Sung-Ko Li

Measuring the Research Performance of Chinese Higher Education Institutions: An Application of Data Envelopment Analysis

Utilizing data from 84 key Chinese higher education institutions, the present study attempts to examine the effectiveness of the Education Reform implemented in the mid-1980s in China. With focus on the research performance of the institutions, individual institution efficiency is computed by the method of data envelopment analysis. Regional differences in the efficiency of institutions are also addressed. It is found that research performance of institutions across regions has improved, although the institutions as a whole have remained inefficient from 1993 to 1995. Institutions located in the East region turn out to have out-performed those in the Central and the West regions. In addition, the decomposition of the group efficiency measure indicates that, for the 3 years under study, the 84 key institutions suffered from technical, allocative and reallocative inefficiency.

Measuring the productive efficiency of a group of firms

While the conventional Farrell-Färe approach to efficiency measurement can identify the most inefficient firms, it fails to consider the efficiency of a group of firms thoroughly. This paper introduces efficiency measures that can be used to find the efficiency of a group of firms and pinpoint whether the group inefficiency is due to inefficiency inside or outside individual firms. Furthermore, a new way of finding the revenue maximum shadow price vector is introduced to compute the allocative efficiency of individual firms when price data are not available.

Solving the puzzles of structural efficiency

Abstract Since the concept of “structural efficiency” first appeared in Farrell [Farrell, M.J., 1957. The measurement of productive efficiency. Journal of the Royal Statistical Society, Series A, Part III 120, 253–281], attempts have been made to derive measures for the performance of a group of production units (often referred to an industry with many firms). Many empirical studies used the technical efficiency of an average unit to measure the structural efficiency of a group, but researchers have been puzzled by the discrepancies between the average of individual efficiency scores and the performance of the group as a whole. In this paper, we point out that the “shadow price model” provides a useful framework for understanding the economic meaning of the structural efficiency as well as its components. By recognizing these components, the puzzles related to the inconsistencies between the individual and group performance can be solved readily.

Relations between convexity and homogeneity in multioutput technologies

Abstract Homogeneity and convexity are two important concepts in economics. With homogeneity, this paper shows that the technology set is convex if and only if the input sets are convex. Furthermore, even if individual technology sets are not convex, the aggregate technology can be made convex by considering at most two firms only.

Aggregating Capacity-Limiting Separable Inputs

The concept of separability is closely related to aggregation. Previous studies on separability are important mainly for theoretical contributions, not for empirical results. This paper provides a theoretical justification for using the maximum capacity of a subset of inputs as an aggregate index in empirical settings. The method is demonstrated with two examples, aggregating inputs within a firm and across firms.

Inner and outer approximations of technology: a data envelopment analysis approach

This paper discusses inner and outer approximation to a technology given by a data set of prices and quantities.

A nonparametric test of translation homotheticity

Abstract Recently Chambers and Fare (1998) [Chambers, R.G., Fare, R., 1998. Translation homotheticity. Economic Theory 11, 629–641] generalized the Shephard type of “multiplicative” homotheticity into an “additive” type, which they termed translation homotheticity. This type of homotheticity has parallel expansion paths and its cost function is additively separable. In this letter a nonparametric test for translation homotheticity is developed.