In 2022, an entire line of the Boston subway system was shut down for about a month for track repairs and related work. The shutdown took place from August 20th through September 18th during the time when students (a significant part of the Boston population) typically move in. The entirety of the Orange line and a portion of the Green line were affected by the shutdown. The Orange line is a heavy rail subway line separated from traffic along its entire extent, and is typically the second busiest line in the city’s subway system. The Green line has smaller, light-rail style streetcars running underground in downtown and above ground further out, mixing with street traffic along some of its branches.

During the subway closures, the local authorities put in effect a number of mitigation measures, including free shuttle buses and free access to the commuter rail lines that runs along the same right of way as the Orange Line for much of its extent. One additional initiative was availability of free, month-long BlueBikes memberships. BlueBikes is the local Bike Sharing System (BSS), and the regular cost of monthly membership was $26.75 at the time. This analysis examines whether the free bike share memberships were successful in helping to address transportation deficits caused by the subway closures. We find that while BSS ridership in the areas most affected by the Orange line shutdown did increase substantially in proportional terms, overall, the lion’s share of the ridership increase during this period went to rides that were unrelated to subway closures. At the same time, the large boost in BlueBikes ridership indicates that BSS consumption is price-elastic, and that appropriate reductions in ride cost may generate more BSS ridership in the future.

Background

A previous post detailed an analysis of Boston Bike Share System ridership via a regression model. One period where the BSS ridership predicted by the model does not match the actual ridership is the period of late August through early September 2022:¹

The difference is not insignificant: the actual number of rides exceeded the number of rides predicted by the model by over 50%. To appreciate the ridership increase, we can also compare the the number of BSS rides between August 19th and September 19th in 2022 with another four-week period such as of the following:

• Same dates in the previous year, 2021
• Same dates in the next year, 2023
• Thirty days immediately before August 20th²
• Thirty days after September 18th
• Fifteen days before August 20th and fifteen days after September 18th

Using any of the above comparison periods, the number of rides during the subway closure in 2022 exceeds expectations by at minimum 153 thousand rides or 31%. In fact, the number of rides in this period in 2022 has not been matched in the Boston Bike Share system before or since:

As noted in the introduction, two things happened during this period of increased BlueBikes ridership: the Orange line and parts of the Green line were closed, and the city of Boston offered free 30-day BlueBikes memberships as a subway closure mitigation measure. In the next section we try to consider in more detail whether mitigation strategy of giving away free BSS memberships was successful.

Did the intervention work?

We can start by establishing that the BSS ridership increase during the Orange line shutdown was related to the free 30-day BlueBikes memberships intervention. Looking at the trip data we can see that it was the membership rides that accounted for the ridership increase, not the “causal” (non-member) rides:

The above graph also shows that the number of non-member rides during the Orange line shutdown period actually decreased relative to the previous thirty days. This suggests that some of the occasional, non-member BlueBikes riders may have become members, possibly due to the availability of free membership.

Before digging further into the data, we should note that in some ways, the intervention (free BSS memberships plus whatever changes were made in the rebalancing schedule and bike availability) was very successful: the Boston bike share system experienced an unprecedented utilization that has not been matched since. Having noted that, the next thing to consider will be the location of the bike share rides during the relevant period.

We will start by identifying BSS stations that were likely to be used for rides that can substitute for trips involving the Orange line subway and the inoperative Green line stations. Here we define these “substitution stations” as BSS stations that match both of the following criteria:

• Are within 0.5 miles of a closed subway station
• Are closer to a closed subway station than any open subway station

Do note that the distances referred to here are measured linearly, or “as the crow flies,” not via walking paths, so more realistic distances are likely to be longer in many cases.

In contrast to the “substitution stations”, the BSS stations that do not fit the above criteria can be called “non-substitution stations.” The following map illustrates the distribution of substitution versus non-substitution stations near the subway terminals that have been closed: large color circles are open subway stations, large white circles are closed subway stations with the circle border color indicating the closed station’s line. Small gray circles are non-substitution BSS stations and small black circles are substitution BSS stations:

Having made a distinction between substitution and non-substitution BSS stations we can examine the difference in the number of rides during the subway shutdown period versus different comparison periods. The following graph shows the relative increase in rides by BSS station type (“substitution” or “non-substitution”) across several comparison periods:

Looking at the above graph, we can see that proportionally, the number of rides starting or terminating at substitution BSS stations increased more than the number of rides between other stations, regardless of which comparison period is used to establish the baseline. This suggests that free BlueBikes memberships may have helped generate rides that could have been used to mitigate the shutdown of the Orange line subway. On the other hand, the substitution stations are not the most popular stations in the system, so if we look at the absolute increases in ridership, we see that the majority of the additional bike share rides during the Orange line shutdown went to trips that are less likely to be Orange line replacement trips, also regardless of the comparison period:

To put this in perspective, it is instructive to see how the increase in BSS ridership was distributed among substitution versus non-substitution BSS stations:

We can see that regardless of what comparison period we use for our baseline numbers, during the Orange line shutdown, rides that did not involve substitution stations outnumbered the rides that did by a factor of more than 2. The basic reason for this was hinted at above: in general, the number of increased rides at a BSS station during the Orange line shutdown is roughly proportional to the station utilization prior to the subway shutdown, and stations near the Orange line are not the busiest stations in the system. The following plot illustrates this by plotting BSS station popularity (the number of rides originating or terminating at the station) on the x-axis against the increase in the number of rides the station saw during the Orange line shutdown (versus the comparison period, which here is the same dates in 2021):

We can see that substitution stations “outperformed” non-substitution stations during the subway closure period: relative to their popularity, they experienced a larger ridership increases during the subway shutdown; however, as we saw above, even using the most favorable comparison set, more than 60% of the additional rides likely to have been generated by the free BSS membership mitigation went to rides that are not likely to have been replacing the lost subway trips.

Conclusions

In August/September of 2022, Boston gave away free bike share system memberships as a part of a mitigation strategy during a partial subway shutdown. During this time, the number of trips taken on BlueBikes reached a record number that has not been matched yet. While the BSS stations located along the inoperative subway stations saw a larger proportional increase in utilization than other stations, the bulk of additional trips during this time were not subway replacement trips.

The relatively modest increase in bike share usage which could be substituting for the Orange line trips may have been due, in part, to the availability of other relatively good subway substituion options:

• The shuttle buses replacement ran along a bus lane delineated with orange traffic cones along much of its length, and therefore relatively free of car traffic.
• Commuter rail access for branches running along the Orange line was essentially free for everyone. While the services on the commuter rail lines is not nearly as frequent as that of the subway,³ the service along the Orange line runs like an express line would, making it for faster service in and out of center of Boston.

As far as the general BSS ridership increase is concerned, while we don’t know what other measures (such as special rebalancing, or additional bikes from other parts of the system) contributed to the BSS “boom” during the Orange line shutdown, the fact that free bike share memberships drove, at least in part, a substantial increase in usage suggests a certain amount of price elasticity on the part of (potential) BSS customers. This indicates that with appropriate price-reducing measures, the BSS ridership can be substantially increased. Conversely, increased prices may result in less system utilization.

The next section offers some speculations on the recent or upcoming factors affecting the BlueBikes pricing.

Early 2024 Pricing Factors

Boston Bike Pass

The record BlueBikes ridership during the free monthly membership program did not go unnoticed by the city of Boston; the city introduced BlueBikes memberships called “Boston Bike Pass” specifically citing BlueBikes ridership during the Orange Line shutdown. The program includes a nearly free tier for income restricted residents and a greater than 50% discount ($60 versus $133.50 yearly) on a yearly membership for non-income restricted denizens. The subsidy to non-income restricted residents is limited to people who have not been BlueBikes members in the previous three years.

The Boston Bike Pass is only available to residents of Boston proper, which excludes the nearby communities of Cambridge, Somerville, Brookline, Malden, and others. These surrounding communities are often thought of as part of “Boston”, participate in the Boston-area BSS, and together have a population that is nearly equal to the population of Boston itself. As the discounted BlueBikes membership initiative is funded by the city of Boston, the restriction to Boston residents makes sense. Surprisingly, it seems that the program should reach a substantial part of BSS ridership: almost half (48%) of the BSS trips during the Orange line shutdown period were entirely inside Boston proper, and another third (29%) were mixed trips, either originating or terminating in Boston itself.

Nonetheless, I have some doubts that the Boston Bike Pass program will have an impact that is anywhere near that of the free one-month membership during the subway shutdown in 2022 for the following reasons:

• Students, who comprise a large part of Boston fall, winter, and spring population, and who often retain their parents’ address as their permanent address, may be put off from obtaining the discounted memberships by the residency requirement. It is true, however, that the Boston Bike Pass signup form does not require any residency verification beyond a Boston address.
• People who have had a BlueBikes membership in the past three years are not eligible. This would be particularly deleterious if people who took advantage of the free membership in 2022 are unable to participate.
• The signup form itself may create the kinds of barriers to discounted memberships that were not in place for the free membership month in 2022.

February/March Green Line Shutdown

In 2024, in advance of the Green line closures BlueBikes was offering a $20 credit to any Boston resident.

Need a ride with the Green Line closure?

Thanks to @BCBSMA and @CityOfBoston, all Greater Boston residents can claim a FREE $20 Bluebikes credit for pedal or ebikes by using code MBTA2024 in Bluebikes app. Take advantage while supplies last. #MBTAclosure https://t.co/HXmQsdAIg4 pic.twitter.com/goOmonwgmF

— Bluebikes (@RideBluebikes) February 21, 2024

The coupon code program was exhausted within two days, and while popular, is also unlikely to replicate the success of free monthly memberships: it is even less targeted (the $20 could be used at any time, not just during the Green line closure), and doesn’t have the “all-you-can-eat” benefits of the monthly membership, which would have cost the sponsors of the discount only a bit more ($29 versus $20 credit)

March Price Increase

While the Boston Bike Pass program and the BlueBikes coupon during Green Line closure are aiming to lower the price of using Boston bike share system, the non-discounted prices for BlueBikes are going up: in March 2024 BlueBikes membership price increase goes into effect. My suspicions are that the negative pressure on ridership from the price increase will outweigh the effect of the price discounts and that BSS ridership will increase slower than it would otherwise, especially if the MBTA makes measurable progress against the “slow zones.”

Small Print

For gory details see the details in the previous post: the data was selected and treated in the same way. I’d like to thank Wes Drew, Matthew Glidden, David Lustig, Stacey Moisuk, Jackson Moore-Otto, and Sabina Wolfson for helpful comments and suggestions on this post.

---

Comments