We revisit the estimation of taxable income responses using tax reforms, and show that the state-of-the-art estimation strategy is in essence a Triple Difference Estimation (DiDiD). This insight makes the identifying assumption clear: Absent tax reforms the difference in income trends across the income distribution would have remained constant.
We show how this assumption of constant trend differentials can be validated both graphically and econometrically by inspecting income dynamics across the income distribution between periods with and without tax reforms, and we see this new approach as an essential tool for evaluating the credibility of elasticity estimates.
We illustrate our new approach using tax and benefit reforms in Denmark from which three key findings emerge. First, we find clear evidence of positive taxable income responses to tax reductions. Second, income shifting is an important driver for the observed effect on taxable income. Third, there is substantial heterogeneity in elasticities across the income distribution and across times.