01 September 2008

Probability of VP Unexpectedly Becoming President

In the wake of Senator McCain selecting Gov. Palin to be his running mate, many words have been devoted to the possibility of Sen. Biden or Gov. Palin unexpectedly becoming President. However, I have seen no discussion of the probability of this actually happening.

Our nation's history provides us with a sample of 42 Presidents and 46 Vice-Presidents (including Bush and Cheney) from which we can estimate this probability. The number of Presidents and Vice-Presidents is unequal because Presidents have had between zero and three Vice-Presidents. Eight VPs have become President upon the previous President's death (most recently Johnson), and one VP has become president upon the President's resignation (Ford).

So, the probability of our next VP unexpectedly becoming President is

Punexp (VP → P) = 9/46 = 19.6%.

I am not sure how to do error propagation on a calculation like this.

We should pay attention to the Vice-Presidential candidates because we have a roughly one in five probability of choosing the next two Presidents of the United States in November.

5 comments:

Anonymous said...

Most of those VPs to become president did so because the president was assassinated. The likelihood of such a situation is now almost nil, since they close down whole states when the Prez comes to visit. So while historically, yes, 1 in 5 VPs have taken the reins, the practical percentage is much lower.

Anonymous said...

Hi Luke,

The 95% confidence interval on your estimate is about 8.2% to 30.9%. (Standard Error for a proportion is sqrt(p*(1-p)/n)

I'd agree with the previous post that assassination is probably less likely; but death (or incapacitation) due to health issues is probably more likely for McCain due to his age and health history than many previous presidents.

Luke C. said...

Thank you, Stephanie, it is always good to have a statistician to comment on discussions like this.

Actually, only 4 of the VPs became president due to assassination. Most of them (5) became president due to the presidents death by natural causes or resignation.

Trying to reduce the sample to account for changes in security and medical technology would reduce the significance of the calculation too much. Besides, I am not trying to make a precision calculation. I am making the point that we should care about the VP candidates.

Anonymous said...

You're specifically interested in this case, since the probability is different for each person. Given that the state of medicine and healthcare has improved significantly since the last President passed away naturally, promoting the VP, and since the President is assumed to receive the finest healthcare the nation can provide, you have to answer the question: what is the probability that a person with a given medical history, of a certain age, under tremendous mental and physical stress, with perfect healthcare, will live 4-8 more years? That's the question that needs to be answered, rather than viewing this as a history driven probability. In fact, given how much things have changed historically with regard to many of these factors, I think that this probability is much smaller. But age and history have a lot to do with this.

I'm significantly more interested, though, in the question of which ticket will try to do a good job making sure all of us live 4-8 years longer.

Unknown said...

This is Russian roulette with worse odds....and consequences. There are 300 million of us playing.