Classification Shifting and Special Items: Evidence of Earnings Management or a Research Design Consequence?

Abstract

McVay (2006, 2008) report compelling evidence that firms reclassify ordinary expense items as special item in order to boost benchmark core earnings numbers. We show that another far more innocuous explanation exists for this evidence. The effects that McVay attributes to classification shifting activities can arise as an artifact of the approaches used in deriving expected core earnings values. This can happen because expected core earnings and expected change in core earnings values are determined, in part, by special item accruals. When we focus on either subsamples where this special item accrual effect on the expected core earnings values is a priori expected to be small or incorporate controls for it into the analysis, there is no reliable evidence of widespread classification shifting activities by managers.