Abstract:     As a place-based intervention experiment, policy makers not only expect the pilot policy for combining technology with finance to play a "pilot role" to enhance the independent innovation capability of pilot areas, but also pay more attention to its "promotion value" and whether it can play a demonstration role to drive the integration of science and technology and finance in more regions and accelerate independent innovation. PSM-DID method has been used to evaluate the "pilot role" of the pilot policy for combining technology with finance comprehensively, but its spatial effect in non-pilot cities has not been carefully investigated. The study of the spatial effect of the pilot policy for combining technology with finance is related to its further promotion and practice, which is an important clue to understand the development of regional science and technology innovation. In particular, how does this spatial effect manifest itself? Siphon or spillover? It is controversial. 
    Based on the panel data of 265 cities from 2002 to 2020, this paper takes the "pilot policy for combining technology with finance" promulgated and implemented by the Ministry of Science and Technology in conjunction with the People′s Bank of China, the China Banking Regulatory Commission, the China Securities Regulatory Commission and the China Insurance Regulatory Commission in 2011 and 2016 as a quasi-natural experiment. For the first time, spatial difference-in-differences (SDID) analysis was used to quantitatively estimate its spatial effect on non-pilot cities. 
    The findings are as follows: (1) Under the premise of considering the spatial effect, the effect of the pilot policy on improving the level of technological innovation in pilot cities is still significant; (2) The pilot policy have spatial spillover effect on the whole, which promotes the scientific and technological innovation level of non-pilot cities; (3) The spatial effect of pilot policies is spillover between local governments at the same level within the same provincial government jurisdiction, which promotes innovation sharing within the province, but the spatial effect is siphon after crossing provincial boundaries, which is not conducive to the scientific and technological innovation of non-pilot cities in different provincial-level regions Robustness test and placebo test both supported the conclusions of this paper. 
    This study empirically tests the spatial effect of the pilot policy for combining technology with finance, and further confirms that the SUTVA hypothesis in the difference-in-differences method is not valid. Ignoring the spatial effect of the pilot policy for combining technology with finance will produce deviation in policy evaluation. The existence of spatial effect has verified and enriched the value of China′s policy pilot and made theoretical contribution to improving the scientific nature of policy evaluation after the event. The heterogeneity of the spatial effects of pilot policies on provincial administrative boundaries is also found, which fills the gap of previous studies. 
    This paper shows that the siphon effect and spillover effect of the central city on the surrounding city emphasized by the New Economic Geography (NEG) theory exist simultaneously, and the relative size determines the shape of the spatial effect of policy. When the siphon effect is greater than the spillover effect, the overall siphon effect dominates, and the policy pilot will inhibit the technological innovation of non-pilot cities. When the spillover effect is greater than the siphon effect, the spillover effect dominates on the whole, and the policy pilot can promote the technological innovation of non-pilot cities. Therefore, in order to accelerate the realization of the innovation-driven development strategy, the government must fully consider the inter-regional interest distribution relationship while promoting the orderly diffusion of pilot policies, pay attention to the influence of the internal political process and administrative structure of local governments on policy innovation and diffusion, and ensure the continuity of policy effects across provincial administrative divisions. 
    This finding further provides theoretical support for the central government to intervene vertically in the cooperative behavior of local governments, break down local protectionism, crack market segmentation and break through regional constraints to build a unified large market. To build bridges for innovation sharing, promote knowledge flow across administrative regions, and promote regional collaborative innovation.

Key words: science and technology finance, scientific and technological innovation, policy assessment, spatial difference-in-differences (DID) method