Richard Florida’s singles map and commentary just didn’t seem right. I was both dissatisfied with the approach to data visualization, as well the likelihood that some communities could be so overwhelmingly favorable to one sex.

First, Florida’s map focused on a total count instead of a ratio. This obscures actual imbalance in pairing opportunities. This is because even a small percentage difference multiplied times a huge base can make it seem like there are a lot of surplus women or men in a big town (i.e. NYC.) Second, it did not control for factors linked to Florida’s “creative class” arguments resulting in low-skill workers being included. And third, I didn’t think that people over 45 years of age really were indicative of the “creative class” singles scene.

I used the 2006 American Community Survey data to examine the top 50 Metropolitan Statistical Areas (MSAs) for a comparison of only females and male residents with a bachelor’s degree or higher. I took education level as a proxy for “creative class” status, though I recognize Florida’s work is more nuanced than that. I only included 18-44 year olds.

(Click on Map for higher quality image)

When we look at the data this way, it turns out that there are not such significant disparities across communities as Florida’s visualization leads one to believe. The map is trafficking in something similar to what visualization authority Ed Tufte calls the “Lie Factor.” With the exception of San Juan, Puerto Rico, no major MSA community breaks a 10% disparity between the sexes. Even in that most imbalanced of the major MSAs, if one where to put ten people in room to replicate its ratio, four of them would be men and six would be women. Obviously this is better than even odds but not as dramatic as Florida’s singles map visually implies. And it certainly challenges Florida’s assertion that: “If you’re a single man or a single woman the odds of meeting that special someone vary dramatically across the country.”

Moreover, this cut and visualization of the data challenges the idea of a substantial East-Coast divide. With the exception of the San Jose MSA, large coastal MSAs aren’t that vastly different from each other. Florida’s visual effects are the result of looking at counts instead of proportions.

As the above table highlights, there is a small but consistent relative scarcity of educated men across all communities. Which leads to my final observation, one of the most quirky things about Florida’s piece is that it doesn’t clarify for the reader even what a desirable scene looks like…is it equal ratios or a tip in one direction or another. What is a policy maker to do? I suggest that it is probably more beneficial for public and social sector leaders to address the lagging rates of male higher education instead of chasing after an elusive or ill-defined”hot singles scene.”

BTW, I am generally a huge fan of Florida’s work.