How important is experience in making a good President?

Well, this was one of the huge arguments in contrasting Barack Obama and Hillary Clinton. She argued that experience was one of the, if not the, most important factor in choosing a President. She even went so far as to say, because of it, that John McCain, because of his experience, and unlike Obama, had passed the "commander-in-chief threshold".

I know numerous arguments had been made about this, but I recently found a very interesting article that details this question on an historical and statistical basis. The article can be found here.

Basically what's done is, based on the rankings of different historical surveys, studies and opinions, articles, etc., a ranking of the best to worst Presidents is made. Then, each of the points of experience that each President had before taking office is made, coming to a cumulative numerical number.

As if rightly pointed out,

All experience is not the same. How do you compare being governor of a small state with being governor of a big state? Is being Vice President worth a pitcher of warm beer or any other fluid? Is being a state senator more or less relevant to being president than being Secretary of Commerce? Impossible to say for sure, so I have justed added up the total number of years of experience as a proxy for experience.

And he gives the reader the ability to weigh each variable accordingly.

But the conclusion is clear: there does not seem to be a correlation between the experience someone has before taking office, and how good a President they are once in office.

In the table below, the Presidents are sorted on experience. The most experienced President was James Buchanan, with Lyndon Johnson and Jerry Ford getting the silver and bronze medals, respectively. Chester A. Arthur had a mere 1.5 years experience before he was elevated to the Presidency upon the asassination of President Garfield. If experience and greatness correlate, one would expect the top half of the table (the most experienced Presidents) to be mostly green (good) and the bottom half (inexperienced) to be mostly red (bad). If there is no correlation, the red and green should be random. Indeed, the latter seems to be the case. The top half in terms of experience (above the gray bar) has 11 bad Presidents and 10 good ones, essentially no correlation between greatness and experience. If you don't like this result and want to try for a better one, just get the spreadsheet and start weighting the columns. Undoubtedly you will be able to get a different result if you try hard enough. But the point remains, the Presidents with a lot of experience have not been more outstanding than those with little experience.

The table can be viewed on the page. It's in HTML format, so it's too much of a pain to transfer here. But here's a scatterplot of the same data:

Click to enlarge

The Presidents (way) under the line might be regarded as overperformers. They didn't have much experience going into office, but did pretty well once there. The ones (way) above the line had lots of experience but were not good Presidents.