## _wp0516 — 1. The PIT Rate Structure Before and After Reform (excerpts)

## Source details

**Canonical URL:** [_wp0516 — 1. The PIT Rate Structure Before and After Reform (excerpts)](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp0516.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp0516.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp0516.pdf.json)

---

### Overview and motivation
- At the start of 2001, Russia unified its marginal rates of personal income taxation—previously at 12, 20, and 30 percent—at the flat rate of 13 percent.
- PIT revenue increased by about 46 percent: about 26 percent in real terms, over the next year.
- As a percentage of GDP, PIT revenues increased by nearly one-fifth.
- Paper purpose: assess impact of reform on tax revenue, work effort, wage rates, and taxpayer compliance using macro evidence and RLMS individual- and household-level panel data.

### PIT rate schedule: before and after (summary)
- Before Reform (2000):
  - Below 3,168 rubles — Marginal Rate: 0
  - 3,168 to 50,000 rubles — Marginal Rate: 12
  - 50,000 to 150,000 rubles — Marginal Rate: 20
  - Above 150,000 rubles — Marginal Rate: 30
- After Reform (2001):
  - Below 4,800 rubles — Marginal Rate: 0
  - Above 4,800 rubles — Marginal Rate: 13
- Notes:
  - Basic exemption grew by 30 percent in real terms between 2000 and 2001; remained roughly unchanged relative to average wage (about 12 percent).
  - Certain income types taxed at 35 percent (gambling, lottery prizes, some insurance payments, loans at less than market rates, ‘excessive bank interest’). Dividends taxed at 30 percent (up from 15 percent in 2000).
  - Bank interest taxable if paid at rate exceeding 75 percent of Central Banks’ refinancing rate on ruble deposits or nine percent on foreign currency deposits.
  - Some PIT exclusions eliminated; simplified deductions introduced (standard, social, property, professional).
  - PIT sharing: in 2000 regional governments received 80 percent; from 2001 they received 100 percent.

### Social tax reforms and combined marginal rates
- Before Reform (2000):
  - Employee any income — Marginal Rate: 1
  - Employer any income — Marginal Rate: 38.5
- After Reform (2001):
  - Employee any income — Marginal Rate: 0
  - Employer below 100,000 rubles — Marginal Rate: 35.6
  - Employer 100,000-300,000 rubles — Marginal Rate: 20
  - Employer 300,000-600,000 rubles — Marginal Rate: 10
  - Employer above 600,000 rubles — Marginal Rate: 5 (from 2002: 2)
- Composition of 35.6 percent employer contribution: Pension Fund (28 percent), Social Insurance Fund (5.4), State Employment Fund (1.5), Medical Insurance Fund (3.6).
- Policy implications:
  - For incomes below 100,000 rubles (about 254 percent of the average wage), marginal and average social insurance rate fell by 7.3 points.
  - For those initially paying PIT at 12 percent, net effect was reduction in combined marginal rate of PIT and social insurance of about 1.3 percentage points.
  - Changes calculated as differences between (0.12+0.01+ 0.385)/(1.385) and (0.13+0.356)/(1.356).

### Other concurrent tax and administrative changes
- Corporate income tax: combined federal and regional rate remained at a maximum of 30 percent in 2001 (federal 11 percent plus regional up to 19 percent). Local municipalities could impose profit tax up to 5 percent (used in practice), raising combined maximum to 35 percent. Maximum CIT rate reduced to 24 percent in 2002 (federal 7.5 percent, regional between 10.5 and 14.5 percent, local up to 2 percent).
- Dividend tax: increased in 2001 to 30 percent (non-refundable credit for underlying CIT paid); lowered to 6 percent in 2002 when imputation credit was eliminated.
- Turnover and other taxes: Social Infrastructure Maintenance Tax abolished; Road User tax reduced from 2.5 to 1.5 percent and abolished start of 2003; VAT rates unchanged (20 percent basic, 10 percent reduced), some exemptions narrowed; shift from origin to destination basis for CIS trade (mid-year) except Belarus and energy.
- Tax administration: Part I of new tax code (effective January 1, 1999) strengthened legal powers (common taxpayer ID, indirect assessment), greater authority to State Tax Service for allocation across related taxpayers and enforcing debt repayments.
- Part I eliminated ceiling on interest accrued on overdue taxes; some provisions allowed tax obligations deemed discharged once payment order provided to bank.

### Theoretical model and revenue decomposition
- Revenue identity: R = L w ... λ τ, where:
  - τ = (tax-exclusive) tax rate;
  - λ = ratio of declared taxable income to true taxable income (compliance);
  - w = gross wage rate;
  - L = level of employment.
- Proportionate changes approximation:
  - ˆR ≈ ˆL + ˆw + ˆλ + ˆτ
- Key question: could ˆL, ˆw, and ˆλ responses offset reduction in τ to account for strong post-reform PIT performance?
- Theory on gross wage: in formal sector reduced tax wedge expected to cause gross wage w to fall unless compliance increases by greater proportion than tax rate falls; revenue rise more likely the greater is e_D (labor demand elasticity) and e_S (labor supply elasticity), the higher initial tax rate and initial compliance.

### Work effort and compliance theory
- Work effort effects:
  - Substitution effect (lower marginal tax rate) increases pre-tax income and tax base.
  - Income effect (higher net income) can reduce work effort if leisure normal.
  - Individuals at kink points or with discrete contract choices more likely to increase labor supply.
  - Starting marginal rate (inclusive of social insurance) fell by 1.3 percentage points; effects at higher rates more diverse.
- Compliance theory:
  - Allingham-Sandmo (1972) and Yitzhaki (1974): tax reduction can reduce compliance if fines proportional to evaded tax; lower tax increases desired riskiness leading to more evasion under decreasing absolute risk aversion.
  - Other models show evasion may be independent of τ or increase when fines proportional to evaded tax; empirical evidence is mixed (Andreoni et al. (1999) review; Schneider and Enste (2000); Friedman et al. (2000)).
- Recharacterizing income:
  - Possible incentive to shift corporate to personal income given corporate/dividend tax changes and PIT/social tax cuts, but adoption of imputation and coverage differences made a priori effect unclear.

---

### Macro evidence on revenues, wages, and labor share

### Revenue performance (selected observations)
- General government total revenue (percent of GDP): 1994: 34.6; 1995: 36.8; 1996: 35.8; 1997: 39.3; 1998: 34.4; 1999: 33.6; 2000: 36.9; 2001: 37.4; 2002: 37.6; 2003: 36.6
- o/w Personal income tax (percent of GDP): 1994: 2.9; 1995: 2.6; 1996: 2.8; 1997: 3.2; 1998: 2.7; 1999: 2.4; 2000: 2.4; 2001: 2.9; 2002: 3.3; 2003: 3.4
- Profit tax (percent of GDP): 1994: 8.0; 1995: 8.2; 1996: 4.9; 1997: 4.4; 1998: 3.7; 1999: 4.5; 2000: 5.5; 2001: 5.8; 2002: 4.3; 2003: 4.0
- VAT (percent of GDP): 1994: 7.0; 1995: 6.9; 1996: 7.6; 1997: 7.3; 1998: 6.4; 1999: 5.9; 2000: 6.3; 2001: 7.2; 2002: 6.9; 2003: 6.6
- Resource taxes (percent of GDP): 1994: 0.0; 1995: 0.9; 1996: 1.1; 1997: 1.5; 1998: 0.9; 1999: 0.9; 2000: 1.1; 2001: 1.4; 2002: 3.1; 2003: 3.0
- Notes:
  - Energy prices increased from late 2000; oil- and gas-related revenues accounted for about one-third of federal government revenues; Kwon (2003) attributes about 80 percent of post-crisis revenue recovery to oil and gas sector (which accounted for about 20 percent of GDP).
  - Breadth of revenue increases suggests common underlying causes beyond PIT reform.

### Wage and compliance aggregate indicators (table highlights)
- Billions of rubles (selected years and series):
  - Gross wage income: 1999: 1934; 2000: 2937; 2001: 3819; 2002: 4995
  - Reported wage income: 1999: 1408; 2000: 2126; 2001: 2826; 2002: 3749
  - Hidden wage income: 1999: 526; 2000: 811; 2001: 993; 2002: 1246
  - PIT revenue: 1999: 117; 2000: 175; 2001: 256; 2002: 358
  - UST revenue: 1999: 373; 2000: 561; 2001: 652; 2002: 865
  - Reported wage income base: 1999: 1035; 2000: 1565; 2001: 2175; 2002: 2883
  - Net wage income: 1999: 918; 2000: 1390; 2001: 1919; 2002: 2525
- Average effective rates and compliance:
  - Average effective PIT rate: 11.3 (1999); 11.2 (2000); 11.8 (2001); 12.4 (2002)
  - Average effective UST rate: 36.0 (1999); 35.8 (2000); 30.0 (2001); 30.0 (2002)
  - Average effective tax rate inclusive of social taxes: 34.8 (1999); 34.6 (2000); 32.1 (2001); 32.6 (2002)
  - Compliance (reported share of wages): 72.8 (1999); 72.4 (2000); 74.0 (2001); 75.1 (2002)
- Decomposition of PIT revenue growth (approximation from official data):
  - ~13 percent of increase reflects increase in effective rate of PIT itself
  - ~10 percent due to lower rate of social taxation (enlarging PIT base)
  - ~5 percent reflects improved compliance
  - Over 70 percent associated with increase in gross incomes
- Wage developments:
  - After-tax real wage income grew by 18.5 percent in 2001
  - Gross real wages grew by 11.6 percent in 2001
  - GDP growth in 2001 = 5.1 percent

---

### Micro evidence: RLMS panel analysis, identification, and principal results

### Data and empirical strategy
- Data: Russian Longitudinal Monitoring Survey (RLMS), ~3,500 adults per year; analysis uses 2000 and 2001 waves.
- Measurement limits: no direct tax payments or pre-tax incomes; inferred from reported after-tax incomes; limited deduction detail; best/worst-off under-represented.
- Main income variable: reported average monthly wage after taxes over last 12 months from primary employer (pjpayt); alternative wagelm used as robustness.
- Identification: difference-in-differences comparing treatment (initially paying higher marginal PIT rates, 20–30 percent) versus control (0–12 percent). Also household-level analysis using consumption to proxy true net income and to infer compliance.

### Panel results (individual-level, assuming full compliance)
- Distribution of taxpayers by marginal tax rate in 2000 (Table 6): No. taxpayers = 100 (0%), 2130 (12%), 173 (20%), 11 (30%).
- PIT payments by initial marginal rate (real 1995 rubles):
  - Pre-reform mean PIT: control (0–12%) = 356.4; treatment (20–30%) = 2669.9
  - Change (2000–2001): control = 161.8; treatment = -334.9
  - Difference-in-differences (levels): -496.7*** (160.7)
  - Growth rates (mean): control 104.3; treatment -3.2; diff-in-diff = -107.4*** (12.5)
  - Median growth-rate diff-in-diff = -41.9*** (7.0)
- Extended treatment (≥75 percent of threshold) results show similar negative and significant diff-in-diff estimates (levels = -232.3** (104.2); mean growth = -102.0*** (13.2); median = -32.9*** (5.1)).
- PIT, total tax and gross income (Table 9):
  - Levels: change in total tax diff-in-diff = -1860.2*** (452); change in gross income levels not significant.
  - Growth rates: mean gross income diff-in-diff = -41.9*** (3.9); median growth diff-in-diff = -26.0*** (5.7).
- Hours and wage rate (Table 10):
  - Levels: hours worked diff-in-diff = -2.155** (1.174); wage rate change not significant in levels.
  - Growth rates: hours worked mean diff-in-diff = -5.6*** (1.5); wage rate mean growth diff-in-diff = -37.7*** (5.5); median wage-rate growth = -22.4*** (5.4).
  - Interpretation: relative reductions in both wage rate and labor supply contribute to relative decline in gross incomes; adjustments more through wages than employment.

### Household-level results allowing for evasion (consumption-based)
- Treatment defined by estimated gross employment income per working adult ≥ pre-reform higher-rate threshold or by consumption-based criterion.
- Reported-income household results (Table 11):
  - Families: control 1016; treatment 122.
  - Change in total tax paid: control 23.7; treatment -20.7; diff-in-diff = -44.4* (24.0).
  - Growth-rate mean diff-in-diff = -147.8*** (23.9); median = -41.2*** (7.7).
- Consumption-based treatment (Table 12):
  - Families: control 946; treatment 192.
  - Levels: change in total tax paid diff-in-diff = -16.3 (19.0) — not significant.
  - Growth rates: mean diff-in-diff = -67.0** (29.8); median = -18.3*** (3.7).
- Compliance results (Table 14, consumption-based):
  - Pre-reform declared share: control 73.8 percent; treatment 52.0 percent.
  - Change in declared share: control -0.9; treatment +16.4; diff-in-diff = 17.3*** (4.4).
  - Mean growth-rate diff-in-diff = 101.3*** (24.8); median growth diff-in-diff = 34.3*** (5.6).
  - Interpretation: estimated compliance in treatment group increases by about 16 percentage points (from 52 percent to 68 percent) in levels; growth-rate measures indicate large proportional increases (skewed distribution).
- Gross income, wage rate, hours (Tables 15–16, consumption criterion):
  - Pre-reform gross income levels: control 975.0; treatment 3124.8.
  - Change in gross income: control 320.6; treatment -556.7; diff-in-diff = -877.3*** (214.7).
  - Wage rate diff-in-diff (levels) = -12.8*** (3.1).
  - Growth-rate mean diff-in-diff for gross income = -54.3*** (5.9); median = -41.1*** (5.0).
  - Hours worked: no differential change (levels change in hours worked = 1.66 (3.99)).
  - Interpretation: relative gross incomes fell in treatment households driven primarily by lower gross wage rates rather than reductions in hours worked.

---

### Summary of empirical findings and policy lessons

### Principal empirical findings
- The 2001 PIT reform did not account for the observed PIT revenue boom.
  - Stronger PIT growth came from groups little affected by the reform (control).
  - Difference-in-differences estimates for treatment groups on PIT and total tax are negative and often highly significant in both levels and growth specifications.
- Behavioral effects identified:
  - Significant increase in compliance among affected taxpayers: levels estimate ~+16.4 percentage points in declared income share; diff-in-diff = 17.3*** (4.4).
  - Significant reduction in relative gross wage rates for treatment groups; no measurable differential effect on labor supply (hours worked).
- Quantitative highlights:
  - RLMS sample-based PIT revenue growth 2000/2001 = 45.2 percent vs. published 46.3 percent.
  - Pre-reform declared income shares: control 73.8 percent; treatment 52.0 percent.
  - Difference-in-differences for change in PIT (20–30% treatment vs 0–12% control): -496.7*** (160.7) in levels.

### Policy implications
- Tax-cutting reforms of this kind should not be expected to pay for themselves via greater work effort and improved compliance alone.
- In contexts with initially high noncompliance, parametric reforms may be associated with notable compliance improvements; such improvements may be conditional on concurrent enforcement and administrative strengthening.
- Structural labor-market features (elasticities, rigidities) matter for how wage rates and employment adjust to tax changes and thereby mediate revenue effects.
- Caution: data limitations (no direct tax payment observations, under-reporting) mean level estimates should be treated cautiously; however, consistent qualitative patterns across multiple empirical strategies bolster the conclusions.

*Source: _wp0516 — excerpts from IMF working paper content unit.*

### 1. The PIT Rate Structure Before and After Reform ............................................................. 6

### 1. The PIT Rate Structure Before and After Reform

### Overview and motivation
- At the start of 2001, Russia unified its marginal rates of personal income taxation—previously at 12, 20, and 30 percent—at the flat rate of just 13 percent.
- Over the next year, revenue from the personal income tax (PIT) increased by about 46 percent: about 26 percent in real terms.
- As a percentage of GDP, PIT revenues increased by nearly one-fifth.
- The Russian reform has influenced tax policy in other countries (Ukraine, Slovakia, Belarus, Georgia, Guatemala, Kyrgyz Republic, El Salvador, Paraguay, Poland; Estonia and Latvia previously adopted single-rate structures in 1994 and 1995; Bolivia since 1986).
- Purpose of the paper: assess impact of the reform on tax revenue, work effort, wage rates, and taxpayer compliance using macro evidence and individual- and household-level panel data from the Russian Longitudinal Monitoring Survey (RLMS).

### Empirical issues and strategy
- Previous regional-level RLMS-based econometric work (Sinelnikov-Mourylev et al. (2003) and related) suggested a significant revenue effect, attributing about half of the revenue gain to the reduction in marginal rates.
- Potential confounding: regions with larger pre-reform shares of higher-rate taxpayers may also have experienced systematic income changes independent of reform. Micro-level panel data (individual and household) are used here to address such concerns.
- The analysis focuses on positive impacts of the reform (revenue, compliance, labor supply) and does not evaluate efficiency, welfare, or distributional impacts.

### PIT rate schedule: before and after (summary of Table 1)
- Before Reform (2000):
  - Taxable Income (rubles): Below 3,168 — Marginal Rate: 0
  - Taxable Income (rubles): 3,168 to 50,000 — Marginal Rate: 12
  - Taxable Income (rubles): 50,000 to 150,000 — Marginal Rate: 20
  - Taxable Income (rubles): Above 150,000 — Marginal Rate: 30
- After Reform (2001):
  - Taxable Income (rubles): Below 4,800 — Marginal Rate: 0
  - Taxable Income (rubles): Above 4,800 — Marginal Rate: 13
- Notes:
  - The basic exemption grew by 30 percent in real terms between 2000 and 2001 but remained roughly unchanged relative to the average wage (at about 12 percent).
  - Post-reform PIT was not strictly single-rate: certain income types (gambling, lottery prizes, some insurance payments, loans at less than market rates, ‘excessive bank interest’) were taxed at 35 percent. Dividends were taxed at 30 percent (up from 15 percent in 2000).
  - Bank interest became taxable if paid at a rate exceeding 75 percent of the Central Banks’ refinancing rate on ruble deposits or nine percent on foreign currency deposits.
  - Some PIT exclusions were eliminated (military servicemen, expatriates) and a simplified system of deductions introduced (standard, social, property and professional).
  - Sharing of PIT between federal and regional governments changed: in 2000 regional governments received 80 percent of PIT; from 2001 they received 100 percent.

### Social tax reforms and combined marginal rates (summary of Table 2)
- Before Reform (2000):
  - Employee: Any income — Marginal Rate: 1
  - Employer: Any income — Marginal Rate: 38.5
- After Reform (2001):
  - Employee: Any income — Marginal Rate: 0
  - Employer: Below 100,000 rubles — Marginal Rate: 35.6
  - Employer: 100,000-300,000 rubles — Marginal Rate: 20
  - Employer: 300,000-600,000 rubles — Marginal Rate: 10
  - Employer: Above 600,000 rubles — Marginal Rate: 5 (from 2002: 2)
- Composition of the 35.6 percent employer contribution: Pension Fund (28 percent), Social Insurance Fund (5.4), State Employment Fund (1.5), Medical Insurance Fund (3.6).
- Policy implications of social tax changes:
  - For incomes below 100,000 rubles (about 254 percent of the average wage), the marginal and average rate of the social insurance tax fell by 7.3 points.
  - For those initially paying PIT at the lower rate of 12 percent, the net effect of the 2001 reforms was a reduction in the combined marginal rate of PIT and social insurance of about 1.3 percentage points.
  - Because social taxes are charged on a tax-exclusive basis, the indicated changes are calculated as differences between (0.12+0.01+ 0.385)/(1.385) and (0.13+0.356)/(1.356).

### Other concurrent tax and administrative changes
- Corporate income tax:
  - Combined federal and regional rate remained at a maximum of 30 percent in 2001 (federal 11 percent plus regional up to 19 percent).
  - Local municipalities were allowed to impose a profit tax of up to 5 percent (used in practice), raising a combined maximum to 35 percent.
  - The maximum CIT rate was reduced to 24 percent in 2002 (federal 7.5 percent, regional between 10.5 and 14.5 percent, local up to 2 percent).
  - Investment incentives scaled back in 2002 and replaced by accelerated depreciation and other changes.
- Dividend tax:
  - Increased in 2001 (dividends taxed at 30 percent), accompanied by a non-refundable credit for underlying CIT paid.
  - Dividend tax rate lowered to 6 percent in 2002 when the imputation credit was eliminated.
- Turnover and other taxes:
  - Social Infrastructure Maintenance Tax (turnover tax at 1.5 percent) abolished.
  - Road User tax reduced from 2.5 to 1.5 percent; it was abolished at the start of 2003.
  - No changes in VAT rates (basic rate remained at 20 percent, reduced rate 10 percent for basic foodstuffs, children’s goods and some other items), but some exemptions were narrowed (including pharmaceuticals).
  - Shift from origin to destination basis for trade with other CIS countries (mid-year), except trade with Belarus and on energy.
  - Measures adopted to reduce compliance burden on small traders, including threshold for compulsory registration of 1 million rubles of sales (in the preceding month).
- Tax administration reforms:
  - Part I of the new tax code (effective January 1, 1999) strengthened legal powers of tax authorities: common taxpayer identification number, indirect assessment of tax liability in certain cases.
  - Greater authority given to the State Tax Service for allocating income, deductions and credits across related taxpayers and enforcing debt repayments by liquidated companies.

*Source: _wp0516 - 1. The PIT Rate Structure Before and After Reform — excerpt provided.*

### Part I also eliminated a ceiling on interest accrued on overdue taxes. Some of its provisions,

### _wp0516 - Part I also eliminated a ceiling on interest accrued on overdue taxes. Some of its provisions,

### Context and legal changes
- Part I eliminated a ceiling on interest accrued on overdue taxes.
- Some provisions operated in the opposite direction: tax obligations were deemed discharged once the taxpayer had provided a payment order to a bank, allowing taxpayers to claim fulfillment of obligations without actually paying tax.
- The general thrust of the reform was to strengthen the powers of the tax administration; further measures to the same effect were taken in 2002.
- How these legal reforms changed actual practice is harder to judge.
- There were concurrent policy changes at the start of 2001 (notably reductions in social insurance taxes) that complicate isolating the effects of the PIT reform alone.

### Predictions of theory — model and revenue decomposition
- Personal income tax revenue is written as R = L w ... λ τ, where:
  - τ denotes the (tax-exclusive) tax rate,
  - λ denotes the ratio of declared taxable income to true taxable income (degree of compliance, with 1 = λ corresponding to fully truthful reporting),
  - w is the gross wage rate,
  - L is the level of employment.
- Denoting proportionate changes by hats, the revenue effect of any reform is approximated by:
  - ˆR ≈ ˆL + ˆw + ˆλ + ˆτ    (equation (1) as presented).
- For many initially higher-rate PIT payers, the reform corresponded to a substantial reduction in τ, amplified by the reduction in social insurance taxes.
- Key question: could the responses in ˆL, ˆw, and ˆλ have increased the tax base enough to account for the strong post-reform PIT revenue performance?

### Gross wage rates
- In the formal sector, a reduced tax wedge (PIT and social taxes) is expected to cause the gross wage w to fall, reinforcing the direct revenue effect of the tax cut.
- In the informal sector, gross wages might rise to keep take-home pay in line with the formal sector, but this has no direct tax revenue impact.
- The competitive labor-market benchmark (notations preserved) implies:
  - Denoting labor demand and supply elasticities by e_D and e_S, the gross wage falls unless compliance increases by a greater proportion than the tax rate falls.
  - Revenue is more likely to rise (for given ˆλ + ˆτ < 0) the greater is e_D (a necessary condition is that this exceed unity), the greater is e_S, the higher is the tax rate and the higher is initial compliance.

### Work effort
- Consideration focuses on changes in parameters of PIT and social taxes rather than gross wage changes.
- If the reform lowered the top marginal rate to the standard level (with exempt amount and starting marginal rate unchanged), the budget constraint rotates anti-clockwise around the kink point until it becomes a straight line.
- Substitution effect: reduction in marginal tax rate increases pre-tax income (increasing the tax base).
- Income effect: increased net income (including reduced taxation of intra-marginal income) tends to reduce work effort if leisure is normal, decreasing the tax base.
- Cases highlighted:
  - Individuals interior to the standard-rate segment: reform has no effect on work effort.
  - Individuals at the kink point: reform generates only a substitution effect and increases work effort.
  - Discrete choice/hiring-constraint case: individuals choosing between discrete gross income levels may increase labor supply even if previously paying the lower rate, because net income changes can make higher contracts relatively more attractive.
- Note: starting marginal rate—inclusive of social insurance—fell by 1.3 percentage points; effects at higher rates were more diverse.

### Compliance
- The reform might affect the tax base via changes in compliance (λ).
- Classic Allingham-Sandmo (1972) framework and Yitzhaki (1974) result:
  - A tax reduction can reduce compliance when fines on concealed income are proportional to the amount of tax evaded.
  - Intuition: lower tax rate increases after-tax income at any evasion level, increasing desired riskiness and proportion concealed under decreasing absolute risk aversion — so evasion increases.
- Other models (bargaining with corruptible inspectors) predict evasion could be independent of τ or increase as τ is reduced if fines are proportional to evaded tax.
- Opposing considerations:
  - Engel and Hines (1999): higher tax rates may increase evasion if past declarations can be re-opened on audit selection.
  - Legal avoidance: costs of avoidance may depend on amount concealed but not on the tax rate in the same way as evasion penalties; with marginal cost of avoidance increasing, higher tax rates may increase avoidance.
- Empirical evidence is mixed:
  - Andreoni, Erard and Feinstein (1999) review finds mixed empirical conclusions.
  - Schneider and Enste (2000) conclude high tax rates encourage concealment.
  - Friedman et al (2000) find the opposite.
- The reform’s effect on compliance is theoretically ambiguous and empirically unresolved.

### Recharacterizing income
- The reform might induce reclassification of income from corporate to personal (e.g., paying out earnings as salary, pensions, or interest deductible by firms but taxable to individuals).
- At the start of 2001:
  - Maximum corporate tax rate and the tax on dividends were increased,
  - Higher PIT and social taxes were cut.
- Mitigating factors reducing the incentive to recharacterize:
  - Adoption of imputation in 2001 reduced the effective tax rate on distributed corporate earnings from 40.5 percent (=1-(1-0.15)(1-0.3)) to 35 percent (the imputation credit being nonrefundable).
  - The marked PIT and social insurance reductions applied only to income in excess of pre-reform thresholds for the higher rate, whereas the reduction in the rate on distributed corporate earnings applied essentially to all profits.
- Conclusion: a priori unclear whether significant recharacterization occurred.

### The PIT reform in a wider context — revenue performance
- General government revenues (consolidated across all levels) increased substantially in 2001 relative to GDP across most revenue sources; PIT showed the greatest increase but indirect and trade taxes performed almost as well.
- The breadth of revenue increases suggests some common underlying cause beyond PIT reform, notably the increase in energy prices from late 2000.
- Specific observations:
  - Natural gas prices peaked in 2001, declined in 2002 and increased in 2003; oil prices increased and peaked somewhat earlier.
  - Oil-related revenues accounted at this time for about one-third of federal government revenues.
  - Kwon (2003) attributes about 80 percent of the recovery of revenues after the crisis to the strength of the oil and gas sector (which accounted for about 20 percent of GDP).
  - Revenues from resource taxes, and excises and taxes on trade track energy prices closely.
  - End-of-year employment increased by only 1.3 percent in 2001 (and year average employment by only [text truncated in source]).

### Key statistics from Table 3: General Government Revenues, 1994–2003 (In percent of GDP)
- Total Revenue: 1994: 34.6; 1995: 36.8; 1996: 35.8; 1997: 39.3; 1998: 34.4; 1999: 33.6; 2000: 36.9; 2001: 37.4; 2002: 37.6; 2003: 36.6
- o/w Personal income tax: 1994: 2.9; 1995: 2.6; 1996: 2.8; 1997: 3.2; 1998: 2.7; 1999: 2.4; 2000: 2.4; 2001: 2.9; 2002: 3.3; 2003: 3.4
- Profit tax: 1994: 8.0; 1995: 8.2; 1996: 4.9; 1997: 4.4; 1998: 3.7; 1999: 4.5; 2000: 5.5; 2001: 5.8; 2002: 4.3; 2003: 4.0
- VAT: 1994: 7.0; 1995: 6.9; 1996: 7.6; 1997: 7.3; 1998: 6.4; 1999: 5.9; 2000: 6.3; 2001: 7.2; 2002: 6.9; 2003: 6.6
- Excises: 1994: 1.2; 1995: 1.7; 1996: 2.8; 1997: 2.7; 1998: 2.7; 1999: 2.2; 2000: 2.3; 2001: 2.7; 2002: 2.4; 2003: 2.6
- Taxes on trade: 1994: 1.0; 1995: 1.7; 1996: 1.1; 1997: 1.2; 1998: 1.3; 1999: 1.8; 2000: 3.1; 2001: 3.7; 2002: 3.0; 2003: 3.4
- Payroll taxes 1/: 1994: 8.9; 1995: 8.1; 1996: 8.2; 1997: 9.7; 1998: 8.4; 1999: 7.7; 2000: 7.7; 2001: 7.3; 2002: 8.0; 2003: 7.8
- Resource taxes: 1994: 0.0; 1995: 0.9; 1996: 1.1; 1997: 1.5; 1998: 0.9; 1999: 0.9; 2000: 1.1; 2001: 1.4; 2002: 3.1; 2003: 3.0
- Other tax revenue: 1994: 5.1; 1995: 3.4; 1996: 3.6; 1997: 4.9; 1998: 3.6; 1999: 2.9; 2000: 2.8; 2001: 2.4; 2002: 2.8; 2003: 2.3
- Non-tax revenue: 1994: 0.0; 1995: 1.5; 1996: 1.2; 1997: 1.2; 1998: 1.5; 1999: 1.7; 2000: 1.8; 2001: 2.3; 2002: 2.5; 2003: 2.6
- Budgetary funds 2/: 1994: 0.5; 1995: 1.9; 1996: 2.4; 1997: 3.2; 1998: 3.1; 1999: 3.4; 2000: 4.0; 2001: 1.8; 2002: 1.3; 2003: 0.9
- Notes from table:
  - Source: Ministry of Finance, CBR, Goskomstat, and IMF staff estimates.
  - 1/ Payroll taxes include annual accumulation of a fully-funded state pension system.
  - Budgetary funds inclusive of on-budget and off-budget regional road funds.

*Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp0516.pdf*

### 0.3 percent) mostly due to an increase in employment by small businesses, which account for

### _wp0516 - 0.3 percent) mostly due to an increase in employment by small businesses, which account for

### Summary of PIT and other tax revenue developments
- Personal Income Tax (PIT) revenues increased strongly through 2002 and 2003 while other sources declined in 2003.
- Revenues from three sources fell relative to GDP between 2000 and 2001:
  - Budgetary fund revenue decline attributable mostly to reduction in turnover taxes.
  - "Other tax revenue" (including small business taxes, property taxes, and many other small items) declined with no single obvious explanation.
- Payroll taxes (levied on a similar base as PIT) fell by about 5 percent relative to GDP (increasing in nominal terms by about 16 percent) at the same time PIT revenues rose markedly.
- The combined rate of the social tax was markedly reduced, lowering rates throughout the entire income range.
- Profit tax revenue performed relatively poorly in 2001 and declined thereafter, even as PIT continued to increase. Multiple opposing influences (energy price effects, change in maximum profit tax rate from 30 to 35 percent, extension of deductions, investment timing) make interpretation difficult.
- Continued marked PIT growth in 2002 and 2003 despite significant cuts in tax rates on both profits and dividends (while the PIT structure remained unchanged) suggests limited importance of income-shifting from corporate to personal incomes.

### Wage developments and compliance
- The bulk of PIT revenue comes from wages and salaries.
- Official estimates imply the average effective PIT rate increased slightly from 11.2 to 11.8 percent (table reports 11.3 in 1999, 11.2 in 2000, 11.8 in 2001, 12.4 in 2002).
- The average effective UST rate dropped from 35.8 to 30.0 percent between 2000 and 2001.
- The average effective tax rate inclusive of employer paid taxes decreased by only 2.5 percentage points despite dramatic reductions in marginal tax rates.
- Official statistics include an estimate of hidden wages and degree of compliance:
  - Compliance (ratio of reported to total wages) implied 72.4 percent reported in 2000, rising to 74.0 percent in 2001—an improvement of a little over 2 percent.

### Decomposition of PIT revenue growth (as implied by official data and approximation)
- Approximate decomposition of the increase in PIT revenue:
  - About 13 percent of the increase in PIT revenue reflects the increase in the effective rate of the PIT itself.
  - About 10 percent is due to the lower rate of social taxation (through enlarging the PIT base for any level of gross income).
  - About 5 percent reflects improved compliance.
  - Over 70 percent is associated with an increase in gross incomes.
- The modest increase in UST revenues reflects the dominance of gross income increases over the large cut in the average effective UST rate.

### Key statistics (from Table 4: Billions of rubles and percentage changes)
- Billions of rubles (Gross wage income; o/w Reported; Hidden; PIT revenue; UST revenue; Reported wage income base; Net wage income):
  - 1999: Gross wage income 1934; Reported 1408; Hidden 526; PIT revenue 117; UST revenue 373; Reported wage income base 1035; Net wage income 918.
  - 2000: Gross wage income 2937; Reported 2126; Hidden 811; PIT revenue 175; UST revenue 561; Reported wage income base 1565; Net wage income 1390.
  - 2001: Gross wage income 3819; Reported 2826; Hidden 993; PIT revenue 256; UST revenue 652; Reported wage income base 2175; Net wage income 1919.
  - 2002: Gross wage income 4995; Reported 3749; Hidden 1246; PIT revenue 358; UST revenue 865; Reported wage income base 2883; Net wage income 2525.
- Average effective rates and compliance:
  - Average effective PIT rate: 11.3 (1999); 11.2 (2000); 11.8 (2001); 12.4 (2002).
  - Average effective UST rate: 36.0 (1999); 35.8 (2000); 30.0 (2001); 30.0 (2002).
  - Average effective tax rate (inclusive of social taxes): 34.8 (1999); 34.6 (2000); 32.1 (2001); 32.6 (2002).
  - Compliance: 72.8 (1999); 72.4 (2000); 74.0 (2001); 75.1 (2002).
- Percentage changes (selected):
  - Gross wage income percentage change: 52.9 (1999); 51.8 (2000); 30.0 (2001); 30.8 (2002).
  - Reported wage income percentage change: 41.7 (1999); 50.9 (2000); 32.9 (2001); 32.6 (2002).
  - Hidden wage income percentage change: 94.1 (1999); 54.2 (2000); 22.4 (2001); 25.5 (2002).
  - PIT revenue percentage change: 64.5 (1999); 49.3 (2000); 46.3 (2001); 40.1 (2002).
  - UST revenue percentage change: 67.9 (1999); 50.4 (2000); 16.2 (2001); 32.8 (2002).
  - Average effective PIT rate percentage change: 22.6 (1999); -1.2 (2000); 5.3 (2001); 5.7 (2002).
  - Average effective UST rate percentage change: 25.1 (1999); -0.5 (2000); -16.4 (2001); 0.2 (2002).
  - Average effective tax rate percentage change: 17.9 (1999); -0.5 (2000); -7.2 (2001); 1.7 (2002).
  - Compliance percentage change: -7.3 (1999); -0.6 (2000); 2.2 (2001); 1.4 (2002).

### Observations on real wages and growth
- After-tax real wage income grew by 18.5 percent in 2001.
- Gross real wages grew by 11.6 percent in 2001, reflecting reduction in tax rates.
- Both gross and net wages outpaced GDP growth, which amounted to 5.1 percent, and average labor productivity growth (value reported but truncated in source).

*Source: Goskomstat and authors' estimates.*

### 2.3 percent in 2001, implying an increase in the labor share in this year.

### _wp0516 - 2.3 percent in 2001, implying an increase in the labor share in this year.

### Macro evidence on labor share and PIT revenues
- Labor share dynamics:
  - Hidden wages remained constant at about 10 percent of GDP.
  - Reported wages drove fluctuations in the labor share.
  - An increase in the labor share in 2001 fits a pattern previously observed for Russia, with real wages tending to overshoot real GDP growth.
- PIT revenue dynamics:
  - Strength of PIT revenues between 2000 and 2001 owes overwhelmingly to a marked increase in gross incomes, with any gain in compliance being very modest (aggregate data suggest).
- Figures referenced (descriptive):
  - Figure 2: average labor productivity, output and wages (1995–2002) — average labor productivity closely follows real GDP growth.
  - Figure 3: share of wages and net profits in GDP (1994–2002) — reported gross wages, hidden wages (~10 percent of GDP), net profits.

### Micro evidence: data, measurement limits, and methodology
- Data source and scope:
  - Russian Longitudinal Monitoring Survey (RLMS), Carolina Population Center, University of North Carolina.
  - RLMS provides information on around 3,500 adults for every year (except 1997 and 1999) between 1994 and 2002; analysis uses only data for 2000 and 2001.
- Key data limitations:
  - No direct data on tax payments or pre-tax incomes; tax payments and gross incomes inferred from reported after-tax incomes.
  - Survey does not provide enough information to calculate all tax deductions.
  - Best- and worst-off individuals are under-represented.
- Main income variables used:
  - “What was your average monthly wage after taxes over the last 12 months from the primary employer regardless of whether it was paid on time or not” (pjpayt).
  - “How much money in the last 30 days did you receive from your primary job after taxes?” (wagelm) — used as robustness check.
- Sample construction and cleaning:
  - Retained individuals between 20 and 60 years of age throughout 2000 and 2001; dropped those reporting >84 hours/week, not reporting hours, not reporting income from primary employment, and/or owning their own business.
  - Initial cleaned sample left 3,722 individuals; regressions use individuals present in both years with required variables.
- Representativeness check:
  - RLMS sample matches official aggregates closely.
  - Growth in PIT payments in the sample over 2000/2001: 45.2 percent (sample estimate) vs. 46.3 percent (published data).
  - Table 5 entries (selected): Average wage/salary 2000 = 2174 (estimated) vs. 2223 (published); 2001 = 3310 vs. 3282; 2002 = 4332 vs. 4426.

### Empirical strategy and hypotheses
- Difference-in-differences approach:
  - Compare individuals/households strongly affected by the reform (treatment) with those little affected (control).
  - Treatment: those initially paying a higher PIT marginal rate (20–30 percent); control: low-rate payers (0–12 percent).
- Regression specifications:
  - Levels: y_it = β0 + β1 P_t + β2 T_i + β3 (T_i × P_t) + u_it; β3 is the difference-in-differences estimator.
  - Growth rates: Δy_i = γ0 + γ1 T_i + u_i.
- Null hypotheses:
  - On levels: R_L^T − R_L^C ≤ 0 (no differential increase in tax paid in treatment vs. control).
  - On growth rates: analogous R_G.
  - Additional hypotheses on reported income, true gross pretax income, compliance, hours worked, and wage per hour.

### Panel data results (assuming full compliance)
- Sample sizes and group splits:
  - Table 6: taxpayers by marginal tax rate in 2000: No. taxpayers = 100 (0%), 2130 (12%), 173 (20%), 11 (30%).
- PIT payments by initial marginal rate (Table 6 highlights):
  - Pre-reform mean PIT (real 1995 rubles): 0 -> 0; 12% -> 373.1; 20% -> 2355.0; 30% -> 7621.1.
  - Change in PIT (2000–2001): 0 -> 80.6; 12% -> 165.6; 20% -> -104.3; 30% -> -3961.0.
  - Growth rates (median): 12% -> 34.0 percent; 20% -> -6.3 percent; 30% -> -35.1 percent.
  - Interpretation: tax payments fell for all groups except those initially paying low tax rates; larger tax cuts associated with larger falls in tax payments.
- Difference-in-differences (Table 7, treatment 20–30% vs. control 0–12%):
  - Levels: Pre-reform PIT — control 356.4, treatment 2669.9; Change — control 161.8, treatment -334.9; Difference-in-differences = -496.7*** (standard error 160.7).
  - Growth rates: Mean growth rate, PIT = control 104.3, treatment -3.2; Difference-in-differences = -107.4*** (standard error 12.5); Median = -41.9*** (standard error 7.0).
  - Statistical significance: *** denotes significance at conventional levels.
- Extended treatment group (≥75 percent of threshold) — Table 8:
  - Difference-in-differences in levels = -232.3** (std. error 104.2).
  - Mean growth rate diff-in-diff = -102.0*** (std. error 13.2); median = -32.9*** (std. error 5.1).
- PIT, total tax (PIT + social insurance) and gross income (Table 9):
  - Levels: Change in PIT diff-in-diff = -496.7*** (160.7) for 20–30% treatment; Change in total tax = -1860.2*** (452); Change in gross income = -158.8 (100.9) — not significant in levels.
  - Growth rates (means): PIT = -107.4*** (125.3); total tax = -51.1*** (3.8); gross income = -41.9*** (3.9).
  - Median growth rates: PIT = -41.9*** (7.0); total tax = -28.3*** (4.7); gross income = -26.0*** (5.7).
- Hours worked and wage rate (Table 10):
  - Levels: Change in hours worked diff-in-diff = -2.155** (1.174) for 20–30% treatment; Change in wage rate = -0.491 (0.564) — not significant in levels.
  - Growth rates (means): hours worked = -5.6*** (1.5); wage rate = -37.7*** (5.5).
  - Median growth rate, wage rate = -22.4*** (5.4).
  - Interpretation: relative reductions in both wage rate and labor supply contribute to relative decline in gross incomes; labor adjustments more through wages than employment.

### Allowing for tax evasion and underreporting (household-level using consumption)
- Rationale:
  - Use household non-durable consumption as proxy for true net income to infer gross income and compliance.
  - Treatment households: those whose estimated gross employment income per working adult ≥ pre-reform higher-rate threshold.
- Household-level results (reported income basis, Table 11):
  - Number of families: control 1016, treatment 122.
  - Change in total tax paid: control 23.7; treatment -20.7; Difference-in-differences = -44.4* (24.0).
  - Growth rates: mean growth rate diff-in-diff = -147.8*** (23.9); median = -41.2*** (7.7).
- Consumption-based treatment definition (Table 12):
  - Number of families: control 946, treatment 192 (treatment group almost doubles in size versus reported-income definition).
  - Levels: change in total tax paid diff-in-diff = -16.3 (19.0) — not significant.
  - Growth rates: mean growth rate diff-in-diff = -67.0** (29.8); median = -18.3*** (3.7).
  - Interpretation: controlling for evasion/misreporting yields mixed results — reject null in growth rates but not in levels.
- PIT, total tax, declared income (Table 13, consumption criterion):
  - Levels: Change in PIT diff-in-diff = -16.3 (19.0); Change in total tax = -63.3 (61.4); Change in declared income = -20.9 (167.9) — none significant in levels.
  - Growth rates: Mean growth rate, PIT = -67.0** (29.8); median growth rate, PIT = -18.3*** (3.7); median growth rate, total tax = -9.4** (4.0); median declared income = -8.2* (4.6).
- Compliance estimates (Table 14, consumption-based household analysis):
  - Pre-reform declared share of income: control 73.8 percent; treatment 52.0 percent.
  - Change in declared share: control -0.9; treatment +16.4; Difference-in-differences = 17.3*** (4.4).
  - Growth rates: mean growth rate — control 22.0, treatment 123.3; Difference-in-differences = 101.3*** (24.8). Median growth rate diff-in-diff = 34.3*** (5.6).
  - Interpretation: estimated compliance in treatment group increases by about 16 percentage points (from 52 percent to 68 percent) using levels; weighted average growth rate of compliance in treatment ~ one-third; unweighted average and median growth rates are 101 percent and 34 percent respectively, indicating skewness.
- Gross income, wage rate, and hours (Tables 15–16, consumption criterion):
  - Gross income levels: pre-reform control 975.0; treatment 3124.8. Change: control 320.6; treatment -556.7; Difference-in-differences = -877.3*** (214.7).
  - Growth rates: mean diff-in-diff for gross income = -54.3*** (5.9); median = -41.1*** (5.0).
  - Wage rate diff-in-diff (levels) = -12.8*** (3.1).
  - Growth rates: mean wage rate diff-in-diff = -56.8*** (7.6); median = -38.8*** (7.0).
  - Hours worked: no differential change (levels change in hours worked = 1.66 (3.99)).
  - Interpretation: relative gross incomes fell in treatment households driven primarily by lower gross wage rates rather than reductions in hours worked.

### Summary of empirical findings
- Robust finding across specifications: the 2001 reform did not cause the observed PIT revenue boom.
  - Stronger growth in PIT payments systematically came from those little affected by the reform (control group).
  - Difference-in-differences estimates show negative and often highly significant effects for treatment groups on PIT and total tax in both levels and growth-rate specifications.
- Behavioral effects identified:
  - Significant increase in compliance among affected taxpayers: estimated increase of around 16 percentage points in declared share of income in treatment households (levels) and large increases in growth-rate measures.
  - Significant reduction in relative gross wage rates for treatment groups; no measurable differential effect on labor supply (hours worked).
- Quantitative highlights:
  - Sample-based PIT revenue growth 2000/2001 = 45.2 percent (RLMS estimate) vs. published 46.3 percent.
  - Pre-reform declared income shares: control 73.8 percent; treatment 52.0 percent.
  - Estimated compliance change in treatment group = +16.4 (levels) with diff-in-diff = 17.3*** (4.4).
  - Difference-in-differences for change in PIT (treatment 20–30% vs control 0–12%): -496.7*** (160.7) in levels.

### Conclusions and policy lessons
- Attribution of revenue boom:
  - It is hard to attribute the very strong performance of PIT revenues after the reform to the flat-rate reform itself; evidence points to other factors (e.g., marked increase in real wage rates unrelated to the reform).
- Key behavioral effects of the reform:
  - Strong improvement in compliance around the reform period, albeit possibly also reflecting contemporaneous strengthening of tax administration.
  - Reduction in gross wage rates for those most affected; negligible differential labor supply response.
- Policy implications for other countries:
  - Tax-cutting reforms of this kind should not be expected to pay for themselves by greater work effort and improved compliance alone.
  - In settings with initially high noncompliance, parametric reforms may be associated with notable compliance improvements; however, such improvements may be conditional on concurrent enforcement measures.
  - Structural features of labor markets (elasticities, rigidities) matter for how wage rates and employment adjust to tax changes and thus mediate revenue effects.
- Cautionary note:
  - Given data limitations (no direct tax payment observations, potential under-reporting), estimates—particularly of levels—should be treated with caution; nonetheless, the qualitative patterns across multiple empirical strategies are consistent.

*Source: IMF working paper content unit _wp0516 - 2.3 percent in 2001, implying an increase in the labor share in this year.*

### REFERENCES

### _wp0516 - REFERENCES

### Academic articles and theoretical analyses
- Allingham, Michael, and Agnar Sandmo, 1972, “Income Tax Evasion: A Theoretical Analysis,” Journal of Public Economics, Vol. 1, pp. 323–38.
- Andreoni, James, Brian Erard, and Jonathan Feinstein, 1999, “Tax Compliance,” Journal of Economic Literature, Vol. 36, pp. 818–60.
- Blundell, R., A. Duncan, and C. Meghir, 1998, “Estimating Labor Supply Responses Using Tax Reforms,” Econometrica, Vol., 66, pp. 827–62.
- Eissa, N., 1995, “Taxation and Labor Supply of Married Women: The Tax Reform Act of 1986 As a Natural Experiment,” NBER Working Paper No. 5023 (Cambridge, Massachusetts: NBER).
- Engel, Eduardo E., and James R. Hines, Jr., 1999, “Understanding Tax Evasion Dynamics,” NBER Working Paper 6903 (Cambridge, Massachusetts: NBER).
- Feldstein, M., 1995, “The Effect of Marginal Tax Rates on Taxable Income: A Panel Study of the 1986 Tax Reform Act,” Journal of Political Economy, Vol. 103, pp. 551–72.
- Friedman, Eric, Simon Johnson, Daniel Kaufman, and Pablo Zoido-Lobaton, 2000, “Dodging the Grabbing Hand: The Determinants of Unofficial Activity in 69 Countries,” Journal of Public Economics, Vol. 76, pp.459–93.
- Gordon, Roger, and Jeffrey Mackie-Mason, 1994, “Tax Distortions to the Choice of Organizational Form,” Journal of Public Economics, Vol. 55, pp. 297–306.
- Gordon, Roger, and Joel Slemrod, 1998, “Are ‘Real’ Responses to Taxes Simply Income Shifting Between Corporate and Personal Tax Bases?” NBER Working Paper 6576 (Cambridge, Massachusetts: NBER).
- Hall, Robert E., and Alvin Rabushka, 1995, The Flat Tax, 2nd edition (Stanford: Hoover Institution Press).
- Johnson, S., D. Kaufmann, and P. Zoido-Lobaton, 1998, “Regulatory Discretion and the Unofficial Economy,” American Economic Review, Vol. 88 (2), pp. 387–432.
- Konings, Jozef, and Hartmut Lehmann, 2002, “Marshall and Labor Demand in Russia: Going Back to Basics,” Journal of Comparative Economics, Vol. 30, pp.134–59.
- Lacko, M., 1996, “Hidden Economy—An Unknown Quantity? Comparative Analysis of Hidden Economies in Transition Countries in 1989–1995, Working Paper No. 9905 (University of Linz: Department of Economics).
- Schneider, Friedrich, and Dominik Enste, 2000, “Shadow Economies: Sizes, Causes, and Consequences,” Journal of Economic Literature, Vol. 38, pp. 77–114.
- Slemrod, Joel, 2001, “A General Model of the Behavioral Response to Taxation,” International Tax and Public Finance, Vol. 8, pp.119–28.
- Slemrod, Joel, and Schlomo Yitzhaki, 2002, “Tax Avoidance, Evasion, and Administration,” in Handbook of Public Economics, ed. by Alan Auerbach and Martin Feldstein (North-Holland: Amsterdam) pp. 1423–70.
- Yitzhaki, Schlomo, 1974, “A Note on ‘Income Tax Evasion: A Theoretical Analysis’,” Journal of Public Economics, Vol. 3, pp. 201–02.

### Empirical studies, shadow economy, and measurement methods
- Eilat, Y., and C. Zinnes, 2000, “The Evolution of the Shadow Economy in Transition Countries: Consequences for Economic Growth and Donor Assistance,” CAER II Discussion Paper No. 83 (Cambridge, Massachusetts: Harvard Institute for International Development).
- Feige, E.L., and I. Urban, 2003, “Estimating the Size and Growth of Unrecorded Economic Activity in Transition Countries: A Reevaluation of Electric Consumption Method Estimates and Their Implications,” William Davidson Institute Working Paper 636.
- Engel, Eduardo E., and James R. Hines, Jr., 1999, “Understanding Tax Evasion Dynamics,” NBER Working Paper 6903 (Cambridge, Massachusetts: NBER).
- Friedman, Eric, Simon Johnson, Daniel Kaufman, and Pablo Zoido-Lobaton, 2000, “Dodging the Grabbing Hand: The Determinants of Unofficial Activity in 69 Countries,” Journal of Public Economics, Vol. 76, pp.459–93.
- Sinelnikov-Mourylev, S. S. Batkibekov, P. Kadochnikov, and D. Nekipelov, 2003, “Evaluation of the Results of Personal Income Tax Reform in Russian Federation” (Institute for Economies in Transition Working Paper), at http://www.iet.ru/papers/52/index.htm
- RLMS, 2004, “Russian Longitudinal Monitoring Survey” (Carolina Population Center: University of North Carolina).

### Russia-specific tax reform and fiscal studies
- Chua, Dale, 2003, “Tax Reform in Russia,” in Russia Rebounds, ed. by David Owen and David Robinson (Washington: International Monetary Fund).
- Glavatskaya, N., and K. Seryanova, eds., 2003, “Tax Reform in Russia: Problems and Solutions,” Institute for Economies in Transition Working Paper No. 67. (In Russian; available at http://www.iet.ru/papers/67/.)
- International Monetary Fund, 2000, “Russian Federation: Selected Issues and Statistical Appendix” (Country Report 02/75), April 2002.
- Kwon, Gohoon, 2003, “Post-Crisis Fiscal Revenue Developments in Russia: From an Oil Perspective,” Public Finance and Management, Vol. 3, pp. 503–30.
- Rabushka, Alvin, 2003, “The Flat Tax in Russia and the New Europe,” Brief No. 2003 (Washington: National Center for Policy Analysis).

*Source: _wp0516 - REFERENCES*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp0516.pdf_