How does mandatory IFRS adoption affect tax planning decision? Evidence from tax avoidance distributions

ABSTRACT

The extant literature has provided much evidence on the consequences of adopting the International Financial Reporting Standards (IFRS), but studies that pertain to taxes and the IFRS are rare. This study investigates whether IFRS adoption affects corporate tax avoidance and how the impact of IFRS varies with country-level institutions. We find that the effect of an IFRS mandate on corporate tax avoidance is conditional; that is, firms with a lower (higher) initial level of tax avoidance tend to be more (less) tax aggressive after IFRS adoption. In addition, this conditional impact is more pronounced in regimes with higher levels of investor protection. We also find that the strength of country-level tax enforcement might erode (bolster) the positive (negative) effects of IFRS adoption.

KEYWORDS:

• IFRS adoption
• tax avoidance distribution
• investor protection
• tax enforcement

JEL CODES:

• M42
• M48

Acknowledgements

The authors appreciate helpful comments and suggestions from Jason Xiao, In-Mu Haw, Charlie Sohn and participants at seminars at University of Macau and Macau University of Science and Technology.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 The IFRS Foundation, Use of IFRS Standards around the world: http:/go.ifrs.org/iasb-profiles.

2 Brüggemann et al. (2013) and Emmanuel et al. (2016) review studies on IFRS adoption.

3 Firms in 13 countries that have not adopted IFRS on a mandatory basis during the sample period (2003–2009) comprise the benchmark sample in the PSM-DID analysis. The results hold even if we include the tax avoidance variable in the matching process, as shown in Table 13.

4 The existing literature also provides mixed evidence on whether the adoption of IFRS has improved earnings comparability (e.g. Lang et al., 2010; DeFond et al., 2011; Fargher & Zhang, 2014). Moreover, if IFRS improves earnings quality, the increased informativeness of earnings for evaluation of management’s effort might reduce the need for the use of RPE (Sloan, 1993).

5 We manually collected these statutory rates from the official website of the Organization for Economic Co-operation and Development (https://data.oecd.org/), Ernst & Young Worldwide corporate tax guide and KPMG LLP online summary. This represents the basic combined central and sub-central (statutory) corporate income tax rate given by the adjusted central government rate plus the sub-central rate.

6 CTP = total tax expense (Item #01451) – deferred taxes (Item #04199) – changes of tax payable (Item #03063). If missing, it is replaced with Current taxes paid (Item #04150) in the Worldscope Database which represents the amount of income taxes paid as reported on the cash flow statement.

7 We exclude: (1) firms that have voluntarily used IFRS before they are adopted on a mandatory basis in their resident country, (2) firms that switched from IFRS to local standards or the U.S. GAAP, and (3) firms that switch between IFRS and the U.S. GAAP from time to time. The specific accounting standard code and the classification criteria to identify IFRS users in Worldscope are shown in Appendix B.

8 The dependent variable, TaxAvoid, is a long-run tax avoidance measure which is calculated by using 3 years of financial data. For example, TaxAvoid in 2008 captures the 3-year average tax avoidance level (from 2006 to 2008) after IFRS adoption.

9 The cut off points to categorize the groups are −0.021, 0.068 and 0.214 respectively.

10 In other words, we exclude firms that do not show up in the pre- or post-IFRS periods. As firms may show up once or twice in the pre-IFRS (or post-IFRS) period, the sample constructed here is unbalanced. In addition, we also use the firms that show up in all four years of our sample period to construct a balanced sample in a robustness check.

11 Our sample period covers two years before and after the transition to IFRS. Hence, if we only implement propensity score matching year by year, the firms in the control group may not be consistent. To deal with this concern, we use the financial data of 2004 to conduct the matching procedure to determine the paired firms to construct the PSM sample. Consequently, the above one-to-one nearest neighbor matching will result in the same number of firms as opposed to the firm-years in the treatment and control groups.

12 We also use a balanced PSM sample, i.e. to be included in the balanced PSM sample, a firm must show up in all the four years of our sample period.

13 As we have defined above, all the firms in the IFRS mandatory countries will be categorized based on the 25th, 50th and 75th percentiles of the mean tax avoidance level before IFRS adoption, and the main sample is split into 4 sub-groups in which the hypotheses are tested respectively.

14 We realize that the coefficients on Mandatory are significant in Columns (2) to (4), indicating the difference in TaxAvoid between the treatment and control groups prior to the adoption. This is because we do not control TaxAvoid in the matching process. We address this issue in Table 13 by considering TaxAvoid in the matching process.

15 Countries with ADRI value of 4 are included in “Low ADRI” group. We obtain consistent results if we use Revised Anti-Direct Rights from Djankov et al. (2008) to proxy for investor protection (untabulated).

16 Detailed description of GAAP differences of each IAS item is provided in Appendix C.

Additional information

Funding

This work is funded by the Multi-Year Research Grant of University of Macau [MYRG2018-00203-FBA] and Faculty Research Grants of Macau University of Science and Technology [FRG-20-009-MSB].