政府科技投入对企业R&D支出影响的再审视——基于分位数回归的实证研究

The Impact of Government S&T Input on the Enterprise's R&D Expenditure——Empirical Research Based on a Quantile Regression

本文基于1997—2009年中国省级面板数据,运用分位数回归方法考察了政府科技投入对企业R&D支出在各区域以及各分位点处的影响效果.研究表明:政府科技投入对企业R&D支出的影响效果存在鲜明的区域与分布差异性.从全国范围看,政府对高校、科研院所的研发资助会对企业R&D支出产生杠杆效应,而对企业的研发资助则存在显著的挤出效应,且随着条件分布向高分位点处移动,二者之间的差距将逐渐减小;三大经济区域中,政府对中西部高校、科研院所的研发资助均对企业R&D支出产生杠杆效应,而仅有对东部企业的研发资助表现为杠杆效应,对中西部企业研发资助的效果要视分位点位置而定.

Based on the provincial panel data of China from 1997 to 2008, it examines the impacts of government S＆T input on enterprise＇s R＆D expenditure in each region and typical quantile by using the quantile regression methods. The results show that the government S＆T input has distinctly different influences on enterprise＇s R＆D expenditure due to different region and distribution. Nationally, the public R＆D funds to universities and research institutes produce leverage effects on enterprise＇s R＆D expenditure, but those to enterprises have significant crowding-out effects, and the gap between the two effects gradually decreases with the conditional distribution moving to the higher quantile. Regionally, the public R＆D funds to universities and research institutes have positively leverage effects on enterprise＇s R＆D expenditure in the central and western China, but exert merely the same effect on the eastern enterprise R＆D funds. In addition, the subsidized impact on the central and western enterprises depends on the location of specific quantile.