Why Some Requested Divvy Stations Get Installed and Others Don’t

Screen shot from the Divvy suggestion map. Blue pins are existing stations and green bubbles are requested locations.
Screen shot from the Divvy suggestion map. Blue pins are existing stations and green bubbles are requested locations.

A new study by Greg P. Griffin and Junfeng Jiao of the University of Texas at Austin, published in the journal of the American Planning Association, found that while Chicago and New York have asked for suggestions from residents on where to install new bike-share stations, that input has had little influence on where the docks are ultimately placed.

The study looked at Divvy’s second major expansion in 2015, when 175 new stations were installed, bankrolled by city and federal money. The Chicago Department of Transportation had previously asked for recommendations on where the docks should go via public meetings, workshops, and an online suggestion map. “Chicagoans gave great suggestions for the locations of new stations, and we look forward to placing them where they were requested,” Transportation Commissioner Rebekah Scheinfeld said at the time.

In the end, however, the researchers found that only ten percent of the Divvy stations were installed at or near locations requested by residents on the online map.

The researchers didn’t blame CDOT for the discrepancy, but rather noted that there were practical reasons why the department didn’t honor many of the requests. They pointed to the need for the baby-blue bikes to compliment the existing transit network, and mentioned that issues with the built environment, such as too-narrow sidewalks and obstacles like fire hydrants and utilities, can prevent installation.

And then there are human factors that can prevent installation, such as opposition to replacing on-street parking with a station, and Not In My Back Yard resistance to having a public amenity near one’s home. In one case, early in Divvy’s history, a Lakeview condo association actually sued to have a “hideous” station removed from in front of the building.

Griffin and Jiao found that CDOT compliance with Divvy siting requests varied by neighborhood. For example locations requested downtown were twice as likely to get docks than those in outlying neighborhoods – 12 percent versus six percent of suggested spots — possibly because centrally located stations get more use. The study pointed out that the Divvy suggestion map doesn’t offer any info on guidelines for station placement, such as the rule-of-thumb that residents are most likely to use bike-share if the stations are within a five-minute walk of their starting points and destinations.

The percent of Divvy station requests honored downtown in 2015 was twice as high as the percent honored in outlying neighborhoods.
The percent of Divvy station requests honored downtown in 2015 was twice as high as the percent honored in outlying neighborhoods.

But ultimately, the researchers argued that CDOT “made great efforts” to get input from Chicagoans, even if the final station placements don’t seem to reflect that. “The online [map] enabled residents to take direct action in planning their cities, rather than just commenting on the ideas of planners — or waking up to discover a docking station had been built outside their door.”

However, they also asserted that CDOT should have done a better job of informing residents about the factors guiding their final decisions on station placement. “That helps residents make informed recommendations that are more likely to be implemented.”

While public input may have played a relatively minor role in the 2015 Divvy expansion, it was a significant factor in where the city ultimately decided to place the latest wave of 36 stations, installed in late 2018. However, this wasn’t due to CDOT honoring more of the online requests this time.

While in 2017 the department had announced plans to add more station density in low-income communities on the South and West sides, in the end most of the new docks went to parts of town that already have plenty of stations, including many affluent or gentrifying neighborhoods.

The 36 new station locations (black dots.)
The 36 new station locations (black dots.)

Why? The decision was made during the thick of the summer 2018 Divvy theft crisis, when some 500 bikes went missing. (Divvy has since reinstalled security hardware in the docks, which seems to have solved the problem.)

“During the coordination process with all of the stakeholders, aldermen, and [the Illinois Department of Transportation], some of those locations were pulled from this list, which is why it may not look exactly like what we had wanted it to when we started,” CDOT consultant Amanda Woodall told me last August.

Factors in the community resistance to getting new stations included some residents’ views that stolen Divvy bikes are a distraction for the police in high-crime neighborhoods. Others may have felt that the then-unsecure bikes were a temptation that could lead to more local residents being jailed, as well as innocent Divvy riders being unfairly stopped by the police.

It goes to show that the factors behind the city’s decisions on siting Divvy stations can be complex. But it would be great if CDOT could make that decision-making process more transparent.

donate button
Did you appreciate this post? Consider making a donation through our PublicGood site.

  • what_eva

    “In the end, however, the researchers found that only ten percent of the Divvy stations were installed at or near locations requested by residents on the online map.”

    I don’t think that is a correct statement. I think the correct statement is:
    Only ten percent of the suggested locations got a Divvy station installed within 100 ft.

    The way you’ve phrased it, you’re stating that CDOT ignored the suggestions completely and installed 90% of the stations nowhere near where people wanted them. That is completely wrong.

    The reality of the suggestion map is that people put pins at pretty much any intersection and even mid-block. With over 4500 suggestions, obviously CDOT couldn’t put a station at every pin.

    Which brings me to the study’s biggest flaw. Using 100 feet for “nearby” is ludicrously short. You mention the rule of thumb that people want a station within a 5 minute walk to find the system useful. Even at a drunk zombie shuffle I can cover 100 ft a lot faster than that… 100ft is a distance you cover in seconds. A “short” block in Chicago is 330 ft, so even pins with stations a short block away would fail by this metric. eg, my own closest station which I would refer to as “around the corner” is a short block away.

    To give a specific example, there are a bunch of pins around the intersection of Clark and Diversey. For likely a number of reasons, CDOT decided that putting a station at this very busy intersection wasn’t possible (3 bus lines with stops and morning bus lanes on Clark made on street difficult and busy but not super wide sidewalks made that a bad idea as well). So they put a station as close as they could. . Currently it’s south a short block at Schubert (this station has moved around a bit due to construction as well). ie, it’s within 500 feet, plenty close for anyone going to the Landmark theaters or the axe throwing in the Century mall or the shopping all over.

    I also hope the study wasn’t using the station placements on the suggestion map, as those haven’t been updated.

    Given the 100ft distance that was chosen by the study’s authors, I have to question the entire thing. Selecting a distance that short makes me think the authors are trying to make the systems look bad.

    As to downtown vs outlying, since the stations downtown are often 1/8 mile apart vs 1/4 mile outlying, that makes perfect sense, especially given the short distance. I suspect the numbers would even out a bit if a more reasonable 1500-2000 ft was selected

  • Tooscrapps

    There really is no reason that there shouldn’t be a Divvy station at Clark/Diversey/Broadway. The fact that three bus lines converge there is more a reason to place a station there than not. While the current station isn’t that long of a walk, sometimes 1-2 minutes is the difference for catching a bus that might have 10 minute lead times.

    Some nuances, since I’ve seen this station migrate all over the place:
    – Technically the Clark AM bus lanes start after Diversey, so the stop can and should be placed north of Diversey on Clark. There is open curb (No stopping/standing) the entire length of the theater to Stan’s donuts. I would say right by the Aveda Institute is the sweet so as not to interfere with the right turn lane.

    – The sidewalk south of the old Walgreens certainly is wide enough and there was a station there for some time. It was moved for construction this summer, but never moved back. I’m not sure why. That is the best spot at this intersection and I hope it moves back.

  • Jeff H

    The Clark & Drummond station has moved quite a bit. It was originally the Hampden Court station. It was moved from Hampden Court to the corner of Broadway & Diversey due to the construction of those bazillion dollar condos. Then it was moved from Broadway & Diversey to Clark & Drummond because of the upcoming construction of Lincoln Park Plaza at the old Walgreens site. I was told by Divvy support that it will eventually be moved back to Hampden Court once construction has been completed there.

  • Tooscrapps

    That’s unfortunate. Really wish they’d move it Century or back to Walgreens.

  • Greg Griffin

    Hi what_eva, original study author here! Our definition of the 100 foot distance is as a “critical distance for a station location to meet the intent of a participant suggesting a station on the PPGIS platform”. So, it’s really at not nearby. I looked at a ~1/4 mile distance as part of the study (not documented), but this also blurs the whole notion of allowing the public to pinpoint station locations. We’re evaluating public involvement, not overall access to stations.

    The study goes into more detail on how we developed the criteria, and another geostatistical method we used to show whether the proximity of stations to suggestions was statistically significant. Original journal article is online here and a free pre-print is here. The study also goes into more detail about the challenges and constraints planners had to work with, in addition to recommendations for improving the overall practice as a form of engagement.

  • johnaustingreenfield

    In the study authors’ own words: “Ten percent of docking stations built through 2015 were located at or near the spots residents identified on the interactive map.” https://www.chicagotribune.com/news/opinion/commentary/ct-perspec-divvy-bike-share-docking-stations-public-input-0115-20190114-story.html

  • what_eva

    Apologies for implying they were your words. Because they’re *way* off.


Condo Association Sues to Have “Hideous” Divvy Station Removed

Update Friday, 15:09: Judge Kathleen Kennedy denied the plaintiffs’ request for a temporary restraining order because they didn’t state a clearly ascertainable, protectable right and cannot sustain the merits of their complaint (if it went to trial, as they requested). The protectable rights they tried to ascertain and state were that the Divvy station would […]

Why Did Divvy Stations Dance Around River West, Lincoln Square?

Divvy bike-share stations were designed to be easy to move around, with their modular construction and off-grid solar power. Sure enough, plenty of Divvy members have had their routines disrupted by station moves lately: 8,000 Divvy members received word this year that stations they’d recently used were on the move. One Divvy member forwarded two […]