University of Southern California USC
Peter Gordon

Peter Gordon's Blog

A blog exploring the intersection of economic thinking and urban planning/real estate development and related big-think themes.

Recommended Books

Recommended Films

Riding the Sunset Coast Line

John W. Dyckman
James Irvine Professor
Urban and Regional Planning
University of Southern California

I. Introduction: The June, 1976 Proposal
For the third time since 1968, Los Angeles County voters will confront a rapid transit ballot when they go to the polls in the June primary election. This proposal, like its predecessors, emanates from the Southern California Rapid Transit District. By the present version originated with County Supervisor Baxter Ward who is also an RTD director. Ward has dubbed his 232 mile all-rail proposal the "Sunset Coast Line".

How does this proposal differ from other rapid transit plans which have been offered to the citizens of Los Angeles County? Without entering into all the details of the several versions of the various plans that have been proposed, we can distinguish all important features that distinguish the Sunset Coast line from other plans. First, unlike the 1974 RTD proposal which supplemented its 140 miles of high-speed rail with extensive express-bus lines and local feeder bus services to the high-speed rail, this is an all rail system. The rationale behind this emphasis, as expressed by Mr. Ward in a letter to the Los Angeles Times on April 6, 1976, that "All over the country people are being told they must ride when they really want rail systems." In the Ward proposal, much of the "feeder" service would be provided by light rail and monorail not buses. The amount of heavy rail on freeway right-of way alone (170 miles) would exceed the total heavy rail mileage of the 1974 proposal.

The use of freeways to carry the heavy rail is in itself a distinctive feature of the Sunset Line plan. Where other proposals have proposed exclusive bus or carpool lanes, the Ward plan contemplates rail lines in the medians or along the freeway shoulders. Stations would appear at the intersections between freeways. Under this proposal, actual station and line location would be chosen by the municipalities, who would receive a fixed subsidy from the District scaled to the type of station to be built (freeway, aerial, or atgrade). If the local government would elect to underground its line, it would need to find other financing for the differential cost between the amount allowed in the Sunset Line formula and the actual costs of subway construction. Thus the cities of Beverly Hills and Los Angeles, should they wish to build a subway along the heavily-traveled Wilshire Corridor (as was contemplated in the 1974 proposal), could not expect much help from the RTD.(1)

The Sunset Coast Line is planned to serve 44 cities of the County when completed. Thus it is designed to act as a comprehensive suburban rail system. The reason for this coverage is fundamentally that of political equity. As Ward put it in his April 6th letter to the Times, "You need the money of the whole county to build any system at all; fairness demands that you give service to the county in return". The ballot argument states "Within 12 to 14 years, lines will be operating to West Covina, Encino, Pasadena, Long Beach, Santa Monica and LAX, plus a connecting western line from Redondo Beach to Van Nuys - and that is less than half the system." In the end, the line is planned to reach the borders of Ventura and Orange Counties, and to stretch well into the San Gabriel Valley. And unlike BART, it will go to the principal airport.

How long will it take to get from one place to another? Once you get on a train, the Sunset Line proposes to move you swiftly. It aims at a theoretical average speed of 60 miles per hour. This is roughly 33% faster than the two highest speed systems now in operation in the U.S., which average 40-45 miles per hour. (These are BART and Lindenwald line outside Philadelphia). It would achieve this high line-haul speed by reaching a maximum operating speed of 85 mph, with a 20 second stopping time per station, and an average station spacing of just over 2.5 miles. Whether these specifications can be in fact attained is questionable, but if they could be, the rider would spend 20 minutes in the train from Santa Monica to a downtown station, and 25 minutes from the South Bay to the same destination. That this is not total travel time is, of course, obvious.

What will the residents pay for this system? Originally, the Ward plan proposed a $7.5 billion bond program. This has been modified in the present proposal in two important ways: 1) the amount has been scaled down to an estimated $5.8 billion, and 2) in place of the bond issue the present plan offers a "pay-as-you-go" financing based on the sales tax. The present 6¢ sales tax would be upped by 1¢ - which the proponents estimate would amount to about 10¢ a day for each resident of the County. This sales tax is anticipated by the designers to cover not only construction costs, but a substantial portion of operating costs. For the goal, according to the proposal, is to "approach no-fare, subsidized rate structure".(2) Thus the user charges paid by the actual riders are intended to be well below the costs resulting from the use.

When will the Sunset Line be in business? This is one of the thornier questions, whose answer awaits favorable coincidence of the stars. Certainly an advantage of the system is to be found in its avoidance of the time-consuming scheduling problems of subway construction. One clear advantage of the plan is the provision (30836.1) of the Public Utilities Code which provides time limits for the negotiations between the RTD and local communities over stations and lines. (After 60 days the question at issue goes to arbitration, where 30 days are provided for agreement). The bugaboo of all transit development in the present era is inflation, and speed is essential if the forecast expenditures are not to greatly exceed expectations. As designed, the plan does not anticipate strikes, lawsuits, technical obstacles, and program "line loss". If all goes well, the essential framework of the system could be in place in about 12 years, with eventual completion in about 25 years. As we shall see, the timing and scheduling of the work is of great importance in the assessment of its costs and benefits, for the experience of other transit developments undertaken in recent years makes abundantly clear, the delays - - particularly when the sales tax commitment is fixed by the vote in June - - could impose a high cost on the citizens. This is all the more telling when opportunity costs of the funds are charged at their real rate - not less than 8% or 9% at present interest. (If the project is delayed by as much as five years at these rates, the real costs could be virtually doubled). To minimize this effect, a "starter line" has been proposed to run to San Pedro-Long Beach. The issue of the value of this "corridor" will be a point of contention in the plan. For the more important the corridor first opened, the more valuable the operating income realized by the line.

In the venerable tradition of engineering project analysis, one should properly ask the questions: why do it?, why here?, why now? For the citizens of the County who will pay for the system every time they go shopping, these questions take on special force. Let us now consider these questions.

II. Does Los Angeles have a Transportation Problem?
On the eve of voting decision on the issue of a roughly $6 billion commitment to a rapid transit system, Los Angeles area citizens are well justified in asking, "do we need it?". Certainly there is justification if the present system does not work, if there is widespread dissatisfaction with transport in the region, and if a clearly better system can be devised which would merit the cost. (A cost which may well prove to be substantially higher, in real cost terms, than the estimated $6 billion in construction costs).

An initial question is: how efficient is the present transportation system? One efficiency measure of an urban transportation system might be "mean travel time" of journey to work, or over a selected package of trip purposes. This is not a complete measure, for transportation studies show that time, while important, is not the only relevant measure. Comfort, convenience, and economy to the user, among other factors, enter the value system of the users. Nevertheless, travel time is important. Here the evidence suggests that average travel time in Los Angeles, when compared to other cities of the world or of the United States, is not high, and is even less than that in many cities where transit is a more important mode. Given present conditions in the region, with its dispersed pattern of jobs and other destinations, and given its extensive network of freeways and abundant access to automobiles, the Los Angeles area is relatively efficient on grounds of mean travel times.

What, then, are the grounds on which this arrangement fails to meet public needs or to satisfy public purposes? Certainly, for the individual users, the freeway system works. But there are some possible areas of failure. Two of these come to mind. First, there are questions of social efficiency arising from congestion, energy costs, environmental pollution, and the land use issue of the sunk investments in areas that are badly served. Second, there are equity problems of those left behind, not served, and who cannot participate in the private solution of the dominant pattern. These are worthy of attention.

Given the linear and essentially grid pattern of the freeway system, there are certain congestion problems at the central intersections of the system. (In this case, the celebrated downtown "stack"). Some observations on congestion are in order. First, from the public point of view, congestion is not an unmixed liability. For while the social costs of congestion exceed the private costs at the margin, the volume of use on a public facility such as a freeway is also a measure of the return on the investment. To clarify this point, one need only consider the public cost of under-use of the freeways, which would represent a burden on the public which had paid for this facility. Since the cost of the freeways is a sunk cost, long ago committed, there would be a public interest in maximizing use of this investment.

To make it worthwhile for the public to "write off" a portion of this investment there would need to be very substantial public benefits offsetting this cost. One such source of benefits would be the enhanced mobility and accessibility arising from the transit development. If, for example, total travel times would be reduced by the development, users would have a value equal to that time or to the value of the alternative opportunities created by the freed time. But we will argue that the gain in travel time is unlikely, due to the time lost in getting to the stations, to the destinations on arrival, and to the waiting times at the stations. At this point, let us concede that this saving is problematical.

A second public concern is energy use. Almost half the liquid fuel used in the U.S. annually is consumed in transport, and the rate of growth of energy use in transport has been higher - over the past decade - than in all other uses. If an urgent public interest in energy conservation can be asserted, the problem is worth attention. At present, however, there is much confusion on this score. The Administration's position that fuel consumption should be regulated by prices might lead to some overall savings, but these are unlikely to be great in their effects on automobile use short of very steep changes. The ability of consumers to absorb substantial price increases in gasoline in most recent times suggest that demand is relatively price-inelastic. Certainly the overall effect of price changes on total vehicle miles traveled has not been appreciable. Much of the effect has been felt in shift to smaller cars, but recent sales figures show a weakening of this trend as well. And some industry forecasts hold out the prospect of medium term energy oversupply as new sources come into the market and as energy producers speed up their rate of exploitation. And in any event, if the EPA standards for automobile mileage were in fact to be achieved, the relative advantage of the rail system over the automobile would be somewhat reduced, although not eliminated. This is all the more the case if the automobile remains necessary as a feeder to the rail system.

Similarly, unless the Sunset Coast Line would succeed to divert a substantial portion of automobile trips to rail, the impact on air quality would not be great. Studies for the 1974 RTD proposal indicated that a 100% increase in transit ridership by 1985 would produce a 7% decline in total vehicle miles traveled. If vehicle miles traveled are used as an indicator of vehicle contributions to air pollution, admittedly crude, this is a rough measure of the environmental effect. On ground of air pollution reduction alone, this effect would scarcely justify the expenditure of the proposed transit investment. Citizens of the Los Angeles basin will welcome any reduction in air pollution; it is an open question, however, how much they are willing to pay for a small reduction.

Finally, let us turn to the question of equity. Los Angeles is a city with many centers which make up its "galaxy" of desired destinations. It has the highest automobile ownership of any large city and has a dense and well-connected freeway system (a feature which the Sunset Coast Line seeks to exploit). What of those who do not have the ability to drive and to use this system? According to the LARTS survey in 1967 there were 3 automobiles for every 7 persons in the Los Angeles region; 41 percent of the households had access to one vehicle; 44 percent had access to two or more vehicles, and only 15 percent of the households lacked a car. This distribution of auto ownership has certainly not "thinned out" in the intervening nine years. But if we take an upper limit of the national average of 60 percent of the old and poor who do not have cars, a class which needs other modes can be identified. If the old and poor constitute as much as 30 percent of the region, only 18 percent of the population, using the 1967 data, can possibly be considered to be in dire need of a non-auto system. Without analyzing the desire lines of such a group, one can conclude that less than 18 percent of the people of the District would need transit for lack of access to other modes, and of this group, a certain percentage can be well-served by buses. On equity grounds, the case for transit investment is not compelling. More modest alternative investments in jitneys, special taxi services or other public subsidies would have a clearer redistributive or rectifying effect.

Land use effects of the system have implications both for equity and efficiency criteria. New transportation systems have inevitable developmental impacts. Because they do not so much "create" values as they "shift" them, there are a number of distributional changes resulting from such developments. The case of BART illustrates this point. In the period of operation of the BART system there has been no net population growth in the counties of the District, but there have been population shifts adjusting to the transit opportunities. The main changes have been a growth in employment density at the principal destination downtown San Francisco, and a growth in residential development in the outer ends of the line, particularly in Contra Costa County (which has the largest "comparative advantage" in use of the system).

To the extent that the Sunset Coast Line will be "centered" on downtown Los Angeles we may expect some densification of employment in the center. This could conceivably help present plans to speed the pace of downtown renewal. But we should not expect the pattern of San Francisco to be repeated. And if a comparable densification of downtown were to develop, it would not necessarily be efficient. Los Angeles does not have the concentration in finance and headquarters functions of San Francisco, but has a more developed - - as it is presently - - in industrial districts at some distance from the center. Since the Sunset Coast Line would essentially follow the freeway pattern, it might encourage denser development of employment at certain of these subcenters. The layout of the system, as presently designed, would in fact strengthen the large-scale grid system of the present freeway network. (An important link in the grid is the proposed Century Freeway, which would provide the route for the connector to LAX.) A characteristic of grid networks is their flexibility of routing and the absence of a clear locational advantage of any single point. The current Sunset Line proposal has a large number of potentially "favored" locations at the intersections of lines in the large-scale grid. A number of these points might be expected to benefit at least as much as the downtown, especially since the proposed financing of the transit system offers little in the way of assistance to the Wilshire corridor, which would almost certainly need to be underground.

III. Will fixed rail transit help Los Angeles County?
The Sunset Coast Line has come down firmly on the side fixed rail over other types of transit. As a first step to answering the question "what can the system do fro the country?" it may be useful to examine certain properties of such systems. The key elements that must be addressed by fixed-rail transit are these: 1) access time to the system - the collector problem; 2) waiting time for the trains - the "haedway" problem; and 3) transfer time and convenience between modes - the interface problem. For the most part, these are properties of rail transport and its technology, but to some extent they are also properties of urban form. (As Wilfrid Owen once remarked, the transportation technology itself is only half the solution.)

This is immediately apparent in the task of passenger collection. Mean access time to the rail line is functionally related to the concentration or dispersion of people and activities. And the distribution of activities over space also effects feasible transfer arrangements and transfer times, albeit more indirectly. In considering the tasks faced by BART in 1965 I wrote: "Within metropolitan areas industries have moved increasingly towards the outskirts in search of larger sites; this movement has tended to disperse places of work and so reduce the usefulness of the highly centered, radial transit systems. Circumferential systems moving through predominantly low density areas have been less attractive to the transit companies. Within the downtown areas dispersal of places of work and of central points of attraction (brought about by changes such as the shift of a department store to the fashionable fringe of the area) has greatly lengthened that portion of the trip between arrival at the terminal and arrival at the final destination. The lengthening of the walk or taxi ride from station to destination has made the whole transit ride less attractive. These developments can be summed up in the observation that the general dispersal of activities and functions within metropolitan areas has made the fixed-rail system less efficient in point-to-point delivery of passengers."

The waiting time-headway task is also a trade-off between the economics of rail transport and the attractiveness to the rider. In this case, the rider wishes the headway, which depends on the frequency of trains, and so, on the number of cars at the disposal of the transit line, to be as short as possible as to minimize waiting time and to reduce his dependency on the schedule. For the transit operator, however, an opposing set of incentives operate. He wishes to maximize the loading on any particular train, so that his stock is more fully utilized.

In a certain respect, the headway problem is interrelated with the collection task. Given its stock of vehicles, a transit line must continually balance overall time performance against the need for frequent stops, both to collect the payload and to reduce waiting time for passengers at the intermediate locations. An ancillary routing problem (express trains versus locals) thus emerges. As the project review by the Legislative Analyst's office observes, "From a service standpoint, such things as nonstop airport services and nontransfer service for many origin-destination pair is desirable. However, the expense, including higher numbers of required cars and more yard and storage space, may not economically justify such service."

When peak load considerations are added the problem is more onerous, for frequent service at the peaks is necessary just to move the traffic from the busiest stations. "Peak loads" are more pronounced for transit than for other modes. A 1972 study by the Department of Transportation found that peak hour trips on commuter railways accounted for 20-50 percent of total daily travel nationally; while more "local" rapid transit trips at peak were 15-20 percent of the total; buses 10-15 percent; and autos 8-10 percent. For many transit companies, 80 percent of the volume of travel is concentrated in 20 hours of the week. Such high peaks lead to high operating costs, since the capacity for meeting peak loads without breakdown is far in excess of the average capacity needed to meet the demand of the system. Since transit is most heavily used fr commuter journeys, the concentration of journeys in narrower bands of time accompanies the movement toward fewer workdays in the week and less work in shifts.

The far-flung extent of the Sunset Line raises another difficulty. For the greater the distance from the periphery to the desired destination (e.g. the center), the greater the advantage the express trains to riders, or the higher the cost of stops. These are not insoluble programming tasks, but operating costs are increased by factors such as additional tracks needed fr express bypass, more sophisticated switching controls, and more cars. Even if the Sunset Line does a good job of compromising conflicting operational requirements, the very process of trade-off means that the system either favors ceriian classes of users or arrives at a program which is acceptable to many users but is not optimal for any. BART, for example, has essentially opted for better service to the farther-outpoints at the expense of servicing some intermediate points, or of requiring high transfer time. In that case, higher income users are favored over lower income users at intermediate locations.

The transfer time is obviously a function of the compromises in routing and weight given to various origin-destination pairs. Just as obviously, it depends upon the total amount of rolling stock available to serve the routes. But as we have seen, the efficiency criteria fro the operation require carefully programming to maximize load factors, or at least to trade-off carefully with the travel time measures of service. The extensive grid system of the Sunset Line proposal, and the large number of desired destinations, a re likely to require extensive transfer time. This, in turn, will add to total travel times impressively and will weaken the attractiveness of the service.

The conclusion that one reaches from examining the Sunset proposal is that it has undervalued the costs to the user of waiting, transferring, and interacting with other "feeder" modes. The solution of the overall transportation needs of Los Angeles is not likely to be found simply in heavy rail. As a first step, one would want to match different local conditions with the most suitable mode in each case. For example, one would match fixed rail to line haul connection of high density origins and destinations; automobiles to very low density collection and distribution points; buses or other modes to intra-district (e.g. CBD) high density needs; jitneys, subsidized cabs, or dial-a-bus to flexible route collection fr certain groups, such as the elderly, children, or handicapped.

The Sunset Line resembles a suburban railway system, and suburban railways have long depended on a mixture of ample parking, taxis, and "kiss and ride" services to connect their stations to the low-density residence of riders. Similar connections between rapid transit lines and rider destinations--often achieved by interface with the more flexible buses- are important to overall transit performance. Put simply, there is a severe task of user accommodation in most system interface, ranging from considerations of shelter from unwanted intrusions risks, and weather. Finally, we should recognize that on-line performance may need to have provision of uninterrupted off-line switching to accommodate the user concerns so that the overall system of performance is not dragged down by the needs of the entering or leaving users. These concerns have not been adequately addressed in the proposal now before the voters.

When we view the benefits of the Sunset Line from the point of view of potential users, these features of fixed-rail transit loom large. For it is soon apparent that the advantage to the prospective rider, if it can be established at all fro those having an automobile option, is quickly eroded by the costs of waiting, transferring, and interfacing.

IV. Vales, Demand and Choice
A paradox of the rail solution is that such advantages as it has, are largely in evidence at rush hours. But the concentration of riders in the peak hours is, as we have seen, notoriously costly to transit management. If the Sunset Line is to be essentially a commuter railroad, the citizens of the county are being asked to subsidize the 30-35 percent of all trips (at the maximum) which are home-work related. And if the Sunset Line succeeds to divert as much as 30 percent of work commuting to rail (a generous share, considering that there are very few cities where all public transit carriers achieve this level, and that buses can be expected to continue to carry some of the commutation), then the multi-billion dollar expenditure will be charged against no more than 10-20 percent of all commuter trips!

It may be worthwhile to consider the incentives to the individual to use this facility. First, let us look at the time spent on travel. If the Sunset Line will deliver a rider in 20 minutes fro Santa Monica, that rider must allot time to get to the station, to wait for the train, and to get from the station to his destination in calculating his total trip time. As little as 10 minutes to drive to a station and park, 5 minutes waiting, and five minutes from the station to his destination doubles the total travel time. Once in his car, the commuter may very well choose to stay in it to his destination, especially since there are no clear time savings in changing modes.

There are, moreover, some incentives to use an automobile once it is owned; the license fee is paid, and insurance costs are not a linear function of miles traveled. Automobile ownership represents heavy fixed costs, but even at present prices of fuel, relatively lower operation costs, so that the economic rationale of the owner favors intensive use of his investment. It is doubtful that even relatively low fares will prove an inducement to many drivers, who see their operating costs as insignificant.

In some cities, notably New York and Washington, the inability to park the car at a reasonable cost to the driver, if parking is available near the destinations at all, may be an important factor in his choice of mode. But in Los Angeles, 24 percent of the CBD land is devoted to parking, almost twice the average in the central areas of most large cities, and almost three times that in cities such as Chicago. The present pattern in downtown Los Angeles favors lot parking over structures, with resulting lower prices. Facilities in the subcenters and industrial districts of the County are even more favorable to convenient parking.

The advantages of the Sunset Line, like those of BART, would be greatest fro the user who is farthest away from his desired destination, and who would have available an express train that would connect him reasonably closely with his destination. There would be, for example, some attractiveness to a resident of the South Bay communities, and those in the southern portion of Orange County, who work downtown, assuming that the total number of stops is not too costly of travel time, and that transfers can be expedited where necessary. Such locations would enjoy some of the advantages that Walnut Creek and Concord have on the BART line. This advantage is reinforced by the proposed fare policy of the Sunset Line, which promises not to gear fares to miles traveled.

For most users of the Sunset Coast line, some sort of feeder mode will be needed to bridge the distance between the stations of the main rail lines and the origins and destinations of the journeys. As noted, the use of these feeders is an important factor in total travel time. Yet feeders must, by definition, collect and distribute at a number of locations. As the January prospects for the Sunset Coastline says, "The feeder and distribution systems, of course, would operate at lower average speeds than the Main Line, since those routes are designed to facilitate local travel, and their station spacing average .44 miles between stops." To the extent that the Sunset Coast line relies on buses to provide this feeder service and to the extent that it will supplant buses over many of the most heavily-used corridors, it will place a heavy burden on the operational profitability of existing bus lines. For by pre-emptying the most expensive and poorly patronized role. In the case of BART, it has been necessary to have the rail system subsidize the feeders. This may also be necessary for the Sunset Line if it wishes to avoid loss of essential feeders by abandonment or service reductions.

When all these factors are considered, it is apparent that the Sunset Coast Line proposal is an expensive solution to a limited portion of the total transportation task. In what a sense, then can we say with Supervisor Ward that the people of Los Angeles County "want a rail system"? If the proposal were to charge fares fro riding the line which reflected its real cost, a fairer test then that before the voters might be made. For then each prospective user could compare his marginal benefit from transit use with the cost of the service. If the Sunset Coast Line were a business proposition, it would have to bring its customers a service which warranted the cost.

In fact, however, the proposal would tax all purchases made in the County, whether by transit riders or non-transit riders. Fares are to be deliberately held below real costs. Non-users will subsidize users. And heavy sales tax payers will bear the brunt of the subsidy. The reliance on these tax sources suggests that the Sunset Coast Line is considered to be some sort of public good. What is the case for considering this rail transit as a public good?

Those goods and services fro which market price systems cannot effectively be devised by reason of consumption externalities and the inability of market prices to exclude non demanders are generally called public goods. Transportation facilities normally partake of this characterization to a very limited degree. What are some of the consumption externalities which are considered to be likely in the Case of the Sunset Coast Line? Riding on the line is not itself an external benefit, for we have seen that, in principle at least, pricing could be used for inclusion or exclusion. Yet, some external benefits may be generated. If enough drivers of autos are attracted to transit use, for example, other driers may enjoy the benefits of reduced congestion and decreased travel times. In the case of BART, it has been calculated that opening of the system has reduced vehicle trips on the heaviest corridors by about 5 percent. This has certainly not been enough to eliminate congestion, but it might be argued that the congestion would be worse without the system. So there are clearly some gains of this type, although these gains have been limited to certain routes and have not been dramatic even on these routes.

An additional benefit might be realized in the furtherance of other public policies, such as downtown renewal. Traditionally, interurban rail systems have been good for the downtowns, whose function is dependent on concentration of large numbers. If there is justification for public policy favoring downtown renewal, there may be related benefits in investing in transit to further those objectives. But it cannot be argued that total benefits from land development will be increased by the rail line; its function will be to redistribute benefits from land use to the locations best served by the transit. There may be advantage to the functioning of the city from redistribution in space of activities. If this would be the case, there would be some positive benefit from the development. The amount of uncertainty surrounding such outcomes, however, would cause this benefit to be liberally miscounted in a present calculation.

If all the probable benefits are summed: impact on travel times, reduction of congestion, improvement of land use patterns it is improbable that they will be of an order of magnitude to justify the cost of the line at any positive interest are over any reasonable expected life. Of course, we need not be thorough-going utilitarians about all aspects of life. We may find we have an overwhelming nostalgia for the Red Cars, for the time when Big Reds ran over 1100 miles of the basin, and when the smog was minimal, the prices of cars and gas were cheap, and grass grew along the tracks. Some among us may have a long unrequited passion for the scene around the old Union Station, and may wish to restore it to a wino-free hub again. Or we may just hate cars, the freeways we drive, or the frictions of driving. Many of us probably would like to say that Los Angeles is just as up-to-date as any big city (particularly our rival to the north), and everyone knows that a modern city must have a "Metro". To the extent that these feelings are prevalent in the community, voters will rally around the Sunset Coast Line.

V. Where do we go from June?
If the ballot measure passes, the emphasis will move to the urgent need for a detailed plan of constructions. This is an exacting Task for many reasons. The experience of recent rail transit development, such as Washington's , shows that delays in construction can cost dearly in inflation is decidedly positive - delays could soon lead to a doubling of initial cost estimates. Furthermore, the frictional costs of construction are certainly high. Since the Sunset Line proposes to make extensive use of the freeways, we can expect very substantial disruption of freeway traffic during the period of construction. Any "stretch-out" of this time will soon lead to time losses of an order of magnitude which dwarf the putative gains from the system. (Consider the cost to surface traffic from the construction on the Washington Metro, or to downtown San Francisco from the Market Street Building.) Prolonged construction of the rail line along the San Diego Freeway, for example, could cost more in time lost to the motorist user of that route than the rail system could be reasonably expected to save along the same alignment. For if motorists were delayed an average of 5 to 10 minutes each day by construction activities over a period of 3 years, it would take about 30 to 50 years of operation (at the rate of time-saving achieved fro freeway drivers by the development of recent transit lines) to compensate this loss.

The sales tax increase will also impose a differential disadvantage on businesses in the Los Angeles County. While the impact of this tax is difficult to estimate on business in general, it would certainly disadvantage businesses in the District which border on Orange County, for example. Retailers selling large ticket items would be most disadvantaged, for comparative shoppers could realize appreciable savings by crossing the county line. Beyond a certain distance, travel costs would outweigh the differential, but the impact is not negligible.

Indeed, it is regrettable that the political strategy of "selling" the system has dictated a district-wide sales tax and planned lines reaching every part of the county. At the least, a compromise utilizing a half-cent sales tax, as was done by BART, would have the advantage of greater equity, since the property tax would presumably have the advantage of reflecting the differential impact of the lines on local property values. It is also important that the District could not begin with a more modest trial on selected corridors offering the greatest patronage and revenue possibilities by virtue of their high density of origins and desired destinations. If such limited lines could be coupled with special assessments on properties a fairer test of the financial feasibility and indirectly the equity of the burden would be realized.

Should the Sunset Coast Line proposal not be approved, will the Los Angeles transportation system break down? I think not. The system has demonstrated great flexibility. The availability of alternative routes, as much as the adaptability of the users, has permitted the Diamond Lane experiment to avoid chaos. The notion of reserved lanes for buses is not in itself a bad one. The limited success of the Diamond Lane is largely attributable to the lack of preparation and to the inexplicable neglect of those supporting activities which would make it work. Among these we should include all the needed actions to increase and improve bus service, organization of subscription buses, solution of the interface between the freeway routes, encouragement of jitney fleets, and other devices to encourage multiple-rider carriers. Without these efforts to reduce the transaction costs of multiple ridership the reservations of a lane is scarcely an adequate solution.

There is evidence, moreover, that freeway use in the Country is nearing its peak. Population in the country has shown signs of stabilizing, and even of decreasing. While the existing population is constantly redistributing itself, the emergence of denser concentrations at certain locations in the future may justify the more limited solutions on selected portions of the freeway system. Furthermore, the automobile ownership rates in Los Angeles have approached saturation. It is probable that the presently high rates will be increased in the future, although total automobile ownership is to some extent a function of the number of households in a given population, and to the age composition of those households. But at the least, total freeway capacity is adequate; the difficulty is in the peak hour concentrations. Some staggering of work hours in certain areas could spread this load over excess capacity at other times.

In the very long run the freeways may suffer attrition, but the long run future of the automobile as we know it may change. If vehicle characteristics change, for example, the presently rated capacity of the freeways could increase. In the very long run the possibilities fro numerous adaptations increase dramatically. What is clear is that the urgency of an enormous investment in fixed rail at this time is not such as to warrant the cost, even at the cost estimates of the proposal now before us. For these numerous reasons, I cannot endorse the ballot proposition on June 6th next year.



Harry W. Richardson
University of Southern California

David Freeman, the energy specialist once said: "There are only two things wrong with coal. One, we can't mine it. Two, we can't use it". Analogously, there are only two things wrong with the Sunset Coast Line. "One, we can't afford it. Two, we won't ride it".

Clearly, the question of cost requires a little qualification. However, we should not delude ourselves. No one with any experience of the differences between projected and ex-post costs of large-scale capital projects believes that the actual capital costs will not be many times greater than the current figures (even after deflating for increases in the general price level. Whether it is nuclear power plants or Concordes the litany of cost overruns is depressingly familiar. Unfortunately, that is not all. The more frightening burden is the eternity of operating losses, a tribute levied on all to sustain what I believe will be the mobility of a minority. Indeed, to paraphrase Winston Churchill: "Never in the field of human transportation will so much have been paid by so many to carry so few".

The ridership question is, of course, critical. We can make some "guesstimates" if we assume we know the routes. We forecast commuter flows along the selected corridors, and then estimate the modal split. Let existing freeway system. Thus, the rail and the freeway will be in direct competition. Since commuters are very sensible people, they will be very conscious of door-to-door travel time and of the time losses involved in transferring and waiting at park-and-ride facilities. Thus, many of those further than walking distance from stations (and the American commuter's concept of walking distance is very different from that of his European counterpart) will choose to drive by car. Moreover, to the extent that it attracts riders a rail system that duplicates the freeways will make auto commuting faster and less congested. All the evidence everywhere points to a very small diversion rate from autos to rail. And with duplicated routes the freeway will even be able to compete for newly generated traffic. The likely equilibrium is where both systems run at well below maximum capacity.(1) To pay billions of dollars for half-empty trains and half-empty freeways hardly seems a wise investment.

Apart form doubts about the ability of the train to capture enough of the freeway traffic (the modal split issue), another problem with the ridership estimates is whether it is reasonable to assume a marked increase in corridor traffic flows, since newly generated traffic is critical to the viability of the transit system. Two points are relevant here. First, population projections for the area are being almost continuously revised downwards. It becomes increasingly dubious, for instance, whether we can expect much increase in population of Los Angeles County over the next two or three decades. Second, the trend towards decentralization and the diffusion of workplaces and employment centers continues. With the exception of the route from LAX I do not see the current proposal making much of a contribution towards dealing with this dispersion trend. In particular, the priority axes seem to be overwhelmingly downtown - directed at a time when the competitive position of downtown is becoming, inexorablky I think, progressively weaker. By the time the first phase is built I would estimate that downtown will no longer be the leading employment center in the metropolitan region.

That raises another issue. Mass transit supporters and downtown revivalists are hand in glove (e.g. Mayor Bradley's support for the project). In fact, the relationship between an express transit route and downtown revitalization is very obscure (on this point I am less convinced than professor Dyckman about the benefits for downtown business). The common assumption in this region is that mass transit would be a shot in the arm for downtown. In fact, the only way in which this might happen would be if hasty and speculative office construction decisions were made prior to the system being completed. If downtown was strengthened in this way (an extension of the San Francisco paradigm to a very different metropolitan area), there is a risk of additional road congestion unless the modal transfer is much higher than that achieved by BART.

However, revival of downtown is a most unlikely consequence. A transportation improvement tends to make the suburbs more attractive and flattens the density gradient (e.g. promotes population dispertion). Dennis Capozza (of this university) showed this result very clearly in the case of a mass transit facility in a paper published a few years ago. If the suburbs gain by the way, it becomes clear that mass transit is against the poor rather than for the poor. The gainers are the relatively well-off suburbanites. The losers are the general taxpayers, and especially the poor if the project is financed out of repressive sales taxes.

Another myth is that a rail system, or indeed any other form of new transportation system, is needed because of congestion on existing roads. This argument assumes that the goal of zero congestion is desirable. In fact, there is a level of optimal congestion. Any efficient transportation system will be congested in peak travel hours. If rush hour traffic were not congested, this would be a symptom of gross inefficiencies, and would indicate that we had spent far too much on the transportation system. Unless we could equalize the number and distribution of trips in each hour of the day (clearly we cannot), the more we aim for uncongested peak travel the greater the cost in terms of chronic underutilization of the transportation network at other times of the day. Professor Dyckman's paper correctly emphasizes the severe utilization problem with mass transit.

Talking about myths, there is wild talk about the influence of energy supply constraints on travel behavior. In fact, there is no "energy crunch" of a kind that might induce a transfer to rail. First, there is little possibility of a kind that might induce transfer to rail. First there is little possibility of an oil shortage over the next 10-15 years, while beyond the 1990's petroleum may be obsolete as an automobile fuel. Second, in the shorter run, it is unlikely that oil prices will rise much faster, if at all, than the general price level. Even if they do, there is no evidence that higher oil prices would induce a shift to other modes. Consider European experience, for instance. Gasoline prices are more than double the US level with a negligible impact on travel behavior. Also, the urban rail systems in European countries remain in poor shape. Furthermore, the most likely medium-term response to high gas prices in the US would be a shift to smaller cars. Considerable scope for this remains. Indeed, the current revival in the sales of mid-sized cars is convincing evidence of what people in general think of the energy constraints argument.

Almost all transportation experts agree that a fixed-route rail system is much too inflexible in view of the continuous changes in metropolitan structure. The gestation period of this, or indeed any other, regional-scale mass transit project is so long that we have no idea at all that the later routes to be completed will be appropriate to the needs of their time.

To sum up, there is no evidence anywhere in the United States, and very little evidence elsewhere, that a new mass transit system is effective in the sense of accommodating a sufficient share of total traffic (even peak traffic) to justify the heavy capital costs and the large operating subsidies. If mass transit systems do not work elsewhere, how can they be expected to work in Los Angeles? Residents here tend to underrate the high quality of the existing transportation system (i.e. the freeways); newcomers with experience of other cities and other countries do not make this mistake. Also, the decentralization spatial structure with highly dispersed workplaces and residences and very few really high density corridors is a less than ideal environment for a mass transit facility. A more sensible, less costly and more viable strategy is to work withe the existing system but to introduce improvements, e.g. relaxing restrictions on paratransit modes such as subscription buses, commercial vanpools and jitneys; better traffic management; incentives to staggered working hours; more express lanes (although not by cheapjack methods such as taking away existing lanes - - unless stronger and more palatable incentives are offered).

In Europe as the rail lines have been closed down they have been frequently bought by train buffs who spend their own money on running the trains at weekends. Baxter Ward reminds me of these, but with a not insignificant difference - - he wants the taxpayer to finance his hobby.

Mr. Beaty's paper is disappointing in the sense that he fails to address the issues raised by Professor Dyckman in regard to the economic and social justification for rail transit in general and the Sunset Coast line in particular. Instead, he assumes that the time for such argument is past. Consultants have been paid $5 million to reveal the hidden benefits of rail transit so they must be true. The critical question in the strategy of the rail transit so they must be true. The critical question in the strategy of the rail transit protagonists is how to gain political support for the system: hence one that runs through all cities regardless of their potential corridor densities; hence a system built on freeway medians to avoid the obstacles of environmental impact evaluation and the need to purchase rights of way; hence the need for quick action to prevent available Federal funds being absorbed in other cities; hence the stress on what can be built with a 1 percent sales tax, regardless of whether or not this is an appropriate scale.

Focusing on these issues begs the question of whether we need a rail transit system. Telling us that we need it is insufficient. In fact, it is "topsy turvey" politics. The primary objective of politics in a democratic system should not be to tell us what we must do, but to provide the arrangements under which we can decide what we want and to make sure that our representatives do what we, in effect, tell them. Nor is a referendum in itself a big enough "sop" for democratic decision-making unless the issues are freely and openly discussed beforehand. The "general will" is not necessarily equivalent to the people's will.

The argument that we have hesitated for so long (fifteen, twenty-five or how many years it may be) that it is imperative to act now is also weak. Rather the politicians are behaving as if their minds are riding on a fixed rail track and they are unable to get off. As Los Angeles has debated the pros and cons of a rail transit system, the world has changed and such a system is less viable than ever.

In the closing minutes of his talk Mr. Beaty speeded through a list of justifications for a rail transit system. His analysis raised far more questions than it answered. In what sense are the existing freeways "overused"? What is "urban sprawl", and how would the rail system eliminate it? Why should increasing energy prices induce a major shift from automobile to rail, and what are the grounds for believing that incremented energy costs justify a multi-billion dollar rail system? What is the justification for a one million per day ridership estimate in view of the failure of other rail transit system to reach their forecast targets? Why does the size of the project strengthen to reach their forecast targets? Why does the size of the project strengthen the case for building it except to soothe the massive (but unjustified) inferiority complex of Los Angeles relative to other major world cities? The criterion of "the biggest and the best" has no place in cost-benefit evaluations. As for arguing that building the system would create jobs, of course $6 billion of expenditures will boost employment. But if maximizing employment was the goal, it would be easy to find a way of spending that sum to create even more jobs. Also, of course, the higher sales tax would have adverse income and employment effects. Finally, justifying the project in terms of induced business investment is mere speculation. No one knows what the investment effects would be, but it is probable that most of them would not be newly generated but merely relocated from other sites in the metropolitan region. If there is a case for a rail transit system, it is not to be found in Mr. Beaty's paper.

Top | Home | Contents | Next