工业用地价格、选择效应与工业效率

Industrial Land Price, Selection Effect and Industrial Efficiency

本文基于新新经济地理(NNEG)理论,首次从"选择效应"的视角对工业用地价格影响工业效率的作用机制进行了理论与实证探讨。利用中国工业企业调查数据库和土地市场动态监测与监管系统数据库,从区县和企业维度实证分析了2007—2012年中国35个省会、副省级及以上城市工业用地价格对工业效率的影响,验证了选择效应的存在性和作用机制。研究发现:第一,工业用地价格对工业效率具有提升效应,且在城市内部呈现明显的空间差异性,中心城区的选择效应强于外围地区,在考虑不同效率测度方法和空间识别策略的情况下,结果仍然稳健;第二,工业用地价格上升带来的工业效率优势来源于选择效应,其作用机制体现在更低的低效率企业进入比例和更高的低效率企业淘汰风险。

The development of cities involves not only scale expansion, but also structural upgrading and layout optimization. In the process of industrial development, structures are transformed and upgraded and spaces are optimized and adjusted. Based on the public land ownership and land reserve system, local governments in China have implemented land supply as the main policy control method. Land is an important input factor in industrial production. The change in production costs resulting from a change inland prices is a major driving force for industrial efficiency and layout optimization. The present study explores a reasonable price strategy for industrial land supply. This strategy reveals the "black box" of regional industrial efficiency changes and provides a reference for the precise allocation of land resources.Adopting the "new" new economic geography(NNEG) theoretical framework, we use the Chinese Industrial Enterprises and Dynamic Monitoring and Supervision System of Land Market databases. We comprehensively apply spatial positioning technologyand the S-SBM-DEA model to construct districts and counties of 35 cities in China from 2007 to 2012. Using the urban master plans of cities as a reference, we divide the "central-periphery" space within each city. We then include the effect of prices on the level of industrial efficiency from multiple dimensions, along with the existence of a selection effect and its acting mechanism. We consider the slope of the sample as an instrumental variable of industrial land price, and an unconditional distribution characteristic-parameter correspondence function is used to solve endogeneity issues.The main findings are the following. First, overall, the land price has a considerable effect on improving industrial efficiency;a higher price leads to higher industrial efficiency. Second, in terms of urban interior space, the improvement effect shows strong spatial heterogeneity according to market access and distance from downtown. Central areas have the strongest effect, followed by suburban areas, and then outer suburbs. Because the improvement effect is mainly due to the difference in affordability of land costs for heterogeneous companies, the effect is unsustainable. Finally, the industrial efficiency advantages arising from higher land prices are mainly due to the selection effect;the stronger selection effect in higher-priced zones left-truncates the productivity distribution. The selection effect in higher-priced zones can mainly be explained by the lower entry proportion and higher exit probability of inefficient companies. One of the implications of these findings is that the government should fully exploit the selection effect of land prices on firm efficiency and further optimize the market-oriented mechanism of industrial land bidding/selling. In addition, the degree of government intervention in the land market should be reduced, and the incentives of local governments to "manage land" based on fiscal and taxation goals should be eliminated.This paper contributes to the literature in several ways. First, using micro-data on parcel land transactions, we accurately characterize the spatial evolution features of industrial land prices in inner cities through spatial location technology using Xgeocoding and semiparametric estimation. This study also uses micro-data to support an in-depth study of the effect of industrial land prices on industrial efficiency at the district and county levels, and develops the spatial scale of existing empirical research. Second, by applying the NNEG theory and unconditional distribution characteristic-parameter correspondence function, we creatively explore and verify the selection effect of higher-priced land zones. We identify the mechanism of the selection effect in higher-priced zones from the perspective of the entry and elimination of heterogeneous firms. Further, we distinguish the acting mechanism between the lower proportion of inefficient companies entering the market and higher corporate elimination risk. Third, unlike studies that use population density or employment density to identify urban central-peripheral areas, the present work is the first to define central urban areas, suburban areas, and outer suburbs based on the urban master plans of major cities. This approach intuitively reflects the spatial heterogeneity of the selection effects of industrial land prices under government guidance.