This week part of the blog will begin a day early, so as to include Bob McMullan’s prediction for the 2020 American Presidential Election before the vote is counted in America. After the prediction post I have also added some of the material from his weekly reports on the election.
Brilliant & Bold! Bold & Brilliant!
Conversations with fabulously fascinating, always interesting, enviably engaging women … Sunday 15 November join the mayor of Broken Hill, New South Wales, a Spanish student returned to study at Anglia Ruskin, Cambridge, an Italian environmental fashion designer, and an Israeli former librarian who spends time in Italy, Israel & Australia … Plus women from all over the world – Join our global conversation … ‘Courage Speaks to Courage Everywhere … Women’s Voices Must be Heard …’
Book Reviews for the next few weeks will concentrate on Agatha Christie’s work, and related nonfiction sources. My Agatha Christie Indulgence for a couple of weeks, although possibly dated (I say possibly because her novels continue to be published) is based on the pleasure I have had in revisiting the novels through membership of the Goodreads site. I joined the group having enjoyed a trip to Greenway, Agatha Christie’s home in Devon. Years before I had been taken by friends to her home in Wallingford and grave in Cholsey. The indulgence will start with two nonfiction books, Laura Thompson – Agatha Christie: A Mysterious Life and Agatha Christie’s Complete Secret Notebooks: Stories and Secrets of Murder in the Making John Curran. Before I begin my own train travel (travel after the borders of South Australia opened, what joy!) two of Christies train novels will be added on Friday 6th November. They are Murder on the Orient Express and The Blue Train mystery.
Television Comments will continue with A Country Podcast to be added to the November 6 blog. Update…I was so busy watching the Presidential Election coverage I didn’t have time to listen to the podcast. More on both next week.
Bob McMullan, prediction for the American Presidential Election
As promised I will conclude this post with my prediction. Before I do that I will outline what I see as the range of scenarios depending on the underlying assumptions and data. Of course, there are factors an observer cannot forecast. relating to voter suppression, electoral college manipulation and Supreme Court decisions. A decisive Biden victory will make the probability of any of these affecting the choice of president by the people extremely unlikely.
Scenario A
This is what I call the mainstream position. Based on polling averages RCP or 538 would have Biden winning 336 or 351 EC votes to 202 or 187 for Trump. These projected results are based on Biden winning all the states Clinton won and adding Arizona, Florida, Georgia, Michigan, Pennsylvania and Wisconsin plus the two Congressional Districts in Maine and Nebraska. 538 would also suggest Biden winning North Carolina. These assessments give the “toss-up” states of Iowa, Ohio and Texas to Trump. These would add 62 EC votes if Biden wins them all.
Scenario B
This is the final prediction from the Virginia Center for Politics and seems a reasonable indicator of the middle-ground. Professor Sabato and his team have forecast Biden winning 321 to 217 for Trump. It differs from Scenario A in that it suggests Trump will win Florida and the Maine 2nd District. They have been consistently pessimistic about the Democrat’s capacity to actually deliver on the ground in Florida.
Scenario C
For those still scarred by 2016, this scenario is based on discounting Scenario A by the distance the state polling was out in 2016 in each state. This would result in Biden winning 335-203. The discount factor would mean Trump would win North Carolina and Maine 2nd district.
Scenario D
This scenario is based on the minority polling stream from the pollsters which have delivered results more favorable to Republicans. This would result in a Trump win or a very narrow Biden win. All pollsters agree that Biden will win Wisconsin. Some of these firms give Biden a win in Michigan or Pennsylvania or Michigan. The Trafalgar group is the outlier whose results would give Biden only Wisconsin and would result in a 242/296 win for Trump. I don’t think this is likely for the reasons I have given previously but it is unwise to ignore the results you don’t like.
Bob’s forecast
Given that the national polling gives Biden a lead greater than any Presidential candidate since 1996 and that no candidate has ever lost from a position even remotely close to Biden’s 8.6% lead I am confident that Joe Biden will win the national vote and a substantial Electoral College advantage. The hard part is predicting State by State. It is similar to picking Australian elections seat by seat rather than overall. Local variations always catch you out. However, here goes my best assessment. My best assessment is that Biden will win at least 306 Electoral College votes and possibly as many as 375. This is based on Biden winning : Wisconsin, Michigan and Pennsylvania. This is the conventional path to victory, In addition I think he will win Arizona and the Nebraska 2nd district.
My more courageous prediction is that Biden will pull off a victory in Georgia. He has the wind in his sails there. He could get to 375 by winning North Carolina, Florida, ME2, Ohio and Iowa. I have listed them in the order of what I consider to be their probability of going to Biden. I have not included Texas, which looks to me to be a bridge too far but will be worth watching on the night.
Bob McMullan, early posts American Presidential Election
16th June
According to my analysis of the polling in the US, Joe Biden is on course to win by a big margin with an outside chance of a landslide victory. I understand that many people will be dubious because of the failure of polling to predict the last presidential election and the 2019 Australian election. And I know there are still 22 weeks to go and a lot can happen in that time. But the signs are promising for Biden. This is not a prediction, just a description of the current situation. There are no states that Hillary Clinton won that Trump at this stage looks close to winning. Of the states which Trump won in 2016 Biden has been ahead in every recent poll in Wisconsin, Florida and Michigan. These states have 55 electoral college votes. If Biden won them he would have 287 electoral college votes and Trump 251. Furthermore, Real Clear Politics averages of the recent polls have Biden ahead in Arizona and Pennsylvania. These states would add 31 votes and result in Biden 318 and Trump 220.The polling results are also very close in North Carolina, Iowa, Ohio, Georgia and Texas. I have not included any of these in my estimates as Trump is ahead on the averages of the polls in all five states. I intend to update this analysis each week until November.
23 June
As promised a campaign update. A lot has happened this week, but as yet it is not showing up in any significant change in the polling data. Such data as has come out has tended to reinforce the trends I outlined last week. That is, no states which Hillary Clinton won in 2016 are in any doubt at the moment and five states which Trump won last time showing a clear (more than 4%) lead for Joe Biden. The other major source of polling data, the 538 website, is even more positive. 538 polling averages have Biden ahead in Ohio, Georgia and North Carolina as well. I will wait for more data before I include these states in my electoral college analysis. So I still have Biden on 318 electoral college votes to Trump on 220. Several people raised the important question of how this polling information compares to 2016. The average gap is comparable although Biden’s lead in key states tends to be slightly larger. The key difference is the stability of the polling. Hillary Clinton’s lead between January and June 2016 averaged 5.7% but varied from 10% to 2.5%. Biden’s average lead is 5.9% but it has never been less than 5% on average in any month this year. The most likely explanation of this relative stability is Trump’s incumbency. For better or worse most people have made up their mind about him. This is particularly so for self-described “Independent” voters. Polls show Biden leading among these voters by from 9 to 17 points. In 2016 Trump won them by 4.
1 July
Another US election update. The polls overall have moved more in favour of Joe Biden, although there are some amber lights flashing to make sure there is no complacency about the result. The only state whose status has changed over the last week has been North Carolina. The Real Clear Politics averages now show a 2.2% lead for Biden. He has been ahead in 8 of the last 10 polls in North Carolina. This state added to Biden’s column would see him win 333 votes in the electoral college to 205 for Trump. This is a state which trump won by 3.7% and Mitt Romney won against Obama in 2012. The warning signs are small at this stage but remind everyone that there is still a long time to go. For example, the pollster who got Michigan closest to right in 2016 has Biden ahead by only 1% in that state while the average of the polls has Biden ahead by 8.6%. There was also on recent poll which had Trump ahead in Wisconsin. The averages have Biden still ahead there by 6.2%. The other signs are positive for Biden. RCP now has him in a tie with Trump in Ohio. The 538 polling remains better for Biden than RCP. It has Biden ahead in Ohio and in a tie in Iowa. A remarkable outlier was recorded in Missouri this week. Trump won Missouri in 2016 by 18.5%. One poll this week had Biden ahead! And 538 records only a narrow Trump lead on the average of a number of polls. I am inclined to doubt these numbers at the moment.
8 July
This week has seen the disruption of 4th July weekend. Therefore the pace of polling has been down somewhat. What there has been has tended to reinforce the two key trends: Joe Biden has a substantial lead in the national polls and in sufficient “battleground” states to win by a large margin if the results are reflected in the November election and there are occasional “outlier” results in both directions to ensure that there will remain doubt about the ultimate outcome. The current RCP averages still point to a 333 to 205 victory for Biden. For those to whom this is their first look at this update I reiterate the explanation I included in the first report. At this stage I am not making a prediction merely describing the current situation. Other reputable outlets are attempting to measure the probability of the alternative outcomes. The Economist magazine has developed a complicated formula based on historical trends. They are saying that the current situation suggests a 99% chance that Biden will win the national vote and 90% chance he will win the electoral college vote. CNN has just published an article by their polling analyst which claims that polls taken around Independence Day are highly correlated with November results in incumbent contests. The other event this week which could be significant (although it probably won’t ultimately happen) is the announcement by Kanye West that he will run. One feature of the polling results so far has been the absence of a significant third party candidate. West will certainly not win but given his closeness to Trump his announcement may indicate an attempt by the Trump campaign to split the vote. Even 1-2% can make a difference as Ralph Nader showed when he tipped George Bush over the edge to a narrow win in Florida in 2000.
15 July
Polling trend emerge gradually so each week tends to show minor changes. This week was no different in aggregate. Each week I will try to go into more detail about one State. There was a significant improvement for Biden in Florida. The latest poll had him ahead by 10! This is an outlier but there are two significant indicators which suggest Trump is in deep trouble in a state he won by 1.2% in 2016 and which has 29 electoral college votes. The average lead for Joe Biden is now 6% and Trump has not led in any of the last 16 polls taken since March. There is no guarantee of victory in November but there will need to be a big turnaround for Trump to win Florida in 16 weeks time. Other minor movements in the averages were mixed in their impact. Biden slightly improved his position in North Carolina and Texas (this is now rated as a tie!) However Trump improved his position slightly in Arizona although he is still behind. The cautious assessment of the current situation remains: Biden 333-Trump 205.
29 July
It has been a milestone week. We now have fewer than 100 days until the election, and only 60-70 days until people can start voting. There has been a flurry of polling data, but the predominant trend is stability. In the key states the results vary slightly as polling data does, but the averages and trends are remarkably stable. I thought it might be worth going over the background to the week by week data rather than just focussing on the variations. The assessments I have been giving are based on Joe Biden winning all the 232 electoral college votes Hillary Clinton won in 2016. This is not a given but all the signs point in that direction. There is not so much polling in these states but so far there have been no danger signs. In Minnesota, which reports suggest Trump has been targetting, this week polls suggest Biden is leading by an average of more than 11%! On top of these 232 Biden needs another 38 to win. 270 is the magic number. The key states I have identified to get him over the line are Michigan; Florida; Pennsylvania and Wisconsin. The current RCP averages for these states are: Wisconsin Biden +6.4 10 votes Michigan` Biden +8.4 16Penn Biden +7.4 20Florida Biden +7.8 29These states alone, or any three of them, would give Biden enough votes in the electoral college. In addition he is leading clearly in: North Carolina by 3% Ohio by 1.5% Arizona by 4%. Some data suggests Biden is also leading in Georgia and Texas, but at this stage the evidence is slightly against him in these states. The fact that they are even close is a significant signpost towards the election outcome at this stage.
See my October posts for later information on how the election polls have moved.
Robin's Room 