The rising cost of gasoline has become a problem for many people who drive to work every day, or even drive in general, for that matter. Everyone talks about doing something about it, but is it actually having an effect on people? In order to find out, the top fifty public transportation website domains were taken from the "Bus and Train" category, sites such as amtrak.com, greyhound.com, and the ever-popular fungwahbus.com, where you can book a Chinatown bus from Boston to New York for only fifteen dollars. The number of weekly unique visitors (UVs) to these sites were found for the first 27 full weeks of 2008, from January 6th to July 12th, and compared to the average gas price in the United States for each week.
Several portions of the data seemed very interesting:
- At first glance, one might think it was odd that the UV count actually decreased through January, but upon getting a closer look, it turns out that gas prices fell slightly that month, as well, dipping back under $3.00 and in turn inspiring more people to drive.
- One of the more notable occurrences is the huge UV spike in Week 6. Week 6 covers February 10-16, so with Valentine’s Day being the 14th and President’s Day being the following Monday, February 18th, its obvious why so many people were looking to take a bus or train that weekend. Not to mention that gas had just jumped back over the $3.00 mark.
- Throughout April and May the relation seems very close, until gas passes $4.00 per gallon in Week 23, causing a small jump in public transportation UV count.
The most compelling sign of the correlation between gas prices and use of public transportation would be the aptly-named "correlation coefficient," which indicates the strength of a linear relationship between two variables, falling somewhere between 1 and -1. For example, a correlation coefficient of 1.00 would mean that the two sets of data were exactly the same, and a correlation coefficient of —1.00 would mean that the data sets are exactly opposite. When comparing these particular two sets of data, the correlation coefficient between them was found to be an astonishing 0.9725! So there is a direct correlation; people are responding and actively doing something about gas prices. It makes you wonder why, in places like Rhode Island, they are talking about cutting the daily number of buses running because they aren’t fuel-efficient and gas is too pricy. With demand for buses up, cutting them would just put more drivers on the road and waste even more gas.
You have to wonder though; will people continue to react so directly to the increase in gas prices? It is interesting to note that in the first two weeks of July, the UV count begins to drop. Could this be because people have become used to paying over $4.00 for a gallon of gas and are starting to accept it? In my experience, this is unfortunately true.