Uberdata: How the city of Washington responds to the government shutdown
Greetings humans! Bradley Voytek coming back with a look at the US government shutdown of 2013. Man, politics, am I right? With your side and my side fighting back and forth. The nerve of those people doing that thing.
In all seriousness, the US government shutdown has had an obvious effect on a whole lot of people, but quantifying those effects can be difficult. We at Uber were musing about how such a major financial and employment shift such as this might affect the way people move around a city. And of course, what better city to examine these effects than the very epicenter?
In this post I look at how the US government shutdown has affected the ridership patterns in Washington, DC.
The first thing I wanted to do was look at a “normal” week’s worth of data for Washington, DC. Here’s a plot of all of the trips going between the different DC neighborhoods:
Briefly, neighborhoods are outlined in grey with the size of the circle in neighborhood center representing the proportion of rides going into that neighborhood. Circles with the same color are part of a related subnetwork. The weight of each line reflects the proportion of rides going between the connected neighborhoods. (For detailed methods, look at our post here.)
As can be seen, most trips move between just a few neighborhoods, however a large amount of the business day trips move between Capitol Hill and Downtown.
Here’s what a week’s worth of trips that go to Capitol Hill and Downtown look like, pre- and post- shutdown:
Note that the shutdown (the red line) occurred on Tuesday, 01 October 2013. This means that the red Monday (30 September 2013) is before the shutdown. See how much it looks like the black line Monday (23 September 2013)? I include this here to highlight just how big of an effect the shutdown had on the proportion of rides going into Capitol Hill or Downtown.
Also, see those horned spikes every workday morning and afternoon? That’s people using Uber to and from work. But as soon as the shutdown kicks in, the number of Uber work trips drops dramatically.
There’s another way to look at this, as well.
We can look at how the number of rides between neighborhoods changed as a function of the shutdown. In the graph below, blue links mean that the number of trips between those neighborhoods increased after the shutdown. Red links mean that the number of trips decreased.
The US government shutdown significantly changed Washington’s habits, with far fewer people moving between Capitol Hill and Downtown, and fewer people taking workday rides.
Obviously this stalemate has affected a huge number of people. It’s interesting to be able to use our data to uncover some of these effects. But, for everyone’s sake, let’s hope this “interestingness” doesn’t last too much longer.