Conference Agenda

This paper revisits the empirical evidence on multinational profit shifting using a unique confidential administrative dataset from the General Financial Directorate of the Czech Republic, which includes firm-level Country-by-Country Reporting from every international enterprise operating in Czechia. The analysis reveals a shift in the selection of profit-shifting destinations in recent years. Between 2016 and 2023, the share of global profits shifted into European profit centres decreased from 78.7% to 40.9%, while Caribbean destinations rose in importance. Additionally, our results affirm the emerging pattern in contemporary literature, in which micro-level datasets systematically yield more conservative estimates of the magnitude of profit shifting than previous macro-level studies. Lastly, the exceptional level of granularity in our data enables us to confirm the extreme nonlinearity of the semielastic relationship between booked corporate profits and effective tax rates, with the cubic specification providing the best fit.

Beneficial ownership (BO) registers make ultimate owners visible and raise the expected detection risk of intra-group payments routed through opaque ownership chains. Using confidential Czech firm-level country-by-country reporting and the staggered adoption of BO registers across jurisdictions, I estimate the effect of ownership transparency on within-multinational profit allocation. Heterogeneity-robust difference-in-differences estimates show that BO adoption modestly narrows downward profit misalignment: reported profits move closer to levels predicted by local employment, assets, and revenue. The effect concentrates among multinational-jurisdiction pairs with lower pre-reform exposure and fades within two years. The public-access split is null, consistent with a mechanism operating through expected authority detection rather than public exposure. Related-party transaction data corroborate a restructuring reading: BO reforms abroad bring extensive-margin entry of new firm-counterparty disclosure relationships, with little change within continuing pairs. The pattern suggests unilateral transparency reallocates rather than eliminates shifted profits, so durable benefits depend on the breadth of adoption.

This paper studies how profit shifting affects real investment by multinational corporations (MNCs). We model three main profit-shifting channels and show that two influence investment through the user cost of capital. For all channels, we identify conditions under which an MNC’s minimum tax rate affects investment in other affiliates. Using detailed micro-level data on foreign affiliates, we test these predictions with instrumental variable regressions and event study analyses. We find that incentives to shift profits to low-tax locations reflect in the user cost of capital in high-tax countries. A one percentage point increase in the local statutory tax rate reduces investment by about 0.55%, while the response to the MNC-specific minimum tax is considerably smaller. Beyond user costs, transfer pricing of intermediate goods also affects investment through the minimum tax rate. Overall, our findings inform evaluations of policies such as the Global Minimum Tax.

Tax authorities in developing countries face risks of corporate tax base erosion by MNEs but lack practical tools to identify high-risk firms. This paper develops a data-driven framework for detecting profit shifting based on administrative data. The approach combines persistent abnormal profitability relative to industry peers with anomalous use of established profit-shifting channels. Applying the framework to firm-level data from Kenya, Uganda, and South Africa, we find that only a small share of MNEs are consistently flagged as potential profit shifters, around 5% in Kenya and Uganda and 1.7% in South Africa, yet the associated fiscal risks are highly concentrated. In South Africa, 44 firms account for over 60% of an estimated ZAR 5.29 billion in potential foregone corporate income tax revenue, with similar patterns observed in Kenya and Uganda. The results show that profit shifting is driven by a narrow subset of firms and illustrate how administrative data can support risk-based audit.