Replies

For your consideration...

For purposes of this post (and for the rest of my life), I have put an asterisk next to Biden's "election." In fact I may simply call him President Asterisk from here on out.

The national popular vote is not a legit, Constitutionally-loyal way of determining how "close" is an election. Focusing on the national popular vote is akin to saying that the Yankees won the 1960 World Series because they outscored the Pirates 55 runs to 27. While that's true, the the Pirates won the World Series because they beat the Yankees four times out of seven games.

So, fed-up, I built a couple of objective, accurate, and Constitutionally-loyal ways of assessing "closeness."

Bottom line: Biden's win* is hardly a mandate; it is amongst the six closest across the 39 Presidential elections we've had since the end of the Civil War. Biden's election* was about as close as Kennedy's 1960 election (where there were rumors of fraud as well), and in close proximity to that of the tumultous 1876 Hayes election that brought about the Compromise of 1877, as well as the 1880 Garfield election which featured irregularities.

By contrast, Trump's 2016 win was less close than Biden's win*. Finally, all elections since 1980 have been more close than average except for Reagan's win in 1984 and, to a much lesser extent, Clinton's win in 1996.

How does this work? I chose to focus on the margin (i.e., difference between winner and first runner-up) in the state races that were closest through that from the "tipping-point state."

A tipping-point state is the state that pushes the candidate over the winning EC vote total WHEN all states are sorted from lowest to highest winning vote margin.

For example, in 2012, Obama secured 332 EVs over Romney's 206. This EV total is 23% over the 270 needed to win. That feels like a wide margin of victory.

But, that doesn't tell you how close was the election. What if, for example, Obama won those close state races by a margin of one vote? Therefore, if we sort Obama's state-level winning vote percentage margin (i.e., Obama's state-level vote share minus Romney's state-level vote share), from lowest to highest, we can get a better feel for the "closeness."

Here are the actual data:

State	Obama winning margin (votes)	Obama winning margin (% diff from Romney)	Electoral Votes	Obama cumulative EVs with flipping
Florida	74,309	0.88%	29	303
Ohio	166,272	2.98%	18	285
Virginia	149,298	3.87%	13	272
Colorado (tipping point state)	137,858	5.37%	9	263

We now see that Obama beat Romney in Florida by only 0.88% of the Florida votes, followed 2.98% of the votes in Ohio, and 3.87% of the votes in Virginia. Had Romney flipped those three states, Obama would still have 272 EVs and the win.

But, if Romney also flipped the "tipping-point state" of Colorado where the margin was 5.37%, Obama would have had 263 EVs and the Presidency would be Romney's (he'd have 275 EVs).

Across these four states, Obama's vote margin was 527,737 votes, which is 2.577% of those states' aggregate votes and 0.409% of the national vote total of 129,085,410.

Now, we are getting somewhere. Those percentages are lower than the 3.86% national vote difference, and much lower than the 23% of excess Electoral Votes that Obama got over the 270 needed to win. Perhaps more importantly, they reflect the Constitution and the Electoral College (e.g., we're not counting runs scored in the World Series but how close were the games).

I believe these two measures - the cumulative winning vote difference in close states through the tipping-point state relative to (1) those respective state vote totals and (2) national vote total - are more accurate, reflect how our Constitution work, AND can be compared across all elections, regardless of the number of citizens voting and EVs. Finally, they are based on facts, and not some 'journalist' cherry-picking a cutoff for including this or that state in his/her analysis.

I went to Wikipedia and pulled election vote data from 2020 to 1868 (I cut if off at this point, because from 1864 backwards, southern state non-participation and other issues make the analysis a little difficult) and sorted the state-level winning vote percentage difference between the winner and runner-up from lowest to highest, and summed the winning vote margins from the closest state up to and including the tipping-point state. I divided that aggregate margin by State-level total votes for the closest through tipping point states, and aggregate national vote totals. Both metrics have advantages - the pure State Closeness Metric (SCM) is agnostic toward other states outside of this analysis, while the closeness metric based on the national vote (NCM) is a general barometer of the close states relative to the entire race.

I also took the national popular vote difference between the winner and first runner up and the Electoral Vote margin (excess electoral votes divided by the amount needed to win, which varies across time) for comparison purposes.

Here are the data:

Election & Winner	Popular Vote Margin %	NCM: Sum of Tipping Point Popular Margins to Total Popular Vote %	SCM: Sum of Tipping Point Popular Margins to Total Tipping Point Popular Vote %	EV Margin/Total Evs %
2020 Biden*	4.45%	0.027%	0.367%	6.691%
2016 Trump	−2.10%	0.057%	0.558%	6.320%
2012 Obama	3.86%	0.409%	2.577%	11.524%
2008 Obama	7.27%	0.757%	3.607%	17.658%
2004 Dubya	2.46%	0.110%	1.706%	2.974%
2000 Dubya	−0.52%	0.001%	0.009%	0.186%
1996 Clinton	8.52%	1.442%	5.886%	20.260%
1992 Clinton	5.56%	0.588%	2.864%	18.587%
1988 GHW Bush	7.73%	1.345%	4.010%	28.996%
1984 Reagan	18.22%	6.419%	12.230%	47.398%
1980 Reagan	9.74%	1.856%	4.273%	40.706%
1976 Carter	2.06%	0.057%	0.746%	5.019%
1972 Nixon	23.15%	8.763%	16.102%	46.468%
1968 Nixon	0.70%	0.235%	1.992%	5.762%
1964 Johnson	22.58%	6.587%	15.777%	40.149%
1960 Kennedy	0.16%	0.028%	0.276%	6.134%
1956 Eisenhower	15.40%	4.233%	10.857%	35.970%
1952 Eisenhower	10.85%	3.071%	7.787%	33.145%
1948 Truman	4.48%	0.051%	0.359%	6.968%
1944 FDR	7.49%	1.501%	3.363%	31.262%
1940 FDR	9.95%	2.296%	4.282%	34.463%
1936 FDR	24.25%	10.676%	16.801%	48.399%
1932 FDR	17.76%	5.041%	9.653%	38.795%
1928 Hoover	17.42%	2.860%	8.070%	33.522%
1924 Coolidge	25.22%	2.357%	9.601%	21.846%
1920 Harding	26.17%	4.989%	16.064%	21.846%
1916 Wilson	3.12%	0.151%	0.352%	2.072%
1912 Wilson	14.44%	2.772%	6.755%	31.827%
1908 Taft	8.53%	1.007%	4.227%	16.356%
1904 T. Roosevelt	18.83%	3.297%	10.803%	20.378%
1900 McKinley	6.12%	1.698%	6.273%	15.213%
1896 McKinley	4.31%	0.509%	2.844%	10.515%
1892 Grover Cleveland	3.01%	0.374%	1.984%	12.162%
1888 B. Harrison	-0.83%	0.147%	0.901%	7.980%
1884 Grover Cleveland	0.57%	0.011%	0.098%	9.214%
1880 Garfield	0.11%	0.307%	1.754%	7.859%
1876 Hayes	-3.00%	0.011%	0.487%	0.000%
1872 Grant	11.80%	2.167%	5.952%	30.966%
1868 Grant	5.32%	1.015%	3.776%	22.449%
*Average*	*9.40%*	*2.031%*	*5.283%*	*20.462%*
*Median*	*7.49%*	*1.015%*	*3.776%*	*18.587%*
*Avg 1980-2020*	*7.53%*	*1.183%*	*3.462%*	*18.300%*
*Median 1980-2020*	*7.27%*	*0.588%*	*2.864%*	*17.658%*

Here is the data sorted by NCM and SCM:

Election & Winner	NCM: sorted lowest to highest	Election & Winner	SCM: sorted lowest to highest
2000 Dubya	0.001%	2000 Dubya	0.009%
1876 Hayes	0.011%	1884 Grover Cleveland	0.098%
1884 Grover Cleveland	0.011%	1960 Kennedy	0.276%
2020 Biden*	0.027%	1916 Wilson	0.352%
1960 Kennedy	0.028%	1948 Truman	0.359%
1948 Truman	0.051%	2020 Biden*	0.367%
1976 Carter	0.057%	1876 Hayes	0.487%
2016 Trump	0.057%	2016 Trump	0.558%
2004 Dubya	0.110%	1976 Carter	0.746%
1888 B. Harrison	0.147%	1888 B. Harrison	0.901%
1916 Wilson	0.151%	2004 Dubya	1.706%
1968 Nixon	0.235%	1880 Garfield	1.754%
1880 Garfield	0.307%	1892 Grover Cleveland	1.984%
1892 Grover Cleveland	0.374%	1968 Nixon	1.992%
2012 Obama	0.409%	2012 Obama	2.577%
1896 McKinley	0.509%	1896 McKinley	2.844%
1992 Clinton	0.588%	1992 Clinton	2.864%
2008 Obama	0.757%	1944 FDR	3.363%
1908 Taft	1.007%	2008 Obama	3.607%
1868 Grant	1.015%	1868 Grant	3.776%
1988 GHW Bush	1.345%	1988 GHW Bush	4.010%
1996 Clinton	1.442%	1908 Taft	4.227%
1944 FDR	1.501%	1980 Reagan	4.273%
1900 McKinley	1.698%	1940 FDR	4.282%
1980 Reagan	1.856%	1996 Clinton	5.886%
1872 Grant	2.167%	1872 Grant	5.952%
1940 FDR	2.296%	1900 McKinley	6.273%
1924 Coolidge	2.357%	1912 Wilson	6.755%
1912 Wilson	2.772%	1952 Eisenhower	7.787%
1928 Hoover	2.860%	1928 Hoover	8.070%
1952 Eisenhower	3.071%	1924 Coolidge	9.601%
1904 T. Roosevelt	3.297%	1932 FDR	9.653%
1956 Eisenhower	4.233%	1904 T. Roosevelt	10.803%
1920 Harding	4.989%	1956 Eisenhower	10.857%
1932 FDR	5.041%	1984 Reagan	12.230%
1984 Reagan	6.419%	1964 Johnson	15.777%
1964 Johnson	6.587%	1920 Harding	16.064%
1972 Nixon	8.763%	1972 Nixon	16.102%
1936 FDR	10.676%	1936 FDR	16.801%
*Average*	*2.031%*	*Average*	*5.283%*
*Median*	*1.015%*	*Median*	*3.776%*
*Avg 1980-2020*	*1.183%*	*Avg 1980-2020*	*3.462%*
*Median 1980-2020*	*0.588%*	*Median 1980-2020*	*2.864%*

I know there is a LOT to unpack here, so I'll be brief:

1. As noted earlier, Biden's election* is far from a mandate: it is the fourth tightest using the SCM and sixth tightest using the LCM. The only contest tighter than Biden's since 1980 was Dubya's 2000 election.

2. While Trump's 2016 election was also tight, historically it is line with Truman in 1948 and Carter in 1976 the SCM. On a NCM basis, Trump's election is comparable in closeness to Biden's contest* but is not as close.

3. At the other end of the scale, Nixon's 1972 election is the second-widest blowout using both metrics, even bigger than Reagan's 1984 blowout.

4. About two-thirds (Seven out of eleven) of the elections from 1980-2020 are in the closet half of all elections, with only 1988, 1996, 1980 and 1984 landing in the widest half.

5. Obama and Wilson are the only re-elected presidents in this analysis where their re-election was closer than the initial election. FDR's first re-election was a blowout, but this third and fourth re-elections were sequentially closer.

Comments, corrections, and critiques are welcome. Thank you.