Fiscal decentralization, preference for government innovation and city innovation Evidence from China

Purpose The purpose of this paper is to examine whether fiscal decentralization has impacts on city innovation level and to examine the moderating effects of the preference for government innovation in China. Design/methodology/approach Using a panel data of China's 278 cities from 2003 to 2016, the authors first use fixed-effect model and quantile regression to analyze the impact of fiscal decentralization on city innovation level and the variations of impacts conditional on different innovation levels, followed by a mediating effect model to test the moderating effects of the preference for government innovation and its temporal and spatial heterogeneity. Findings The paper finds that fiscal decentralization significantly inhibited city innovation, and with the improvement of city innovation level, the inhibition demonstrated characteristics of "V" type variation. When the degree of fiscal decentralization is between 0.377 and 0.600, the inhibition of fiscal decentralization on city innovation level is the weakest. We further show that fiscal decentralization also inhibits the government's preference for innovation, reduces the proportion of fiscal expenditure on innovation and has a negative impact on city innovation. In addition, the influence of fiscal decentralization on city innovation present clear heterogeneity in space and in time. On one hand, the inhibition of fiscal decentralization on city innovation level in eastern China is significantly weaker than that in central and Western China; on the other hand, after the implementation of China's innovation-driven development strategy in 2013, the negative impact of fiscal decentralization on city innovation disappeared. Originality/value This paper extends prior research by bringing the decentralization system reform into the study of city innovation system and analyzing its mechanism and its temporal and spatial heterogeneity.