The distribution-free approach (DFA) to estimating X-efficiency assumes that individual firms exhibit constant inefficiency across time, and that inefficiency can be revealed by estimating a panel cost (or profit or production) function and averaging together the annual residuals for individual firms. However, the existing DFA literature may not pay enough attention to the consequences of including too many, or too few, time series observations in the data panel. This paper derives a test of whether additional annual data improve or worsen estimated X-efficiency, and demonstrates the test using a DFA cost efficiency model and US commercial bank data from 1984 through 1994.
