Tipping Points

I want to follow up on a tangent from my last point…tipping points in complex systems. This brief discussion is prompted by the following statistic:

In 2008 versus 2007, average vehicle miles travelled on urban highways in the US decreased by about 3%. During the same time period, congestion decreased by an average of 30%. See here and here. How did such a small change in the number of cars cause such a large decrease in congestion?

The answer lies in the way that complex systems handle operating at, near, or beyond their “capacity”. When a road is half empty, adding another car does little to nothing to affect other vehicles’ speed and travel time. But at rush hour, in stop and go traffic, even a single car entering your lane can disrupt or even halt movement for a large number of cars for a significant time. Congestion, speeds, and travel times are subject to a tipping point or “non-linearities” in response to increasing demand. Great. So why is this important?

It is well known that traffic congestion has numerous ill effects. The time wasted has a high economic cost, cars in congestion pollute more and suffer more wear and tear, the added stress and exposure have been shown to have measurable health affects for both the driver and surrounding community, and time spent commuting continuously shows up as a major factor in people’s self reported happiness and life satisfaction. So we know congestion is bad. Isn’t that why we’re doing all these road projects? To reduce congestion?

Where poor road design has resulted in a true bottleneck, it makes sense to invest in improvements to fix the problem. But when the entire system is overwhelmed, does it make sense to expand the complete road system? Even ignoring the other arguments for shifting auto traffic to alternative forms of transport, a simple cost/benefit calculation questions the sense of investing in much of the road “improvements” that occur. Criticisms of most mass transit, bike, and pedestrian projects often center around the idea that money is being spent to serve a ‘tiny minority’ of people who use them. It is true, at least for bicycle projects, that a small percentage of trips occur by bike in even the most bicycling friendly cities in the US. What is ignored, even often by bicycle advocates, is the huge benefit to drivers that even a tiny percentage shift of transport mode can have.

With limited resources, it makes sense to spend your infrastructure dollars where they’ll have the biggest impact. Portland is looking at spending $100 million to build 123 miles of bike lanes. Los Angeles just began a project to spend $1 billion to build 10 miles of car lane. Direct comparisons may not be completely fair. But for the price of 1 mile of freeway construction, Portland will get 123 miles of bike lane. And if those 123 miles shift only 1% of road traffic to bikes (Portland currently has about 7% of commutes going by bike), the entire road system could see a reduction in congestion by a full order of magnitude better.

2 comments so far

  1. Herbie Huff on

    But don’t forget that any improvement in capacity on a congested roadway will be subject to ”induced demand.” So if bike lanes get some people out of their cars and that frees up scarce road space, people will switch from other routes, modes and times to fill that road space back up.

    This is the conventional planning wisdom with regards to capacity increases. Doesn’t mean they are all a bad thing. I most certainly believe that shifting 1% of current drivers onto bike lanes would be a very GOOD thing. But as long as the road is free and there is high demand for travel in that corridor, new drivers will come to fill up the road space.

    See the Orange Line in LA in 2005 and the BART line under the Bay Bridge for real life examples of this ”induced demand” phenomenon.

  2. thomasthethinkengine on

    Great article! thoughtful and well written. I suspect however, that the tipping point is an argument for congestion pricing as much as for bike lanes.

    If the change in the number of road users needed to seriously diminish congestion is not large, then perhaps it follows that the congestion charge need not be large either. This could make congestion charging more politically palatable.

    To Dr Huff I would say that the statistics quoted above suggest that moving cars off the roadway did not move cars on to the roadway… Or at least not in quite the same volumes.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: