To curb tax-motivated income shifting and limit tax competition, the G20/OECD Inclusive Framework on Base Erosion and Profit Shifting (BEPS) has pursued a global reform of the taxation of large multinational enterprises (MNEs), which 141 jurisdictions have endorsed. The agreement in principle includes a minimum tax of 15 percent for the largest MNEs (Pillar II).
Ongoing discussions generally focus on its implications for corporate income tax revenues. Less is known about the effect of a minimum tax on the overall tax rate paid by MNEs on foreign direct investment (FDI) income, which ultimately drives investment decisions. We contribute to filling this gap.
EXISTING METRICS OF CORPORATE INCOME TAX RATES
There are 2 broad classes of corporate income tax rates:
statutory tax rates (STRs), established by law,
effective tax rates (ETRs), indicating the tax rate at which profits are “actually” taxed.
ETRs are are best suited for studying the taxes paid on FDI for 2 related reasons. First, ETRs give a more accurate picture of corporate income taxation. Unlike STRs, they absorb credits, deductions, exemptions, and other tax breaks granted by governments to lighten corporate income taxation. Second, ETRs better reflect the tax aggressiveness of offshore financial centers (OFCs). The difference between STRs and ETRs is most pronounced in OFCs due to greater resort to fiscal incentives and preferential tax treatment (see figure 1).
ETRs can be either backward-looking or forward-looking. In our analysis, we use backward-looking ETRs (for reasons explained in the paper). We also disregard within-country variation in ETRs and focus on average ETRs.
Backward-looking (average) ETR in host country c is defined as:
A NEW METRIC: THE FDI-LEVEL EFFECTIVE TAX RATE
An extensive body of research shows that MNEs engage in large-scale tax avoidance and profit shifting. They artificially book profits generated in high-tax countries in low-tax countries, especially in OFCs. Therefore, profit shifting disconnects the location of economic activities and the location of profits, and the ETR ultimately paid by MNEs on the income generated in host country c does not align with the ETR reported in c.
To complement standard ETRs and offer greater insight into the taxes paid on FDI income, we introduce a more comprehensive notion that encompasses the entire income generated by FDI (including shifted profits): the FDI-level effective tax rate (FDI-level ETR). The FDI dimension implies a shift in the analytical focus from the foreign affiliate’s country of operations (host country c) to the underlying, value-creating FDI project itself:
In the absence of profit shifting, profits are fully reported where they are generated. The ETR and the FDI-level ETR are equivalent.
In the presence of profit shifting, however, some profits generated in c are moved across borders for tax saving purposes. The overall amount of taxes paid on the FDI income generated in c is smaller than the ETR reported in c.
The FDI-level ETR in country c depends not only on the ETR in the host country c, where production takes place and profits are made, but also on ETRs in OFCs where some profits are (potentially) transferred. The weights are given by bilateral profit shifting shares (denoted γ), i.e., by the share of profits shifted from host country c to each OFC h:
The first term in equation (1) refers to the ETR in host country c. The second term represents profit shifting gains, i.e., the taxes saved by MNEs on the income generated by their FDI in host country c due to profit shifting. The difference between ETRs and FDI-level ETRs widens as profit shifting shares and ETR differentials between host countries and OFCs increase.
We calibrate ETRs and bilateral profit shifting shares for 208 jurisdictions. Having a quasi-exhaustive sample is challenging, unique, but crucial. It not only enriches our understanding of corporate tax and profit shifting patterns, but it also allows us to better grasp the impact of a tax reform at the worldwide level.
Armed with these ETRs and bilateral profit shifting shares, we then construct FDI-level ETRs (see the paper for technicalities). Figure 1 displays the results. Key takeaways:
The average ETR faced by foreign affiliates of MNEs in non-OFCs stands at 17 percent, but ETRs differ markedly across groups.
The difference between ETRs and FDI-level ETRs lies between 2 and 3 pp. Profit shifting activities are thus sizable. They reduce the tax rate paid on FDI income by more than 13 percent.
Incorporating profit shifting dynamics is critical to assess the impact of Pillar II. FDI subject to income taxes below 15 percent is indeed significantly higher once profit shifting is accounted for.
FIGURE 1 - CORPORATE INCOME TAX RATES FACED BY MNEs ON FDI INCOME
Interpretation – FDI income generated by MNEs in developing countries is taxed at 19.6 percent on average, after accounting for profit shifting schemes of MNEs.
Notes – LAC: Latin America and the Caribbean. LDCs: least developed countries. OFCs: Offshore financial centers. FDI-weighted averages.
Source – Casella and Souillard (2022).
THE IMPACT OF PILLAR II ON TAXES PAID BY MNEs ON FDI INCOME
We now turn to the impact of a global minimum tax applied to the foreign affiliates of large MNEs. In doing that, we focus on FDI-level ETRs as the most comprehensive and realistic measure of the total tax liability faced by MNEs on their FDI income.
MECHANISMS
A global minimum tax exerts 2 types of effects on FDI-level ETRs of large MNEs:
An ETR effect (1st term in equation (1)): ETRs in host countries below the threshold increase (if they are below the minimum).
A profit shifting effect (2nd term in equation (1)):
some profits initially transferred to OFCs are no longer shifted,
profits still shifted toward OFCs are subject to higher taxation.
The 2 channels can be disentangled and quantified.
SCENARIOS
In the paper, we consider 2 scenarios:
In the 1st scenario, we assume that profit shifting of large MNEs vanishes after the reform. This scenario provides an upper bound for the impact of Pillar II on FDI-level ETRs.
In the 2nd scenario, we assume that profit shifting of large MNEs partially declines as a result of the reform. The share of profits that remains in OFCs after Pillar II then rests on empirical and modelling considerations. Here, we assume that profit shifting linearly decreases after the reform (see the paper for technicalities). Given the recent literature on the non-linearity of profit shifting, this exercise is likely to provide a conservative estimate of the increase in FDI-level ETRs post-Pillar II.
We also discuss in the paper the introduction of a substance-based carve-out, whose objective is to reduce the tax base to which the Pillar II top-up tax applies (see the paper for technicalities). We expect the carve-out to mitigate the impact of Pillar II on FDI-level ETRs, exclusively through the ETR channel.
MAIN RESULTS
The main results can be visualized in figure 2. Key takeaways:
The increase in FDI-level ETRs induced by Pillar II is estimated to be between 2 and 3 pp globally. This implies a growth relative to pre-Pillar II levels between 14 percent (2st scenario with carve-out) and 20 percent (1st scenario without carve-out).
The carve-out plays a minor role, notably for jurisdictions with relatively high ETRs such as developing countries.
Another important result not graphically showed here is that the effect of Pillar II on FDI-level ETRs mainly passes through the profit shifting channel. The profit shifting channel is particularly strong for developing economies, owing to the combination of high pre-Pillar II ETRs and greater exposure to profit shifting.
FIGURE 2 - IMPACT OF PILLAR II (15%) ON FDI-LEVEL EFFECTIVE TAX RATES OF MNEs
Notes – PS: profit shifting. LAC: Latin America and the Caribbean. LDCs: least developed countries. OFCs: Offshore financial centers. FDI-weighted averages.
Source – Casella and Souillard (2022).
FOR MORE DETAILS, DISCUSSIONS AND RESULTS 👇
