An analysis of “The Density of Smart People”

Posted by steve | Education | Sunday 30 May 2010 8:56 am

Over in The Atlantic, this article appears that compares cities by density of degreed people.  I was delighted to find the included link to the actual spreadsheet data used to create the prominent graphic associated with the report.  After reading it I wanted to perform my own analysis.

When reading the article my immediate reaction was “Of course that is the outcome, but it does not demonstrate anything useful.”  Which is why I was glad to find the raw data.  My reasoning goes like this:

  • The highest “rated” cities in the report also appear to be the cities having overall highest population density.
  • I would expect that in cities having a high density of population to have a high density of degreed people.
  • This would only be useful if I believed that the distribution of educated people is linear.

A much more interesting analysis, in my opinion, would contrast the density of the city to the density of the educated in that city.  So I did my own analysis using the same input data, assuming that would be a valid starting point.

Here is the graph of my own analysis.   Lower is better.

Comparing Density of City to Density of Degrees

What the analysis tells me is that a city like Seattle is a good place to be if you want to meet more educated people per population density. And Detroit ranks at the bottom.