地理约束、合作与劳动力知识分配空间偏好

Geographical Constraints,Collaboration and Spatial Preference of Labor Knowledge Allocation

经济高质量发展阶段,知识资源的空间再分配有着重要的现实意义。然而,在当前地理条件的约束下,对运输成本的改善将如何影响不同区位间技能劳动力跨空间知识合作行为,以及其在不同区位合作对象间进行知识分配的空间偏好及其倾向与策略变化,却鲜有研究。为此,本文首先构建了一个纳入不同地理约束条件、运输成本变化以及基于异质性劳动力知识合作的区位选择模型;其中,城市间交通网络设施的改善降低了技能劳动力面对面交流的时间成本,提高了知识合作的匹配质量,从而促进了更远距离的知识合作及其绩效,同时伴随着知识配置从区域内部向外部的转变。另外,一个城市知识资源配置的空间偏好会随着该城市所达可交流市场量的增加而先递减再递增,并呈现知识资源集中分配趋势。进一步,本文利用2006-2015年的我国论文合作数据,基于多期双重差分法,估计了由高铁线路连通带来的劳动力运输成本的降低对知识资源空间配置行为的影响。本文的实证结果与理论模型相一致。

Although the spatial reallocation of knowledge resources is crucial for high-quality economic development,the related studies are limited. This paper attempts to bridge the gap by answering the questions:How does the reduction of transportation costs, subject to geographical constraints, affect knowledge collaboration between skilled labors of different locations,and how does it alter the spatial preferences,tendencies,and strategies for knowledge allocation among collaboration partners in different locations? To this end,we first construct a location selection model that incorporates different geographical constraints,changes in transportation costs,and knowledge collaborations among heterogeneous labor. The model shows that the improvement of inter-city transportation networks reduces the time cost of face-to-face communication between skilled labor,and increases the matching quality of knowledge collaborations,which in turn promotes the formation and performance of distant collaborations. Consequently,knowledge resources of a region shift from the inside to the outside of the region. In addition,when the markets accessible to a city increase,the city’s spatial preference for knowledge allocation will decrease first and then increase,making the distribution of its knowledge resources more centralized. Finally,we use China ’s co-publication data from 2006 to 2015 to estimate the effect of high-speed rail connection between cities on the spatial allocation of knowledge resources via the multi-period DID method,and the empirical results verify our model.