DOES TRANSIT REALLY WORK?
Thoughts on the Weyrich / Lind "Conservative Reappraisal"
by Peter Gordon
University of Southern California
Table of Contents
Conservative commentators Paul Weyrich and William Lind had recently achieved some notoriety with their Conservatives and Mass Transit: Is It Time for a New Look? [1], offering a purportedly novel case for public transit. Because most of transit’s support had been from the political left, this was almost a "man bites dog" story, receiving considerable attention from editorial writers and other pundits. W&L have now expanded their argument in a new paper, Does Transit Work? A Conservative Reappraisal [2] that is prominent on the American Public Transit Association’s website and likely to garner similar attention. Is there anything new to what they say? What is a "conservative reappraisal"? Does it change anything? Or, is it just the latest of many efforts to make a bad idea look good?
Waving aside a mountain of findings that document the poor performance of public transit in the United States, W&L note that transit’s critics have been asking the wrong questions. Rather, W&L recommend, people should be asking: What share of "transit competitive trips" does transit really serve? "Transit competitive" has three parts: "First, transit must be available. Second, the available transit must be high quality. And third, the trip purpose must be one for which transit can compete." W&L add that high-quality transit usually means rail transit; availability usually means easy pedestrian access, presumably at both ends of the trip; worktrips are the logical trip purpose – although the authors also see a new recreation market, people riding trolley lines for fun, such as at the Atlanta Olympics.
Is transit less and less widely available because it is little used? Or, is it little used because it is less and less widely available? Inevitably, both are true to some extent. Yet, which matters most? W&L emphasize the latter in spite of the fact that most of us are glad to live in a world of consumer sovereignty where resource allocation and use is shaped by demand rather than the reverse.
Demand still matters even where provision is highly politicized, as in public transit, although the link between quantities supplied and those demanded is weak. In the United States, transit use per capita is now at a historic low despite more than $350 billion of public subsidies since the mid-1960s. The recent spate of rail transit building has been a porkbarrel exercise that has added significant capacity but not in forms or in places where people seem want it. In spite of all the new capacity, "availability," as W&L define it, has been receding -- and will continue to recede.
W&L’s approach, which emphasizes availability over demand, could be used to obscure cause and effect in any declining industry. People are no longer buying any of the items in the hundred-year old Sears Roebuck catalogue because they just can’t seem to get them any more. But, in fact, most of these items have been replaced because the public’s demand for them was trumped by more popular substitutes. The wonderful thing about significant new technologies is how they change the world. Airplanes, automobiles, cellular telephones, computers, microwaves, the internet and many others are so auspicious that they profoundly affect and shape lifestyle choices. Thereby these and many other new technologies do prompt ever greater demand for them. Personal transportation via the auto-highway system and suburban living are mutually reinforcing.
Concurrently, an opposite downward spiral affects conventional transit. It worked best in a world with high concentrations of origins and destinations (where employment and population densities are high) and with large numbers of people too poor to own and operate an automobile. Both conditions have been declining in the United States and in many other developed and developing nations, explaining the decline of conventional transit both here and abroad. The three preconditions that W&L insist on are declining precisely because transit competes so poorly. Population and employment densities are falling in most places, and most people and jobs are choosing not to locate near transit stations because collective transportation is inconvenient and expensive in terms of what really counts, people’s precious time. In low-density settings where origins and destinations are dispersed, transit that best serves high-volume corridors competes poorly. High-capacity rail systems are, thus, inevitably underutilized, ever more expensive and ever more difficult to provide more of.
How does rail transit compete in the circumscribed realm that W&L insist on? They rest their case on three examples, the light-rail systems in St. Louis and San Diego and one commuter rail system, Chicago’s Metra. The authors argue that commuter rail and light rail "represent the future"; they are the only plausible alternatives. Citing the Los Angeles Red Line subway, they exclude such heavy rail projects because they are too costly. In each case, the authors seek to apply their standard: How many "transit competitive trips" are served? They admit that their answers are imprecise because ridership data are not necessarily compiled with their definition in mind. Nevertheless, a look at the three examples shows that they are not a basis for dismissing the voluminous evidence that the era of rail transit is long gone.
Chicago’s Metra. Commuter rail systems bring suburban commuters downtown. In modern America, downtowns are the slowest-growing job market. As a result, suburb-to-downtown commuting represents a small and declining market. In 1990, metro areas with a population of 1 million or more had less than 10 percent of their jobs in the traditional downtown; Chicago’s 336,313 downtown workers represented 8.7 percent of the metropolitan area’s total employment. Chicago ranked second after New York in the proportion of downtown workers that used public transit, 60.7 percent. But both of these CBDs are exceptions. They are two of just four U.S. downtowns with more than a quarter-million jobs.[3] Yet, between the economic censuses of 1987 and 1992, downtown Chicago’s retail and service jobs actually fell, a slightly worse performance than for the other large downtowns.[4] Fewer origins and destinations near transit stations will diminish W&L-accessibility.
In summarizing Metra performance, W&L report that in 1996, 92.7 percent of Metra’s riders were on business or work-related trips (not many recreation trips). They also report that Metra’s 1996 ridership was 73.4 million. This means that 68 million of its boardings were worktrips. At two trips per day and 260 workdays per year, this represents almost 130,850 Metrans commuters per day. Yet, W&L also report that Metra carried 21% of Chicago’s CBD commuting in 1990. The census data suggest that this would have been 70,626 commuters. This is a huge disparity. Six years of growth cannot resolve it because downtowns do not grow that fast, if at all. At least one of W&L’s descriptors must be wrong.
The authors say little about costs, simply that, "Metra recovered 58 percent of its operating expenses from the farebox in 1997 – the highest in the country for commuter rail service – plus 5 percent of passenger revenues for capital financing." The true capital costs on commuter rails that share substantially grade separated rights-of-way are hard to identify. Whatever they are for Metra is not revealed but, 95 percent is apparently paid by nonriders while 42 percent of operating costs are also subsidized. Comparisons with long-haul bus operations are not made. Indeed, these cannot be made as long as true costs per Metra boarding are not specified. W&L simply conclude that "it works!". Claims for more systems on this scale require more detail and a much more solid case than the one provided. True costs per rider would be a good start.
Fortunately, the 1998 FactBook of the area’s Regional Transit Authority reveals that Metra’s capital asset value in 1997 dollars was $7.2 billion.[5] The standard way to annualize this is to assume a life of 35 years and a capital cost of 7 percent per year. If so, the approximate annual capital cost is $710 million. The FactBook also reports 1997 ridership of 72.3 million. Capital costs per boarding, then, were $9.82. Also, 1997 revenues per passenger were reported to be $2.93 while operating costs per passenger were $5.44, leaving an operating subsidy of $2.52 per passenger. Adding capital costs, the total subsidy was $12.34 per boarding, almost $25 per round-trip.
San Diego’s Trolley. Of the ten U.S. light rail systems put into operation in the years 1980-95, only four are attached to area transit systems that did not decline. San Diego’s is one of these four. San Diego’s MTDB served 34.5 million transit boardings in 1980 and 50.4 million in 1995, 46 percent growth while the areas’s population grew by 41 percent. 15.6 million of the 1995 boardings were on the trolley, and 12 million of these boardings were on the system’s major line (the Blue Line) which runs from downtown San Diego to the Mexican border.[6] W&L admit that this market is not exactly what their three criteria are about. Many riders on this leg are tourists and Mexican residents going to and from the Mexican border. Most of the former certainly do have an automobile available to them, prompting W&L to conclude that, "The trolley has taken single-occupant vehicles off the road."
While the Blue Line accounts for most of the Trolley system’s riders, it accounts for only a small share of the costs. The 15.9-mile Blue Line capital costs per mile were $7.3 million (current dollars) while the rest of the system’s 30.7-miles accounted for $635.5 million, $20.7 million per mile. Thus, 85 percent of capital costs generate only 23 percent of ridership; put another way, 15 percent of the capital costs generate 77 percent of the ridership—on the peculiar border-to-downtown line. Clearly, beyond that unique case, there is very little demand for light rail.
The same story emerges from the study of operating costs. The Blue Line’s 91.7 percent farebox recovery (1996) is far above the Orange line’s 38.4 percent. The latter’s performance also lags the rest of the San Diego bus system.
Basing their case on 1998 Super Bowl week, W&L also see substantial recreational use of the San Diego Trolley. This appears to be another thin reed. Tourists going to Tijuana do use the Trolley everyday. Super Bowls come to San Diego far less often.
St. Louis MetroLink Taken together, the ten U.S. transit systems that added light rail served fewer riders in 1995 than in 1980. St. Louis’s system is one of the six individual systems with new light rail that lost ridership over that 15-year period, despite the addition of $348 million of light-rail capacity. Indeed, system ridership loss in these circumstances has occurred more often than not. For the case of St. Louis, the drop has been the most precipitous, from 84.2 million boardings in 1980 to 51.2 million in 1995. Light-rail served 12.5 million of the latter figure.
Bus patronage in many cities has suffered when service was cut to pay for rail and/or when routes were aligned to serve as rail feeders. Both appear to have occurred in St. Louis. Harvard’s Jonathan Richmond reports that, "The attraction of passengers to light rail can be partly explained by the removal of bus alternatives and the restructuring of fares. Prior to light rail inception, express bus service was priced at $1.25 and it cost $1.15 to cross the Mississippi. Local bus service cost $1 in Missouri and $0.85 in Illinois, with a $0.10 transfer. The current fare is $1 and the $0.10 transfer fee can newly be voided by buying books of 12 tickets which include transfers) for $10. The change in fare structure has made longer journeys proportionately cheaper ... a substantial discount is being given to rail riders compared to bus-only riders."
W&L celebrate the many "upmarket" customers (including those using it on weekends to attend professional football games) using light rail. They ignore the much larger, and probably much less well-to-do group induced to abandon the bus system.
W&L’s three cases are probably not as auspicious as the authors would like to believe. But, even if they were, could they possibly be a guide to public policy? Chicago and St. Louis have been losing population for many years; only San Diego is among the nation’s growth leaders. Even with a new light-rail line, St Louis’ transit use is down 40 percent. San Diego’s numbers hinge on one line that serves a unique border access market. Moreover, all three cities have not been able to avoid the continuing national decline of transit’s mode share. In 1980, the Chicago’s transit mode shares for commuting was 16.43 percent; it dropped to 13.38 percent by 1990. San Diego’s transit mode share was 3.23 percent in 1980 but 3.20 percent in 1990. And in St. Louis, transit mode share dropped from 5.58 percent in 1980 to just 2.83 percent in 1990. The U.S. Department of Transportation’s 1995 NPTS data for Chicago showed that its transit commuting mode share had dropped further to 10.3 percent; unfortunately, NPTS samples for the other two cities were too small to get a 1995 reading. [7] Even if the three projects that W&L single out are stand-outs, they appear to have made little or no difference.
On the basis of the three case studies, W&L conclude that "transit works". Even if we restrict attention to just their three projects, why not ask the economic question? All things considered, were these investments cost-effective? How else could the money have been spent? What benefits could have been bought if the same sums had been spent on transitways or exclusive busways? Would rail’s costs per passenger round-trip stand up to that kind of comparison? [8]
There are other problems with W&L’s attempt to make a bad idea look good. They fall back on the canard that the abandonment of transit for autos is a peculiarly American phenomenon to be pinned to U.S. policies (highway construction, favorable tax treatment of residential mortgage interest, zoning codes, low gasoline taxes, etc.). This glosses over the fact that the same trends are now apparent in most of the developed and many of the developing countries. Interestingly, this occurs in spite of the fact that many of these places have adopted tough policies designed to generate compact urban forms and "get people out of their cars." The demand for personal transportation is powerful and universal. [9]
W&L also claim that transit serves "economic development". Yet, areas stuck with huge construction bills and subsidies to pay into the indefinite future are clearly the losers. The authors also believe that moving the poor off welfare and into jobs depends on transit. If joblessness were an access problem, then New York City would long have had the lowest inner city unemployment. Far from it. New York usually has the highest of the big-city unemployment rates. Besides, W&L celebrate the "upmarket" use of their three transit systems but have nothing to say about large numbers of displaced bus users.
The authors would fix the many underperforming LRTs with, "... five words: more parking at transit stations!" Again, at what cost? How much extra ridership predicated on transfers will there be? It is well known, that once in their cars, most people drive right past the park-and-ride lots. W&L would cut rail transit’s costs by looking for volunteer labor (even an ROTC-style "police corps" for riders’ security concerns). More voluntarism is always a good idea. Yet, on any appreciable scale, it might spark the attention of transit workers’ unions and their political allies. W&L would (somehow) compel sufficient retail and services to develop near transit stations so that transit could also become the mode of choice for trip-chaining commuter-shoppers. This is the dream of many city planners who advocate Transit Oriented Development. To date, and unless offered huge carrots, investors have shown very little interest in locating near transit stations. When betting their own money, they are realistic about expected demand in such places and go elsewhere. Planners may want to rearrange cities and lifestyles in order to fill the trains but this confuses ends and means. Most astonishingly, W&L expect local government to mend its ways, abandoning the planning responsible for transit’s awful numbers and adopting the authors’ simple, if unpromising, precepts.
The unpleasant truths about transit in America, which W&L claim to dispose of, remain. Wishful thinking or an overly optimistic treatment of selected cases can not change some hard facts. The trouble is that the wishful thinkers, W&L included, are often enlisted by the special interests. Their paper is prominent on the APTA website. Thanks to the recent TEA-21 law, there are now more light-rail proposals on the drawing boards than ever.[10] If built, these will surely worsen an already woeful picture of transit waste.
Because public transit is clearly important, policies must change drastically. Resources and interest should be redirected at proposals that are cost-effective and make sense. If there are to be subsidies, they ought to go to transit users, perhaps the elderly and the poor. Deregulation to allow new van services to emerge (and to bring the "gypsy" cabs out of the shadows) ought to be pursued at all levels of government. The most congested roads and highways ought to be decongested via proper pricing. "HOT" lanes that accommodate express buses, new (deregulated) van services, carpools and those willing to pay for time savings ought to replace all the rail transit proposals. Compared to the W&L approach, these approaches would save taxpayers lots of money while serving many more travelers.
- Available at www.apta.com/info/online/.
- Ibid.
- Central Business District employment and commuting data from the 1990 census are available at www.publicpurpose.com.
- Peter Gordon and Harry W. Richardson, "The Destiny of Downtowns: Doom or Dazzle?" Lusk Review, vol. 1 (1997).
- FactBook of the Regional Transportation Authority (1998).
- Extensive Light Rail performance data are available in Jonathan Richmond, New Rail Transit Investments – A Review. Harvard University: John F. Kennedy School of Government (1998).
- www-cta.ornl.gov/cgi/npts/
- Many of the comparisons were made in 1965. They are updated in John F. Kain "The Urban Transportation Problem: A Reexamination and an Update" in Jose Gomez-Ibanez, William B. Tye and Clifford Winston (eds.) Essays in Transportation Economics and Policy. Washington DC: Brookings Institution Press (1999).
- See, for example, Christian Gerondeau Transport in Europe. Norwood, MA: Aretch House, Inc. (1997) and Genevieve Giuliano "Land Use Policy and Transportation: Why We Won’t Get There From Here" Transportation Research Board Circular (forthcoming).
- Cliff Henke "More Cities Prepare to Join Rail Club" Metro (July, 1999).
Peter Gordon is a Professor of Urban Planning and Development and Economics at the University of Southern California. He has conducted research on urban transportation policy and related matters for more than 25 years. Gordon's findings have been published in the major urban and regional planning journals. He has consulted for various private groups and government agencies as well as the World Bank and the UN. Gordon received the Ph.D. from the University of Pennsylvania in 1971.
3415 S. Sepulveda Blvd., Suite 400
Los Angeles, CA 90034
(310) 391-2245