Avoiding cash taxes can serve as a significant source of additional cash flows for firms, though how managers utilize these funds and the resulting consequences remain open empirical questions. We provide answers by examining the association between the amount of cash tax savings and two uses of cash -investment and dividend payout -for an international sample of firms. We find that firms are more likely to invest cash tax savings rather than distribute them in the form of dividends and that this results in inefficient overinvestment. We find that our results hold for an international sample of domestic-only firms, distinguishing our study from US-only studies, which focus on constraints and distortions of multinational corporations in a worldwide tax system. We find nuanced results when partitioning on country-level governance. Our results suggest cash tax avoidance has real effects on firm decisions, namely investment and payout policies, and this effect varies based on the country in which the firm operates.JEL classifications: M41, H20, G31, G30
Investment:Invest The sum of research and development expenditures, capital expenditures and acquisitions less dispositions of property plant and equipment, deflated by total assets in t-1 (Biddle et al. 2009). absInvestEff The absolute value of the error term from a regression of Invest on the percentage change in sales (Biddle et al. 2009).
CapExCapital expenditures less dispositions of property plant and equipment, deflated by total assets in t-1.
Non-CapExThe sum of research and development expenditures and acquisitions, deflated by total assets in t-1.
IntangiblesIntangibles deflated by total assets in t-1.
DivPay:CashDiv Cash dividends paid scaled by total assets in year t.
PayoutDividends less gain/loss on the sale of stock scaled by lagged total assets.
Cash Tax Savings: CTSThe country's statutory tax rate less the cash effective tax rate (computed as cash taxes paid divided by pre-tax income). CTS is set to missing if pretax income is negative.
Governance: LowGovIndicator variable set equal to one if the firm is located in a country in the lowest tercile of Bushman, Piotroski and Smith (2004) governance index.
Control Variables:CloseShares Closely-held shares at t divided by total shares outstanding at t.
AQAccrual quality, based on Dechow and Dichev (2002) and Biddle et al. (2009), computed as -1 times the standard deviation of the residual (from t-5 to t-1) obtained from regressing a change in working capital on cash from operations in t+1, cash from operations in t, cash from operations in t-1, change in revenue, and property plant and equipment, with all variables scaled by total assets in t-1. The model is estimated by country, industry and year, provided the country-industry-year has 10 or more observations. σCFO Standard deviation, computed from year t-5 to year t-1, of cash flow from operations scaled by total assets in t-1. Does investment efficiency improve after the disclosure of material weaknesses in internal control over financial reporting? Jou...