In the background of the gradual economy recovery from the financial crisis, economic development in China is still in a period of superposition and transition, and the key to solving development problems lies in innovation. The tax burden of enterprises is an important factor affecting economic development. At the same time, the corporate tax burden became a hot topic at that time, and the corporate tax burden was an important factor affecting economic development. In 2017, when the supply-side structural reform deepened, China would further implement the policy of tax reduction and fee reduction. The higher tax burden reduces the expected income and raises the financing cost, which is especially bad for enterprises′ innovation. How to stimulate R&D investment through tax policies has also become a competitive activity practiced by the government.

In order to promote R&D investment, the Chinese government has issued a series of tax incentives. Among them, the pre-tax deduction policy on R&D expenses is the most impressive, and it has played an active guiding role in promoting scientific and technological innovation of enterprises, upgrading industrial structure, and building an innovative country. In 2006, the Ministry of Finance and the State Administration of Taxation(SAT) carried out the "Notice on Preferential Corporate Income Tax Policies on Enterprise Technological Innovation". Except for full deduction for businesses′ R&D expenses, it is allowed extra 50% deduction in the current year before the enterprise income tax. In 2008, with the reform of tax, the SAT has clarified the definition of R&D activities, but mainly for high-tech enterprises, and then it expanded the accounting extent in 2013. In 2016, the range of benefit object was expanded to resident enterprises that have sound accounting, conduct auditing and collection, and can accurately collect R&D expenses. Since corporate income tax is the basis for the extra deduction of R&D expenses, it not only directly affects the retained earnings of reinvestment, but also affects the motivation and intensity of R&D investment. Therefore, it is particularly important to reveal the mechanism and effect of corporate tax incentives affecting R&D investment.

Considering the existing research, there are still some limitations. First, few studies have analyzed the mechanism of corporate tax incentives on R&D investment from the perspective of financial decision-making, such as capital costs, required return and owned capital. Therefore, it is necessary to conduct research on the financial decision-making path of corporate tax incentives on R&D investment. Second, the relevant research mainly focuses on the impact of corporate tax incentives on the "quantity" of R&D investment, but there is little discussion on the "quality" of innovation output, that is, whether tax incentive improves innovation output due to the increase of R&D investment or efficiency?

This paper analysis how tax incentives affect R&D investment through multiple paths. First, tax incentives can help reduce the capital cost of R&D investment, increase the company′s after-tax profits, and mobilize the enthusiasm of the company′s R&D investment. Second, tax incentives can increase or inject R&D investment resources, reduce the marginal cost and uncertainty of R&D efforts, and diversify R&D investment risks, which promote R&D investment. Third, tax incentives can convey a positive signal to external investors to attract external financing, which in turn helps companies obtain R&D investment resources. Finally, both expenses and capitalized expenditures in R&D investments can be regarded as non-debt tax shields, reducing corporate tax burdens. Therefore, policies such as extra deductions for businesses R&D expenses have strengthened incentives for tax avoidance.

The research expects to clarify the mechanism and effect of corporate taxation on R&D investment. First, by constructing an agency model of corporate shareholders and external investors, we reveal the transmission path of tax incentives on R&D investment in different scenarios of no tax, normal tax rate and super deduction policy, and further explore the mechanism on innovation output; Second, this paper use panel regression model to empirically analyze the effect of corporate tax incentives on R&D investment, identify effectiveness boundary conditions, and further discuss the intermediary path to innovation output.

Through constructing the agency model about corporate shareholders and external investors, this paper finds that tax incentives can lower the required rate of return and financing costs, and exhibit income effects increasing R&D investment and outputs, especially additional R&D deduction policy has dual effects of value guidance and joint incentive. Using the data of listed companies in China from 2007 to 2015, panel regression results show that: effective tax rates display a downward trend while R&D investments upward; the incentive of corporate tax on R&D investment has a boundary that is effective within the areas of low tax rates and high R&D intensity and less with others. The promoting effect of corporate tax incentives on output primarily results from R&D investment increasing. The conclusions provide references for the effectiveness of tax policies, increasing R&D investment and stimulating corporate innovation.