Standard precision and aggressive financial reporting: the influence of incentive horizon

Abstract

The extant literature on precision in accounting standards suggests that financial statement preparers are less likely to make aggressive financial reporting decisions under less precise, principles-based accounting standards as compared to under more precise, rules-based accounting standards. We extend this line of research by examining how the incentive horizon of financial statement preparers influences earnings management behaviour. Consistent with prior literature, we find evidence that more precise standards lead to more income-increasing earnings management behaviour than do less precise standards when the incentive horizon is short-term in nature. However, when the incentive horizon is long-term, more precise standards are associated with financial reporting decisions that reduce current income relative to less precise standards. Importantly, the findings demonstrate that the effects of standard precision are changed by the incentive time horizon, and the effects of standard precision on financial decision makers cannot be fully understood when precision is studied without considering the timing of management incentive structures.

Keywords:

• incentive time horizon
• principles-based standards
• rules-based standards
• standard precision

JEL Classification:

• M41
• M48
• G38

Acknowledgement

We thank Ali Abdolmohammadi, Chris Agoglia, Jean Bedard, participants of workshops at Bentley University, University of Rhode Island and Fairfield University for their helpful comments. We also wish to thank the experienced professionals who gave their time to participate in this research.

Disclosure statement

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

Notes

1 It was important to include participants from countries that operate under different standard regimes to ensure that results are not driven by familiarity with a particular set of standards. We control for the potential effects of cultural difference by including a country indicator variable in our statistical models, and all results are insensitive to a participant's country.

2 Examples of both rules-based and principles-based standards exist within a single standard system and may co-exist within a single standard. For example, while IFRS is generally regarded as more principles-based, both IFRS 15 – Revenue from Contracts with Customers and IFRS 16 – Leases contain a number of specific rules.

3 For example, Imhoff and Thomas (1988) document a significant decline in capital leases and a corresponding increase in operating leases by companies that were previously capital-lease intensive following the release of Financial Accounting Standards Board (FASB)'s Statement of Financial Accounting Standard (SFAS) No. 13 Accounting for Leases. This standard required capital leases to be treated as assets and debt, which moved the treatment of such transactions from the footnotes to the balance sheet.

4 The term ‘economic substance of the transaction’ is used frequently in the literature on standard precision but is rarely defined. This term borrows from tax law, where a transaction's economic substance is measured by the change in firm financial position (other than tax effects) due to a specific transaction (Schreiber 2014).

5 The original participant pool contained 146 financial statement preparers. As an attention check, participants were asked to identify both the accounting standard and the incentive structure that were presented in the case facts. Ninety-two percent of the participants were able to do so. Eleven participants who were unable to do so were removed from further analysis. 

6 The survey asked the participants to indicate their current positions in an open-ended format. Thirty-five responses were coded as heads of finance, which included the following: Director of Finance, Head of Accounting, Head of Finance, VP of Finance or Accounting, Senior VP of Finance or Accounting, Executive VP of Finance or Accounting.

7 With the exception of IFRS 9 (which is presently being considered for adoption in Bangladesh), all Bangladesh Accounting Standards (BAS) have adopted all International Accounting Standards and International Financial Reporting Standards as of 2015.

8 Using participants from two different countries introduces a number of considerations, including cultural differences between the two countries, institutional setting, strength of the regulatory regime, etc. To address these issues, we included a country control variable in our analysis. None of our results are affected by the participants’ countries of origin.

9 With respect to the demographics of our participants (e.g., gender, number of years work experience) there were no significant differences between conditions and, when included in our analyses, the demographic variables were neither significant nor altered the conclusions that we draw in the study.

10 This study received proper approval from an institutional review board prior to data collection.

11 We also manipulated the psychological construct of future self-continuity for the financial statement preparers using a prime developed by Hershfield et al. (2012) as a second method of lengthening participants’ incentive horizon. We anticipated a potential interactive effect of future-self continuity and standard precision on a financial statement preparer's decision to report aggressively. Preliminary analyses revealed there is no significant main or interactive effects of self-continuity, and the self-continuity measure is not considered further and collapsed within our analyses.

12 ASC 810–10 – Consolidation uses a threshold of 50 percent to determine control of an entity, whereas IFRS 10 – Consolidated Financial Statements uses principles-based terminology referencing the effective power of the parent company.

13 Folsom, Hribar, Mergenthaler and Peterson (2016) analyse U.S. accounting standards to determine the extent to which each standard includes principles. They find that U.S. accounting standards vary in precision from heavily rules-based (e.g. ASC 815 – Derivatives, ASC 715 – Compensation – Retirement Benefits, and ASC 410 – Asset Retirement and Environmental Obligations) to largely principles-based (e.g. ASC 330 – Inventory and ASC 606 – Revenues from Contracts with Customers).

14 The instrument also contained information about the Research & Development budget of the firm. This portion of the instrument was designed to detect any potential for preparers to switch between real and accruals-based earnings management under our experimental conditions. Detailed analyses revealed no evidence of switching between forms of earnings management, and there were no statistically significant effects for the alternative dependent variable that captured attempts at real earnings management. Therefore, these results are not tabulated. On average, participants in all treatment conditions cut approximately $1 million from R&D, indicating that participants used R&D to reach earnings targets, and then they employed the lease classification decision to reach CEO bonus targets.

15 As supplemental analysis, we created a dichotomous variable to represent whether or not participants chose to classify the lease as operational (and thus engage in earnings management). We performed a binary logistic regression using the dichotomous measure as a dependent variable and STANDARD_PRECISION, INCENTIVE_HORIZON, and the interaction term as independent variables. The interaction term is again significant (p=0.020), which provides further support for our hypothesis.

16 It is possible that a participant in the less precise standard condition may interpret the phrase ‘for the major part of’ in a manner that creates a different decision context relative to the more precise treatment condition. Under both treatment conditions, participants must use professional judgment to determine whether the renewal option represents a bargain, and thus whether it should be included in the lease term for the purposes of determining the appropriate lease classification using the provided standard. Within the more precise condition, once the decision regarding the renewal option has been made, the bright-line threshold of 75 percent must be applied to arrive at the appropriate lease classification. Thus, participants who make the same judgment regarding whether the renewal option represents a bargain should reach the same lease classification decision. Within the less precise condition, a participant's interpretation of what constitutes ‘for the major part of’ the lease term will influence their ultimate classification. For example, a participant that interprets ‘for the major part of’ to mean 60 percent of the lease term will arrive at a different classification than a participant who interprets the phrase to mean 90 percent, given the same set of case facts. Accordingly, participants who make the same judgment regarding whether the renewal option represents a bargain may not necessarily arrive at the same lease classification decision.

17 We also analysed a model where we examined whether country of origin interacts with either standard precision or the incentive horizon. There are no interactive effects, which again supports the argument that results are robust for preparers from different standard regimes. Further, we included the country dummy in moderated mediation models to determine whether country influenced the mediating effects of economic substance on the lease classification decision. The country dummy was also not statistically significant in any of these models.