sight. That would imply that the major structural changes that have occurred in this
industry during the last four decades are likely to continue unabated. In addition to
seeking answers to these three growth and diversification questions, we examine
incumbent firms and new entrants separately.
Method of Analysis
Typically, cost economies are analytically derived from either primal or dual
econometric approaches. We approach the topic in a nonparametric way. Rather than
econometric modeling, we track farms in 10 initial size cohorts through three successive
censuses, determine differences in growth rates, level of diversification, and industry exit
rates. We also track new entrants to determine similarity to incumbent firms. While our
findings do not provide explanations about the causes of structural change, they do
contribute essential missing links in understanding how structural change is being
implemented at the firm level. They also create an informational base that can help focus
subsequent econometric analysis of causal factors.
The first question is addressed by examining the relationship between initial cohort
size and mean growth rate of each incumbent cohort. This relationship will provide
inferential evidence concerning whether an equilibrium firm size exists to which firms
are converging, and if it does exist, whether it is stable. Cohorts that are growing the
most rapidly are likely operating under increasing returns to scale and/or scope.
We also test two hypotheses relevant to the first question from the dynamic firm
growth literature: Gibrat’s law and mean reversion. Under Gibrat’s law, firms are
hypothesized to face the same distribution of possible growth rates independent of their
size. If they do, they follow a random walk growth pattern. No convergence to steady