Academic journal article Economic Inquiry

Efficiency, Growth, and Concentration: An Empirical Analysis of Hospital Markets

Academic journal article Economic Inquiry

Efficiency, Growth, and Concentration: An Empirical Analysis of Hospital Markets

Article excerpt


Taking an evolutionary view, Harold Demsetz hypothesized that firms. differ persistently in efficiency and that industry concentration results from growth of efficient firms at the expense of inefficient ones. We test this idea with microdata from the hospital industry. Initial hospital efficiency and subsequent growth (and profitability) are significantly and positively related. Also, greater initial variation in hospital efficiency within local markets is positively related to subsequent growth in market concentration. Our findings support the evolutionary efficiency hypothesis, though they cannot confirm the stronger idea that variation in efficiency is the dominant explanation for changes in concentration. (JEL L11, L84, I11, L31, L20)


In 1973, Harold Demsetz provided aggregate, cross-industry statistical support for his hypothesis that industry concentration is endogenous, largely the result of growth of relatively efficient firms. In this article, we provide the first known test of this hypothesis using microdata from a single industry: the hospital industry. The hospital industry is a good industry to study for both scientific and policy reasons.

Recently, hospital mergers have received much attention, and efficiencies have been claimed for them. But the courts have remained skeptical, perhaps due to contradictory findings regarding hospital scale economies (Frech and Mobley [1995]; Lynk [1995]). The Demsetz hypothesis does seem to explain cross-sectional results from older studies--that costs are lower in more concentrated hospital markets. [1] We apply the hypothesis to the hospital industry, using California data, 1983/84-1990/91. The first part of the article places the Demsetz hypothesis in context; the second part is an empirical test. Using this excellent data, we estimate firm-specific efficiency early in the sample period and relate it to subsequent growth, the persistence of profits, and change in market concentration.

We use several kinds of statistical analysis, including cross-tabulations (closely following Demsetz), and two different methods of efficiency assessment. One method uses a deterministic frontier, while the second uses a newer stochastic frontier technique. We employ different output measures and geographic market definitions as sensitivity tests.


The structure-conduct-performance paradigm dominated industrial organization in the early 1970s and was the subject of many empirical investigations (Weiss [1974]). It largely ignored efficiency explanations for concentration, perhaps because existing theoretical literature assumed homogeneous firms and the existing empirical literature concluded that minimum efficient scale is generally small (McGee [1988]).

Influential dissenters began to be heard in the mid-1960s (Bork and Bowman [1965]; Bork [1967]; McGee [1971]). Among them was Demsetz [1973], who argued that concentration is largely endogenous, and results from more efficient firms growing faster. Contrary to the older tradition, Demsetz stressed persistent heterogeneity among firms. His analysis is similar in spirit to the survival analysis of Stigler [1958] and to the evolutionary models of Nelson and Winter [1982].

Demsetz [1973] conducted an indirect, cross-industry test to distinguish efficiency from market power effects. He reasoned that if tacit or explicit collusion caused high rates of return in concentrated industries, it would benefit all firms therein--implying a positive correlation between the rate of return and industry concentration across all size groups. In a 1963 sample of firms from 95 industries, no such correlation was found for small firms, while the largest firms exhibited higher rates of return, more so in the most concentrated industries. He concluded that the overall correlation between concentration and profits must be caused by superior efficiency in larger firms. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.