The Distribution of Profit shifting and the Global Minimum Tax
28 August 2025
Sarah Clifford (University of Oxford), Jakob Miethe (LMU Munich), Camille Semelet (LMU Munich)
International tax avoidance by multinational enterprises (MNEs) is one of the most dominant issues on the tax policy agenda of many countries. On January 1, 2024, the Global Minimum Tax (GMT) came into force with the objective of creating a global tax floor of 15% on the corporate income tax of large internationally active firms. This reform represents one of the most profound changes to the global tax system in recent history and impacts hundreds of tax administrations and thousands of large MNEs worldwide.
To alleviate undue administrative and compliance burdens, the GMT is targeted only at multinational groups above a threshold of EUR 750 million in revenue. In a recent paper (Clifford et al., 2025), we study whether the threshold strikes the right balance between targeting profit shifting and keeping compliance costs low. We use administrative data from Germany with no reporting gaps for tax havens to construct group-level estimates of shifted profits and relate them to group size.
Profits shifted at the MNE group level
Our methodology for calculating group-level estimates of shifted profits relies on a fundamental comparison between the profitability of tax-haven and non-haven multinational affiliates. In contrast to studies that compare domestic firms and multinationals, we compare affiliates of similar size belonging to similarly sized groups, limiting the influence of underlying productivity differences unrelated to tax incentives. The central assumption is that, in the absence of profit shifting, affiliates of MNEs in tax havens would earn the same amount of profit as similar affiliates in non-haven countries.
Using this straightforward methodology, we document that German owned MNEs shifted a total of EUR 19 billion into tax havens in 2022. The pattern of profits shifted is exponential across the group-size distribution, with the Global Minimum Tax covering less than 30% of MNE groups but 95% of all profits shifted. This result is displayed in Figure 1, panel (a), which plots the average profits shifted per group within each decile of group-level turnover. The dashed vertical line shows the placement of the GMT threshold. Profits shifted is relatively stable for the first eight deciles of the total turnover distribution, however, in the last three deciles, we see an exponential gradient reaching EUR 10 million in the 9th decile and more than EUR 100 million per group in the top decile. These results show that the bulk of profit shifting is done by the very largest groups and that these are located above the GMT threshold.
Figure 1: Profit shifting across group size
(a) Average Profits Shifted
(b) Shifting Aggressiveness
Larger groups might shift more profits simply because they are large, or they might shift more profits because they are more aggressive than smaller groups. In Figure 1, panel (b) we scale the average profits shifted per group with the size of the group to get a measure of shifting aggressiveness. Panel (b) shows that both of these margins matter: Larger groups shift more profits than smaller groups, even relative to their size. However, note that this aggressiveness gradient is much flatter than that of profits shifted. On average, a group in the 10th decile shifts more than 10 times the profits of a group in the 9th decile. Relative to its employment, it only shifts twice as much. The bulk of the shifting gradient documented in Figure 1, panel (a) is therefore attributable to larger groups simply being much larger and consequently having more profits to potentially shift.
Potential tax revenue gain and compliance costs of the GMT
To be able to evaluate the merits of the GMT we use our results above to compare potential tax revenue gains from the policy to compliance costs across the group-size distribution. We start by computing, for each turnover decile, total profits shifted by all groups. We then assume a 15% GMT tax rate on this excess profit and plot the resulting tax revenue potential in Figure 2 below. The distribution mechanically mirrors the exponential figure seen in average profits shifted. If all excess profits of firms in the top decile had been taxed at the GMT rate in 2022, the total potential revenue gain would have been EUR 2.4 billion.
Figure 2: Aggregate tax revenue and compliance costs across group size
To compare the potential tax revenue gains to compliance costs of a global minimum tax, abstracting from costs of the tax administration, we use a survey conducted by Gaul et al. (2022). They surveyed German MNE groups above the GMT threshold to gauge expected compliance costs of the policy. The black dots in Figure 2 represent the total compliance costs of firms within the top three deciles calculated using the size categories reported in the survey directly. The dashed line represents an alternative measure of compliance costs assuming a fixed cost per subsidiary and using the average compliance cost reported in the survey.
In the two highest deciles, potential revenue clearly dominates compliance costs regardless of the measure used. Some revenue potential exists below the threshold; we estimate a total of EUR 44 million potential taxes gained on low taxed excess profit if the next decile below the cutoff were included. But lowering the threshold to cover this decile would also generate EUR 27 million in costs for firms, according to our estimate. Potential welfare gains of moving the threshold and including more groups within scope therefore appear modest.
Policy Discussion
Overall, our results suggest that the GMT is well targeted and support policy consistency in the face of current unilateral challenges against it. With an objective of balancing firm compliance costs with government tax revenue for the marginal firm entering the scope of the GMT, the current threshold appears well placed. With an objective of only maximizing total tax revenue gain, the current threshold also appears sensible with a coverage of 95% of all profits shifted. Level playing field and fairness concerns remain but are mitigated by the observation that the MNEs above the GMT threshold are also the most aggressive profit shifters relative to their size.
References
Clifford, S., Miethe, J., and Semelet, C. (2025). The distribution of profit shifting. CBT Working Paper Series, 25/02.
Gaul, J., Klein, D., Mueller, J. M., Pfrang, A., Schulz, I., Spengel, C., Weck, S., Wickel, S., and Winter, S. (2022). Significant costs, limited benefits: A global minimum tax in Germany. ZEW policy brief No. 07.
