The Citi Bike bike-share system in New York City has been aggressively expanding its coverage in 2016. Bike-share systems work by building clusters of bicycle docks within its 'service area'. This cluster has a border outside of which there are no stations. To understand where pent-up demand might still exist, we investigated whether there was increased usage of bike share stations along the border in Northern Manhattan as of June 2016. We selected 10 stations roughly along 84th Street. We compared average usage at these stations with 20 stations directly south of this border, what we call the 'non-boundary' stations. We hypothesize that boundary stations will see higher average usage than non-boundary stations since anyone living north of the border will likely use and drop off bikes along the system's edge.
We use Citi Bike ridership data for June 2016 downloaded from Citibike's official data platform. We reduce the data to only include rides starting midnight, Thursday June 3rd to Thursday, June 30th at 23:59:59 so as to only have 4 instances of every day of the week (4 Mondays, 4 Tuesdays, etc). We then create a count of trips originating per station for our border and non-border stations. To control for varying sizes of stations (different stations have a different number of bike docks), we take the count of rides originating per station and divide by the number of docks at that station. This gives us normalized count of usage.