Differentiated tax regimes assume that a class of firms can be identified in advance and that membership predicts the constraint the policy is meant to relieve. This paper examines that premise using the accounting effective tax rate, defined as income-tax expense over pre-tax accounting profit, which measures the charge recognised in the accounts rather than taxes paid. The sample comprises 67,899 firm-years of Italian certified innovative start-ups, innovative SMEs and ordinary small firms in 2014–2025, restricted to pre-tax profit of at least 1% of assets and rates of 0–100%. One equation is estimated with panel estimators, a partition and thirteen learners. Capitalisation and public contributions rank highest in permutation importance on held-out data (39% and 31% of summed permutation loss, a predictive ranking, not policy importance); median rates are about four points lower among certified firms than among ordinary SMEs, a gap that widens once size, region and profitability are held fixed. Partitioning half the firms on baseline accounts, the other half reject a common coefficient vector in fifteen of twenty-one Holm-corrected comparisons. Tax positions vary widely within registers, so fiscal measures need evaluation with firm-level tax, loss and take-up data. Results cover profitable firms; no policy counterfactual is estimated.
No takes yet. Share an insight, caveat, or question.
Leogrande et al. (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: