Determinants and consequences of non‐audit service fees

Purpose - This paper aims to provide preliminary evidence on the determinants of non‐audit service provision and whether the provision of such services impairs auditors' independence in the context of Bangladesh. Design/methodology/approach - Both univariate and multivariate regression methodologies are used to test the hypotheses. Signaling theory is used to test whether auditors from a non‐litigious environment are concerned about signaling their quality to the market by constraining clients' earnings management behavior. Findings - Results show that larger firms, firms with more liquidity, firms having an audit firm with an international link as their auditor and multinational subsidiaries purchase more non‐audit services, while highly‐leveraged firms purchase less. With respect to non‐audit fees (NAF) causing auditors to sacrifice their independence, this study fails to find any relationship between NAF and discretionary accruals (DACCR) after controlling for factors that are expected to drive DACCR behavior. Research limitations/implications - Use of DACCR as an earnings management device has been criticized on the ground that managers could use DACCR to signal their own private information. Hence, future research should look at alternative proxies of earnings management to investigate the relationship between non‐audit services and earnings management. Practical implications - The findings of this paper will be of use to financial reporting regulatory authorities in Bangladesh, such as the Securities and Exchange Commission of Bangladesh as well as the Institute of Chartered Accountants of Bangladesh, regarding the corporate governance role played by the auditors. Originality/value - Use of DACCR to detect earnings management behavior in the context of Bangladesh is a novel idea. The findings that the purchase of non‐audit services does not lead to increased earnings management behavior in a developing country context further strengthens the notion that there is a spill‐over effect of providing non‐audit services.