Introducing "The Dataset of the Day" – Feeling the Pinch at the Pump

With all the data we are pushing into Finder! sometimes it is easy to loose track of the interesting pieces unless you are searching for something specific. So, we thought it would be fun to post up a “dataset of the day” on the blog. The goal is to let folks know about new datasets or content that is relevant to a current event.

Since Memorial Day officially kicked off the summer driving season we thought it would be useful to map the current state of gas prices. We all know that gas prices are sky rocketing, but who is getting hit the hardest and suffering the largest increases. Today Bill (one of our resident data gurus) grabbed the latest data from AAA, calculated percent change from a year ago and loaded it into Finder! You can access the data here.

Next he took the data out as a shapefile and whipped up a quick thematic map in a GIS application. The first map is of prices with red being the highest and green being the lowest:

For the next map he took the percent change he calculated to illustrate the states with the largest increase in price (shaded red) and those experiencing the least price increases (green):

When it comes to the most expensive gas California, New York, Connecticut, Michigan and Illinois lead the pack and they are spread fairly evenly across the country. On the other hand if you look at percent increase in prices there is a large concentration of reds in the Midwest and South. This is especially stark in the manufacturing rust belt states of Michigan, Ohio, West Virginia and Indiana. The unfortunate things is these are the same places that have been hardest hit the recession and unemployment as you can see in the map below of percent unemployed for February 2008 (data here):

While there is no cause and effect between “percent unemployed” and “percent change in gas prices” it does illustrate the sad reality that those with the least ability to pay for an increase in their gas budget are being hit the hardest. Although if you wanted to run a correlation analysis you could pull the files out as .CSV and go to town.