Big Data In Tax Enforcement And Corporate Earnings Management

With the application of emerging information technologies such as big data,the tools of tax enforcement have been continuously upgraded,which has led to a change in the model of tax enforcement.Digital Governance Theory emphasizes the important role of information technology and its operating system in the reform of government departments is not only reflected in the upgrading of management tools,but also in the the ability of collaborative governance from a cross-domain perspective,and the ability to optimize the allocation of economic resources for the whole society.Based on digital governance theory,we use Golden Tax-III as a quasi-natural experiment,and choose a Multi-Period Difference-in-Difference(DID)model to empirically test the impact of Big Data in Tax Enforcement on corporate earnings management.The research shows that:(1)The implementation of Golden Tax-III can significantly inhibit the accrual-based earnings management.(2)Golden Tax-III has a stronger restraint on the accrual-based earnings management in non-state-owned enterprises than in state-owned enterprises.Among state-owned enterprises,Golden Tax-III has a stronger restraint on the accrual-based earnings management in local state-owned enterprises than in central state-owned enterprises.Among non-state-owned enterprises,Golden Tax-III restrains the accrual-based earnings management of non-state-owned enterprises without political connections more strongly than non-state-owned enterprises with political connections.(3)Golden Tax-III has stronger restraint on the accrual-based earnings management of enterprises in the eastern region than in the western region in China.(4)Golden Tax-III has stronger restraint on the accrual-based earnings management and real earnings management,while significantly increased the use of classification shifting of earnings management.