A warning to my casual readers: this post is going to get more than a wee bit nerdy, and probably a bit political, too.

OK, with that out of the way, let me note that one of the things that’s been bugging me ever since John McCain’s announcement of Sarah Palin as his VP choice last week is that while there’s been a veritable chorus describing her as “inexperienced” and “unqualified”, no one has really bothered to set down what they thought a VP’s experience should be. I had this discussion with a guy who is a big Obama supporter over on a forum I frequent, and even he was unwilling to suggest a criteria other than that it should be “the same as if they were running for president”.

I decided it was time to give myself a history lesson. How much experience, exactly, did our various candidates for president and vice president have? Geof suggested plotting that data against their presidential ratings to see how it panned out. So I did that, too. To bound the problem a little bit, I decided to limit my study to the more modern presidential era (starting with 1960). Then I headed off to Wikipedia to do some data collection.

The Setup

A person’s experience is, in some ways, difficult to quantify, but I settled on the following categories of experience:

  • Years of college education (I also tracked whether it was Ivy League and whether they got a law degree)
  • Years of military service
  • Years in a state legislature
  • Years as a state governor
  • Years in other federal government service (i.e. cabinet or civil service positions)
  • Years in Congress
  • Years as Vice President
  • Years as President

The tricky part, then, is how you choose to sum these up; let’s just agree that, for instance, years served as Vice President or as a governor are more valuable, year-for-year, than those served in the military or in a state legislature. I settled on some multipliers to try to help even things out. Feel free to argue over these if you want to.

  • Years of college education (I also tracked whether it was Ivy League and whether they got a law degree) - 0.25
  • Years of military service - 0.25
  • Years in a state legislature - 0.25
  • Years as a state governor - 1.0
  • Years in other federal government service (i.e. cabinet or civil service positions) - 0.5
  • Years in Congress - 0.75
  • Years as Vice President - 1.0
  • Years as President - 2.0

So, for example, George H. W. Bush, in 1984, had 4 years of college, 4 years in the military, 5 years in government service, 4 years in congress, and 4 years as VP. That gives him a score of ((4*0.25)+(4*0.25)+(5*0.5)+(4*0.75)+(4*1.0)) = 11.50.

With those multipliers in place it was easy enough to get Excel to do some sums and give me some totals.

What I found was fairly interesting.

The Data

The average experience score for a presidential candidate: 16.8. The average experience score for a VP candidate: 12.9. Highest score for a presidential candidate: 28.75, shared by Bob Dole in 1996 and Gerald Ford in 1976. Highest score for a VP candidate: also 28.75, Joe Biden this year.

Lowest score for a presidential candidate: 5.25, Barack Obama, this year. (second lowest: George W. Bush’s 7.50 in 2000.) Lowest score for a VP candidate: 3.00, Sarah Palin, this year. (second lowest: Spiro Agnew’s 3.75 in 1968.)

Highest POTUS/VP combined score: Dole/Kemp in 1996 (45.75) Lowest POTUS/VP combined score: Reagan/Bush in 1980 (17.25)

So that’s a lot of data, how about some analysis?

Analysis

I did a plot of the experience ratings against some presidential performance ratings (as found here, which claim to be amalgamated from several different ratings on Wikipedia), but found that to be a mixed bag. There were experienced presidents who ranked poorly (Nixon) and well (LBJ) and inexperienced presidents similarly (Reagan ranked high, Jimmy Carter much lower). Result: Inconclusive.

Next, I noticed an interesting trend. If you throw out the few elections where strong incumbents were running for second terms (LBJ in 1964 after finishing JFK’s term, Nixon in 1972, Reagan in 1984), in each of the other cases, the POTUS/VP pair with the lower experience score won the election. Result: If that trend holds through this election, McCain/Palin will win.

If you want to do a little more hardcore statistical analysis,

POTUS Standard Deviation: 6.59 VP Standard Deviation: 5.82

Just for sake of argument, this means that Obama’s POTUS score (5.25) is 1.75 standard deviations below the mean, and that Palin’s VP score (3.00) is 1.70 standard deviations below the mean… which means that, per these ratings, Obama is slightly more relatively inexperienced as a presidential candidate than Palin is as a VP candidate. (Only slightly, though.)

Conclusions

Well, this is great data for us dataheads who like to ponder such things. What it really shows, I think, is that there are far more factors that play into the election (and the subsequent job performance) than just experience.

I’ll also conclude that I still haven’t answered the question regarding “how much experience is enough?”. Yes, Palin is the least-experienced VP candidate in the past 50 years. But Obama is also the least-experienced POTUS candidate. Hey, the nature of number is that somebody will have to be least-experienced. So until somebody can give me some quantifiable other measures, I think it’s still gonna come down to gut feel and politics… like usual.