From Atlantic Cities:

The more cities spend on bike infrastructure, the more important it becomes to make sure that money is spent wisely. One way to measure success in this area is to lay down bike lanes or paths and see if ridership grows. Another is simply to ask riders what facilities they prefer. Both approaches have their drawbacks: The former assumes transportation officials know best and relies on correlations that hopefully reflect causations; the latter may put too much emphasis on hypothetical options and not enough on actual behavior.

A potentially more instructive way to see what riders want from a bike route is to follow riders, in real-time, as they choose a bike route. A trio of transportation researchers led by Joseph Broach of Portland State University recently did just that. In an upcoming issue of Transportation Research Part A, Broach and company report a series of nuanced rider preferences that could help designers create more comprehensive bike facilities and help cities implement these facilities more efficiently.

The researchers selected 164 experienced riders in the metropolitan Portland area and clipped a G.P.S. unit to their bikes. The G.P.S. network was rigged to follow them no matter where they might go — from major streets to informal cut-throughs — and to document route details, such as overpasses. Broach and colleagues also worked with Oregon Metro and the City of Portland to gather information on traffic volumes, turns, elevations, bike infrastructure, intersection details in various corridors.

Altogether the researchers compiled data on nearly 1,450 actual bike trips. (They tracked commutes and non-commutes, but not pure recreational riding.) But the information on where riders went wasn’t terribly useful unless the researchers also knew where riders hadn’t gone. So for every trip they came up with an average of about 20 alternatives, to see if they could determine what riders really looked for in a route.

Into list mode:

  1. The shorter the better. This one isn’t much of a surprise: all else being equal, cyclists preferred shorter routes. Every 1 percent increase in route distance reduced the chances that route was taken by 5 or 9 percent (non-commute and commute, respectively). Cyclists also showed a natural nose for distance; half of all observed trips were only slightly longer than the shortest possible route from A to B (within 10 percent).
  2. Turns are a turn-off. A notable exception to the shortness rule was a route that involved turns. Each additional turn per mile was like increasing route distance (adding 7 percent for non-commutes; 4 percent for commutes) in terms of rider preference. In general it didn’t matter whether the turn was a left or a right — with the exception of difficult lefts, which were a notably large detractor.
  3. Mixed signals. Broach and company found that traffic signals generally decreased a route’s attractiveness but that wasn’t the case in high-traffic corridors. In these cases lights presumably either decrease delay or increase safety (or both). For every mile of a non-commute trip, cyclists were willing to go 10 percent farther to avoid an unsignaled crossing at a busy intersection, and 16 percent farther to avoid a left turn at such intersections (6 percent and 9 percent farther, respectively, for commutes).
  4. Uphill battles. Hills are another exception to the distance rule. The researchers were notably surprised at the lengths cyclists will travel to avoid an upslope: non-commuters routinely went 1.7 miles flat rather than 1 mile on a mild (2-4 percent) upslope; commuters went about 1.4 miles. Even greater slopes invited even greater detours.
  5. Everyone hates traffic. Where bike lanes didn’t exist, traffic played a major role in discouraging bike riders. For non-commute trips, riders used a road with heavy traffic (20,000 cars a day) only if the best alternative was more than twice as long or on a steep hill. This wasn’t such a huge factor in Portland — where most high-traffic streets have bike lanes — but in other cities it might be a major deterrent to people biking at all.
  6. Paths over lanes. The most intriguing find was how much riders preferred off-road bike paths or traffic-calming bikeways (a.k.a. “bike boulevards”) to traditional arterial bike lanes. In terms of rider preference, a path was like reducing a route 26 or 16 percent (non-commute versus commute) and a boulevard was like reducing it 18 or 11 percent (same). A bike lane on a high-traffic street, meanwhile, was no more attractive than riding a low-traffic street with a lane.

The last point might offer cities the most guidance in terms of expanding bike infrastructure. Bike boulevards or off-road paths, which in many cities will be less expensive than bike lanes to implement, may be very attractive to riders. That isn’t to say bike lanes aren’t valuable — on the contrary, riders might avoid busy streets entirely if not for lanes — but it does suggest that, even for experienced riders, routes away from car traffic are extremely desirable:

In the final specification, bike lanes more or less exactly offset the negative effects of adjacent traffic but had no residual value of their own. …

All else equal, the estimation suggests cyclists are willing to go considerably out of their way to use a bike boulevard or bike path rather than an arterial bike lane.

Now for the caveats. The biggest was that the current sample of riders was far from random, and, in fact, hand-selected for experienced riders. While it’s good to know why regular riders ride, cities hoping to encourage bike use must also learn what’s important to potential riders. In other words, what is it about some routes through a city that keep people from riding at all. It’s also good to keep in mind that Portland’s bike network is far more expansive, and therefore offers far more route options, than those in other U.S. cities.

On the whole, though, the work provides a blueprint for metros that want to understand their own network of riders, and illuminates key subtleties that might otherwise go overlooked when cities direct their limited bike resources. At a more individual level the research might inspire cyclists to develop more comprehensive digital route-finders. Calling all app makers.