When it comes to mobile emissions, not all bike rides are created equal.
The cyclist who drives her bike into downtown to take part in Critical Mass or rides along the Towpath on a Saturday afternoon does not actually eliminate vehicle miles traveled (VMT) or reduce greenhouse gas (GHG) emissions to any extent. (This is why the National Bike Challenge’s methodology tends to irk me).
None of this is to say that these rides are somehow inferior or less than those taken for transportation; they’re not. Recreational riding is good for public health, enjoyable, and it increases the number and visibility of cyclists on roads. But it is somewhat disingenuous to claim they improve air quality or mitigate climate change.
How do we calculate the emissions savings from bike projects?
Now, we already know that shifting people from cars to bikes can go a long way towards promoting these ends. The problem is that we lack good tools to let us demonstrate this on the small scale. How do we prove, definitively, that investing in a particular piece of bike infrastructure gets people to change their travel mode? And how can we calculate the associated emissions reductions?
In some ways, recreational cycling may make this process more difficult. Traditional methods, like bike counts, don’t distinguish between those who are riding for recreation and those who are riding for transportation. Knowing the difference between the two and being able to isolate that segment of the latter who would have otherwise driven is essential for cycling advocates. We need to be able to quantify the demonstrable benefits of bike infrastructure in order to get funding for projects under certain programs, particularly the Congestion Mitigation and Air Quality (CMAQ) Improvement program.
In part because we struggle to get accurate data, bike infrastructure projects remain a small sliver of overall CMAQ projects. The Federal Highway Administration (FHWA) estimates, for instance, that CMAQ needs to invest $3.5 million in bike projects to reduce one ton of fine particulate matter (PM2.5), compared to just $38,000 for diesel vehicle retrofits and $76,000 for idle reduction programs. Perhaps the cost-benefit ratio for bike projects would improve if we had better data on how bike infrastructure directly affects mode choice.
New research may provide an answer
Fortunately, researchers are beginning to develop better tools to do just this. In a new study (paywall) in the journal Transportation Research Part A, researchers Seyed Amir H. Zahabi, Annie Chang, Luis F. Miranda-Moreno, and Zachary Patterson explore how the built environment and accessibility to bike infrastructure affects mode choice and GHG emissions among commuters in Montreal.
In the study, the authors broke Montreal down into a series of 500-square meter neighborhoods based on population density, employment density, cycling network density, transit accessibility, and land use mix. It defined neighborhoods using one of five typologies: downtown, urban, urban-suburb, inner suburb, and outer suburb.
Using this approach, they sought to answer two main questions. First, what are the effects of the built environment and the network connectivity of the transportation system on cycling rates during the period in question (1998-2008)? Second, how did cycling rates and the associated GHG emissions change over this period?
In order to study the first question, they estimated the effects of the neighborhood typologies on cycling rates. However, this isn’t as straightforward as it may seem. One cannot directly estimate this effect on mode choice, as people often self-select into certain types of neighborhoods that fit their preferred mode. For example, I consciously looked for apartments in certain parts of Washington, DC so that I could be within a short walk of a Red Line station. The same holds for cyclists, who may choose to live in more bike-friendly areas.
When your independent variable (in this case, neighborhood type) is not completely independent from your dependent variable (commute mode choice), we say they are endogenous. The researchers employed a statistical approach, known as a simultaneous equation model, which allows them to control for this endogeneity.
Drawing the link between bike lanes and GHG reductions
To study their second research question, they utilized a variable that allowed them to measure the distance a person lives from the nearest bike path or lane. This enabled them to consider how increasing or decreasing that distance may affect commute mode choice and GHG emissions.
It’s this second question that I want to focus on, as it gets to the heart of the issue I raised earlier. Fortunately, the authors provide some concrete evidence that investing in bike infrastructure does foster mode shift. When it comes to bike lanes, if you build it, people really do come.
Based on their results, they found that reducing the distance that a person lives from the nearest bike facility increase the odds that s/he will bike to work by 3.7%. In Montreal, the city expanded its bike network to 648 kilometers (402.6 miles) in 2014, from 603 km (374.7 miles) in 2008. The expansion directly led to a 1.7% decrease in vehicle GHG emissions within the city.
This reduction stacks up well with alternative emissions control options. As the authors conclude,
As in other studies, it is found that cycling infrastructure accessibility is positively linked to bicycle usage, playing a positive role in reducing transportation GHG emissions, by shifting the mode share of bikes. Although this effect may appear small (about 1.7%), it is as big as the estimates we have found in our previous research when converting all the transit diesel buses to hybrid technology and electrifying the commuter trains in Montreal at the same time. This is to say that the GHG benefit from adding low-cost new cycling infrastructure can be as important as other more costly strategies.
Hopefully this type of research can provide further, tangible justification for incorporating bike infrastructure in the urban toolkit to tackle climate change. We need to build real (preferably protected) bike lanes in order both to increase the number of people biking regularly and broaden the type of people biking from hardcore recreational cyclists to normal people using bikes as a transportation mode. Because, while recreational biking is great, only transportation biking can help us solve these pressing crises