Let me add my two cents to what Noah and Dominic and the Editors podcast panel have said about the rankings of “presidential greatness” by a survey of 154 “current and recent members of the Presidents & Executive Politics Section of the American Political Science Association, which is the foremost organization of social science experts in presidential politics, as well as scholars who had recently published peer-reviewed academic research in key related scholarly journals or academic presses.” There are all manner of quibbles I could offer with the specific rankings, which placed Joe Biden at number 14 and Donald Trump dead last. Franklin Roosevelt above George Washington? Teddy fourth? Reagan behind Biden, Obama, Bill Clinton, and LBJ? Grant behind James Madison, who was a great American but not a very good president? Woodrow Wilson above . . . anybody? Coolidge 34th and Harding 40th? Polk below Jimmy Carter? I could go on.
I don’t have a problem, in theory, with ranking presidential performance. But these exercises suffer from a number of intractable problems of execution.
First, what are the criteria? Does “greatness” mean that the person did consequential things, or good things? The latter calls for inherently ideological judgments about what is good, and the former does not — although it embeds some assumptions. It still values people who made changes and exercised power rather than people who prevented changes, refrained from action, or avoided crises — even if those were the wiser courses. There are ways we can measure presidents by somewhat objective metrics — popular and electoral vote, growth of their parties down-ballot, legislation signed, etc. Some of these have more serious defects than others. If one person ranks the presidents, that person can state their criteria. But 154 people? They are likely to be answering different questions, so combining their answers is like comparing how tall a tree is to how heavy a rock is.
Consider Polk. On one hand, he massively expanded the country by means of one fairly easy war and one diplomatic negotiation. He accomplished all of his campaign promises in a single term, the only president to do so. Does that make him great, because he accomplished so much? Or does it make him a pro-slavery Democrat whose addition of territory in the West by means of an unjust war lit the fuse that led to the Civil War? There are such questions with many presidencies.
Second, who are the voters? One-hundred and fifty-four people isn’t much of a scientific sample size. An electorate of academics is inherently a left-leaning group, and an electorate of political scientists will have particular biases towards certain types of activism. Wilson himself was president of the APSA from 1909-1910, just before entering electoral politics, so of course, he did well with his own professional clique.
How left-leaning is this sample? We are told that Trump, for example, received an average rating of 10.92 out of 100. His average rating from self-reported Republicans is 31.38, self-reported Democrats is 6.66, and self-reported independents is 13.96. By self-reported conservative/liberal/moderate, it’s 26.76, 6.97, and 14.21, respectively. We’re not given a breakdown of how many people self-reported each category or whether all respondents self-reported — a fairly shocking lack of crosstabs, for a survey of political scientists. I will spare you the math, but to get close to a 10.92 rating with those inputs, you’d need a sample that looked something like 90 Democrats and liberals (58.4 percent), 53 independents and moderates (34.4 percent), and 11 Republicans and conservatives (7.1 percent). So, this is obviously a very skewed sample.
Third, how much do the voters know about all 45 presidents? I’m a professional political writer and history buff who majored in history in college. I have devoted quite a lot of my time to studying American presidents. Offhand, I’ve read full biographies of at least 14 presidents (I spent a vast amount of time with Wilson’s presidential papers while writing my senior thesis in college), and I’ve read either parts of biographies or multiple-subject bios of nearly as many others. But coming fully to an assessment of a presidential administration requires covering a lot of ground. For example, writing up a two–part review of Grant’s presidency, I covered Reconstruction and the formation of the Justice Department; corruption, fiscal, tariff, and monetary policies; his policies towards Britain, the Dominican Republic, and Spain; his treatment of the Sioux, Chinese immigration, Mormons, Catholics, Jews; and his approach to religious schools, abortion, and pornography. That’s a lot of topics, and I didn’t even get to, say, his Supreme Court nominees.
Now, do that for 44 more terms. I’d guess that a whole lot of these respondents — even ones with deep and profound expertise on some particular issue or time period — know, at most, two or three aspects of the typical presidency, and less than that for a bunch of the less-remembered presidencies. How do you compare the trade policies of Chester Arthur and Gerald Ford, or the judicial nominations of Rutherford B. Hayes and Herbert Hoover, or the Pacific Rim policies of Warren Harding and John Quincy Adams? How many people, even historians — and many political scientists are not historians, especially in this age when much political science is dominated by quantitative metrics — have handy in their heads a full assessment of all 45 presidents across all of the areas of politics, policy, and personnel that matter?
Fourth, there is the matter of recency bias and perspective. It’s impossible to ask anybody to set aside current politics in rating Trump and Biden, both of whom are currently running for president this year. It’s hard to have a historical perspective of any sort on Obama, Clinton, or George W. Bush, and hard even to do so for Reagan or Nixon.
So, combine a small and skewed sample, no comprehensible criteria, a highly uneven base of knowledge, and the gravitational pull of an ongoing election, and you get a survey that is basically just clickbait.