Wednesday, July 20, 2016

Cars parking in bike lanes isn't the urbanist meme we need

A specific point of focus for many internet bicycle advocates is the widespread problem of cars parking in bike lanes. The classic example is a double-parked delivery vehicle in an unprotected bike lane.

I'm hesitant to critique this trend--bike advocates are doing good and generally deserve more power and support--but recently I've noticed a streak in some of the bike lane media that annoys me. Every interest group has an inclination and an incentive to catastrophize their issue and to demonize opponents, and I've started seeing hints of this coming from some cars-in-bike-lanes advocacy.


This issue's relative prominence is largely due to its useful strategic qualities. Most obviously, cars parked in bike lanes are much easier to photograph and publicize compared to unsafe behavior by moving cars. The violation is easy to understand, unambiguous (cars are physically occupying space set aside for bikes), and have a nice channel leading to formal action (via license plate numbers). In a way, the rise of the cars-in-bike-lanes issue is a result of the movement's own success: cities are aggressively building-out their bike lane infrastructure, and the increase in cycling means more car-bike conflicts are captured on social media.

The core problem being highlighted is that cars parking in bike lanes temporarily negate the entire purpose of bike lanes, namely to make cycling safer and more convenient by creating exclusive road space for bikes and also formalizing car-bike-pedestrian interactions. A second-order benefit of advocacy is to further shift people's attitudes about cycling public policy, and to get them to see the urban world a bit differently. Constant reminders to advocates about the risks of cycling, enforcement bias, the importance of more action/support, and arousing anti-outgroup solidarity all probably help further the cause.

That's all great. Police traffic enforcement that skews anti-bike despite the obvious asymmetry in lethality compared to cars is particularly egregious. But it's not really the sociology or interest group tactics that bother me. On these grounds cars-in-bike-lanes is brilliant. Symbolism is a powerful thing. I get it.

My concern is more about the intellectual mentality that flows from advocates' arguments and style, and what it means for the broader cyclist and urbanist movements. Maybe that's unfair or silly, but this is my blog, so let's take a look.



First and foremost: do cars parking in bike lanes really make cyclists worse off? These incidents certainly make cycling marginally more inconvenient, and to the extent that annoyance and perceived risk affects hard outcomes like cycling adoption, behavior, and public opinion, they could indeed matter quite a bit on the whole.

And to the extent that cars-in-bike-lanes indicate deeper problems with enforcement bias and motorist attitudes about cycling, tackling the issue could be a smart inroad towards more systemic positive changes.

But what about the immediate, direct effect of a car parked in a bike lane on cyclist safety? On this point I'm unsure. (This is admittedly a small slice of the total issue, but much of the issue's rhetorical firepower hinges on it). What percentage of crashes/injuries/fatalities involving cyclists also involve a car parked in a bike lane? Very few. In fact, depending on what counterfactual you compare against, I'd speculate that a double-parked car occupying an unprotected bike lane can actually make a section of street temporarily safer for cyclists.


The geometry of bike lanes

If the choice is between a section of street with a double-parked car in an unprotected bike lane (let's call it scenario A) vs. a section of street with no double-parked car at all (scenario B), clearly the existence of the car decreases safety. This is the implicit comparison that most advocates make: get rid of the offending car altogether.

But if you acknowledge that some amount of temporary illegal parking is going to exist no matter what, we must consider another outcome: double-parked cars will occupy the rightmost car lane but leave the unprotected bike lane free, potentially creating a weird tunnel with parked cars on both sides of the bike lane (scenario C). Comparing the direct safety values of scenarios B and C isn't clear at all.

(In scenarios B and C, the negative safety effects are overstated. The presence of a double-parked car causes other passing cars to slow down a bit, and temporarily jolts motorists into a more focused mental state because of the narrowed lane and increased blind spots.)

In my experience, scenario C seems more dangerous than B. It increases ambiguity and blind spots: some cyclists will try and shoot through the tunnel and risk double exposure to dooring and step-outs (though temporarily eliminating the risk of being hit from behind), while some will swing wider than they otherwise would to pass the double-parked car on the left (in the case of a narrow street this can expose the cyclist to an oncoming traffic conflict). Even if there are no parked cars to the right of the bike lane to create the tunnel, having a cyclist pass the parked car on the right breaks the norm and creates a higher dooring risk.

In scenario B, however, the cyclist has only one option: pass on the left. This situation allows plenty of space and visibility to avoid hazards and oncoming traffic, and it eliminates ambiguity over a cyclist's path from the perspective of drivers.

So how do we know the relative likelihood of each scenario in response to cars-in-bike-lanes advocacy efforts? A good start would be to look at the preferences and choices of motorists.


The economics of double-parking

Direct safety-based advocacy, at its core, wants to prompt enforcement and norm changes that make parking in bike lanes more costly to motorists. For enforcement, this means either shifting police department resources (explicitly or through internal cultural change) to increase the probability of catching offenders, or increasing the severity of the parking infractions that are caught. In terms of norms, shaming bike-lane-parking motorists and changing their habits have the same effect of shifting people's utility function and decreasing their willingness to park in bike lanes.

To properly analyze the direct safety effects of increasing the cost of bike lane parking, we must recognize that motorists can respond in a number of ways (scenario B moving to either A or C).

On one margin, a higher bike lane parking cost will make some motorists simply double-park less. This is a move from scenario B to scenario A (an objective safety win). One imagines these folks would spend more time cruising for legal parking, or perhaps some amount of low-value activities previously undertaken while being double-parked would disappear altogether. To really stretch the idea, some motorists might be bothered enough to shift their overall driving habits, perhaps visiting certain areas less or opting for different transportation options. If a city were composed entirely of drivers like this, direct-safety advocacy would be very effective and justified.

The other margin concerns those motorists whose desire or need to double-park on a specific section of street is more intense (scenario B to scenario C). These might be people like: commercial and delivery drivers, taxis, police cruisers, people moving things, etc. Not coincidentally, these are the folks who are most targeted by cars-in-bike-lanes advocates. If motorists lived entirely on this margin, direct-safety advocacy would be ambiguously beneficial, and possibly counterproductive.

Figuring out how each of these margins actually plays out in the real world is ultimately an empirical question that in principle could be studied. But it's probably not necessary: there's a better way that largely eliminates the cars-in-bike-lanes dilemma altogether.


 Protected bike lanes are the answer

The impossibility of having unprotected bike lanes and simultaneously having no double-parked cars in them is glaringly obvious. The geometry of streets means double-parked cars will gravitate to the edge, which is where unprotected bike lanes are. Obsessing over motorists who have no practical alternatives, or cops who try to enforce the contradiction is just silly.

Both groups could and should do better, but the revelation in urbanism that physical design shapes driver behavior and road safety outcomes far more than education and enforcement is highly applicable to cars parking in bike lanes.

The good news is that protected bike lanes exist, and they largely solve the dilemma in a way that makes everyone better-off. Cyclists are physically protected from fast-moving cars, and the opportunities for cars to park in them are drastically reduced. There is a higher risk of pedestrian-cyclist conflict, but this comes in direct proportion to a reduction in car-cyclist conflict--a net safety win. For motorists with a reason to double-park, the available street edge is still the most logical place, but now the sequencing is changed to disrupt a full travel lane instead of half a lane. Not a huge loss, because in both cases a passing car's ability to drive on through without changing lanes is blocked by a double-parked car--whether a bike lane is protected or not.


Happily, many internet bike lane advocates are shifting their focus to protected bike lanes, and emphasizing the almost hilariously high return of thrifty bike barriers like traffic cones and flower pots. I hope this continues among cycling advocates, but even if it doesn't, more unprotected lanes are probably inevitable.

In the most desirable outcome for car-in-bike-lane advocates, enforcement and norms are shifted so much so that commercial drivers (and streetside landowners that rely on them) get exasperated and begin to lobby for more protected bike lanes in order to eliminate the problem altogether (double parking would then occur in the interior car lanes of the street). The negative flipside of this will be to dramatically increase the incentive to oppose new unprotected bike lane construction on a street. Depending on your view about the relative value of putting in protected vs unprotected vs nothing, this might be a good thing or a bad thing (there's a pretty compelling 'heighten the contradictions' argument against new unprotected bike lanes, but that's for another blog post).


Advocacy and understanding require different attitudes

In a nutshell, the core weakness of advocates is the implicit denial that at any given time there exists some optimal amount of cars parked in bike lanes. Their arguments and style are often geared around condemning motorist offenders and the police departments that enable them, which limits the ability to accept nuance. As I've discussed, this may be a smart political strategy, but the worldview it engenders rankles me somewhat.

The reality is, cities would be intolerable without double-parking. At any given time in a busy city, some percentage of driving lanes are occupied by double-parked cars. Of the streets with unprotected bike lanes, some percentage are being violated by parked cars. Vibrant, crowded urban streets are necessarily chaotic, noisy and messy. The constant little one-off violations of traffic rules are crucial lubrication for any functioning streetscape. Rules governing traffic are useful patterns of behavior, but they are not sensitive to the specific circumstances of time and place. Information on the ground matters, and being overly-attached to bad rules and bad designs is a recipe for intellectual rigidity.

The irony is that cyclists intuitively understand this: more than driving a car and walking, cycling is ambiguous with regard to existing traffic rules; an example like illegally rolling through an empty intersection on a stop sign is case-in-point. Advocates surely understand the downsides of excessive attachment to status-quo traffic rules: the entire history of bicycle advocacy--ongoing--is one of brutal and bitter knockdown fights with car-obsessed groups over objectively modest infrastructure and policy changes.

The attitude which leads cars-in-bike-lanes advocates to jump at every small violation is skirting the edge of this familiar intellectual trap. Viewing unprotected bike lanes as some transcendental category of road-space is not only wrong--it assumes a level of civic control, police enforcement, and driver behavior that is ultimately just not realistic. If a specific bike lane has an ongoing problem with parked cars, public action is awesome. But instead of calling-out every random violation, let's keep in mind that healthy streets are complex and chaotic, motorists and police are making decisions also, and to really fix the issue we'll need to scrap a broken design paradigm and shift to protected bike lanes and slower streets.