Abstract
In many well known studies, unit values are often used in place of observed prices when constructing price indexes or modeling demand systems. This study shows that one cannot get consistent estimates by replacing prices with their unit values. When computing price indexes, the use of unit values or unit value indexes will misspecify substitution effects and mismeasure the true price. The use of unit values as a price regressor in an aggregate demand model will misspecify the model even though the functional form of the demand model is correct. I then present empirical investigations using cereal sales data to evaluate a well known paper that estimated the welfare gain of a new cereal brand. I show that the use of unit values in this study biases the estimate of the welfare gain of the new brand. JEL Codes: C43, C51. Keywords: Unit value, cost of living index, aggregation consistency.
Get full access to this article
View all access options for this article.
