研發外溢效果對高科技廠商表現之貢獻-台灣新竹科學園區之實証分析

Empirical Examination of the Impact of R&D Spillovers on the Performance of High-technology Firms in Taiwan's Hsinchu Science Park

宋佩珊(Pei-Shan Song)

指導教授 ： 祁玉蘭

國立清華大學/科技管理學院/經濟學系/碩士(2007年)

https://doi.org/10.6843/NTHU.2007.00479

摘要

本文以新竹科學園區中公開發行的公司為研究對象，探討研發外溢 (R&D spillover) 效果是否對高科技廠商的表現有正面影響。本文蒐集這些公司2003-2005的年報資料，及2001-2006的台灣專利核准資料，並採用Griliches (1979) 提出的擴張的Cobb-Douglas 生產函數 (extended Cobb-Douglas production function) 為本文實證模型。
本文將研發外溢效果區分成知識外溢 (knowledge spillover) 效果和租金外溢 (rent spillover) 效果，分別依據Jaffe (1986) 及Terleckyj (1974) 建立變數。在知識外溢效果部份，根據分群結果把總知識外溢效果 (total knowledge spillover) 分成群內知識外溢效果 (local knowledge spillover) 及群外知識外溢效果 (external knowledge spillover)，此外，本研究根據竹科公司的奈米專利資料建立奈米知識外溢效果 (Nano knowledge spillover)，並討論工研院在奈米方面的研發對竹科廠商銷售表現的影響。本研究主要的貢獻在於建立廠商的投入產出表，並從兩方面探討租金外溢效果；原料提供者方面 (input rent spillover) 和產品購買者方面 (output rent spillover)，並同時討論廠商對於研發外溢的吸收能力 (absorptive capacity) 是否也會影響廠商的表現。
本研究在知識外溢效果的部份，樣本為130家竹科公開發行公司，期間為2003至2005年，在租金外溢效果部份，樣本為77家竹科中半導體和光電產業的公司在2005年的資料。估計方法分別採用固定效果 (fixed effect)、隨機效果 (random effect)、兩階段最小平方法 (two-stage least squares) 及普通最小平方法 (ordinary least squares)。
本研究的實證結果如下：(1) 知識外溢效果對廠商表現有正面的影響，但其顯著程度會隨估計方法的選擇而有所不同。 (2) 租金外溢效果對廠商表現的影響並不顯著。

關鍵字

研發外溢 ； 追蹤資料 ； 新竹科學園區

並列摘要

This study investigates the contribution of R&D spillover to firm performance in high-tech clusters, and attempts to explain the rapid growth of the Hsinchu Science Park (HSP).
R&D spillover is distinguished into knowledge spillover and rent spillover. Knowledge spillover variables are constructed based on Jaffe (1986), while rent spillover variables are developed according to the method proposed by Terleckyj (1974). Total knowledge spillover contains local and external knowledge spillovers, which are constructed based on the clustering result. Additionally, nano knowledge spillover variables are developed using the distribution of nano patents. The construction of pseudo input-output table in firm level is the main contribution of this thesis. Input and output rent spillovers which from potential suppliers and customers and the R&D absorptive capacities are considered together to estimate their influence on HSP firm performance.
For the analysis of knowledge spillover effects, the data set consists of 130 HSP high tech firms over the period 2003-2005. However, the data set for analyzing rent spillover effects is collected form 77 HSP IC and optoelectronics firms in 2005.
This study adopted panel data models to test the impact of knowledge spillovers. Also, 2SLS is used to generate robust estimate. For the influence of rent spillovers, ordinary least squares is the method applied in this work.
Empirical results of this study are summarized as follows. Total, local, external, and nano knowledge spillover stocks positively influence HSP firm performance. However, the significances of these knowledge spillover effects depend heavily on the adoption of econometric methods. Nevertheless, the rent spillover effects are obscure.

並列關鍵字

R&D spillovers ； Panel data ； Hsinchu Science Park

參考文獻

Anderson, G. (1994). Industry Clustering for Economic Development. Economic Development Review, 12(2), 26-32.

Arrow. (1962). The Economic Implications of Learning by Doing. The Review of Economic Studies, 29(3), 155-173.

Basberg, B. (1987). Patents and the Measurement of Technological Change: A Survey of the Literature. Research Policy, 16, 131-141.

Cincera, M. (2005). Firms' Productivity Growth and R&D Spillovers: An Analysis of Alternative Technological Proximity Measures. Economics of Innovation and New Technology, 14(8), 657-682.

Cohen, W. M. and Levinthal, D. A. (1989). Innovation and Learning: The Two Faces of R & D. The Economic Journal, 99, 569-596.