One of the many myths propagated around transit and housing is that if we can increase adoption of transit, getting people to switch from their cars, that we can reduce congestion. Another is that by investing more in transit we can “increase mobility”.
It almost sounds like common sense? But it’s important to look at the evidence objectively. The figures for the San Francisco – Oakland Urbanized Area tell a very different story than the common sense that some might presume to tell us.
The chart on the right shows two variables over the period 1982 – 2011:
- In blue is the transit passenger miles per capita for the entire Bay Area population, this includes people who do and do not take transit.
- In red is shown the travel time index. This represents the travel time in the peak period to travel time at free-flow conditions. E.g. a Travel Time Index of 1.35 indicates that a 20 minute free-flow trip takes 27 minutes at peak.
Source: Texas Transportation Institute, Urban Mobility Report 2009, Exhibit 1., “Major Findings for 2009″
The key question is, does an increase in transit usage, measured here by transit passenger miles per capita (blue line), cause a decrease in congestion, here measured by the Travel Time Index (red line)?
To ascertain this a statistician applies a coefficient of determination. A figure of 1.00 reflects a direct correlation and 0.00 no correlation. The coefficient for the data in the chart is just 0.28 – most statisticians would say that there is no statistically significant relationship. However, what should also be considered is that the transit-reduces-congestion theorists expect, as transit utilization increases, congestion decreases, and vice versa – and, what we have here is a weak relationship showing that, as transit utilization increases, congestion increases. Why Doesn’t Increased Transit Utilization Decrease Congestion?
The problem here is that transit represents only a tiny fraction of total Bay Area people mobility. To achieve any meaningful increase in transit usage would require a prohibitive increase in expenditures.
This is best illustrated by how much impact the SMART train could have on highway 101. During peak highway 101 carries 15,400 vehicles per hour at the N San Pedro Road exit in San Rafael in Marin County (Source: Caltrans 2012 Traffic Volumes on California State Highways, page 142). The national average vehicle occupancy is 1.67. So this translates to 25,718 travelers.
By comparison the SMART train, a heavy rail project costing $1.6 billion, will provide a train running every 30 minutes in peak direction is capable of carrying 156 passengers per carriage in 2 carriages (Note that train length is limited so that the train does not block multiple city blocks at railroad crossings in San Rafael). Assuming the train is 80% full (it is near impossible to achieve 100% of capacity), this translates to 499 passenger per hour for SMART. However we cannot assume that every passenger on SMART switched from driving – realistically 50% of those passengers already used transit, so only 250 passengers per hour are removed from 101 (149 cars are removed of 15,400).
This means that the cost of of removing one car (1.67 travelers) from highway 101 to SMART during the peak hour is $10.7m. Consider by comparison that for under half that amount, at just $4m, we could pay people currently commuting on 101 at peak $80,000 / year not to commute over 30 years.
It must be noted that the the CalTrans vehicle figure of 15,400 is rounded to the nearest hundred vehicles. So it is almost accurate to state that for $1.6 billion, the reduction in traffic congestion achieved by SMART is a rounding error.
Note however that the above comparison still gives SMART an unfair advantage as it assumes, invalidly, that all SMART passengers were previously driving cars – when in reality many, perhaps half, may have used another form of public transit such as a bus. So we really are in the rounding error territory.
So Why is Travel Time Increasing?
The first graph in this article clearly demonstrates that travel time is increasing in the San Francisco Oakland metropolitan area. The math demonstrates that transit represents a tiny portion of Bay Area people mobility.
More likely, what is occurring is not a cause-effect relationship that transit advocates would tell us is common sense – this is more a result of land use policy and of towns and cities becoming built out to their capacity. As Bay Area cities have become more built out it has pushed residents to live farther and farther from jobs. During this long term expansion policy makers have funded an increasing proportion of funding in transit rather than road expansion. The result: increased congestion.
The figures also tell a story of fewer people taking transit since the early 80s (see chart on the right), but those who do travel further – increasing trip time.
What Does this Mean?
This article is not intended to offer a direct solution. Many good solutions, based on market forces rather than expensive government and ineffective intervention funded by taxpayers, are covered in the solutions area of this website.
This article provides a grounding in the fundamentals – so that instead of basing thinking on what some might presume to be common sense, we can have a framework where we can really plan based on reality.
References & Credits:
Much of the data and analysis used by this article was collected and synthesized by Thomas Rubin, a mass transit consultant in Oakland, California. He served as Controller-Treasurer of the Southern California Rapid Transit District from 1989 until the SCRTD/LACTC merger that formed the Los Angeles County Metropolitan Transportation Authority in 1993. Prior to joining the SCRTD, he was a partner in and National Transit Services Director for Deloitte Haskins & Sells (now Deloitte & Touche). He earned his BSBA from the University of Nebraska-Lincoln and his MBA from Indiana University-Bloomington.