政府治理與國家研發效率：隨機距離函數法之應用

Abstract

本篇文章採用隨機距離函數法來衡量政府治理與國家研發之間的關係。資料數據涵蓋59個國家，而資料年限為1998年、2000年、2002到2008年。在多投入與多產出的模型中，我們使用了R&D人員與R&D費用存量做為投入項；產出項則為專利數、科學期刊文章與技術授權金。研究結果發現六個政府治理指標對研發效率有正面的影響，尤其是政府效率指標對研發效率的影響最大。進一步把六項政府治理指標區分成三個面向時，同樣顯示公部門治理對研發效率有正面影響效果。

This paper adopts the distance function approach for stochastic frontier analysis (SFA) to estimate the relationship between six governance indexes and research and development (R&D) efficiency across 59 nations during the years of 1998, 2000, and 2002–2008. The multiple inputs and outputs framework which includes R&D manpower, R&D expenditure stocks as inputs and patents, scientific journal articles, royalties and licensing fees (RLF) as outputs. We find that all six indexes have positive effects on R&D efficiency; moreover, the government effectiveness especially plays the most important role. Grouping the six indexes into three dimensions still shows the same result that the public governance significantly improves national R&D efficiency.