Consumer preference change and the generalized axiom of revealed preference

This thesis assesses the efficacy of using revealed preference axioms to detect changes in preferences. The generalized axiom of revealed preference is applied to simulated datasets that are derived from consumer optimization of a Cobb-Douglas utility function. Parameters of the utility function change while generating the data and so the resultant datasets reflect a structural change in consumer preferences. Results show that in most scenarios examined, the datasets rarely violated revealed preference axioms even though the true process that generated the demand observations involved unstable preferences. Violations of revealed preference axioms were most frequent when the number of observations was large and when the number of products in the choice set was small or when the products that were affected by the preference change accounted for a large share of the consumer's budget. Revealed preference violations also depended on the nature of the preference shift. Abrupt and permanent shifts to preferences were most likely to lead to violations of the axioms.