Harald J. Amberger, Eva Eberhartinger, and Matthias Kasper

A growing body of research suggests that financial decisions with tax implications (tax decisions) may be subject to decision biases. For example, prior work shows that individuals overestimate average tax rates relative to marginal tax rates (Bartolome 1995), which can lead to suboptimal outcomes (Graham, Hanlon, Shevlin, and Shroff 2017). However, it is unclear whether decision makers evaluate tax-rate effects and tax-base effects differently. Moreover, the drivers of decision biases are largely unknown.

In this study, we examine whether and under which conditions decision makers exhibit a systematic bias when evaluating tax-rate information and tax-base information. In particular, we investigate the effects of resource constraints, variation in the financial consequences of the tax-base effect, and professional experience. Our main hypothesis is that resource constraints induce a tax-rate bias in tax decisions because the economic effects of tax rates are more intuitive and simpler to identify than the economic effects of the tax base. Moreover, we expect that a large tax-base effect mitigates this bias, because decision makers are less likely to ignore information that has a meaningful effect on the outcome of a decision, a phenomenon termed “rational inattention” (Sims 2003; Mackowiak and Wiederholt 2009; Gabaix 2014). Finally, we hypothesize that professional experience reduces biased decision making.

We test these hypotheses using four laboratory experiments with 303 students and 62 tax professionals where participants must choose the form of financing that minimizes the tax burden of a firm. We model tax-rate effects through variation in statutory tax rates and tax-base effects through variation in the tax loss carryforward. To account for resource constraints, we include a time-constraints treatment in which we limit the time per decision. We test for rational inattention by introducing variation in the size of the tax-base effect. Finally, we investigate the effect of professional experience by comparing the responses from the student sample to the responses from the tax-professional sample.

Our results reveal a systematic tax-rate bias in decisions made under time constraints. We find that decision makers overestimate the relevance of less complex tax-rate information compared to more complex tax-base information, leading to suboptimal tax decisions. We also find that decision making, on average, is unaffected by professional experience. However, senior tax professionals show more awareness of the magnitude of the tax-base effect. These decision makers allocate their inattention more rationally and are thus less likely to exhibit a tax-rate bias when exhibiting such bias is relatively costly. To complement our experiments, we conduct interviews with eight senior tax executives. The interviews indicate potential for tax-rate biases in real-world tax decisions. Interviewees also discuss several strategies to reduce the risk of biased decision making, providing directions for future research.

Our results have implications for tax policy, real-world tax decisions, and future research. First, our findings suggest that decision makers might insufficiently adapt their tax strategies to tax-rate cutting and tax-base broadening reforms. This could explain why corporate tax revenues in the OECD remain constant despite falling statutory corporate tax rates. Second, our findings indicate that overestimating the importance of simple (tax-rate) information has the potential to cause suboptimal tax decisions. This finding provides a new, behavioral perspective for the observation that many firms appear not to minimize their tax burden (Dyreng, Hanlon, and Maydew 2008). Finally, our results suggest that even experienced decision makers could overestimate tax-rate effects. Thus, future research could, for instance, examine to what extent tax-base effects are reflected in firms’ income-shifting strategies.

Brooke D. Beyer, Jimmy F. Downes, Mollie E. Mathis, and Eric T. Rapley

This study investigates whether key aspects of the Tax Cuts and Jobs Act of 2017 (TCJA) affect U.S. multinational corporations’ (MNCs) domestic spending and investment behavior. It also examines whether U.S. MNCs change their foreign capital investment behavior due to the newly enacted global intangible low-taxed income (GILTI) inclusion regime.

We examine changes in firm spending and investment behavior by looking at the difference in shareholder payouts (i.e., dividends and repurchases), debt reductions, and capital expenditures following the TCJA. Given that the TCJA resulted in many broad changes to the corporate tax landscape, our sample includes U.S. MNCs and U.S. domestic-only firms to use a difference-in-differences design to isolate the changes related to the taxation of U.S. MNCs’ foreign earnings. This full sample consists of 3,681 firm-year observations for 2010 to 2019. We find that dividends increase post-TCJA for U.S. MNCs with low domestic liquidity and low domestic investment opportunities, which suggests U.S. MNCs pay dividends when there are limited investment opportunities and previous domestic financial constraints. We also find that share repurchases and domestic capital expenditures increase post-TCJA for U.S. MNCs with low domestic liquidity and high domestic investment opportunities, consistent with these U.S. MNCs lacking the prerequisite domestic resources in the pre-TCJA period. We find that U.S. MNCs with a high cost of debt and low domestic investment opportunities decrease their outstanding debt post-TCJA.

To examine the consequences of the GILTI regime, we hand-collect required segment disclosures of net property, plant, and equipment for domestic and foreign operations and partition the sample based on high and low potential GILTI inclusions. Our findings show that, for U.S. MNCs potentially subject to greater GILTI inclusions, higher levels of pre-TCJA foreign cash are associated with increased post-TCJA foreign property, plant, and equipment investments. However, we do not find a similar increase in domestic property, plant, and equipment. These findings suggest that while the TCJA changed investment opportunity sets for U.S. MNCs, confounding incentives created by the GILTI inclusion counteract domestic investment incentives.

We make at least two contributions to the literature. First, Clemons and Shevlin (2016) argue that policymakers only consider academic research when it is directly useful to those involved in the policymaking process, and the authors argue that the most effective way for researchers to affect tax policy is to discuss tax policy directly in research papers. This paper accomplishes both of those objectives. Second, this study builds on prior literature examining the spending and investment of internal capital resulting from a reduction in internal capital market frictions. Prior literature used the American Jobs Creation Act of 2004 as a setting of temporary repatriation cost reduction and documented that, on average, firms increased shareholder payouts and did not increase domestic investment (Blouin and Krull 2009; Dharmapala et al. 2011). Our findings suggest that post-TCJA investment depends on at least four factors: domestic liquidity, domestic investment opportunities, cost of debt, and whether the U.S. MNC is subject to the GILTI inclusion regime. Our results suggest that the permanent rather than temporary change in a U.S. parent access to its foreign subsidiary cash flow and the introduction of differing asset ownership incentives lead to nuanced results following the TCJA.

Candice T. Hux, Jean C. Bedard, and Tracy J. Noga

Research generally finds positive audit and tax outcomes associated with auditor-provided tax services (APTS), attributing knowledge sharing (KS) between tax and audit as the underlying cause. However, archival methods cannot observe how KS occurs in practice. Our study contributes to understanding of cross-service interactions by investigating experienced audit and tax professionals’ perceptions of when and how KS occurs throughout the APTS process, factors enhancing or impeding KS, and how KS can benefit both audit and tax services. Importantly, our design is comprehensive in capturing KS in both directions between auditors and tax professionals and including activities both before and after the inception of APTS.

Results from our interviews of 33 highly ranked audit and tax professionals from three Global 6 public accounting firms imply that KS occurs in two phases and is enhanced or inhibited by multiple factors related to knowledge relevance, motivation, opportunities, and culture. In Phase 1 (KS prior to APTS, often occurring during the audit), auditors and tax personnel share client information and their respective domain expertise, enabling both parties to gain a deeper understanding of the client and identify potentially valuable APTS (e.g., tax savings and solutions to tax problems). Collaboration on the audit can foster relationships between audit and tax professionals and with clients, motivating and enhancing opportunities for KS about client tax issues, and giving the firm a first-mover advantage for potential nonaudit tax work. Proactive KS about potential APTS is motivated by auditor and tax professionals’ trust and understanding of each other’s respective expertise and a culture that values other service lines, enabling “locked arms” between audit and tax in approaching clients with APTS opportunities. However, KS in this phase can be impeded by power struggles, “frenemy” relationships across service lines, tax specialists not being onsite, and competing incentives in managing the client relationship (e.g., auditors’ unwillingness to share and tax professionals’ “fishing expeditions” for nonaudit fees).

In Phase 2 (during/after APTS performance), further KS can occur. Tax professionals perceive benefits of acquiring knowledge from auditors, which can provide clients with a holistic view of financial reporting impacts of possible tax strategies considered, and/or book treatment of items in the tax return. As tax professionals perform APTS, they gain an understanding of how the client’s tax positions or strategies could affect the audit; sharing this APTS outcome knowledge with auditors can benefit the current or following year audit by identifying risky or judgmental areas for auditors to consider and/or potential or known misstatements. KS in this phase continues to be enhanced by the need for domain-specific expertise, strong relationships, an integrated firm culture, and proactive check-ins across service lines. However, some auditors indicate reasons as follows for less motivation and opportunities for KS in this phase: knowledge may have already been shared during the audit (i.e., relevant facts are already known from prior interactions), auditors may have rolled onto different clients so the goals of audit and tax services are no longer aligned, and different tax personnel may work on the APTS project relative to the audit engagement.

In summary, this study contributes to the APTS literature by providing a rich understanding of when and how KS occurs prior to, during, and/or after APTS performance. KS is not a discrete event after APTS but rather an extended process in which input and outcome knowledge are shared across service lines. We outline many future research opportunities and implications for practice based on our findings. We also provide an Online Appendix containing perceptions of clients’ APTS contracting considerations, supplementing our results on KS.

Yoojin Lee, Terry Shevlin, and Aruhn Venkat

In this study, we examine the effects of tax avoidance on firms’ capital structure decisions. Prior studies focus on the effects of tax avoidance on cost of equity and debt capital. These studies find different associations between tax avoidance and cost of equity capital but generally find a negative association with cost of debt (Goh, Lee, Lim, and Shevlin 2016; Hasan, Hoi, Wu, and Zhang 2014). We extend this literature by focusing on a direct and important consequence of cost of capital effects: capital structure decisions.

Tax avoidance may affect financing decisions in at least two ways. First, tax avoidance may affect the relative costs of equity and debt capital. Tax avoidance may increase future expected cash flows and thereby decrease cost of equity (Goh et al. 2016). However, tax avoidance may engender increased risk, which may increase both the cost of equity and debt (Hutchens and Rego 2013; Shevlin, Urcan, and Vasvari 2020). Second, tax avoidance may affect financing due to managerial biases. Survey evidence suggests that managers mistakenly focus on GAAP effective tax rates (ETRs) rather than marginal tax rates in making financing choices (Graham, Hanlon, Shevlin, and Shroff 2017). If so, managers will issue equity rather than debt as the firm increases tax avoidance because tax avoidance lowers GAAP ETRs, leading managers to (incorrectly) estimate a higher after-tax cost of debt and exhibit a bias toward equity.

Our main dependent variable captures the propensity to issue equity rather than debt. Our main independent variable is the three year cash effective tax rate. We use several covariates known to affect tax avoidance based on prior studies. Using logistic regressions, we document positive associations between tax avoidance and firms’ propensity to issue equity relative to debt after controlling for firms’ marginal tax rates. Using path analyses, we find that tax avoidance is (1) positively associated with cost of debt and (2) negatively associated with cost of debt. More importantly, we find that these effects explain the positive association between tax avoidance and the propensity to issue equity rather than debt.

Next, we use the Ninth Circuit’s Xilinx decision in 2010 to mitigate endogeneity concerns. The decision invalidated regulations intended to reduce income shifting, resulting in greater tax avoidance among Ninth Circuit firms. We document evidence that Ninth Circuit firms (1) increased tax avoidance following the decision relative to firms in other Circuits and (2) were more likely to issue equity rather than debt following the decision and relative to firms in other Circuits.

Finally, we turn to behavioral effects. We find evidence that one-year GAAP effective tax rates are positively associated with the propensity to issue equity relative to debt. We use an annual GAAP effective tax rate because survey evidence is consistent with managers using the annual GAAP ETR in making capital structure decisions (Graham et al. 2017). Next, we isolate managerial bias toward GAAP ETRs by controlling for permanent, discretionary tax avoidance and continue to find a positive association between GAAP ETRs and the propensity to issue equity relative to debt. Overall, our study contributes to the literature on tax avoidance and cost of capital effects by documenting the way and extent to which these effects affect capital structure, an important firm outcome. Moreover, we extend the literature on tax rates and corporate finance by documenting the financing consequences of managerial focus on GAAP ETRs instead of marginal tax rates.

Stevanie S. Neuman

A tax strategy is a pattern of decision-making and actions related to tax planning intended to maximize after-tax returns and firm value (Scholes et al. 2014). Most recent tax research examines the level of firms’ effective tax rates (ETRs), focusing on tax avoidance. However, theoretical work and research on book-tax tradeoffs and reputational costs indicate that some firms have tax planning goals besides minimizing cash taxes paid. Moreover, anecdotal evidence suggests that consistent tax outcomes are important (e.g., Tax Executives Institute (TEI) 2005; KPMG LLP 2007; Deloitte LLC 2013); therefore, the volatility of ETRs may be an alternative aspect of firms’ tax planning. In this study, I seek to establish maintaining low ETR volatility as a second, distinct approach to tax strategy and to test both tax strategies’ implications for firm value.

I find systematic differences in firm characteristics associated with each tax strategy approach and a predictable shift in characteristics when firms change tax strategies. In combination, these results identify at least two distinct approaches to tax strategy—low volatility of ETRs and low levels of ETRs. These results are important because they identify tax strategy approaches beyond the level of ETRs and offer an avenue for future research to further explore firms’ tax planning. Moreover, using a holdout sample analysis, I demonstrate that tax strategies can be predicted from firm characteristics, which could be useful for researchers seeking to identify tax strategies without using firms’ tax outcomes (e.g., because their research question seeks to examine ETRs). Finally, I find firms exhibiting low ETR volatility earn significantly higher median buy-and-hold returns than firms exhibiting low ETR levels. These results inform managers and practitioners of the importance of considering volatility when choosing a tax strategy. Furthermore, my results suggest that firms can derive valuation benefits from alternative approaches to tax strategy, which could explain why relatively few firms minimize taxes (Maydew 2001).

Adam J. Olson and Paul Ordyna

Accounting for income taxes (AFIT) is a complex area of accounting associated with many control weaknesses and restatements (Bedard and Graham 2011; Cheffers, Whalen, and Usvyatsky 2011; Audit Analytics 2016). Articles in the popular press suggest that CEOs and CFOs are held uniquely accountable for problems in AFIT, leading to executive turnover (Frieswick 2005; Murphy 2012; Bennecke 2018). A quick examination of CFO and CEO turnover around AFIT control weaknesses and restatements gives some credence to these popular press articles. To see whether these are simply anecdotes or actually some unstudied phenomena, we examine CFO and CEO turnover around AFIT control weaknesses and restatements. Specifically, we compare CFO and CEO turnover around AFIT control weaknesses and restatements to CFO and CEO turnover around non-AFIT control weaknesses and restatements.

We find no differences in CFO or CEO turnover around AFIT control weaknesses compared to non-AFIT control weaknesses. However, we find that CFO and CEO turnover is higher around AFIT restatements compared to non-AFIT restatements, consistent with popular press articles. This suggests that boards see AFIT restatements as different from other restatements.

To further explore this result, we examine auditor turnover and auditor provided tax services around AFIT problems. We find that auditor turnover and auditor provided tax services are no different around AFIT problems compared to non-AFIT problems. Further, we find weak evidence that CEOs are not uniquely held accountable for AFIT restatements if the firm was previously purchasing tax services from their auditor. Next, we examine executive turnover around problems in a number of other unique and complex accounts. We do not find unique executive turnover around these other accounts. Finally, we specifically examine observations we had previously removed, control weaknesses and restatements in both AFIT and other accounts, and observations with both a control weakness and restatement in the same year. Our examination of these samples supports our main results and suggests that CFOs and CEOs are uniquely held accountable for AFIT restatements compared to non-AFIT restatements.

This paper helps us understand that CFOs and CEOs (and possibly lower level employees) are uniquely held responsible for AFIT problems. Our results also show that boards of directors do not view all accounts within in a firm as identical.