## wpiea2024078-print-pdf

## Source details

**Canonical URL:** [wpiea2024078-print-pdf](https://www.imf.org/-/media/files/publications/wp/2024/english/wpiea2024078-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2024/english/wpiea2024078-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2024/english/wpiea2024078-print-pdf.pdf.json)

---

### Executive Summary
- Presents a novel approach: a single-rate—broad-base VAT that feels de facto progressive to consumers — a Progressive VAT.
- Key claims:
  - The “traditional approach” (exclusions from the VAT base) is flawed: at best modest progressivity gains; at worst it can increase regressivity.
  - The “modern approach” (welfare transfers to compensate VAT effects) faces political economy, cash-flow, and welfare stigma limitations.
  - A Progressive VAT would re-pay VAT paid on consumption to lower-income households in real-time, at the moment of purchase, effectively eliminating regressivity while minimizing political economy, cash-flow, and welfare stigma obstacles.
  - The Progressive VAT also offers significant compliance incentives and can improve equity without significant efficiency losses, overcoming the typical trade-off between efficiency and equity.
- Empirical assessment uses household consumption survey data from Mozambique (low income) and South Africa (upper middle income).

### Background and Motivation
- VAT coverage and role:
  - VAT is applied in over 170 countries worldwide and collects approximately a fifth of global tax revenues.
- Public perceptions and behavioral framing:
  - Cognitive biases (framing, tax disaggregation bias, mental accounting) shape fairness perceptions, making addressing regressivity within VAT itself policy-relevant.

### Limitations of Traditional and Modern Approaches
- Traditional approach (exemptions, zero or reduced rates):
  - Nearly universal use of base exclusions to address regressivity.
  - Significant costs: revenue loss, qualification problems, loss of neutrality, distortions to competition, tax planning/avoidance, increased compliance and administrative costs.
  - Effectiveness depends on pass-through to consumer prices; empirical evidence casts doubt on full pass-through.
  - To improve incidence efficiently via exclusions would require targeting inferior goods, which are rare in consumption baskets.
  - High revenue forgone often results in larger subsidies to better-off households; informality exacerbates the problem.
- Modern approach (welfare transfers on the expenditure side):
  - Technically superior in principle but limited in practice due to political economy difficulties and practical implementation problems (cash-flow timing, stigma).

### Progressive VAT: Concept and Design
- Core idea:
  - Use real-time electronic invoicing/tracking technology with a validation element so VAT charged at purchase is re-paid in real-time to lower-income households.
  - Legal VAT base remains single-rate, broad-base; progressivity comes from point-of-sale compensatory refunds.
- Technological foundations:
  - Real-time invoice tracking repurposed from anti-fraud validation to delivering distributive transfers in real-time. Systems in South Korea, Israel, Portugal, Russia, Slovenia, Uzbekistan, and several Latin American countries provide precedents.
- Political economy and behavioral advantages:
  - Consumers experience VAT as progressive, addressing fairness perceptions influenced by framing and tax disaggregation biases.
  - Minimizes welfare stigma by delivering compensation at moment of purchase.
  - Reduces cash-flow problems for low-income households compared with ex-post transfers.

### Empirical Assessment: Mozambique and South Africa (consumption patterns and VAT reliefs)
- Consumption shares and patterns:
  - Food accounts for 54 percent of total expenditures on average in Mozambique.
  - Food accounts for 30 percent of total expenditures on average in South Africa.
  - Food’s relative importance by decile:
    - Mozambique: from 73 percent of total consumption for the lowest decile to 14 percent for the top decile.
    - South Africa: from 41 percent for the lowest decile to 8 percent for the top decile.
- Share of consumption relieved from tax (VAT reliefs):
  - Mozambique average: 54 percent of consumption benefits from VAT reliefs; declines from 65 percent for the lowest decile to 33 percent for the top decile.
  - South Africa average: 34 percent of consumption benefits from VAT reliefs; declines from 42 percent for the lowest decile to 11 percent for the top decile.
- Incidence under typical exclusions:
  - Mozambique: VAT imposes almost a flat burden of around 6.1 percent across the income distribution with a slight uptick for the top decile — to 7.1 percent — implying only modest progressivity.
  - South Africa: slightly higher level of progressivity is achieved (detailed income-profile figures presented in the body of the paper).
- Impact of VAT reliefs on incidence and distribution:
  - Reliefs can appear progressive as a share of income while in absolute terms higher-income households receive larger benefits.
  - Examples:
    - Mozambique: over 35 percent of the potential VAT revenue is forgone due to VAT reliefs; the top decile receives 26 percent of the subsidy while the bottom four deciles receive 22 percent.
    - South Africa: the top decile receives 44 percent of the total VAT subsidy, more than the bottom seven deciles (26 percent).

### Design Scenarios (three) and Fiscal-Equity Trade-offs
- Scenario definitions:
  - Scenario 1 — Cut-off threshold: households with income below a threshold bear no VAT (threshold aligned with median income of a chosen decile). This is the “true” Progressive VAT.
  - Scenario 2 — Universal subsidy VAT: all consumers receive VAT compensation equal to VAT charged on purchases up to a threshold amount.
  - Scenario 3 — Negative VAT: all consumers receive a VAT subsidy equal to the threshold amount regardless of income or consumption.
- Trade-off principle:
  - Revenue and equity objectives are inversely correlated — “the more revenue is forgone, i.e., used for equity-improving subsidy, the higher levels of progressivity can be achieved.”
- Preferred option:
  - Cut-off threshold (Scenario 1) is preferred because it can yield inequality gains while improving or preserving revenue yield; universal subsidy or negative VAT may deliver higher inequality gains but at worse fiscal cost.

### Stylized Revenue and Distributional Impacts (selected empirical results)
- Threshold = median income of the third decile (stylized results):
  - Scenario 1 (cut-off threshold):
    - Mozambique: increases revenue yield by 40 percent (equivalent to 3 percent of GDP).
    - South Africa: increases revenue yield by 25 percent (equivalent to 2.6 percent of GDP).
    - Progressivity: Scenario 1 achieves a small degree of progressivity relative to current VAT.
  - Scenarios 2 and 3:
    - Revenue yield: would not materially change revenue yield compared with benchmark in the third-decile threshold example.
    - Progressivity (Gini coefficient changes):
      - Mozambique: Scenario 2 changes Gini by one percentage point; Scenario 3 changes Gini by nearly 1.5 percentage point.
      - South Africa: Scenario 2 changes Gini by 0.9 percentage point; Scenario 3 changes Gini by 1.25 percentage point.
- Sensitivity to a fifth-decile threshold:
  - Scenario 1 with fifth-decile threshold:
    - Mozambique: revenue gain reduced from 40 to 27 percent of current collections.
    - South Africa: revenue gain reduced from 25 to 15 percent.
    - Progressivity: cut-off at fifth decile changes the Gini coefficient by one percentage point in Mozambique and by 0.9 percentage point in South Africa.
- Country-specific modeled impacts (selected results):
  - Mozambique:
    - Negative VAT (Scenario 3) would hurt government revenue by 31 percent.
    - Progressive VAT design can improve the Gini coefficient by 2.5 percentage points under Scenario 3.
    - A cut-off threshold approach yields revenue gains and an inequality reduction example referenced as 1.3 percentage points in Mozambique.
  - South Africa:
    - Scenarios 2 and 3 would lead to revenue loss of 14 and 25 percent, respectively, compared to current collections.
    - Progressivity improvements of 1.2 and 2.4 percentage points for Scenario 2 and 3, respectively.

### Main Gains, Spillovers, and Risks
- Main advantages of Progressive VAT:
  - Overcomes trade-offs between efficiency and equity by improving equity with minimal efficiency losses.
  - Minimizes political economy obstacles faced by ex-post transfers and base exclusions.
  - Reduces welfare stigma by providing compensation at the point of purchase.
  - Strengthens compliance incentives through integrated real-time invoice validation and matching.
  - Potentially lower administrative and operational costs relative to broad, permanent exemptions that require complex definitions and enforcement.
- Compliance and formality incentives:
  - Incentivizes customers to request invoices, extending self-enforceability to B2C transactions.
  - Competitive pressure may encourage formalization over time.
- Fraud, behavioral, and operational risks:
  - Risk of sharing/sale of digital wallets or PIT numbers by eligible households; limited by maximum compensation thresholds.
  - Increased incentive to under-report income because VAT advantage supplements income tax benefits; empirical evidence on bunching around VAT registration thresholds suggests risk may be significant.
  - Mitigation: adopt progressive compensation with multiple thresholds (full compensation for lowest deciles, partial for medium incomes).
  - Privacy and cybersecurity risks: real-time technologies are intrusive—requires strong cyber-security and legal safeguards to limit data misuse and biased algorithms.
- Other spillovers:
  - Cross-border shopping: Progressive VAT covers domestic supplies and is available only to residents, which protects and may export the consumption tax base.
  - Potential gender-equality gains inferred from digital-money experiences (e.g., Kenya).

### Modeling Approach, Key Assumptions, and Limitations
- Methodology:
  - Microsimulation reconciling household survey data with private final consumption from national accounts through extrapolation.
  - Steps include deriving net household expenditure by subtracting VAT from gross spending per spending category and income decile, using VAT treatment data, value added estimates from supply-use tables, and assumptions on informality.
  - Calculate back gross expenditure under three VAT scenarios: (i) full taxation; (ii) VAT with exclusions as currently imposed; (iii) Progressive VAT in three designs.
- Important assumptions and limitations:
  - Analysis is static: no behavioral responses, no changes in consumption patterns, and no changes in non-compliance are modeled.
  - Income measurement challenge: household survey income is unreliable, especially in lowest deciles.
  - Income imputation: reported income substituted with incomes derived from consumption using marginal propensity to consume assumptions:
    - Assumption: marginal propensity to consume implied to be 1 for lower-band deciles and a declining propensity for upper five deciles to a weighted average of 0.7.
  - Extrapolating survey consumption to national accounts improves VAT base approximation but may still underestimate revenue gains.
- Policy-relevant modeling implication:
  - Progressive compensation with multiple thresholds can reduce incentives for income under-reporting.

### Policy Implications and Operational Considerations
- Digital tax technologies reduce technical and political obstacles to implementing a broad-based single-rate VAT while protecting lower-income households in real time.
- Where targeting capacity is weak, a universal subsidy or simple cut-off may be pragmatic; where targeting is feasible, targeted real-time compensation can improve progressivity while preserving revenue gains.
- Merging removal of VAT concessions with immediate compensatory transfers (real-time subsidy at purchase) can mitigate status quo bias and loss aversion, improving political feasibility.
- Real-time systems offer additional benefits for compliance and fraud reduction, reinforcing revenue outcomes.
- Caveats:
  - Political economy choices (eligibility cut-offs, unit of assessment) matter.
  - Advances in digital tax technology are a pre-condition for feasible implementation.
  - Risk management is essential: cybersecurity, data protection, fraud mitigation, and privacy safeguards are required.

### Conclusion
- A Progressive VAT — a single-rate, broad-base VAT with real-time, point-of-sale compensation to lower-income households — is presented as a practical, technology-enabled solution to VAT regressivity.
- Empirical modeling for Mozambique and South Africa suggests significant distributive improvements and favorable revenue outcomes relative to traditional exclusions.
- The approach promises to minimize political economy, cash-flow, and stigma barriers while delivering compliance and neutrality benefits.

*Source: IMF Working Paper — Designing a Progressive VAT (excerpted content).*

### 2023. We are grateful to the organisers of these events, and to the many participants for the comments received therein.

### Designing a Progressive VAT

### Executive Summary
- Presents a novel approach to addressing VAT incidence: a single-rate—broad-base VAT that feels de facto progressive to consumers — a Progressive VAT.
- Key claims:
  - The “traditional approach” (exclusions from the VAT base) is flawed: at best modest progressivity gains; at worst it can increase regressivity.
  - The “modern approach” (welfare transfers to compensate VAT effects) faces political economy, cash-flow, and welfare stigma limitations.
  - A Progressive VAT would re-pay VAT paid on consumption to lower-income households in real-time, at the moment of purchase, effectively eliminating regressivity while minimizing political economy, cash-flow, and welfare stigma obstacles.
  - The Progressive VAT also offers significant compliance incentives and can improve equity without significant efficiency losses, overcoming the typical trade-off between efficiency and equity.
- Empirical assessment uses household consumption survey data from Mozambique (low income) and South Africa (upper middle income).

### Background and Motivation
- VAT is applied in over 170 countries worldwide and collects approximately a fifth of global tax revenues.
- VAT’s popularity rests on revenue mobilization, perceived efficiency, and neutrality; its chief weakness is perceived regressivity.
- Public perceptions of tax fairness are shaped by cognitive biases (framing, tax disaggregation bias, mental accounting), explaining why addressing regressivity within VAT itself is policy-relevant even when tax-mix arguments exist.

### Limitations of the Traditional and Modern Approaches
- Traditional approach (exemptions, zero or reduced rates):
  - Nearly universal use of base exclusions to address regressivity.
  - Significant costs: revenue loss, qualification problems, loss of neutrality, distortions to competition, tax planning/avoidance, increased compliance and administrative costs.
  - Effectiveness depends on pass-through to consumer prices, but empirical evidence casts doubt on full pass-through.
  - To improve incidence efficiently via exclusions would require targeting inferior goods, which are rare in consumption baskets.
  - High revenue forgone often results in larger subsidies to better-off households; informality exacerbates the problem.
- Modern approach (welfare transfers on the expenditure side):
  - Technically superior in principle but limited in practice due to political economy difficulties and practical implementation problems (cash-flow timing, stigma).

### Progressive VAT: Concept and Design
- Core idea:
  - Adapt existing real-time electronic invoicing/tracking technology, introducing a validation element so that VAT charged at purchase is re-paid in real-time to lower-income households.
  - The legal VAT base remains a single-rate, broad-base VAT; progressivity arises from real-time compensatory refunds at the point of sale rather than from differentiated tax rates.
- Technological foundations:
  - Real-time invoice tracking has spread as an anti-fraud mechanism (first in South Korea and Israel), and is now used in countries such as Portugal, Russia, Slovenia, Uzbekistan; several Latin American countries use advanced e-invoicing.
  - The Progressive VAT repurposes this technology from anti-fraud validation to delivering distributive transfers in real-time.
- Political economy and behavioral advantages:
  - Consumers/taxpayers will experience VAT as progressive, addressing fairness perceptions influenced by framing and tax disaggregation biases.
  - Minimizes welfare stigma by delivering compensation at the moment of purchase.
  - Reduces cash-flow problems for low-income households compared with ex-post transfers.

### Empirical Assessment: Mozambique and South Africa (consumption patterns and distributional impact of base exclusions)
- Data and modeling:
  - Uses household consumption survey data for Mozambique and South Africa; model and assumptions detailed in Appendix.
- Consumption patterns:
  - Spending on food accounts for:
    - 54 percent of total expenditures on average in Mozambique.
    - 30 percent of total expenditures on average in South Africa.
  - Food’s relative importance declines with income:
    - Mozambique: from 73 percent of total consumption for the lowest decile to 14 percent for the top decile.
    - South Africa: from 41 percent for the lowest decile to 8 percent for the top decile.
  - Share of consumption relieved from tax (VAT reliefs):
    - Mozambique average: 54 percent of consumption benefits from VAT reliefs; declines from 65 percent for the lowest decile to 33 percent for the top decile.
    - South Africa average: 34 percent of consumption benefits from VAT reliefs; declines from 42 percent for the lowest decile to 11 percent for the top decile.
- Incidence under typical exclusions:
  - Mozambique: VAT imposes almost a flat burden of around 6.1 percent across the income distribution with a slight uptick for the top decile — to 7.1 percent — implying only modest progressivity.
  - South Africa: a slightly higher level of progressivity is achieved (detailed income-profile figures are presented in the body of the paper).

### Revenue, Distributional Impact, and Stylized Results
- The paper estimates distributional and revenue impacts of Progressive VAT versus traditional base exclusions using the same microsimulation approach.
- Stylized outcomes (described in figures and text):
  - Progressive VAT can eliminate regressivity while preserving or improving revenue yield compared to high levels of exemptions.
  - Compensation under Progressive VAT can be targeted in real-time to households below chosen cut-offs (modeled in the paper with examples such as 3rd and 5th decile cut-offs).
  - Figures illustrate Progressive VAT incidence for Mozambique and South Africa and changes in revenue yield and Gini coefficient under different cut-offs.

### Main Gains and Positive Spillovers
- Key advantages of a Progressive VAT:
  - Overcomes trade-offs between efficiency and equity by improving equity with minimal efficiency losses.
  - Minimizes political economy obstacles faced by ex-post transfers and base exclusions.
  - Reduces welfare stigma by providing compensation at the point of purchase.
  - Strengthens compliance incentives through integrated real-time invoice validation and matching.
  - Potentially lower administrative and operational costs relative to broad, permanent exemptions that require complex definitions and enforcement.
- Caveats and design considerations discussed:
  - Political economy considerations remain important (design choices, eligibility cut-offs).
  - Advances in digital tax technology are a pre-condition for feasible implementation.
  - The paper evaluates design options and presents trade-offs (e.g., cut-off choices and revenue vs. distributional outcomes).

### Conclusion
- A Progressive VAT — a single-rate, broad-base VAT with real-time, point-of-sale compensation to lower-income households — is presented as a practical, technology-enabled solution to VAT regressivity.
- Empirical modeling for Mozambique and South Africa suggests significant distributive improvements and favorable revenue outcomes relative to traditional exclusions.
- The approach promises to minimize political economy, cash-flow, and stigma barriers while delivering compliance and neutrality benefits.

*Source: IMF Working Paper — Designing a Progressive VAT (2023).*

### 6.8 for the bottom decile to 8.2 percent for the top decile. Such outcome is due to the amount of subsidy

### wpiea2024078-print-pdf - 6.8 for the bottom decile to 8.2 percent for the top decile. Such outcome is due to the amount of subsidy

### Impact of VAT reliefs on incidence and distribution
- Reliefs can appear progressive when measured as a share of income (amount of subsidy as a share of income decreases with income level), while in absolute terms higher-income households receive larger benefits (absolute amount increases with income).
- Examples:
  - Mozambique: over 35 percent of the potential VAT revenue is forgone due to various VAT reliefs; the top decile receives more (26 percent) than the bottom four deciles (22 percent).
  - South Africa: the top decile receives 44 percent of the total VAT subsidy, more than the bottom seven deciles (26 percent).
- These empirical findings align with other studies cited in the source.

### Modern approach to addressing regressivity (welfare transfers instead of exemptions)
- Core idea: address VAT regressivity outside the VAT system via progressive income taxes or welfare transfers.
- Evidence/experience:
  - New Zealand implemented the ‘Modern VAT’ approach relying on welfare transfers.
  - Variants: some countries apply cash transfers (public finance mix) while others use tax-system transfers (tax mix), e.g., Canada’s income tax credit.
- Observed practice: many countries adopt hybrid systems (some exclusions remain while welfare transfers are introduced).

### Political economy constraints to adopting the modern approach
- Two main constraints explain persistence of exemptions/reduced rates:
  - Targeting difficulties: limited administrative capacity in some middle and low-income countries.
  - Political dynamics: information asymmetry and trust issues between voters and policymakers, cognitive biases, status quo bias, and lobbying by special interests.
- Behavioral factors:
  - Voters are often rationally ignorant about complex consumption tax effects.
  - Fairness-centric intuition favors reduced rates on essentials; counterintuitive evidence that concessions may be regressive is less persuasive.
  - Two-step reforms (remove benefit then add compensation) evoke loss aversion; merging steps can mitigate resistance.

### Advances in digital tax technology enabling alternative solutions
- Real-time systems and data warehouses linking B2C transactions allow identification and validation of taxpayers and purchases at point of sale.
- Notable outcomes and system features:
  - Russia: reported decrease in VAT gap from 20 to 1 percent.
  - Portugal: reported decrease in VAT gap from 16 to 12 percent; Portuguese system links invoices to personal tax numbers and offers consumer incentives (monthly lottery and VAT deduction from income tax assessment).
  - South Korea, Latin America, Israel, Slovenia, Uzbekistan: examples of invoice/e-invoicing and real-time reporting systems; Uzbekistan’s AI generates individual risk factors for VAT-registered businesses.
- Digital welfare delivery successes during COVID-19:
  - Mobile money adoption (e.g., M-PESA): credited with lifting 2 percent of Kenyan households out of poverty and decreasing gender inequality.
  - Pandemic-era transfers: Argentina, Pakistan and Peru—new support programs covered one-third of the population; Philippines—over 70 percent received emergency transfers; Brazil—beneficiaries rose from 14 to 50 million.
  - Thailand’s State Welfare Smartcard: five percentage points of the 7 percent VAT returned to e-wallets, one percentage point to National Savings Fund or savings, rebate capped at around US$16 per month; five months after launch, 14.5 million smartcard holders and the Government had transferred around US$650 thousand to e-wallets and around US$120 thousand to savings accounts.
  - Colombia: March 2020 welfare transfer program using mobile money to compensate a VAT base broadening reform; early evidence shows positive effects on food access and financial health.

### Concept and core design of the Progressive VAT
- Core design elements:
  - Full taxation of all consumption at a single VAT rate (no merit exemptions or reduced rates).
  - Payment of a VAT compensation subsidy.
  - Digital mechanism allowing real-time payment of the subsidy at the moment of purchase.
- Key design options (examples):
  - Unit of assessment: individual or household.
  - Objective scope: targeted to low-income citizens or universal if targeting not possible.
  - Payment method: integrated personalized tax system with centralized bank transfers where possible; or digital wallet/mobile devices where bank integration is infeasible.
  - Compensation design choices: full compensation for VAT paid up to a ceiling; different compensation levels by decile; thresholds based on median income of a decile.

### Three design scenarios considered
- Scenario 1: Cut-off threshold — households with income below a threshold bear no VAT (threshold aligned with median income of a chosen decile).
- Scenario 2: Universal subsidy VAT — all consumers receive VAT compensation equal to VAT charged on purchases up to a threshold amount.
- Scenario 3: Negative VAT — all consumers receive a VAT subsidy equal to the threshold amount regardless of income or consumption.
- Note: Scenario 1 reflects the “true” Progressive VAT; Scenarios 2 and 3 resemble alternative targeted social programs.

### Revenue and distributional impacts (stylized applications and sensitivity)
- General relation: higher threshold amounts increase cost of compensation and progressivity; revenue cost lowest for Scenario 1 and highest for Scenario 3.
- Stylized and empirical results for Mozambique and South Africa (threshold = median income of the third decile):
  - Cut-off threshold (Scenario 1):
    - Mozambique: increases revenue yield by 40 percent (equivalent to 3 percent of GDP).
    - South Africa: increases revenue yield by 25 percent (equivalent to 2.6 percent of GDP).
    - Progressivity: Scenario 1 achieves a small degree of progressivity relative to current VAT.
  - Universal subsidy VAT (Scenario 2) and Negative VAT (Scenario 3):
    - Revenue yield: would not materially change revenue yield compared with benchmark in the third-decile threshold example.
    - Progressivity (Gini coefficient changes):
      - Mozambique: Scenario 2 changes Gini by one percentage point; Scenario 3 changes Gini by nearly 1.5 percentage point.
      - South Africa: Scenario 2 changes Gini by 0.9 percentage point; Scenario 3 changes Gini by 1.25 percentage point.
- Sensitivity to threshold (cut-off at fifth decile example):
  - Cut-off threshold (Scenario 1) with fifth-decile threshold:
    - Mozambique: revenue gain reduced from 40 to 27 percent of current collections.
    - South Africa: revenue gain reduced from 25 to 15 percent.
    - Progressivity: cut-off at fifth decile changes the Gini coefficient by one percentage point in Mozambique and by 0.9 percentage point in South Africa.
  - The universal subsidy VAT (Scenario 2) at higher thresholds becomes closer in cost to Scenario 1; at lower thresholds it is closer to Scenario 3.

### Policy implications and operational considerations
- Digital tax technologies reduce the technical and political obstacles to implementing a broad-based single-rate VAT while protecting lower-income households in real time.
- Where targeting capacity is weak, a universal subsidy or simple cut-off may be pragmatic design choices; where targeting is feasible, targeted real-time compensation can improve progressivity while preserving revenue gains.
- Merging removal of VAT concessions with immediate compensatory transfers (real-time subsidy at purchase) can mitigate status quo bias and loss aversion, improving political feasibility.
- Real-time systems also offer additional benefits for compliance and fraud reduction, which can reinforce revenue outcomes.

*Source: https://www.imf.org/-/media/files/publications/wp/2024/english/wpiea2024078-print-pdf.pdf*

### 1.3 percentage points in Mozambique, while the negative VAT (Scenario 3) would hurt the government revenue

### wpiea2024078-print-pdf - 1.3 percentage points in Mozambique, while the negative VAT (Scenario 3) would hurt the government revenue

### Design scenarios and fiscal-equity trade-offs
- Three Progressive VAT designs modeled:
  - (1) threshold cut-off: consumers below a chosen threshold (corresponding to income decile) do not bear VAT; those above the threshold do not receive any VAT subsidy.
  - (2) universal subsidy: all consumers receive a subsidy equal to the VAT actually paid but no more than a chosen ceiling amount.
  - (3) negative VAT: all consumers receive the same amount of subsidy, potentially exceeding VAT actually paid by low-income households.
- Trade-off across designs: revenue and (tax) equity objectives are inversely correlated—“the more revenue is forgone, i.e., used for equity-improving subsidy, the higher levels of progressivity can be achieved.”
- Preferred option: a cut-off threshold with possible differentiated compensation levels depending on tax administrative capacity, because:
  - It yields inequality gains and revenue gains, improving the public finance mix.
  - Universal subsidy or negative VAT may deliver higher inequality gains but at the cost of less favorable public finance mix (less revenue → less scope for public expenditure).

### Country-specific modeled impacts (selected results)
- Mozambique:
  - Negative VAT (Scenario 3) would hurt government revenue by 31 percent.
  - Progressive VAT design can improve the Gini coefficient by 2.5 percentage points under Scenario 3.
  - A cut-off threshold approach yields revenue gains and inequality reduction (example reference: 1.3 percentage points improvement mentioned).
- South Africa:
  - Scenarios 2 and 3 would lead to revenue loss of 14 and 25 percent, respectively, compared to current collections.
  - Progressivity improvements of 1.2 and 2.4 percentage points for Scenario 2 and 3, respectively.
- General statement: “Regardless of what option is taken, however, there is ample scope to improve progressivity compared to the current VAT design.”

### Main gains of the Progressive VAT
- Equity gains:
  - Progressive VAT reduces inequality and goes “beyond the mere decrease in regressivity.”
  - Potential gender-equality gains inferred from Kenyan experience with digital money.
  - Real-time compensation reduces welfare stigma because it feels like an exemption rather than a welfare payment and is not observable at point of sale.
  - Compensation via real-time technology is de facto indexed to prices, addressing deficient indexation problems in some welfare systems.
- Revenue and efficiency gains:
  - Compared to the traditional VAT, Progressive VAT would result in significant revenue gains.
  - Compared to the modern VAT, short-term administrative costs may increase, but medium-term small gains are conceivable from automation.
  - Progressive VAT is de jure a broad-base, single-rate VAT with compensation mechanism; de facto it feels like targeted zero-rating to low-income households.
- Political economy gains:
  - Less exposure to loss-aversion and distrust from consumers because loss and benefit are concurrent.
  - Potentially smaller and easier-to-overcome political resistance to targeted protection for the poorest.

### Spillovers, compliance incentives, and risks
- Compliance and formality incentives:
  - Progressive VAT incentivizes customers to request invoices, extending self-enforceability to B2C transactions.
  - Incentive against informal suppliers: competitive pressure may encourage formalization over time.
- Fraud and behavioral risks:
  - Risk of sharing/sale of digital wallets or PIT numbers by eligible households; limited by maximum compensation thresholds.
  - Increased incentive to under-report income because VAT advantage supplements income tax benefits; empirical evidence on bunching around VAT registration thresholds suggests the risk may be significant.
  - Mitigation: adopt progressive compensation with multiple thresholds (full compensation for lowest deciles, partial for medium incomes).
- Other spillovers:
  - Cross-border shopping: Progressive VAT covers domestic supplies and is available only to residents, which protects and may export the consumption tax base.
  - Privacy and cybersecurity risks: real-time technologies are intrusive—requires strong cyber-security and legal safeguards to limit data misuse and biased algorithms.

### Modeling approach and key assumptions
- Method: microsimulation reconciling household survey data with private final consumption from national accounts through extrapolation.
- Steps:
  - Derive net household expenditure by subtracting VAT from gross spending per spending category and income decile.
  - Use publicly available VAT treatment data, value added estimates from supply-use tables (assumed otherwise), and assumptions on informality.
  - Calculate back gross expenditure under three VAT scenarios: (i) full taxation; (ii) VAT with exclusions as currently imposed; (iii) Progressive VAT in three designs described above.
- Important modeling choices and limitations:
  - Analysis is static: no behavioral responses, no changes in consumption patterns, and no changes in non-compliance are modeled.
  - Income measurement challenge: household survey income is unreliable, especially in lowest deciles.
  - Income imputation approach: substitute reported income with incomes derived from consumption data using marginal propensity to consume.
    - Assumption: marginal propensity to consume implied to be 1 for lower-band deciles and a declining propensity for upper five deciles to a weighted average of 0.7.
  - Extrapolating survey consumption to national accounts improves approximation of VAT base but may still underestimate revenue gains.
- Policy-relevant modeling implication: progressive compensation with multiple thresholds can reduce incentives for income under-reporting.

### Conclusion and policy implications
- Progressive VAT aims to address VAT regressivity without the efficiency costs of traditional base exclusions or the political-economy/implementation costs of modern welfare transfers.
- Digital real-time technologies enable de facto progressive consumption taxes that can be implemented without compromising efficiency or neutrality, though risk management (cybersecurity, data protection, fraud mitigation) is essential.
- The paper signals that consumption taxes need not be inherently regressive and presents a practicable path to design taxes that are both efficient and equitable.

*IMF WORKING PAPERS Designing a Progressive VAT — excerpted content*

### References

### wpiea2024078-print-pdf - References

### VAT design, incidence, and redistribution
- Ainsworth, R.T. (2012), “Refund Fraud? Real-Time Solution! Digital Security Borrowed from the VAT (Brazil, Quebec & Belgium)”. Boston University School of Law Working Paper 12-15.  
- ⎯ (2006), “The Digital Vat (D-Vat)”. Boston University School of Law Working Paper 06-22.  
- Alatas, V., A. Banerjee, R. Hanna, B. Olken and J. Tobias (2012). American Economic Review 102(4), 1206-1240.  
- Alt, J., I. Preston and L. Sibieta (2010), “The Political Economy of Tax Policy”. In S. Adams et al. (eds), Dimensions of Taxation (Oxford: Oxford University Press), Ch.13.  
- Bachas, P., L. Gadenne and A. Jensen (2021), “Informality, Consumption Taxes and Redistribution”. NBER Working Paper 27429.  
- Barreix, A., M. Bes and R. Jeronimo (2012), “Solving the Impossible Trinity of Consumption Taxes Personalized VAT”. Inter-American Development Bank.  
- ⎯ M. Bès, O. Fonseca, M.B. Fonteñez, D. Morán, E. Pineda and J. Roca (2022), “Revisiting Personalized VAT: A Tool for Fiscal Consolidation with Equity”. Inter-American Development Bank, Discussion Paper IDB-DP-939.  
- Benedek, D. et al. (2020), “Varieties of VAT Pass Through”. International Tax and Public Finance 27, 890.  
- Besley, T. and S. Coate (1992), “Understanding Welfare Stigma: Taxpayer Resentment and Statistical Discrimination”. Journal of Public Economics 48(2), 165-183.  
- Bird, R. (2010), “Visibility and Accountability — Is Tax-Inclusive Pricing a Good Thing?”. Canadian Tax Journal 58, 63.  
- Bird, R. and P.P. Gendron (2007), The VAT in Developing and Transitional Countries (Cambridge University Press).  
- Blasco, J., E. Guillaud and M. Zemmour (2023), “The inequality impact of consumption taxes: An international comparison”. Journal of Public Economics 222, 104897.  
- Boeters, S. et al. (2010), “Economic effects of VAT reforms in Germany”. Applied Economics 42(17), 2165.  
- Buettner, T. and B. Madharova (2021), “Unit Sales and Price Effects of Pre-Announced Consumption Tax Reforms: Micro-level Evidence from European VAT”. American Economic Journal: Economic Policy 13(3), 103-134.  
- Caspersen, E. and G. Metcalf (1994), “Is a Value Added Tax Regressive? Annual Versus Lifetime Incidence Measures”. National tax Journal 47(4), 731.  
- Carlson, G.N. and M.K. Patrick (1989), “Addressing the Regressivity of a Value-Added Tax”. National Tax Journal 42(3), 339.  
- Correia, I. (2010), “Consumption Taxes and Redistribution”. American Economic Review 100, 1673-1694.  
- Cnossen, S. (1989), “The Value-added Tax: Key to a Better Tax Mix”. Australian Tax Forum 6(3), 265.  
- Deaton, A. (2019), The Analysis of Household Surveys : A Microeconometric Approach to Development Policy. Washington, DC: World Bank.  
- de la Feria, R. (2023), “The UK VAT at 50: The Good, the Bad and the Ugly”. British Tax Review 30(3), 307-321.  
- ⎯ (2020), “Tax Fraud and Selective Law Enforcement”. Journal of Law and Society 47(2), 240-270.  
- ⎯ and A. Grau Ruiz (2022), “The Robotisation of Tax Administration” in A. Grau (ed), Interactive Robotics: Legal., Ethical., Social and Economic Aspects (Springer Nature), Ch 20, 115-123.  
- ⎯ and M. Walpole (2020), “The Impact of Public Perceptions on General Consumption Taxes”. British Tax Review 67/5, 637-669.  
- ⎯ and R. Krever (2013), “Ending VAT Exemptions: Towards a Post-Modern VAT”. In R. de la Feria (ed.), VAT Exemptions: Consequences and Design Alternatives (The Hague: Kluwer Law International), 3–35.  
- de Mooij, R. and M. Keen (2013), “Fiscal Devolution and Fiscal Consolidation: The VAT in Troubled Times”. In A. Alesina and F. Giavazzi (eds), Fiscal Policy after the Crisis (University of Chicago Press), 443–485.  
- Ebrill, L. et al. (2001), The Modern VAT (International Monetary Fund).  
- Fenochietto, R. and J. C. Benitez (2021), “Encouraging Formal Invoicing and Reducing the VAT Impact on Low-Income individuals”. IMF Working Papers WP/21/40.  
- Finkelstein, A. (2009), “EZ-Tax: Tax Salience and Tax Rates”. Quarterly Journal of Economics 124, 969.  
- Gamage, D. and D. Shanske (2011), “Three Essays on Tax Salience: Market Salience and Political Salience”. Tax Law Review 65, 23.  
- Grinberg, I. (2006), “Implementing a Progressive Consumption Tax: Advantages of Adopting the VAT Credit-Method System”. National Tax Journal 59, 926-954.  
- Harju, J. et al. (2018), “Firm types, price-setting strategies, and consumption-tax incidence”. Journal of Public Economics 165, 48.  
- Krever, R. (2014), “Combating VAT Fraud: Lessons from Korea”. British Tax Review 3, 329-341.  
- Leahy, E., S. Lyons and R. Tol (2011), “The Distributional Effects of Value Added Tax in Ireland”. The Economic and Social Review 42(2), 213–235.  
- Liu, L., B. Lockwood, M. Almunia and E. Tam (2021). "VAT Notches, Voluntary Registration, and Bunching: Theory and U.K. Evidence". The Review of Economics and Statistics 103(1), 151-164.  
- ⎯ B. Lockwood, and E. Tam (2022). “Small Firm Growth and the VAT Threshold: Evidence for the UK”. Oxford University Centre for Business Taxation Working Paper 2022-21.  
- Lustig, N., C. Pessino and J. Scott (2014), “The Impact of Taxes and Social Spending on Inequality and Poverty in Argentina, Bolivia, Brazil, Mexico, Peru, and Uruguay: Introduction to the Special Issue”. Public Finance Review 42(3), 287-303.  
- McCaffery, E. and J. Baron (2003), “The Humpty Dumpty Blues: Disaggregation Bias in the Evaluation of Tax Systems”. Organizational Behaviour and Human Decision Processes 91, 230-242.  
- McNulty, J. (2000), “Flat Tax, Consumption Tax, Consumption-Type Income Tax, Proposals in the United States: A Tax Policy Discussion of Fundamental Tax Reform”. California Law Review 88(6), 2095-2185.  
- Moore, J., L. Stinson and E. Welniak, Jr. (1997), “Income Measurement in Surveys: A Review”. U.S. Census Bureau Working Paper SM97-05.  
- OECD (2020a), Consumption Tax Trends - VAT/GST and Excise Rates, Trends and Policy Issues (Paris: OECD Publishing).  
- OECD (2010), “Choosing a Broad Base — Low-Rate Approach to Taxation”. OECD Tax Policy Studies 19 (OECD Publishing).  
- Pessoa, M., A. Okello, A. Swistak, M. Muyangwa, V. Alonso-Albarran and V. de Paul Koukpaizan (2021), How to Manage Value-Added Tax Refunds. IMF How-To Note No. 2021/004.  
- Pomeranz, D. (2015), “No Taxation Without Information: Deterrence and Self-Enforcement in the Value Added Tax”. American Economic Review 105, 2539.  
- Ramsey, F.P. (1927), “A Contribution to the Theory of Taxation”. Economic Journal 37, 47.  
- Ruiz, N. and A. Trannoy (2008), “Le caractère régressif des taxes indirectes: les enseignements d’un modèle de microsimulation”. Economie et Statistique 413, 21–46.  
- Sandmo, A. (1987), “A Reinterpretation of Elasticity Formulae in Optimum Tax Theory”. Economica 54(213), 89.  
- Sausgruber, R. and J.R. Tyran (2005), “Testing the Mill Hypothesis of Fiscal Illusion”. Public Choice 122(1), 39.  
- Schenk, A., V. Thuronyi and W. Cui (2015), Value Added Tax: A Comparative Approach (Cambridge University Press).  
- Sorensen, P.B. (2007), “The Theory of Optimal Taxation: What is the policy relevance?”. International Tax and Public Finance 14, 383.  
- Stiglitz, J. (1998), “The private uses of public interests: incentives and institutions”. Journal of Economic Perspectives 12(2), 3.  
- Swistak, A., S. Wawrzak and A. Alinska (2015), “In Pursuit of Tax Equity: Lessons from VAT Rate Structure Adjustment in Poland”. Financial Theory and Practice 39 (2), 115-137.  
- van Oordt, M. (2018), “Zero-Rating vs Cash Transfers under the VAT”. Fiscal Studies 39(2), 1.  
- Viswanathan, M. (2022), “Implementing A (Modern) Progressive Consumption Tax”. Virginia Tax Review 41(2), 241-288.  
- Warren, N., A. Harding and R. Lloyd (2005), “GST and the changing incidence of Australian taxes: 1994-95 to 2001-02”. eJournal of Tax Research 3(1), 114.  
- Warwick, R, T. Harris, D. Phillips, M. Goldman, J. Jellema, G. Inchauste, K. Goraus-Tanska (2022), “The redistributive power of cash transfers vs VAT exemptions: A multi-country study”. World Development 151, 105742.  
- Waseem, M. (2020). The Role of Withholding in the Self-Enforcement of a Value-Added Tax: Evidence from Pakistan. Review of Economics and Statistics 104(2), 336-354.  

### Digitalization, fraud, and tax administration
- Brockmeyer, A. and M.S. Samarriba (2022), “Electronic Payment Technology and Tax Compliance – Evidence from Uruguay’s Financial Inclusion Reform”. World Bank Policy Research Working Paper 9947.  
- Herbain, C. (2018), “Fighting VAT Fraud and Enhancing VAT Collection in a Digitalised Environment”. Intertax 46(6/7), 579-583.  
- IMF (2020b), Enhancing Digital Solution to Implement Emergency Responses. IMF Special Series on COVID-19, June.  
- IMF (2020a), Digital Solutions for Direct Cash Transfers in Emergencies, April.  
- Giles, C. (2019), “Russia’s role in producing the taxman of the future”. Financial Times, July 29, 2019.  
- Pomeranz, D. (2015), “No Taxation Without Information: Deterrence and Self-Enforcement in the Value Added Tax”. American Economic Review 105, 2539.  
- Pessoa, M., A. Okello, A. Swistak, M. Muyangwa, V. Alonso-Albarran and V. de Paul Koukpaizan (2021), How to Manage Value-Added Tax Refunds. IMF How-To Note No. 2021/004.  
- Waseem, M. (2020). The Role of Withholding in the Self-Enforcement of a Value-Added Tax: Evidence from Pakistan. Review of Economics and Statistics 104(2), 336-354.  
- Wilks, D., J. Cruz and P. Sousa (2019), “Please give me an invoice”: VAT evasion and the Portuguese tax lottery”. International Journal of Sociology and Social Policy 39(5/6), 412-426.  

### Behavioral, political economy, and welfare stigma
- Bursztyn, L. and R. Jensen (2017). “Social Image and Economic Behaviour in the Field: Identifying, Understanding and Shaping Social Pressure”. Annual Review of Economics 9, 131-153.  
- Bolton, R., J. Whelan and F. Dukelaw (2022). “What Can Welfare Stigma Do?”. Social Policy and Society 21(4), 632-645.  
- Boston, J. (2019). “Redesigning the Welfare State. Rethinking the indexation of cash and non-cash assistance”. Policy Quarterly 15(1), 3-8.  
- Boyer, P. and M.B. Peterson (2018), “Folk-economic beliefs: An evolutionary cognitive model”. Behavioral and Brain Sciences 41, 158.  
- Bursztyn, L. and R. Jensen (2017). “Social Image and Economic Behaviour in the Field: Identifying, Understanding and Shaping Social Pressure”. Annual Review of Economics 9, 131-153.  
- Celhay, P., B. Meyer and N. Mittag (2022). “Stigma in Welfare Programs”, NBER Working Paper 30307.  
- Downs, A. (1957), “An economic theory of political action in a democracy”. Journal of Political Economy 65(2), 135.  
- Fernandes, R. and D. Rodrik (1991), “Resistance to Reform: Status quo bias in the presence of individual uncertainty”. American Economic Review 81, 1146.  
- Konow, J. (2000), “Fair Shares: Accountability and Cognitive Dissonance in Allocation Decisions”. American Economic Review 90(4), 1072.  
- McCaffery, E. and J. Baron (2003), “The Humpty Dumpty Blues: Disaggregation Bias in the Evaluation of Tax Systems”. Organizational Behaviour and Human Decision Processes 91, 230-242.  
- Mercier, H. (2020), Not Born Yesterday — The science of Who we trust and What we believe (Princeton University Press).  
- Stantcheva, S. (2020), “Understanding Tax Policy: How Do People Reason”. NBER Working Paper 27699.  
- Sunstein, C. (2015), Simpler: The Future of Government (Simon & Schuster).  
- Thaler, R. (1999), “Mental Accounting Matters”. Journal of Behavioral Decision Making 12, 183-206.  
- ⎯ (1985), “Mental Accounting and Consumer Choice”. Marketing Science 4, 199-214.  
- Bursztyn, L. and R. Jensen (2017). “Social Image and Economic Behaviour in the Field: Identifying, Understanding and Shaping Social Pressure”. Annual Review of Economics 9, 131-153.  

### Methodology, theory, and broader tax policy literature
- Bradford, D. (2021), “The X Tax in the World Economy”. NBER Working Paper w10676.  
- Caspersen, E. and G. Metcalf (1994), “Is a Value Added Tax Regressive? Annual Versus Lifetime Incidence Measures”. National tax Journal 47(4), 731.  
- Corlett, W.J. and D.C. Hague (1953), “Complementarity and the Excess Burden of Taxation”. Review of Economic Studies 21, 21.  
- Kleven, H.J. (2004), “Optimum Taxation and the Allocation of Time”. Journal of Public Economics 88, 545.  
- Konow, J. (2000), “Fair Shares: Accountability and Cognitive Dissonance in Allocation Decisions”. American Economic Review 90(4), 1072.  
- Klen, M. (2013), “Targeting, Cascading, and Indirect Tax Design”. IMF Working Papers WP/13/57.  
- Keen, M. and J. Slemrod (2021), Rebellion, Rascals and Revenue – Tax Follies and Wisdom through the Ages (Princeton University Press).  
- Kleven, H.J. (2004), “Optimum Taxation and the Allocation of Time”. Journal of Public Economics 88, 545.  
- Liu, L., B. Lockwood, M. Almunia and E. Tam (2021). "VAT Notches, Voluntary Registration, and Bunching: Theory and U.K. Evidence". The Review of Economics and Statistics 103(1), 151-164.  
- McCaffery, E. and J. Baron (2003), “The Humpty Dumpty Blues: Disaggregation Bias in the Evaluation of Tax Systems”. Organizational Behaviour and Human Decision Processes 91, 230-242.  
- Ramsey, F.P. (1927), “A Contribution to the Theory of Taxation”. Economic Journal 37, 47.  
- Sandmo, A. (1987), “A Reinterpretation of Elasticity Formulae in Optimum Tax Theory”. Economica 54(213), 89.  
- Sorensen, P.B. (2007), “The Theory of Optimal Taxation: What is the policy relevance?”. International Tax and Public Finance 14, 383.  
- Weisbach, D. (2002), “Does the X-Tax Mark the Spot?”. Chicago John Olin Law & Economics Working Paper 163.  

*Designing a Progressive VAT — Working Paper No. WP/2024/078*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2024/english/wpiea2024078-print-pdf.pdf_