## 1. Current ‘Flat Taxes’

## Source details

**Canonical URL:** [1. Current ‘Flat Taxes’](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06218.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06218.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06218.pdf.json)

---

### Introduction
- Sparked by Estonia (1994) and propelled by Russia’s reform in 2001; Russia’s reform combined rate flattening and sharp reductions in highest marginal rates of tax on labor income and was followed by a marked increase in revenue from the PIT: "over one-quarter, in real terms."
- Paper purpose: broad review of principles and experience on the impact of the “flat tax,” focusing on flatness as the distinctive feature while acknowledging that rate cuts (particularly for higher incomes) are often integral to observed outcomes.
- Structure (as in source):
  - Section II: definition and international experience with flat taxes.
  - Section III: implications from optimal tax considerations and impacts on equity, work incentives, compliance and administration, automatic stabilization, and political economy.
  - Section IV: conclusions.

### What is a “Flat Tax”?
- Definition used in the paper: any tax system under which the personal tax on labor income takes the form
  - ]0),.(max[)( FF AYtYT−=  (equation (1) in source),
  - where )(YT F denotes liability on labor income Y, t is the single strictly positive marginal tax rate, and FA is some allowance (possibly varying with taxpayer circumstances).
- Common feature: a single strictly positive marginal tax rate on labor income.
- Caveats:
  - Hall and Rabushka (HR) proposal (cash-flow tax on business income and single-rate tax on workers with personal allowance) not the practical subject here.
  - Effective tax on labor income depends on social insurance contributions; many countries with flat personal taxes raise substantially more revenue via social contributions, so effective labor tax schedules may be far from flat.
  - Many flat tax introductions were part of wider reform packages including administrative improvements; paper focuses on parametric flat tax aspects and directly related administrative/compliance consequences.

### The “Flat Taxes” in Practice — overview and headline country features
- Eight flat tax introductions reviewed (definition per equation (1)); short-term revenue developments (PIT, CIT, indirect taxes, total tax revenue relative to GDP) reported (Figure 1; Appendix II gives numbers).
- Headline country features (Rates in Percent relate to year before and after adoption; change in basic allowance summarized):
  - Estonia — Adopted: 1994; Personal Income Tax: After 26; Before 16-33; Corporate Income Tax Rate, After Reform: 26; Change in Basic Allowance: Modest increase.
  - Lithuania — Adopted: 1994; Personal Income Tax: After 33; Before 18-33; Corporate Income Tax Rate, After Reform: 29; Change in Basic Allowance: Substantial increase.
  - Latvia — Adopted: 1997; Personal Income Tax: After 25; Before 25 and 10; Corporate Income Tax Rate, After Reform: 25; Change in Basic Allowance: Slight reduction.
  - Russia — Adopted: 2001; Personal Income Tax: After 13; Before 12-30; Corporate Income Tax Rate, After Reform: 37; Change in Basic Allowance: Modest increase.
  - Ukraine — Adopted: 2004; Personal Income Tax: After 13; Before 10-40; Corporate Income Tax Rate, After Reform: 25; Change in Basic Allowance: Increase.
  - Slovak Republic — Adopted: 2004; Personal Income Tax: After 19; Before 10-38; Corporate Income Tax Rate, After Reform: 19; Change in Basic Allowance: Substantial increase.
  - Georgia — Adopted: 2005; Personal Income Tax: After 12; Before 12-20; Corporate Income Tax Rate, After Reform: 20; Change in Basic Allowance: Eliminated.
  - Romania — Adopted: 2005; Personal Income Tax: After 16; Before 18-40; Corporate Income Tax Rate, After Reform: 16; Change in Basic Allowance: Increase.

- Selected observed country outcomes (preserved exactly where quoted):
  - Estonia
    - Initial flat rate: 26 percent (later scheduled to reach 20 percent in 2009).
    - Personal allowance increased from EEK 2,400 to EEK 3,600 (described as a modest increase in real terms given inflation).
    - Corporate tax previously 35 percent set at 26 percent; undistributed profits untaxed since 2000 with distributions taxed at the regular flat rate.
    - PIT and CIT revenue fell relative to GDP after adoption.
  - Lithuania
    - Flat tax introduced in 1994 at 33 percent (highest of prior marginal rates).
    - Threshold increased from LTD 35 to LTD 115 ("more than doubling relative to GDP, though still reaching only a very modest 2.6 percent of per capita GDP").
    - Corporate tax maintained at 29 percent (CIT later reduced to 15 percent).
    - PIT revenue rose after reform; CIT revenue showed a marked fall for which no obvious explanation is provided.
  - Latvia
    - Flat tax introduced in 1997 at 25 percent.
    - Prior structure degressive (25 then 10 percent); adopting flat tax increased liability at very highest incomes.
    - Personal allowances slightly reduced but remained high (around 19 percent of per capita income).
    - PIT and CIT revenue rose somewhat following the reform.
  - Russia
    - 2001 unified marginal rates 12, 20, and 30 percent into 13 percent.
    - Threshold for taxable income increased only modestly in real terms; allowance withdrawn discontinuously, creating points where effective marginal rate is, in principle, infinite.
    - Personal tax on interest income reduced from 15 percent to flat 13 percent; special tax of 35 percent introduced on "excessive" bank interest.
    - Maximum corporate tax (inclusive of municipal add-ons) increased from 30 to 35 percent; dividend taxation increased from 15 to 30 percent.
    - Structural VAT changes: elimination of some exemptions and application to individual entrepreneurs.
    - Excises imposed on diesel and lubricants; tobacco and petroleum excises substantially increased (tobacco excise for instance more than doubled in nominal terms).
    - Reform accompanied by major change in social contributions: various flat-rate contributions (~39.5 percent) were replaced by a degressive "unified social tax," with the marginal rate falling from (text cuts off).
  - Ukraine
    - Effective January 2004, flat tax at 13 percent (slated to increase to 15 percent in 2007) replaced a PIT with marginal rates ranging from 10 to 40 percent.
    - Personal allowance increased from around Hrv 204 to Hrv 738 (annual equivalent), available only to those with incomes less than 1.4 times a subsistence amount; allowance removed entirely above that point.
    - CIT rate reduced from 30 to 25 percent; VAT remained at 19 percent; outcome: "PIT revenue fell significantly following reform."
  - Slovak Republic
    - 2004 reform established single common rate of 19 percent for PIT, CIT, and VAT.
    - Personal allowance more than doubled from SK 38,760 to SK 80,832 (reaching around 60 percent of the average wage); spouse allowance increased almost sevenfold from SK 12,000 to same value as personal allowance.
    - Aggregate social security tax reduced from 51 to 48.6 percent; revenue effect partly offset by increasing contribution ceilings.
    - Revenue outcomes:
      - PIT revenue fell by around 0.7 points of GDP (less than the 1.2 points forecast by the authorities).
      - CIT revenue fell by 0.4 percent of GDP.
      - Social contributions fell by 1.5 percent of GDP.
      - These revenue losses were largely offset by higher indirect tax revenue, mainly from VAT.
  - Georgia
    - Flat tax on personal income of 12 percent introduced at start of 2005; no basic allowance after reform.
    - Minimum pension more than doubled.
    - Social security contribution rate reduced from 33 to 20 percent.
    - Revenue outcomes:
      - PIT revenue fell from around 2.7 to 2.5 percent of GDP.
      - CIT revenue rose.
      - Indirect tax revenue rose.
      - Total revenues remained broadly unchanged.
  - Romania
    - Effective January 2005, flat rate of 16 percent replaced a PIT schedule with five marginal rates rising from 18 to 40 percent.
    - Profit tax reduced from 25 to 16 percent.
    - Revenue outcomes:
      - Revenue from both the PIT and CIT fell following the reform by a total of around one percentage point of GDP.

### Common features and cross-country differences
- Differences
  - First wave (Baltics): rates set at moderately high levels—at or close to the highest pre-reform marginal rate.
  - Second wave (starting in Russia): rates closer to lowest pre-reform rates; in Romania even lower—"less than 20 percent in all cases."
  - In half the cases (Estonia, Latvia, Slovak Republic, Romania) the single positive rate of labor income tax was set equal to the corporate income tax rate at adoption.
  - Three countries taxed corporate income more heavily than labor income; Lithuania taxed corporate income less heavily.
  - Slovak Republic uniquely set the single VAT rate equal to the common labor and CIT rate (19 percent).
  - Treatment of capital income varies: second wave generally taxes interest and capital gains at or close to flat rate; Latvia exempts them entirely; Estonia taxes interest at below flat rate.
  - Personal allowances generally increased (especially in second wave); Georgia eliminated threshold.
- Recurring themes
  - Except in Latvia and Georgia, adoption associated with increases in personal allowances.
  - Reforms often accompanied elimination/reduction of exemptions under PIT and CIT (bringing capital income into tax).
  - Except in Latvia and Russia, adoption was followed by a reduction in PIT revenue.
  - Many cases saw reductions in the CIT rate; reductions in CIT have been more marked than reductions in flat labor tax.
  - Most second wave reforms included reductions in social contributions.
  - Second wave typically accompanied by substantial increases in indirect taxation, especially excises, with "a strong increase in revenue from this source."

### Analytical framing: can a "flat tax" be fully optimal?
- Optimal non-linear income tax problem (Mirrlees (1971)): balances efficiency costs of distortions to work effort against distributional objectives.
- General property: marginal tax rate lies everywhere between zero and one; this provides no direct implication about how marginal tax rate should vary with income.
- Mirrlees simulations (utilitarian objective, Cobb-Douglas preferences, log-normal ability distribution) suggested optimal tax schedules "clos[e] to linearity." Quoted: "[p]erhaps the most striking feature of the results is the closeness to linearity of the [optimal] tax schedules" (p.206).
- Robustness caveat: conclusion not robust under more inequality-averse social welfare functions, alternative labor supply responses, or different ability distributions — optimal schedule can be highly nonlinear.
- If ability distribution is bounded above, Pareto efficiency requires marginal tax rate on highest earner be zero → decreasing marginal tax rate in upper range.
- No overwhelming theoretical presumption for or against flat taxes; linearity may be reasonable proxy given administrative and compliance difficulties with non-linear schedules.

### Flat tax, evasion and corruption (HKM)
- Hindriks, Keen and Muthoo (1999) (HKM) stylized result: a linear proportional tax, set at an appropriate rate and combined with appropriate penalty structure and simple wage payment to inspectors, can maximize revenue without inducing tax evasion or payment of bribes (Corollary 1).
- HKM intuition (stylized):
  - If collusion detected with probability π and collective penalty equals full income Y, misreporters report zero liability.
  - Expected collective payment if misreport: πY; collective payment if truthful: T(Y).
  - Revenue-maximizing, truth-preserving T implies T(Y) = πY → proportional tax at rate equal to probability of detection.
- Caveats:
  - Result is highly special and stylized; non-linear corruption-/evasion-proof schedules also exist with inspector commissions.
  - Requires penalties stronger than society commonly applies for tax offences.
  - HKM supports claim that flatness provides some protection against evasion/corruption but level of taxation and enforcement matter.

### Political economy argument for linearity
- Brennan and Buchanan (1977): flat tax as commitment device limiting government size by restricting policymakers to a single marginal tax rate.
- Critics: institutional devices (fiscal rules) can limit government size without constraining tax instruments; restricting efficient instruments may reduce consumer welfare if policymakers value it.
- Empirical context: many adopting governments were new administrations seeking to signal regime change (examples: Estonia 1992, Russia 2000, Georgia 2003, Romania 2004).

### Equity and progressivity: definitions and comparative results
- With basic exemption, a flat tax is progressive in the sense average rate increases with income.
- More relevant question: whether a flat tax is more or less progressive than the replaced scheme.
- Definition: M more progressive than L iff concentration curve of tax liabilities under M lies everywhere below that of L.
- Single-crossing results:
  - If schedules cross once and raise same revenue, ordering is unambiguous (Hemming and Keen, 1983).
  - Two crossings can make comparison ambiguous (Dardoni and Lambert (1988)).
- Revenue-neutral comparisons (flat F vs pre-reform progressive P):
  - If A_F < A_P → T_F crosses T_P once from above → flat tax unambiguously less progressive.
  - If A_F > A_P:
    - If flat marginal rate ≥ highest pre-reform marginal rate → crossing at most once → flat tax more progressive (Panel A case; matches Lithuania).
    - If flat rate between lowest and highest pre-reform marginal rates → schedules cross twice → neither unambiguously more progressive (Panel B case; matches Estonia, Russia, Slovak Republic, Ukraine).
  - For any given inequality measure, a critical constant marginal rate exists where flat tax with higher (lower) rate than critical yields less (more) inequality in after-tax incomes.
- Non-equal-yield comparisons: must distinguish distribution of tax payments from distribution of after-tax income; normalization procedures described in source.
- Country-specific outcomes:
  - Lithuania: movement toward flat tax had strongly progressive elements.
  - Georgia: strongly regressive elements.
  - Many reforms: effect intrinsically ambiguous.

### Progressivity implications of noncompliance (Slemrod-style model)
- Individuals choose concealment E to maximize )()(ECEYTY−−− with C(E) convex; first-order condition (3): )()(ECEYT′=−′.
- Effective schedule T*(Y) ≡ T(Y − E(Y)).
- For flat tax:
  - All taxpayers declaring above exempt amount conceal same amount → effective schedule equivalent to statutory with increased tax-free amount.
  - Proportionate reduction in tax liability due to noncompliance is higher at lower income levels → tends to improve vertical equity: T*_F more progressive than statutory T_F.
- For pre-existing progressive tax:
  - Better off conceal more in absolute terms; if ln(T_P) is convex, rich escape larger proportion of liability → noncompliance worsens vertical equity.
- Net implication: noncompliance tends to make flat tax appear more progressive and pre-existing taxes less progressive.

### Work incentives and labor supply evidence
- Effects depend on changes in Average Tax Rate (ATR) and Marginal Tax Rate (MTR).
  - Substitution effect: increased effort iff MTR falls.
  - Income effect (if leisure normal): increased effort if ATR rises.
- Stylized panel patterns produce mixed incentive effects across income ranges; highest incomes face ambiguous net effects.
- Empirical literature: broad consensus that effects on effort of primary workers are modest.
- Russia (Ivanova, Keen and Klemm (2005); IKM): difference-in-differences approach found no evidence of strong effects on work effort—neither gross income nor hours worked increased more among treatment than control.

### Compliance, administration, and simplicity
- Claimed merits: simplification, improved compliance, reduced administration costs.
- Caveats:
  - Tax-free thresholds imply effectively two marginal rates; withholding/multiple-job issues persist.
  - Georgia eliminated threshold to facilitate wage withholding.
  - Major sources of complexity often arise from defining/monitoring tax base and exemptions rather than rate structure alone.
- Evidence on perceptions and compliance:
  - Russian survey: proportion rating perceived complexity as important/very important to compliance was virtually the same in 2002 as in 1998 (Ivanova, Keen, and Klemm (2005)).
  - Theoretical models give mixed predictions: Slemrod-style frameworks suggest higher MTR increases concealment; ASY-style gamble models produce different implications.
- Russian empirical indicators (IKM):
  - Ratio of consumption to income for the control group remained at around 75 percent.
  - That for the control group increased from around 52 to 70 percent.
  - Interpretation: suggests increase in compliance; attribution to parametric reform uncertain due to concurrent administrative changes and public statements.

### Automatic stabilizers (built-in flexibility)
- Common belief: flat tax weakens automatic stabilizers.
- Analytical points:
  - Aggregate marginal tax rate defined as B_i ≡ ∫ dY Y f(Y) T_i′(Y) (equation (4) in source).
  - If A = 0 (pure proportional tax), built-in flexibility unambiguously weaker under flat tax.
  - With threshold/exempt amount, built-in flexibility effect is ambiguous: revenue neutrality may require higher marginal rate above threshold, potentially increasing flexibility.
  - Appendix 4: built-in flexibility under the flat tax is greater the lower the ratio of average incomes above the threshold to the level of the threshold itself (i.e., more income concentrated just above threshold tends to raise flexibility under flat tax).
  - Empirical note: income shocks often concentrated at lower levels; if these lie just above threshold, built-in flexibility may be stronger under flat tax.

### Political economy and adoption drivers
- Some flat tax moves were Pareto-improving (e.g., Russia: no change for those paying 12 percent; gain for higher-rate payers).
- Rapid implementation can hinder opposition mobilization (e.g., Russia: "less than 50 days" from announcement to presidential approval).
- Flat tax adoption commonly part of broader reform packages; political acceptability often tied to whole-package effects.
- Voting-theory insight (Bohn and Stuart (2003)): linear taxes are unlikely political equilibria; median-voter logic predicts redistribution toward the middle—opposite common flat tax pattern.
- Contextual enablers: rapid income inequality increases during transition, weak tax administration, low pre-reform PIT revenue, and new pro-reform governments helped adoption and potential compliance gains.

### Key empirical findings and lessons (summary)
- Revenue effects:
  - Except in Russia, second-wave low-rate flat tax reforms associated with reduction in PIT revenue.
  - No case found of Laffer-effect where reform paid for itself.
  - Russia’s strong revenue performance appears to reflect wider macroeconomic recovery rather than flat tax alone.
- Compliance and administration:
  - Evidence of improved compliance after Russian reform exists but attribution to parametric reform alone is uncertain.
- Labor supply:
  - Impact on work incentives ambiguous in principle; household-level evidence (Russia) shows no detectable significant impact on work effort.
- Distributional effects:
  - Complex and country-specific; no general presumption that flat tax reduces progressivity.
  - Compliance effects can plausibly increase effective progressivity.
  - Summary indices in few studies show increased progressivity but may overstate implications.
- Complexity:
  - Flatness simplifies rate structure but major complexity sources remain exemptions and special treatments.
- Base-broadening and fiscal packages:
  - Rate cuts often packaged with base-broadening (e.g., Ukraine base broadening estimated to increase revenue by around one point of GDP).
  - Base broadening contributes to horizontal equity, efficiency gains, and simplicity.
- Automatic stabilizers:
  - Movement to flat tax may plausibly strengthen automatic stabilizers in some cases; empirical study limited.
- Political economy and signaling:
  - Flat tax often used to signal regime shift toward market-oriented policy; spread of flat tax can undermine signaling value over time.

### Sustainability, technical tensions, and open questions
- Diversity of design across countries (first wave vs second wave differences; varying treatment of corporate rates, allowances, social contributions, VAT).
- Empirical limitations: paucity of household-level analyses outside Russia and Slovak Republic; need for more studies.
- Taxing internationally mobile capital: adopted flat taxes do not provide coherent framework; pressures may push toward dual income tax logic or decoupling taxation of capital from labor.

### Appendix II — Composition of Tax Revenues Before and After Introduction of the Flat Taxes (In percent of GDP)
- Estonia
  - Personal Income Tax: Year Before the Reform 8.5; Year of the Reform 8.1
  - Corporate Income Tax: Year Before the Reform 4.8; Year of the Reform 3.5
  - Indirect Taxes: Year Before the Reform 11.1; Year of the Reform 13.3
- Georgia
  - Personal Income Tax: Year Before the Reform 2.7; Year of the Reform 2.5
  - Corporate Income Tax: Year Before the Reform 1.6; Year of the Reform 1.9
  - Indirect Taxes: Year Before the Reform 8.1; Year of the Reform 11.0
- Latvia
  - Personal Income Tax: Year Before the Reform 5.4; Year of the Reform 5.6
  - Corporate Income Tax: Year Before the Reform 2.0; Year of the Reform 2.4
  - Indirect Taxes: Year Before the Reform 12.6; Year of the Reform 12.5
- Lithuania
  - Personal Income Tax: Year Before the Reform 5.0; Year of the Reform 5.4
  - Corporate Income Tax: Year Before the Reform 5.3; Year of the Reform 2.5
  - Indirect Taxes: Year Before the Reform 6.2 1/; Year of the Reform 6.3 1/
- Romania
  - Personal Income Tax: Year Before the Reform 3.0; Year of the Reform 2.3
  - Corporate Income Tax: Year Before the Reform 2.7; Year of the Reform 2.4
  - Indirect Taxes: Year Before the Reform 10.2; Year of the Reform 10.9
- Russia
  - Personal Income Tax: Year Before the Reform 2.4; Year of the Reform 2.9
  - Corporate Income Tax: Year Before the Reform 5.5; Year of the Reform 5.8
  - Indirect Taxes: Year Before the Reform 8.6; Year of the Reform 9.9
- Slovak
  - Personal Income Tax: Year Before the Reform 3.3; Year of the Reform 2.6
  - Corporate Income Tax: Year Before the Reform 2.8; Year of the Reform 2.4
  - Indirect Taxes: Year Before the Reform 9.8; Year of the Reform 11.4
- Ukraine
  - Personal Income Tax: Year Before the Reform 5.1; Year of the Reform 3.8
  - Corporate Income Tax: Year Before the Reform 5.0; Year of the Reform 4.7
  - Indirect Taxes: Year Before the Reform 8.8; Year of the Reform 8.6
- Unweighted average
  - Personal Income Tax: Year Before the Reform 4.4; Year of the Reform 4.1
  - Corporate Income Tax: Year Before the Reform 3.7; Year of the Reform 3.2
  - Indirect Taxes: Year Before the Reform 9.4; Year of the Reform 10.5
- Footnote
  - 1/ Includes only VAT.

*Source: IMF staff working paper excerpt, "1. Current ‘Flat Taxes’."*

### 1. Current ‘Flat Taxes’ ................................................................................................

### 1. Current ‘Flat Taxes’

### Introduction
- Recent years saw widespread interest in the “flat tax,” sparked by Estonia (1994) and propelled by Russia’s reform in 2001, which combined rate flattening and sharp reductions in highest marginal rates of tax on labor income and was followed by a marked increase in revenue from the PIT: "over one-quarter, in real terms."
- The paper’s purpose is to provide a broad review of principles and experience on the impact of the “flat tax,” focusing on flatness as the distinctive feature of these reforms while acknowledging that rate cuts (particularly for higher incomes) are often integral to observed outcomes.
- Structure of analysis:
  - Section II: definition and international experience with flat taxes.
  - Section III: implications from optimal tax considerations and impacts on equity, work incentives, compliance and administration, automatic stabilization, and political economy.
  - Section IV: conclusions.

### What is a “Flat Tax”?
- Definition used in the paper: any tax system under which the personal tax on labor income takes the form
  - ]0),.(max[)( FF AYtYT−=  (equation (1) in source),
  - where )(YT F denotes liability on labor income Y, t is the single strictly positive marginal tax rate, and FA is some allowance (possibly varying with taxpayer circumstances).
- Common feature across variants: a single strictly positive marginal tax rate on labor income.
- Distinctions and caveats:
  - The Hall and Rabushka (HR) proposal (a cash-flow tax on business income and a tax on workers’ income both levied at a single rate with a personal allowance) is not the practical subject of this paper; no country had adopted the HR flat tax at the time of writing.
  - Effective tax on labor income also depends on social insurance contributions; many countries with flat personal taxes raise substantially more revenue via social contributions, so effective labor tax schedules may be far from flat in practice.
  - Several flat tax introductions were part of wider tax reform packages including administrative improvements; the paper focuses on parametric flat tax aspects and administrative/compliance consequences only insofar as directly related.

### The “Flat Taxes” in Practice — overview and empirical patterns
- The paper reviews eight flat taxes introduced in recent years (definition per equation (1)). Short-term revenue developments (PIT, CIT, indirect taxes, and total tax revenue relative to GDP) around the time of reform are reported (Figure 1; underlying numbers in Appendix II).
- Caution: revenue developments around reform dates are influenced by many factors beyond the reform itself; nevertheless, short-term experiences shaped public policy debate.

Key summarized headline features (as presented in Table 1; "Rates in Percent" relate to year before and after adoption):
- Estonia
  - Adopted: 1994
  - Personal Income Tax: After 26; Before 16-33
  - Corporate Income Tax Rate, After Reform: 26
  - Change in Basic Allowance: Modest increase
- Lithuania
  - Adopted: 1994
  - Personal Income Tax: After 33; Before 18-33
  - Corporate Income Tax Rate, After Reform: 29
  - Change in Basic Allowance: Substantial increase
- Latvia
  - Adopted: 1997
  - Personal Income Tax: After 25; Before 25 and 10
  - Corporate Income Tax Rate, After Reform: 25
  - Change in Basic Allowance: Slight reduction
- Russia
  - Adopted: 2001
  - Personal Income Tax: After 13; Before 12-30
  - Corporate Income Tax Rate, After Reform: 37
  - Change in Basic Allowance: Modest increase
- Ukraine
  - Adopted: 2004
  - Personal Income Tax: After 13; Before 10-40
  - Corporate Income Tax Rate, After Reform: 25
  - Change in Basic Allowance: Increase
- Slovak Republic
  - Adopted: 2004
  - Personal Income Tax: After 19; Before 10-38
  - Corporate Income Tax Rate, After Reform: 19
  - Change in Basic Allowance: Substantial increase
- Georgia
  - Adopted: 2005
  - Personal Income Tax: After 12; Before 12-20
  - Corporate Income Tax Rate, After Reform: 20
  - Change in Basic Allowance: Eliminated
- Romania
  - Adopted: 2005
  - Personal Income Tax: After 16; Before 18-40
  - Corporate Income Tax Rate, After Reform: 16
  - Change in Basic Allowance: Increase

Observed country experiences (selected findings preserved exactly from source text):
- Estonia
  - Initial flat rate: 26 percent (later scheduled to reach 20 percent in 2009).
  - Personal allowance increased from EEK 2,400 to EEK 3,600 (described as a modest increase in real terms given inflation).
  - Corporate tax previously 35 percent set at 26 percent; undistributed profits untaxed since 2000 with distributions taxed at the regular flat rate.
  - PIT and CIT revenue fell relative to GDP after adoption.
- Lithuania
  - Flat tax introduced in 1994 at 33 percent (highest of prior marginal rates).
  - Threshold increased from LTD 35 to LTD 115 ("more than doubling relative to GDP, though still reaching only a very modest 2.6 percent of per capita GDP").
  - Corporate tax maintained at 29 percent (CIT later reduced to 15 percent).
  - Transformation of previous general excise tax into full VAT at 18 percent; new excise law and excise rate increases on gasoline, diesel, and beer in late 1994/early 1995.
  - PIT revenue rose after reform; CIT revenue showed a marked fall for which no obvious explanation is provided in the text.
- Latvia
  - Flat tax introduced in 1997 at 25 percent.
  - Prior structure was degressive: starting marginal rate of 25 percent followed by marginal rate of 10 percent on highest incomes; adopting flat tax increased liability at very highest incomes.
  - Personal allowances slightly reduced but remained high (around 19 percent of per capita income).
  - Corporate tax was 25 percent retained at reform (later diverged to 15 percent); dividend and interest income exempt.
  - PIT and CIT revenue rose somewhat following the reform.
- Russia
  - 2001 unified marginal rates 12, 20, and 30 percent into 13 percent.
  - Threshold for taxable income increased only modestly in real terms; allowance is withdrawn discontinuously, creating points where the effective marginal rate is, in principle, infinite.
  - After allowance exhaustion, tax functions as proportional; in highest income range, 13 percent is both average and marginal rate.
  - Base broadening measures included elimination of exemptions for military servicemen.
  - Personal tax on interest income reduced from 15 percent to flat 13 percent; special tax of 35 percent introduced on "excessive" bank interest (chosen to broadly equate with effective tax—PIT plus social security—on labor income).
  - Maximum corporate tax (inclusive of municipal add-ons) increased from 30 to 35 percent; dividend taxation increased from 15 to 30 percent.
  - Structural VAT changes: elimination of some exemptions and application to individual entrepreneurs.
  - Excises imposed on diesel and lubricants; tobacco and petroleum excises substantially increased (tobacco excise for instance more than doubled in nominal terms).
  - Russian reform accompanied by major change in social contributions: various flat-rate contributions (~39.5 percent) were replaced by a degressive "unified social tax," with the marginal rate falling from (text cuts off here in source).
- General notes on omitted or borderline cases:
  - Serbia: sometimes included among flat tax countries but not by the paper’s definition because it applies an additional tax on total income above a threshold, resulting in two strictly positive marginal rates; functioned much like a flat tax in practice due to high threshold (though threshold was significantly lowered in July 2006).
  - Paraguay: introduced a 10 percent flat PIT in 2006 largely intended to improve VAT compliance; taxpayers may deduct practically all consumption expenditures (provided invoices are shown) and some types of savings.

### Analytical themes highlighted for later sections (as previewed in source)
- The paper will analyze:
  - Implications of optimal tax theory for the shape of the tax schedule.
  - Detailed examination of impacts of flat taxes on equity, work incentives, compliance and administration, and automatic stabilization, drawing on available evidence.
  - Political economy factors underlying flat tax introductions.
- Appendices noted for further empirical detail:
  - Appendix I: detailed account of pre- and post-reform systems for each country.
  - Appendix II: underlying numbers for revenue developments reported in Figure 1.
  - Appendix III: equity aspects of the flat tax.
  - Appendix IV: automatic stabilizers and the flat tax.

*Source: IMF staff working paper excerpt, "1. Current ‘Flat Taxes’."*

### 35.6 to 5 percent.

### _wp06218 - 35.6 to 5 percent.

### Country case studies: key changes and outcomes

- General note:
  - "In the year following the reform, revenue from the PIT rose by nearly one-quarter in real terms."

- Ukraine
  - Effective January 2004, a flat tax at 13 percent (slated to increase to 15 percent in 2007) replaced a PIT with marginal rates ranging from 10 to 40 percent.
  - Personal allowance increased from around Hrv 204 to Hrv 738 (annual equivalent), available only to those with incomes less than 1.4 times a subsistence amount; allowance removed entirely above that point.
  - PIT base broadened by bringing interest income (previously exempt) into tax at the flat rate, eliminating various occupation-specific exemptions (replacing them with direct payments), and adopting a more comprehensive definition of capital gains.
  - CIT rate reduced from 30 to 25 percent; significant preferences and exemptions removed in 2003 and 2004.
  - VAT remained at a single rate of 19 percent, but several sectoral exemptions were eliminated (including for housing construction, aircraft, and shipbuilding).
  - Outcome: "PIT revenue fell significantly following reform."

- Slovak Republic
  - 2004 reform established a single common rate of 19 percent for PIT, CIT, and VAT.
  - Replaced a PIT schedule of five non-zero marginal rates ranging from 10 to 38 percent (with five other rates applying to specific types of income).
  - Personal allowance more than doubled from SK 38,760 to SK 80,832 (reaching around 60 percent of the average wage); spouse allowance increased almost sevenfold from SK 12,000 to the same value as the personal allowance.
  - Child allowances replaced by fixed allowances and a refundable tax credit conditional on labor market participation.
  - CIT reductions accompanied by more rapid depreciation and more generous carry-forward provisions.
  - Dividend tax (15 percent final withholding, prior to reform) abolished; inheritance and gift taxes abolished.
  - Significant scaling back of exemptions under both PIT and CIT (examples: removal of some allowances for interest and capital income, additional private pension contributions, exemptions for soldiers and judges; limits on tax holidays, deductibility of charitable giving, tighter rules for provisioning and reserves).
  - Aggregate social security tax reduced from 51 to 48.6 percent, with revenue effect partly offset by increasing contribution ceilings.
  - Indirect tax reforms: unification of a 20 percent standard VAT and a reduced rate of 14 percent into a flat rate of 19 percent; excises increased to align with EU requirements.
  - Labor market and benefits reforms accompanied tax changes (reduction in out-of-work benefits, reduction in loss of benefit on taking employment, reduction in implicit tax rate on those working beyond retirement age, labor law changes to increase flexibility).
  - Revenue outcomes:
    - PIT revenue fell by around 0.7 points of GDP (less than the 1.2 points forecast by the authorities).
    - CIT revenue fell by 0.4 percent of GDP.
    - Social contributions fell by 1.5 percent of GDP.
    - These revenue losses were largely offset by higher indirect tax revenue, mainly from VAT.

- Georgia
  - Flat tax on personal income of 12 percent introduced at the start of 2005, replacing rates of 12, 15, 17, and 20 percent.
  - No basic allowance after reform; tax on labor income is proportional at all income levels.
  - Rationale for eliminating threshold: to facilitate wage withholding for holders of multiple jobs.
  - Minimum pension more than doubled (providing some protection to retired).
  - CIT rate unchanged at 20 percent.
  - Social security contribution rate reduced from 33 to 20 percent.
  - Standard VAT rate reduced from 20 to 18 percent.
  - To partly offset revenue loss, exemptions eliminated under PIT (e.g., those in mountainous regions and self-production by farmers) and VAT; most excises were increased.
  - Revenue outcomes:
    - PIT revenue fell from around 2.7 to 2.5 percent of GDP.
    - CIT revenue rose.
    - Indirect tax revenue rose, largely reflecting increased excise rates.
    - Total revenues remained broadly unchanged.

- Romania
  - Effective January 2005, flat rate of 16 percent replaced a PIT schedule with five marginal rates rising from 18 to 40 percent.
  - Monthly basic personal allowance raised modestly in real terms.
  - Profit tax reduced from 25 to 16 percent.
  - Dividends paid to individuals, interest income and capital gains were subject to final withholding taxes at lower than the flat rate, but since 2006 have also been subject to the flat 16 percent.
  - VAT rates (19 and 9 percent) left unchanged; in 2005 several relatively minor exemptions were eliminated and excises increased.
  - Revenue outcomes:
    - Revenue from both the PIT and CIT fell following the reform by a total of around one percentage point of GDP.

### Common features and differences across reforms

- Differences highlighted
  - First wave (Baltics): tax rates set at moderately high levels—at or close to the highest marginal tax prior to reform.
  - Second wave (starting in Russia): tax rates closer to the lowest of the pre-reform rates; in Romania even lower—"less than 20 percent in all cases."
  - In half the cases (Estonia, Latvia, the Slovak Republic, Romania) the single positive rate of labor income tax was set equal to the corporate income tax rate at adoption.
    - Equating labor and corporate rates removes distortions between activities conducted in corporate form or as an individual, so long as dividend and capital gains are untaxed.
    - Estonia eliminated its CIT altogether (taxing dividends at the flat rate); Latvia reduced the CIT rate far below the flat tax on labor income.
  - Of the remaining countries, three tax corporate income more heavily than labor income; Lithuania taxes corporate income less heavily.
  - Most retain withholding on dividends; Russia introduced a credit against personal liability to mitigate withholding effects.
  - Slovak Republic uniquely set the single VAT rate equal to the common labor and CIT rate (19 percent).
  - Treatment of capital income varies:
    - Second wave generally taxes interest and capital gains at or close to the flat rate.
    - Latvia exempts them entirely; Estonia taxes interest at below the flat rate.
  - Differences in personal allowances:
    - Generally increased, especially in the second wave.
    - Georgia is an exception with no threshold after reform.

- Common themes that recur
  - Except in Latvia and Georgia, adoption of the flat tax associated with increases in personal allowances (motivation: limit tax burden increase for less highly paid).
  - With exceptions (Slovak Republic for those active in labor market; Georgia for retired) measures to protect lower-income groups were uncommon.
  - Reforms often accompanied elimination or reduction of exemptions under PIT and CIT (bringing capital income into tax, eliminating preferences).
  - Except in Latvia and Russia, adoption was followed by a reduction in PIT revenue. Base-broadening and behavioral responses often did not offset rate reductions and increased allowances.
  - Many cases saw reductions in the CIT rate; ongoing reductions in CIT have been more marked than reductions in the flat labor tax.
  - Most of the second wave reforms (Russia, Slovak Republic, Georgia) included reductions in social contributions.
  - Second wave reforms typically accompanied by substantial increases in indirect taxation, especially excises, with "a strong increase in revenue from this source."

### Analytical considerations: can a "flat tax" be fully optimal?

- Theoretical framing
  - Optimal non-linear income tax problem (Mirrlees (1971)): balancing efficiency costs of distortions to work effort against distributional advantages of collecting revenue mainly from the better-off.
  - General property: marginal tax rate lies everywhere between zero and one.
  - This provides no direct implication about how marginal tax rate should vary with income.

- Mirrlees (1971) insight
  - Simulations for a utilitarian objective function, Cobb-Douglas preferences in consumption and leisure, and a log-normal distribution of abilities suggested optimal tax schedules "clos[e] to linearity."
  - Quoted result: "[p]erhaps the most striking feature of the results is the closeness to linearity of the [optimal] tax schedules" (p.206).
  - Practical difficulties of non-linear schedules (administration, compliance, tax arbitrage) imply a flat tax could be a close proxy for the best possible income tax.

- Subsequent work
  - Later simulation work (notably Tuomala (1990)) challenged or extended the Mirrlees result (text indicates further discussion follows).

*Source: _wp06218 - 35.6 to 5 percent.*

### conclusion is not robust: with more inequality-averse social welfare functions than the

### _wp06218 - conclusion is not robust: with more inequality-averse social welfare functions than the

### Optimal tax schedule: shapes and robustness
- Conclusion is not robust: with more inequality-averse social welfare functions than the utilitarian, alternative patterns of labor supply response, and other assumptions on the distribution of underlying abilities, the optimal schedule can be highly nonlinear.
- No overwhelming theoretical reason that the optimal tax schedule is progressive in the unqualified sense (average rate increases at all income levels) or that the marginal tax rate is everywhere increasing.
- If the distribution of abilities is bounded above:
  - Pareto efficiency requires the marginal tax rate on the highest earner be zero.
  - This implies a decreasing marginal tax rate in the upper income range and a decreasing average rate (since a marginal tax rate of zero must lie below the average rate15).
- With an unbounded distribution, the optimal tax schedule may take a variety of shapes; special cases identified where marginal tax rate either increases or decreases with income (Boadway, Cuff and Marchand (2000)), with an inverted U shape often plausible.
- The optimal tax literature offers no clear presumption for or against flat taxes; linearity may be reasonable given uncertainty over forms of non-linearity and lack of guidance on which best represents reality.

### Flat tax, evasion and corruption (Hindriks, Keen and Muthoo (1999); HKM)
- Context: income exogenous; tax inspectors observe income costlessly and may collude with taxpayers in understating income for bribes; inspectors may over-report if not bribed; government maximizes revenue and can set tax schedule T(Y), penalty structure, and reward system for inspectors.
- HKM (Corollary 1) result: a linear (proportional) tax, set at an appropriate rate and combined with an appropriate penalty structure and a simple wage payment to inspectors, can maximize revenue without inducing tax evasion or payment of bribes.
- Stylized argument as presented:
  - If collusion detected with probability π and collective penalty equals full income Y (stronger than needed16), misreporters report zero liability.
  - Expected collective payment if misreport: πY.
  - Collective payment if truthful: T(Y).
  - For revenue-maximizing, truth-preserving T, it must be that :)(YTY=π which implies a simple proportional tax at rate equal to the probability of detection.
- Caveats and limitations:
  - Result is highly special and stylized.
  - Non-linear revenue-maximizing, corruption-proof and evasion-proof schedules also exist if commissions are paid to inspectors on collected tax revenue17.
  - The result requires penalties higher than society commonly appears willing to apply for tax offences.
  - HKM thus supports the claim that flatness provides some protection against evasion and corruption, but also stresses that the level of taxation matters and is ultimately restricted by detection and punishment likelihood.

### Political economy argument for linearity
- Brennan and Buchanan (1977) argument: flat tax limits government size by restricting policymakers (assumed to pursue self-interest and revenue maximization) to a single marginal tax rate, thereby limiting revenue and potential government excess.
- Critics and counterpoints:
  - Institutional devices (fiscal rules, simple revenue limits) can limit government size without forcing inefficient tax instruments.
  - If policymakers derive selfish benefit from increased consumer welfare (e.g., higher chance of retaining office), restricting them to less efficient tax instruments reduces real income available for goods they value, potentially reducing consumer welfare (Edwards and Keen (1996)).

### Equity and progressivity: definitions and preliminaries
- With basic exemption, a flat tax is progressive in the sense that the average rate increases with income.
- More relevant question: whether a flat tax is more or less progressive than the tax scheme it replaces.
- Definition: tax scheme M with schedule )(YT_M is ‘more progressive’ than L with )(YT_L iff the concentration curve of tax liabilities under M lies everywhere below that of L19 — i.e., the poorest p percent pay a smaller share of total tax revenue under M than under L, for all p20.
- Single-crossing results:
  - If M and L raise same revenue and )(YT_M crosses )(YT_L once from below, M is more progressive (Hemming and Keen, 1983).
  - If they cross twice on the support of pre-tax income, concentration curves cross once and no unambiguous progressivity comparison can be made (Dardoni and Lambert (1988))21.

### Progressivity comparison between a flat tax (F) and a pre-existing progressive tax (P)
- Pre-reform tax P assumed to satisfy:
  - 0)(=YT_P for Y ≤ A_P (tax-free amount).
  - 0)(),(>′′′YTYT for all Y > A_P: marginal tax rate strictly positive and everywhere increasing above A_P22.
- Revenue-neutral comparisons (F and P raise same revenue):
  - If A_F < A_P then T_F crosses T_P only once from above; flat tax is unambiguously less progressive than predecessor (matching Georgian case, though that reform was not revenue neutral).
  - If A_F > A_P:
    - Panel A case: flat tax marginal rate ≥ highest pre-reform marginal rate → schedules cross at most once → flat tax unambiguously more progressive than predecessor (best matches Lithuanian flat tax; reform was not revenue neutral).
    - Panel B case: flat rate between lowest and highest pre-reform marginal rates → schedules cross twice → concentration curves cross twice → neither tax unambiguously more progressive. This typifies Estonia, Russia, Slovak Republic, Ukraine (typical second-wave reforms).
  - For any given inequality measure, there exists a critical constant marginal rate where flat tax with higher (lower) rate than critical generates less (more) inequality in after-tax incomes26.
- Non-equal yield comparisons:
  - Must distinguish distribution of tax payments (progressivity) from distribution of after-tax income (inequality in disposable incomes).
  - Normalize schedules by dividing by aggregate yield; compare pre-reform schedule with a notional flat tax having the actual threshold but marginal rate = actual × (pre-reform revenue / flat tax revenue).
  - For revenue-loss flat tax reforms:
    - Georgian reform: notional steepened flat tax intersects pre-reform only once from above → distribution of tax payments unambiguously less equal after flat tax.
    - Typical second-wave reforms (higher threshold, net revenue loss): ambiguity remains; steepening unlikely to remove higher crossing.
    - Lithuanian revenue-increasing reform: comparison with a flatter notional schedule leads to double-cross → impact ambiguous.
- Overall: country-specific impacts varied:
  - Lithuania: movement towards flat tax had strongly progressive elements.
  - Georgia: had strongly regressive elements.
  - Many reforms: effect intrinsically ambiguous.

### Progressivity implications of noncompliance (Slemrod-style model)
- Setup: individuals choose amount E to conceal to maximize )()(ECEYTY−−− with C(E) increasing and convex; first-order condition (3): )()(ECEYT′=−′.
- Define E = E(Y) and effective tax schedule T*(Y) ≡ T(Y − E(Y)), the tax actually paid.
- For flat tax:
  - Marginal tax rate equal for all declaring above exempt amount → (3) implies all such taxpayers conceal the same amount → same tax escape.
  - Range of lower-true-income taxpayers choose to pay no tax → effective schedule equivalent to statutory schedule with increased tax-free amount.
  - Proportionate reduction in tax liability due to noncompliance is higher at lower income levels → tends to improve vertical equity: T*_F is more progressive than statutory T_F27.
- For pre-existing progressive tax:
  - Better off conceal more: )1,0()(∈′YE28 and they face higher marginal tax rates → for each dollar concealed, rich escape larger absolute tax amount.
  - If ln(T_P) is convex (stronger condition), rich escape a larger proportion of their tax liability → noncompliance worsens vertical equity: T*_P less progressive than statutory T_P.
- Net implication: noncompliance tends to make the flat tax appear more progressive and pre-existing taxes less progressive — mitigating concerns that flat tax adoption necessarily reduces progressivity.

### Other considerations affecting distributional impact
- Changes in tax base (personal and corporate) and accompanying tax/spending measures are important:
  - Scaling back personal exemptions/deductions (other than basic personal allowances) increases progressivity; Peichl (2006) shows elimination effects monotonically increase proportionate income loss across deciles.
  - Movement to a flat tax equalizes value of allowances across taxpayers, reducing regressive distributional effects of some allowances.
- Distributional impact must be evaluated for the entire tax-and-benefit system:
  - Many reforms accompanied by indirect tax changes (excise increases, VAT exemption reductions) to achieve revenue neutrality; distributional effects vary by country and depend on preference patterns and income distribution.
  - Evidence that VAT/excises evaluated relative to lifetime consumption can be broadly neutral29.
  - Gasoline tax increases may be progressive in some lower income countries where car use is concentrated among better-off (Slovak Republic example)30.
- Treatment of capital income matters:
  - Changes in taxation of capital income (interest, dividends) can be ambiguous: retirees reliant on capital income may be negatively affected; distributional outcomes depend on portfolio holdings and opportunities for tax avoidance or capital flight.
- Incidence assumptions:
  - Analytical results assume labor income taxes incidence falls entirely on workers. If higher-paid employees are more mobile or have easier avoidance options, they may bear less of the burden and benefit less from flat tax reforms than static analysis suggests.

### Evidence from enacted flat tax reforms
- Only two detailed distributional studies of enacted flat taxes identified:
  - Russia (Sinelnikov-Mourylev et al (undated)): report 2001 reform associated with increased progressivity in the Kakwani sense for wage distribution; cross-region revenue regressions against tax base measures also indicate increased progressivity.
  - Slovak Republic (World Bank (2005)): simulations using household survey data; report that the Kakwani index (K) increased (signaling greater progressivity) quite substantially.
- Both studies focus on global summary measures (Kakwani index for distribution of tax payments; Reynolds-Smolensky index for distribution of after-tax income) to produce clear rankings; drawback is dependence on a single-number summary that embodies particular normative choices.
- Overall conclusion from these studies: effect of reform on global progressivity has been modest and perhaps even beneficial in the cases examined.

*Source: _wp06218 - conclusion is not robust: with more inequality-averse social welfare functions than the (excerpt).*

### 0.114 to 0.192. The Reynold-Smolensky (RS) index, on the other hand, was broadly

### _wp06218 - 0.114 to 0.192. The Reynold-Smolensky (RS) index, on the other hand, was broadly

### Distributional indices: Kakwani (K) versus Reynold-Smolensky (RS)
- Reported change in K in the range "0.114 to 0.192."
- The Reynold-Smolensky (RS) index was broadly unchanged.
- Mathematical relation reported exactly as in source: KttRS))1/((−=, where t is the overall average tax rate.
- Interpretation:
  - If t falls (as in the Slovak Republic), K may rise even though RS does not.
  - Implication: distribution of tax payments can become more unequal (K rises), while the equalizing impact on after-tax income (RS) can remain essentially unchanged.

### Key conceptual point on progressivity and perception
- A substantial tax reduction for the highest earners can coincide with an increase in summary indices of progressivity (as measured by K), but this is counterintuitive and hard to explain to non-economists.
- Caution: Changes in summary indices do not tell the whole story; distributional impacts of flat tax reforms are often complex and not unambiguously adverse for the least well-off.

### Work incentives: principle and qualitative effects
- Effect of moving to a flat tax depends on changes in:
  - Average Tax Rate (ATR)
  - Marginal Tax Rate (MTR)
- Substitution effect: increased effort iff MTR falls.
- Income effect (under leisure normality): increased effort if ATR rises.
- For the stylized reform (panel B of Figure 2) the pattern across income ranges:
  - Those with incomes to right of point α or in region B: reduction in MTR.
  - Incomes above higher crossing point, or below lower: lower ATR.
  - Only in ranges C and E are effects clearly toward less and more effort respectively; other ranges are ambiguous.
- For highest income range F: substitution effect → greater effort; income effect → lower effort; net effect ambiguous.
- Empirical literature does not suggest especially strong labor supply responses among high earners.

### Work incentives: participation decisions
- Participation impact is ambiguous:
  - Increasing tax threshold encourages participation (more can be earned tax-free), likely increasing participation among lowest paid workers (ranges B and C).
  - Increased MTR could discourage participation at somewhat higher income levels (D or E).
- If flat tax adoption is accompanied by measures tilting social support to in-work benefits (as in the Slovak Republic), participation increases among lowest paid are reinforced.

### Evidence on labor supply responses (empirical)
- Broad consensus: effects of tax changes on effort of primary workers are modest.
- Empirical study cited: Ivanova, Keen and Klemm (2005; IKM) on Russian flat tax reform:
  - Reform feature: lower-income group largely unaffected (threshold broadly unchanged in real terms; marginal tax rate increased only from 12 to 13 percent), while another group had marginal rate fall from 20 or 30 percent.
  - IKM approach: treat unaffected group as control and affected group as treatment; use differences-in-differences.
  - Findings: no evidence of strong effects on work effort—neither gross income nor hours worked increased more among treatment than control; when differences are significant they point to increased effort in the control group.

### Compliance, administration, and simplicity: general considerations
- Claimed merits of flat tax: simplification, improved compliance, reduced administration costs.
- Caveats:
  - Tax-free threshold often implies effectively two marginal tax rates (one zero), so issues of arbitrage, withholding, averaging persist.
  - Complications remain for withholding from multiple jobs.
  - In Georgia, elimination of basic allowance aimed to remove withholding difficulties; perceived compliance improvement is largely anecdotal and confounded by concurrent administrative changes (penalties, reregistration).
  - Major sources of tax complexity often stem from defining/monitoring tax base and exemptions, not just rate structure. Base-broadening that cleans exceptional treatments (e.g., Slovak Republic, Ukraine) may yield significant simplicity gains.

### Evidence on perceptions and compliance
- Russian survey evidence: proportion of respondents rating perceived complexity as important/very important to compliance was virtually the same in 2002 as in 1998 (Ivanova, Keen, and Klemm (2005)).
- Theoretical frameworks on how rates affect concealment:
  - Slemrod (2001) style: with convex concealment cost C, increase in marginal tax rate increases amount concealed → higher MTR → more concealment.
  - Allingham and Sandmo / Yitzhaki (ASY) gamble model: an increase in proportional tax rate increases the proportion of income concealed if penalties increase with the amount of tax attempted to escape; conversely, tax cuts can increase risk-taking and thus evasion.
- Empirical review (Andreoni, Erard and Feinstein (1998)): evidence on taxation and income concealment is mixed.

### Russian evidence on compliance post-reform
- IKM methodology: estimate ‘true’ gross income as sum of consumption and taxes paid; compare ratio of declared income to estimated ‘true’ income for treatment versus control.
- Findings reported exactly:
  - Ratio of consumption to income for the control group remained at around 75 percent.
  - That for the control group increased from around 52 to 70 percent.
  - (Median growth rates tell a similar story.)
- Interpretation:
  - Evidence suggests an increase in compliance in Russia.
  - If attributable to parametric reform, this aligns better with the Slemrod framework than ASY.
  - However, concurrent administrative changes and high-level statements (e.g., by President Putin) make attribution uncertain.

### Laffer-type effects via improved compliance
- Theoretical possibility: taxpayers at crossing points of pre-existing and flat tax schedules could optimally conceal less under the flat tax and thus pay more tax in equilibrium, even if statutory rates fell.
- Continuity implies nearby incomes may also pay more after reform due to increased compliance incentives.
- Empirical observations: revenue experiences, including Russia’s, suggest such behavioral compliance gains were not large enough to offset direct revenue loss from rate reductions.

### Automatic stabilization (built-in flexibility)
- Common belief: flat tax weakens automatic stabilizers.
- Analytical points:
  - Movement to flat tax reduces MTR for highest and lowest wage earners but increases MTR at intermediate levels; aggregate marginal tax rate (built-in flexibility) effect is ambiguous.
  - Aggregate marginal tax rate defined as B_i ≡ ∫ dY Y f(Y) T_i′(Y) (equation (4) in source).
  - No threshold (pure proportional tax): aggregate marginal tax rate falls → conventional view correct (stabilizers weaken).
  - With threshold/exempt amount under flat tax:
    - Revenue neutrality implies higher marginal rate above threshold to raise same revenue, tending to increase built-in flexibility.
    - Built-in flexibility may increase or decrease depending on pre-reform liability progression, threshold level, and income distribution shape.
    - Appendix 4 result (as summarized): built-in flexibility under the flat tax is greater the lower is the ratio of average incomes above the threshold to the level of the threshold itself (i.e., the more income concentrated just above the threshold).
  - Empirical note: income shocks tend to be concentrated at lower income levels; if these lie just above the threshold, built-in flexibility may be stronger under the flat tax.

### Political economy: drivers of adoption and sustainability
- Some flat tax moves were broadly Pareto-improving (e.g., Russia—no change for those paying 12 percent; gain for those paying higher rates).
- Rapid implementation (e.g., "less than 50 days" from announcement to presidential approval in Russia) may have hindered mobilization of opposition.
- Flat tax adoption commonly occurred as part of broader reform packages; political acceptability often tied to whole-package effects.
- Voting theory (Bohn and Stuart (2003)) suggests linear taxes are unlikely political equilibria; median-voter logic predicts redistribution toward the middle—opposite of the typical flat tax pattern in Panel B.
- Common characteristics of adopting countries:
  - Rapid increases in income inequality during transition from planned economies (Gini coefficient in Eastern Europe and the Former Soviet Union rose from around 0.25 to 0.35 during the 1990s).
  - Empirical finding: more unequal societies sometimes redistribute less (Lindert (2000) — the ‘Robin Hood paradox’).
  - Explanations include social preferences for relative consumption (Persson (1995)) or emergence of strong elites and distrust in government capacity to redistribute.
- Structural and contextual factors that may have enabled adoption:
  - Market economies with little tradition of individual tax payment and dysfunctional tax administrations—conditions conducive to compliance gains from rate reductions.
  - Flat tax adoption often followed fundamental political change (examples in source): pro-liberalization government in Estonia in 1992; election of Vladimir Putin in 2000; the 2003 ‘Rose Revolution’ in Georgia; 2004 coalition government in Romania.
  - These governments had strong mandates for reform, weak traditions of tax compliance/enforcement, and PITs that raised relatively little revenue.

*Source: Excerpt from the IMF working paper content unit provided.*

### 4.1 percent of GDP on average in these eight countries in 2003, for example, compared to an

### _wp06218 - 4.1 percent of GDP on average in these eight countries in 2003, for example, compared to an

### Background and motivation for low-rate flat taxes
- 4.1 percent of GDP on average in these eight countries in 2003, for example, compared to an OECD average of 6.4.
- A key reason for adopting a low-rate flat tax, especially in the second wave of reforms, has been to signal a fundamental regime change toward more market-oriented policies.
- The Brennan-Buchanan view (Section II.A) frames the flat tax as a commitment device intended to tie government hands and limit the size of government.
- Marketing and signaling advantages were explicitly noted in evaluations such as Ministry of Finance (2005; p.79) for the Slovak reform.
- Successful cases combined rate reduction with:
  - commitments to improved tax enforcement and encouragement of voluntary compliance (examples: Russia and Georgia), and
  - scaling back of tax preferences and exemptions (examples: the Slovak Republic and Ukraine).

### Key empirical findings and lessons
- Revenue effects:
  - Except in Russia, the second wave of low-rate flat tax reforms have been associated with a reduction in revenue from the PIT.
  - Behavioral responses may have mitigated the revenue loss, but in no case does there appear to have been a Laffer effect: these reforms have not set off effects strong enough for them to pay for themselves.
  - In Russia, little evidence suggests the strong revenue performance after the reform was due to the flat tax itself; it appears to have reflected wider macroeconomic recovery.
- Compliance and administration:
  - There is evidence that compliance improved after the Russian reform.
  - This improvement may fit circumstances where administration is extremely weak, but no firm evidence ties the compliance improvement solely to the parametric tax reform rather than to contemporaneous changes in enforcement.
- Labor supply and work incentives:
  - The impact of the flat tax on work incentives is not clear cut in principle.
  - For highest income groups, second wave reforms reduce marginal tax rates (increasing incentives) but also reduce average tax rates (blunting incentives); ambiguous effects may appear elsewhere in the wage distribution.
  - The only household-level study to date (for Russia) does not detect any significant impact on work effort.
- Distributional effects and progressivity:
  - Distributional effects of movement towards a flat tax are potentially complex.
  - Second wave reforms that increase the basic tax-free amount are beneficial to both the lowest and the highest earners.
  - Compliance effects may plausibly lead to an increase in effective progressivity.
  - There is no general presumption that movement to a flat tax is associated with a reduction in progressivity.
  - Commonly used summary indices of progressivity—in the few studies—show an increase in progressivity but may overstate the point.
- Complexity and simplification:
  - Flatness itself is a simplification and can eliminate some tax arbitrage, but the primary sources of complexity are exemptions and special treatments.
  - Limited survey evidence for Russia does not suggest the system was widely seen as significantly less complex after adoption of the flat tax.
- Base-broadening and fiscal packages:
  - The rate-cutting aspect of second wave reforms has sometimes been packaged with significant base-broadening via elimination of exemptions and preferences.
  - Example: In Ukraine, base broadening measures were estimated to have increased revenue by around one point of GDP.
  - Base broadening likely contributed to improved horizontal equity, efficiency gains, and greater simplicity.
- Automatic stabilizers:
  - Movement to a flat tax may plausibly strengthen the automatic stabilizers, not weaken them, though this has received little attention and little empirical study.
- Political economy and signaling:
  - The flat tax has commonly—almost universally—been adopted by new governments anxious to signal a regime shift toward more market-oriented policies.
  - Where no such reputation needs to be acquired, the appeal of the flat tax is less.
  - The spread of the flat tax undermines its signaling value: it may prove too easy to mimic.
  - Political economy considerations point toward adopting rate schedules that tend to benefit middle income earners—the group that tends to lose most from second wave flat taxes.
  - The more interesting question may be whether there will be any defections from the flat tax community.

### Sustainability, technical tensions, and open questions
- Diversity of design:
  - First wave flat taxes typically set the single rate at the highest of pre-reform marginal tax rates; the second wave typically set it at the lowest.
  - Variations include whether the same rate applied to corporate earnings, changes in personal allowances (e.g., Slovak Republic increased allowance; Georgia eliminated it), and differing treatments of social contributions, benefits, and indirect taxation.
- Empirical limitations:
  - Except for Russia and the Slovak Republic, there is a paucity of careful empirical analysis; few reform analyses are based on household-level data.
  - There is an evident need for studies in other flat tax countries using household-level data.
- Taxing internationally mobile capital:
  - Structurally, adopted flat taxes do not provide a coherent framework for taxing internationally mobile capital income.
  - Tensions are most evident in first wave flat taxes, which may be viewed as special cases of a dual income tax.
  - Addressing capital income taxation pressures may point toward:
    - further pursuing the dual income tax logic (for example, in treatment of the self-employed), and
    - nudging second wave reformers with higher flat rates toward decoupling taxation of capital income from taxation of labor income.

### Appendix I: summary of post- and pre-adoption tax system snapshots (selected country highlights)
- Estonia:
  - 1994: General Income Tax rate 26; Personal allowances EEK 3,600 3/; Corporate Tax and Dividends 26; Interest: 10 percent final withholding, 4/; Capital gains usually taxed as ordinary income; Employer Social Contributions 33; Employee  zero; VAT Rates 18.
  - 1993: Rates 16, 24, 33; Personal allowances EEK 2,400; Corporate Tax 35; Capital gains usually untaxed; Employer Social Contributions 33; Employee zero; VAT Rates 18.
- Lithuania:
  - 1994: General Income Tax 33; Personal allowances LTD 115; Corporate Tax and Dividends 29 2/; No withholding system; Capital gains taxed as ordinary income; Employer Social Contributions 30; Employee 1; VAT Rates 18.
  - 1993: Rates 10, 18, 24, 28, 33; Personal allowances LTD 35; Corporate Tax 29; NIA for interest/capital gains treatment; Employer Social Contributions 30; Employee 1; VAT Rates 18.
- Latvia:
  - 1997: General Income Tax 25; Personal allowances LVL 252, plus 126 for spouse and per child; Corporate Tax and Dividends 25 1/; No change in interest/capital gains; Employer Social Contributions 33; Employee 5; VAT Rates 18, 9.
  - 1996: Rates 25, 10; Personal allowances LVL 270, plus 138 for spouse and per child; Corporate Tax and Dividends 25; Dividends paid by Latvia company exempt; Deposit interest paid by Latvian bank exempt; Employer Social Contributions 37; Employee 1; VAT Rates 18, 9.
- Russia:
  - 2001: General Income Tax 13; Personal allowances RU 4,800 per annum; Corporate Tax and Dividends 35 1/; 30 percent withholding, with credit for underlying CIT 2/; Interest: 13 percent, except 35 percent on 'excessive' bank interest; Capital gains: taxed as ordinary income; Employer Social Contributions Marginal rates decreasing from 35.6 to 5 percent; Employee Zero; VAT Rates 20, 10.
  - 2000: Rates 12, 20, 30; Personal allowances RU 3,168 per annum; Corporate Tax and Dividends 30; 15 percent final withholding; Interest: 15 percent final withholding; Capital gains: taxed as ordinary income; Employer Social Contributions 38.5; Employee 1; VAT Rates 20, 10.
- Slovak Republic:
  - 2004: General Income Tax 19; Personal allowances SK 80,832; spouse SK 80,832 1/; Corporate Tax and Dividends 19; Exempt dividends; Interest: 19 percent withholding; Capital gains taxed as ordinary income; Employer Social Contributions 35.2; Employee 13.4; VAT Rates 19.
  - 2003: Rates 10, 20, 28, 35, 38 2/; Personal allowances SK 38,760; spouse SK 12,000; Corporate Tax and Dividends 25; 15 percent final withholding; Interest: 15 percent final withholding; Capital gains: taxed as ordinary income; Employer Social Contributions 38.2; Employee 12.8; VAT Rates 20, 14.
- Ukraine:
  - 2004: General Income Tax 13; Personal allowances Hrv 61.5 per month for those earning less than 1.4 minimum monthly living amounts; Corporate Tax and Dividends 25; 13 percent final withholding; Interest: final withholding at 13 percent 2/; Capital gains: taxed as ordinary income; Employer Social Contributions No change 2-3.5; Employee 20; VAT Rates 20.
  - 2003: Rates 10, 15, 20, 30, 40; Personal allowances Hrv 17 per month; Corporate Tax and Dividends 30; 30 percent withholding in advance; Interest: exempt; Capital gains taxed as ordinary income; Employer Social Contributions 37 plus sector dependent accident 9; Employee 1.75-3; VAT Rates 20.
- Georgia:
  - 2005: General Income Tax 12; No basic exemption; Corporate Tax and Dividends 20; No change in interest/capital gains treatment; Employer Social Contributions No change; Employee 20; VAT Rates zero; VAT Rate column value 18 appears for 2004.
  - 2004: Rates 12, 15, 17, 20; Personal allowances GEL 108 basic exemption; Corporate Tax and Dividends 20; 10 percent final withholding; Interest: 10 percent final withholding; capital gains taxed as ordinary income; Employer Social Contributions 31; Employee 2; VAT Rates 20.
- Romania:
  - 2005: General Income Tax 16; Personal allowances ROL 2.5 million per month, with additional 1 million per dependent; Corporate Tax and Dividends 16; 10 percent final withholding 1/; Interest: 10 percent final withholding; capital gains: 10 percent; Employer Social Contributions no change; Employee no change; VAT Rates 19, 9.
  - 2004: Rates 18, 23, 28, 34, 40; Personal allowances ROL 2 million per month, with additional 1 million for spouse and each dependent child; Corporate Tax and Dividends 25 2/; 5 percent final withholding; Interest: 1 percent final withholding; capital gains: 1 percent final withholding; Employer Social Contributions 22-32 3/; Employee 9.5; VAT Rates 19, 9.

- General notes from the appendix:
  - 1/ Rates in percent.
  - 2/ “NIA” indicates no information available.
  - 3/ Social contribution rates are in several cases aggregates of distinct charges.
  - 4/ Corporation tax is classical (that is, no allowance made for underlying corporate tax paid) unless otherwise indicated.
  - 5/ Description of interest and capital gains taxation relates to individual taxpayers.

*Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06218.pdf*

### Appendix II. Composition of Tax Revenues Before and After Introduction of the Flat Taxes

### Appendix II. Composition of Tax Revenues Before and After Introduction of the Flat Taxes

### Composition of tax revenues (In percent of GDP)
- Estonia
  - Personal Income Tax: Year Before the Reform 8.5; Year of the Reform 8.1
  - Corporate Income Tax: Year Before the Reform 4.8; Year of the Reform 3.5
  - Indirect Taxes: Year Before the Reform 11.1; Year of the Reform 13.3
- Georgia
  - Personal Income Tax: Year Before the Reform 2.7; Year of the Reform 2.5
  - Corporate Income Tax: Year Before the Reform 1.6; Year of the Reform 1.9
  - Indirect Taxes: Year Before the Reform 8.1; Year of the Reform 11.0
- Latvia
  - Personal Income Tax: Year Before the Reform 5.4; Year of the Reform 5.6
  - Corporate Income Tax: Year Before the Reform 2.0; Year of the Reform 2.4
  - Indirect Taxes: Year Before the Reform 12.6; Year of the Reform 12.5
- Lithuania
  - Personal Income Tax: Year Before the Reform 5.0; Year of the Reform 5.4
  - Corporate Income Tax: Year Before the Reform 5.3; Year of the Reform 2.5
  - Indirect Taxes: Year Before the Reform 6.2 1/; Year of the Reform 6.3 1/
- Romania
  - Personal Income Tax: Year Before the Reform 3.0; Year of the Reform 2.3
  - Corporate Income Tax: Year Before the Reform 2.7; Year of the Reform 2.4
  - Indirect Taxes: Year Before the Reform 10.2; Year of the Reform 10.9
- Russia
  - Personal Income Tax: Year Before the Reform 2.4; Year of the Reform 2.9
  - Corporate Income Tax: Year Before the Reform 5.5; Year of the Reform 5.8
  - Indirect Taxes: Year Before the Reform 8.6; Year of the Reform 9.9
- Slovak
  - Personal Income Tax: Year Before the Reform 3.3; Year of the Reform 2.6
  - Corporate Income Tax: Year Before the Reform 2.8; Year of the Reform 2.4
  - Indirect Taxes: Year Before the Reform 9.8; Year of the Reform 11.4
- Ukraine
  - Personal Income Tax: Year Before the Reform 5.1; Year of the Reform 3.8
  - Corporate Income Tax: Year Before the Reform 5.0; Year of the Reform 4.7
  - Indirect Taxes: Year Before the Reform 8.8; Year of the Reform 8.6

- Unweighted average
  - Personal Income Tax: Year Before the Reform 4.4; Year of the Reform 4.1
  - Corporate Income Tax: Year Before the Reform 3.7; Year of the Reform 3.2
  - Indirect Taxes: Year Before the Reform 9.4; Year of the Reform 10.5

- Footnote
  - 1/ Includes only VAT.

*Sources: IMF country documents; and Fund staff estimates.*

### Equity aspects of the flat tax: key analytical conclusions
- The effective flat tax schedule is more progressive than the statutory schedule.
  - The comparison uses the statutory schedule T_F(Y) = max[0, (F - A - tY)] and the effective schedule T*_F(Y) = max[0, (F - A - tY - E)], where E denotes the amount of concealment, the same for all taxpayers.
  - Normalized schedules are defined as T̂_F ≡ T_F / R_F and T̂*_F ≡ T*_F / R*_F.
  - The marginal rate under the normalized effective flat tax schedule is higher than that under the normalized statutory because the effective schedule raises less than the statutory.
  - Where the normalized statutory and effective schedules cross (as they must, since they raise the same revenue), the effective cuts from below—and so is more progressive.

- The effective pre-existing tax schedule is less progressive than the statutory schedule.
  - The argument uses Jakobsson (1976): tax M is more progressive than L, whatever the distribution of income, iff the elasticity of tax payments under M is greater than under L.
  - Elasticity under the effective schedule T* (omitting the subscript P) yields inequalities showing elasticity under the effective schedule is lower because 1 < E′ and because of log convexity of T.
  - Conclusion: the effective pre-existing (pre-reform) schedule is less progressive than the statutory pre-existing schedule.

### Automatic stabilizers and the flat tax: key results
- Revenue-equality condition
  - The condition that the two tax schedules raise the same revenue R implies (normalizing population to unity):
    - ∫0∞ ∫0∞ (]0),.(max[)(dY Y f A Y t ) = R (as expressed in the text).
    - A substitution yields ∫0∞ (1 - A F t) dY = ... (equations A4.1–A4.2 summarize the normalization and revenue identity).
- Difference in built-in flexibility (BB) between pre-reform and flat tax
  - Defined as BB_PF ≡ ∫0∞ ∫0∞ [T_P(Y) - T_F(Y)] dY f(Y).
  - Manipulations yield expressions (A4.3–A4.6) showing BB_PF can be written in terms of integrals involving A, F, t, and Y.
- Comparative claims
  - If A = 0, built-in flexibility is unambiguously greater under the pre-reform tax.
  - If the progressivity of the pre-reform schedule is sufficiently weak (i.e., liability progression—the elasticity of tax payments with respect to pre-tax income—is sufficiently close to unity) and/or if [AYYE] is sufficiently close to A, then built-in flexibility can be higher under the flat tax.
  - The text notes that progressivity of the pre-reform system (marginal rate > average rate) is central to these comparative results.

*Appendix II. Composition of Tax Revenues Before and After Introduction of the Flat Taxes — Source PDF unit _wp06218.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06218.pdf_
