سرریز دانش و عملکرد شرکت در خوشه های صنعتی با فن آوری پیشرفته
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|20287||2012||9 صفحه PDF||سفارش دهید||7440 کلمه|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Research Policy, Volume 41, Issue 3, April 2012, Pages 556–564
This paper attempts to empirically investigate the mechanisms underlying growth in Hsinchu high-tech clusters. We emphasize knowledge spillovers as one of the potential factors contributing to agglomeration benefits. This paper sheds light on the impact of external and internal spillovers on firm performance in Hsinchu high-tech clusters. The empirical results provide supporting evidence that the external R&D spillover is statistically significant in explaining net sales of firms in Hsinchu high-tech clusters.
This paper empirically investigates the mechanisms underlying growth in the Hsinchu Science Park (HSP). The Taiwanese government established HSP in 1980 to attract high-technology (high-tech) firms, including start-up firms. The primary goal was to create HSP as a potential Silicon Valley of the East. The park is located near the Industrial Technology Research Institute (ITRI) and two major research universities in Taiwan: National Tsing Hua University and National Chiao Tung University. To encourage overseas Taiwanese engineers to return home, the HSP administration as well as ITRI opened branch offices in Silicon Valley to provide information and local contacts. In addition, the Taiwanese government has invested US$1679 million in infrastructure and facilities in HSP since its founding a quarter of a century ago. At the end of 2006, 388 tenants occupied HSP, growing at an annual rate of 12% during the past two decades. Until 2006, overseas Taiwanese returnees had constituted almost one third of these tenants. Total sales in HSP reached US$310 million in 2006, representing an annual growth rate of 38%. The number of employees has increased more than tenfold from 8275 in 1986 to 121,762 in 2006.1 Agglomeration economies and institutional networks may explain the success of the HSP high-tech clusters. On the one hand, economists generally define agglomerations and clusters as a geographical and sectoral concentration of firms (Krugman, 1991). Hence, proximity and specialization could be key sources of collective efficiency in HSP. On the other hand, researchers in economic geography focus on the synergy formed by firms in cooperative networks (Saxenian, 1994). The competitive advantages of firms or industries in HSP may well thank backward and forward linkages, labor pooling, and knowledge spillovers via inter-firm or inter-industry linkages. Among these factors contributing to agglomeration benefits, knowledge spillovers have attracted the most attention in the recent literature on new growth theory and industrial geography. In empirical studies surveyed comprehensively by Nadiri (1993), spillover effects in general have significantly positive influences on productivity at both industry level and firm level. However, the magnitude of these effects varies substantially according to the different measuring methods used in assessing the inter-industry spillover. We apply the technology-based approach proposed by Jaffe (1986) to assess the proximity between HSP's firms in a technological space. Specifically, patents classified across various technological categories allow us to characterize firms’ positions in the technological space. The total number of patent applications filed by firms in HSP to the Taiwan Intellectual Property Office (TIPO) from 2001 to 2005 is classified into 23 technological classes to reflect the distribution of patents and accordingly, the locations of firms in the technological space.2 Based on firms’ patent distributions among 23 technological sectors, we classify them into technological clusters by the k-means clustering technique. Consequently, measures of external and internal knowledge spillovers can be constructed. Both internal and external knowledge pools relevant to advancing firms’ performance in HSP are established. An econometric model for estimating the effect of knowledge spillovers on net sales is founded on the Cobb-Douglas production function. The data set we use is comprised of 92 HSP firms listed publicly in Taiwanese stock markets during 2000–2004. The empirical results suggest that the HSP high-tech clusters show evidence of important knowledge spillovers since the total R&D spillover and the external R&D spillover are both statistically significant in explaining net sales. In particular, the estimated total R&D spillover and external R&D spillover elasticities are higher than own R&D elasticities. Furthermore, we examine the influence of domestic versus international knowledge spillovers on the production of the HSP firms in the semiconductor industry. Using panel data from firms in the U.S. and Taiwan, our empirical results of international knowledge spillovers show that the foreign knowledge stock has a positive impact on net sales. Section 2 provides a literature review of our study. The empirical methodology is presented in Section 3, the data description and empirical results are reported in Section 4, and the concluding remarks follow in Section 5.
نتیجه گیری انگلیسی
This paper attempts to empirically investigate the mechanisms underlying growth and change in the Hsinchu high-tech clusters. The growth of a cluster can be explained by agglomeration economies and institutional networks. This paper sheds light on the impact of internal and external spillovers in Hsinchu high-tech clusters on firms’ performances in net sales. We apply the technology-based approach proposed by Jaffe (1986) to assess the proximity between HSP's firms in the technological space. Specifically, patents classified across various technological categories allow us to characterize firms’ positions in the technological space. The total number of patent applications filed by firms in HSP to TIPO during 2001–2005 is classified into 23 technological sectors to reflect the distribution of patents, and accordingly the locations of firms in the technological space. Then, we classify firms into technological clusters by the k-means clustering technique. As a result, the external and internal knowledge spillovers can be measured. The data set we use comprises 92 HSP firms listed on Taiwanese stock markets during 2000–2004. We apply the system GMM estimation method to our panel data. The empirical results suggest that Hsinchu high-tech clusters show evidence of positive knowledge spillovers in terms of total R&D spillover and external R&D spillover. In particular, the estimated total and external R&D spillover elasticities are higher than own R&D elasticities. Furthermore, this paper gauges the impact of domestic and international knowledge spillovers on the production of HSP firms in the semiconductor industry. We find that the foreign knowledge stock has a positive impact on net sales. Hence, in addition to the external and internal knowledge spillovers created domestically, this paper sheds light on assessing the international knowledge spillovers of the Hsinchu high-tech clusters. Due to data availability, the panel data only include time series with a length of five years. That being said, it is difficult to fit a meaningful time structure from the short panel data, especially when we apply the system GMM estimation in the empirical study. Our future work will focus on extending our sample size and dataset. When taking into account other available data sources, we may improve the construction of measures of both domestic and international knowledge spillover pools relevant to firms’ performance in the HSP. In addition, not only USPTO's patent citations of Taiwanese high-tech companies can be considered as a potential indicator of technological proximity; our future study may also explore the geographical proximity concept proposed by Aldieri and Cincera (2009).