Monday, November 2, 2020

OT - Final Pre Election Observations

Since it is obvious that the outcome of the US election will have some kind of impact on markets and also the longer term fiscal policies of the US, it is important to at least take the election results into account for personal financial decision making. Just for my own information, I have done a fairly deep dive analysis into the early voting numbers for this election and will share that today for anyone interested.

The first observation I would make is that I still believe that all five possible outcome scenarios I listed in this earlier election outcome analysis are still possible after looking at the early voting numbers. Below I will recap a possible path to victory for each side based on the nearly 90 million votes that have already been cast. For anyone who wants to see how those votes split up by party affiliation, you can easily do that on this site (even down to individual counties). If you do that, these are some things you will see:

- 5.4 million more people who identify as Democrats have voted early compared to Republicans

- 9 million people (10%) have voted who don't want to say they identify with either party have voted early

- early voting trends suggest at least 150 million people may vote up to as high as 155 million

- most of the 5.4 million vote lead for Democrat identified voters comes from about 15 states that are heavily Democratic. Everywhere else the vote is either close or Republican identified voters lead. This is the same dynamic we saw in the 2016 election, so no surprises there. 

-in all the key swing states that should determine who wins the electoral vote, either side can still win

A "conventional wisdom" observation;

- More of the Democratic ID voters will vote early than Republican ID voters who are more likely to turnout heavier than  Democrats on election day

Using the above information, below is how each side can look at the early voting data and make a credible case to win the electoral college vote:

The Democratic Path

Democrats would look at their lead in the early voting (including some of the key swing states in the US Rust Belt) and say that that lead combined with a larger than normal number of Republican ID voters voting against President Trump will be enough to get the 270 votes needed. They would point to several suburban counties around major cities and say they see Trump under performing this time in those areas which will negate the expected election day turnout advantage Republicans are counting on to win. Of course they also point to most mainstream polls showing they will win the election. 

The Republican Path

Republicans would say they feel good because in most key swing states, they are doing better in the early voting than they did in 2016 (they are either slightly ahead, even, or slightly behind). The conventional wisdom was that Democrats would have a larger lead after early voting than they do. They then expect to get the usual election day turnout advantage to pull ahead and win the 270 votes they need based on that turnout. They agree that Trump is doing a little worse in the suburbs this time, but they say he will outperform his share of votes from the African American and Hispanic demographics this time and more than offset his drop off in the suburban vote. 


I have dug very deep into all the early voting numbers and I can say that both of these are credible takes on the numbers. It's really not possible to say who will turn out to be right at this point because all we know is what the early voters say is their party identification (not everyone will vote their party ID). Also, in some key states (like Michigan, Wisconsin, and Minnesota) there is a large % of voters who don't say they identify with either party. I do see indications that Trump is not performing as well in some suburban areas (not all though). I do see some indications Trump is probably going to improve his share of the African American and Hispanic vote. Some polls show him possibly adding 5-10% in those voting demographics. So it is reasonable to say this happening. How many votes are moving in each direction is clearly important, but impossible to determine just looking at the early vote data.


So who do I predict will win? I cannot make a prediction. If I thought there was an overwhelming case to interpret the information we have now to make a clear prediction, I would do that. But there are too many millions of votes this time that could be going either way for me to try and predict how those people will vote. This year there are also millions of new voters and millions new first time voters who have no history that you can use to assess their voting patterns. You credibly can say Trump will win if he does get a majority turnout on election day looking at these numbers, but that has not happened yet.

One prediction I can offer is that it is likely that we will not have a clear winner that is agreed upon on election day and perhaps not for several days after the election (possibly even several weeks). This is because 3 key states (Nevada, Pennsylvania, and North Carolina) will still be accepting votes up to as much as 10 days after the election. So no final vote counts will be available in those states and it is very possible no one will be able to claim those 41 electoral votes until at least several days after the election. It is also possible other vote totals in key swing states could be very close and legally contested by either side or both sides. If that happens, even more electoral votes could be held up for some time. 

Readers should expect that the outcome of this election is likely to be delayed for some period of time and that this will create significant uncertainty in markets. So don't be surprised if that does happen. 

I will attempt to analyze the election outcome in terms of any impact it may have on the monetary system once a final outcome is determined. That could be in a few days or a few weeks. Right now, I don't think the outcome will have any major impact on things in the short term (less than one year) unless some kind of major market crash is triggered for some reason.

No comments:

Post a Comment