At a virtual presentation on September 15, 2026, organized by the Institute of Public Finance, Hermes Morgavi (OECD, Paris, France) presented his article published in the journal Public Sector Economics where he proposes a new empirical model to estimate Laffer curves for personal income tax, corporate income tax, and value-added tax across OECD countries.

The model builds on the elasticity-of-taxable-income literature and introduces a machine-learning approach, using LASSO regularisation to let the curve's shape depend on each country's structural, institutional, and macroeconomic characteristics. 

The results reveal substantial heterogeneity in revenue-maximising tax rates across OECD countries: several countries already tax personal and corporate income beyond their estimated peak, leaving limited scope for additional revenue through higher rates. By contrast, most countries remain below their revenue-maximising rate for value-added tax, pointing to comparatively greater fiscal space in consumption taxation. 

Because the shape of the Laffer curve depends on institutional quality and economic structure, reforms that strengthen governance, reduce informality, broaden tax bases, and improve compliance could expand revenue capacity without raising statutory tax rates. 

The presentation was moderated by Anja Špoljarić, researcher at the Institute of Public Finance, and a recording of this interesting lecture is also available.