Economies of scale in the US computer industry: An empirical investigation using data envelopment analysis

Up to recently, economists have had no good tools to measure the returns to scale of individual corporations in an industry. Data envelopment analysis (DEA) is a linear programming technique for determining the efficiency frontier (the envelope) to the inputs and outputs of a collection of individual corporations or other productive units. While DEA offers an avenue for calculating the returns to scale of individual corporations, the approach has been riddled by mathematical complications arising from the possibility of alternate optima. The present paper develops theory for calculating the entire range of these alternate optima. Furthermore, in a quite ambitions empirical application, DEA is employed to determine the time path of returns to scale of all publicly held U.S. computer companies over the time period 1980–1991. For the great majority of companies, a unique time path is obtained; only in less than 4 percent of the linear programming calculations is an entire range of alternate optima obtained. The results indicate that the computer industry was polarized into two camps: large aging corporations with decreasing returns to scale, and swarms of small upstart companies with advanced technology exhibiting increasing returns to scale.     

Chance-constrained data envelopment analysis

Data envelopment analysis (DEA) is extended to the case of stochastic inputs and outputs through the use of chance-constrained programming. The chance-constrained frontier envelops a given set of observations ‘most of the time’. As an empirical illustration, we re-examine the pioneering 1981 study of Program Follow Through by Charnes, Cooper and Rhodes.     

Choosing Between Different Time‐Varying Volatility Models for Structural Vector Autoregressive Analysis

The performance of information criteria and tests for residual heteroscedasticity for choosing between different models for time‐varying volatility in the context of structural vector autoregressive analysis is investigated. Although it can be difficult to find the true volatility model with the selection criteria, using them is recommended because they can reduce the mean squared error of impulse response estimates substantially relative to a model that is chosen arbitrarily based on the personal preferences of a researcher. Heteroscedasticity tests are found to be useful tools for deciding whether time‐varying volatility is present but do not discriminate well between different types of volatility changes. The selection methods are illustrated by specifying a model for the global market for crude oil.     

Fundamental analysis of stocks by two‐stage DEA

Fundamental analysis of stocks links financial data to firm value in two consecutive steps: a predictive information link tying current financial data to future earnings, and a valuation link tying future earnings to firm value. At each step, a large number of causal factors have to be factored into the evaluation. To effect these calculations, we propose a new two‐stage multi‐criteria procedure, drawing on the techniques of data envelopment analysis. At each stage, a piecewise linear efficiency frontier is fitted to the observed data. The procedure is illustrated by a numerical example, analyzing some 30 stocks in the Spanish manufacturing industry in the years 1991–1996. Copyright © 2004 John Wiley & Sons, Ltd.     

Using data envelopment analysis to quantify ILO objectives and identify policies conducive to decent work in a globalizing world

The International Labor Office, an arm of the UN based in Geneva, has as its goal the promotion of opportunities for women and men to obtain decent and productive work, in conditions of freedom, equity, security and human dignity. Since 1999, the ILO has conducted a series of studies of decent work. In 2001, the organization posed the global challenge of reducing the decent work deficit as measured by an employment gap, rights gap, a social protection gap, and a social dialogue gap. Using standard economic terms, “decent work” may be seen as an efficiency point along a generalized input-output function, dependent upon variables of both economic performance and economic and social policy. The decent work deficit of a given country (if any) is then obtained as the difference between an observed point and its projection on the efficiency frontier. Using data envelopment analysis (DEA), we fit a piecewise linear frontier to observations for 61 countries from all continents. Importantly, 27 of these countries lie on the decent work frontier; the remaining ones reveal conditions of decent work deficit. The possibilities of reducing such deficits by appropriate control of policy variables are discussed.     

DEA of financial statements data: The U.S. computer industry

DEA (data envelopment analysis) is a technique for determining the efficiencyfrontier (the envelope) to the inputs and outputs of a collection of individual corporations or other productive units. DEA is here employed to estimate the intertemporal productive efficiency of U.S. computer manufactures, using financial data brought from earnings statements and balance sheets. The results indicate that a few corporations, including Apple Computer Inc., Compaq Computer Corp., and Seagate Technology were able to stay at the productivity efficiency frontier throughout the time period investigated. But not all successful corporations did; sometimes subefficiency (=disequilibrium) actually goes together with very rapid growth. A new Malmquist type productivity index is calculated for each corporation, measuring shifts of the estimated intertemporal efficiency frontier.