https://gabriel-zucman.eu/files/AJZ2019.pdf
https://gabriel-zucman.eu/files/GLRRZ2021.pdf
A lot of attention is paid to his inequality papers that build on Piketty and Saez, but some of his best work is on tax evasion. Whatever you think of his inequality work, I think his tax evasion work should be lauded.
Zucman linked the Panama Papers, HSBC Swiss Leaks, and tax amnesties to Scandinavian tax records to estimate tax evasion among high wealth persons. He also has another paper examining this in the context of IRS audits. In general a picture emerges of significant tax evasion among high wealth individuals. That shouldn't come as a surprise to anyone familiar with taxation.
Where his work builds on existing knowledge is it refines the admittedly noisy estimates from random audits. The IRS conducts random audits as part of the National Research Program and then uses Detection Controlled Estimation to estimate the size of tax evasion. DCE is an imperfect tool that leaves a lot to be desired. Zucman finds that it generally underestimates tax evasion among the highest income taxpayers.
Zucman uses operational audits (non-random) to show that generally random audits of high income earners underestimate the extent of tax evasion of this group. This deficiency is known by the IRS, and even with DCE adjustment is believed to be an underestimate. Zucman incorporates his work on offshore tax evasion to argue that this form of tax evasion is generally missed.
This all ties back to his inequality work as inequality metrics don't incorporate this tax evasion.