محاسبات از میزان سهم پیشرفت های علمی و فن آوری به رشد اقتصادی در مناطق چینی

According to the new economic growth theory, a new method of computing the contribution rate of scientific and technological (S&T) progress to economic growth based on the Cobb–Douglas production function and the Solow residual value method is proposed in this paper. This method includes three steps: Firstly, according to their levels of S&T progress, fuzzy soft clustering of thirty one Chinese regions is performed to obtain the membership degree of these places to the categories. Secondly, to calculate the contribution rates that different categories of levels of S&T progress contribute to economic growth. Thirdly, to multiply the obtained contribution rate of each category by the membership degree of the place belonging to this category, from which the contribution rate of S&T progress to economic growth in each place is obtained. Finally, this method is used to calculate the contribution rates of S&T progress to economic growth in thirty one Chinese regions during the period from 1998 to 2007. Last but not least, some reasonable suggestions and conclusions are proposed by analyzing the computing results.

Economy is a product of the co-operation of science, techniques, labors and capital. Seeing from the strategic angle of the economic development, in order to realize the economic development in a continual, stable, and highly efficient way, we have to adjust the proportion of various input factors in the economy. It is important to obtain the contribution rate of each factor to economic growth, especially the contribution rate of S&T progress, which can help the leader make reasonable macro policies. The contribution rate of S&T progress to economic growth, in the broad sense, refers to the sum of contributions rate of other factors to output increase, excluding those of the increase in labor force and capital (Robert, 1996, Li, 1995 and Sun, 1998). Mokhtarul Wadud (2004) and Lewis (1954) had proposed the total factor productivity (TFP). Since then, the idea that S&T progress has been an important factor in economic growth which attracted the attention of economists. In 1957, Solow (1957) adopted residual value method in computing the contribution rate of S&T progress, which created the measurement of S&T progress operational. After that economists have made continuous improvement in studying the contribution rate of S&T progress to economic growth (Hilbrink, 1989, Jorgenson, 2001, Sengupta, 2004, Shipley et al., 2004 and Tallon and Kraemer, 1999), among which Denison’s research in the 1970s is relatively significant. Denison, when analyzing economic growth reasons from the year of 1929 to 1969 in the USA, divided the total input and total factors productivity into several small factors, and based on this classification he made a quantitative measurement of the effect of each individual factor on economic growth (Denison, 1962). The study on S&T progress in China starts relatively late. In the 1980s, Jia (1997) initiated the potential analysis method with Chinese characteristics to estimate contribution rate of S&T progress to economic growth, which is based on the active decision theory. In 1983, the research group guided by Shi and Qin the first time completed the analysis of the effect of industrial technological progress in China (Shi & Qin, 1985). In 1998, the Department of Science and Technology of the former State Planning Committee in China launched research program on measuring the effect of S&T progress on economic growth (Jia, 1997). In 2000, Lu, Fan, Wei, and Xu (2000) had adopted such methods as the Solow function, the Denison’s analysis on economic growth factors, the Jogenson’s analysis of production efficiency and the production function to measure the effect of S&T progress on economic growth and had made empirical analysis. In 2002, Li (2003) had utilized the input and output method to measure the contribution rate of S&T progress to economic growth. Song (2003) had improved the measuring method of S&T progress by adopting the potential analysis theory. Zhou (2008) had applied the Cobb–Douglas production function in measuring the contribution rate of S&T progress to economic growth in 10 regions in Henan Province of China in the year of 2005. Liu (2006), by taking the Cobb–Douglas production function as the basic method, had measured the contribution rates of S&T progress, capital and labor force to economic growth respectively during the period of 1985–2002 and the period of 1995–2002 in Hebei province of China. Wu (2008) had measured the contribution rate of S&T progress in agriculture in Henan Province of China from the year of 1996 to 2005, and used the gray production function metabolizing model (Liu, 1997, Liu et al., 2004 and Liu et al., 1999) in predicting the contribution rate of S&T progress in agriculture in Henan province of China from the year of 2006 to 2015. It is obvious that Chinese and overseas scholars have made encouraging achievements in measuring the contribution of S&T progress to economic growth, but these researches are limited to use data in one specific country or region when at a given period of time. The limited data may have negative impact on the accuracy of parameter estimation in the production function. What’s more, the contribution rate of the same S&T progress in different regions may be significantly different and incomparable. In addition, the social and economic system is complicated and nonlinear, and the same S&T progress may make different contributions in different regions. So the new method proposed in this paper aims at measuring the contribution rates of S&T progress to economic growth in thirty one regions of China from the year of 1998 to 2007, by adopting the soft computing approach, the Cobb–Douglas production and the Solow residual value method. In the new method, we first make soft clustering of S&T progress in thirty one regions of China, then compute the contribution rates of S&T progress at various classifications to economic growth. And then calculate the contribution rate in each region by summing up the products of the contribution rate of each classification and the membership degree of this region to each classification. Finally, comparison of the rates in different regions is provided and rational suggestions to the economic development in various regions are presented.

In 1970s, Denison analyzed the courses of economic growth of the United States through the accounting method of the sources of economic growth. Since then, many scholars have not only started to study the contribution rate of S&T progress to economic growth, but also proposed many measuring methods. In the late 1980s, China began to explore thoroughly and systematically the relationship between S&T progress and economic growth. Based on longitudinal data, however, current researches mostly measure the contribution rate of S&T progress to economic growth on certain countries or areas at certain period. Unfortunately, due to the limitation of this kind of data sample, the preciseness of the parameter estimation in the production function can be affected and the comparability of the contribution rate of S&T to economic growth failed to be achieved. This paper proposed a new approach measuring the contribution rate of S&T progress to economic growth. The first step is to conduct fuzzy soft clustering on the 31 provinces and municipalities of China according to the level of S&T progress. The second step is to measure the contribution rate of S&T progress of each cluster. The third step is to calculate the contribution rate of each region according to the principle of membership. Therefore, we can successfully compare horizontally the contribution rate of S&T progress to economic growth of different regions in China. Through the analysis of reasons, corresponding conclusions can be reached and the results can provide guidance for the country and the all-level government foundations to make new counter-measures and policies. But in the process of estimating contribution rate, there are two key aspects which need to be improved. One is that the elasticity coefficient of each factor is different as each factor is measured differently, which affects the estimating results. The other one lies in the methods of measuring elasticity coefficient. Each method has advantages and disadvantages. For instance, the classical regression method is sensitivity to samples. These two aspects are the difficult parts in the process of estimating. In the future, we will consider applying neural network to measure elasticity coefficient.