Friday, May 4, 2012

Bikes, Transit and Traditional Urbanism

Back in 2010, Nathan Lewis published what is one of the few pro-urban critiques of transportational bicycling available on the net. In the piece, he calls into question the assumption that increasing rates of urban bicycling, or increased provision for bicycle infrastructure, are necessarily beneficial for pedestrian-centric traditional urbanism.

One of the greatest dangers, Nathan writes, is that a city consciously designs itself around the bicycle, while neglecting the needs of people on foot. It is not a groundless concern: the blog Half the Fun critiques (with photos) the Dutch city of Houten, which was designed entirely around bicycles, for having "lost sight of the forest for the trees" by its focus on cycling rather than on "creat[ing] better, more livable communities."

The article concedes that "more bikes on the road would lessen just about every transportation problem you can
A growing storage problem: bicycle parking by a transit
station, Copenhagen (Leif Jørgensen).
think of," but I think that is true only where a new bike trip replaces a trip by car. Replacing a walking trip or a transit trip is less obviously beneficial from a citywide perspective, since the bicycle introduces potential conflicts with pedestrians and presents parking issues that differ only in scale from those of cars. A review of European transportation initiatives, however, shows a frequent focus on increasing bicycle use and modal share, rather than simply decreasing automobile share, with the unstated assumption that the added bicycle riders will be drawn largely or exclusively from the pool of car drivers.*

That may not always be the case. Although evidence isn’t abundant, one study in the Danish city of Odense found that, during the mid to late-1990s, although the share of all trips taken by bicycles rose from 22.5 to 24.6%, this was accompanied by decline in the share of public transportation from 8.2 to 6.6%. By contrast, in Portland, a recent increase in bicycling share appears to have been drawn at least as much from the driving population as from transit riders. This question of whether increased funding for bike infrastructure indirectly results in the decline or stagnation of a city's own transit ridership should be of interest to municipal transportation agencies.

Based on commuting modal share statistics, it is not always easy to tell the extent to which new cycling trips have displaced transit and walking trips, car trips or both. Consider the dramatically different commuting mode shares of four European cities, each of which has taken a somewhat different approach toward transportation, while all having near-identical rates of car commuting:
On first glance, it's not easy to explain the differences between cities like Copenhagen and Vienna. Both are national capitals of similarly sized countries with similar metro populations and mass transit networks that include extensive subway systems.  Vienna is denser than Copenhagen, which may in part explain the much higher walking share, but we see that despite Copenhagen's immense pro-cycling efforts (so extensive that they have given their name to a popular pro-bicycling movement), its car share is only slightly less.  Vienna, by contrast, has a much more modest goal of increasing bike share to eight percent, but has primarily focused on improving transit share.

In fact, the results over time for both cities show that transit share hasn't budged, even though Copenhagen's metro entered service in the middle of the time period below (Vienna's U-Bahn has undergone recent expansion but much of the system, which opened in 1976, predates 1993).  Is it possible that pro-cycling efforts siphoned off potential transit riders in Copenhagen?  And if so, was this beneficial for the city?

Vienna street: pedestrian-centric design,
not friendly to bikes (Flickr/PaulLamere)
Vienna met its own earlier cycling share goals, but even more significant was the increase in walking share, which far exceeded expectations even though walking is hardly mentioned in Vienna's 1993 transport agenda.  What Vienna had in abundance, after all, was dense, traditional urbanism the ultimate pedestrian infrastructure.  Through restrictions on cars, the city streets again became pleasant places to be, inducing walking far more than expected.  Copenhagen implemented similar policies, yet saw a decline in its already low walk share. 

Still, it is difficult to be too critical of increased bicycle use.  By the standard of almost any American city, Copenhagen's achievement in reducing modal share for cars is remarkable.  The bicycle, too, is an almost magical technology, the sole transportation method devised by man which has improved on the energy efficiency of walking, and one which expands the range of choices for city dwellers.

But is it enough to consider the impact of bikes on the natural environment? Shouldn't their cumulative impact on the urban environment be considered as well?  This will involve issues not only of parking ever-larger numbers of bikes as their popularity grows, but of compatibility with walking and mass transit, and of a city's vision for the interaction between the various modes of urban transport.

One final mode share chart, for New York:

The cycling share may seem surprisingly low, but consider what a bicycle is worth in the city: although it may improve mobility as compared to walking alone, it also essentially locks the bicyclist out of New York's entire public and private transit system (MTA and taxi service), all of which is implicitly designed around the person on foot (sure, there is the Metrobike and other folding bikes, but their appeal is limited and practical difficulties remain).  With the transit system running 24 hours a day, even the ready availability of the bicycle ceases to be an advantage.

All this suggests that as a city's mass transit system improves its frequency, coverage and hours, the value of a bicycle for urban mobility should decrease until, in the case of New York, it reaches close to zero for many neighborhoods.  There's room for difference of opinion, but I think this must be seen as a good thing.  Bikes can be an excellent transit gap-filler, in limited number, but may not be as well suited to being the central element of a transit strategy.

*See e.g. Copenhagen ("it is municipal policy that cycling mode share should go up to 40% by 2012 and 50% in 2015"); Groningen (city "promot[es] cycling as the main mode of transportation" with "vast expansion of the cycle network"); Charter of Brussels: (cities pledging to "set of target of at least 15% for the share of cycling in the modal split of trips for 2020").


  1. Glad to see you back with a post.

    One thing I did see in NY was those small kick scooters that can fold to the size of a large umbrella. On flat ground and gradual declines they let you move easy and on inclines they aren't too difficult.

    1. Thanks, Seth. I saw a guy riding one of these scooters along a Midtown NY sidewalk earlier today, attempting to weave through lunchtime foot traffic without much success. After a few minutes of that I think most people would be strongly tempted to just step off and walk. In other contexts I'm sure it would work better.

  2. Pro-cycling initiatives here in the near suburbs of Boston pulled this transit user off the buses and onto my bike... resulting in a reliable commute of 30 minutes each way instead of an unpredictable hour to hour and a half depending on traffic conditions and whether the bus headways lined up just right or just wrong at the transfer point. I have to say I can't bring myself to care whether that mode shift was a net benefit for anyone else ;) (but then, for me, a car wasn't an option at all, I couldn't afford one).

    Still, though, in general I have to agree with you. In a lot of ways bikes have all the same advantages and disadvantages as cars, just to a lesser extent (except pollution). While bicycling was probably THE most appropriate solution for my particular commute (from one streetcar suburb to another), there are other problems that it's really not ideal for. Around here, if enough people switch from bus or subway to bike to get to, say, Harvard Square, that... could actually cause some pretty severe bike parking problems, and at times make it a less pleasant walking environment. And given that parking there costs an arm and a leg, I'm pretty sure that the most people who drive in are coming from quite a ways away, or are irrationally attached to their cars- mode shifts towards biking there are going to come from the transit and walk shares. I hate to admit that I'm a contributor to that problem, but the bus I take there is generally overcrowded and doesn't come very often, especially on weekends. So I ride my bike instead, and take up a chunk of sidewalk to lock my bike up. Sigh.

    1. Thanks for the comment, anonymous. I don't think you or anyone else can be faulted for commuting by bike when it is clearly the cheaper, faster and more reliable option. I did the same when I lived in Nashville -- although there was a bus stop two blocks from my apartment, it had 40 minute headways at rush hour and was frequently off schedule. It then took 15 minutes to end up downtown, a four-minute walk from my workplace. Door-to-door by bike (a downhill ride) was about 10 minutes. In fact, for that particular bus route, a bike was *always* faster, even starting from the end of the line.

      The reason it worked so well is because Nashville has a bike mode share of essentially zero, so when I arrived at work, I was never in competition for a "parking space." If, say, ten percent of the population within a two-mile radius did the same thing, the crush of bikes in such a small area would be overwhelming.

    2. A parked bike uses much less space than a parked car. If 10% of the population biked to work, there would be a lot of empty car parking spots or lots which could be used for bikes.

    3. Hi Eric -- there might be more open spaces, but part of the appeal of the bike is the door-to-door convenience it offers. The first choice for most people appears to be tying the bike up to the nearest tree or street sign, but it takes only a few bikes before all those spots are gone (this often obstructs sidewalks as well). On-street parking, on the other hand, is so convenient that I don't think decreasing the number of cars will open up any spaces, but a city might well decide to reserve several of them for bikes anyways, if bike share is high enough. Given the bike to car space ratio you describe, that would probably be the best and most efficient use of space.

    4. The first preference is also to tie the bike up to a good, solid, bike stand. Bikes tied up to the nearest tree or street sign is a sign that either an inadequate number of bike stands have been provided or that the ones provided have been badly located. And an advantage of solid permanent bike stands is that they can be located to ensure that they do not obstruct pedestrian paths. A good urban ratio between high density attended / locker parking at main transit interchanges, major parking sites with air pump facilities and distributed permanent staple parking would be about 15:35:40.

  3. Cycling can actually cause a lot of problems - and I've been bicycling all of my life and won't trade it against some car commute. In my hometown Uppsala, Sweden (~150 000 inh.) most people ride their bikes occasionally and many people take their bike on their daily commute. And it's not just students who can't afford transit or car, also well paid lawyers, bankers etc.

    But the big bicycle culture here also has some drawbacks... It drownes many of the public spaces in the city-core in bicyclestands - a problem that is almost as big as surface parking for cars in most inner cities. By the trainstation right in the middle of the city, there's a bicycle parking area as big as a full sized fotball field! I think that's a very bad use of perhaps the most valuable piece of land in the city.

    Copenhagen is a very interesting example. For me it's fascinating to see the bike-culture that they have there. And I must say that they have done a very good job on organizing bicycle-parking. When I was there last summer I didn't notice any bike-parkings that messes up the public space. Even in the midieval city-core, with quite narrow streets I didn't see any problems...

  4. "By the trainstation right in the middle of the city, there's a bicycle parking area as big as a full sized fotball field! I think that's a very bad use of perhaps the most valuable piece of land in the city."

    Surely a multi-story bike parking garage could be constructed, possible with an actual building (housing/offices) above it... This would be much cheaper than a comparable car garage.

  5. Replies
    1. Thanks for those links, alai! I had no idea about the underground bike lockers.

  6. Wow, your blog is brilliant. I was just pondering this question yesterday after the cyclist in front of me cut off a pedestrian who had the right of way at a stop sign, just like a driver would do (I was also cycling despite preferring transit + walking).

    It's interesting to consider your assertion that Vienna is denser than Copenhagen - I'm assuming it's due to the much greater number of American-style suburbs surrounding the latter city? (it's hard for me to call them American-style when they are on smaller lots than the typical house in the core of Minneapolis and their streetscapes feature traffic-calming that would be called socialistic in suburbs here) Just from my experience it seems like their old cities and 18th-19th century suburbs are pretty comparable.

    1. Thanks, Alex -- a lot of these points were inspired by reading Nathan Lewis' site. As for Vienna and Copenhagen, I think the difference in density is mainly due to the time at which the cities experienced growth: in 1900, the Copenhagen metro had only 700,000 inhabitants, while Vienna had 1.7 million (Vienna declined after 1918), which means that Copenhagen experienced the majority of its population growth during the automobile era. In addition, northern European urbanism generally seems to have been less dense than contemporaneous central and southern European urbanism, which would accentuate the difference.

      The early 20th century Copenhagen suburbs, in turn, are denser than their American counterparts, as you mention, due to shallower lot depths, narrower streets and what looks like an almost complete absence of rear alleys. It is still closer to the American model, though, than, say, the single-family suburbs of Tokyo.

  7. One advantage of increasing the cycling mode share is that it provides extra mobility independently of motorized transportation. There are advantages if people bike circumferential trips for which direct transit is not an option, instead of taking transit and changing at the core. Likewise, for medium-distance trips, having a bicycle available and usable for the entire trip could radically extend urban mobility into the suburbs, i.e. do the opposite of what urban renewal programs and suburban park-and-rides do.

    1. Yes, as a gap filler for routes that transit doesn't serve, a bicycle can be ideal. The suburban carfree area of Freiburg (Vauban) shows that this can work well in practice. What I'm still a bit conflicted about is the provision of bicycle infrastructure that mirrors mass transit routes -- the bike is so efficient that (in smaller cities especially) it's often possible to have it beat out transit on all but the longest routes. This directly undermines transit mode share, possibly without reducing auto mode share. The answer to whether this is, on balance, a good thing I suppose depends on one's view of the purpose transit is intended to serve in a particular city, how bikes are integrated, etc.

    2. The question is whether the target is increased transit mode share or the target is reduced automobile mode share. Clearly if the latter is the target, a well integrated system that provides opportunity for cycle / transit transfers at every strategic point and provides an effective separated transport cycleway system as well as effective parking infrastructure is going to recruit cyclists for higher frequency, rapid transit routes to key target destinations.

      But the critical question for the leverage that support for cycling infrastructure can play if pursued as part of an integrated transport system is not what role it can play where urbanist development is easiest, in the largest, most densely settled cities, but what role it can play where urbanist development is harder.

      As far as, "will we get the best outcome if we pursue cycle mode share willy-nilly, without investing any effort in fitting it into an integrated transport system", the answer is, of course not.

      But consider what would happen if you started out designing the system for a smaller American city that presently has an excessively high auto mode share, taking a local cycle catchment area as part of the design principle for a system of Express routes (of whatever specific transit mode). That would require a system of local circulator buses, running routes such as one way loops to the express route network or alternate route cycles between express route network nodes, to provide access to the system to those who are NOT on bikes.

      To me, that system with as effective an express route network as possible and the four alternative means of accessing it ~ locating to withing walking distance of an express stop, cycling to an express stop, catching a local bus to an express stop, and driving to a parking lot in walking distance to an express stop ~ is more likely to result in drivers making the transition to making some use of the network to abandoning the car and going car free. An effective cycle alternative for local transport reduces the anxiety involved in giving up the perceived freedom of "being able to jump in the car and go".

      As far as the folding bikes remark, I don't follow it, since part of my cycle commuting experience in a smaller city with fairly mediocre public transport was with an aluminum Dahon folder which I used to commute to work in Newcastle, NSW. I could put the bike in the overhead luggage rack in the Hunter trains, between my legs in the electric V-Sets that I took down to the Central Coast or to Sydney, it fit in the luggage space in the local city buses, and it fit in the trunk of a taxi if I was going home after the local bus system had largely shut down and for whatever reason I didn't want to ride. It freed me from reliance on the inadequate Newcastle Bus Service ~ but since it also freed me from having to have a car at all, I did in fact ride the bus more than I would have done if I had a car, and I rode the bus more than I would have done if I was relying on an unfoldable bike.

    3. Yes, and furthermore when people (car)park and ride there is a strong temptation to just keep driving. That temptation is much weaker with bike and ride.

    4. Yes, Park and Ride needs parking limits on the other side of the trip to really drive it, but it still involves having to own a car to commute, and having a car at home for non-commuter transport. Pedal and Ride with either a bike locker at the train station or other express transport stop, or else a folding bike brought with the commuter, is a steady transit customer and an opportunity to go car free as the commuter gains confidence as a cycle commuter. If you don't need your car to get to work, then that car insurance payment and that unexpected transmission job are not costs of getting to work but a cost of the driving you do for your shopping and leisure.

      Park and Ride is useful training wheels to get motorists into transit, but to be part of reducing car dependency and its attendant costs in blood, health and treasure, there has to be a strategy to transition a steady stream of those park and riders into car-free individuals.

    5. Thanks for the comments, Bruce. My remark about folding bikes was based on a belief that these bikes are likely to remain a niche product because their added utility, as compared to standard bikes, only comes into play in limited situations. In Nashville, for instance, city buses are already outfitted with frontal bike racks designed for standard bikes. On the other hand, in a city with bike-unfriendly subways or other commuter rail, transit is likely to be good enough that a bike is not strictly necessary (or at least not necessary on the destination end, eliminating the need to carry the bike on board). These bikes also do not eliminate the storage issue, although they do mitigate it.

      Your scenario with the small American city is an interesting one. For people to find it beneficial to bike to an express bus stop (either parking their bike or hauling it in), the bus must, I think, be both fast and frequent. Using the example of Nashville again, it would not be very useful to "pedal-and-ride" today, since the bike will almost always be faster to one's destination from the stop. I can think of circumstances in which a cyclist, on a given day, might want to take the bus, but as a matter of everyday use it would be uncommon.

      To make pedal-and-ride a viable proposition, the city would have to eliminate many stops, which would lessen the walkability of the bus system – potentially leading to a net loss in ridership. On the other hand, it could layer an express service over the local service, but this would be a large expenditure for a relatively small return. That money might be more effectively spent improving bike infrastructure to encourage people who live too far from a stop to walk to simply ride all the way to their destination – or adding new lines to increase the number of people within walking distance of the transit system. The most efficient means of reducing auto mode share is not readily apparent.

    6. The most effective means of reducing auto mode share is the means that you can obtain funding to cover its third party benefits so that third party beneficiaries are not taking a free ride on the farebox of the public transport riders.

      But as far as efficiency, in Newcastle, NSW, an urbanized area of about 300,000 in a five "city council" region of about 600,000, where I was first cycle commuting, the Newcastle Bus Service aimed to have a bus serve within a quarter mile of every resident over a daytime service period and within three quarter miles over the full service day.

      What "service" meant was a long service interval and meandering routes that took many multiples of the time of driving, with steadily declining ridership in the decade I lived in Oz, even during the period when train ridership was rising.

      A cross-urban and local structure places longer cross city bus routes along the normal main driving arteries through the city, with connecting local circulators serving local centers that are also the main connecting points with the main cross-urban routes. With short enough local routes, many routes can be provided as one way loops, since going the "wrong way" around the loop is not much more time than going the right way around, and the one-way running is reinvested in higher local route frequency.

      The quasi-BRT Express Bus system is then a high speed, possibly relatively low frequency overlay on the intercity bus route system, with the regular cross-city routes providing the fallback for missing the Express Bus.

      From the perspective of a cycle commuter, the local routes fade into the background, and the relevant routes are the cross city routes, with a strong preference for the Express buses along those routes.

      The threshold frequency for an Express bus route to be appealing to a pedal and ride commuter depends in part whether there is free wifi at the express bus stop, but the service must be fast, must be reliable which implies investment in bus priority at point bottlenecks, must connect important clustered centers together, and if the frequency is not very high should have time to next service display, and if the frequency is not very high, the corridor should host a city bus with stops at regular city bus separation as a fall back for the Express service.

      The frontal bike racks are handy for the regular transport cyclist who is an occasional bus riders, but they are not useful if there are, eg, ten regular pedal and rider commuters on that particular service. Per month rental bike lockers and luggage racks that can be used among other things for folding bikes are more appealing.

  8. I'm going to post here, because newworldeconomic's guy doesn't have comments and never responded to my email, and he reads this blog: I just finished reading "Imagine" by Jonah Lehrer. He has a whole chapter (7) about how walkable urban environments (traditional cities with really narrow streets?) measurably produce more creative output than suburbia - another reason oldurbanism is more optimal. I recommend the book, it lead me to some interesting new research.

  9. Any numbers on how weather affects bike ridership in various places? I could imagine biking to work on dry days when it was cool enough that I wouldn't arrive at work covered in sweat, but not when it was freezing cold, precipitating or warmer than about 75 degrees. (Even if I were willing to endure the sweat film all day, it would not be acceptable in my office to arrive sweaty enough to create even a hint of odor for the rest of the day.) That said, I don't think of Coppenhagen as a great-weather city so, clearly, some people look past this. Any numbers?

    1. Part of the cycling promotion policies of the University of Newcastle in Australia is the requirement of shower facilities in all buildings with staff offices.

      Heat is definitely a more challenging problem than cold or rain. Appropriate layering and cold weather gear handles temperatures well below freezing ~ after all, people go out in those temperatures on purpose for hours at a time on a winter sports holiday. And a rain cape and for more severe rain storms a pair of rain pants handles rain at cool temperatures quite well ~ though, as above, cycling in rain gear on a warmer day generates a sweat.

    2. No precise numbers, but there are some studies that have addressed this, generally finding evidence to support the conclusion that transit policy and bike culture override the disincentives of weather and topography:

  10. A simple viewing of videos from Dutch cities and towns show that bikes provide a solid backbone to urban-designs. All users go with the flow at the speeds everyone is expected to. No need for excess signage, policing, maintenance or costs in general. You all just keep arguing about cycle-scaled places, the Dutch will just keep-on keepin-on. I can't believe this post isn't just bait. :)