Biden's Electoral College win was narrow in the tipping-point state; Labor surges in Victoria

With all states called by US media^[1], Joe Biden won the Electoral College by 306 votes to 232 over Donald Trump, an exact reversal of Trump’s triumph in 2016, ignoring faithless electors. Biden gained the Trump 2016 states of Michigan, Pennsylvania, Wisconsin, Arizona and Georgia; he also gained Nebraska’s second district.

While Biden’s win appears decisive, he won three states – Wisconsin (ten Electoral Votes), Arizona (11) and Georgia (16) – by 0.6% or less. Had Trump won these three states, the Electoral College would have been tied at 269-269.

If nobody wins a majority (270) of the Electoral College, the presidency is decided by the House of Representatives^[2], but with each state’s delegation casting one vote. Republicans hold a majority of state delegations, so Trump would have won a tied Electoral College vote.

Wisconsin (Biden by 0.6%) will be the “tipping-point” state. Had Trump won Wisconsin and states Biden won by less (Arizona and Georgia went to Biden by 0.3% margins), he would have won the Electoral College tiebreaker.

The national popular vote^[3] has Biden currently leading Trump by 50.9% to 47.3%, a 3.6% margin for Biden. This does not yet include mail ballots from New York that are expected to be very pro-Biden.

Biden is likely to win the popular vote by 4-5%, so the difference between Wisconsin and the overall popular vote will be 3.5% to 4.5%. That is greater than the 2.9% gap between the tipping-point state and the popular vote in 2016.

Prior to 2016, there had not been such a large gap, but in both 2016 and 2020 Trump exploited the relatively large population of non-University educated whites in presidential swing states compared to nationally.

Read more: US 2016 election final results: how Trump won^[4]

This article, written after the US 2016 election, has had a massive surge in views recently.

Relative to expectations, Democrats performed badly in Congress. In the Senate, Republicans lead by 50-48 with two Georgian runoffs pending on January 5. In the House, Democrats hold a 218-203 lead with 14 races uncalled. Republicans have gained a net seven seats so far, and lead in ten of the uncalled races.

Before the election, Democrats were expected^[5] to win the Senate and extend their House majority. The House would be worse for Democrats if not for a judicial redistribution^[6] in North Carolina that gave Democrats two extra safe seats.

US polls understated Trump again

New York Times analyst Nate Cohn^[7] has an article on the polls. Biden was expected to greatly improve on Hillary Clinton’s 2016 performance with non-University educated whites and seniors, but the results indicate that Trump held up much better with these demographics than expected.

Trump also had large swings in his favour in heavily Latino counties such as Miami Dade, Florida^[8]; polls suggested a more modest improvement for Trump with Latinos.

After the 2016 election, most polls started weighting by educational attainment, but this did not fix the problem. Cohn has some theories of what went wrong. First, Republican turnout appears to have been stronger than expected. Second, Trump’s attacks on the mainstream media may have convinced some of his supporters to not respond to polls.

A third theory is that coronavirus biased the polls’ samples, because people who followed medical advice and stayed home were more likely to respond to pollsters and more likely to be Democrats. Polls had suggested Biden would win Wisconsin, a coronavirus hotspot, easily, but he only won by 0.6%.

While US polls understated Trump in both 2016 and 2020, it is not true that international polling tends to understate the right. At the October 17 New Zealand election^[9], polls greatly understated Labour’s lead over National. Polls also understated UK Labour at the 2017 election^[10].

Victorian Labor surges after end of lockdown

In a privately conducted Victorian YouGov poll reported by The Herald Sun, Labor led by 55-45 from primary votes of 44% Labor, 40% Coalition and 11% Greens. Premier Daniel Andrews had a strong 65-32 approval rating, while Opposition Leader Michael O'Brien had a terrible 53-26 disapproval rating. The poll was conducted from late October to early November from a sample of 1,240. Figures from The Poll Bludger^[11].

A Victorian Morgan SMS^[12] poll, conducted November 9-10 from a sample of 818, gave Labor a 58.5-41.5 lead, a seven-point gain for Labor since the mid-October Morgan poll. Primary votes were 45% Labor (up five), 34.5% Coalition (down 5.5) and 11% Greens (up two). In a forced choice, Andrews had a 71-29 approval rating, up from 59-41 in mid-October. Morgan’s SMS polls have been unreliable in the past.

Labor wins Queensland election with 52 of 93 seats

At the October 31 Queensland election^[13], Labor won 52 of the 93 seats (up four since 2017), the LNP 34 (down five), Katter’s Australian Party three (steady), the Greens two (up one), One Nation one (steady) and one independent (steady). Labor has an 11-seat majority.

Primary votes were 39.6% Labor (up 4.1%), 35.9% LNP (up 2.2%), 9.5% Greens (down 0.5%), 7.1% One Nation (down 6.6%) and 2.5% Katter’s Australian Party (up 0.2%). It is likely Labor won at least 53% of the two party preferred vote. The final Newspoll^[14] gave Labor a 51.5-48.5 lead – another example of understating the left.

Labor gained five seats from the LNP, but lost Jackie Trad’s seat of South Brisbane to the Greens. Two of Labor’s gains were very close and went to recounts, with Labor winning Bundaberg^[15] by nine votes and Nicklin by 85 votes.

Federal Newspoll: 51-49 to Coalition

In last week’s federal Newspoll, conducted November 4-7 from a sample of 1,510, the Coalition had a 51-49 lead, a one point gain for Labor since the mid-October Newspoll. Primary votes were 43% Coalition (down one), 35% Labor (up one), 11% Greens (steady) and 3% One Nation (steady). Figures from The Poll Bludger^[16].

Scott Morrison is still very popular, with 64% satisfied with his performance (down one) and 32% dissatisfied (up one), for a net approval of +32. Anthony Albanese’s net approval jumped eight points to +4, but he continued to trail Morrison as better PM by 58-29 (57-28 previously).

A YouGov poll in former Labor frontbencher Joel Fitzgibbon’s seat of Hunter^[17] had a 50-50 tie; this would be a three-point swing to the Nationals from the 2019 election. Primary votes ^[18]were 34% Labor, 26% National, 12% One Nation, 10% Shooters and 8% Greens.

References

1. ^{^} US media (www.nytimes.com)
2. ^{^} decided by the House of Representatives (en.wikipedia.org)
3. ^{^} national popular vote (cookpolitical.com)
4. ^{^} US 2016 election final results: how Trump won (theconversation.com)
5. ^{^} Democrats were expected (edition.cnn.com)
6. ^{^} judicial redistribution (about.bgov.com)
7. ^{^} Nate Cohn (www.nytimes.com)
8. ^{^} Miami Dade, Florida (theconversation.com)
9. ^{^} New Zealand election (en.wikipedia.org)
10. ^{^} 2017 election (en.wikipedia.org)
11. ^{^} The Poll Bludger (www.pollbludger.net)
12. ^{^} Morgan SMS (www.roymorgan.com)
13. ^{^} Queensland election (en.wikipedia.org)
14. ^{^} final Newspoll (www.pollbludger.net)
15. ^{^} Labor winning Bundaberg (kevinbonham.blogspot.com)
16. ^{^} The Poll Bludger (www.pollbludger.net)
17. ^{^} Hunter (twitter.com)
18. ^{^} Primary votes (twitter.com)