Is Bus Versus Rail Investment a Zero-Sum Game?

Abstract

Problem, research strategy, and findings: Bus and rail transit are often assumed to compete for the same pot of transit dollars, with one, therefore, substituting for another. Advocates for each mode lobby accordingly. However, this article presents evidence that spending on the two transit services may be unrelated or even complementary, based on a multivariate analysis of transit spending in the 55 largest U.S. metropolitan statistical areas from 1992 through 2006.

In all the estimated models, using a robust set of controls and statistical strategies, bus spending is positively associated with rail spending; that is, spending on the two modes rise or fall together. This supports the argument that, over the long term, rail investment does not come at the expense of bus investment; the existence of a tradeoff between the modes in investment spending is not supported.

Takeaway for practice: Many conventional transportation policy debates presume bus and rail spending is a zero-sum game, with investment in rail transit, therefore, coming at the cost of lower bus investment. By contrast, this analysis suggests an alternative scenario: Spending on one transit mode is, in general, positively associated with spending on other modes. This may occur because of needed interaction between the modes, as with feeder bus service, through the fostering of transit-supportive land uses or through broadening of transit coalitions into more affluent sectors of society. Assertions that rail investment inherently requires cutbacks in buses stems from a rhetorical misuse of the opportunity-cost concept.

Research support: Environmental Protection Agency project RD-83334901-0.

Keywords:

• National Transit Database
• transit
• finance
• investment
• bus
• train

Acknowledgments

Thanks to David Bieri for advice on modeling and Patrick Dieter for invaluable research assistance.

Notes

1. Los Angeles, New York, San Francisco, and Washington, DC, entered as consolidated metropolitan statistical areas.

2. New Orleans was excluded from the change models because of the extraordinary impact that Hurricane Katrina in 2005 had on reported transit spending during the study period: from $191 per capita in 2004 to $26 in 2005.