## wp18166

## Source details

**Canonical URL:** [wp18166](https://www.imf.org/-/media/files/publications/wp/2018/wp18166.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2018/wp18166.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2018/wp18166.pdf.json)

---

### Introduction: scope and types of spillovers
- Major U.S. tax reform adopted on December 20, 2017 affected corporate income, personal income, and estate taxes and has:
  - Macroeconomic spillovers: stronger U.S. growth boosts export demand in trading partners.
  - Tax-related spillovers: changes to U.S. tax rates and structure affect MNEs’ location, profit reporting, and headquarters decisions.
- The U.S. 2017 reform reduced the U.S. statutory corporate rate by 14 percentage points and shifted aspects of the system toward territoriality, with potential to intensify tax competition.
- The reform included novel provisions intended to address international profit shifting; some measures are more aggressive than BEPS minimum standards and recommendations.

### Background: the pre-reform U.S. CIT regime
- Pre-reform characteristics:
  - Federal statutory tax rate: 35 percent; combined with state taxes: average around 40 percent.
  - ‘Domestic production activities deduction’ reduced the federal rate to 32 percent for many lines of business.
  - Most qualifying dividends and long-term capital gains taxed at rates of up to 20 percent; other gains taxed at rates of up to 39.6 percent.
  - About 95 percent of U.S. businesses were organized as pass-through entities; C-corporations accounted for slightly less than half of net business income.
  - Generous depreciation allowances and ‘bonus depreciation’ (expensing of 50 percent of many assets’ value in the first year) produced a narrow tax base despite high statutory rates.
  - Worldwide taxation with deferral: foreign earnings not taxed until repatriated; foreign tax credit available and excess credits carry forward or offset other active foreign income.
  - Accumulation of undistributed foreign earnings estimated at $2.6 trillion by 2017.
- Pre-reform tax planning practices included transfer pricing, thin capitalization, capitalizing subsidiaries in low-tax jurisdictions, ‘check-the-box’ hybrid arrangements, and corporate inversions.

### Main features of the 2017 reform
- Rate cut:
  - Federal statutory tax rate fell from 35 to 21 percent (a 14-percentage point cut).
  - Combined rate including state taxes fell on average from 40 to 26 percent.
  - 26 percent is close to the 2018 OECD average of 24 percent.
- Accelerated depreciation (expensing):
  - Expensing for most capital investment introduced for 5 years.
  - Between 2023-2027, expensing is scheduled to be phased out, dropping by 20 percentage points per year.
  - Pre-reform 2017 “bonus depreciation” allowed expensing of 50 percent of many assets’ value; in 2018 this was scheduled to fall to 40 percent absent reform.
- International tax provisions (uncertainty until regulations finalized):
  - Global Intangible Low-Taxed Income (GILTI):
    - Creates a minimum U.S. tax on foreign active income without deferral; oil and gas sector exempt.
    - Minimum rate: 10.5 percent; tax base: foreign income exceeding 10 percent of foreign tangible assets.
    - Nonrefundable tax credit granted for 80 percent of foreign taxes.
    - Under current legislation, the minimum rate will increase to 13.125 from 2026, raising maximum creditable foreign tax to 16.4 percent.
    - Expense allocation rules (including interest allocated to foreign income) can make GILTI apply even when average foreign tax rates are higher.
  - Foreign Derived Intangible Income (FDII):
    - Preferential rate for certain foreign earnings of U.S.-resident firms: 13.125 percent.
    - Base defined as earnings in excess of a 10 percent return on tangible assets multiplied by the proportion of net income from export sales.
    - Rate scheduled to rise to 16.4 percent from 2026.
  - Base Erosion Anti-Abuse Tax (BEAT):
    - Alternative minimum tax adding back most foreign related-party service costs on a gross basis into the U.S. tax base.
    - Rate: 5 percent in 2018, rising to 10 percent in 2019 and to 12.5 percent from 2026.
    - BEAT does not provide credit for foreign taxes or take foreign tax rates into account.
- Interest deduction limits:
  - Interest deductions must not exceed 30 percent of earnings before interest, tax, depreciation, amortization, and depletion (EBITDA) plus business interest.
  - From 2022, limit applies to earnings before interest and tax (EBIT).
  - Regulated public utilities exempt; real estate and farming can opt out (then must use the Alternative Depreciation System).
- Anti-hybrid rules:
  - Disallow deductions for related-party hybrid transactions (not taxable in recipient country) or payments to an entity characterized as taxable in one jurisdiction and transparent in another.
- Transition (deemed repatriation):
  - Existing unrepatriated profits deemed repatriated and taxed at a one-time rate of 15.5 percent for funds held in cash or equivalent assets and 8 percent for all other assets.
  - Payment is spread over 8 years and is independent of actual repatriation.

### Impact of GILTI on average tax rates (authors’ calculations)
- GILTI could be more onerous than the previous worldwide system for some firms because it applies without deferral and provides a more limited foreign tax credit.
- Simplified industry-level estimates (authors’ estimates based on BEA data) of foreign CIT, GILTI, and combined average rates (in percent):
  - All industries: Foreign CIT 7.9, GILTI 2.9, Combined 10.8
  - Mining: Foreign CIT 80.8, GILTI 0.0, Combined 80.8
  - Manufacturing: Foreign CIT 11.8, GILTI 0.0, Combined 11.8
    - Food: Foreign CIT 21.9, GILTI 0.0, Combined 21.9
    - Chemicals: Foreign CIT 14.0, GILTI 0.0, Combined 14.0
    - Primary and fabricated metals: Foreign CIT 20.5, GILTI 0.0, Combined 20.5
    - Machinery: Foreign CIT 10.0, GILTI 1.9, Combined 11.9
    - Computers and electronic products: Foreign CIT 7.6, GILTI 3.1, Combined 10.7
    - Transportation equipment: Foreign CIT 15.0, GILTI 0.0, Combined 15.0
  - Wholesale trade: Foreign CIT 11.7, GILTI 0.5, Combined 12.2
  - Retail trade: Foreign CIT 17.7, GILTI 0.0, Combined 17.7
  - Information: Foreign CIT 8.5, GILTI 1.4, Combined 9.9
  - Finance and insurance: Foreign CIT 10.4, GILTI 1.9, Combined 12.3
  - Professional, scientific, and technical services: Foreign CIT 9.0, GILTI 2.6, Combined 11.6
  - Other industries: Foreign CIT 2.6, GILTI 7.9, Combined 10.5
- Calculation note: The GILTI liability (expressed as a ratio to corporate pre-tax income) is calculated for each industry as the maximum of zero and [(Foreign net income + foreign income tax – 0.1*PP&E)*0.105 – 0.8*foreign income tax]/(foreign income + foreign income tax).

### Effective tax rates: definitions and measurement (Appendix I)
- EATR definition used: EATR = (R* − R)/p(τ + r) [equations (11)–(12) define R*, R and parameters], where R* is the PDV of the economic rent in absence of taxation, R is PDV in presence of taxation, p is pre-tax net profit, r is the real interest rate, and δ is true economic depreciation.
- γ = (1 − m_d)/(1 − z) measures difference in treatment of new equity and distributions; ρ = (1 − m_i)/(1 − z)i is the investor’s discount rate. A is the PDV of depreciation allowances. Financial effect F applies for new equity (NE) or debt (D) finance (zero for retained earnings).
- EMTR is obtained by setting post-tax rent R = 0 and solving for the corresponding cost of capital.
- Backward-looking empirical measures (tax payments divided by profits) are less suitable for assessing incentives for new investments.

### Effects of the 2018 U.S. tax reform on EATR and EMTR (standard tax system, 2018)
- Calculations based on the tax system in place in 2018; no effective tax rates reported for later years due to uncertainty about scheduled changes.
- Equity-financed investment in plant and equipment (standard tax system, including average state taxes that raise the federal rate by 5 percentage points):
  - Table 2 (plant and machinery, equity-finance) — EATR and EMTR under alternative debt-limit binding cases:
    - EATR: 30.8, 19.5, 20.5, 12.8, 12.8, 12.3
    - EMTR: 5.2, 0.0, -173.7, -56.7, -16.9, -16.9
    - Assumptions: Asset: plant and machinery; Inflation: 2 percent; Real rate of interest: 5 percent; True depreciation: 5 percent; Profit rate (EATR): 20 percent.
  - Expensing drives EMTR to 0 for equity-financed marginal investments (under end-of-period timing assumption).
- Debt-financed investment:
  - Pre-reform: debt-financed investment faced lower effective tax rates (debt bias) and could produce negative EMTRs.
  - Post-reform: interest deductibility limit (30 percent of EBITDA; from 2022 measured against EBIT) changes outcomes:
    - If limit not binding, lower statutory rate reduces value of interest deductibility -> EMTR increases relative to pre-reform.
    - If marginally binding, deductibility of new interest depends on project profitability.
    - If binding, no additional interest deduction from issuing more debt; EATR can still fall because additional profit creates scope for interest deductibility.
  - Reform reduced—though did not eliminate—debt bias. Interest limit is less restrictive for the financial sector (netted against business interest received).

### Incorporating FDII, GILTI, and BEAT into effective tax rate calculations
- FDII:
  - Applies only to foreign sales of domestic firms; impact highly case specific and depends on firm-wide returns relative to 10 percent of tangible capital.
  - Table 3 (Effective Tax Rates under FDII, Equity Finance, 2018) — EMTR and EATR under four cases:
    - Return on K_{t-1} << 10%: EMTR 0.0; EATR 19.5 — "Standard tax system applies"
    - Return on K_{t-1} < 10%: EMTR -3.5; EATR 12.9 — "Expense against full tax rate"
    - Return on K_{t-1} >> 10%: EMTR 13.7; EATR 16.8 — "Expense against FDII tax rate"
    - Intangible investment: EMTR 0.0; EATR 13.6 — "Entire return subject to FDII rate"
    - Assumptions: Inflation: 2 percent; Real rate of interest: 5 percent; True depreciation: 5 percent; Profit rate (EATR): 20 percent.
  - From 2022 R&D will become depreciable over 5 years (15 years for foreign R&D) rather than being expensed, which would raise tax rates relative to 2018.
  - Overall: EATR is reduced by FDII under most reasonable assumptions; EMTR patterns vary by case.
- GILTI:
  - Move to territoriality reduces effective tax rates for many foreign investments previously subject to residual U.S. tax.
  - For firms covered by GILTI some investments’ effective tax rates may rise because immediate GILTI liability could be more onerous than deferred U.S. tax.
  - For additional real investments with relatively low rates of return, effective tax rate could be negative because increases in foreign tangible capital raise the GILTI tax-free amount.
  - Incorporation into formulas replaces FDII reduced rate by GILTI rate (0.105) and adjusts bases; foreign tax credits and different depreciation bases interact; GILTI applies globally.
- BEAT:
  - Primarily affects deductibility of certain related-party payments (services), bringing them into U.S. tax base; for a typical investment BEAT cannot be straightforwardly incorporated into the effective tax rate framework because the share of costs nondeductible under BEAT is unknown.

### Spillovers from the U.S. tax rate cut — channels and conceptual framework
- Three reaction margins considered in quantifying fiscal spillovers:
  - Profit shifting decisions of MNEs.
  - Real location decisions of MNEs (investment reallocation).
  - Policy decisions of other governments (tax rate reactions).
- Profit shifting:
  - MNEs exploit statutory tax differences via transfer pricing, intra-group loans, ownership of intangibles; profit-shifting entails costs.
  - The U.S. rate cut strengthens incentives to report profits in the United States and can lead to profit shifting out of a broad range of countries, including some low-tax jurisdictions.
- Location of real investment:
  - Lower U.S. CIT rate makes the U.S. more attractive; tax affects net return to capital so a rate cut should foster investment in the United States.
  - Channels affecting investment elsewhere include macroeconomic effects, capacity limits at MNEs, and anticipated profit shifting.
- Policy spillovers:
  - Other governments may react via reduction in their own CIT rates (resource-flow and yardstick models); empirical evidence shows countries react to some average of other countries’ rates.

### Conceptual model and parameterization
- Profit reporting and tax revenue approximations based on Hines and Rice (1994) and Huizinga and Laeven (2008).
- Parameter choices:
  - Semi-elasticity of capital w.r.t. tax rate (ε_K): set to 2.4.
  - Semi-elasticity of profits holding capital constant (ε_S): set to 1.5.
  - Instantaneous reaction coefficient (β) for other countries’ CIT response to average foreign rates: set to 0.6.
  - Country-specific capital intensity α_c estimated from ORBIS micro regressions; sample: slightly more than 1.8 million observations.
  - Country-specific micro-weights ω_c defined by presence of direct ownership link to the United States; robustness check uses bilateral FDI-based weights.
- Summary country-specific parameters (Table 4 exact values):
  - Capital intensity: Mean 0.54; Median 0.54; Min 0.23; Max 1
  - Micro-weight: Mean 0.03; Median 0.02; Min 0; Max 0.19
  - FDI-weight: Mean 0.11; Median 0.06; Min 0; Max 1

### Simulation results — revenue spillovers (57 countries)
- Using micro weights (Table 5; losses are percentage changes of MNE profits), no policy response (other countries maintain CIT rates):
  - Profit shifting: Mean 0.85; Median 0.7; Lowest percentile 0.22; Highest percentile 2.91
  - Investment response: Mean 0.73; Median 0.59; Lowest percentile 0.17; Highest percentile 3.11
  - Investment response (in percent of capital stock): Mean 1.36; Median 1.11; Lowest percentile 0.18; Highest percentile 4.66
  - Total change in MNE tax revenue: Mean 1.58; Median 1.34; Lowest percentile 0.39; Highest percentile 6.02
- Using micro weights, with policy response (other countries react):
  - Profit shifting: Mean 0.77; Median 0.66; Lowest percentile 0.22; Highest percentile 2.23
  - Investment response: Mean 0.65; Median 0.53; Lowest percentile 0.16; Highest percentile 2.4
  - Investment response (in percent of capital stock): Mean 1.23; Median 1.06; Lowest percentile 0.18; Highest percentile 3.6
  - Tax rate response: Mean 3.76; Median 2.89; Lowest percentile 1.24; Highest percentile 13.38
  - Tax rate response (in p.p. of tax rate): Mean 0.84; Median 0.69; Lowest percentile 0.1; Highest percentile 3.95
  - Total change in MNE tax revenue: Mean 5.17; Median 4.25; Lowest percentile 1.57; Highest percentile 17.14
- Using FDI weights (Table 6), no policy response:
  - Profit shifting: Mean 2.4; Median 1.69; Lowest percentile 0.24; Highest percentile 12.87
  - Investment response: Mean 2.14; Median 1.3; Lowest percentile 0.19; Highest percentile 12.08
  - Investment response (in percent of capital stock): Mean 3.84; Median 2.7; Lowest percentile 0.06; Highest percentile 20.59
  - Total change in MNE tax revenue: Mean 4.54; Median 2.76; Lowest percentile 0.47; Highest percentile 24.2
- Using FDI weights, with policy response:
  - Profit shifting: Mean 1.48; Median 1.28; Lowest percentile 0.21; Highest percentile 3.4
  - Investment response: Mean 1.31; Median 0.89; Lowest percentile 0.16; Highest percentile 3.74
  - Investment response (in percent of capital stock): Mean 2.49; Median 2.08; Lowest percentile 0; Highest percentile 6.26
  - Tax rate response: Mean 10.69; Median 7.1; Lowest percentile 1.31; Highest percentile 53.63
  - Tax rate response (in p.p. of tax rate): Mean 2.12; Median 1.49; Lowest percentile 0; Highest percentile 18.5
  - Total change in MNE tax revenue: Mean 13.48; Median 9.58; Lowest percentile 1.77; Highest percentile 57.51
- Key summary findings:
  - When other countries do not change CIT rates, corporate tax revenue from MNEs decreases by 1.58 percent on average across countries (micro weights).
  - Changed profit-shifting incentives account for 0.85 percent average contraction; investment response 0.73 percent average contraction (micro weights).
  - Incorporating policy response increases average revenue shortfalls to 5.17 percent (micro weights), as tax rate responses amplify revenue losses.
  - Using FDI weights yields larger spillovers: average losses around 4.54 percent with no policy response and 13.48 percent with policy response.
  - Depending on parameter assumptions, reform leads to average revenue losses between 1.5 and 13.5 percent of the MNE tax base.
- Sensitivity and uncertainty:
  - Large uncertainty due to wide ranges of elasticities; parameter β (policy reaction) is particularly influential.
  - Table 8 (estimated tax reactions by coefficient, GDP weighted, full sample 182 economies):
    - β 0.3: Minimum 1.3; Mean 1.4; Maximum 1.5
    - β 0.4: Minimum 2.0; Mean 2.0; Maximum 2.2
    - β 0.5: Minimum 2.8; Mean 2.8; Maximum 3.0
    - β 0.6: Minimum 3.8; Mean 3.8; Maximum 4.1
    - β 0.7: Minimum 5.2; Mean 5.2; Maximum 5.4
    - β 0.8: Minimum 7.0; Mean 7.1; Maximum 7.3
    - β 0.9: Minimum 9.7; Mean 9.7; Maximum 9.9

### Spillovers from other reform features and policy implications
- FDII, GILTI, and BEAT likely increase profit reporting in the United States beyond what is implied by the statutory rate cut alone; the size of these effects is hard to gauge and depends on perceptions of policy permanence and firm-specific circumstances.
- Counter-intuitive effects:
  - GILTI can increase incentives to invest in tangible capital outside the United States because foreign tangible investment raises the tax-exempt GILTI amount.
  - FDII can make the United States more attractive for keeping IP and profitable export activity; other countries may respond by introducing similar provisions or by cutting main CIT rates depending on targets.
  - BEAT reduces profit shifting out of the United States for services; other countries might adopt similar anti-base-erosion measures or alter location incentives to avoid BEAT effects.
- Territoriality may raise sensitivity of U.S. MNEs to other countries’ tax rates (potentially intensifying tax competition); GILTI may dampen competition among very low-tax countries by imposing a minimum taxation effect.
- Policy reactions by other countries can both mute relative tax differentials (reducing shifts) and amplify global revenue loss (through lower rates elsewhere). Tentative estimate: tax rates elsewhere may fall by on average around 4 percentage points based on simulations.

### Conclusions (selected)
- The statutory U.S. CIT rate cut is the clearest and possibly largest source of spillovers; simulated average revenue losses range between 1.5 and 13.5 percent of the MNE tax base depending on assumptions and weighting scheme.
- Novel provisions (FDII, GILTI, BEAT, expensing, interest limitation) complicate incentives: in most cases they encourage more profit reporting in the United States, but can sometimes encourage real investment outside the United States.
- Large spillovers raise broader policy questions on whether the international CIT system can be made less prone to profit shifting and tax competition; ongoing multilateral initiatives (e.g., BEPS) may not be sufficient and further reforms are likely to be needed.
- Despite falling statutory rates, historical evidence shows CIT revenues have held up relatively well due to base-broadening; however, with base-narrowing measures like expensing, there is concern that lower rates may not be indefinitely offset by rising profits, and pressures on revenue could increase appetite for more fundamental reform.

*Source: wp18166 — References and Appendix I.*

### References .............................................................................................................

### wp18166 - References

### Introduction: scope and types of spillovers
- Major U.S. tax reform adopted on December 20, 2017 affected corporate income, personal income, and estate taxes and has:
  - Macroeconomic spillovers: stronger U.S. growth boosts export demand in trading partners.
  - Tax-related spillovers: changes to U.S. tax rates and structure affect MNEs’ location, profit reporting, and headquarters decisions.
- The U.S. 2017 reform reduced the U.S. statutory corporate rate by 14 percentage points and shifted aspects of the system toward territoriality, with potential to intensify tax competition.
- The reform included novel provisions intended to address international profit shifting; some measures are more aggressive than BEPS minimum standards and recommendations.

### Background: the pre-reform U.S. CIT regime
- Pre-reform characteristics:
  - Federal statutory tax rate: 35 percent; combined with state taxes: average around 40 percent.
  - ‘Domestic production activities deduction’ reduced the federal rate to 32 percent for many lines of business.
  - Most qualifying dividends and long-term capital gains taxed at rates of up to 20 percent; other gains taxed at rates of up to 39.6 percent.
  - About 95 percent of U.S. businesses were organized as pass-through entities; C-corporations accounted for slightly less than half of net business income.
  - Generous depreciation allowances and ‘bonus depreciation’ (expensing of 50 percent of many assets’ value in the first year) produced a narrow tax base despite high statutory rates.
  - Worldwide taxation with deferral: foreign earnings not taxed until repatriated; foreign tax credit available and excess credits carry forward or offset other active foreign income.
  - Accumulation of undistributed foreign earnings estimated at $2.6 trillion by 2017.
- Pre-reform tax planning practices included transfer pricing, thin capitalization, capitalizing subsidiaries in low-tax jurisdictions, ‘check-the-box’ hybrid arrangements, and corporate inversions.

### Main features of the 2017 reform
- Rate cut:
  - Federal statutory tax rate fell from 35 to 21 percent (a 14-percentage point cut).
  - Combined rate including state taxes fell on average from 40 to 26 percent.
  - 26 percent is close to the 2018 OECD average of 24 percent.
- Accelerated depreciation (expensing):
  - Expensing for most capital investment introduced for 5 years.
  - Between 2023-2027, expensing is scheduled to be phased out, dropping by 20 percentage points per year.
  - Pre-reform 2017 “bonus depreciation” allowed expensing of 50 percent of many assets’ value; in 2018 this was scheduled to fall to 40 percent absent reform.
- International tax provisions (noting uncertainty in detailed operation until regulations are finalized):
  - Global Intangible Low-Taxed Income (GILTI):
    - Creates a minimum U.S. tax on foreign active income without deferral; oil and gas sector exempt.
    - Minimum rate: 10.5 percent; tax base: foreign income exceeding 10 percent of foreign tangible assets.
    - Nonrefundable tax credit granted for 80 percent of foreign taxes.
    - Under current legislation, the minimum rate will increase to 13.125 from 2026, raising maximum creditable foreign tax to 16.4 percent.
    - Expense allocation rules (including interest allocated to foreign income) can make GILTI apply even when average foreign tax rates are higher.
  - Foreign Derived Intangible Income (FDII):
    - Preferential rate for certain foreign earnings of U.S.-resident firms: 13.125 percent.
    - Base defined as earnings in excess of a 10 percent return on tangible assets multiplied by the proportion of net income from export sales.
    - Rate scheduled to rise to 16.4 percent from 2026.
  - Base Erosion Anti-Abuse Tax (BEAT):
    - Alternative minimum tax adding back most foreign related-party service costs on a gross basis into the U.S. tax base.
    - Rate: 5 percent in 2018, rising to 10 percent in 2019 and to 12.5 percent from 2026.
    - BEAT does not provide credit for foreign taxes or take foreign tax rates into account.
- Interest deduction limits:
  - Interest deductions must not exceed 30 percent of earnings before interest, tax, depreciation, amortization, and depletion (EBITDA) plus business interest.
  - From 2022, limit applies to earnings before interest and tax (EBIT).
  - Regulated public utilities exempt; real estate and farming can opt out (then must use the Alternative Depreciation System).
- Anti-hybrid rules:
  - Disallow deductions for related-party hybrid transactions (not taxable in recipient country) or payments to an entity characterized as taxable in one jurisdiction and transparent in another.
- Transition (deemed repatriation):
  - Existing unrepatriated profits deemed repatriated and taxed at a one-time rate of 15.5 percent for funds held in cash or equivalent assets and 8 percent for all other assets.
  - Payment is spread over 8 years and is independent of actual repatriation.

### Impact of GILTI on average tax rates (authors’ calculations)
- GILTI could be more onerous than the previous worldwide system for some firms because it applies without deferral and provides a more limited foreign tax credit.
- A simplified industry-level estimate (authors’ estimates based on BEA data) of foreign CIT, GILTI, and combined average rates (in percent):
  - All industries: Foreign CIT 7.9, GILTI 2.9, Combined 10.8
  - Mining: Foreign CIT 80.8, GILTI 0.0, Combined 80.8
  - Manufacturing: Foreign CIT 11.8, GILTI 0.0, Combined 11.8
    - Food: Foreign CIT 21.9, GILTI 0.0, Combined 21.9
    - Chemicals: Foreign CIT 14.0, GILTI 0.0, Combined 14.0
    - Primary and fabricated metals: Foreign CIT 20.5, GILTI 0.0, Combined 20.5
    - Machinery: Foreign CIT 10.0, GILTI 1.9, Combined 11.9
    - Computers and electronic products: Foreign CIT 7.6, GILTI 3.1, Combined 10.7
    - Transportation equipment: Foreign CIT 15.0, GILTI 0.0, Combined 15.0
  - Wholesale trade: Foreign CIT 11.7, GILTI 0.5, Combined 12.2
  - Retail trade: Foreign CIT 17.7, GILTI 0.0, Combined 17.7
  - Information: Foreign CIT 8.5, GILTI 1.4, Combined 9.9
  - Finance and insurance: Foreign CIT 10.4, GILTI 1.9, Combined 12.3
  - Professional, scientific, and technical services: Foreign CIT 9.0, GILTI 2.6, Combined 11.6
  - Other industries: Foreign CIT 2.6, GILTI 7.9, Combined 10.5
- Calculation note: The GILTI liability (expressed as a ratio to corporate pre-tax income) is calculated for each industry as the maximum of zero and [(Foreign net income + foreign income tax – 0.1*PP&E)*0.105 – 0.8*foreign income tax]/(foreign income + foreign income tax).

### Impact on effective tax rates and considerations
- Net impact on effective tax rates is ambiguous due to simultaneous changes; key influences include the rate reduction and expensing (tending to reduce effective rates) and limits on interest deductions or restrictions on loss carryovers (which can raise effective rates, especially for cyclical businesses).
- Effective average tax rates (EATR) and effective marginal tax rates (EMTR) depend on assumptions about:
  - Type of asset invested.
  - Source of finance (equity vs debt).
  - Whether interest limitation is binding.
  - Whether GILTI, FDII, or BEAT apply.
- For technical implementation and measures, references are made to Devereux and Griffith (2003) for EATR/EMTR methodology and to sector-specific calculations described above.

*Source: wp18166 - References (IMF working paper content).*

### Appendix I. For an intuitive understanding, however, note that the EATR is a measure of the

### wp18166 - Appendix I. For an intuitive understanding, however, note that the EATR is a measure of the

### Effective tax rates: definitions and measurement
- EATR definition (as used in the paper): EATR = (R* − R)/p(τ + r) [equations (11)–(12) define R*, R and parameters], where R* is the PDV of the economic rent in absence of taxation, R is PDV in presence of taxation, p is pre-tax net profit, r is the real interest rate, and δ is true economic depreciation.
- γ = (1 − m_d)/(1 − z) measures difference in treatment of new equity and distributions; ρ = (1 − m_i)/(1 − z)i is the investor’s discount rate. A is the PDV of depreciation allowances. Financial effect F applies for new equity (NE) or debt (D) finance (zero for retained earnings). Exact formulae are given in equations (13)–(14).
- EMTR is obtained by setting post-tax rent R = 0 and solving for the corresponding cost of capital; EMTR simplification noted in Appendix I.B.
- Note on empirical effective tax rates: backward-looking approaches (tax payments divided by profits) are less suitable for assessing incentives for new investments because they reflect past decisions and losses carried forward.

### Effects of the 2018 U.S. tax reform on EATR and EMTR (standard tax system, 2018)
- Calculations reported are based on the tax system in place in 2018; no effective tax rates are reported for later years due to uncertainty about scheduled changes.
- Equity-financed investment in plant and equipment (standard tax system, including average state taxes that raise the federal rate by 5 percentage points):
  - Table 2 (plant and machinery, equity-finance) — EATR and EMTR under alternative debt-limit binding cases (columns correspond to Finance Reform pre, post (2018) and Debt limit not binding | marginally binding | binding):
    - EATR: 30.8, 19.5, 20.5, 12.8, 12.8, 12.3
    - EMTR: 5.2, 0.0, -173.7, -56.7, -16.9, -16.9
    - Assumptions: Asset: plant and machinery; Inflation: 2 percent; Real rate of interest: 5 percent; True depreciation: 5 percent; Profit rate (EATR): 20 percent.
  - Expensing drives EMTR to 0 for equity-financed marginal investments (expensing keeps the normal rate of return free of tax under the assumption investment at end of period; strictly zero only under that timing assumption).
- Debt-financed investment effects are mixed:
  - Pre-reform: debt-financed investment faced lower effective tax rates (debt bias) and could produce negative EMTRs because interest deductibility + depreciation allowances exceeded true economic depreciation.
  - Post-reform: the new limitation on interest deductibility (30 percent of EBITDA, from 2022 measured against EBIT) complicates outcomes:
    - If limit not binding (firms with low leverage/high profits), all interest deductible but lower statutory rate reduces the value of interest deductibility -> EMTR increases (relative to pre-reform) though EATR may not if profits high.
    - If marginally binding (new firms or existing interest exactly equal to permissible deduction), deductibility of new interest depends on project profitability; higher profit -> more interest deductible; in the EATR example (net profit 20 percent), all interest can be deducted.
    - If binding (existing leverage so interest exceeds allowed amount), no additional interest deduction from issuing more debt; EATR can still fall because additional profit creates scope for interest deductibility exceeding interest cost for profitable investments. This effect can also occur with equity finance for such firms.
  - Overall: reform clearly reduced—though did not eliminate—debt bias. Interest limit is less restrictive for the financial sector (netted against business interest received).

### Incorporating FDII, GILTI, and BEAT into effective tax rate calculations
- FDII (foreign-derived intangible income):
  - Applies only to foreign sales of domestic firms; impact is highly case specific and depends on firm-wide returns relative to 10 percent of tangible capital (the FDII threshold).
  - Table 3 (Effective Tax Rates under FDII, Equity Finance, 2018) — EMTR and EATR under four cases (exact rows and values):
    - Return on K_{t-1} << 10%: EMTR 0.0; EATR 19.5 — "Standard tax system applies"
    - Return on K_{t-1} < 10%: EMTR -3.5; EATR 12.9 — "Expense against full tax rate"
    - Return on K_{t-1} >> 10%: EMTR 13.7; EATR 16.8 — "Expense against FDII tax rate"
    - Intangible investment: EMTR 0.0; EATR 13.6 — "Entire return subject to FDII rate"
    - Assumptions: Inflation: 2 percent; Real rate of interest: 5 percent; True depreciation: 5 percent; Profit rate (EATR): 20 percent. Note: K_{t-1} represents the existing capital stock.
  - Key mechanics:
    - Firms with average returns below 10 percent: standard statutory and effective tax rates apply.
    - Firms pushed above 10 percent by new high-return investment: expensing reduces current tax liabilities at the standard rate, while the additional profit above 10 percent can be taxed at the reduced FDII rate (leading to possible negative EMTR in extreme cases).
    - Firms already above 10 percent: new investment may raise profit subject to the standard rate; EMTR positive.
    - For intangible investment where FDII applies to the entire return: EMTR returns to 0 (neutral taxation under expensing).
  - From 2022 R&D will become depreciable over 5 years (15 years for foreign R&D) rather than being expensed, which would raise tax rates relative to 2018.
  - Overall: EATR is reduced by FDII under most reasonable assumptions; pattern of reduction follows EMTR pattern.
- GILTI (Global Intangible Low-Taxed Income):
  - Move to territoriality reduces effective tax rates for many foreign investments that previously bore residual U.S. tax liability.
  - For firms covered by GILTI:
    - Some investments’ effective tax rates may rise because immediate GILTI liability could be more onerous than previous deferred U.S. tax.
    - For additional real investments with relatively low rates of return, effective tax rate could be negative because increases in foreign tangible capital raise the GILTI tax-free amount (mirror image of FDII).
  - Incorporation in effective tax rate formulas: replace FDII reduced rate by GILTI rate (0.105) and adjust A and bases; complications arise because foreign tax credits and different depreciation bases interact and GILTI applies globally across jurisdictions.
- BEAT:
  - BEAT primarily affects deductibility of certain related-party payments (services), bringing them into U.S. tax base; for a typical investment BEAT cannot be straightforwardly incorporated into the effective tax rate framework because the share of costs nondeductible under BEAT is unknown.

### Spillovers from the tax rate cut — channels and conceptual framework
- Three reaction margins considered in quantifying fiscal spillovers from the U.S. tax rate cut:
  - Profit shifting decisions of MNEs.
  - Real location decisions of MNEs (investment reallocation).
  - Policy decisions of other governments (tax rate reactions).
- Profit shifting:
  - MNEs exploit statutory tax differences via transfer pricing, intra-group loans, ownership of intangibles; profit-shifting entails costs (consultants, structures, legal risks).
  - The U.S. rate cut strengthens incentives to report profits in the United States and can lead to profit shifting out of a broad range of countries, including some low-tax jurisdictions due to lower relative costs of keeping profits in the United States and close operational links.
- Location of real investment:
  - Lower U.S. CIT rate makes the U.S. a more attractive location; tax affects net return to capital so a rate cut should foster investment in the United States.
  - Channels that affect investment elsewhere include:
    - Macroeconomic: higher U.S. investment raises global required rate of return and may reduce pre-tax global return.
    - Capacity limits at MNEs: profitable discrete projects may be allocated to the U.S., reducing investment elsewhere.
    - Anticipated profit shifting: changes in profit-shifting opportunities affect global cost of capital and hence investment.
- Policy spillovers:
  - Other governments may react via reduction in their own CIT rates (resource-flow and yardstick models); empirical evidence shows countries react to some average of other countries’ rates.

### Conceptual model and parameterization
- Basic log-linear profit reporting and tax revenue approximations are derived from Hines and Rice (1994) and Huizinga and Laeven (2008); key equations summarized (equations (1)–(6) in source).
- Parameter choices drawn from literature and data:
  - Semi-elasticity of capital with respect to tax rate (ε_K): set to 2.4 (FDI increases on average by 2.4 percent for a one percentage point cut in domestic statutory tax rate); authors regard simulated investment effects as upper-bound estimates.
  - Semi-elasticity of profits holding capital constant (ε_S): set to 1.5 (profits decrease on average by 1.5 percent in response to a one percentage point increase in group statutory tax rate).
  - Instantaneous reaction coefficient (β) for other countries’ CIT response to average foreign rates: set to 0.6 (in line with literature).
  - Country-specific capital intensity α_c estimated from micro regressions; sample: slightly more than 1.8 million observations from ORBIS (unconsolidated accounts, firms with profit and capital stock for at least 3 years).
  - Country-specific micro-weights ω_c defined by presence of direct ownership link to the United States (equation (10)). Robustness check uses bilateral FDI-based weights.
- Summary of country-specific parameters (Table 4 exact values):
  - Capital intensity: Mean 0.54; Median 0.54; Min 0.23; Max 1
  - Micro-weight: Mean 0.03; Median 0.02; Min 0; Max 0.19
  - FDI-weight: Mean 0.11; Median 0.06; Min 0; Max 1

### Simulation results — revenue spillovers (57 countries)
- Using micro weights (Table 5; losses are percentage changes of MNE profits):
  - No policy response (other countries maintain CIT rates):
    - Profit shifting: Mean 0.85; Median 0.7; Lowest percentile 0.22; Highest percentile 2.91
    - Investment response: Mean 0.73; Median 0.59; Lowest percentile 0.17; Highest percentile 3.11
    - Investment response (in percent of capital stock): Mean 1.36; Median 1.11; Lowest percentile 0.18; Highest percentile 4.66
    - Total change in MNE tax revenue: Mean 1.58; Median 1.34; Lowest percentile 0.39; Highest percentile 6.02
  - With policy response (other countries react):
    - Profit shifting: Mean 0.77; Median 0.66; Lowest percentile 0.22; Highest percentile 2.23
    - Investment response: Mean 0.65; Median 0.53; Lowest percentile 0.16; Highest percentile 2.4
    - Investment response (in percent of capital stock): Mean 1.23; Median 1.06; Lowest percentile 0.18; Highest percentile 3.6
    - Tax rate response: Mean 3.76; Median 2.89; Lowest percentile 1.24; Highest percentile 13.38
    - Tax rate response (in p.p. of tax rate): Mean 0.84; Median 0.69; Lowest percentile 0.1; Highest percentile 3.95
    - Total change in MNE tax revenue: Mean 5.17; Median 4.25; Lowest percentile 1.57; Highest percentile 17.14
- Using FDI weights (Table 6):
  - No policy response:
    - Profit shifting: Mean 2.4; Median 1.69; Lowest percentile 0.24; Highest percentile 12.87
    - Investment response: Mean 2.14; Median 1.3; Lowest percentile 0.19; Highest percentile 12.08
    - Investment response (in percent of capital stock): Mean 3.84; Median 2.7; Lowest percentile 0.06; Highest percentile 20.59
    - Total change in MNE tax revenue: Mean 4.54; Median 2.76; Lowest percentile 0.47; Highest percentile 24.2
  - With policy response:
    - Profit shifting: Mean 1.48; Median 1.28; Lowest percentile 0.21; Highest percentile 3.4
    - Investment response: Mean 1.31; Median 0.89; Lowest percentile 0.16; Highest percentile 3.74
    - Investment response (in percent of capital stock): Mean 2.49; Median 2.08; Lowest percentile 0; Highest percentile 6.26
    - Tax rate response: Mean 10.69; Median 7.1; Lowest percentile 1.31; Highest percentile 53.63
    - Tax rate response (in p.p. of tax rate): Mean 2.12; Median 1.49; Lowest percentile 0; Highest percentile 18.5
    - Total change in MNE tax revenue: Mean 13.48; Median 9.58; Lowest percentile 1.77; Highest percentile 57.51
- Key summary findings:
  - When other countries do not change CIT rates, corporate tax revenue from MNEs decreases by 1.58 percent on average across countries (micro weights).
  - Changed profit-shifting incentives account for 0.85 percent average contraction; investment response 0.73 percent average contraction (micro weights).
  - Incorporating policy response increases average revenue shortfalls to 5.17 percent (micro weights), as tax rate responses amplify revenue losses.
  - Using FDI weights yields larger spillovers: average losses around 4.5 percent with no policy response and 13.5 percent with policy response.
  - The paper reports that depending on parameter assumptions, reform leads to average revenue losses between 1.5 and 13.5 percent of the MNE tax base.
- Sensitivity and uncertainty:
  - Large uncertainty due to wide ranges of elasticities in the literature; policy reaction parameter (β) is particularly influential.
  - Table 7 (tax rate reactions, GDP weights, β = 0.6):
    - Full sample (182 economies): GDP in US$ — Minimum 3.8; Mean 3.8; Maximum 4.1
    - Full sample (182 economies): GDP in PPP US$ — Minimum 2.7; Mean 2.7; Maximum 3.0
    - Largest 30 economies: GDP in US$ — Minimum 4.2; Mean 4.3; Maximum 4.6
    - Largest 30 economies: GDP in PPP US$ — Minimum 3.2; Mean 3.3; Maximum 3.5
  - Table 8 (estimated tax reactions by coefficient, GDP weighted, full sample 182 economies — exact rows):
    - β 0.3: Minimum 1.3; Mean 1.4; Maximum 1.5
    - β 0.4: Minimum 2.0; Mean 2.0; Maximum 2.2
    - β 0.5: Minimum 2.8; Mean 2.8; Maximum 3.0
    - β 0.6: Minimum 3.8; Mean 3.8; Maximum 4.1
    - β 0.7: Minimum 5.2; Mean 5.2; Maximum 5.4
    - β 0.8: Minimum 7.0; Mean 7.1; Maximum 7.3
    - β 0.9: Minimum 9.7; Mean 9.7; Maximum 9.9

### Spillovers from other reform features and policy implications
- FDII, GILTI, and BEAT are likely to increase profit reporting in the United States beyond what is implied by the statutory rate cut alone; size of these effects is hard to gauge and depends on perceptions of policy permanence and firm-specific circumstances.
- Some provisions can have counter-intuitive effects:
  - GILTI can increase incentives to invest in tangible capital outside the United States because foreign tangible investment raises the tax-exempt GILTI amount.
  - FDII can make the United States more attractive for keeping IP and profitable export activity; other countries may respond by introducing similar provisions (e.g., patent boxes) or by cutting main CIT rates depending on their investment targets.
- BEAT reduces profit shifting out of the United States for services; other countries might adopt similar anti-base-erosion measures or alter incentives for locating production to avoid BEAT effects.
- Territoriality may raise sensitivity of U.S. MNEs to other countries’ tax rates (potentially intensifying tax competition); GILTI may dampen competition among very low-tax countries by imposing a minimum taxation effect.
- Policy reactions by other countries can both mute relative tax differentials (reducing shifts) and amplify global revenue loss (through lower rates elsewhere). Tentative estimate: tax rates elsewhere may fall by on average around 4 percentage points based on simulations.

### Conclusions (selected)
- The statutory U.S. CIT rate cut is the clearest and possibly largest source of spillovers; simulated average revenue losses range between 1.5 and 13.5 percent of the MNE tax base depending on assumptions and weighting scheme.
- Novel provisions (FDII, GILTI, BEAT, expensing, interest limitation) complicate incentives: in most cases they encourage more profit reporting in the United States, but can sometimes encourage real investment outside the United States.
- Large spillovers raise broader policy questions on whether the international CIT system can be made less prone to profit shifting and tax competition; ongoing multilateral initiatives (e.g., BEPS) may not be sufficient and further reforms are likely to be needed.
- Despite falling statutory rates, historical evidence shows CIT revenues have held up relatively well due to base-broadening; however, with base-narrowing measures like expensing, there is concern that lower rates may not be indefinitely offset by rising profits, and pressures on revenue could increase appetite for more fundamental reform.

*Source: IMF Working Paper wp18166 — Appendix I (extracts and tables reproduced from the supplied content).*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2018/wp18166.pdf_