Why Growth Rates? Which Growth Rate? Specification and Measurement Issues in Estimating Elasticity Values

Abstract

In theoretical settings, growth rates are generally based on derivatives, which relate infinitesimal changes in time to changes in a variable's value. In applications of theory, growth occurs over discrete intervals, and a variety of discrete growth measures exist. Therefore, a key question is which discrete measure provides estimates most consistent with true parameter values. In estimating elasticity values, results show that the standard growth rate can be highly inaccurate. The average-base and logarithmic growth rates do a much better job. In addition, when the process generating data is unknown, a linear in growth rates regression equation is found to be more accurate than the widely used linear, logarithmic, logarithmic-linear, and linear-logarithmic specifications.