企业之间创新联系的行政区边界效应

行政区边界作为一种城市单元界线分割了地理空间，造成企业创新联系的边界效应。论文基于2008—2017年上海市高新技术企业微观统计数据，构建空间断点回归模型，检验了行政区边界对专利合作、引用、代理、转让和许可5类跨区创新联系的抑制作用。研究表明：① 创新联系存在明显的边界效应，较代理、转让、许可3类市场途径，合作、引用2类非市场途径的边界效应更强；② 随着时间的推移，边界效应呈明显的下降趋势；③ 政府对企业创新的较多干预和本地特有的创新环境会导致这种边界效应增强；④ 行政区合并降低了被合并的2个行政区的边界效应。实证结果可为促进企业创新联系、提高创新绩效提供政策参考。

Border effects of innovation connections between firms

Innovation has a geographic dimension that promotes economic growth and technological change, and is closely related to regional spatial characteristics. The innovation connections formed by firms are key elements in understanding the spatial dynamics of innovation processes. As a geographic element, administrative boundaries divide urban space, resulting in the border effects of firm innovation connections. Based on the micro statistical data of Shanghai high-tech enterprises from 2008 to 2017, this study used a spatial regression discontinuity design to test inhibitory effects of administrative boundaries on five types of cross-district innovation connections, including patent cooperation, citation, agency, transaction, and licensing. The empirical results are as follows: 1) Administrative boundaries have a significant inhibitory effect on the five types of cross-district innovation connections, called border effects, and this inhibitory effect still exists when controlling for distance and firm locations. Compared with market approaches (agency, transaction, and licensing), border effects are higher on non-market approaches (cooperation and citation). 2) Temporal trend analysis showed that there is a significant downward trend on the intensity of border effects from 2008 to 2017, which may be related to the fact that in recent years, the governments have continuously encouraged firms to establish interregional innovation connections through various political instruments. 3) Mechanism analysis showed that government intervention in firm innovation and local innovation environment result in the increase of border effects. Administrative district consolidation reduces border effect between two merged districts. The empirical results of this study provide policy references for promoting firm innovation connections and improving innovation performance.