especially if the driver leases his taxi. A fare structure that does not produce the necessary revenue requirement will result in poorer service and the exit of drivers from the industry.

Equity in the rate structure is nearly as important as In the taxi industry, this means that there should not be "good" trips and "bad" trips. In economic terms, a good trip is one that produces revenues above average cost, and a bad trip is one that produces revenues below average cost. While costs will vary by trip according to length, traffic conditions, and other factors, the fare determination system should recognize these differences as closely as possible. As long as fares are close to the costs experienced on individual trips, drivers will not prefer to make certain trips and not others. When there are good and bad trips, the situation does not balance out, as is sometimes assumed. Consumers paying for good trips will be overcharged, and consumers wanting the trips that are bad may not find a taxi available at all; in a sense no taxi is paying the highest price of all.

the recoupment of the revenue requirement.

The third objective, as stated above, relates to the proper allocation of economic resources in the District of Columbia. Taxi trips that do not cover their economic costs should not be encouraged through low fares, and taxi trips that do cover economic costs should not be discouraged through high fares. If the fare structure does not provide the proper incentives for the taking of cost-justified trips, taxi service will not fulfill its proper role in the District of Columbia. Certain trips that should go by taxi will be made by other modes of transport and vice versa. Thus, it should be fully recognized that the fare structure used in the taxi industry plays an important role in the use of taxicabs in the overall transport mix in D.C.

The fare determination system should be easy to understand, perhaps even at the expense of some simplification. While the proper following of costs is essential, as described above, so is a proper understanding by consumers. Consumers should know how they are being billed, why the billing system is designed as it is, and even what the cost of a specific trip will be. Then they will feel comfortable in knowing that they have the money to pay for their trip and not feel they can be cheated by the system.

The fare determination system should be as low in cost to administer as possible. In other words, there should not be a significant cost associated with administering the fare system, as this will add to costs and increase the price of riding a taxi.

Finally, the fare determination system should be as stable as possible so customers can come to know and understand it. This is usually an advantage of an existing system. Any method of fare determination takes time to learn and understand. Thus it is important that there be as few changes as possible in the system so that customers do not have to go through a learning process any more than necessary.

This section discusses the available methods of fare determination and includes analyses of the fixed zone, circular zone, and meter systems. It does not cover possible systems that are not now in use and that would require measuring equipment that is not yet available.

The fixed zone system is the one that is currently used in the District of Columbia. Under this method of fare determination,

a passenger is charged on the basis of the areas or zones in which his trip originates and ends. The boundaries of the fixed zone system should be set to divide a city into areas that take about equal amounts of time to pass through. Thus it would be expected that zones would be somewhat smaller in the more congested central business district than in the outlying residential areas.

The fixed zone system is low cost in its administration since all that is required to determine a fare is a zone map of the area. But its other disadvantages may outweigh this advantage. Zone boundaries encourage drivers to look for trips that cross these lines, while encouraging customers to walk or take a bus across zones. At the same time, drivers will not wish to make trips within a single zone unless they are very short, and customers will wish to use cabs for one-zone trips and for trips that go through very long zones. Incentives that result in drivers looking for one kind of trip and passengers looking for another are not desirable since driver unhappiness about the fare structure may result in poor service and customers feeling that they are not getting what they pay for.

Thus the disadvantages of the fixed zone system are that it does not follow cost very well and it encourages the taking of uneconomic trips while discouraging the taking of certain trips that would be made if rates were cost-based. There should be no difficulty in recouping the revenue requirement under this system if the fare level is set high enough. It is a difficult system to understand because there are many zones and zone boundaries that are necessary in a large area such as Washington, D.C. It is not hard to understand why it has been dropped in so many areas.

A very different kind of zone system, the circular zone system, was proposed by a Washington, D.C. taxi driver, Irving Schlaifer, some 15 years ago. The taxi fare under this system is

based on the number of airline miles between the origin of the trip and the termination point. It is currently the basis for determining interstate fares by District of Columbia taxis and is the basis used to determine parcel post rates and long-distance telephone rates. The rationale for this system is that it relies on distance, it does not require a meter for implementation, it is easy to understand, and it encourages the driver to take the lowest cost route from origin to destination.

To be more specific, a circular zone system follows cost better than a fixed zone system because the rate is more closely based on the distance between origins and destinations. A disadvantage here is that taxicab costs are incurred on the basis of miles actually traveled and not airline miles. This disadvantage may be more than offset, however, by the incentive that drivers are given to take the lowest cost path. Passengers will know that rates are based on mileage and react accordingly. They will probably try to take more trips that are just short of the break points in the system, perhaps even at the expense of doing a certain amount of walking. But, if there is a small enough distance increment in the system, such as the half-mile break proposed by Schlaifer, this might not be a big problem. The circular zone system does not require meters and is therefore low cost to administer. Finally, once implemented, it should be very easy for customers to understand. Under this method, customers will know from the map in the taxi exactly how much their fares will be before they start their trips.

The meter system, when carefully administered and patrolled, probably follows costs the best. This happens because fares are based on miles actually driven and the time actually spent in providing the service. There is a circular zone element in this system in that the meter registers additional amounts on a discrete rather than a continuous basis, such as every fifth of a mile.

Meters are more costly than zone systems to administer. They cost about $300 and require maintenance, which may be timeconsuming. They can be tampered with and do not encourage drivers to take the lowest cost route. If a driver doesn't know where he is going, it will cost the consumer. Drivers will prefer longer and faster routes while passengers will prefer shorter routes. Areas with poorer roads and greater traffic congestion will not be frequented by drivers unless there is a high enough waiting time charge. Customers will easily understand the basis upon which their fares will be determined but will not know what their fares will be until their trips are over.

While the different fare determination systems discussed above produce different incentives to passengers and drivers based on each wanting to get the best bargain for himself, there is no reason to expect that any of the above systems would be more likely than any other to affect ownership patterns or levels of compensation. Meters are a more costly method of fare determination and this would probably be reflected in terms of a slightly higher cost level for meter cabs over zone cabs. Rate levels appear to be independent of fare structure and depend more on entry policies and regulatory control of revenue requirements than anything else. This question will be discussed in more detail when the comparable cities study is discussed.

The method of fare determination is not likely by itself to affect overall demand and service levels, but will very likely affect demand and service levels for certain trips and in certain areas of the city. What must be recognized at the outset is that fare level and fare determination are problems that are largely independent of one another. The fare level can be high or low with