Answers are grounded in this paper and cite their sources.
Financial-statement studies of tax avoidance estimate taxable income by dividing tax expense by a statutory rate, usually the top rate for every firm. Under a progressive schedule such as Korea’s, many firms lie below the top bracket, yet the effect of this choice has rarely been examined. Using 30,209 firm-years of Korean non-financial listed firms over 2011–2024 (main regression sample: 16,772), this study computes book-tax-difference (BTD) measures with the top rate, the bracket rate implied by prior-year income and the inverted schedule, and compares coefficients in otherwise identical regressions. In all fifteen models with donations, advertising, research and development, capital intensity and leverage as outcomes, the top-rate coefficients are larger in absolute value, significantly so in twelve (nine when standardised), and, where the bracket-rate coefficient is significant, 1.66 times as large per standard deviation at the median. The rate alone changes estimated effects. The difference arises mainly from firms without positive prior-year income: elsewhere it is much smaller, and rates based on loss firms’ past income or loss carryforwards give estimates close to bracket-rate estimates. Researchers using BTD under progressive schedules should state the rate used, re-estimate with such rates for loss firms and avoid comparing coefficient magnitudes across conventions.
No takes yet. Share an insight, caveat, or question.
Kwon Geejung (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: