## wpiea2021161-print-pdf

## Source details

**Canonical URL:** [wpiea2021161-print-pdf](https://www.imf.org/-/media/files/publications/wp/2021/english/wpiea2021161-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2021/english/wpiea2021161-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2021/english/wpiea2021161-print-pdf.pdf.json)

---

### I. Introduction — problem, trends, and objectives
- Problem: Multinational enterprises (MNEs) shift profits and erode domestic tax bases via intra-group transactions, financial arrangements, and corporate structuring (transfer pricing; inter-affiliate debt), leaving CIT burdens on domestic firms.
- Recent trends:
  - CIT rates have fallen by around 20 percentage points on average since 1980 (IMF 2019).
  - Despite falling rates, CIT revenues have remained robust; in some countries (for example, the United States), revenues have started to increase even as rates have fallen.
  - CIT revenues have grown markedly as a share of GDP in countries at all income levels since the early 1980s.
- Rationale for minimum taxes (MTs):
  - MTs deter evasion/avoidance, shore up domestic revenue—important for developing economies with weaker tax administrations and broad tax incentives.
  - MTs may act as a “recovery contribution” after large stimulus packages.
  - Tradeoffs: MTs can distort production efficiency and slow recovery; they may be third-best optimal when public funds are scarce.
- International coordination:
  - OECD GloBE proposal (late 2019) is a global anti-base erosion GMT with outbound and inbound rules; paper focuses on local minimum taxes (LMTs).
- Paper objectives:
  - Construct database of past and current MTs.
  - Use macro and micro data to study MT impacts on revenues and corporate behavior.
  - Paper structure: definitions (Section II), efficiency/equity (Section III), stylized facts and empirical analysis (Section IV), conclusions (Section V).

### Definitions, classification, and bases of MTs
- What is an MT:
  - Guarantees a floor on tax liability; can be fixed payment or modified rules. Alternative minimum taxes (AMTs) operate in parallel to standard regime.
  - MTs may be creditable against CIT or payable even if loss-making.
- Classification used:
  - Local Minimum Tax (LMT): applies to income sourced locally (domestic resident corporations and inbound activities).
  - Global Minimum Tax (GMT): applies to outbound activities (domestic multinationals’ foreign earnings).
- Base choices (defining for MT design):
  - Turnover (gross income/receipts)
  - Assets (net or gross)
  - Modified net income with fewer deductions/exemptions

- Empirical prevalence (as of end-2018):
  - 63 countries operating an MT.
  - Of these 63 countries: 31 use turnover; 10 use assets; 9 use modified income.
  - Countries may operate parallel systems using multiple bases.

### Turnover-based MTs — features and facts
- Properties:
  - Common in presumptive taxation; reduces scope for evasion because under-reporting sales is harder than fabricating input costs.
  - Drawbacks: blind to net margins; identical tax for firms with different costs; can force payment when small profits or losses; can cascade along production chain.
- Empirical facts (as of 2018):
  - In use across 31 countries.
  - Rates vary between 0.2 percent (Tunisia) to 3 percent (Bolivia, Equatorial Guinea, Guinea, and Madagascar).
  - Average tax rate is 1.2 percent.
- Comparative illustration (US$ values):
  - Turnover tax rate 2 percent vs Profits tax rate 20 percent example:
    - Revenue: 100,000 / 100,000
    - Costs of Goods Sold: 50,000 / 70,000
    - Gross Profit: 50,000 / 30,000
    - Expenses: 20,000 / 20,000
    - Net Operating Income: 30,000 / 10,000
    - Other Income: 5,000 / 5,000
    - Other Expenses: 7,500 / 7,500
    - Net Other Income: -2,500 / -2,500
    - Net Income before tax: 27,500 / 7,500
    - Taxes payable (Turnover): 2,000 / 2,000
    - Taxes payable (Profits): 5,500 / 1,500
    - Net income after tax (Turnover): 25,500 / 5,500
    - Net income after tax (Profits): 22,000 / 6,000
    - Effective Tax Rate (Turnover) percent: 7.8 / 36.4
    - Effective Tax Rate (Profits) percent: 25.0 / 25.0
    - Pretax Profit Margin percent: 27.5 / 7.5

### Assets-based MTs — features and facts
- Properties:
  - Base often = book value of assets, sometimes net of debt (tax on equity).
  - Robust in high inflation environments; gross-assets tax neutral with respect to risk-bearing if assets marked to market.
  - Penalizes capital-intensive firms; net-assets base can encourage over-leveraging.
  - Recommended gross-assets rate formula: τ_a = τ_π (p − r) where p = expected real marginal return, r = real rate of interest, τ_π = statutory CIT rate.
- Empirical facts:
  - In place in only 10 countries.
  - Rates vary between 0.4 percent (Peru) to 2 percent (Panama).
  - Average tax rate is 1.2 percent.

### Modified-income MTs — features and facts
- Properties:
  - Base similar to standard CIT but with fewer deductions/allowances.
  - Rates typically much higher (on average around 10 percent) because base closely matches standard CIT base.
  - More sophisticated, more likely in advanced economies.
- Example (Korea):
  - 10 percent if tax base <KRW 10 billion;
  - 12 percent if tax base >KRW 10 billion and <KRW 100 billion;
  - 17 percent if tax base >KRW 100 billion.

### Carry-forward and foreign tax credit considerations
- Carry-forward rules:
  - Loss-making firms typically subject to MT; carry forward availability varies widely.
  - Where allowed, MT payments can be carried forward and applied against future excess of regular income tax liabilities over MT, but cannot reduce payments below MT in any year.
  - Illustrative two-year scenarios (Table 2):
    - Year 1 / Year 2 (Scenario 1) / Year 2 (Scenario 2)
    - MT liability: 100 / 100 / 100
    - Income tax liability: 80 / 140 / 110
    - Amount paid: 100 / 120 / 100
    - “Excess” MT available to be carried forward: 20 / 0 / 10
  - Carry back of MT payments does not appear to be permitted under MTs.
- Foreign tax credit considerations:
  - Design should allow multinationals to credit MT payments against home country income tax liabilities.
  - Two mechanical approaches:
    - Approach 1: characterize entire liability as the greater tax—complete denial of foreign tax credits when MT bites.
    - Approach 2: MT allowed as credit against regular CIT—tax imposed first qualifies for credit; excess MT carried forward.
  - Practical note: one design to maximize U.S. foreign tax credit availability is to require taxpayers to pay income tax in full, then refund excess MT carried forward rather than reducing foreign tax credit availability.

### Efficiency, equity, and theoretical insights
- In a frictionless environment no MT is needed; MTs rationalized as third-best when enforcement limited or tax preferences proliferate.
- MTs can limit inefficient tax preferences and tax planning that distort financing, asset, or location decisions.
- METR and AETR effects depend on MT design:
  - MTs can equalize tax treatment across assets and reduce dispersion of METRs.
  - MTs more likely to increase cost of capital for debt-financed investments; can reduce debt bias by increasing METRs on debt-financed investments while reducing METRs on equity-financed investments.
  - Extending carry forward provisions can mitigate MT impact on investment incentives; with sufficiently long carry forward, METR with MT converges to no-MT level.
- Location decisions: MTs unambiguously increase AETR on projects earning economic rent.
- Equity and political economy:
  - MTs can improve horizontal equity by countering avoidance and differing effective tax rates across similar firms.
  - Vertical equity impact ambiguous; MTs can be viewed as increasing aggregate taxation of capital.
  - Assets-based MTs likely more equitable since assets better predict economic income than turnover.
- Compliance: MTs may boost taxpayer morale and compliance by enhancing perceived fairness.

### Database construction and macro stylized facts
- Novel database of LMTs across 196 countries since 1980 by base and rate; compiled manually from IBFD, Deloitte, and national legislation.
- Merged with IMF WEO series on “taxes on income, profits, and capital gains, payable by corporations” and IMF FAD statutory CIT rate series.
- Stylized aggregate findings:
  - Turnover-based MTs are the most popular historically and recently.
  - Over past 25 years, CIT rates tend to be higher on average in countries with MTs.
  - CIT rates highest on average in countries with turnover-based MTs; modified-income-base countries have CIT rates closer to those without MTs.
  - Revenue performance:
    - Only countries with turnover-based MTs have outperformed countries without MTs since 2000, collecting CIT revenues of up to almost 1 percentage point of GDP more on average.
    - Assets-based MT countries broadly in line with countries without MTs (only more recently outperforming).
    - Modified income-based MT countries have tended to underperform in terms of revenues.

### Macro identification and results (local projections)
- Method: local projection regressions (Jordà 2005 et al.) estimating impulse response functions of CIT revenues to MT changes; controls include country/time fixed effects, lags/forwards of MT shocks, lagged revenue growth; supplemental controls in alternatives.
- Data limitations:
  - Of 86 country-year observations for MT introduction/reform, only 48 observations available under broader WEO income tax revenue variable and only 25 under the CIT revenue variable.
- Estimated equation summary:
  - ∆^h y_{i,t} = α_i^h + γ_t^h + β^h ∆s_{i,t} + Σ_{j=1}^p β_{1j}^h ∆s_{i,t−j} + Σ_{j=1}^p θ_{1j}^h ∆y_{i,t−j} + Σ_{j=0}^p θ_{2j}^h x_{i,t−j} + ε_{i,t}^h
  - ∆^h y_{i,t} ≡ y_{i,t+h} − y_{i,t−1}; ∆s_{i,t} ≡ s_{i,t} − s_{i,t−1}; y_{i,t} ≡ ln(CITrev_{i,t}); s_{i,t} ≡ ln(MT_{i,t})
- Macro results:
  - Panels (Figure 5) show advanced-economy CIT revenues fall marginally on impact after MT introduction before increasing by the second year by almost two percentage points and remaining positive.
  - Repeal responses largely negative—averaging around 5 percentage points before falling further.
  - Revenue-to-GDP responses are asymmetric but not statistically different from zero (confidence intervals encompass the x‑axis).
  - Alternative specifications and subsamples produced similar, statistically insignificant results.
- Identification caveats:
  - Endogeneity between revenues and MT introduction/repeal not fully resolved.
  - Cannot fully disentangle MT impact on numerator (revenues) vs denominator (GDP).
  - Instrument (informality measure) tested; impulse responses remained not significant.

### Firm-level micro evidence — data and approach
- Data: Bureau Van Dijk Orbis (unconsolidated accounts) for 176 countries (1980–2018), merged with statutory CIT rates and MT dataset.
- Firm variables: gross sales, operating profit before taxes, taxes payable, total assets, employees, sector descriptors.
- AETR definition: taxes payable divided by operating revenues from sales (turnover).
- Event window: years t−1, t, t+1 around MT introduction, reform, or repeal; restrict to economies with >70 firms in those years and other firm-level filters.
- Sample limitations: reporting quality in Orbis poor in many MT countries; final usable sample smaller.

### Micro-level findings — AETRs, investment, employment, profits
- Pooled event-study estimate:
  - MT introduction associated with a 1.6 percent higher average ETR (Table 8).
- Heterogeneity by MT type (pooled estimates):
  - Modified income-based MTs: increase in average ETR up to around 2.7 percentage points.
  - Assets-based MTs: increase in average ETR around 1.2 percentage points.
  - Turnover-based MTs: increase in average ETR around 1.0 percentage points (up to 1.5 percent for Slovakia’s largest threshold).
- Slovakia (turnover-based “tax license”) specifics:
  - Companies above largest turnover threshold EUR 500,000: 1.5 percent average ETR increase; effect distributed across sectors for large companies.
  - Companies below threshold: larger AETR effect concentrated in leasing, real estate, and retail sales.
  - Smaller companies: small positive employee growth, significant contraction in assets, no effect on pre-tax profitability.
- Investment/employment/reported profits:
  - Association with greater investment and employment in balanced panel for asset and modified income-based MTs (possible omitted variable bias).
  - Increased reported profits may reflect more truthful reporting or compositional exit of less productive firms.
- Asymmetry on repeal:
  - Repeal or rate reduction does not show immediate deterioration in firm-level AETRs (coefficients negative but statistically insignificant), except balanced panel for assets-based MTs.
  - Possible persistence due to improved administration/documentation.

### Sectoral heterogeneity and identification attempts
- Sectors with thinner profit margins disproportionately burdened by turnover-based MTs; asset-intensive sectors penalized by assets-based MTs.
- Augmented specification introduces sectoral structural AETR τ_jk,t^MT(AETR) = f(X_jk,t), with X_jk,t = Profit_jk,t̄ / Revenue_jk,t̄.
- Findings:
  - Introducing an MT leads to average increase in AETR with respect to turnover of around 2.5 percent controlling for binding liability and sector structural profit margins.
  - Marginally stronger effect on firm AETR for firms bound by MT when sectoral AETR increases.
  - Instrumenting sectoral variation suggests MT impact on AETRs not primarily driven by sector-specific variation.

### Counterfactual revenue-yield back-of-the-envelope estimates (Orbis population)
- Simulated policies (excluding certain offshore jurisdictions):
  - (i) 0.5 percent MT on turnover.
  - (ii) 1 percent MT on total assets.
- Median country estimated yields relative to current levels (latest available year for median country):
  - Turnover-based MT (0.5 percent): could yield 7 percent additional tax revenues relative to current levels.
  - Assets-based MT (1 percent): could yield as much as 33 percent additional tax revenues relative to current levels.
- Scaled to national accounts (assuming Orbis firms represent similar proportion of overall corporate tax liability measured by WEO CIT revenues):
  - Median revenue increment: 0.2 percent of GDP for turnover-based MT; 0.9 percent of GDP for assets-based MT.
  - Median corporate income tax to GDP ratio in sample: 2.7 percent.
- Assumptions/caveats:
  - Assumes no behavioral changes and maintenance of current compliance.
  - Sample excludes low-tax jurisdictions; differs from OECD Pillar 2 population.

### Case study — Honduras (administrative panel data)
- 2013 dual MT introduced:
  - Turnover-based MT: 1.5 percent of turnover if turnover > HNL 10 million (selected sectors 0.75 percent).
  - Net assets tax (NAT): 1 percent on net assets if net assets > HNL 3 million.
  - Standard CIT: 25 percent of reported taxable profits.
  - Ultimate liability: CIT_liability_t = max(MT_t, NAT_t, CIT_t)
- Counterfactual simulation using 2013 AETRs held constant for firms continuously present (2014–16):
  - Tax revenues from firms that became liable to the MT after introduction increase by almost 1 percent of GDP in 2016.
- Behavioral/documentation effects:
  - Bunching of reported profits around indifference threshold (profit margin of 6 percent).
  - More firms truthfully reporting higher profit margins; previously evading firms faced higher ETRs over 2014–16.
- Sectoral distribution:
  - Gross-sales MTs disproportionately burden low-margin sectors (e.g., food retail).

### Policy-relevant implications and design recommendations
- Comparative effectiveness:
  - Modified income-based MTs most effective at increasing average ETRs; assets-based second; turnover-based third but most common and administratively simple.
- Design recommendations:
  - Base selection:
    - Turnover: simple and observable; poor predictor of profitability.
    - Cash flow: less distortionary but potentially complex.
    - Assets/modified income: better proxies for economic income and more effective against profit shifting.
  - Rate setting:
    - MT rate should proxy effective tax rate under regular income tax but kept relatively low (except modified income MTs).
    - Turnover-based MT rate should proxy CIT burden on low gross-margin firms.
    - Assets-based MT rate should mimic regular tax burden on a firm earning a return covering its cost of capital.
    - Consider sector differentiation where margins or rates of return vary, weighing targeting benefits against complexity and classification ambiguity.
  - Crediting and carryforward:
    - MT payments should be creditable against future CIT liabilities with long carry forward periods to mitigate negative investment impacts.
    - Carry forward provisions help offset METR increases and support investment, especially under high uncertainty.
- Aggregate implications:
  - Unilateral, administratively simple LMTs can effectively raise CIT revenue in source countries without large aggregate adverse effects on corporate investment/profitability.
  - Findings support inbound-rule segment of OECD Pillar 2 proposal (local perspective), while not addressing outbound rule.
  - Back-of-the-envelope median revenue gains of 0.2 and 0.9 percent of GDP (turnover- and assets-based LMTs) justify policy consideration where tax capacity is imperfect.

### Limitations and avenues for further research
- Identification challenges: endogeneity in MT timing; macro estimates statistically imprecise.
- Partial equilibrium micro estimates: illustrative, not causal; small sample and data quality issues in Orbis for many MT countries.
- Research priorities:
  - Welfare implications of MTs within optimal taxation models, conditional on specific frictions.
  - Deeper empirical identification using richer panels and administrative data across countries.
  - Evaluate assets-based MT impacts on investment in fixed assets and on average ETRs.

*Source: wpiea2021161-print-pdf*

### References _______________________________________________________________________________________ 41

### I. INTRODUCTION

### Problem Statement: Profit Shifting and Base Erosion
- Multinational enterprises (MNEs) can exploit differences in domestic tax laws to shift profits across borders and erode domestic tax bases using intra-group transactions, financial arrangements, and corporate structuring.
- Specific avoidance activities include transfer pricing (e.g., charging relatively lower prices for goods and services transferred from high-tax to low-tax affiliates) and the strategic use of inter-affiliate debt (e.g., financing high-tax affiliates using debt issued by low-tax affiliates).
- Policymakers are concerned that MNEs with large earnings can pay little or no corporate income taxes locally, leaving the CIT burden to fall primarily on domestic (non-MNE) firms and less mobile tax bases.

### Recent Trends in CIT Rates and Revenues
- CIT rates have fallen by around 20 percentage points on average since 1980 (IMF 2019).
- Despite falling rates, corporate income tax (CIT) revenues have remained robust on average over time; in some countries (for example, the United States), revenues have started to increase even as rates have fallen.
- CIT revenues have grown markedly as a share of GDP in countries at all income levels since the early 1980s.
- Contributing factors to strong CIT performance include revenues from natural resources and the growing share of corporate profits in national income.
- CIT revenues remain volatile for most countries and are at risk from globalization and pervasive use of tax preferences in some regions.

### Rationale for Minimum Taxes (MTs)
- MTs have re-emerged as an instrument to deter tax evasion and avoidance—particularly for developing economies—by limiting MNEs’ ability to exploit domestic tax system differentials (IMF 2019).
- MTs can help shore up tax systems and ensure a certain amount of domestic revenue, which is particularly important for developing economies with weaker tax administrations and for countries offering broad tax incentives to attract mobile investment.
- MTs may act as a “recovery contribution” to recover lost fiscal resources following large stimulus packages.
- Tradeoffs noted: MTs can distort production efficiency on the margin and potentially slow down recovery, but they may be third-best optimal in a world with a high marginal value of public funds (Mayshar 1990) and scarce resources.

### International Coordination: OECD GloBE Proposal and Concerns
- In late 2019, the OECD formulated a global anti-base erosion (“GloBE”) proposal to stop a harmful race to the bottom on corporate taxes (OECD 2019).
- The GloBE proposal aims to shield developing economies from pressure to offer inefficient tax incentives and to stem forgone revenue losses for countries with extensive CIT incentives.
- Critics of both multilateral (GloBE) and unilateral MT approaches worry these instruments may impair automatic stabilizers during sharp economic shocks (e.g., the pandemic-driven recession).
- Introducing MTs could shift international competition for mobile investment toward subsidies, raising potential WTO issues.

### Research Gap and Paper Objectives
- There is limited literature reviewing and exploring the potential impact of MTs on revenues and economic activity.
- This paper constructs a database of past and current MTs and uses macro- and micro-level data to study their impact.
- Structure of the paper:
  - Section II: defines MTs and reviews different types in operation.
  - Section III: examines efficiency and equity implications.
  - Section IV: presents stylized facts from the created database and studies the impact of MTs on revenues and corporate behavior using macro, firm-level, and administrative data.
  - Section V: concludes.

### Appendix and Supplementary Material (listed in source)
- Appendix I. - The Rationale For A Corporate Minimum Tax
  - A. Restricting the use of corporate tax preferences
  - B. Countering tax planning and avoidance
  - C. Complexity and Compliance Costs
  - D. Reduced Volatility and Cyclicality of Corporate Income Tax Revenues
- Selected figures referenced in the source (examples):
  - Figure 1. Determining the Scope of Minimum Taxes
  - Figure 5. Impulse Responses of CIT Revenues-to-GDP to the Introduction and Repeal of MTs
  - Figure 7. Distribution of Firms’ Profit Margins, 2013-2016
- Selected tables referenced in the source (examples):
  - Table 1. AETRs under different turnover tax rates and margins.
  - Table 3. Summary of Minimum Tax Rates by Base, 2018
  - Table 8. Estimates on Impact of MT Introduction on Firm Average ETR

*Source: wpiea2021161-print-pdf - References _______________________________________________________________________________________ 41*

### appendix explores in greater depth several of the rationales for the use of MTs by countries.

### wpiea2021161-print-pdf - appendix explores in greater depth several of the rationales for the use of MTs by countries.

### What is a Minimum Tax?
- Corporate minimum taxes (MTs) guarantee a floor on the tax liability of businesses.
- MTs can be:
  - a fixed payment, or
  - a modified set of rules that simplify the regular tax system (example: Austria).
- Alternative minimum taxes (AMTs) represent a parallel liability to the “standard” regime; taxpayers compute tax under both regimes and pay the larger amount.
- MTs can:
  - use broader definitions of income and a less generous set of deductions, or
  - have separate bases and rates.
- In some countries, MTs can be creditable against a company’s corporate income tax (CIT) liability; in others, MTs must be paid even if the company is loss-making.
- Classification used in this paper:
  - Local Minimum Tax (LMT): applies to income sourced locally (domestic resident corporations and inbound activities where foreign corporations have income/activities in the domestic economy).
  - Global Minimum Tax (GMT): applies to outbound activities (domestic multinationals’ foreign earnings). Examples include CFC rules and GILTI.
- The OECD’s GloBE proposal is a GMT with an “outbound rule” (income inclusion rule) and an “inbound rule” (denial of deductions or treaty benefits). The paper focuses on local minimum taxes (LMTs).

### The Base of the Corporate Minimum Tax
- Choice of base is defining for MT design; bases can match the regular CIT or be a rationalization (AMT).
- Common simplified bases: turnover (gross income/receipts), assets (net or gross), or a modified form of net income with fewer deductions/exemptions.

A. Turnover-based Minimum Taxes
- Turnover-based MTs:
  - are common in presumptive taxation and as a simpler alternative where compliance/admin costs of general income tax are regressive.
  - reduce scope for evasion relative to profits taxes because under-reporting sales is typically harder than fabricating input costs.
- Drawbacks:
  - blind to net margins; firms with identical turnover but different cost structures pay identical turnover tax.
  - can force firms to pay tax when making small profits or losses; high turnover rates can discourage investment and new firm growth.
  - lead to production chain cascading and vertical integration incentives; thus rarely used in more developed economies.
- Empirical facts (as of 2018):
  - turnover-based MTs were in use across 31 countries.
  - rates vary between 0.2 percent (Tunisia) to 3 percent (Bolivia, Equatorial Guinea, Guinea, and Madagascar).
  - average tax rate is 1.2 percent.
- Example from comparative illustration (values in US$ unless otherwise indicated):
  - Turnover (Tax rate 2 percent) vs Profits (Tax rate 20 percent).
  - Revenue: 100,000 / 100,000
  - Costs of Goods Sold: 50,000 / 70,000
  - Gross Profit: 50,000 / 30,000
  - Expenses: 20,000 / 20,000
  - Net Operating Income: 30,000 / 10,000
  - Other Income: 5,000 / 5,000
  - Other Expenses: 7,500 / 7,500
  - Net Other Income: -2,500 / -2,500
  - Net Income before tax: 27,500 / 7,500
  - Taxes payable (Turnover): 2,000 / 2,000
  - Taxes payable (Profits): 5,500 / 1,500
  - Net income after tax (Turnover): 25,500 / 5,500
  - Net income after tax (Profits): 22,000 / 6,000
  - Effective Tax Rate (Turnover) percent: 7.8 / 36.4
  - Effective Tax Rate (Profits) percent: 25.0 / 25.0
  - Pretax Profit Margin percent: 27.5 / 7.5

B. Assets-based Minimum Taxes
- Assets-based MTs:
  - base often = book value of firm’s assets, sometimes net of debt and liabilities (tax on equity).
  - have been popular in Latin America (examples: gross assets in Argentina and Mexico; net assets in Peru, Ecuador, Colombia, Guatemala).
- Rationale and properties:
  - more robust in high inflation environments where income taxes may be eroded by inflation and loss carryforwards.
  - gross assets tax is neutral with respect to risk-bearing; it gives a predictable liability.
  - market value of assets approximates net present value of future after-tax cash flows—if assets are regularly marked to market.
- Recommended base content:
  - ideally include fixed assets: land, structures, machinery, equipment.
  - exclude current assets (cash balances, accounts receivable, inventories) to attain relatively acyclical revenues.
  - ideally include intangibles in principle, but practically often excluded due to valuation difficulties.
- Caveats:
  - penalizes capital-intensive firms; net assets base can encourage over-leveraging (increased debt to raise liabilities and lower MT).
  - gross vs net assets choice affects rate setting; recommended gross-assets rate formula: τ_a = τ_π (p − r) where p = expected real marginal return, r = real rate of interest, τ_π = statutory CIT rate.
- Empirical facts:
  - assets-based MTs are in place in only 10 countries.
  - rates vary between 0.4 percent (Peru) to 2 percent (Panama).
  - average tax rate is 1.2 percent.

C. Modified-income Minimum Taxes
- Modified-income MTs:
  - deviate less from standard CIT; base similar but with variations in deductions, credits, allowances.
  - rates typically much higher (on average around 10 percent) because base more closely matches standard CIT base.
  - more sophisticated and more likely in advanced economies.
- Example (Korea):
  - 10 percent if the tax base is <KRW 10 billion;
  - 12 percent if the tax base is >KRW 10 billion and <KRW 100 billion;
  - 17 percent if the tax base >KRW 100 billion.
- These MTs reflect targeting of identified base erosion practices rather than wholesale base substitution.

### Carry-forward Rules
- Loss-making firms are typically subject to MT, which can exacerbate cash-flow problems.
- Countries vary widely in availability and length of carry forward provisions; carry forward provisions influence investment incentives.
- Mechanics:
  - where allowed, MT payments can be carried forward and applied against future excess of regular income tax liabilities over MT, but cannot reduce payments below the MT liability in any year.
- Two illustrative scenarios (Table 2):
  - Year 1 / Year 2 (Scenario 1) / Year 2 (Scenario 2)
  - MT liability: 100 / 100 / 100
  - Income tax liability: 80 / 140 / 110
  - Amount paid: 100 / 120 / 100
  - “Excess” MT available to be carried forward: 20 / 0 / 10
- Carry back of MT payments does not appear to be permitted under MTs.

### Foreign Tax Credit Considerations
- Design should allow multinationals to credit MT payments against home country income tax liabilities.
- Problem: some countries allow credit only for foreign taxes explicitly levied as income taxes.
  - Pre-2018 U.S.-based MNEs encountered difficulty claiming U.S. foreign tax credits for assets-based taxes paid in Latin America.
- Mechanical approaches to crediting:
  - Approach 1: taxpayer pays the greater of regular income tax and MT and entire liability is characterized as the greater tax—complete denial of foreign tax credits when MT bites.
  - Approach 2: MT allowed as a credit against regular CIT—when two taxes overlap, the tax imposed first is the tax for which credit is available; excess MT carried forward to offset future CIT.
- Practical note:
  - To maximize availability of U.S. tax credits and avoid legal challenges, one design is to require taxpayer to pay income tax in full, then receive a refund for excess MT carried forward, rather than reducing U.S. foreign tax credit availability.

### Efficiency and Equity
- In a second-best environment with minimal distortions and no evasion, no MT is needed; MTs are rationalized as a third-best policy when enforcement is limited or tax preferences proliferate.
- MTs can:
  - limit inefficient tax preferences,
  - limit tax planning that distorts financing, asset, or location decisions of MNEs.
- MT impact on Marginal Effective Tax Rate (METR) and portfolio distortions depends on MT design (turnover-, asset-, or modified-income-based).
- Key theoretical insights:
  - King-Fullerton (KF) model aids in assessing METR impacts; Bernheim (1989) provides results on modified-income MTs.
  - Equation (1) summary: P = γ(1 − D*), where P = D − D* = value of tax preference; γ = [τ − τ*]/[(1 − τ*) τ]. Lower P and lower τ* relative to τ increase likelihood MT reduces METRs.
- Examples and implications:
  - For a 30 percent CIT, abolishing depreciation allowances under MT can permit MT rate on buildings at 16 percent to leave METR unchanged, but vehicles would require MT rate of 5 percent.
  - MTs can equalize tax treatment of assets and reduce dispersion of METRs across assets.
- Debt vs equity effects:
  - MTs more likely to increase cost of capital for debt-financed investments because interest deductibility value is diminished.
  - MTs can reduce debt bias by increasing METRs on debt-financed investments while reducing METRs on equity-financed investments.
- Risk and uncertainty:
  - MTs increase METR as return uncertainty increases; MT provides asymmetric hedge to governments against CIT revenue shortfalls.
  - Extending carry forward provisions can mitigate MT impact on investment incentives; with sufficiently long carry forward, METR with MT converges to no-MT level.
- Location decisions:
  - Average Effective Tax Rate (AETR) matters for location of discrete projects; MTs unambiguously increase AETR on projects earning economic rent.
- Equity and political economy:
  - MTs can improve horizontal equity by countering tax avoidance and differing effective tax rates across similar firms.
  - Vertical equity impact is ambiguous and depends on incidence; MTs can be viewed as increasing aggregate taxation of capital (typically owned by wealthier individuals).
  - Assets-based MTs are likely to be more equitable since assets better predict economic income than turnover.
- Compliance effects:
  - By enhancing perceived fairness, MTs may boost taxpayer morale and improve compliance.

### Assessing the Impact of Minimum Taxes
- The paper constructs a novel database identifying local minimum taxes (LMTs) across 196 countries since 1980 by base and rate.
  - Database compiled manually from annual international tax law reports (IBFD, Deloitte) and digitized national legislation where needed.
  - Recorded applicable benchmark MT rate and binned regimes by base (turnover, assets, modified-income).
  - Merged with time series of CIT revenues (% of GDP) from IMF WEO (“taxes on income, profits, and capital gains, payable by corporations”) and statutory CIT rates from an IMF FAD database based largely on IBFD archives.
- Stylized aggregate findings:
  - Turnover-based MTs are the most popular historically and in recent years; number of countries with turnover-based MTs more than double those with other MTs.
  - Over the past 25 years, CIT rates tend to be higher on average in countries with MTs.
  - CIT rates are highest on average in countries with turnover-based MTs; countries with modified-income bases have CIT rates closer to those without MTs.
  - In terms of CIT revenues:
    - Only countries with turnover-based MTs have outperformed countries without MTs since 2000, collecting CIT revenues of up to almost 1 percentage point of GDP more on average.
    - Assets-based MT countries have broadly performed in line with countries without MTs (only more recently outperforming).
    - Modified income-based MT countries have tended to underperform in terms of revenues.

*Italicized source attribution.*

### 2. Number of countries with or without minimum taxes

### 2. Number of countries with or without minimum taxes

### Prevalence and types of minimum taxes (MTs)
- As of end-2018, 63 countries were identified as operating an MT.
- Of these 63 countries:
  - 31 countries use turnover as the favored tax base.
  - 10 countries have an assets base.
  - 9 countries have a modified income base.
- Countries may operate parallel systems using multiple bases (countries counted separately in each category in such cases).
- Examples of country-specific design choices:
  - Morocco differentiates MT rates by sector.
  - Ecuador’s (recently repealed) MT was used as a form of advance tax for the following year and was nonrefundable if not taxable income was generated.
  - Argentina restricted their MT base to 20 percent of gross assets for financial sector firms (until recently).
  - Colombia exempts financial sector firms from its MT.

### Typical MT rates and economic equivalence
- MT rates on turnover typically range between 0.1 and 3 percent.
- Asset-based MTs are typically levied at a rate of 1 percent.
- Equivalence examples provided:
  - A one percent MT rate on turnover is equivalent to a 30 percent tax on profits for a firm with a gross margin of around 3 percent.
  - A one percent MT on gross assets is equivalent to a 30 percent profits tax for a firm with a rate of return of three percent.
- Policy caution noted: firms incurring losses for reasons unconnected to excessive use of tax preferences or tax planning (for example, startups) can still be liable to pay MT; MTs are indifferent to investment incentives when setting the rate.

### Do MTs have any impact on Corporate Income Tax (CIT) revenues? — macro evidence approach
- The paper attempts to quantify the impact of imposing a local MT on CIT revenue, acknowledging potential endogeneity in the timing of MT implementation or repeal.
- Methodology used: local projection methods (Jordà 2005; Kilian & Lütkepohl 2017; Nakamura and Steinsson 2018; Ramey 2016; Stock and Watson 2018; Teulings and Zubanov 2014). This method:
  - Regresses contemporaneous and future changes in CIT revenue growth on changes in MT rates to estimate impulse response functions at various horizons.
  - Controls for country and time fixed effects, lags and forwards of MT shocks, and lagged revenue growth.
  - Includes additional controls in alternative specifications (e.g., changes in personal and corporate income tax rates, and real house prices).
- Data source: IMF’s WEO database (contains country-level series on “taxes on income, profits, and capital gains, payable by corporations” (CIT revenues) and broader “taxes on income, profits, and capital gains” (IT revenues)).
- Data limitations:
  - Of 86 country-year observations for MT introduction or reform, only 48 observations are available under the broader WEO income tax revenue variable and only 25 under the CIT revenue variable.

### Local projection specification (equation and variables)
- Estimated equation (as described in the source) uses:
  - ∆^h y_{i,t} = α_i^h + γ_t^h + β^h ∆s_{i,t} + Σ_{j=1}^p β_{1j}^h ∆s_{i,t−j} + Σ_{j=1}^p θ_{1j}^h ∆y_{i,t−j} + Σ_{j=0}^p θ_{2j}^h x_{i,t−j} + ε_{i,t}^h
  - Definitions:
    - ∆^h y_{i,t} ≡ y_{i,t+h} − y_{i,t−1}
    - ∆s_{i,t} ≡ s_{i,t} − s_{i,t−1}
    - y_{i,t} ≡ ln(CITrev_{i,t})
    - s_{i,t} ≡ ln(MT_{i,t})
  - Controls x_{i,t} include trade-weighted output growth of trading partners, political regime transition, and conflict in the domestic economy.
  - Regressions include country fixed effects α_i^h and time fixed effects γ_t^h.
  - β^h measures the contribution to the cumulative increase in the level of y_{i,t} at horizon h from a 1 percentage point increase in ∆s_{i,t}.

### Macro results and interpretation
- Figure 5 (described) plots responses of total tax revenues-to-GDP, IT revenues-to-GDP, and CIT revenues-to-GDP to:
  - (i) introduction and reforms of MTs; and
  - (ii) repeal of MTs.
- Average patterns described for advanced economies:
  - CIT revenues fall marginally on impact after MT introduction before increasing by the second year by almost two percentage points and remaining positive (Figure 5, panels 5 and 6).
  - Response of CIT revenues to repeal of the MT is largely negative—averaging around 5 percentage points before falling more dramatically over subsequent years.
  - Repealing an MT can more than offset the gains from its introduction.
- Statistical significance:
  - The responses of revenue-to-GDP to introduction or repeal of MTs are asymmetric but are not statistically different from zero (confidence intervals encompass the x‑axis).
  - Alternative specifications and subsamples (e.g., focusing just on turnover-based MTs) produced similar, statistically insignificant results.
- Identification caveats:
  - Results are unable to fully correct for endogeneity between revenues and MT introduction/repeal.
  - Cannot fully disentangle MT impact on numerator (revenues) and denominator (GDP), which could bias results.
  - An instrument for MT introduction (a measure of tax evasion related to informality from the International Labor Organization) was tested; impulse response functions remained statistically not significant.

### Firm-level analysis: data, definitions, and approach
- Motivation: with limited macro-level evidence, firm-level microdata is used to assess correlations between MT reforms and firm tax liability, profitability, and investment behavior.
- Caution: firm-level estimates are illustrative and should not be interpreted as causal due to potential endogeneity and unobserved confounders.
- Data sources and coverage:
  - Bureau Van Dijk’s Orbis database on historical industry financials (cashflow and global financials sub-databases).
  - Unconsolidated accounts only, for 176 countries from 1980 to 2018.
  - Merged with statutory corporate income tax rates and a novel historical dataset on MT bases, rates, and dates of introduction and repeal created by the authors.
- Core firm variables used (U.S. dollar values unless otherwise stated):
  - Gross sales
  - Operating profit before taxes
  - Taxes payable
  - Total assets
  - Number of employees
  - Primary sector descriptors
- Definition of average effective tax rate (AETR) used in the analysis:
  - AETR primarily defined as the ratio of taxes payable (from the income statement) to operating revenues from sales (turnover).
  - Rationale for this AETR definition:
    - Comparability across systems with and without MTs (not always based on corporate income measures).
    - Facilitates interpretation when reported profits are negative but assessed tax liability is positive.
    - Minimizes variability from manipulation of reported profits, as revenues are more consistently reported.
- Event window and sample restrictions:
  - Consider changes one year on either side of year t when MT was introduced, reformed, or repealed (years t−1, t, and t+1).
  - Restrict to economies with a firm count of more than 70 in those years.
  - Firm-level inclusion criteria:
    - At least 3 years of non-missing observations in Orbis.
    - AETRs with respect to turnover between 0 and 1.
    - Profit margins smaller than one in absolute value.
    - Operating revenues from sales greater than USD 10,000.
- Sample limitations: taxpayer data quality in countries with MTs (especially pre-treatment years) is poor and often inconsistent across years in Orbis; sample size was consequently restricted.

### Micro analysis scope and next steps
- Tables 4–7 (referenced) list countries by MT type satisfying criteria and present median values and summary statistics; after data cleaning and truncation the sample included a total of 17 MT-related observations (text cuts off at that point).
- Authors note intent for future research:
  - Expand analysis by exploiting variation in impact across industries as an instrument to address endogeneity in timing of reforms.

*Source: wpiea2021161-print-pdf - 2. Number of countries with or without minimum taxes*

### introduction events (country-year instances) and 12 MT repeal events spanning 1993-2018.

### wpiea2021161-print-pdf - introduction events (country-year instances) and 12 MT repeal events spanning 1993-2018

### Sample and data coverage
- Sample includes introduction events (country-year instances) and 12 MT repeal events spanning 1993-2018.
- Turnover-based MTs with available firm-level coverage in the merged Orbis dataset: Pakistan and Slovakia only.
- Reasons for limited coverage of turnover-based MTs:
  - Several turnover-based MT introductions/reforms date back to before 2007, when Orbis coverage was substantially smaller (e.g., Bangladesh and Bolivia).
  - Many turnover-based MT countries have little Orbis coverage (e.g., Guatemala and Ecuador).
- Reporting requirements introduced with MT reforms often produce a spike in the number of reporting companies (example: Morocco coverage over 50,000 firms starting in 2011, the year of MT reform).
- Most businesses in sample are SMEs:
  - turnover less than US$ 800 million and fewer than 3,000 employees.
  - average turnover typically between US$ 12–20 million.
  - average gross assets between US$ 20–30 million.
  - average profit margin between 6.5–11.5 percent.

### Identification strategy and empirical specification
- Narrow event-window approach: restricts sample to three years around MT reform date (introduction or change in rate/base). Identifying assumption: the MT is the only key change affecting companies in those years (acknowledged as unlikely due to concurrent fiscal/macroeconomic reforms).
- Baseline pooled specification:
  - Y_ik,t = β1 M_k,t + γ_i + ε_ik,t
  - Y_ik,t is firm outcome; M_k,t is dummy for presence of minimum tax regime; γ_i are country fixed effects.
- Samples considered:
  - Balanced panel of firms (smaller sample).
  - Unbalanced panel / repeated cross-section (only requirement: country-year with >70 non-missing observations).
- Sample exclusions:
  - Firms with reported negative operating revenues.
  - Firms with tax liability greater than 100 percent of operating revenues (AETR between 0 and 1 preserved).
  - Peru and Morocco excluded from some event study estimations due to insufficient pre-reform firm counts.

### Impact on firm average ETRs (AETRs)
- Pooled event-study average estimate:
  - Introduction of MTs associated with a 1.6 percent higher average ETR (Table 8).
  - Effect present in both balanced and unbalanced samples.
- Heterogeneity by MT type (pooled estimates and disaggregated):
  - Modified income-based MTs: increase in average ETR up to around 2.7 percentage points.
  - Assets-based MTs: increase in average ETR around 1.2 percentage points.
  - Turnover-based MTs: increase in average ETR around 1.0 percentage points (up to 1.5 percent for Slovakia’s largest threshold).
- Slovakia (turnover-based "tax license"):
  - For companies above largest turnover threshold of EUR 500,000: 1.5 percent average ETR increase.
  - Effect well distributed across sectors for large companies.
  - For companies below threshold: much larger AETR effect concentrated in leasing, real estate and retail sales.
  - Smaller companies: small positive employee growth, significant contraction in assets, no effect on pre-tax profitability.

### Effects on investment, employment, and reported profits
- Association with greater investment and employment in balanced panel for asset and modified income-based MTs (possible omitted variable bias or macro expansion timing).
- Two possible interpretations of positive reported profit effect:
  - MT introduction leads to increased truthful reporting of profits (firms formerly underreporting now constrained by MT).
  - Increased competition and exit of less productive firms, leaving more productive and profitable firms in sample.
- Repeated cross-section samples show slightly larger profitability parameter estimates than balanced panel, consistent with compositional change.

### Asymmetry: repeal/reduction events
- Robustness tests around MT repeal or rate reduction (Table 10):
  - Removal of MTs does not show immediate deterioration in firm-level AETRs (coefficient negative but statistically insignificant), except for balanced panel for assets-based MTs.
  - Possible explanations:
    - Improved revenue administration capacity during MT implementation (increased documentation and audit scrutiny) with persistent effects.
    - MTs introduced when likely to be binding; repealed/reduced when no longer binding relative to standard CIT.
  - Sample for repeal effects is sparser, especially for employee data, limiting parameter estimation.

### Exploiting sectoral structural characteristics
- Rationale:
  - Sectors with thinner profit margins likely disproportionately burdened by turnover-based MTs.
  - Sectors with large fixed assets relative to operating profits penalized more under an assets-based MT.
- Augmented specification incorporates firm sector j and structural AETR under MT based on sectoral profitability:
  - Y_ijk,t = β0 + β1 M_k,t + β2 Z_ijk,t + β3 τ_jk,t^MT(AETR) + β12 Z_ijk,t·M_k,t + β23 Z_ijk,t·τ_jk,t^MT(AETR) + β13 τ_jk,t^MT(AETR)·M_k,t + β123 Z_ijk,t·τ_jk,t^MT(AETR)·M_k,t + γ_k + ε_ijk,t
  - Y_ijk,t includes Profit_i,j,k,t / Revenue_i,j,k,t and other firm outcomes.
  - τ_jk,t^MT(AETR) = f(X_jk,t) is structural AETR under MT based on sectoral profitability X_jk,t = Profit_jk,t̄ / Revenue_jk,t̄.
  - Z_ijk,t is firm-specific indicator for liability to the LMT (binding).
- Findings from sectoral approach (repeated cross-section):
  - Introducing an MT leads to an average increase in AETR with respect to turnover of around 2.5 percent, controlling for binding liability and sector structural profit margins.
  - Profit-based AETR coefficients are much larger (by construction, profit-based AETR is multiple of turnover-based AETR).
  - Marginally stronger effect on firm AETR from increases in sectoral AETR for firms bound by the MT.
  - Instrumenting sectoral variation using structural profit margins suggests MT impact on AETRs not primarily driven by sector-specific variation.

### Counterfactual revenue-yield estimates (back-of-the-envelope)
- Hypothetical policies simulated on Orbis population (excluding certain offshore jurisdictions as classified in Beer et al. (2019)):
  - (i) 0.5 percent MT on turnover.
  - (ii) 1 percent MT on total assets.
- Estimated yields (median country, relative to current levels in latest available year for the median country):
  - Turnover-based MT (0.5 percent): could yield 7 percent additional tax revenues relative to current levels.
  - Assets-based MT (1 percent): could yield as much as 33 percent additional tax revenues relative to current levels.
- Scaled to national accounts (assuming Orbis firms represent similar proportion of overall corporate tax liability measured by WEO CIT revenues):
  - Median revenue increment: 0.2 percent of GDP for turnover-based MT; 0.9 percent of GDP for assets-based MT.
  - Median corporate income tax to GDP ratio in sample: 2.7 percent.
- Assumptions and caveats:
  - Assumes no behavioral changes and maintenance of current compliance levels.
  - Sample excludes low-tax jurisdictions; affected sample differs from OECD Pillar 2 population.
- Comparative notes:
  - Devereux and others (2020) estimate OECD Pillar 2 country-by-country could raise around 14 percent of existing corporate taxes paid by foreign controlled corporations (magnitude between the local minimum tax estimates here); blended multinational implementation estimated at approximately 4 percent of current corporate taxes on average.
  - OECD’s October 2020 estimates of Pillar 2 yield up to USD 70 billion (1.7–2.8 percent of total global CIT revenues), broadly within range when considering different denominators and inclusion of low-tax jurisdictions.

### Case study: Honduras (administrative panel data)
- Honduras introduced a dual MT in 2013 as part of anti-evasion measures:
  - Turnover-based MT: 1.5 percent of turnover (gross sales) if turnover > HNL 10 million (selected sectors subject to 0.75 percent).
  - Net assets tax (NAT): 1 percent on net assets if net assets > HNL 3 million.
  - Standard CIT liability: 25 percent of reported taxable profits.
  - Ultimate liability for firms with turnover > HNL 10 million and net assets > HNL 3 million:
    - CIT_liability_t = max(MT_t, NAT_t, CIT_t)
- Counterfactual simulation using 2013 AETRs held constant for firms continuously present in administrative data (2014–16):
  - Tax revenues from firms that became liable to the MT after introduction increase by almost 1 percent of GDP in 2016.
- Behavioral/documentation effects observed:
  - Clear bunching of reported profits around the indifference threshold between MT and CIT liabilities (profit margin of 6 percent).
  - Firms previously evading taxes by underreporting taxable profits faced higher ETRs over 2014–16.
  - More firms started truthfully reporting higher profit margins (Figures 7 and 8).
- Sectoral burden concern:
  - Gross-sales based MTs disproportionately burden sectors with structurally lower profit margins (e.g., food retail sales) relative to sectors with higher profit margins.

### Policy-relevant implications and interpretive notes
- Modified income-based MTs appear most effective at increasing average ETRs in the sample, followed by asset- and turnover-based MTs.
- MT introduction can raise reported profits via increased truthful reporting or compositional exit of less productive firms.
- MT repeal does not necessarily reverse AETR increases immediately, suggesting persistence of administrative and behavioral effects.
- Sectoral structural characteristics matter: turnover-based MTs can overweight low-margin sectors; assets-based MTs penalize asset-intensive sectors.
- Back-of-the-envelope revenue estimates indicate significant potential revenue gains (median increments of 0.2 and 0.9 percent of GDP for turnover- and assets-based MTs respectively, given assumptions), warranting policy consideration in revising international corporate tax architecture.
- Case evidence (Honduras) shows MTs can mobilize revenue and reduce underreporting but also raise distributional/sectoral burden concerns.

*Source: wpiea2021161-print-pdf - introduction events (country-year instances) and 12 MT repeal events spanning 1993-2018.*

### introduction of the MT induced more truthful declaration of taxable profits, especially in sectors

### wpiea2021161-print-pdf - introduction of the MT induced more truthful declaration of taxable profits, especially in sectors

### Summary of main findings
- The introduction of the MT induced more truthful declaration of taxable profits, especially in sectors that had previously declared low margins.
- Evidence suggests that at least part of the lower reported profits for several industries was due to underreporting to tax authorities, rather than truly lower profit margins.
- Scot and others (2020) corroborate these findings.
- Though the rates have been gradually reduced since March 2018 due to high lobbying pressure, the minimum tax remains in place in the country today and is recognized by the revenue administration as having assisted in their efforts to improve tax compliance.

### Empirical impacts and magnitudes
- In a partial equilibrium setting (ignoring behavioral implications), firm-level data suggests the introduction of the MT is associated with an increase in the average ETR at the firm level of just over 1.5 percentage points.
- Baseline results are stronger when controlling for variation in sector’s profit margins and firm-specific binding constraints — with an estimated increase in the average ETR with respect to turnover of 2.5 percent.
- Tax revenues from firms in Honduras that became liable to the MT following its introduction increased by almost 1 percent of GDP in 2016.
- Simple estimates of potential revenue yield of local minimum taxes (LMTs):
  - Turnover-based LMTs: median of 0.2 percent of GDP.
  - Assets-based LMTs: median of 0.9 percent of GDP.

### Comparative effectiveness by MT base
- Modified income-based MTs: most effective at increasing average ETRs.
- Assets-based MTs: second most effective; less susceptible to transfer pricing abuses and effective where tax administrations lack resources and expertise.
- Turnover-based MTs: third in effectiveness but common; simple and readily observable tax base that can limit tax avoidance and evasion.
- Assets- and modified taxable income bases are likely to better reflect true economic income than turnover, with potentially beneficial effects for neutrality and horizontal equity.
- Turnover and cash-flow MTs share the disadvantage of base erosion during periods of high inflation; assets-based MTs avoid this intra-year erosion.

### Design considerations and recommendations
- Base selection:
  - Turnover: simple and observable; poor predictor of profitability and ability to pay.
  - Cash flow: less distortionary due to investment expensing; may be too complex for smaller businesses.
  - Assets or modified taxable income: better proxy for regular income tax and more effective against profit shifting.
- Rate setting:
  - The rate of an MT should proxy the effective tax rate under the regular income tax but be kept relatively low (except for modified income-based MTs).
  - Turnover-based MT rate should proxy for the CIT burden on firms at the lower end of gross profit margins.
  - Assets-based MT rate should mimic the regular tax burden on a firm earning a rate of return sufficient to cover its cost of capital.
  - Consider sector differentiation where margins or rates of return vary significantly, weighing targeting benefits against added complexity and classification ambiguity for multi-activity firms.
- Crediting and carryforward:
  - MT payments should be creditable against future CIT liabilities with long carry forward periods to mitigate negative investment impacts and the “rough justice” nature of MTs.
  - Carry forward provisions help offset increases in marginal effective tax rates and support investment incentives, especially under high uncertainty over future returns.

### Aggregate implications and policy relevance
- Growing prevalence of minimum taxes worldwide, particularly turnover-based, associated with higher statutory CIT rates and revenue/GDP ratios.
- Unilateral, administratively simple local minimum taxes may effectively raise CIT revenue in source countries without significantly impairing corporate investment or profitability on aggregate.
- Findings provide support to the inbound rule segment of the OECD’s Pillar 2 proposal, while not addressing the outbound rule.
- Significant potential revenue yield (median 0.2 and 0.9 percent of GDP for turnover- and assets-based LMTs, respectively) warrants serious policy consideration under settings of imperfect tax capacity and tax competition.

### Limitations and avenues for further research
- Partial equilibrium analysis ignores behavioral implications; richer panels across countries and MT reform episodes could improve identification strategies.
- Few studies use administrative data to evaluate turnover-based MT regimes; this paper provides initial empirical evidence (including Honduras case).
- No prior studies identified estimating the impact of assets-based MTs on investment in fixed assets or average ETRs; this analysis provides an initial inroad.
- Future research should:
  - Explore welfare implications of MTs within optimal taxation models, contingent on specific sources of friction in second-best settings.
  - Integrate minimum taxes into more structured normative policy frameworks.
  - Combine the novel database of MT regimes with other tools to deepen empirical assessment.

*Source: wpiea2021161-print-pdf - introduction of the MT induced more truthful declaration of taxable profits, especially in sectors*

### APPENDIX I. - THE RATIONALE FOR A CORPORATE MINIMUM TAX

### APPENDIX I. - THE RATIONALE FOR A CORPORATE MINIMUM TAX

### A. Restricting the use of corporate tax preferences
- Purpose: guarantee a minimum CIT liability to shore up revenues undermined by excessive tax incentives and to guarantee a minimum average tax rate for all taxpayers.
- Observations on prevalence and trends:
  - 30 years ago no low-income Sub-Saharan African countries had tax free zones; now 50 percent do so.
  - 80 percent of sub-Saharan African countries offered tax holidays in 2005, compared to 40 percent in 1980 (Keen and Mansour 2008).
  - Nearly 5,400 special economic zones in operation in 147 countries as of 2018 (UNCTAD 2019).
  - More than 1,000 SEZs were established in the last five years and at least another 10 per cent more are expected to open in the coming years.
- Revenue cost evidence (estimates cited):
  - Corporate tax incentives in 15 Caribbean countries between 1995 and 2003 cost 5.5 percent of GDP in foregone revenues (Cubeddu et al. 2008).
  - Revenue forgone from CIT holidays may have exceeded 4 percent of GDP annually in the Eastern Caribbean Currency Union; abolishing these tax holidays would increase revenues by 3.3 percent of GDP (Chai and Goyal 2008).
  - Estimates for Latin America put the cost of preferential treatments under the income tax at 0.5–6 percent of GDP (Villela, Lemgruber, and Jorratt, 2010).
- Policy note: Eliminating unproductive and cost-ineffective tax preferences is first-best for base broadening, discouraging rent seeking, eliminating leakage, and improving CIT simplicity and transparency, but political economy factors often prevent such reforms in developing economies.

### B. Countering tax planning and avoidance
- Rationale: MT imposes a floor on a firm’s average effective tax rate and can limit revenue loss from tax planning/avoidance (e.g., corporate tax shelters), assuming the MT base is less manipulable than taxable earnings.
- Bases and vulnerability:
  - Turnover-based MT: base determined by gross earnings; unaffected by tax-deductible expense strategies (e.g., intra-group debt, excessive royalties/management fees) that reduce accounting or taxable profit, though susceptible to transfer-price manipulation that reduces recorded turnover.
  - Gross assets base: unaffected by profit shifting but introduces horizontal inequity across firms with different returns on investment.
  - Net assets MT: introduces bias toward debt financing and is vulnerable to related-party debt similar to standard CIT with interest deductibility.
- Multinational context:
  - Stock of FDI accounted for just over 30 percent of the GDP of developing economies in 2018, compared to 14 percent in 1990 and 25 percent in 2000 (UNCTAD 2019).
  - Identification and audit of transfer pricing is challenging for developing-country tax administrations with limited resources and expertise.
- Alternative or complementary measures:
  - Restrictions on interest deductibility (thin capitalization rules, maximum debt-equity ratios).
  - More comprehensive neutrality reforms: disallow interest deductibility (CBIT) or include allowance for corporate equity (ACE); note: no real-world examples of CBIT, only a handful apply ACE variants.
- Tailoring MT for compliance incentives:
  - Reduced MT rate for firms with strong compliance/record-keeping.
  - Examples: Romania repealed its MT of 5 percent of turnover for most firms but maintained it in sectors with prevalent non-compliance; Sierra Leone (before 2013) and Mauritania offered reduced MT rates to incentivize formal accounting or compliance.

### C. Complexity and compliance costs
- Trade-off: MTs can increase compliance and administrative complexity when operated alongside the regular CIT, requiring multiple sets of tax records.
- Evidence of compliance cost impact:
  - U.S. AMT complex base estimated to increase tax compliance costs for large corporates by 18 to 26 percent compared to those not subject to the tax (Slemrod and Blumenthal 1996). This likely understates total costs since it omits record-keeping burdens for firms potentially liable but not paying the AMT.
- Distributional effects of compliance costs:
  - Compliance costs tend to be regressive with respect to firm size; smaller firms face higher fixed compliance costs (European Commission 2018).
- Design guidance:
  - MT for small and medium sized firms should use a turnover base to avoid exacerbating regressivity; turnover is readily understood and often recorded for VAT/GST.
  - Larger firms could be subject to MTs based on more sophisticated measures (cash flow or modified taxable income) that better proxy economic income.
- Welfare trade-offs:
  - In second-best settings with limited tax capacity, switching from profit to turnover taxation can increase revenue collection efficiency without reducing aggregate profits (Best and others 2015). MTs have been effective in reducing tax evasion in the short term without tangible negative effects on aggregate firm profitability.
- Administrative implication: complexity of the AMT base determines the scale of compliance cost increases; if compliance costs are tax deductible they can reduce the revenue potential of the AMT.

### D. Reduced volatility and cyclicality of corporate income tax revenues
- Problem: CIT revenues are highly pro-cyclical in developing economies, increasing volatility of public spending and depressing public investment (Ebeke and Ehrhart, 2011).
- MT smoothing effect:
  - MT with indefinite carryforward turns the MT into a minimum payment rather than a pure tax, smoothing corporate tax payments over the business cycle (van Wijnbergen and Estache, 1999).
  - This smoothing can reduce pro-cyclicality of fiscal policy and aggregate uncertainty facing risk-averse investors, potentially encouraging investment.
- Cyclicality nuance:
  - MTs can be counter-cyclical from a firm perspective during recessions (firms pay higher effective tax rates than they otherwise would) and acyclical during expansions (standard CIT applies).
  - From a government revenue perspective, MTs can appear relatively acyclical, providing a stable revenue source in absolute value irrespective of the business cycle.
- Empirical example:
  - U.S. AMT payments rose dramatically during the 1990–91 recession, peaking at US$ 20.7 billion or 8.4 percent of total taxes paid that year (Carlson, 2005; Lyon, 1997).

*Source: APPENDIX I. - THE RATIONALE FOR A CORPORATE MINIMUM TAX (wpiea2021161-print-pdf)*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2021/english/wpiea2021161-print-pdf.pdf_