How Uber Works Together With Southern California’s Growing Rail Network

Rik Williams
Uber Under the Hood
4 min readJun 30, 2017

Think of the last time you took a train or the subway. How did you get to the station?

If you were in a dense urban area like Manhattan or central London, where stations can be found every few blocks, you probably walked there. On the other hand, if you started out in a suburban or rural area where the nearest station was a few miles away, you may have taken the bus to the station, driven to a park-and-ride lot, been dropped off by someone else, or even biked over. In the past few years, ridesharing services like Uber have emerged as another way to bridge the last few miles between public transit and the places where people live, work, and play.

Uber’s Policy Research team is keenly interested in (and has written several previous posts on) how ridesharing complements public transit. However, the data are not always straightforward: train and bus stops are often located near homes, apartment complexes, restaurants, retail, and other destinations. When our data show Uber riders being dropped off near a subway station, are those riders actually going to the subway, or to the office building above the subway?

One way to solve this puzzle is to look for clues such as substantial changes to the transportation network — for instance, when new train stations open — and see if any simultaneous changes appear in Uber’s trip data. Last year, Southern California served up not one but two such “natural experiments”: six new stations were added to the Gold Line in the San Gabriel Valley northeast of Los Angeles, while seven new stations extended the Expo Line from Culver City to Santa Monica on the coast.

Map of Los Angeles Metro Rail extensions in 2016 [1]. The six new Gold Line stations opened on March 5, and the seven Expo Line stations opened on May 20.

If Metro Rail passengers use Uber to bridge the last few miles between stations and their destinations, we’d naturally expect to see an increase in Uber pickups near the newly-opened stations, and perhaps some other changes at existing stations. To investigate this, we measured the number of Uber pickups within 100 meters of the new stations (along with the previous “end of the line”) 1 month after each extension opened, and compared these to the number of pickups the week before the opening [2]. What do the data show?

Difference in the number of weekly Uber pickups within 100 meters of stations 1 month after the Expo Line extension opened on May 20, 2016, compared to the week before and corrected for long-term growth. New stations are to the left of the dashed vertical line.

As expected, there was a surge in Uber pickups at the new stations, indicating that passengers indeed combine their light rail trips with Uber. Interestingly, we also see a substantial decrease in Uber pickups at Culver City, the last stop on the Expo Line before the extension opened. One likely explanation is that some passengers can now get much closer to their destinations thanks to the extension — in other words, someone who lives in Rancho Park no longer has to find a ride from Culver City to their home, but can take the Expo Line nearly all the way.

The Expo Line traverses the relatively high-density (and heavily congested) Downtown LA-Santa Monica Corridor. What about the Gold Line extension, serving the San Gabriel Valley suburbs [3]?

Same as the previous chart, for the Gold Line extension which opened on March 5, 2016. In this case, new stations are to the right of the dashed line.

In this case, total Uber pickup volume is lower (likely due to the lower population density of the surrounding neighborhoods), but we see essentially the same pattern: pickups increase at new stations and decrease at the previous last stop (Sierra Madre Villa). While the similar pattern in the two very different urban environments is notable, what’s particularly striking is how quickly travelers on both lines adopted this combination of rail and ridesharing: just a month after each extension opened, hundreds of riders per week were taking Uber from the new stations. In both cases, the increase in Uber trips to new stations far outweighed the decrease seen at the previous last stop, suggesting a net increase in ridership for both Uber and Metro Rail.

Southern California’s reputation as the epicenter of car culture belies the degree to which it depends on transit: in terms of ridership, Los Angeles County has the second-largest light rail and bus systems in the United States [4]. In the coming decades, as the region dramatically expands its light rail network, there will be numerous opportunities to develop multimodal options that seamlessly integrate with rail. The development of more attractive alternatives to driving (and, crucially, more walkable neighborhoods around transit) will, in turn, empower more Southern Californians to give up a car. Alongside buses and bicycles, services like Uber provide rail passengers who need to travel farther than a short walk from a station with another reliable option to bridge the “last mile” — facilitating easier access to rail, lowering parking demand, and promoting more livable, sustainable cities.

Uber trips starting within 100 meters of Expo and Gold Line Extension stations during the first week of May, 2017. Trip endpoints have been jittered for privacy.

Notes
[1] Geospatial data courtesy Los Angeles County Metropolitan Transportation Authority,
http://developer.metro.net/introduction/gis-data/download-gis-data/
[2] Pickup data before and after the Expo Line opened are from May 8–14 and June 19–25, 2016, respectively.
[3] Pickup data before and after the Gold Line opened are from Feb. 21–27 and Apr. 3–9, respectively.
[4] American Public Transportation Association,
Fourth Quarter 2016 Ridership Report

--

--

Rik Williams
Uber Under the Hood

Data scientist @Uber Policy Research. Time also spent in US foreign assistance, astronomy, hiking, silicon wafers, fast food, and poorly-played music.