As a recent article by The Atlantic shows, the men who have been elected to the highest office of government have generally been wealthy. By combining The Atlantics data on presidential wealth with a 2009 CSPAN poll which ranked the presidents by their job performance, we have attempted to analyze whether wealth affects Presidential effectiveness.

Very generally, the trend line on the bottom graph shows that as wealth increases, so does the president's ranking. However, when outliers like George Washington, Thomas Jefferson and JFK are removed, the correlation is significantly reduced. There are other exceptions: Abraham Lincoln tops the chart, but was essentially penniless his entire life.

What we like about this viz

Pictures: Although it would have been too complex to put pictures on the visualization itself, having thumbnails on the side achieves the same result without harming the analysis.


Ross, what are you even doing with a trend line on that viz. :) What I find really interesting is that there were only 100 years where we had some (relatively) "poor" presidents. My guess is that before 1850, you had to be more rich/noble to *get* elected, and in the modern era post 1950 being president *makes* you rich quickly. The data is *peak* net worth, so it makes sense.

I do not think the average wealth reference line adds any info, because of the large outliers. What if you were to turn on quartiles or add a median instead of average, and as a line chart instead of shapes?

I wish the wealth axis was logarithmic for better detail close to the zero wealth line. It's pretty crowded down there.

Meli, yeah he did a logarithmic wealth axis on one, and not on the other.

Thanks for your comments everyone!

The trend line is there to show that there is very little trend. As I clearly state in the post, I am not trying to attribute causation, but rather illuminate how little correlation there is.

There is no average wealth. As James points out, these values are for peak net worth.

I tried a log axis on the bottom graph, but I wanted to keep it linear because it was so easy to read it the wrong way (there was no option on the top graph).


I have a few questions about your last comment,

1. What do you mean "There is no average wealth"? When I look at the top chart, there is a yellow reference line, and when I hover over it, the tool-tip says "Average: $61M". The point I was making was that averaging the amounts does not add value to the viz because of the large outliers. I am guessing that you are also saying that the average of the wealth amounts adds no value because of what the wealth amount means.

2. What do you mean "there was no option on the top graph"? If you think it is easy to read a logarithmic incorrectly, why did you use it on the top chart? I would interpret the phrase I quoted as: a linear scale was not an option on the top graph. Is it not an option because it would not clearly display the marks? Because I know an axis can be made either logarithmic or linear. So I am confused why you chose logarithmic for one and linear for the other.

I realize I am splitting hairs, but I am very interested in understanding your point of view, and thoughts on the merits of adding a reference line and scaling the axes as you did.

Is this data adjusted for inflation?

Yes, wealth is inflation-adjusted.

Thomas Jefferson was a founder of what became the Democratic party under Jackson. The visualization incorrectly lists him as associated with the Republican party, which didn't even exist during his lifetime. And yes, both parties have changed over the years so its hard for either party to claim President's prior to Jackson.

Note that the vertical axis on the second graph is ordinal rank and not a score for effectiveness. The graph appears to depict Lincoln and Washington as "45 times as effective" as Buchanan, and the trend line assumes this. In the CSPAN survey their actual scores were 902 and 227, so Lincoln was only four times as effective. Computing the trend line on score, rather than rank, would make the impact of wealth on effectiveness even less.

Ross, teach me how you were able to get the pictures to update and change depending on the president you were referencing? I have a similar application that would work very well for. Appreciate the help.

Assine nosso blog