#96-04
"Process Design and Efficiency: Evidence from Retail Banking"
Summary: Traditional efficiency studies measure the performance of a firm or decisionmaking unit (DMU) by its ability to transform inputs to outputs. The actual way in which these inputs are transformed to outputs is often overlooked. Each DMU's operation is conceptualized as a black box: inputs go in, outputs come out, and little analytical attention is paid to the inner workings of the transformation process. This paper examines the black box and argues that the actual design of the transformation process is a critical component in the performance of the DMU. The design of the transformation mechanism, which the authors refer to as the process design, must be fully studied and integrated int he performance analysis in order to provide useful managerial recommendations.
The authors use a variation on frontier estimation (DEA-like) techniques, to address the question of how much inefficiency is due to the wrong process design for the desired output maximization, and how much is due to poor execution of the correct design. They demonstrate the approach using a service delivery process in the retail banking industry.
Holding the effects of scale and scope effects constant, the study focuses on why x-efficiency varies among financial institutions. Their goal is to understand how technology, human resources, and process management methods vary across organizations, and how this variation affects performance.
By focusing on the process as the unit of analysis, the authors consider how technology, human resources, and most importantly, the interaction between the two, contribute overall performance. Comparisons of performance are based on intermediary measures which evaluate the drivers of performance from the perspective of all participants in the service delivery process. That is, the outputs of the process do not explicitly represent the amount of money produced by the process but rather, consist of measures that are thought to move in coincidence with profitability.
The empirical analysis presented shows that there are not only statistical differences in the performance of the design groups, but that there are potentially vast differences in the recommendations resulting from analyzing the entire set of banks versus analyzing a single design group. For an individual bank, the effect of its performance projection amidst all banks as compared to solely with its design group is critical. By performing both types of analysis, the authors can predict for an individual bank the level o efficiency overall, the amount of improvement that might be achieved within its own design group, and the amount of improvement that will like require radical change. The bank will then be in a better position to analyze the tradeoffs involved in incremental improvement within a group versus a radical change required to move to another design group.
Download the paper