…

Matt Berger7 Comments

Biden vs. Trump: The view from three months out

On November 3, 2020, the presidential election between incumbent Republican Donald J. Trump and his Democratic challenger, former Vice President Joseph R. Biden, Jr., will mark the end of a weeks-long electoral process. Three months ago, I analyzed all publicly-available polls of the presidential election—nationally and by state, recognizing presidential elections are determined by the Electoral College—conducted since January 1, 2019.

Now the elections end in three months. As we wait for Biden to announce his vice-presidential running mate and the start of the Democratic National Convention on August 17, here is an update.

Table 1 lists the number of national polls assessing Biden vs. Trump conducted in 2019 and in each month in 2020. Sixty pollsters, with an average FiveThirtyEight pollster rating of B-, have assessed the 2020 presidential election at least once since January 1, 2019; only 38 have assessed the election more than once (mean B-).

Table 1: Number of 2020 Monthly National Polls Assessing Biden vs. Trump

Month Biden
All of 2019 106
January 2020 20
February 2020 23
March 2020 34
April 2020 49
May 2020 48
June 2020 62
July 2020 45
TOTAL 387

Just eight pollsters (average pollster rating: B-) account for 53% of these polls, as well as 51% of the 281 polls conducted in 2020:

  • YouGov (B-), 55 polls (40 in 2020)
  • Morning Consult (B/C), 36 polls (31 in 2020)
  • Ipsos (B-), 30 polls (23 in 2020)
  • HarrisX (C), 21 polls (12 in 2020)
  • Emerson College (B+), 18 polls (6 in 2020)
  • Fox News (A-), 16 polls (7 in 2020)
  • Optimus (B/C), 14 polls (13 in 2020)
  • Change Research (C-), 14 polls (11 in 2020)

Figure 1 shows how Biden has fared monthly against Trump in 2020, using my weighted-adjusted polling averages (WAPA). I use pollster rating data to adjust for partisan lean (tendency to err more Democratic or Republican than other pollsters in analogous races) and quality. I weight more recent polls higher, using this ratio: number of days since January 1, 2019 divided by 673, the number of days between January 1, 2019 and November 3, 2020. I then average two versions of WAPA: one treating polls by the same pollster as statistically independent, and one treating all polls by the same pollster as a single, time-weighted value; differences between estimates are negligible.

Figure 1: 2020 Monthly weighted-adjusted average margins for Biden vs. Trump

Biden v Trump since Jan 2020

Using all polls conducted since January 1, 2019, Biden leads Trump nationally by 7.2 percentage points (“points”). Biden’s margin rose from just over four percentage points in January and February, when he was fighting for the 2020 Democratic presidential nomination, to just under six points in the three months after Biden’s decisive win in the 2020 South Carolina Democratic presidential primary, to nearly nine points in June and July. The latter averages track closer to the FiveThirtyEight national polling average.

**********

Again, though, presidential elections are fought across 50 states and the District of Columbia (“DC”), with the plurality winner in each state/DC winning every electoral vote (“EV”) from that state. Table 2 lists the number of polls within each state assessing Biden vs. Trump since January 1, 2019, plus that state’s 3W-RDM, my estimate of much more or less Democratic than the nation a state tends to vote; 10 states[1] and DC have not been polled.

Table 2: Number of state-level polls assessing Biden vs. Trump since January 1, 2019

State 3W-RDM Overall 2020
Michigan 2.2 61 47
North Carolina -6.0 52 39
Wisconsin 0.7 49 36
Pennsylvania -0.4 45 35
Florida -3.4 43 35
Arizona -9.7 41 33
Texas -15.3 41 26
Georgia -9.6 24 19
California 23.2 17 11
Ohio -5.8 14 10
Iowa -4.7 16 9
New Hampshire 0.1 14 9
Colorado 2.2 10 8
Virginia 1.5 11 7
Minnesota 1.5 8 7
Montana -18.6 8 7
South Carolina -15.7 8 6
Missouri -15.9 8 6
New York 21.6 6 6
Washington 12.1 7 5
Kentucky -28.7 7 5
Utah -33.1 6 5
New Jersey 12.0 5 5
Maine 5.9 7 4
Alabama -28.4 4 4
Nevada 2.0 7 3
Connecticut 12.8 5 3
Massachusetts 22.1 4 3
Alaska -19.2 4 3
New Mexico 6.5 3 3
Mississippi -18.5 3 3
Kansas -23.4 3 3
Tennessee -25.8 3 3
Oklahoma -38.1 3 3
North Dakota -29.4 4 2
Maryland 22.6 2 2
Indiana -16.3 2 2
Delaware 12.5 1 1
Arkansas -28.2 1 1
West Virginia -35.5 1 1
TOTAL D-6.7 556 418

Fourteen states have been polled at least 10 times since January 1, 2019, of which 10 have been polled at least 10 times in 2020. Four of the top five, along with suddenly-swing North Carolina, are the closest states won by Trump in 2016: Michigan, Wisconsin, Pennsylvania, Florida. Four other Republican-leaning states have been frequently polled: Arizona, Georgia, Texas and Ohio, reflecting their status as ongoing or emerging battlegrounds. California, with 54 EV, rounds out the top 10.

National averages still matter, though. Combined with 3W-RDM, they provide the “expected Democratic-minus-Republican margin” in each state in 2020, all else being equal. Comparing polling averages to this expected value tells us where Biden may currently be under- or over-performing.

For example, Biden currently leads Trump nationally by 7.2 points. North Carolina has recently been 6.0 points less Democratic than the nation as a whole. Adding those two values together (7.2 – 6.0 = +1.2) suggests Biden could easily win North Carolina in 2020. Indeed, Biden leads Trump by an adjusted mean of 1.9 points in 52 polls conducted in North Carolina, implying Biden is “outperforming” expectations there by about 0.7 points. Table 3 lists every state’s expected value and WAPA.

Table 3: Expected and actual polling margins for Biden over Trump in each state in November 2020

State 3W-RDM Expected WAPA WAPA-Expected
DC 82.0 89.2    
Hawaii 34.3 41.5    
Vermont 27.7 34.9    
California 23.2 30.3 27.5 -2.8
Maryland 22.6 29.8 24.6 -5.2
Massachusetts 22.1 29.3 32.1 2.8
New York 21.6 28.8 24.3 -4.5
Rhode Island 18.0 25.2    
Illinois 14.7 21.9    
Connecticut 12.8 20.0 16.8 -3.2
Delaware 12.5 19.7 16.3 -3.4
Washington 12.1 19.3 24.6 5.4
New Jersey 12.0 19.1 18.1 -1.1
Oregon 8.7 15.9    
New Mexico 6.5 13.7 11.3 -2.4
Maine 5.9 13.1 10.2 -2.9
Michigan 2.2 9.4 7.0 -2.4
Colorado 2.2 9.4 12.4 3.1
Nevada 2.0 9.2 3.5 -5.7
Minnesota 1.5 8.7 8.8 0.1
Virginia 1.5 8.6 9.1 0.5
Wisconsin 0.7 7.9 3.9 -3.9
New Hampshire 0.1 7.3 5.0 -2.3
Pennsylvania -0.4 6.8 5.2 -1.6
Florida -3.4 3.8 4.0 0.2
Iowa -4.7 2.5 -2.0 -4.5
Ohio -5.8 1.4 0.6 -0.8
North Carolina -6.0 1.2 1.9 0.7
Georgia -9.6 -2.4 -0.6 1.8
Arizona -9.7 -2.5 2.7 5.3
Texas -15.3 -8.1 -2.1 6.0
South Carolina -15.7 -8.5 -8.1 0.3
Missouri -15.9 -8.8 -5.7 3.1
Indiana -16.3 -9.1 -11.6 -2.5
Mississippi -18.5 -11.3 -12.0 -0.7
Montana -18.6 -11.4 -9.2 2.2
Alaska -19.2 -12.0 -4.3 7.6
Louisiana -22.2 -15.0    
Kansas -23.4 -16.3 -11.4 4.8
Nebraska -25.8 -18.6    
South Dakota -25.8 -18.7    
Tennessee -25.8 -18.7 -14.5 4.2
Arkansas -28.2 -21.0 -3.5 17.5
Alabama -28.4 -21.2 -17.0 4.2
Kentucky -28.7 -21.5 -18.2 3.3
North Dakota -29.4 -22.2 -20.4 1.8
Utah -33.1 -25.9 -12.0 13.9
Idaho -34.2 -27.0    
West Virginia -35.5 -28.3 -34.3 -6.0
Oklahoma -38.1 -31.0 -23.7 7.2
Wyoming -45.7 -38.5    
Average D-6.4 Biden+0.5* Biden+1.5 +1.0

        * Only for the 40 states with both measures

The correlation between the expected margin and WAPA is +0.993, meaning polling matches expectations nearly perfectly—as one increases or decreases, so does the other. Still, Biden is polling ahead of those fundamentals by an average of about one percentage point, meaning state-level polling as a whole is even better for Biden than his excellent national polling; that said, the difference vanishes once you adjust for a state’s 2016 presidential election turnout.

Biden is underperforming expectations in some states, most notably woefully-under-polled Nevada. Biden leads there by 3.5 points, nearly six points lower than the 9.2 points by which he “should” be leading. Biden is also underperforming expectations in Iowa (-4.5) and Wisconsin (-3.9). By the same token, Biden is strongly overperforming in the traditionally Republican states of Arkansas, Utah, Oklahoma, Alaska, Texas and Arizona, as well as in reliably-Democratic Washington; the first four have only been polled 14 times in total, however. There is a partisan split in Biden’s over-and under-performance: in states with 3W-RDM>-5.0, Biden is underperforming by 1.7 points, on average. In states with 3W-RDM≤5.0, Biden is overperforming by 3.7 points. Many grains of salt are in order here, though. In recent elections, “fundamentals” have missed the final margin by an absolute value average of 5.4 points.

Still, the near-perfect correlation between the two values allows us to combine them into a single estimate of Biden’s margin over Trump on November 3, 2020, assuming polls become more predictive as an election gets closer:

  1. Assign expected value and WAPA equal weight as of January 1, 2020.
  2. WAPA weight increases, by day, with proximity to November 3, 2020.

I also calculated how likely Biden is to win the EV from each state, assuming this likelihood is distributed normally:

  1. For expected margins, I use mean = -0.8 and standard error = 7.1[2]
  2. For WAPA, I use standard error = 3.0, roughly the margin of error in most quality polls; this is likely an over-estimate, as pooling reduces the standard error of the resulting polling average.
  3. Combined probability Biden wins a state’s EV calculated the same as for predicted final margin

While the means and standard errors I use are arguably arbitrary, albeit defensible, the final EV probabilities shown in Table 4 are in line with what other forecasters are saying.

Table 4: Estimated final state margins and probability of winning EV, Biden vs. Trump, November 2020

State EV P(EV): Expected P(EV):

WAPA

 P(EV):

Overall

 Predicted Margin
DC 3 100.0%   100.0% 89.2
Hawaii 4 100.0%   100.0% 41.5
Vermont 3 100.0%   100.0% 34.9
California 55 100.0% 100.0% 100.0% 28.3
Maryland 10 100.0% 100.0% 100.0% 26.0
Massachusetts 11 100.0% 100.0% 100.0% 31.6
New York 29 100.0% 100.0% 100.0% 25.2
Rhode Island 4 100.0%   100.0% 25.2
Illinois 20 99.9%   99.9% 21.9
Connecticut 7 99.7% 100.0% 99.9% 18.4
Delaware 3 99.6% 100.0% 99.8% 18.0
Washington 12 99.5% 100.0% 99.9% 22.0
New Jersey 14 99.5% 100.0% 99.9% 18.6
Oregon 7 98.3%   98.3% 15.9
New Mexico 5 96.5% 100.0% 99.2% 11.8
Maine 4 95.8% 100.0% 99.3% 10.7
Michigan 16 88.8% 99.1% 97.4% 7.4
Colorado 9 88.6% 100.0% 98.1% 11.9
Nevada 6 88.0% 87.9% 87.9% 5.2
Minnesota 10 86.6% 99.8% 97.6% 8.8
Virginia 13 86.5% 99.9% 97.6% 9.1
Wisconsin 10 84.1% 90.5% 89.5% 4.6
New Hampshire 4 81.9% 95.2% 92.9% 5.4
Pennsylvania 20 80.1% 95.8% 93.2% 5.4
Florida 29 66.4% 90.9% 86.9% 4.0
Iowa 6 59.2% 24.8% 30.5% -1.3
Ohio 18 53.1% 57.7% 56.9% 0.7
North Carolina 15 52.1% 73.4% 70.7% 1.8
Georgia 16 32.7% 41.8% 41.2% -0.8
Arizona 11 31.9% 81.8% 73.6% 1.9
Texas 38 10.6% 24.7% 22.3% -3.1
South Carolina 9 9.6% 0.3% 1.9% -8.2
Missouri 10 8.9% 2.9% 4.2% -6.3
Indiana 11 8.2% 0.0% 2.2% -11.0
Mississippi 6 4.4% 0.0% 1.1% -11.8
Montana 3 4.3% 0.1% 0.8% -9.6
Alaska 3 3.6% 7.4% 6.8% -5.6
Louisiana 8 1.3%   1.3% -15.0
Kansas 6 0.8% 0.0% 0.2% -13.8
Nebraska 5 0.3%   0.3% -18.6
South Dakota 3 0.3%   0.3% -18.7
Tennessee 11 0.3% 0.0% 0.1% -16.6
Arkansas 6 0.1% 12.2% 9.3% -12.3
Alabama 9 0.1% 0.0% 0.0% -19.1
Kentucky 8 0.1% 0.0% 0.1% -21.2
North Dakota 3 0.1% 0.0% 0.0% -21.1
Utah 6 0.0% 0.0% 0.0% -16.0
Idaho 4 0.0%   0.0% -27.0
West Virginia 5 0.0% 0.0% 0.0% -31.4
Oklahoma 7 0.0% 0.0% 0.0% -27.3
Wyoming 3 0.0% 100.0% 0.0% -38.5

Three months before Election Day 2020, and with every caveat about voting during a pandemic, Joe Biden is the prohibitive favorite to be elected the 46th president of the United States.

  • He is at least an 86.9% favorite in enough states—and by margins of at least four points—to earn him 308 EV, or 307 depending on what happens in Maine, which, along with Nebraska, allocates two EV to the statewide winner and one each to the winner of its Congressional districts. Moreover, Biden could lose Florida (+4.0, 86.9%), Nevada (+5.2, 87.9%) and one EV in Maine and still win 272 EV, two more than he needs
  • He is a 70-75% favorite to win in Arizona (+1.9) and North Carolina (+1.8), for an additional 26 EV, increasing Biden’s total to 333/334 EV.
  • The 34 combined EV of Ohio (+0.7) and Georgia (-0.8) are essentially toss-ups, meaning Biden has a roughly 75% chance to win at least one of them, putting him somewhere between 349 and 352 EV, with a maximum of 368 EV (or 369 with one EV in Nebraska).

Plus, it might take only a sharp break by undecided voters and a modest polling error for Biden to win the 44 combined EV of Iowa (-1.3) and the ultimate prize—Texas (-3.1). Thus, while something in the low-to-mid 300’s appears the most likely EV total for Biden, 413 EV cannot be discounted.

Using the simplistic—perhaps even simple-minded—method of multiplying Biden’s probability of winning each state by its EV and summing yields a “projected” EV total of 347.1, essentially adding Michigan, Pennsylvania, Wisconsin, Florida, Arizona, North Carolina, and one of Ohio/Georgia to the states 2016 Democratic presidential nominee Hillary Clinton won.

Biden’s lead looks even more robust when you make either of two historically-valid assumptions:

Polls systematically overestimate Biden’s margins by 3.0 points.

In this scenario, Biden’s projected EV drops to 299.7, which is still 29.7 more than required. He would be favored at least 80% to win in enough states to win 239 EV, though he would also be favored by at least 63.8% in three states totaling 34 EV, putting him over the top. Thus, even if Biden “only” wins the national popular vote by 4.2, he would likely still prevail.

Polls systematically underestimate Biden’s margins by 3.0 points.

In this scenario, Biden’s projected EV are a landslide-level 387.0, 117 more than necessary. He would be favored at least 80% to win enough states to earn 352 EV, while being a 74.6% favorite in Georgia, for a total of 368 EV. He would also be a 64.6% favorite in Iowa, with Texas essentially a toss-up at 55.3%–and a projected Trump victory of just 0.1%! Based on only one poll, Biden would even have a 33.2% chance of winning Arkansas’ 6 EV, with an 18.4% chance of winning Missouri’s 10 EV and a 10.9% chance of winning Alaska’s 3 EV. The last presidential candidate to come close to 433 EV was Republican George H. W. Bush, who won 426 EV in 1988.

**********

None of this is to say Biden is guaranteed to be the next president of the United States; even three months from Election Day, it would be monumentally foolish for me to conclude that. Indeed, there are worrisome signs this year’s elections will not be conducted as efficiently and transparently as they could be. In an election in which a record number of voters are expected to cast their ballots by mail, delays in mail delivery—allegedly orchestrated by a newly-confirmed Postmaster General—could leave millions of votes uncounted because they did not arrive by November 3. Moreover, while Biden’s national polling lead has consistently ranged between four and 10 points over the last 19 months, a late-recovering economy or a last-minute “October surprise” could upend that trajectory.

All that being said, however, unlike Clinton in 2016, Biden has a sufficiently-wide range of paths to 270 EV that I estimate he is at least a 90% favorite to be elected president of the United States on November 3, 2020—or whenever ballots are ultimately counted.

Until next time…please stay safe and healthy…

[1] Hawaii, Vermont, Rhode Island, Illinois, Oregon, Louisiana, Nebraska, South Dakota, Idaho, Wyoming

[2] The former value is the mean arithmetic difference between “expected” and actual D-R margins across 153 state-level contests in 2008, 2012 and 2016, while the latter value is the standard deviation of these values. I recognize this is not a standard error. However, using the value 13.6—the range of values covering 95% of all values divided by 1.96, the final EV projection changes by only 1.0.

Matt Berger6 Comments