(Mis)measurement of Income Shifting

We examine the reliability of financial statement data for measuring cross-border income shifting. Using a novel dataset of Austrian corporate tax returns matched with financial statement data, we estimate a tax semi-elasticity of taxable income reported on tax returns of-0.9. When we use financial statement profits instead of taxable income, the semi-elasticity decreases by more than 50%. This result suggests that financial statement profits understate the extent of income shifting observed in tax return data. As a result, studies that rely on financial statement data are likely to underestimate income shifting at the firm level. We offer practical guidance on how to correct for biases arising from the disparity between financial statement profits and taxable income.