DEA as a tool for auditing: application to Chinese manufacturing industry with parallel network structures

Yande Gong,Joe Zhu,Ya Chen,Wade D. Cook

DOI: https://doi.org/10.1007/s10479-016-2197-1

IF: 4.82

2016-04-21

Annals of Operations Research

Abstract:Performing a high-quality manufacturing audit can be time consuming and costly given the large number and scale of manufacturing firms and enterprises. Chinese economy has experienced some rapid and significant growth over the past 30 years, which is largely due to contributions from the development of manufacturing industry. Auditing tools are very much needed in auditing the Chinese manufacturing industry. Data envelopment analysis (DEA) has been used as an auditing tool in selecting audit objects that are treated as decision making units (DMUs). These DMUs are characterized by a set of performance measures. DEA then uses data on these performance measures to identify potential audit objects. However, conventional DEA treat each DMU or system as a “black-box” and does not take the operations of individual components within the “black-box” into consideration. For example, a large number of firms or enterprises exist in a manufacturing industry. When the conventional DEA is applied to the industries, the performance of the firms is often ignored. This paper proposes a new parallel DEA model where each input/output of the system is not the sum of those of all its components. Such a situation arises from the need of auditing firms or enterprises in Chinese manufacturing industries. For example, the cost margin of a particular industry may not equal to the sum of that of all its component firms within the industry because this metric is measured in percentage. The proposed approach is applied to the manufacturing industry of China.

operations research & management science