The real effects of income tax preference have called great attention in both practical and academic world（Liu, et al., 2016）. Some studies find that tax preference could attract capital inflow and improve corporate donation, investment, financing, as well as R&D investments（Jennings, et al., 2012; Chen, et al., 2018）. However, some other studies find that tax preference could also restrain enterprise value increase（Zhang, et al., 2012; Zhang, et al., 2015）. Besides these consequences on the corporate level, the social effect of tax preference is directly connected with social welfare. Therefore, this paper examines the influence of tax policy from the perspective of corporate income tax preference allocation.

Our paper discusses whether labor could benefit from corporate income tax preference using non-executive employee salary. And we further investigate how ownership and the source of tax preference affect the process of allocation. The empirical results show that corporate income tax preference increases non-executive employees’ salary and this effect exists only in state-owned enterprises. Relation-based tax preference would even decrease employees’ salary in non-state-owned companies mainly because managers with political connection might use the connection as bargaining power in the compensation negotiation with non-executive employees. We conduct instrumental variable analysis（2SLS）and construct a one-year lag in dependent variable to deal with the potential endogeneity problem, and our main results remain the same. We further conduct cross-sectional tests to investigate what factors affect tax preference allocation by dividing samples based on manager shareholding, provinces’ GDP per capita, local market competition, and the intensity of tax collection practice. Our results show that non-executive employees benefit less from tax preference in companies with more manager shareholding, higher GDP per capita, fiercer competition, and less intense tax collection practice. In all, our cross-sectional examinations show that managers’ bargaining power play a key role in tax collection allocation.

This paper has the following possible contributions. First, our results prove that the theory of corporate income tax incidence（Harberger, 1962; Feldstein, 1974）is also applicable in China, and we show that ownership and the source of tax preference could affect this process. Second, this paper finds that non-state-owned companies’ managers with political background are less likely to increase non-executive employees’ salary when tax preference is relation-based, which enriches the research on the influencing factors of employee compensation. Third, the results show that reducing tax burden for enterprises could improve the quality of employment, which is in line with the requirements of " improving the institutional mechanism of distribution according to factors” proposed by the 19th National Congress of CPC. However, the fact that ownership and the tax preference source have a significant influence in deciding non-executive employees’ salary is enlightening to policy-makers.