How Well Do You Know American History?
Or, how biased is your American history?
Think you know American history? I’m sure many of you can name most or all of the American presidents, but how well can you rank them?
I was curious how well my perception of US presidents lines up with expert rankings, so I ranked the US presidents and then plotted those rankings against the experts.
Here are the results:
The expert rankings are plotted on the x-axis and my personal rankings are on the y-axis. Any president plotted above the green 45-degree line in the center was ranked higher by me than by the experts. Any president below the line was ranked higher by the experts.
So, how did I do?
Overall, I had a correlation of .89. I have no idea how that would compare to others, or even how it would score against an individual expert’s ranking. I’d love to get standard deviation data on each president, but unfortunately C-SPAN only publishes the aggregated results.
I got the top 5 presidents correct and nearly got them in the correct order, having just switched the two Roosevelts. The next ~15 were fairly in line with expert ranks, with roughly half receiving better ranks and the other half receiving worse ranks.
Most of my variance comes from mid-to-late 19th century presidents that I frankly don’t know a ton about. The first president that I was more than 10 ranks away from the experts was Taft, and Taylor and Harding were the only other two that eclipsed a difference of 10 or more. I’d be really curious to see if the experts had more variance for politically charged presidents like Reagan and Wilson or for some of the more obscure presidents like Arthur and Garfield.
I was also curious what the data would show about my political biases, so I colored Republicans and Democrats red and blue starting with FDR to see if I could spot a pattern. In general, I gave a slight advantage to Republicans, giving them an average of 2 spots higher than the experts did, and a slight discount to Democrats, ranking them an average of 4 spots lower than the experts did. This small but consistent bias is pretty reflective of my center-right leaning, so I was interested to see that reflected in the data so clearly.
I was also curious how the presidents have fared over time, so here are the presidents in order on the x-axis, with expert rankings on the y-axis. The two worst presidents of all time (Buchanan and Andrew Johnson) being on either side of the best president (Lincoln) is fascinating.
Methodology
The expert rankings come from the C-SPAN Presidential Historians Survey, which has had four iterations since the year 2000. The data in the graph represents the latest results, which are from 2021.
The 2021 rankings are based on survey responses from 142 “historians, professors, and other professional observers of the presidency”, who were asked to rate each president on a scale of 1 to 10 on the following 10 qualities of presidential leadership
Public Persuasion
Crisis Leadership
Economic Management
Moral Authority
International Relations
Administrative Skills
Relations with Congress
Vision/Setting an Agenda
Pursued Equal Justice for All
Performance Within the Context of the Times
Survey participants are guaranteed that their individual survey responses will remain confidential, which limits the effect of social desirability bias. The 10 categories are then given equal weights in determining a president’s total score.
There’s a ton to argue about when it comes to presidential rankings, but in general, this type of aggregated scoring from experts is probably as objective of a measure as we can get for something that is inherently so subjective. You might notice I excluded Trump’s rank (4th to last place) from the data. I think the earlier the president, the more concrete their ranking is. More recent presidents are mostly still waiting for the political dust to settle into historical consensus, so I’ve exercised the rate for the dust to settle on Trump a bit more.
How would you do?
Curious how you’d do? Let me know in the comments or shoot me a DM and I can send you the excel. I’d love to get a sense for what a “normal” correlation is and also how well political biases are gauged by this type of ranking.