The Case for Rapid Transit;
Before and After the Fact
Professor Transportation System Planning
Allow me to part company with many avid transit buffs
by saying that my homework and best judgements lead me to conclude
(1) In modern-day urban America, no city not now
having a rail rapid transit system will find one attractive,
appealing, convenient, efficient, economic and useful - - all
(2) Only a handful of cities have economic, demographic
and social characteristics which warrant even study much less
construction of a high-capacity bus or rail rapid transit system;
(3) Myopia, ignorance, and lack of objectivity
(and perhaps integrity) - - on the part of the transit industry,
its consultants, legislators and members of the 4th estate -
- have led to the widespread waste of resources in the transit
field. Moreover, these shortcomings have led collectively, to
gross over-use of rail technologies and to both under-use and
improper use of bus and para-transit technologies; and
(4) In many cities the best proposal for solving
its urban transportation problem may well be no proposal. That
is, many cities may be better off by simply living with and wallowing
in congestion and pollution than by expending enormous resources
to make only a small dent in them.
It seems to me that much of the above is obvious
- - if only one wants to see it. For instance, if rail rapid transit
was even a way - - much less the way - - to cure the
traffic congestion and pollution problem, then why is congestion
and pollution so bad in New York, Chicago, Boston and Philadelphia?
And why, if rail rapid transit is so attractive to people, are
the five older U.S. rail rapid transit systems losing patrons -
- even as both capital and operating subsidies are increasing and
as new lines and stations are being added?
In reply, many would and do say that the problem with existing and pre-BART
type rail transit was its image and its poor service. If only cleaner and more
modern cars and stations were provided, and if only subways with speeds matching
that of autos were built, they - - they say - - transit can and would become
competitive, would attract high ridership, would reduce auto usage and, in
turn, traffic congestion and pollution, would promote better urban development
patterns, and would help the transit dependent. Also, it is argued that such
transit to be effective must be grade-separated (e.g., rapid transit) and have
high capacity, in fact enough to carry 30,000 to 40,000 people per hour per
On the surface, these arguments sound good - - particularly if one is trying
to sell a system or an improvement to one. Or, as BART advertised in the early
1960's, the main purpose of Rapid Transit is to eliminate the oppressive
traffic congestion. To accomplish this, the BART System must be so appealing
that commuters will choose, most willingly, to ride the trains to and from
work each day, instead of struggling along crowded traffic arterials in their
cars. The trains must be competitive with the private automobile in terms of
comfort, speed, cost and convenience.
BARTD sold its idea and got it funded. It was to be the modern-day version
of yesterdays rapid transit and the answer to traffic congestion. It
was designed with high vehicle speeds, with massive capacity, with comfortable
cars and seats and was accorded rave reviews.
However, despite high hopes and fancy predictions by BART consultants and analysts,
almost none of the advertised and promised successes came to pass.
Quickly, what did happen?
(1) BARTs high capacity system - - one which
can carry 30,000 to 40,000 people per hour in each direction
- - is actually carrying only 120,000 per day on all its lines,
and its Transbay tube is actually handling only 26,000 per day
in each direction. The low capacity Bay Bridge, by
contrast, is still carrying 110,000 people each day in each direction
(some 90,000 by car) despite the existence of BART. While some
outlying BART stations have less than ample parking space, and
a few of its trains have more standees than is desirable, it
would be fair to say that low patronage for the most part has
not resulted from low transit capacity.
(2) Traffic congestion even on the Oakland Bay
Bridge and its approaches did not go away. Or as a recent BART
impact study report put it: Traffic levels at the busiest
hours showed only small reductions. This is evidenced by an increase
in average speeds though the Bay Bridge Toll Plaza from about
15 mph at the busiest time of the morning peak (7:30 a.m.) ...
to an average of about 18 mph ....
(3) In terms of being competitive with private
auto vis-a-vis time and cost, full data is not available. But
what there is shows that BART may be cheaper farewise, but certainly
is not quicker. For instance, Transbay BART users who formerly
used auto now lose about 13 minutes per trip but save $1.50.
Also, others crossing the Bay by auto could save $1.70 by using
BART but would lose about 23 minutes per trip if they chose to
(4) The so-called transit dependent seem to be
ill-served by BART. With over 40% of BARTs riders coming
from households with family incomes of $20,000 or more, and over
60% from those in the $15,000 and over category, less than an
impressive argument can be mounted on that score. Moreover, with
over 60% of all westbound Transbay BART users and 87% to 96%
of those coming from the most heavily used home-end stations
getting to BART by auto, one must wonder about how much BART
in fact is helping the transit dependent.
In many respects, BART is no exception. For example,
the new Dan Ryan and Kennedy Lines in high density Chicago have
passenger loads of less thank 20,000 people per hour while the
adjoining low-capacity highways carry many more people on a daily
basis. For the Ryan, 160,000 per day by highway vs 110,000 per
day by rail transit; for the Kennedy, 120,000 per day by highway
vs only 60,000 per day by rail transit.
More blatant examples would include the Lindenwold Line from New Jersey to
Philadelphia, the So. Shore Line in Boston and the Airport extension in Cleveland.
The automated and high speed Lindenwold Line attracts only 20,000 passengers
per day each way, the So. Shore line only 10,000 people daily each way, and
the Cleveland extension only 5,000. Not only do these lines handle far, far
less than high capacity passenger loads, but they also have not brought about
the talked about and promised traffic relief. Again, as the BART impact study
team put it, when examining the So.Shore and Lindenwold cases, No perceptable
(sic) changes in traffic congestion were recorded on [the parallel highway]
One must wonder out loud: Why are we providing these facilities and what can
we do to improve them?
One question at a time:
First, why are we providing such facilities if not to provide high-capacity,
to reduce congestion and to help the transit dependent? The stock answer, of
course, is to conserve resources and to provide the cheapest form of mass transport,
because we all know how cheap rapid transit is, especially relative
to the auto.
However, the story on this score is not dissimilar. In fact, the record is
sordid, to be blunt about it, and it seems high time to blow the whistle on
the high-priced PR experts and fancy analysts who estimate and bandy about
the expected costs and revenues for these modern-day urban saviors.
Let us look squarely at the facts.
During the selling or pre-construction phase, Morgantowns PRT facility
was only going to cost $13 _ million; after the fact, it did cost $64 million.
The Lindenwold Line was to be built for $54 million but did cost $92 million.
The So. Shore Line increased from an estimated $74 million to $111 million.
The most recent and blatant fiascos, though, are D.C.s METRO and San
After D.C.s METRO was sold to congress and to the local and state governments
involved, the estimated costs were revised, in 1969. The unsuspecting public
was told that the capital costs - - including 30% for contingency and inflation
- - would be only 2.5 billion and that, once running, the annual maintenance
and operating costs would be $32 million. Then, only six years later, the costs
were again revised. The $2.5 billion capital outlay figure zoomed upward to
over $4.5 billion and the $32 million figure for annual maintenance and operating
costs skyrocketed to $129 million. Very tidy! Only an 80% mistake on the capital
outlay side in six years, and a 300% slip with respect to maintenance
and operating costs.
What about BART? As you know, this 19th century system, which has been embellished
with a few 20th century garnishments, was peddled to the public and legislature
over 15 years ago; its bonds were first floated in late 1963 and its first
line was opened in 1972. BART was to cost about $1.000 billion, a figure which
included over $200 million or about 30% for contingency and inflation. Moreover,
a year later, Adrien Falk, president of BARTD, was quoted as saying, We
are obligated to complete this system within the funds made available by the
voters. Bill Stokes, the BARTD general manager, only poured oil on the
fire by saying, We directed the engineers to be ultra-conservative in
projections on the income and costs of the project. We added an inflation factor
of $200 million though we know it would be easier to sell a $600 million bond
issue than an $800 million issue.
What did happen? Instead of costing $1.000 billion, BART actually cost $1.600
billion, and a $600 million or 60% cost over-run.
But the story does not end there, sadly enough. As you may know, BART was estimated
- - again by high priced consultants and BARTD personnel - - to have rather
substantial patronage and large operating surpluses. However, after 3 years
operating experience we find that:
(1) Annual patronage is only 36 million, a figure
far below the projected 75 million number, a mere 50% goof.
(2) Annual maintenance and operating M & O
costs were estimated to be only $13 million, but in fact are
about $57 million, an error of only 340% in spite
of lower patronage and car-mileage than anticipated.
Thus, the projected operating surpluses of $10 million
a year vanished and were replaced by annual operating deficits
of almost $40 million.
These estimation bungles, collectively, mean that it is costing the public
at large some $5.40 for each BART trip and not $1.30 as estimated. Put differently,
with an actual average fare running about 60¢ (not 30¢ as planned),
each BART rider is subsidized to the tune of $4.80.
Put in these terms, rail rapid transit no longer looks so cheap - - relative
to private auto, taxicab or what-have-you. Nor is there any reason to believe
that the overall situation for BART will look better rather than worse in years
What, you may ask, has all of this got to do with L.A. and its spate of rapid
transit proposals. A lot.
All of these recently completed facilities and the L.A. starter line alternatives
share four things in common, the first two of which are fatal deficiencies:
(1) The markets being served are too small (in
absolute numbers) and the densities at the two trip ends are
too low to result in reasonable patronage levels. After all,
the be all and end all for a high-capacity rail rapid transit
facility is the patronage. If you fail there, then you are dead;
(2) The stations are so widely spaced that coverage
of both the residential and employment trip ends is poor, thus
resulting in poor accessibility. As a result, access time, walking
and transfers will be too high for most people and most especially
for those not having a car for access. The problem in L.A. will
be particularly acute if you undertake one of the proposed starter
lines. Why? To date, no successful rapid transit facility or
system has more than 1-mile spacing between stations, compared
to 2 miles for BART and 1.4 miles or more for L.A.;
(3) Rail technologies have been provided or proposed
in the examples mentioned and in L.A. As a consequence, a separate
access mode and excessive walking, waiting and transferring are
required, thus making the facilities really unappealing to most
travelers. Nor can you get to and from the transit stations without
an auto, for the most part. Moreover, rail trains and tracks
can not easily or cheaply be moved as living and working and
shopping patterns change. Thus you find yourself tied to a fixed
route pattern that stands in stark contrast to the changing travel
and living patterns of today and tomorrow; and
(4) In all these cases, the costs were seriously
understated and the patronage projections were grossly inflated.
While this error may or may not be fatal, I wonder what the voting
public would say if you and the press, the analysts, the consultants,
and the academicians alike were to tell it like it is.
Rest assured that the figures included in the present
SCAG starter line reports are considerably off-the-mark. Not only
are the capital maintenance and operating costs considerably understated,
but the ridership estimates are ridiculously inflated. On the latter
point a recent SCAG report provides cost and patronage information
for a number of alternative starter lines, one of which (alternative
1) runs eastward and then south from North Hollywood. According
to the study, alternative 1 (with only 29 stations and a station
spacing of 1.8 miles) will garner some 125 million riders a year
while alternative 5 (with 17 stations and a station spacing of
1.4 miles) will have annual ridership of 95 million. Surely those
responsible for preparing these L.A. starter line patronage estimates
are kidding! To appreciate how absurd these ridership projections
really are, merely consider the following comparative data: Chicagos
entire rapid transit system (with over 150 stations) currently
enjoys annual patronage of only 94 million; Bostons system
(with about 70 stations) has only 90 million riders; Philadelphias
(with over 60 stations) has less than 60 million; and San Franciscos
(with 34 stations) has about 36 million.
All of the above - - and much more - - leads me to suggest the following course
of action for SCAG, SCRTD and Los Angeles:
(1) Go back to the drawing board;
(2) Seriously consider bus rapid transit if you
insist on a high-capacity rapid transit system (which you probably
dont need and cant justify);
(3) Do more experimentation with express buses,
either on exclusive lanes or on grade- separated right-of-way;
(4) Experiment more with para-transit modes to
include taxicabs, shared cab, jitney, subscription bus, and private
van pooling. These alternatives, if coupled sensibly with proper
de-regulation and pricing, hold considerable promise.
While I tried but was not successful in helping Boston,
Washington and San Francisco avoid their recent disasters, let
me hope that L.A. will be a different story. It is a city blessed
with many advantages, all of which I hope you make good use of.
Top | Home | Contents | Next