If you want to bet on anything there is a market ready to take your bet. Like if you want to bet on whether President Trump will complete his first full term as President or not.
After President Trump fired FBI Director James Comey the probability for President Trump not finishing his first term jumped up to 50%.
Betting markets have had the probability of Trump not finishing out his term rise to ~50% since the Comey firing. Was around 40% pre-Comey.
— Nate Silver (@NateSilver538) May 11, 2017
Not finishing his first term includes impeachment, resignation, and death by natural causes.
My immediate reaction is these odds seem high. While other people, maybe under the influence of the Halo and Horn Effect, believe these odds to be too low.
Whatever your viewpoint, what this does offer us is a look at the Inside View and Outside View.
The inside view can be described as intuitive forecasting. From Michael Mauboussin in his book Think Twice,
An inside view considers a problem by focusing on the specific task and by using information that is close at hand, and makes predictions based on that narrow and unique set of inputs.
The inputs may include and grossly overweight “anecdotal evidence and fallacious perceptions.”
Michael uses the example of Big Brown and his bid to win the Triple Crown in 2008.
Big Brown won the first 2 legs of the Triple Crown (the Kentucky Derby and the Preakness) with the Belmont Stakes remaining. Based on how handily Big Brown won the first 2 races and how weak the field would be in the Belmont, Big Brown was the overwhelming favorite to win. Even Big Brown’s trainer, Rick Dutrow, proclaimed it was a “foregone conclusion” that Big Brown would win.
Right before the race bettors pushed Big Brown’s odds of winning to 3/10 or a 76% chance to win.
Then Big Brown lost and he lost big. Big Brown came in dead last.
Racing fans and bettors put too much weight on Big Brown’s recent wins. A belief that Big Brown was just too dominant to lose permeated bettors’ minds.
What the majority of bettors never asked was how many has horse been one win away from the Triple Crown? And then how many times has that horse won?
Again, Michael Mauboussin in his book Think Twice describes the outside view as such.
The outside view asks if there are similar situations that can provide a statistical basis for making a decision. Rather than seeing a problem as unique, the outside view wants to know if others have faced comparable problems and, if so, what happened.
Continuing with Michael’s example of Big Brown.
Of the 29 horses that were one win away from the Triple Crown, only 11 have done so. This is a 38% success rate.
But Michael then points out that before 1950 8 out of 9 horses won their Triple Crown bid but after 1950 only 3 out of 20 horses won their bid, a 15% success rate.
The outside view is also known as the base rate.
The base rate for winning the Triple Crown falls within the range of 15-38%. With the post-1950 numbers being the most applicable.
Big Brown’s chances of winning the Triple Crown were pushed up to 76%. A large divergence from the base rate.
What is the Base Rate for a President not Finishing Out a Full 4-Year Term?
We need to know a couple things.
How many President’s have died in office but were not assassinated?
- William Henry Harrison
- Zachary Taylor
- Warren G. Harding
- Franklin Delano Roosevelt
Only 4 out of the previous 44 Presidents or 9%.
How many Presidents have resigned from Office?
- Richard M. Nixon
1 out of the previous 44 or 2.27%.
How many Presidents have been impeached while in office?
- Andrew Johnson
- Bill Clinton
2 out of previous 44 or 4.55%.
How many Presidents were impeached and removed from office?
The House starts the impeachment process but the Senate conducts the trial. No President has been found guilty by the Senate.
For the most part, these events are mutually exclusive and we can simply sum each probability.
(We could make it harder by focusing only on those events that happened in a President’s first 4-year term but let’s keep it simple.)
The base rate for President Trump not finishing his first term because of either death, resignation, or being impeached by the House and found guilty in the Senate is.
9% + 2.27% + 0% = 11.27%
Base Rate is the Starting Point
The lesson from Phillip Tetlock’s book the Superforecasting: The Art and Science of Prediction is any prediction needs to start with the base rate. Then given new information and specific knowledge you will incrementally raise or lower your odds based on the new information.
If you do not have any new knowledge or specific insights then stick with the base rate.
The firing of James Comey is definitely new information. Optically, the timing of his firing – with the FBI opening up an investigation into Trump’s Russia connections – is bad. Even if it is just that, poor timing.
Some people are comparing the firing to Nixon’s Saturday Night Massacre. When President Nixon fired the special prosecutor investigating into the Watergate break-in.
Is it enough new information to move the odds that President Trump doesn’t finish his first term up to 50%?
I don’t know.
What I do know is I don’t have any special insight into such matters. So I’ll stick with 11.27%.