Non-neoclassical microeconomics: A nominal total bill approach to residential telephone usage demand estimation

A model for forecasting consumer demand for compositely priced goods, with a special focus on telephone usage consumer price response, is developed. This nominal total bill (NTB) model hypothesizes that consumers respond to the total bill sticker price of compositely priced commodities and do not attempt to rationally maximize their utility over the commodity's constituent real prices. The NTB model is applied to the estimation of residential telephone usage demand. The existing literature on residential telephone usage demand is critically surveyed, and related, economic, market research and consumer psychology studies which support this hypothesis are discussed. A formal mathematical comparison of the NTB and the traditional Neoclassical models reveals that the two models are logically incompatible. Household level telephone usage data from a two year fixed panel of about 360 Minnesota AT&T residential customers are used to derive empirical estimates for the two models. Unconstrained household level estimates from the NTB model were compatible with the NTB model's underlying hypotheses about consumer behavior and were within the range of aggregate telephone price elasticity estimates from previous studies. Unconstrained household level estimates from the traditional model were inconsistent with basic Neoclassical consumer behavior hypotheses and were incompatible with previous aggregate estimates. On this basis, the traditional Neoclassical model was rejected in favor of the NTB model. Acceptance of the NTB model implies that: (a) inflation should be a separate independent variable in single sector demand modeling, and (b) the effective price for compositely priced commodities is the customer level nominal total bill price, so that the effective behavioral invariants for the estimation of price response for these goods are average customer level nominal composite price elasticities. NTB model estimates using the Minnesota household level data indicate that time-of-day telephone usage nominal price elasticities are highly variable across different tariff changes and overall (not time-of-day) elasticities are stable across tariff changes.