Selection and Firm Survival: Evidence from the Shipbuilding Industry, 1825-1914

Several theories of firm performance can explain the well known observation that survival is positively related to age. However, a more mundane explanation-selection bias driven by variations in firm quality-may also underlie the phenomenon. This paper employs a 90 year plant-level panel data set on the U.S. iron and steel shipbuilding industry of the nineteenth and early twentieth centuries to discriminate between the explanations. The shipbuilding industry exhibits the usual joint dependence of survival on age and size, but this dependence is eliminated after controlling for heterogeneity by using preentry experience as a proxy for firm quality. The evidence points to a dominant role for selection bias in creating the age dependence of survival. At the same time, preentry experience is found to have a large and extremely persistent effect on survival, and this finding is inconsistent with standard explanations for the role of preentry experience on firm performance. © 2005 President and Fellows of Harvard College and the Massachusetts Institute of Technology.