Valuing public transportation systems
Econometricians and Hollywood producers have one thing in common: We make a living pitching cool counterfactuals. My whip smart (and recently tenured) colleague Michael Anderson has a new paper which answers the question of how congested LA highways would be in the absence of LA’s subway (and bus) system. The answer to this question is important, as transit systems receive large amounts of public funding and had previously been thought to provide little congestion relief since they account for only 1% of VMT nationally. The paper shows that the opposite is true. The intuition is simple:
“Transit is most attractive to commuters who face the worst congestion, so a disproportionate number of transit riders are commuters who would otherwise have to drive on the most congested roads at the most congested times. Since drivers on heavily congested roads have a much higher marginal impact on congestion than drivers on the average road, transit has a large impact on reducing traffic congestion.”
Anderson tests this hypothesis by using a 2003 strike by Los Angeles County Metropolitan Transportation Authority (MTA) workers, which shut down bus and rail lines for 35 days. He examines hourly data on traffic speeds for all major freeways in LA and finds that there is a sharp 47% increase in delays during peak periods. The results are consistent with his intuition: he finds the strongest effects on routes near transit lines with heavy ridership during peak periods and no effects in areas with no rail service.
“Our estimates imply that the total congestion relief benefit of operating the Los Angeles transit system is between $1.2 billion to $4.1 billion per year, or $1.20 to $4.10 per peak-hour transit passenger mile.”
While it is hard to imagine that a big time Hollywood producer will pick up on this story, this paper is another example where the mode choice of the marginal commuter has significant implications for everyone else on the road. While we teach the concept of congestion externalities to all economics undergraduates, determining their magnitude is extremely difficult. From a central planner perspective, this paper changes how we should view public transit systems in congested urban areas. I find this topic fascinating and think more of us should work on this important issue.
Categories
Maximilian Auffhammer View All
Maximilian Auffhammer is the George Pardee Professor of International Sustainable Development at the University of California Berkeley. His fields of expertise are environmental and energy economics, with a specific focus on the impacts and regulation of climate change and air pollution.
Energy Institute Blog 
Assuming, for the sake of argument, that the facts and results of the study are true. It begs the question, that Subway is the only mode that would achieve the same results. I would ask if buses or increased carpooling or HOV lanes or tolled roads or any combination of these modes would achieve the same effect at much lower cost.
The LA rail system is very heavily subsidized by all taxpayers in LA county with two half cent sales taxes and large amounts of federal money. The benefit in congestion relief is only in very specifics corridors that carry a very small amount of LA’s total traffic. Therefore most have to pay for something that they see no benefit to. If the benefit was really this great LA communters would be happy to pay the full cost of the rail system as this would benefit them greatly. This usually requires 3-10 times the fare they now pay, so the system is not economically viable. Some claim that businesses benefit from the ability to create dense concentrations of jobs, but if this is the case why are these businesses not willing to pay the full transit subsidy for their employees?
This article needs much better cost/benefit analysis with details on who pays and who benefits.
While I agree a fuller cost-benefit analysis is desirable, it is unfair to criticize a study that looks at one aspect that requires substantial technical analysis. Not all studies need to be complete soup to nuts analyses. We need individual building blocks to create a more complete analysis of the benefits and costs. This study produces one such important building block.
Why can’t Caltrans fix I 80 from the Bay Area to Sacto. It goes from 4 or 5 to 3 lanes in places and creates huge daily bottlenecks and long delays. I guess they spend all that money on themselves – their generous salaries, benefits, pensions, etc., or waste it on expensive boondoggles like the new span of the Bay Bridge – 10 years behind schedule and 5 times over budget! Please lets go to privatize more of their inefficient operations, like going to design ands build by private contractors. It has worked so often in emergency situations!
This question has bugged me ever since I read a few years ago that the Altamont Express [from the Stockton area to San Jose] was ‘built/ upgraded’ at a cost of about $250M, and has a daily capacity of about 125 passengers in each direction during commute hour arrival/ departure. Can that possibly have any significant impact on road congestion even if every one of those 125 drove individually. For that much investment a decent hub might have been created in the East Bay to permanently move over those workers.
We as a species, certainly as a people-nation, seem to focus on symptoms more than causes. We treat the symptoms and the problem shifts.
A question regarding the findings of this paper: they are based on 2003 data. Those were dot.com [and other factors] driven boom times. If the study could be for ‘normal’ times, and also factor in changes in commuting brought about by technology change, we could make projections.
[I recall reading, about 20 years ago, that if the BART system were used to capacity ‘it would take about 700 years to ‘save’ the energy used in building the system. I have not been able to verify or completely contradict it.]
2003 was after the dot.com crash. I found it much easier to drive into SF for meetings after 2000 than before because of the dot.com crash, and I remember seeing stories about how the traffic load had decreased starting in 2001. So 2003 may have been a low-point, not high point.
As for the embedded energy to build BART, having looked at the embedded energy in other cement and steel infrastructure projects, there’s really not very much compared to the overall energy use of the system. On the other hand, the energy used to move heavy rail vehicles is quite high per passenger compared to light rail or buses.