5 a substantial fixed cost and immediately pour flights into the most profitable routes. However, once that were done it would have created much of its infrastructure and would find it profitable to keep adding flights on an incremental basis in a manner very different than the way the model used in the ATA study does in its phase 1.* It would add more until it came to a step in the fixed costs and then hesitate until traffic pressure built up. A good many of the "more" would, it seems to me, be on the routes the ATA suggests are threatened, albeit mostly between cities each of which already had a station (and thus many fixed costs). If one were simulating competitive practice it would mean that, in the above example, the stations would need to be of the same carrier (although interchange flights could be used as well as other techniques). The most important thing however is that in a competitive environment everyone's fixed cost steps are not identical in time and place. One carrier may be reluctant to add a flight of a certain kind but another carrier which has just bought the right aircraft type may well be anxious to do so. In sum, I believe the treatment of costs in the first phase of the model used (the fully allocated cost phase) greatly distorts the models process of adding flights, and does so because of the model's demand/ ++ frequency assumption, in the direction of concentrating flights in the *i.e., as shown in Figure 2 page 4 of the ATA study and expanded upon in the Lockheed brochure CTR 2007, Lockheed Airline System Simulation, revised May 1970, charts 884-09229 through 886-09229. tibid., Lockheed chart 889-05180. 十七 One may ask why this was done. I believe there are a number of reasons major markets.* 6 Revenues Revenues are less clear. Individual segment-by-segment coach fares were used. They were then reduced by a constant percentage to reflect recent system-wide average discount experience. Thus short haul flights and lightly travelled segments were discounted to the same degree as long haul and dense segments. I do not have data immediately at hand from which to analyze discounts including the effect of the fare pro-rate on joint fare routings by segment type; however, a few observations are obvious. 1. By Trunk Carrier, in 1969, there is a recognizable 2. Many of the present day (1975) discount fares do not but in this case it is most probably a combination of the background of the modelers (technical more than economic) and the state of the art when the model was begun (ca. 1959). For further discussion of the problems associated with models not doing what decision makers want models to do, see my book: The Administration of Transportation Modeling Projects, D. C. Heath, Lexington, Mass., 1973. *I do not wish to suggest that I propose the modifications I have suggested as all that is needed to perfect the model under discussion. I recognize full well, for example, that if my modifications were made one could get vastly different answers if you ran the model from zero frequencies upwards and from many frequencies downwards. It would be more realistic, however, as this is exactly what happens in real life. In truth, I favor a mathematical programming approach using origin and destination data rather than segment flow data. Such a model would break the perpetuation of the present route structure just because it is the present route structure. Such a mathematical programming approach is now just within the state of the art, I believe. See the discussion later in this memo. JOHN W. DRAKE • TRANSPORTATION CONSULTANT 7 Thus, given a choice, I would have to guess that the ATA underestimated the "endangered" segment actual gields. The prorate situation is more of an unknown, however, so to be certain one must have the real data. Aircraft Types Clearly another way in which the ATA study suggested such a large number of endangered segments was by its choice of equipment. The list on page 3 is long enough but not diverse enough: Clearly not all need to be included--perhaps only the asterisked ones. Running time would go up slightly but I dare say some endangered segments would suddenly survive, which leads "Voilà!" to the third level carriers and their willingness to step in. To be sure, as the ATA says, Lafayette, Indiana, (where I write this)- Chicago service by Allegheny is JOHN W. DRAKE • TRANSPORTATION CONSULTANT 8 endangered. It's so endangered it's been gone for well over a year. But we have eight (weekday) non-stops to Chicago on fast pressurized Air Wisconsin planes. Who is suffering? Interstate Highways One must also not lose sight of another factor: the interstate highway system. I will use Indiana as an example since I now reside there and travel extensively. I reproduce the map from the ATA study and the same map with the interstate highways added. Having now spent something approaching $50,000,000,000 to build a vast system of interstate highways, is it any wonder that Indianapolis - Fort Wayne service might be threatened, and so should a lot of the others? We should not cry if we sometimes succeed in transportation! This does point to the most important of all aspects of this whole deregulation questions, however, which is simply that of "What is our national transportation policy?" As of now we don't have one. We need one. Otherwise we are in the silly situation of crying over things that either won't happen or that we've spent billions trying to get to happen. With no integrated intermodal transportation policy, we don't know when we are well off! Model Methodology One must still address the question of whether the model used is fundamentally appropriate to the problem. That is, if it were merely improved and had everybody's nit-picking objections taken care of and computer time were no object, could it do the desired job in this *Or in the case of the trains, happened when we didn't want them to, because of disjointed policies toward air, rail, water and highway transportation. JOHN W. DRAKE • TRANSPORTATION CONSULTANT 51-146 O - 76 pt. 1 28 |