Efficiency evaluation of China’s power industry: A data-driven approach by opening two “black boxes”

The network data envelopment analysis (DEA) method was developed to open the “black box” of complex system efficiency evaluation. Also, the by-production method, a DEA method for modeling undesirable output, was developed to open the “black box” of the pollution generation process. This study combines these two approaches and improves them to open two black boxes in China’s power industry. We find that although the by-production method has been developed theoretically, several application issues still exist, such as the fact that it cannot deal with network structure decision-making units, and it ignores the inherent leader-follower relationships between substages. In addition, current models cannot offer improvement benchmarks for inputs, which are crucial for identifying the overutilization of resources. We will discuss these problems in detail and offer solutions. Valuable recommendations for the development of the Chinese power industry are provided.