Purpose
The issue of tax evasion through fraudulent financial reporting committed by companies is a major concern facing the tax authority in Malaysia. As such, the purpose of this study is to propose a cohesive model of predicting tax evasion from the perspective of fraudulent financial reporting amongst small- and medium-sized enterprise (SME) taxpayers.
Design/methodology/approach
The proposed model for this study is designed to explore the extent of the relationships between the independent variables: family ownership firms, company size, presence of tax professionals, company’s duration in business and frequency of tax audits and the moderating variable, i.e. tax audit officers’ level of competence; and the dependent variable, i.e. tax evasion from the perspective of fraudulent financial reporting. This study is grounded on four theories: agency theory, political cost theory, economic deterrence theory and competency theory. Data will be gathered from actual audit cases resolved by the Inland Revenue Board of Malaysia. Ordinary least square regression analysis is proposed for the investigation.
Findings
This study anticipates that family ownership firms, company size, presence of tax professionals, company’s duration in business and frequency of tax audits could be associated with tax evasion amongst SMEs in Malaysia. This study further proposes that highly competent tax audit officers could mitigate the relationship between frequency of tax audits and tax evasion practices amongst SMEs in Malaysia.
Originality/value
This study should be able to provide a cohesive model of predicting tax evasion from the perspective of fraudulent financial reporting amongst SMEs in Malaysia. Research on fraudulent financial reporting amongst SMEs is very limited, especially involving the level of competence of the tax audit officers; therefore, this study should contribute to the tax evasion literature by providing a comprehensive model of predicting tax evasion through fraudulent financial reporting using a Malaysian tax setting.