台灣積體電路產業變革與創新政策互動之研究

Abstract

本研究探討產業變革與創新政策之互動，並以台灣積體電路產業為研究對象。本研究建立一「技術系統」為基礎的層級評估架構，以「模糊積分多準則評估法」 (Fuzzy integral MCDM) 衡量產業變革創新政策之優先順序。並以變異數分析，確定創新政策工具在產業結構變革中呈現不同的效力。影響產業變革的因素包括：「需求」、「供給」、「分工」和「技術」。「技術」以直接的方式影響產業發展，而「技術系統」可解釋技術推動產業變革。

研究結果顯示，在台灣IC產業變革過程中，其中10項創新政策工具在台灣IC產業變革中呈現不同的效用。在台灣IC產業技術輸入期，政府著重「環境面」政策工具，而「科學與技術發展」、「公營事業」主導國外技術獲得之管道。台灣IC產業技術改良期，政府藉「科學與技術發展」、「公營事業」、「教育與訓練」等供給面政策將技術擴散到整個IC產業以達到最大經濟效益。台灣IC產業技術開發期，政策應重視「需求面政策」以協助本國產品拓展市場。其中政策綜效較前兩階段低，政府對IC產業介入與干涉的程度逐年降低。

本研究同時發現，政府執行創新政策工具時需要相互配合以增加政策效果。本研究提出一套可行的創新政策評估模式，明確指出台灣IC產業結構變革及其所需配合的創新政策工具。

This study focuses on an empirical study of Taiwan’s IC industry. A hierarchical fuzzy integral multi-criteria decision-making (Fuzzy integral MCDM) approach for evaluating the effects of innovation policies is presented in this study. One-way ANOVA test is applied to examine policy impacts amid the three phases of industrial development. Four evolution parameters are chosen to investigate the dynamic interactions of industrial evolution and innovation policy, namely, supply conditions, demand conditions, structure of industry’s supply chain, and technology, among which the latter is solely used as an index for the evolution. This study attempts to develop a “technological system” for evaluating the priorities of innovation policy

The results show that ten policy tools were found significantly different in the three phases of evolution. An “Environment side” policy is found to be effective for the phase of technology acquisition. In this phase, the public sector also plays a dominant role in absorption and adaptation of foreign technology, where policies of “Scientific and technical development” and “Public enterprise” are found important in this phase. A “Supply side” policy is significant for the phase of technology absorption, and policies of “Scientific and technical development”, “Public enterprise”, and “Education” were executed intensively in this phase to expedite technology diffusion. The policy impacts for the phase of technology development are found to be lower than the previous two phases. Government involvement in the phase of technology development is diminishing. The results also suggest that there is a non-additive effect among the 12 policy tools during industrial evolution.