A Microeconometric Analysis of Consumer Response to Direct Marketing and Mail Order

Author: Kalsow, Gretchen Anne

Year: 1997

Degree: Dissertation (Ph.D.)

Advisors: Dubin, Jeffrey A.; Davis, Lance E.

Committee Members: Dubin, Jeffrey A.; Grether, David M.; Kiewiet, D. Roderick; Page, Scott E.

Option: Social Science

DOI: 10.7907/1x9m-x446

Abstract

This dissertation builds a series of micro-econometric models of consumer response to direct marketing and mail order. It employs a unique dataset from one of the largest mail order firms in the United States. Prior to building the models a brief history of the mail order industry is included to alert the reader to level and type of information that has been amassed over the last century. The next chapter compares the linear probability models commonly in use in the direct marketing industry to logit discrete choice models. Following that analysis is an implementation of the Dubin and McFadden [11] discrete/continuous model. It supports the hypothesis that the decision to purchase and the purchase amount are correlated, which indicates that the independent regressions for the propensity to purchase and the purchase amount currently being used by direct marketing firms could be improved upon.

The following chapters build and estimate three discrete time/discrete choice discrete/continuous models. The first model is one in which each time period is independent of the other time periods. The second model assumes that each individual has a reservation value, or pain threshold, for ordering products. The last model is a nested logit model in which future periods, i.e., consumer’s expectations, are considered when making their purchase decision.

An appendix describing the data and the process for building the dataset is included at the end. In addition, a second appendix containing an outline of potential future research closes out the dissertation. The third appendix contains a proof of one of the propositions.

Files