Discretionary Classification of R&D Expense

U.S. generally accepted accounting principles allow managers discretion in deciding what period costs reported on the income statement are classified as R&D expense. Our study investigates the incentives that contribute to and the market consequences of managers’ discretionary R&D expense classification. We develop an R&D expectation model that includes variables that proxy for firms’ budgeted and discretionary R&D spending, and use the residuals from the model to proxy for firms’ discretionary R&D expense classification. We predict and find that managers engage in more positive discretionary R&D classi fication, i.e., report more unexpected R&D expense, in order to obtain R&D tax credits. In addition, we find evidence consistent with our conjecture that managers engage in discretionary R&D classification to bias investors’ perceptions of firm performance and value. Additional analysis incorporating increases in tax enforcement and financial reporting oversight provides corroborating evidence supporting our conjecture that managers engage in discretionary R&D classification to meet tax and financial reporting objectives. Turning to our market reaction tests, we find the market’s negative reaction to missed earnings targets is mitigated when managers report more unexpected R&D expense as a result of classification shifting. Our research provides new insights into firms’ reported R&D expense and managers’ incentives to engage in expense classification shifting.

This seminar is organised by the Erasmus Accounting Research Group.