# Fun with Data: A Look at City Budgets

While I was trying to see how far Texas Roadhouses are from downtown in larger cities (they are all really far, but so are most casual sit down restaurants), I came across an interesting data set that contains the 2016 population sizes and annual budget for America’s 100 largest cities. This data also contained the party affiliation of the mayor (Democrat, Republican, Independent or NA). I decided to play around with the data to see if there were any cities spending an absurd amount and if the usual conservative rallying cry that Democrats spend too much is really true (which I expected to find in this 2016 snapshot). But before getting into the politics, let’s just take a look at the raw data below:

Check out New York City!! The Annual budget there is \$73 billion dollars for the largest U.S. city of 8.4 million residents. The black line is the least squares regression line for fitting population to the annual city budget. Since New York City is such a huge outlier in terms of money spent and population, I decided to refit this line and look at the data sans NYC below:

This refit of the data show two huge outliers in terms of money spent per capita: Washington D.C. and San Francisco, which spend \$10.1 and \$15.8 billion dollars annually.  I then wondered since D.C., San Fran and NYC have Democratic mayors and populace, is that the driving force behind this discrepancy? Actually It is not based on the results of a least squares regression with Mayoral party and Populace regressed on annual budget, you can see in the results below that only population is a significant predictor of annual city budget.

Where Y is the city budget, X is the city population and Dem indicate whether the mayor (and the constituency likely) is a Democrat. So it doesn’t appear that there is a party factor in determining how much money a city spends. I leave you with the following plot of all cities (sans NYC and Independent/NA mayors) population and budget (subsetted by blue for Democrat and red for Republican mayors).

When you see if this way, other than San Francisco and D.C., it’s pretty clear that only population is driving city expenditures and not the party of the city’s mayor. Also it just so happens that cities with more people tend to have Democrat mayors, but that’s a story for another time. The takeaway here is that political generalizations are often misleading (at least of the city level), it is more of a counterfactual thing: cities with larger populations spend more and they just so happen to have mayors who identify as democrats more frequently.

If you enjoyed this post, take a look at some of my other Adventures in Statistics.

-Andrew G Chapple