Returns to Scale and Input Substitution for Large U. S. Banks

Abstract

THE UNITED STATES COMMERCIAL BANKING INDUSTRY has experienced significant changes in recent years. The recent wave of mergers and the near-term prospects of interstate banking suggests that the industry is moving toward much larger average banking organizations (Rhoades 1985). Will a more concentrated banking industry result in lower costs due to scale economies? We address this question by examining the cost function for large banks. Numerous studies of economies of scale in banking exist, generally tracing their roots to the classic papers by Benston ( 1965) and Bell and Murphy ( 1968). A good review of the most recent studies can be found in Clark ( 1988). The majority of the papers uses samples of smaller banks (that is, under $1 billion in deposits), finding generally that economies of scale are quickly exhausted and that further increases in bank size occur at constant, or possibly increasing, cost. The extrapolation of previous findings for samples of smaller banks to large banks may be problematic, especially in the current environment. First, large banks service a different mix of markets and customers than small banks, suggesting that large and small banks probably face different cost structures. Second, the process of deregulation may have rendered the results of prior studies moot, since the data pertain to periods before significant deregulation. Since we