Abstract: Variations in property rights explain a wide range of market experience in jitney and bus services. Scheduled bus service entails route specific investments and cultivation of a market. If these investments can be expropriated by interloping jitneys, scheduled service will be dissolved. Property rights in curbspace determine whether scheduled service will be preserved, and whether jitney services will co-exist. We analyze the dynamics of thick and thin transit markets, with and without curb rights. We develop a governance system of "curb rights" that would let bus operators appropriate their own investments in scheduled service, yet would avoid monopoly by letting jitneys and competing scheduled services operate along the same route. A system of curb rights dispenses with government ownership, franchise contracting, and regulation.

Acknowledgments: This paper is a condensation of the authors' book, Curb Rights: A Foundation for Free Enterprise in Urban Transit (Brookings Institution, 1997). The authors would like to thank Pete Fielding, Pia Koskenoja, Charles Lave, James Nolan, and Ken Small for valuable comments. For financial support, the authors thank the California Department of Transportation (contract RTA-65V450) and the University of California Transportation Center.

I. INTRODUCTION

Urban transit in the United States has long been dominated by government ownership and regulation, and has been declining steadily in terms of ridership and productivity (APTA 1995). An economist-cum-policymaker would seek to inject competition and entrepreneurship into the sector by privatizing. The two types of privatization often advocated are contracting out and "free competition" (Department of Transportation 1984; Lave 1985; Gomez-Ibanez and Meyer 1993). Experience however has shown that each approach has serious shortcomings.

Contracting out allows the public sector to maintain the planning decisions of routes and fares, and the types of vehicles to be used, while putting production and operations in the hands of cost-conscious private companies. Small cities and counties have increasingly contracted out bus service. Larger transit agencies have a harder time putting through major contracting programs in part because of privileges granted to transit unions. Contracting has reduced costs significantly (Teal 1988, 218f; Perry et al 1988, 134f), but contracts, even when competitively let, preserve transit monopoly and service regimentation. Transit agencies use various contracting schemes which have been criticized by Williamson (1976) and Goldberg (1976) because they tend in practice to resemble regulated monopoly.

The second proposal, "free competition," promises on-the-road competition, perhaps in the form of freewheeling jitneys, which are small vehicles that pick up and drop off along a route but do not necessarily follow a schedule. The deregulation or "free competition" precept is incomplete, however, when applied to a service that operates on government property, namely the roadway, curbspace, and sidewalk areas where passengers congregate in waiting. Bus operators must invest in cultivating passenger congregations, and must be able to appropriate the value of their investment. Depending on how it is governed, "free competition" might mean parasitic interloping on routes, where jitneys run ahead of scheduled buses to pick up waiting passengers. Such interloping might undermine any scheduled service and inhibit development of transit markets. All this takes place on public property where market mechanisms are lacking.

Calls to merely privatize the buses and to deregulate bus operations have neglected crucial issues rooted in the management and utilization of the public domain. They ignore curbspaces as a fundamental resource of the industry. We argue that the rules -- or property rights -- governing passenger pick-up areas are in fact a determining feature of transit markets. We suggest that variations in curb rights explain the differences in transit markets seen in the United States and elsewhere. An appreciation of curb-rights issues leads to a better understanding of transit markets.

We proceed by first examining four case studies of transit markets with deficient property rights: the jitney episode in the United States, 1914-1916; jitneys and route associations in less developed countries; illegal jitneys in New York City; and the British experience of bus privatization and deregulation. These case studies help us to develop a logic of transit operations, and to formulate a theory of transit markets. Finally, we propose a system of "curb rights" which promises to improve transit markets.

II. TRANSIT MARKETS WITH DEFICIENT PROPERTY RIGHTS:
FOUR CASE STUDIES

The Jitney Episode of 1914-1916

When the automobile came on the scene, so did freewheeling competition in urban transit. Jitneys charging a nickel per ride picked up waiting passengers along the routes of the electric streetcars. The jitneys were usually just the sedans of the day, serving as shared-ride taxicabs along loosely defined routes. They quickly became popular because of their flexibility and speed -- almost twice that of the streetcars. They were more comfortable and less crowded, and sometimes would deviate from the main route to make courtesy drop offs. By 1915 jitneys operated in most major cities and reportedly numbered 62,000 nationally (Eckert and Hilton 1972, 295-296; Saltzman and Solomon 1973, 63).

Streetcar companies immediately reported losses due to jitney competition, and many began laying off employees and cutting back on service. Jitneys not only interloped on streetcar routes, however. They also filled important market niches. Jitneys were used mainly for short trips and provided transportation to many people who would otherwise not have been served by the streetcars. Although jitneys charged no more than the streetcars, their gross revenues far exceeded the streetcars' loss of revenue (Eckert and Hilton 1972, 296; Rosenbloom 1972, 5).

The jitneys were loosely organized and highly spontaneous. Most jitney drivers were independent, some between jobs or working part-time to supplement their income. Many were simply working people who picked up fares on the way to their regular job. Others were teenagers who borrowed their parents' car to earn spare change after school (Eckert and Hilton 1972, 294). Jitneys adapted flexibly to changing demand conditions -- such as the weather, time of day, day of the week, special events, an so on. Despite the decentralized nature of jitneys, there began to emerge customs, voluntary associations, and company fleets. These associations were formed to help drivers obtain insurance, share maintenance services, and protect them from hostile lawmaking, and in some cases to coordinate routes and schedules (Eckert and Hilton 1972, 295-297).

The electric streetcar companies saw the jitneys as an infringement on their exclusive franchises and lobbied the government to regulate the jitneys. The municipalities went along with streetcar demands, in part because the streetcars afforded them tax revenue and free movement of police and fire department personnel (Hilton 1985, 37). Municipalities required jitney drivers to obtain substantial liability bonds, and to obtain operating permits. These measures and other anti-jitney ordinances proved fatal. The jitneys largely disappeared by 1917, after just two years of rapid growth and experimentation.

The jitneys posed a fundamental question of property rights: Is interloping on scheduled service a form of thievery or a form of legitimate competition: The authorities decided it was "thievery," plain and simple, and instead of developing a framework that could accommodate competitive co-existence, they stamped out freewheeling transit in favor of large-scale monopoly.

Transit in the LDCs: Jitneys and Route Associations

Transit services similar to the 1915 jitney experience still operate on the streets of hundreds of cities throughout the less developed world. Takyi (1990, 171) describes the jitney's appeal to riders:

They charge relatively low fares and provide wide coverage across a city, often serving poor areas that get no other service. Their operations are flexible so they can add service at peak times and quickly cover new neighborhoods. Their small size and cheap labor enables them to profitably provide frequent service in smaller neighborhoods and along narrow streets, as well as work the main thoroughfares. With fewer passengers, they often make fewer stops and faster time.

These advantages also marked the American jitneys of 1915, but in that case regulations were imposed to undercut their competitive advantage. In the LDCs, laws are passed to prevent jitneys from interloping on official service and from establishing competing routes, but the rule of law is very attenuated and, as Takyi says (1990, 175), the jitneys "never operate legally." Takyi tells of "the loss of passengers at transit stops to jitneys during lean as well as peak periods."

As Jitney service develops in thick transit markets, various curbside conflicts and confusions start to occur. Any operator who attempts to establish scheduled service will face an interloping problem. Some operators will head run on the scheduled service, others will linger at the curb to fill up, disrupting traffic and taking ridership from the arriving vehicle (Roth and Shepherd 1984, 4; Diandas and Roth 1995, 27-28; Takyi 1990, 167, 175). Consumers may be reasonably well served, but problems of discoordination and lack of trust are often severe (Grava 1980, 285).

A common development is for the jitney operators to form a route association. These are informal organizations created to bring order and regularity to service, by creating an extralegal system of norms and explicit rules. The jitney literature suggests that route associations have in large measure governed transit services in Lima (De Soto 1989), Hong Kong, Istanbul, Buenos Aires, Manila, Calcutta, and Caracas (Roth and Shepherd 1984; Takyi 1990). The route association becomes a regulatory body, somewhat like government, but more local and entrepreneurial in nature. The association lays down rules against interloping and deviating from schedules. Also they fix fares on the route, although these may vary with time of day. Associations create a degree of order sufficient to control destructive conflict, but they also operate as a cartel. Roth and Shepherd (1984, 42), De Soto (1989, 99), Grava (1980, 282), and Cervero (1996, 132, 149) report that associations function to limit entry.

Thus we arrive again at the issue of rights to waiting passengers -- or curb rights. Jitneys initially transgressed the curb rights of the formal bus operators, yet in time organized to establish curb rights for themselves. How, then, do they prevent new interlopers from transgressing their rights? The main answer seems to be physical intimidation and strong-arm tactics. Roth (1987, 224-25) notes that "the methods used by route associations to protect their territory can become criminal, unlawful, perhaps even homicidal." Sigurd Grava (1980, 282) speaks of route enforcement by means "considerably beyond the law" by "district strongmen, ... local bosses, criminal gangs, powerful families, brotherhoods of operators or otherwise legal associations." As is common in black markets everywhere, outlaw entrepreneurs employ violence to maintain their territory. De Soto (1989, 102) tells of route associations in Lima appointing "dispatchers" to monitor compliance with rules, and bribing the police to accost and harass "pirates" who are trying to invade their route.

Once route associations have organized their operations, they often turn to the government for official recognition. Through a long effort of lobbying, bribery, petition gathering, and so on, the route association often acquires official status, and are granted permits or licenses. Along with official recognition, however, come political obligations and regulations. Transit history in Colombo (Diandas and Roth 1995) and Lima (De Soto 1989) shows a cycle of transit governance: once the decentralized private operators gain official recognition they are hamstrung by regulation and suffer invasion by a new generation of interlopers. Without curb rights, established officially or otherwise, orderly scheduled fixed-route service does not last.

Illegal Jitney Vans in New York City

Black-market jitneying is not restricted to the less developed countries. In New York City and Miami, jitney vans have operated extensively, interloping at public bus stops and establishing routes of their own. People who ride the illegal vans give a number of reasons for preferring them to the city buses.⁽¹⁾

By far the most often mentioned is that the jitneys are faster and even cheaper than the city buses. Jitneys also provide a more comfortable ride, with no standing, and many riders enjoy having a driver who speaks their native language. Finally, many riders say that riding the jitney is safer than the public bus. Since jitneys come more often, riders do not have to wait as long at the bus stop, where one runs a risk of being mugged (Levine and Wachs 1986). Also, jitney drivers will not pick up passengers who are drunk and disorderly, or who otherwise bother or threaten the other passengers. Jitney riders, who are mostly minorities, appreciate being able to escape the forced association with all comers that a public bus entails.

Extreme cases of interloping jitneys may develop where market conditions are favorable and enforcement efforts not yet mobilized. To persist once enforcement begins, interloping must expand to the point where the individual illegal operator finds safety in numbers, like someone taking part in a riot. This jitney outbreak either persists as a significant force or disappears. In most U.S. cities, either market conditions have not favored illegal jitneys or enforcement has been effective.⁽²⁾

A notable instance of such jitney outbreaks is New York City (on Miami's jitneys, see Klein et al 1997).

Modern jitney operation in New York City was prompted by the transit strike in 1980. Illegal jitneys emerged to provide local service and feeder service to the Long Island Rail Road station in Jamaica (southeast Queens). As Boyle (1993, 3) explains, "[the jitneys thrived along busy bus routes ... because of the high numbers of people congregated at the bus stops along these routes." Boyle reasons that jitney service has developed especially in neighborhoods of Caribbean immigrants, because those riders became accustomed to relying on jitneys in their native lands. Although the strike had long been settled and regular bus service reinstated, enforcement against the jitneys had been only "sporadic" (Boyle 1993, 3). Jitneys reached the "take-off" point to self-sustained operation. The authorities face the dilemma of cracking down on services which are well regarded by paying customers and treated sympathetically by reporters and news commentators.

To operate legally the vans would have to obtain special permits and a special insurance policy and undergo multiple inspections each year, and the driver would need a special license. The vans could then pickup and discharge passengers only by pre-arranged appointment, and of course not use city bus stops. It is estimated that between 2,500 and 5,000 vans flout these laws (Boyle 1993, 4).

A public transit executive claimed that each year the jitney vans were diverting $30 million in revenue from public transit (Machalaba 1991). Transit police have been assigned to areas near bus stops to crack down on the interlopers. The New York Times reports: "In the 18-months ended December 1991, a special task force issued 6,542 civil notices of violation against the vans and 11,773 criminal summonses, ... [and] 251 arrests" (Mitchell 1992). Still, the vans are thought to be uncontrollable. A police officer remarks that two or three vans sail by for every one he tickets, and van drivers pay small regard to the summonses. The Wall Street Journal (Machalaba 1991) reports that over a one year period the van drivers have been assessed fines over $4 million, but the city collected only $150,000. Considerations of racial flash-points dampen the will to go beyond current enforcement measures, which amount to random delay and hassles for the drivers and their patrons.

The New York jitney experience shows again that unsubsidized private enterprise can supply fixed-route transit, even when having to cope with enforcement efforts against them. We see also that property rights to waiting bus passengers, and the degree of enforcement, is a fundamental component in such operations.

Bus Privatization and Deregulation in Britain

The 1985 Transit Act deregulated the British bus industry everywhere except London.⁽³⁾

(In London competition is required only as competitive contracting; there is no on-the-road competition.) All publicly owned bus companies were reorganized as private corporations. The Act requires operators to register the commencement of, or changes to, bus service at least 42 days in advance. The only grounds for local government to refuse to allow a service is serious safety or traffic congestion problems. Besides privately registered routes, local authorities can supplement services by putting unserved routes out for competitive tender.

Deregulation permits only scheduled services, not unscheduled services, such as jitneys. This restriction, along with the strength of the rule of law in Britain, ensures that there have been no freewheeling jitney activity, and no interloping as seen in the less developed countries.

On-the-road competition was initially strong, but has tapered off to a rather low level (Dodgson 1991, 125; Hibbs 1993, 52). Besides on-road competition, however, is the question of contestability, or the ability of potential entry to discipline incumbent firms. Mackie, Preston, and Nash (1995, 232) and Dodgson and Katsolulacos (1991, 265-6) suggest that contestability is constrained by the sunk costs of establishing a scheduled service, and the "economies of experience" held by incumbent operators. Another constraint of contestability, which they do not mention, is the ability of an incumbent firm to react quickly to a competitive challenge. Contestability theory suggests that if an incumbent firm can quickly and easily reduce its fares when a competitor challenges it, would-be entrants might be reluctant to enter, even in a market with high fares (Bailey 1981; Bailey and Friedlander 1982). The challenger can no longer expect to grab market share by offering a lower price, and the incumbent has the advantage of experience, reputation, and, in most cases, size.

In fact, it has been very rare in the British experience for firms to compete by offering lower fares (Dodgson and Katsoulacos 1991, 271-2). Rather, real bus fares increased 17 percent between 1987 to 1994 (White 1995, 198). Instead of competing by offering lower fares, firms chose to offer more frequent service than their competitors. Free competition does not necessarily generate price cutting, as has also been found in deregulated taxi markets (Frankena and Paulter 1986; Teal and Berglund 1987). It seems that information and coordination problems between drivers and potential riders may push transit markets toward a single, or focal, rate of fare.

Under the British reforms, registering a scheduled service does not secure one a right to the congregating passengers at the curb. One bus operator can interlope, in a manner of speaking, by registering his own scheduled service just minutes before the scheduled service of another. Since the law does not proscribe schedule matching, local authorities are obliged to allow it. Many British bus operators avail themselves of this strategy, which we call schedule jockeying (Dodgson 1991, 126; Dodgson and Katsoloucos 1991, 269; Savage 1993, 146; Gomez-Ibanez and Meyer 1990, 13). Since the established firm has no window of security from the schedules of competitors, congregations of passengers waiting at the curb can be snatched up by competitors offering comparable fares. Waiting time so dominates passengers travel decisions that any reputation and amenity advantages an incumbent may offer are not likely to keep waiting travelers from taking the first bus to arrive (Weisman 1981; Wachs 1992; Dobson and Nicolaidis 1974).

Incumbent bus companies, however, quickly learned to monitor the registration of new services by competitors using this strategy, and often promptly respond in kind. The 42-day registration period makes it easy for firms to see each other's changes in service and to respond, in a potentially endlessly regress. In the face of a mutually destructive battle, the incumbent has often responded by simply scheduling service so frequently that the challenger cannot expect to get enough riders to make a go of it. The practice, known as route swamping, has been very common (Dodgson 1991, 126; Dodgson and Katsoloucos 1991, 269; Savage 1993, 146; Gomez-Ibanez and Meyer 1990, 13). It has a strategic effect in the immediate contest, and in signaling the willingness to use route-swamping against future challenges. The ability of incumbent firms to quickly and easily change their schedules in reaction to entry, by virtue of the 42 day registration period, constrains contestability in the same way that easily and quickly adjustable prices do in standard contestability theory.

Large incumbent firms often conclude a route swamping conflict by buying out small rival firms. These types of mergers are quite common, as are mergers between firms that do not directly compete against each other (Mackie, Preston, and Nash 1995, 235; Savage 1993, 147). Many of the latter mergers have been in the form of holding companies, with their subsidiaries often geographically dispersed (Gomez-Ibanez and Meyer 1990, 12-13). The result has been a clear trend towards oligopolistic and even monopolistic operations in the industry, another important unexpected outcome of deregulation (Banister and Pickup 1990, 81; Savage 1993, 147). In many counties just a few firms control over 80 percent of the market (Banister and Pickup 1990, 81). Small operators have been progressively squeezed out of the competitive market, while finding more success in the tendered contract market.

The literature offers a plethora of explanations for the concentration of the industry. Hibbs (1991, 4) suggests economies of scope and management efficiencies. Mackie, Preston, and Nash, (1995, 235-36) and White (1995, 202-3) point to financial advantages of larger firms, managerial economies of scale, and purchasing power. Gomez-Ibanez and Meyer (1990, 12-13) argue that holding companies offer many advantages, including very low costs and the ability to move vehicles and mangers from subsidiary to subsidiary as market conditions dictate. they add that firms with large networks have a distinct advantage in the growing use of single-rate unlimited-travel fare cards. Nash (1988, 110) indicates that larger firms enjoy considerable economies of scope in scheduling buses and avoiding long layovers between runs. Finally, Dodgson and Katsoulacos (1991, 267) advance the notion that to some extent the managers of formerly public firms may have retained their old habits of output maximization, even though they are inappropriate for the new goal of profit maximization. The issue of integration calls to mind yet another explanation. Dodgson (1991, 124) and Nash (1988) point out that there has been a steady decline in inter-operator ticket availability. White (1992, 56) mentions one case where the removal of schedule coordination and inter-operator ticketing led to a 20 percent reduction in ridership.⁽⁴⁾

Although we think that there is merit to many of these theories, we think that the issue of curb rights is fundamental to the peculiar form of deregulation in Britain. If we accept that the ability to swamp a route is necessary to combat schedule jockeying, it is easy to see advantages in larger firms with broader networks. As Gomez-Ibanez and Meyer (1990, 13) point out, a larger company has more supervisors, drivers, and buses at its disposal, which they can shift about to swamp a route where competition has commenced schedule jockeying. A larger firm will also have greater financial flexibility to maintain such "fighting fleets" (Dodgson and Katsoulacos 1991, 267, 270). Indeed, the very largeness of the firm presents a formidable warning, signaling to potential entrants that entry can and will be met by swamping. Although British deregulation of buses has led to large reductions in costs and public subsidization (White 1995, 194; Mackie, Preston, and Nash 1995, 238; white 1992, 50), it has also yielded a surprising degree of industry concentration, with lackluster competition. Our theory of schedule jockeying and route swamping, rotted in the issue of curb rights, helps to explain these results.

III. A THEORY OF TRANSIT SERVICES

The Market Advantages of Jitneys

The American experience in 1914-16 and that of transit markets in some less developed countries today suggests that jitneys have market advantages over scheduled bus service on both the supply side and the demand side. Because jitney operators follows a route but not necessarily a schedule, they enjoy efficiencies in being flexible with respect to their own schedules and to changes in weather, congestion, time of day, day of week, and so on. They enjoy flexibility in negotiating traffic conditions and can make small deviations from the route. Under a free entry policy for jitneys, one could expect a cascade of irregular, short-term participants.

On the demand side, passengers waiting for a scheduled bus are generally quite happy to ride a jitney that charges a comparable fare and goes to the same destination. The head running jitney offers several advantages. It is available "now," whereas the bus is yet to arrive. It is a smaller, faster, and probably more pleasant, and may offer deviations from its route, perhaps at an extra charge. The bus is cumbersome and dreary; the jitney is entrepreneurial, more personalized, and even somewhat charming (Takyi 1990, 165). Still, patrons may prefer to wait for the scheduled bus because it offers more certainty and trustworthiness, and it is perhaps more comfortable than the jitney (Grava 1980, 285). In what follows we posit that passengers generally prefer to ride in the head running jitney that charges the same fare as the scheduled service.

Appropriability of the Investment in Scheduled Service

If, in the presence of scheduled service, jitneys enjoys inherent market advantages, a fundamental issue affecting the fate of scheduled service is whether jitneys have free run of the streets. It is the issue of curb rights that determines whether jitneying will flourish. In the experience of America in 1915, of illegal jitneys in America today, and of jitneying in some LDC cities, scheduled services do not enjoy curb rights that fully established and exclusive, either because jitneying is legally permitted, or because it is prohibited but not effectively policed against.

Where interloping is both prohibited and effectively controlled, bus companies will invest in establishing routes and schedules, publicizing the information, and running the service in an incipient market, because they will be able to appropriate the value of these efforts at bringing people out to the curb. Although transportation economists have found that there are no economies of scale in merely expanding bus service (Viton 1981; Shipe 1992; Hensher 1988), they have neglected the issue of the appropriability of the investment in setting up and cultivating a route. We maintain that there are specific investments made in cultivating a route and schedule, and that the appropriability of this investment depends on curb rights. We assume that because jitneys enjoy inherent market advantages, if they are free to interlope, they will dissolve any scheduled service. Without the "anchor" of scheduled service, however, it might be that fewer riders will congregate at the curb and thus fewer jitneys ply the route.

Thick and Thin Transit Markets in the Absence of Curb Rights

Another distinction of fundamental importance is whether ridership on the transit route is potentially heavy enough to sustain the cascade of jitneys in the absence of scheduled service. If the market is potentially thick, a situation may develop in which there is no scheduled service, yet jitneys ply the route spontaneously because they have confidence in finding passengers, and passengers congregate at the curb because they have confidence in finding jitneys plying the route. The emergence of conventions that coordinate vehicle services and congregating passengers is found in America today in the cases of some commuter shuttle vans and carpooling practices (Walder 1985; Cervero 1996, 97).

In an inherently thin market this outcome, even if it were to exist at a point in time, cannot be sustained: there will not be an adequate number of passengers for jitney service to be frequent, so waiting times for unscheduled jitney service will be too long to induce passengers to congregate. Because the coordination problem of unscheduled service is severe in thin markets, service might disappear altogether.

The Thick Market: The Jitney Cascade Is Sustained

Consider the case of the potentially thick market with no exclusive curb rights and thus no scheduled service. The horizontal axis of figure 1 counts the number of jitneys per hour that ply the route. The vertical axis counts the number of passengers who congregate at the curb per hour. The figure contains two curves that represent functions which each map reflexively into the other's domain. The thick curve shows the number of jitneys that would come out to serve the route, given a number of congregating passengers. This jitneying function shows that no jitneys serve the route when there are no congregating passengers, but then as congregation grows, the jitneys begin to offer service. In the figure the jitneying response is shown as linear, but it might also be plausible to have the rate of increase declining because of congestion among jitneys. The thin curve shows the number of people who would congregate at the curb, given a number of jitneys serving the route. It also starts at zero, then rises at an increasing rate, but because there are only so many people who have any demand at all for jitney service, the curve eventually flattens out.

Figure 1. Interactions between Congregating Passengers and
Cruising Jitneys in a Potentially Thick Market

The curves show the mutual dependence of the two sides of the market. If only 60 people congregate per hour over the course of a week, about 6.7 jitneys an hour will respond. The next week, people expect about 6.7 jitneys an hour, and therefore only about 50 people will congregate. The next week, jitney operators expect only 50 people per hour, and the jitney function shows that therefore the jitneys come out in even smaller numbers, and so on. For a point to the lower-left of point Y, the system degenerates down to no market at all, point Z. At point Z, a stable equilibrium, it would make no sense for any jitneys to ply the route or for anyone to wait for a jitney.

Suppose that a critical mass develop beyond point Y, perhaps due to a transit strike, a coordinated effort by jitney operators, or an unusual event such as the Olympics or a hurricane. In that case the system will maintain life. If, for example, eight jitneys per hour were to ply the route, that would induce significant congregations, which would induce even more jitneys, and the system would bounce up to the other stable equilibrium at point X. Ten jitneys per hour induce exactly 100 congregating passengers, and 100 congregating passengers induce exactly 10 jitneys. This is the realization of potential in a thick market.

The Thin Market: The Dissolving Anchor

Figure 2 presents the thin-market but also posits that the market begins with scheduled bus service. This is the case of the "dissolving anchor." The scheduled service begins to operate and builds up a market. At first there is no jitneying, perhaps because jitney operators have not seen an opportunity before the development of this market or because they have not tested the powers of enforcement against interloping. With scheduled service and no jitneys, the number of passengers corresponding to point A wait for the bus. This passenger congregation is the "anchor" provided by scheduled service. Assume that for some reason jitneying suddenly becomes possible, perhaps because operators come to recognize that enforcement against interloping is weak or nonexistent. They begin to head run on the scheduled service, and many passengers are willing to take whichever vehicle comes first. The relationship between the upper congregation function (with scheduled service) and the jitneying function implies that the system will be driven to point B, where 9 jitneys ply the route and 100 passengers wait for service. Passengers and jitneys like this outcome, but there is one problem: the scheduled bus is now not getting enough ridership, and it pulls out. The anchor dissolves. Now passengers are less enthusiastic about congregating at curbside, for two reasons. First, they do not have the guarantee of anchor service, so they may have to wait longer or with more uncertainty for a carrier. Second, without scheduled service, there is no longer a focal schedule for arrival times at the stops. Jitney arrival times become less predictable. Hence, when an individual goes to the curb he goes with less certainty of when a jitney will arrive, and waits longer at the curb.⁽⁵⁾

The decrease in passenger enthusiasm is shown by the shift downward of the congregation function. Nine jitneys per hour now attract much fewer passengers. This in turn reduces the number of jitneys, which in turn reduces the number of passengers, and so on. Finally the system settles at point Z, for zilch. Thus, the progression is as follows: We begin at point A with scheduled service; jitneys come to interlope and the system moves to point B, the anchor is dissolved, and then the system moves to point Z, or market disintegration. In a thin market the jitney cascade cannot be sustained.

Figure 2. Interloping Jitneys Dissolve the Scheduled Service
and Destroy the Market

A Typology of Route-Based Transit Markets

We suggest a typology of fixed-route urban transit, shown in Figure 3. The top left cell represents unsubsidized buses with exclusive monopolies on routes with moderate to thin patronage. Exclusive rights are established: there is no interloping and no competition. Therefore the scheduled service is preserved. But there are some potential problems, the chief of which is inadequate competition and inert monopoly: the incumbent knows that entry is unlikely and consequently skimps on service or increases fares. Another problem is that potential operators will waste resources seeking the "rent" associated with monopoly privilege (Tullock 1967).

Figure 3. A Typology of Fixed-Route Urban Transit

In the bottom left cell the story is not much different. Again, scheduled service is preserved because interloping is not tolerated. Because the market is thick, it would be better able to support multiple scheduled services, more frequent service, and more consumer choice, but still competition is not tolerated. The problem of inert monopoly is more severe.

The middle cells present cases in which there are no exclusive rights, either because they are not granted or not enforced. The entire route is essentially a pure commons. Assuming that there is no impediment to headrunning or interloping, in a thin transit market, shown in the top middle cell, interlopers will headrun on any scheduled service and collect the waiting passengers. This is the case of the dissolving anchor. The lack of property rights in the waiting passengers results in the "tragedy of the commons" (Hardin 1968). The entire market may be destroyed, for, once the anchor has dissolved, people no longer wait and jitneys no longer ply the route.

The case of the thick market is shown in the bottom middle cell. In this case the lack of curb rights may not be a serious problem. Indeed, any scheduled service will be dissolved, but in a thick market scheduled service may not function as an "anchor." Combining elements of figures 1 and 2, visualize the lower congregation function (without scheduled service) in figure 2 as intersecting the jitneying function as does the congregation function in Figure 1. The market is thick enough to sustain the cascade of jitneys, and riders will be satisfied by flexible low-cost, and frequent service. This outcome was sometimes found in the American Jitney experience of 1915, a few markets in America today (both illegal and legal), and many of the LDC transit markets. In any country, however, there are possible problems with the jitney cascade outcome, such as low quality irregularity of service, high uncertainty over terms, and lack of trust.

A case that does not fit into the typology, that would go between the first and second columns can be imagined. In the British deregulation experience, bus operators enjoy neither exclusive monopolies (column one) nor operate on a pure commons (column two). Rather, free competition is permitted among providers who register schedules in advance. The situation is not that of the pure commons, since freewheeling is not permitted. The British example suggests that nuanced approaches can be fashioned to go between the two extremes of exclusive monopoly and pure commons. We will pursue this idea and propose a property rights framework that avoids monopoly by refining rights along a route.

The cases considered so far have assumed that any scheduled provider could enter the market, and that those that do receive no subsidies. Now consider the case in which scheduled service does receive subsidies (notably government subsidies, but much of the reasoning will work also for cross-subsidies). The existence of subsidies usually leads to very low fares. When subsidized service is combined with exclusive curb rights, we get cases similar to those we have placed in the first column of figure 3. The scheduled service, because it charges low fares, is now even more immune to interloping, so again it is preserved. The likely problems are inadequate competition and the familiar problems attendant to government subsidization.

The results of subsidized, low-fare service without exclusive curb rights are shown in the third column of Figure 3. Interlopers are free to headrun on the scheduled service, but in this case it is to no avail because patrons decide they will continue to wait for the scheduled bus, which offers a lower fare. For example, in Los Angeles in 1983 private jitneys were allowed to operate on thirty public transit routes. Matching the 85 cent public bus fare, they were successful initially, but promptly withdrew once the city lowered bus fares to 50 cents (Teal and Nemer 1986).

In a thick market, shown in the bottom right cell, the low fares of the scheduled service again will attract riders, but demand might exceed supply. One of the present authors has witnessed transit operations in Shanghai, where low-fare buses are packed sardine-style, and jitneys and taxis cater to the excess demand. In this case, jitneys survive because of excess demand and because they offer superior quality (less crowding, speedier service), even though they charge higher fares. Further, jitneys charge according to trip distance, so someone traveling a short distance might find the jitney fare competitive to the undifferentiated bus fare.

Now, imagine a decision to privatize and deregulate. If public transit subsidies are eliminated, there remains the first two columns of Figure 3. These two options represent the horns of a dilemma. In one case a provider of scheduled service has an exclusive monopoly over the entire route. Because there is no competition, there is little incentive for service improvement and innovation, and fares will be higher. In the other case, no exclusive rights exist at all. The anchor of scheduled service would be dissolved by jitneys, and markets may never come to be. If policy makers are confined to choosing between these two horns, they should choose on the basis of whether the market is thin or thick. If the market is thin, they should choose monopoly because the alternative results in no service at all. If the market is potentially thick, they should choose not to grant exclusive rights to the route and simply allow the jitney cascade to burgeon. This will bring freewheeling service and competitive energy to the market, whereas the alternative would be inert monopoly.

What would be even better, however, would be an option that avoids either horn of the dilemma, an option which entail a limited degree of exclusive rights, to prevent the anchor from dissolving and yet permits freewheeling competition on the route.

IV. GOVERNANCE FOR BUS AND JITNEY SERVICES: CURB RIGHTS

The answer lies in a system of curb rights that both guarantees some exclusivity to those who successfully cultivate passenger congregations, and gives play to the jitney cascade. There is no specific system of curb rights that is necessarily best for all transit conditions. An important part of our idea is that each case is unique, and that local officials ought to use their knowledge of local conditions to create a suitable curb-rights system.⁽⁶⁾

Spatial Demarcation of Curb Rights

A simple case would combine a scheduled service provider with the jitney cascade. Figure 4 is a schematic diagram showing curb rights as they are demarcated in both space and time. Consider first just the spatial component, where exclusionary zones are separated by a distance. The 8:00 column shows four curb zones. When we speak of "curb zones" or "curb rights," it should be understood that we mean not only the curb, but also adjoining space on the sidewalk and road - in other words, a complete bus stop. The column shows how Company A is granted two exclusionary curb zones where no other operators are allowed to pick up passengers. Think of each exclusionary zone as being 200 feet in length, with the bus-stop situated at the mid-point. Company A has every incentive to invest in creating passenger congregations at its bus-stops. It would establish a route and schedule, and be free from interloping. Yet along the same route jitneys meeting minimal safety and insurance requirements would operate, picking up passengers at non-exclusive zones, or commons. At the commons, passengers have an alternative to Company A, since most anyone may stop and offer service (including Company A).

Figure 4. Property-Rights Assignment to Curb Zones
A denotes a curb right held by Company A.
B denotes a curb right held by Company B.

Temporal Demarcation of Curb Rights

The notion of exclusionary zones may also be defined according to time intervals. Consider now the two peak-period columns, at 8:00 and 8:15. These illustrate the idea that curb zones may be exclusive for Company A during fifteen minute intervals, but then they become the "property" of Company B. This system may make enforcement more difficult, but time-elapsed video evidence still could show curb rights violations. This principle of exclusionary intervals speaks to the central failing of the British experience of bus deregulation: the ability of one company to insert its service just ahead of the competition's. This has led to schedule jockeying and route swamping, which disrupt service and diminish competitiveness in the industry.

We have argued that in a thin market, giving free play to the jitney cascade could possibly dissipate all service. Off-peak periods often correspond to thin markets. The off-peak times in Figure 4 shows an arrangement that precludes jitneys but accommodates competition on the route by granting exclusionary zones first to Company A and then to Company B. Instead of temporal alternation, local authorities might deem it wiser to have spatial alternation of A and B in the same column. Either way, this competitive arrangement would avoid monopoly, unless the two providers were to collude, and would give each provider an incentive to invest in building its ridership. It forgoes, however, the creative and highly efficient input of the freewheelers.

Auctioning Curb Zones

Now, look at figure 4 and in place of the As and Bs envision dollar signs --$$$. The authorities could set out exclusive curb zones and simply put them up for sale, perhaps in the form of five-year leases. The leases could be sold at a set price or auctioned off. Auctioning off the curb rights would avoid the hazard of monopoly power that could arise from a maldistribution of initial rights (Hahn 1984). The curbspace holder could then run its buses with stops in its leased zones. Under this plan, individuals with local advantages and knowledge of local opportunities could negotiate to make the most of the resource, and the one with the highest valuation would get the curbspace. Further possibilities emerge if the curb rights may be sublet or resold, which we think should be permitted. The holder may then wish to authorize other carriers to pick up in its curbspace, and require a monthly rental payment. Or it could resell the lease altogether to a provider with a higher valuation of the curbspace.

We can well imagine the emergence of professional curb zone entrepreneurs who lease available zones, sublet pickup rights to carriers, stage passengers and carriers, and monitor and police the curb rights. These leaseholders could even sublet to jitney associations, but they would manage this competition to protect their interest in their dealings with scheduled buses. Leaseholders could also profit by using the advertising opportunities on transit benches and shelters (Weisman 1984). In our visualization of a curb rights system, there would be at least four categories of participants: local officials, curb zone leaseholders, transit operators (bus companies, jitneys, and so forth), and passengers.

The scenario of having a market in curb rights might raise the specter of holding companies or "robber barons" buying up all the curb zones and exercising monopoly power over the route. The local authorities could mitigate this problem, however, in a variety of ways. In a thick market the most powerful method is for them to reserve certain curb zones as jitney commons, giving the jitney cascade play to compete with scheduled service. In a thin market with a monopoly problem, authorities could see to it that competing service providers each have their own curb rights.

Enforcement of Curb Rights

Any curb rights scheme will depend crucially on curb rights being enforceable. Enforceability may not with certainty be feasible, but there seems to be good grounds for optimism. Today in the United States it is only in very exceptional instances, such as in New York City and Miami, that the curb rights of official services are transgressed at all. In Britain where the scheduled service is typically unsubsidized, there is no interloping. Americans are mostly law-abiding, and local government can mobilize to protect curb rights.

Harold Demsetz (1967) has explained how one's effort to establish and enforce one's own property rights depends on the costs and benefits of doing so. Thus the local authorities ought to take measures that will reduce the costs and increase the benefits of enforcing curb rights. The holder of the curb rights should be encouraged to monitor the "property." Violations of the rights should be treated as a private torts as well as a municipal violation. Therefore, in addition to municipal enforcement efforts by traffic or transit police, curb rights holders could do such things as set up enclosed video cameras to watch for repeated trespass. The video footage of trespassing jitneys would simplify identification and apprehension and serve as evidence. Technology and official practice are advancing the photographic enforcement of traffic laws (Turner 1995; Blackburn and Gilbert 1995). Furthermore, the drivers of scheduled service could provide eye-witness accounts of headrunning. Suit could also be brought against riders of trespassing jitneys. The curb right holder could then put up signs at its bus-stops: "Mounting an unauthorized vehicle in this zone is a trespass and subject to civil suit." Travelers would find this reasonable because they could simply walk away from the exclusionary zone to wait legally for a jitney. By creating a sense of curbspace proprietorship and fair competition, both jitney operators and passengers would be likely to respect curb rights.

Emergence of Staging Areas on Private Property

The jitney commons zones are designated in Figure 4, but it might not be necessary to pre-establish such zones. It might be sufficient to say that jitneys cannot pick up in the A zones, and to let their own curb zones and staging areas emerge spontaneously. Local officials may wish to manage the emergence of such pickup spots to avoid sidewalk congestion or to provide transit focal points, but if certain places seem to be emerging as workable jitney spots, they ought to smile on the development. They may wish to alter parking or standing rules at such spots, and perhaps even provide turnouts, benches, and shelters. Imagine a McDonald's restaurant emerging as a jitneying point, where travelers can buy breakfast and organize shared rides on innovative McTransit (Rehmke 1991). If the McDonald's began to charge for day-time parking or to cooperate in announcing or arranging jitney departures, this ought to be regarded as legitimate private enterprise. Throughout the city, entrepreneurs may find it profitable to develop jitney staging areas on private property, and jitney associations may want to negotiate a system of such spots with local businesses. In our scheme, local officials are not primarily regulators; they are creators and enforcers of property rights. Provided that jitney operators and staging entrepreneurs do not tread on the property rights of others, they should be allowed to operate unencumbered.

IX. CONCLUSION

We have developed a theory of scheduled bus service which recognizes the importance of generating passenger congregations. Furthermore, the investment in cultivating passenger congregations must be appropriable, or protected from interloping. We have reviewed the literature of numerous transit markets, showing that they tend to be gored by one of the two horns of a transit dilemma. Some markets enable scheduled operators to appropriate the value of passenger congregations, but this is achieved by granting exclusive rights, not only to the waiting passengers, but to the entire route. Thus the first horn is transit monopoly. The other horn is the pure commons, giving rise to freewheeling competition like that found in some LDC cities. In this case there is no cultivation of passenger congregations for scheduled service, because interloping will expropriate the investment. In consequence, thin markets are especially poorly served.

A nuanced approach based on property rights can maneuver between the horns of this dilemma. We can have the best of both cases, schedule service and freewheeling jitneys. Figure 5 revises Figure 3, by inserting our solution between exclusive monopoly and the pure commons (and by eliminating the case of subsidized service).

Figure 5. A Typology of Unsubsidized Fixed-Route
Urban Transit Incorporating Curb Rights

Back to Dan Klein's Homepage