In the development process of Chinese enterprises, there are many internal factors, such as the "modularization" trap, homogenization innovation, "surge" of industry and other independent innovation difficulties, and many external factors, such as imperfect intellectual property system and financial market system, all these factors together breed the enterprise innovation "inertia". Although the government′s direct financial support can alleviate the financial constraints faced by enterprises′ innovation investment, the hidden information problem in the allocation of government innovation subsidies makes the enterprises prone to rely on inertia, resulting in enterprises′ tendency to "cater" to government policies. They tend to "cater" to government policies through rent-seeking and preferential policy grabbing behaviors, resulting in the dilemma of "focusing on quantity rather than quality" in enterprise innovation. Therefore, utilizing government intervention to achieve the positive incentive effect on enterprises′ innovation willingness and high-end innovation preference is of great significance for enhancing enterprises′ innovation vitality and enabling local high-quality regional economic growth.
   In this paper, we use the data from the CSMAR database and list 2480 companies from 2008 to 2017 as the research objects, finally we test the multi-dimensional impact of government innovation funding on enterprise innovation laziness by using the method of 2SLS and IV Probit. The results show that: (1) after controlling the fixed effects such as time, industry, city size, and enterprise′s characteristic variables, the government′s innovation funding has a U-shaped nonlinear incentive effect on the output of enterprise′s innovation activities and enterprise′s innovation intention. It means that when the scale of government′s innovation funding exceeds a certain critical value, can it show the "crowding in" of enterprise′s innovation intention effect, enterprises are willing to increase the proportion of R&D expenditure in revenue. Otherwise, it will lead to "crowding out" effect on innovation intention of enterprises, and enterprises will reduce the proportion of R&D expenditure in revenue. (2) The government innovation funding has a significant "Inverted U" effect on the innovation preference of enterprises, which means, while government subsidies within a reasonable range can encourage enterprises to choose "higher level" innovation, excessive subsidies can easily encourage enterprises to "Rent-seeking", grab preferential policies and other "inert" behaviors, resulting in the "degradation" of innovation preference to the lower-level. (3) The above results are heterogeneous between state-owned / non-state-owned enterprises and high/low innovation intensity industries.
    The above results show that the government′s innovation subsidy policy should take into account the differences of different equity nature and innovation intensity among different industries. While the focus on alleviating the capital shortage of private enterprises and high innovation intensity enterprises, it should "spur" the innovation "inertia" of state-owned enterprises, and encourage the innovation enthusiasm of private enterprises in high innovation intensity industries at the same time, otherwise, it will further aggravate the "incentive distortion" among different type enterprises. Besides, local governments should pay more attention to stimulating the basic role of market fair competition mechanism in promoting the independent innovation ability of micro-enterprises. The intervention behavior should play the role of "icing on the cake" rather than "overstepping". Otherwise, it will aggravate the twisted allocation of enterprise innovation resources, promote the innovation inertia of enterprises, and is not conducive to the technological progress and development of the whole industry and economy high-quality transformation and development.