In forming economic development policies the assessment of sectoral economic performance and production interdependence are both very important issues. While, sectoral interdependence is one of the most important source of economic expansion in a competitive economy, efficiency is the most important control parameter for assessing the utilization of inputs in the production process. A large and strongly interdependent sector may be seen as a good candidate for the economic development of a particular country or region. The expansion of this sector will have a significant impact, increasing output, income or employment domestically.
However, and equally important, a sector that operates more efficiently (i.e. producing as much output as the inputs at its disposal permit under the current state of technology) compared to the same sector in other countries or regions could be a better candidate for long term growth and development. That’s because efficiency is one of the main factors determining the overall competitiveness of a sector. The higher the degree of efficiency, the lower will be the unit cost of production and as a result, industries will be able to supply their products at lower prices. Consequently, more efficient industries would have better chances of surviving and prospering in the future than less efficient ones. Furthermore, improvements in sectoral efficiency can increase productivity and thus output growth within the national or regional economy, thus providing a more cost-effective way for stimulating internal economic growth.
In the short-run, an economic sector may be large enough to attract policy attention but if it is not operating efficiently compared to the same sector in other nations or regions, sooner or later it will become less competitive and begin to diminish from external competitive pressures. It is evident therefore, that it is important to know how much a given sector can stimulate economic growth within the economy without resource waste, thereby improving its competitive status. In a long-term competitive environment the overall efficiency and thus productivity of a national or regional economy determines the general well-being of its people (Krugman, 1991).
Traditionally, input-output analysis and the subsequent measurement of linkage coefficients has been used excessively for the identification of key economic sectors in both national and regional economies. Since the pioneering work of Rasmussen (1956), Chenery and Watanabe (1958) and Hirschman (1958), a number of studies employing input-output techniques have relied on linkage analysis to describe the interdependent relationships between economic sectors and to assist in the formulation of economic development strategies. The economic rationale behind the empirical significance of linkage analysis is rather simple: the expansion of sectors with strong linkages and thus production interdependencies within the economy is likely to promote overall economic development.
Through the years the methodological framework has been improved and expanded in several ways, and many different analytical methods have been proposed for the measurement of interindustry linkage coefficients (i.e. Jones, 1976; Cella, 1984; Heimler, 1991; Sonis et al., 1995; Dietzenbacher and Van der Linden, 1997). Although it seems that the problem of the appropriate linkage indicator has been resolved, there still remains an important practical question: would the expansion of the “key” sectors identified by means of linkage coefficients would indeed promote economic development?
The aim of this paper is to explore this unknown relationship between sectoral interdependence and technical efficiency. The empirical analysis is based on the 1996 US input-output tables published by the Bureau of Economic Analysis. Both backward and forward linkage coefficients were computed using the approach suggested recently by Dietzenbacher and Van der Linden (1997), which is an adapted form of Strassert’s (1968) hypothetical extraction method. Measurement of sectoral technical efficiency is based on Shephard’s (1953) input-distance function which can easily accommodate multi-output technologies under the assumption of weak disposability. Finally, the stochastic frontier is modeled according to Battese and Coelli’s (1995) inefficiency effects model.