This paper presents descriptive evidence on economic activity reported under DAC7, the European Union’s newly introduced third-party reporting regime for online marketplaces. Using novel administrative data from Germany, we study platform-mediated sales between businesses and consumers as well as among peers. We document the scale, composition, and distribution of online marketplace participation and revenue volume across platforms and transaction types. The analysis provides a comprehensive empirical characterization of activity reported under DAC7 and establishes a baseline for future research and policy evaluation.

This paper develops empirically implementable algorithms for optimal audit targeting with machine learning. We derive a sufficient statistic-based targeting algorithm that depends on three individualized causal effects: the immediate revenue recovered from an audit, the causal effect of an audit on long-run tax revenue, and the marginal administrative cost of an audit. We show that these effects can be estimated with a variety of machine learners including causal forests, LASSO, gradient boosted trees, and neural networks using the universe of Pakistani income tax returns, exploiting years in which audits were assigned completely at random. We implement our targeting algorithms in out-of-bag years, comparing them to the real-world policy when audits were targeted. We show that the real-world audit program in Pakistan lost almost 173,000 Rs ($1, 700) in net revenue per-audit, while our optimal policy generates 285,000 Rs ($2, 800) in expected net revenue per-audit.

Automatic wage tax withholding is widespread across developing countries and widely viewed as a key prerequisite for enforcing personal income taxes. To limit administrative burdens, withholding systems are often coupled with non-filing options for large taxpayer groups. Using South Africa as a testing ground and population-wide administrative data linking payroll withholding certificates to personal income tax returns, we show that withholding can raise effective tax burdens above legally intended levels due to systematic over-withholding. This effect is particularly pronounced at the lower end of the income distribution and among young workers, consistent with weaker labor market attachment and greater within-year income fluctuations. In 2019, detectable over-withholding amounts to ZAR1.37 billion (USD 85 million) in unclaimed refunds; affected individuals overpay ZAR2,934 (around USD 184) on average, and many face effective tax rates of 5–8% despite zero statutory liability. Overall, over-withholding weakens the effective progressivity of the personal income tax system.

This paper studies whether the timing of income taxation affects welfare when earnings fluctuate within the year. Standard tax systems assess liabilities annually, implicitly treating taxpayers with the same yearly income as equivalent, even if one earns smoothly while another faces sharp monthly swings. We develop a theoretical framework showing that, under a convex tax schedule, shifting from annual to monthly tax adjustment, holding total yearly tax constant, is improving when income is nondecreasing and reduces liquidity risk.

We compare two within-year regimes: Vickrey’s cumulative averaging rule and a new Monthly Compensated (MC) mechanism based on uniform rescaling of monthly tax liabilities. Using monthly data from the U.S. Survey of Income and Program Participation (SIPP), we simulate welfare effects. Income volatility is concentrated at the bottom and linked to employment transitions. The MC system yields substantial gains for low-income, high-volatility individuals across states, while Vickrey delivers smaller but positive gains.