Digital transformation and corporate tax avoidance: An analysis based on multiple perspectives and mechanisms

Abstract

Promoting the seamless integration of the digital economy with the real economy, mitigating the adverse impacts of widespread corporate tax avoidance, and optimizing tax governance are critical imperatives in the era of digital economy. This study examines all A-share listed companies from 2007 to 2022 as research samples. It utilizes multiple perspectives including signal theory, information asymmetry theory, and the T-O-E (Technology-Organisation-Environment) framework to investigate the primary impacts of digital transformation on corporate tax avoidance, along with the intermediate mechanisms and foundational conditions that influence its effectiveness. After conducting both theoretical and empirical analyses, this paper presents the following conclusions. (1) The implementation of digital transformation significantly reduces corporate tax avoidance, a conclusion supported by rigorous robustness tests. Moreover, digital transformation enhances corporate productivity through the suppression of tax avoidance. (2) Digital transformation diminishes corporate tax avoidance through enhanced innovation and efficiency in resource allocation (technology level), improved quality of internal controls (organization level), and decreased industry competition (environment level). (3) The impact of digital transformation in reducing tax avoidance is significantly greater for enterprises in their growth phase, experiencing lower financing constraints, particularly those situated in the central and western regions. (4) Lower business risk is essential for maximizing the effectiveness of digital transformation and reducing corporate tax avoidance. This is crucial for governments seeking to improve tax administration, guide market and regional development, and enhance the impact of corporate digital transformation on mitigating tax avoidance.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 28 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.