Oddities in the high level statistical analysis of the 2020 U.S. presidential elections

There are indeed some significant oddities that make you wonder about election integrity and whether Joseph Biden indeed won fair and square! It is strongly advisable to fully investigate the November 2020 elections!

I am not sure that the impromptu and massive switch to mail-in voting due to Covid-19 excuse can easily explain all these oddities. Or was it the Covid-19 pandemic. Or were there so many GOP voters frustrated with President Trump that they split their votes ...

The only other period since 1960, where we observe comparable dramatic voting results were the 1964 (assassination of John F. Kennedy), 1968 (Vietnam war, assassination of Martin Luther King Jr.), and the 1972 election (Nixon, who faced a then quite extreme and negative news suffering  Democratic opponent whose name barely anyone remembers).

My limited scope high level statistical analysis is based on following data:

1. Total votes for the entire country for the Republican and the Democratic candidate
2. Voting age population as a rough proxy for the total number of eligible voters (As a footnote: actual data of eligible voters (citizens) for the entire U.S. is hard to come by, which surprised me)
3. All presidential elections since 1960 or the 16 most recent elections were included in the study

Some general facts about the 2020 U.S. general elections:

1. The voting age population turnout rate was very high at 61%. Very high in historical context. Only the three 1960s elections reached >= 60%. For comparison, the Obama 2008 election achieved just under 60% turnout
2. Both presidential candidates had excellent results in 2020. Trump achieved the 4th best voting result relative to voting age population of any Republican candidate since 1960 with 29.1%. However, for Biden it was the 2nd best for any Democratic candidate since 1960 with 31.8% (Johnson with 37.9% was the best in 1964, third best was Kennedy with 31.6% in 1960).
3. About 26.5 million more votes were cast in 2020 compared to 2016. That is equivalent to about 10.4% of voting age population. This is by far the largest difference between two elections ever observed since 1960. The 2nd largest change was only 8.9% in 2004. The average is 4.2% since 1960.
4. Biden won with about 81.3 million votes total and by about 15.4 million more votes (or about 6% of the voting age population) than the winner by popular vote of the 2016 election. This is another historically astonishing result. The last time a candidate reached 6% or higher was Reagan's reelection in 1984 with 6.1%. The average over the 16 elections since 1960 is only 4.3%. There were only three other elections over 5%, i.e. 1964 with 7.8%, 1968 with 9.4% and 1972 with 11%.

Further analysis shows:

1. Comparing the ranking of both candidates by political party in terms of total votes received relative to voting age population, one discovers that they were only 2 apart (Trump's 4th best result compared to Biden's 2nd best result, see above). When one of the candidates has an exceptional voting result, the other one usually does not. When you look at the four best results for any candidate since 1960, then you get in 1960 a difference of 1 (i.e. Kennedy ranked 3rd, Nixon ranked 2nd), in 1964 the difference was 13 (Johnson ranked 1st, Goldwater ranked 14th), in 1972 it was 15 (Nixon ranked 1st, McGovern ranked 14th), in 1984 it was 12 (Reagan 3rd, Mondale 15th) and in 2008 it was 6 (Obama 4th, McCain 10th)
2. Biden won 15,4 million more votes than Hillary Clinton in 2016 equivalent to about 6% of voting age population. Only two other Democratic candidates since 1960 pulled off a better result: Johnson (1960 with 7.8%) and Carter (1976 with 7.8%). 

I intend to add a spreadsheet later to document and in support of my statistical analysis. Here it is a link to my table:

U.S. Presidential Election Results - Blog.pdf