Abstract: We examine whether tax-motivated income shifting by U.S. multinational corporations affects information asymmetry. Using a new firm-year measure of income shifting and a two-stage least squares approach, we find income shifting is positively associated with four measures of information asymmetry. Cross-sectional tests reveal that the effect of income shifting on information asymmetry is more pronounced for firms with large differences between foreign and domestic earnings growth. Using SFAS 131 to improve identification and establish evidence consistent with a causal relation between income shifting and information asymmetry, we demonstrate that the negative impact of income shifting on information asymmetry is concentrated in firms that discontinue geographic earnings disclosures. Overall, our study provides evidence that significant consequences of information asymmetry are associated with tax-motivated income shifting.

"Real Consequences of Tax-Motivated Income Shifting: Evidence from Shadow Insurers" with Jaron Wilde and Ryan Wilson (Preparing for re-submission)

Abstract: Tax-motivated income shifting is an escalating concern among policymakers worldwide and a growing area of academic interest. However, we have a limited understanding of the real consequences of the practice. We study one set of consequences using a sample of firms engaged in shadow insurance, a complicated series of transactions designed to free liquid assets typically held to pay life insurance claims. Shadow insurance, which exceeded $354 billion in aggregate at the end of 2013, ostensibly offers relief from unnecessary capital restrictions. However, critics argue it is tax-motivated, resembles the activities leading up to the 2007–2009 financial crisis, and poses risks to the larger economy. Consistent with a tax motivation, we find that five out of every six dollars of shadow insurance are located in tax havens and shadow insurance exhibits economically significant tax savings. We also find that firms using shadow insurance have significantly higher credit default risk and hold considerably lower levels of liquid assets but make significantly greater payouts to shareholders and executives. Taken together, our findings highlight significant consequences to a growing and economically important tax-motivated practice.

"A Re-Examination of the Book-Tax Difference Pricing Anomaly" (Preparing for submission)

Abstract: This study re-examines the mispricing of book-tax differences (BTDs). I provide evidence using multifactor time-series asset pricing tests that investors can generate abnormal returns trading on temporary BTDs, consistent with prior evidence. However, the abnormal returns to a BTD trading strategy are subsumed by the cash flow-to-price (CFO/P) anomaly. Specifically, hedge-control portfolio tests indicate that trading on BTDs does not yield incremental returns relative to trading on CFO/P, whereas trading on CFO/P does generate incremental returns to trading on BTDs. Viewing the pricing of BTDs in isolation, prior research concludes that BTDs contain information for future earnings, accruals, and cash flows that is not fully reflected in stock prices. My results suggest these findings omit an important correlated variable, CFO/P, which jointly captures accrual and value-glamour mispricing. Overall, my evidence casts doubt on the usefulness of BTDs as a mispricing signal, suggests that the BTD pricing anomaly is a subset of the broader CFO/P anomaly, and tempers calls in the literature for added mandatory tax disclosures and book-to-taxable income reconciliations.

Assessing the Construct Validity of Accounting Proxies: 
Evidence from Returns-Based Measures of Conditional Conservatism" with Cristi Gleason, Bruce Johnson, and Sam Melessa (Preparing for submission)

Abstract: We propose and demonstrate a method for validating proxies for accounting constructs. The method consists of identifying observations influential in generating an empirical result of interest and determining whether those observations exhibit accounting attributes consistent with theory. We demonstrate the method by assessing the validity of the Basu (1997) differential timeliness (DT) coefficient and the Lawrence et al. (2017) modified DT coefficient. Results indicate influential observations exhibit some attributes predicted by conditional conservatism. However, controlling for economic news, influential observations are characterized by poorer operating performance and exhibit an accrual-cash flow relation inconsistent with conditional conservatism. For the Basu (1997) model, characteristics inconsistent with conservatism are better able to discriminate influential from non-influential observations than characteristics consistent with conservatism, whereas the opposite result holds for the Lawrence et al. (2017) model. Overall, the evidence favors the Lawrence et al. (2017) modification in capturing the theoretically predicted accounting attributes.

"Tax Function Influence: The Case of Power and Status" with Matt Ege and John Robinson

Abstract: Social hierarchy theory predicts that the power and status of an organizational function affect the influence of the function. We test this theory using the tax function and find that the rank of the title of the top tax executive is positively associated with tax function influence, as captured through multiple measures of tax planning outcomes. Using changes in the size of the c-suite as an exogenous shock to the relative power and status of the tax function compared to other functions, we find evidence of decreasing (increasing) influence when there are c-suite expansions (contractions). Using title changes within the tax function, we find evidence that influence increases (decreases) upon title rank increases (decreases). Overall, our results are consistent with social hierarchy theory and suggest that the influence of the tax function within the organization depends on the position the tax functions holds within the organizational hierarchy. Moreover, we find these effects in the absence of direct compensation incentives, which highlights power and status as a channel through which firms reduce their explicit tax burden.