Using the Survivor Technique to Estimate Returns to Scale and Optimum Firm Size
Abstract The survivor technique for estimating returns to scale and optimum firm size has generated a slow but steady literature since its 1958 pilot presentation by George Stigler. This article (1) integrates advances in its application into a complete demonstration of how the technique works, (2) distinguishes a survivor analysis from the related but different analyses of individual firm growth and size distribution as addressed, for example, by Gibrat's Law of Proportionate Effect, (3) surveys a few exemplary survivor analyses, highlighting their alternative measures of scale and survival, and (4) unifies the scattered discussion of criticisms and qualifications that surround the technique. Accordingly, this essay seeks to reposition the survivor technique as a viable statistical option for research on those industries which meet its criteria.