Please use this identifier to cite or link to this item:
|uhm_phd_8003290_uh.pdf||For UH users only||4.22 MB||Adobe PDF||View/Open|
|uhm_phd_8003290_r.pdf||Restricted for viewing only||4.26 MB||Adobe PDF||View/Open|
|Title:||Transportation demand and catastrophe theory : a comparative analysis of disaggregated choice models|
|Authors:||Schlottfeldt, Carlos Bicalho|
|Keywords:||Freight and freightage -- Mathematical models|
|Abstract:||Previous transportation demand studies reveal that, in addition to freight rates, service and commodity characteristics have a significant influence in modal choice and quantity shipped. Most demand models, however, incorporate these characteristics following behavioral and/or quasi-utility arguments rather than a structural form derived from economic theory. One exception is the inventory-theoretic type model which describes transportation demand in a structural manner according to basic tenets of micro-economic and inventory theories. In this dissertation basic theorems of catastrophe theory were used to re-interpret the transportation problem and develop a generic model which is simpler, but qualitatively equivalent, to the inventory-theoretic model. The resulting model considers simultaneously modal choice and demand, and may be specified for one choice variable (annual demand, order quantity, probability of modal choice, etc.) as determined by any two causal factors in the objective function. Two versions of this model were prepared: 1) a derived demand model relating quantity shipped by individual modes to cost and revenue parameters; and 2) a choice model, relating the probability of specific choice to direct cost of transportation and mean transit time. This second model was compared with the legit model showing that: 1) The legit model would be likely to misclassify some of the observations; and, 2) Misclassification of a significant number of observations is likely to produce inconsistencies in estimated coefficients. even when the model exhibits a strong (apparent) overall explanatory power. Eventual inconsistencies should be more evident in connection with non-linear variables which affect inventory policy of the shipper t notably transit time. A subsequent analysis about the effects of product characteristics on demand revealed that perishable products should provide better logit estimations than more durable products. and that a high product value reinforces this effect. A review of published results from previous empirical studies largely confirmed this implication. All estimations involving durable goods provided inconsistent coefficients (insignificant and/or wrong sign) for the variable transit time. The only perishable product under consideration produced highly consistent coefficients. and one study presented inconclusive results due to non-specification of this variable. Major general conclusions of this study include: 1) The transportation problem may be described through catastrophe theory. and this approach 'appears to promise a good potential for future modeling efforts in the area. Its main advantages include an ability to describe simultaneously modal choice and demand, and its ability to represent simplified versions of more complex structural models. 2) Inventory considerations do play an important role in transportation decisions and must be considered in empirical investigations. In particular, linear logit estimations involving non-perishable products are likely to produce unreliable coefficients, which must be used with utmost caution in policy recommendations. Suggestions for improving future demand estimations included pre-classification and/or pre-treatment of observations prior to their processing through the logit model, modification of this model's formulation, and direct estimation of demand by using the catastrophe theory/inventory-theoretic approach.|
|Description:||Photocopy of typescript.|
Thesis (Ph. D.)--University of Hawaii at Manoa, 1979.
Bibliography: leaves 142-147.
xiii, 147 leaves ill. 29 cm
|Rights:||All UHM dissertations and theses are protected by copyright. They may be viewed from this source for any purpose, but reproduction or distribution in any format is prohibited without written permission from the copyright owner.|
|Appears in Collections:||Ph.D. - Agricultural Economics|
Items in ScholarSpace are protected by copyright, with all rights reserved, unless otherwise indicated.