An examination of the impact of the changes to regulations affecting the scope for income classification shifting in Australia

Abstract

This study examines the impact of amendments to Australian accounting standards governing the classification and reporting of non-recurring components of earnings on firms’ classification shifting practices, and on the propensity to engage in other earnings management such as accrual-based and real earnings manipulation to improve core earnings. The study investigates the impact of three regulatory changes which have the potential to affect firms’ earnings management behaviour and thus the quality of earnings. I first investigate whether the amendments to the standard AASB 1018 Profit and Loss Accounts, effective for financial periods ending on or after 30 June 2002 and which restricted the scope for reporting and classifying abnormal items on the income statement, reduced the incidence of classification shifting. I then investigate whether the implementation of AASB 101 Presentation of Financial Statements, effective for financial periods starting on or after 1 January 2005 (upon IFRS adoption), and which relaxed prior restrictions on classifying and reporting non-recurring items, induced greater classification shifting. I also examine whether AASB 5 Noncurrent Assets Held for Sale and Discontinued Operations (effective at the same time as AASB 101), which increased the scope for classifying and reporting discontinued operations on the income statement, resulted in firms opportunistically misclassifying these items to increase core earnings. I then investigate the impact of each of the three regulatory reforms on firms’ use of alternative earnings manipulation methods, because there is evidence that a change in regulatory regime could impact firms’ alternative earnings management behaviour. For example, Cohen, Dey and Lys (2008) document an increase in the use of real earnings management to improve earnings following the reduction in accrual-based earnings management as a result of the Sarbanes-Oxley Act 2002. Therefore, there is potential for firms to substitute earnings management methods as a result of the changes to the scope for classification shifting. I expect that the restriction on classification shifting provided by AASB 1018 in 2002 may give rise to greater accrual-based and/or real earnings management by firms, whilst the greater scope for classification shifting provided by the later reforms may result in the reduction in use of accrual-based and/or real earnings management. To examine the first three research questions, I develop models of classification shifting from McVay (2006) and Barua, Lin and Sbaraglia (2010). I then use difference in differences designs to investigate the extent of firms’ classification shifting in the post-reform periods relative to the pre-reform periods. I also use a difference in differences design and develop models to test the impact of the regulations on firms’ alternative earnings management behaviour. My results show that there is no evidence of classification shifting following the 2002 amendment while there is indirect evidence of firms engaging in classification shifting using abnormal items, and strong evidence of classification shifting using discontinued operations following the 2005 amendments. However, tests of the impact of the reforms on alternative earnings management behaviour show that the restriction on classification shifting by the 2002 reform resulted in greater accrual-based manipulation to increase core earnings, while there is no such evidence of real earnings manipulation. There is evidence of a reduction in accrual-based manipulation and sales manipulation following the increased scope for classification shifting (post-2005) but there is no such evidence of real earnings management via discretionary expenditure reduction. Show more