Has IFRS Enhanced Accounting Uniformity?

ABSTRACT

I examine the effect of International Financial Reporting Standards (IFRS) adoption on the accounting uniformity of financial statements for a sample of large firms listed on Euronext. Using Taplin’s (A unified approach to the measurement of international accounting harmony*. Accounting and Business Research, 34(1), 57–73) uniformity index, I find that IFRS enhances uniformity of financial statements of firms within the same country (national uniformity) and between countries (international uniformity). The change in uniformity is not, however, homogeneous within and across jurisdictions that are subject to different accounting regulations before IFRS adoption. Those countries whose local GAAP was further from IFRS prior to adoption experience a greater increase in uniformity after IFRS adoption. I also find that international uniformity is increased most for items where IFRS eliminated divergence with local GAAP and for items where no regulation existed under local GAAP; when IFRS preserved the accounting choice set prescribed under local GAAP, uniformity does not increase. The main contribution relies on showing that the various forms of the relationship local GAAP-IFRS prior to the shift matter when examining the direct effect of IFRS adoption on the convergence of financial reporting practices.

Keywords:

• Accounting uniformity
• Local GAAP
• IFRS
• Euronext

Jel Classification:

• M40
• M41
• M4

Acknowledgements

I thank Jennifer Francis, Katherine Schipper, Per Olsson, Christopher Nobes, and two anonymous reviewers for their comments and valuable suggestions. I am also grateful to Carol Huget at Euronext Paris for providing Euronext data and to Ross Taplin to make available the Harmoniser software to calculate the T index.

Disclosure statement

No potential conflict of interest was reported by the author.

ORCID

Gianfranco Siciliano http://orcid.org/0000-0003-0840-7414

Notes

1 For ease of exposition, throughout this paper, I use ‘uniformity’ to refer to ‘accounting uniformity’.

2 As of 31 July 2010, 1,370 of the 1,584 Euronext securities (88%) are French, Belgian, Dutch, and Portuguese. I eliminate Portugal because of the paucity of firms from this country that survive my sample selection design detailed in section 3.

3 It should be noted that comparability does not require absolute uniformity of accounting for the same transactions and events. In this regard, the Conceptual Framework of the International Accounting Standards Board (IASB) does not include uniformity among the qualitative characteristics of financial reporting (‘Comparability is not uniformity’ – IASB Conceptual Framework 2018, p. 18).

4 In prior literature and in this paper, uniformity is based on the concept of ‘consistency’ (i.e., the use of the same accounting method for the same accounting item). The IASB Conceptual Framework states on par. 2.26 that ‘Comparability is the goal; consistency helps to achieve that goal.’

5 The T index generalizes and improves upon other harmonization indices used in international accounting literature (e.g. the H, C, I, and various modifications of these indices). For example, Taplin (2004) documents that the I index has undesirable properties when used to measure uniformity for more than two countries.

6 Previous studies that use a within- and between-country research design include Archer, Delvaille, and Mcleay (1995), Aisbitt (2001), Yip and Young (2012).

7 Defond et al. (2011) investigate uniformity changes at two distinct points in time around the IFRS mandate in 2005 (2003 and 2007). The authors use an industry-country level measure of uniformity that captures the change in the number of firms, within an industry, adopting the same set of accounting standards.

8 Leuz and Wysocki (2016), page. 529.

9 I hereby refer to the directives EC/2001/65 and EC/2003/51 that have modified the directives 78/660/EEC and the 83/349/EEC in order to reduce differences between local GAAP and IFRS. The European Commission (EC) required Member States to transpose these directives in their national jurisdictions before January 2004 and January 2005, respectively.

10 This pattern is also noted by Haller (2002), who documents that, before IFRS adoption, the Netherlands was among those European countries where local GAAPs complied the most with IFRS.

11 This appears to be anecdotally supported by the results from a recent EC’s survey where respondents believed that IFRS led to a significant increase in national uniformity and this change to be more pronounced in countries with the greatest differences between local GAAP and IFRS prior to IFRS adoption (European Commission, 2015).

12 Examples from the Nobes’ survey are: (1) tax-based depreciation is sometimes not corrected in consolidated statements; (2) it is possible not to include indirect production costs in the cost of inventories and contracts in progress; (3) extraordinary items are defined more widely.

13 I examine the following documents (2002 version): the Code de Commerce and the Plan Comptable Général in France; the Code des Sociétés and the Royal Decree 30 January 2001 in Belgium; the Civil Code and the Richtlijnen voor de Jaarverslaggeving in the Netherlands.

14 I acknowledge that my self-constructed measure of distance is not comprehensive, as it does not record all areas of accounting difference that a more detailed study would disclose: there might be other areas of financial reporting which are not included within my measure, but may be more significant for certain companies operating in certain industries or in other countries.

15 Ding et al. (2007) do not provide data with regard to the accounting items included for each country and category, but they display only a summary number for each category. Therefore, I am not able to determine how many items included in my measure overlap with the Ding et al.’s (2007) measure.

16 To ensure that a firm is included in the index at both periods, I cross check the list of firms on Euronext 100 and Next 150 at December 2002 and at December 2008 using monthly data obtained from Euronext Paris.

17 It is important to note that ruling out this option (i.e., equally-weight each country) essentially down weights the comparisons between companies from the two small countries of the sample: Belgium and the Netherlands for the computation of the international T index (see Taplin, 2010, p. 85). I thank one reviewer for pointing this out.

18 However, the reader must interpret the results with caution as the conclusions about national uniformity will be close to statements about national uniformity for France (given that the other two countries of the sample, Belgium and the Netherlands, are relatively small).

19 On the treatment of non-disclosure for the computation of harmonization indices, please see Aisbitt (2001).

20 In this case, the alpha matrix includes a dissimilar method (non-disclosure), thus causing the T index to decrease, especially in cases where non-disclosure is pronounced.

21 Due to this conservative approach, I can be confident that results reported using this design are statistically significant.

22 Note that in France and in Belgium I also find statistically significant decrease in uniformity for two and one accounting items, respectively. This supports the counter-argument that not always national uniformity increases after IFRS adoption.