This paper investigates the productivity change of Japanese credit banks with a Malmquist index and the input technological bias during 2000–2006. Our results indicate that the traditional growth accounting method, which assumes Hicks neutral technological change, is not appropriate for analyzing changes in productivity. Our analysis unambiguously shows that management of Shinkin banks has to be improved. These must be based on the improvement of technical efficiency and/or technological change, emulating the procedures of the best-practice banks, i.e., those banks with Malmquist productivity scores higher than one and simultaneously with technical efficiency and technological change higher than one.
Banks efficiency and performance are important indicators that may identify actual or potential problems at individual banks and in the sector as a whole (Molyneux and Williams, 2005). Such information is also valuable in order to compare competitiveness and efficiency of banking systems. If there is significant inefficiency in the sector, in general, and in different groups of banks, in particular, there may be room for structural changes, increased competition, mergers and acquisitions (Berger and Mester, 1997).
The standard models estimate efficiency via parametric techniques and non-parametric methods. In our study, we apply Data Envelopment Analysis (DEA) and the Malmquist productivity index as in Alam (2001). The advantage of using non-parametric frontier techniques is that they impose no a priori functional form on technology, nor any restrictive assumptions regarding input remuneration. Furthermore, the frontier nature of these techniques allows any productive inefficiency to be captured and offers a “benchmarking” perspective. The DEA method has been widely used to estimate the reciprocal of the Shephard (1970) input distance function. The reciprocal of this distance function serves as a measure of Farrell (1957) input efficiency and equals the proportional contraction in all inputs that can be feasibly accomplished by given output, if a decision-making unit (DMU) adopts best-practice methods.
The objective of our study is the following. We link input efficiency indexes across time in order to estimate the Malmquist productivity index. This index estimates the change in resource use over time that is attributable to efficiency change and due to technological change. Furthermore, we use the approach of Färe and Grosskopf (1996) and decompose technological change into an index of output biased technological change, an index of input biased technological change, and an index of the magnitude of technological change. The introduction of the Malmquist index with biased technological change in banking is a novelty in this context.
Current research studies on productivity change have applied the Malmquist index but without considering biased technology, see, for example, Guzmán and Reverte (2008), Casu et al. (2004), Sturm and Williams (2004). Studies focusing on savings banks and cooperative banks include, for example, Grifell-Tatjé and Lovell (1997), Carbo et al. (2002), Williams (2001), Kumbhakar et al. (2001). However, we apply the innovative frontier models on the segment of financial institutions – credit banks (Shinkin banks) in Japan. Our methodological concept contributes to research studies on Shinkin banks presented by Satake and Tsutsui (2002), Hosono et al. (2006), Fukuyama and Weber, 2008 and Fukuyama and Weber, 2009. The study has clear implications for the definition of productivity strategies, taking into account the previously managerial practices identified in the study at unit levels.
The remainder of this paper is organised as follows. Section 2 presents the contextual setting of credit banks in Japan. Section 3 presents the literature survey. Section 4 presents the data and the detailed methodological concept of our model. Section 5 discusses the results. Section 6 concludes.
The present paper analyses changes in productivity in Japanese credit banks during the period of significant macroeconomic imbalances in the Japanese economy. We emphasize several implications of our findings for economic policy. Firstly, Shinkin banks show, on average, negative productivity change during 2000–2006. We show that individual components of productivity change, i.e., changes in technical efficiency and changes in technological efficiency are negative. The results are rather surprising and raise a number of questions to be asked by managers. If we divide Shinkin banks into prefectures, we see that only 8 out of 47 prefectures achieved positive improvement in productivity, which is attributed to technical efficiency change, signifying that the major cause of the average negative productivity is due to a negative variation in technological change. The intuitive explanation of the negative variation in technological change may be explained in the context of protracted economic slowdown and turbulences in the Japanese financial markets.
This estimation unambiguously shows that management of Shinkin banks has to be improved. These must be based on the improvement of technical efficiency or/and technological change, emulating the procedures of the best-practice banks, i.e., those banks with Malmquist productivity scores higher than one and simultaneously with technical efficiency and technological change higher than one.
Technical change in the majority of Shinkin banks captured by the output bias (OBTECH) and input biased variable (IBTECH), indicates that there is no overall shift in the best-practice frontier between 1999 and 2005. Thus, the marginal rate of substitution between outputs is affected by technical change, i.e., the marginal rate of substitution between loans and securities. The marginal rate of substitution between inputs is then affected by technical change that is in our case marginal rate of substitution between deposits, total number of employees and total fixed assets. Therefore the assumption of parallel neutrality is also rejected for inputs.
