Abstract:    Cities gather all kinds of innovation resources, and the level of urban innovation plays an important role in high-quality economic development. Our government has been exploring the path of urban innovation, and "smart city" is one of them. In 2012, the Ministry of Housing and Urban Rural Development issued the Interim Measures for Administration of National Smart Cities (in pilot), marking the start of smart city construction at the national level. Subsequently, the Ministry of Housing and Urban Rural Development released the list of pilot cities in 2013 and 2014, and a total of 290 smart cities are on the pilot list, setting off a wave of smart city construction nationwide. 
    Smart city is the product of high-quality economic development. The construction of smart city is inseparable from the development of innovative enterprises in the city. In order to explore whether the R&D of registered enterprises can be improved and the channels for improvement after the smart city is rated as a pilot one, this paper tests the relationship between smart city and enterprise innovation. Enterprise innovation is an investment behavior with high risk, long payback period and large investment. One of the obstacles restricting enterprise innovation is the problem of capital. When the cash flow of the enterprise tends to be tight, the management will use the limited cash flow to maintain daily production and operation activities, so as to realize the contracts between the enterprise and stakeholders, but this will inevitably occupy the funds required for enterprise innovation. It is difficult and costly for enterprises to obtain R&D funds through debt and equity. However, government subsidies or tax preference do not require enterprises to pay costs and interests. In terms of smart city construction, the government guides enterprises′ R&D by issuing subsidies and promulgating preferential tax policies. 
    In addition, for enterprises, undertaking government smart city construction projects has many advantages. First, government subsidies or tax preference can be obtained; Second, it can improve its popularity and market share in a certain field; Third, it can increase the degree of association with the government to obtain more future government investment. The experimental results show that the smart city pilot has an innovation spillover effect. When the city is identified as a smart city pilot, the innovation level of registered enterprises will be significantly improved. The promotion channels are as follows: after the implementation of the pilot policy, the government has increased the special activity funds for enterprise innovation through government subsidies or tax preference, which has positively stimulated enterprise innovation. In addition, the heterogeneity test shows that compared with high financing constraint enterprises and state-owned enterprises, low financing constraint enterprises and the innovation spillover effect mainly affect low financing constraint enterprises and private enterprises; At the same time, it is more obvious in the high market competition environment and manufacturing industry.
    Based on the above conclusion, this paper puts forward the following suggestions: first, the government should increase government subsidies or tax preference for smart cities. In the construction of smart cities, the increase of government subsidies or tax preference can enhance enterprise innovation. Therefore, we believe that the government can increase government subsidies or tax preference for smart cities. In particular, we should focus on increasing government subsidies or tax preference for enterprises with high market competition environment, manufacturing enterprises, enterprises with low financing constraints, and private enterprises, so as to give full play to the role of smart cities in promoting enterprise innovation. 
    Second, the government should strengthen the supervision of high financing constraint enterprises or private enterprises. Compared with low financing constrained enterprises and private enterprises, high financing constrained enterprises and state-owned enterprises are more likely to be used for operation or other investment after receiving government subsidies or tax preference. This is because these types of enterprises are more likely to lack funds, and are more likely to use government subsidies or tax preference for daily operations or short-term profit-making projects. Therefore, the government should strengthen supervision and urge these enterprises to use their government subsidies or tax preference for innovation. 
    Third, enterprises should take the initiative to seek government "matchmaking" for scientific research institutes. Scientific research institutes can provide a large number of talents for enterprise innovation, but some small, medium and micro technology enterprises do not have a channel to contact scientific research institutes. They can seek the government to hold contact meetings to strengthen cooperation with scientific research institutes and ensure enterprise R&D.

Key words: smart city, enterprise innovation, government subsidy, tax preference