Taiwan owns the world’s largest manufacturing companies of high-tech components and products. According to the World Economic Forum’s “2007-2008 Global Competitiveness Report,” Taiwan has again taken first place in the world in the “State of Cluster Development” index. HsinChu Science Industrial Pak (HSIP) was established in northwestern Taiwan as a focal point for high-tech R&D and production. Since the park’s beginnings, in 1980, the government has invested approximately US$3.67 billion in it. By the end of March 2007, 475 high-tech companies in six industries (integrated circuits, computers and peripherals, telecommunications, optoelectronics, precision machinery, and biotechnology) were situated within the 625-hectare park. At the end of 2006, the park’s total paid-in capital exceeded US$35.4 billion, and more than 100 park companies were listed on Taiwan's main exchange, the Taiwan Stock Exchange, and on over-the-counter markets (Chen, 2007)
1.2 Research Motivation and Purpose
Industrial clusters can be characterized as being networks of production of strongly interdependent firms (including specialized suppliers), knowledge producing agents (universities, research institutes, engineering companies), bridging institutions (brokers, consultants) and customers, linked to each other in a value adding production chain. It is clear that clusters are dependent upon informal contacts which are based upon trust and reciprocity. Equally the transfer of ideas and a common labor pool enhances competition and reinforces the competitive advantage of the cluster as a whole. The development of successful high-technology industrial clusters is seen as
development, and the local organizations and institutions that evolved to serve them (Cortright and Mayer, 2001). Ketels (2003) argued that clusters develop over time;
they are not a phenomenon that just appears or disappears overnight. Clusters develop because regional proximity among firms promotes learning and competence building.
The industrial cluster will attract similar and related firms because they want to exploit the common knowledge base and take part in the interactive learning that takes place.
Cluster development, however, should be derived from cluster growth. Economic growth can primarily be explained and measured by per capita income output (Marshall, 1920; Rostow, 1960; Hicks, 1946; Lewis, 1955). Likewise, usually the growth of industrial cluster can be evaluated by annual industrial output changes.
There are some examples to measure the growth of industrial clusters in literature. Sull (2001) adopted an embedded case study design that explores the growth of the U.S. automotive tire industry at the level of the industry as a whole, the cluster centered in Akron, Ohio and constituent firms. Sull (2001) takes a historical perspective, covering the period between the emergences of the automotive tire industry in the early 1900s through to 1988, by which time only a single major U.S.
tire manufacturer remained after all others had been acquired by European or Japanese competitors. Moore (1978) proposed a method of characterizing the growth using two parameters, a production index measuring growth, and a structural change index measuring the change in the composition of output. Liu (2004) examined the sources of structural changes in output growth of China’s industrial cluster over 1987–1992 using a decomposition method.
As noted above, the HSIP will be our dissertation topic. Most specifically, this research applied historically quantitative forecasting methods to measure the growth of industrial clusters by observing and analyzing the changes of related industrial
output at HSIP. They are exponential smoothing forecasting model, the GM (1, 1) model, and the Grey-Markov model. We want to forecast the annual output using the exponential smoothing forecasting model, the GM (1, 1) model, and the Grey-Markov model. The period of this research is from 2001 to 2007. The computer and semiconductor industries are the research examples for estimating model. The first part of this research is to understand and estimate the annual output growth trend.
However in this section we offered an extra contribution which relates to offering an example as to how people in the future can apply a good estimation tool for industrial output efficiency. From the research results, the error rates for the exponential smoothing model are 13.48% and 13.49% for the two industries. The relative percentage errors of the GM (1, 1) model are 6.7116% and 7.20% for our surveyed industries. Notably, after the GM (1, 1) was modified using the Markov chain, the semiconductor industry’s annual output of absolute error decreased to 6.54%, while the computer industry’s annual output of absolute error decreased to 7.01%. Thus, our research results indicate that Grey-Markov estimating model is much more accurate for estimating the annual output of the semiconductor and computer industries in the case of HSIP.
From our estimation results, we also understand that the annual output of the semiconductor industry will slow down in the future while that of the computer industry has a decreasing trend. The annual values of semiconductor industry and computer industry account for over 50% for the HSIP. Industrial clusters can be seen as a main source of national competitiveness, serving to upgrade productivity, new business formation and innovation, and advance marketing/customer relations for Taiwan. Therefore, how to improve the alarmingly decreasing annual output and industrial value of HSIP is a critical and urgent task for industrial practitioners and
our research motivation for next step.
Therefore, the second part of this research is to contribute to the understanding of the casual and effect factors among those influencing an industrial cluster. Another core viewpoint anchored in this section is that national competitive advantages can be achieved by industrial clusters. That is, we would like to make use of the concepts of national competitiveness proposed by Porter (1980) and cluster drivers toward cluster value to conduct our second part of research.
We try to examine the impacts of and determine the relationships among different driving forces. Hence, we attempt to find out the impact of the major driving forces behind HSIP clustering and to measure the relationships among those forces. These factors of industrial clusters also exist for improving national competitiveness. To this end, we adopt the Diamond Model (Porter, 1998) in addition to the concept of culture to give the priority to these driving forces. Based on deductions from the prior literature, the driving forces in question are factor conditions, local demand conditions, related and supporting industries, firm structure and strategy and rivalry, government support, and culture. This research then applies the Decision Making Trial and Evaluation Laboratory (DEMATEL) to address the related issues. Discussing the relationship between different drivers and making a causal map that finds out the causal group and effect group, this research provides Taiwan industries and government with some strategic recommendations. It is because we believe that Taiwan can be viewed as an appropriate case demonstrating how her industrial clustering have resulted in particular national competitiveness, and from this perspective we wish to find out how those drivers associate with each other, finally leading to successful forms of clusters. Moreover, we attempt to draw upon our policy analysis results in order to assist government officials or industrial analysts in improving Taiwan’s industrial cluster policy and fostering the growth of the clusters.
1.3 Dissertation Organization
The dissertation is organized as follows. Section 2 is the first part. It presents a hybrid grey forecasting model for Hsinchu Science Industrial Park. Section 3 is the second part. It applies the DEMATEL method to find out the impact of the major driving forces behind HSIP clustering and to measure the relationships among those forces. Then Section 4 displays our conclusions and strategic suggestions.