Working Paper: CEPR ID: DP2562
Authors: Eugenio J. Miravete
Abstract: It is commonly believed that consumers behave irrationally when subscribing optional telephone tariffs. The fact that they show a strong preference for flat rate options has commonly been interpreted as evidence of irrational behavior since such a choice is believed not to be cost-minimizing ex post in most cases. My results, obtained using the data from the 1986 Kentucky tariff experiment, contradict these views and provide strong evidence in favour of the rationality of consumers' choices. I found that expectations on future consumption play a major role in the choice of tariffs but also that consumption forecast errors are more related to the volume of local telephone usage than to any particular demographic profile. More importantly, the evidence shows that there exist important learning effects that induce tariff switching in order to minimize the magnitude of monthly bills even in the short term and responding to very small cost differences. Finally, risk aversion is ruled out as a possible source of consumers' biased taste for flat tariffs.
Keywords: expectation bias; learning; risk aversion; service switching; tariff choice
JEL Codes: D42; D82; L96
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Consumers' expectations of future telephone usage (L96) | Choice of tariff plans (L90) |
| Learning effects (C92) | Tariff switching behavior (F16) |
| Demographic characteristics (J21) | Tariff choice (F14) |
| Consumption forecast errors (E27) | Tariff switching behavior (F16) |
| Tariff choice (F14) | Prediction errors increase with telephone usage (L96) |
| Risk aversion (D81) | Preference for flat tariffs (H29) |