US Election: Mallard Forecast
From now until the end of the US Election, the MallardUK will run a weekly forecasting update on who we think will win the 2020 US Election come Polling Day (or weeks later, when the could be crucial mail ballots are counted).
With both candidates now confirmed, the Incumbent President, Donald Trump, and former Vice President Joe Biden will go head to head in what will be the most virulent battle in American Presidential History.
The forecasting model will be considered by an outside/inside view perspective, as all forecasting models should be, with a consideration of the road ahead. The outside view being the history of the American politics. Whilst not as relevant as the as the inside perspective, which in this case would be polling and approval ratings. Outside views are critical in challenging our biases and ensuring forecasts are not to bullish (being more confident on an outcome than you actually should be).
So, first, let us consider the outside perspective.
The outside perspective is a view that focuses on what we know. Objectively looking at a scenario without knowing any contemporary issues that may impact forecasting. This is known as an anchor forecast and is what the inside perspective must be considered against, to protect from bullish forecasts.
As this Election involves a President in his first term and will be looking for re-election, we will first look to the history books on Presidents seeking re-election. In the History of the US Election, 35/45 Presidents have gone on to win a second term. However, given that the 35th President J.F Kennedy was assassinated just two years into his reign in 1963 and therefore did not have a second election, we will make that 34/44. Meaning that from a purely outside perspective, the chance that an incumbent President will win a second term is around 80%
We will then look at the Presidents who did not serve a second term and see if we can draw any comparisons with the current President. Of the nine who did not get re-elected, there is a trend; they either: endured scandal, economic downturn, or went to war against popular will. Consider John Adams, the second President and the first to not win re-election, who endured a controversial war with France during his first term. Consider Herbert Hoover who failed to win a second term following the Great Crash in 1929. Consider Gerald Ford, who took over the Presidency from Richard Nixon following the Watergate indictment in 1974. Consider the last President to not win a second term, G.H.W Bush, who following an unpopular Gulf War lost to Bill Clinton in 1992.
It is important to point out that many Presidents were re-elected under the conditions above. Consider Bush Jr, whose near entire administration was plagued by the War on Terror, yet still won a second term. Consider the first Black President, Barrack Obama, who won a second term after the Great Recession. But it is fair to say that the more adversity you endure, the higher chances of failing to win a second term are. Therefore, we will adjust the outside perspective on an incumbent President going through economic downturn, having a failed indictment against him, and being involved in countless trade and proxy wars, to 60%
For the inside perspective we need to consider the contemporary positions that will be key in determining, who wins the US Election.
Before we do that, we need to consider how the US Electoral System works. Like Britain, votes are allocated by areas to the American Electoral College. With each state carrying a number of votes depending on its size. The magic number needed in the Electoral College is 270; in 2016, Donald Trump beat Hillary Clinton by 304 College Votes to 227. Trump got just shy of sixty-three million votes, whereas Clinton got just shy of sixty-six million votes, representing a 2.1% national vote victory for Hillary Clinton.
I hope the above demonstrates how looking at national polling doesn’t paint the full picture in regards to the American Election. In 2016, it would have been possible for Trump to win from a national vote defeat as high as 3-4%. As again it is now. The author of the 538, Nate Silver, stated recently that Biden will not be the favourite to win with a 3% national poll lead. It also must be noted that Silver appears to be very left leaning. So an admission from Silver on this speaks volumes. A recent poll found that in strong Democrat states, Joe Biden is as popular, and in some cases more popular by net votes, than Barrack Obama. It is likely that Joe Biden will win strong Democrat States by a higher margin than Hillary in 2016, to no electoral vote impact. I am, therefore, inclined to think that Trump would be favourite to win with a losing vote margin of as high as 4% (with it being possible to win with as much as -10), with anything above 4% moving towards Biden`s favour. See below for a full breakdown of likelihood of Trump victory by losing margin.
(National Vote Losing Margin for Trump)
(Chance Trump wins the Electoral College)
More important than the National Polls when looking for clues as to who will win are the key Battleground states and Swing states: Colorado, Pennsylvania, Virginia, Minnesota, Florida, Wisconsin, Nevada, New Hampshire, North Carolina and Iowa. Strong trends or polling in the majority of the states above gives a much clearer indication than national polls. The only problem is that state polling is widely considered as having a higher margin error. As sample sizes are often very small. An example of state polling swings can be seen in recent polling in Pennsylvania. A poll just a few weeks ago suggested Biden was leading by 16 points. Whereas this week polling suggested it was a tied state. Polling released this week by the Trafalgar Group has Trump coming out on top In Wisconsin (+1), Florida (+3), whereas other polling outlets, including Fox News, showed the opposite.
National Poll averages are widely accepted as being around 6-7% in favour of Joe Biden, which is 2-3 points clear of the +4 he needs to become favourite. We will call the space between the +4 Biden needs, and his polling lead, the ‘buffer’. Giving Biden a buffer of 2-3%. Which, importantly, is inside the margin error of US National Polls, 4%.
Considering both state and national polling. If the election was today, I would favour Joe Biden to win the US Election 63/100.
The Road Ahead
Whilst Joe Biden would be favoured to win if there was an Election today, there isn’t an election today (though some mail votes have already opened), so we have to attempt to forecast the path ahead.
An objective, outside look at poll shifts between 60 days till the US Election to Election day shows that it is very likely that a shift will occur.
Difference between polling averages 60 days out vs winning margin:
60 days: +2.5 Clinton – Actual winning margin: 2.1 Clinton (DT won EC)
60 days: +1.7 Obama – Actual Winning Margin: 3.9 Obama
60 Days: +3.6 Obama – Actual Winning Margin: 7.2 Obama
60 Days: +7.9 Bush – Actual Winning Margin: 2.3% Bush
60 Days: +3.1 Gore – Actual Winning Margin: 0.5 Gore (Bush won EC and Election)
60 Days +15.2 Clinton – Actual Winning Margin: 8.5 Clinton
60 Days +7 Clinton – Actual Winning Margin: 5.6 Clinton
An analysis of the 60 day polling average to the winning margins from the seven elections since 1992, shows the average shift is 3.2%. Which is why, considering that Biden’s buffer margin is 2-3%, forecasts that are 99% certain of a Biden victory are not to be taken seriously.
To try and get an insight into the road ahead from an inside perspective we would look at the momentum, which is certainly in Trumps favour. National Polls have reduced from an average of 11% to an average of 6-7%. With a similar shift in Battleground States, it has been suggested that the recent riots in America have led to a reduced margin for Trump. Whilst the data on this is not clear, it is clear that Law and Order is a key Policy Indicator for Americans, and that support for Black Lives Matter has reduced as the riots have continued. So, whilst it is unclear if they help Trump, they certainly do not help the Democrats who, according to polling, the public see the Democrats on the side of the rioters, with Republicans on the side of the police.
According to a recent poll, the economy is the top priority for Americans, of which Trump holds polling advantages on who the American Public think is best placed to aid economic recovery.
It is Trumps advantage on Americas two biggest concerns, Economy and Law and Order, that give him in the advantage on any perspective debates. A recent poll found that 47% of people thought Trump would come out of the debates on top, compared with Biden, 41%.
The Democrats recognise this; Speaker of the House Nancy Pelosi recently said that Biden should refuse to debate Trump, as to not give him legitimacy.
Whilst Joe Biden has the advantage today, both inside and outside perspectives of the road ahead favour Trump.
Whist the outside perspective gives a 60% chance of an incumbent President like Trump to win a second term. The inside perspective today favours (2-3% buffer) Biden at 63%. Factoring in the road ahead, which favours Trump, and the history of poll shifts from 60 days out (mean shift 3.2%), indicates a shift within the buffer is possible and likely (bear in mind that this could also mean Biden increases his lead).
We at MallardUK think it is evenly poised. The U.K. bookies (Sky Bet) currently have the pair matched at 1.92/1, and we think that is a fair reflection 60 days out. We will watch the election closely and give a weekly update as shifts occur. But for now, we see it as even.