Because Jon Bernstein is recycling his President’s Day “post”:http://plainblogaboutpolitics.blogspot.com/2011/02/against-p-day.html, I will recycle the graph from “mine”:https://themonkeycage.org/2009/02/ranking_presidents.html from 2009. See the post for links to the data.
It’s worth keeping in mind the essential flaws in this kind of ranking process. Individual presidents face very different challenges — many of which are outside of their control. Some of these challenges are very difficult, and others are less so. In essence, we ask some presidents to play the Los Angeles Lakers, and we ask others to play a junior varsity high school team. Or the challenges could be different in kind: some presidents play the Lakers, and some try to climb Everest. But yet, at the end, we feel that presidential “greatness” can be measured accurately despite these challenges.
The more general problem here is that the experiences of presidents are not really comparable. This problem of non-comparability (or “non-equivalence”) is well-known in the social sciences. Zach Elkins and I discuss it in “this paper”:http://home.gwu.edu/~jsides/vodka.pdf (pdf), which is titled “The Vodka Is Potent But the Meat is Rotten.” (If only the paper were as interesting as the title!) Malcolm Gladwell discussed it, although without using the term “non-comparability” in this recent “New Yorker article”:http://www.newyorker.com/reporting/2011/02/14/110214fa_fact_gladwell.
The point is not that all presidents are identical. The point is that ranking their “greatness” conflates their skill with the circumstances they faced.