## 1. Classification of Tax Revenues (wpiea2020094-print-pdf)

## Source details

**Canonical URL:** [1. Classification of Tax Revenues (wpiea2020094-print-pdf)](https://www.imf.org/-/media/files/publications/wp/2020/english/wpiea2020094-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2020/english/wpiea2020094-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2020/english/wpiea2020094-print-pdf.pdf.json)

---

### Purpose and context
- Strengthening resilience to fiscal risks from government revenue volatility is critical for ensuring sustainable delivery of public services.
- Intuition: tax revenue diversification (reliance on more diversified sources for levying revenue) can reduce overall tax revenue volatility because different taxes respond differently to business cycle fluctuations.
- The coronavirus pandemic underscores the vulnerability of tax systems that rely on a concentrated portfolio of tax revenue streams.

### Dataset and RDI construction
- Panel: 127 countries over 2000-15.
- Focus: tax revenue (non-tax revenues left aside).
- RDI constructed from six major categories of taxes (GFSM 2014):
  - corporate income tax (CIT)
  - personal income tax (PIT)
  - property tax
  - tax on goods and services
  - tax on international trade
  - other taxes
- Index choice: RDI built on the Theil index (Theil, 1972) for stability and robustness to outliers; Theil-based RDI range: 0 (perfect diversification) to 1.8 (reliance on one type of tax only).
- Formula (Theil index): T = (1/n) ∑_{i=1}^n (Tax_i / μ) × log(Tax_i / μ), with μ = (1/n) ∑ Tax_i.
- Robustness: HHI-based RDI also computed; correlation coefficient with Theil-based RDI = 0.98.

### Data and tax revenue composition (exact figures)
- Sample coverage: 127 countries (47 AEs, 31 EMEs, 49 LIDCs). Regional counts: 25 SSA, 2 NA, 7 SA, 19 LAC, 14 MENA (of which 5 GCC), 21 EAP, 39 ECA.
- Taxes expressed in percent of GDP.
- Full-sample average tax shares (percent of GDP):
  - Total taxes: 19.0
  - Direct taxes (39% of total): 2.9
  - Corporate taxes: 2.9
  - Personal taxes: 4.3
  - Property taxes: 0.5
  - Indirect taxes (61% of total): 8.5
  - Tax on goods and services: 8.5 (44% of total)
  - Taxes on international trade: 2.5 (15% of total)
  - Other taxes: 0.3 (2% of total)
- Cross-group exact figures:
  - AEs (High income: OECD) total tax revenue: 25.8 percent of GDP.
  - Non-resource-rich countries: 20.2 percent of GDP.
  - Resource-rich countries: 14.9 percent of GDP.
  - Fragile states: 11.6 percent of GDP.
  - Small states: 20.4 percent of GDP; Non-small states: 18.8 percent of GDP.

### Stylized facts and RDI summary statistics
- Full-sample average RDI: 0.51.
- Extremes:
  - Lowest RDI: Japan 0.05 (2000-15 average; table shows 0.053, 0.055, 0.057 across periods).
  - Highest RDI: Kingdom of Bahrain 1.39.
- Country ranking excerpts (2010-2015 exact values):
  - Top performers: Japan 0.06, France 0.15, United Kingdom 0.17.
  - Bottom performers: Bolivia 1.34, Kuwait 1.34, Anguilla 1.32.
- Regional/group patterns:
  - NA and EU: lowest RDI (below full-sample average).
  - GCC, LAC, SSA: highest RDI (least diversified).
  - Income-level pattern: OECD countries most diversified; high-income non-OECD and low-income countries record highest RDI.
- Correlation between log real GDP per capita and RDI: Correlation = -0.43.
- Dynamics: some AEs and EMEs diversified between 2000-04 and 2010-15 (e.g., Austria, Denmark, France, Germany, Japan, Morocco, South Africa); some concentrated further (e.g., Kuwait, Bahrain, Sri Lanka); some LIDCs diversified (e.g., Algeria, Côte d'Ivoire, Ghana, Kenya, Mauritius, Uganda).

### Correlations with macroeconomic indicators (2000-15, exact coefficients)
- RDI and total tax revenue: Correlation = -0.39.
- RDI and tax revenue volatility: Correlation = 0.29.
- RDI and GDP growth volatility: Correlation = 0.13.
- RDI and spending volatility: Correlation = 0.28.
- RDI and income inequality (GINI): Correlation = 0.37.
- RDI and exports concentration: Correlation = 0.46.
- RDI and tax compliance gap: Correlation = 0.54.
- RDI and tax collection efficiency: Correlation = -0.34.
- Interpretation: higher tax concentration tends to coincide with lower revenue, greater volatility (tax, growth, spending), higher inequality, larger compliance gaps, and weaker collection efficiency; RDI positively correlated with export concentration.

### Drivers of tax revenue diversification — econometric findings (system-GMM)
- Persistence: lagged RDI contributes up to 60-74 percent of current RDI.
- Non-linear relationship with development:
  - Log real GDP_pc: negative coefficient; Log real GDP_pc_squared: positive coefficient. Average per capita GDP threshold reported as $ 4222.
- Significant drivers (signs and interpretation):
  - Export concentration index: positive and significant.
  - Natural resource rents: positive and statistically significant (natural resources curse).
  - Net ODA per capita: positive and statistically significant (aid dependency linked to less diversification incentives).
  - Informal sector share: positive.
  - Financial development: negative and significant (deeper financial systems associated with greater diversification).
  - Trade openness: mostly negative (insignificant in many specs; one spec shows significance).
  - IMF program dummy: negative and statistically significant in one specification (IMF programs may help diversify tax sources).
- Political and institutional drivers:
  - Democracy (Polity_2): negative and significant — deeper democracy fosters diversification.
  - Political polarization and government fractionalization: negative and significant — more polarized/fractionalized systems linked to greater diversification.
  - Political stability: negative and significant — greater stability conducive to diversification.
  - Largest government party orientation: negative and significant — more socialist-oriented governments more prone to diversify (Right=1, Center=2, Left=3).
  - Institutional quality indicators (bureaucracy, rule of law, government effectiveness, voice and accountability, control of corruption): negative and significant — stronger institutions support greater diversification.
- Heterogeneities:
  - LIDCs and EMEs have more room to diversify vs AEs.
  - Regions with less diversified tax sources include South Asia, LAC, and MENA.
  - Resource-rich dummy coefficient: 0.095** (positive, less diversified).

### Annex 5 — Alternative estimates (selected exact coefficients and diagnostics)
- Main coefficient estimates across four specifications (t-1 values, standard errors in parentheses):
  - RDI (t-1): (1) 0.611*** (0.011); (2) 0.478*** (0.024); (3) 0.459*** (0.022); (4) 0.453*** (0.023)
  - Log real GDP_pc (t-1): (1) -0.6170*** (0.080); (2) -0.8073*** (0.123); (3) -0.7172*** (0.128); (4) -0.7804*** (0.135)
  - Log real GDP_pc_squared (t-1): (1) 0.038*** (0.005); (2) 0.051*** (0.008); (3) 0.0442*** (0.008); (4) 0.0487*** (0.009)
  - Financial development (t-1): (1) -0.3958*** (0.108); (2) -0.6439*** (0.133); (3) -0.5385*** (0.138); (4) -0.6052*** (0.139)
  - Trade openness (t-1): (1) -0.0318* (0.019); (2) -0.0212 (0.026); (3) -0.0257 (0.026); (4) -0.0116 (0.028)
  - Export concentration index (t-1): (1) 0.0252*** (0.007); (2) 0.0294*** (0.009); (3) 0.0339*** (0.009); (4) 0.0302*** (0.009)
  - VA Services / VA Agri.: (1) -0.0001***; (2) -0.0001***; (3) 0.000; (4) 0.000
  - VA Services / VA Manuf.: (3) 0.0023** (0.001); (4) 0.0020** (0.001)
  - Constant: (1) 2.631*** (0.346); (2) 3.486*** (0.518); (3) 3.137*** (0.540); (4) 3.361*** (0.564)
- Sample and diagnostics:
  - Nb. of observations: (1) 1141; (2) 1089; (3) 1074; (4) 1074
  - Countries: (1) 97; (2) 96; (3) 95; (4) 95
  - AR(1): (1) 0.08; (2) 0.04; (3) 0.05; (4) 0.06
  - AR(2) p-value: (1) 0.3; (2) 0.14; (3) 0.25; (4) 0.25
  - Hansen OID (p-value): (1) 0.07; (2) 0.11; (3) 0.2; (4) 0.22
  - Nb. of instruments: (1) 70; (2) 71; (3) 71; (4) 72
  - Year fixed effects: Yes (all specs)
  - Region fixed effects: Yes (all specs)

### Impacts of tax revenue diversification (Section V — exact reported findings)
- Empirical strategy: GMM estimations of RDI’s effect on (i) tax-to-GDP ratio and (ii) tax revenue volatility (standard deviation over 3-year rolling window); specifications include lagged dependent variable and controls.
- Impact on tax revenue level:
  - Higher RDI score (higher concentration) associated with lower tax revenue.
  - Magnitude (text): "A 10 percent increase in the RDI score can yield additional tax revenue of up to 0.2-0.4 percentage points of GDP."
  - Example RDI (-1) coefficients in Table 7: -0.036***, -0.025***, -0.028*** (negative and statistically significant).
  - Lagged tax revenue persistence: Tax revenue (-1) coefficients range 0.866*** to 0.966*** in reported specifications.
- Impact on tax revenue volatility:
  - Higher tax revenue concentration associated with greater revenue volatility; improving diversification reduces volatility.
  - Magnitude (text): "A one-point improvement in tax revenue diversification is associated with a reduction in tax revenue volatility of up to 0.5-2.8 points."
  - Example RDI (-1) coefficients in volatility regressions (Table 8): 2.422***, 2.801***, 1.644*** (positive and statistically significant).
  - Revenue volatility persistence: Revenue volatility (-1) coefficients e.g., 0.694***, 0.657***.
- Robustness: Standard diagnostic tests for instrument validity generally passed (AR(2), Hansen), with some exceptions noted in footnotes for specific columns.

### Key policy-relevant findings and implications (exact phrasing preserved where provided)
- Tax revenue diversification is associated with:
  - Higher tax-to-GDP ratios — "a 10 percent increase in RDI score can translate into up to 0.2-0.4 percentage points of GDP additional tax revenue."
  - Lower tax revenue volatility — "a one-point improvement in RDI linked to a reduction in tax revenue volatility up to 0.5-2.8 points."
- Determinants amenable to policy:
  - Strengthening institutions (quality of bureaucracy, rule of law, government effectiveness, control of corruption) can foster diversification.
  - Formalization of the economy and financial development support broader tax bases.
  - Reducing aid dependency and managing natural resource reliance can increase incentives to diversify taxation.
  - Political arrangements (democracy depth, polarization, stability) matter for taxation portfolio choices.
- Practical guidance inferred from results:
  - Policies that improve tax administration capacity, deepen financial systems, and broaden the tax base (formalization) are likely to increase diversification and revenue resilience.
  - For resource-rich countries, reforms to reduce over-reliance on resource rents (e.g., diversify economic activity, broaden non-resource tax base) are important for reducing concentration and volatility.
  - IMF-supported programs and policy conditionality may help diversify tax structures where implemented.

### Conclusions and avenues for future research
- Contributions:
  - New national-level RDI for 128 countries (2000-15) based on Theil index; first paper to create such an index at the national level.
  - Systematic differences demonstrated across income groups, regions, fragile/resource-rich status, and over time.
- Main takeaways:
  - AEs exhibit more diversified tax structures than EMEs and LIDCs; resource-rich and fragile states have the most concentrated tax sources.
  - Macroeconomic structure, political and institutional conditions significantly shape tax revenue diversification.
  - Tax revenue diversification materially improves revenue collection and reduces revenue volatility, strengthening fiscal resilience — a critical consideration in contexts such as the coronavirus pandemic.
- Suggested future research directions:
  - Investigate causal links and transmission channels between per capita GDP and tax revenue diversification.
  - Explore the influence of tax revenue diversification on income inequality.
  - Study how diversification affects policymakers’ leeway for implementing countercyclical fiscal policies.

*Source: wpiea2020094-print-pdf (sections 1, V, VI, Annexes 1–5 as contained in the supplied content).*

### 1. Classification of Tax Revenues.......................................................................................

### 1. Classification of Tax Revenues

### H3: Purpose and context
- Strengthening resilience to fiscal risks from government revenue volatility is critical for ensuring sustainable delivery of public services.
- A long-held intuitive view: tax revenue diversification (reliance on more diversified sources for levying revenue) can reduce overall tax revenue volatility because different taxes respond differently to business cycle fluctuations.
- The coronavirus pandemic underscores the vulnerability of tax systems that rely on a concentrated portfolio of tax revenue streams.

### H3: Dataset and RDI construction
- Panel: 127 countries over 2000-15.
- Focus: tax revenue (non-tax revenues left aside).
- Classification: RDI constructed from six major categories of taxes as reported in GFSM 2014:
  - corporate income tax (CIT)
  - personal income tax (PIT)
  - property tax
  - tax on goods and services
  - tax on international trade
  - other taxes
- Index choice: RDI built on the Theil index (as opposed to Herfindahl-Hirschman Index), chosen for stability and robustness to outliers.

### H3: Key stylized facts
- Advanced Economies (AEs) relied on a more diversified structure of tax sources than Emerging Market Economies (EMEs) and Low-Income Developing Countries (LIDCs), "by as high as the double in terms of RDI" over 2000-15.
- Resource-rich countries and fragile states exhibit the most concentrated structure of tax sources, reflecting over-dependence on commodity revenues and weak tax administration capacity.
- Regional patterns: North American and EU countries exhibit the most diversified taxation sources; GCC, South Asian, Latin American, and Sub-Saharan African countries present the least diversified revenue streams.

### H3: Econometric findings and drivers of RDI
- Persistence: up to 60-74 percent of the current level of RDI is predicted by its lagged value.
- RDI reflects not only economic diversification but also macroeconomic, political, and institutional factors.
- Relationship with development: a non-monotone relationship between RDI and economic development — tax source portfolios become more diversified as institutions and tax administration capacity improve, until a tipping point where richer countries find it more difficult to diversify further.
- Effects:
  - Tax revenue diversification mitigates government revenue volatility.
  - Tax revenue diversification improves tax revenue collection.

### H3: Implications highlighted in the text
- Diversifying tax structures can serve as a direct tool to mitigate revenue volatility and strengthen fiscal resilience.
- Policy emphasis on improving institutions and tax administration capacity to achieve greater diversification, recognizing country- and context-specific trade-offs among efficiency, fairness, administrative capacity, and political economy.

*Source: Excerpt from "1. Classification of Tax Revenues" in the provided IMF content unit.*

### section V assesses its effects on both volatility and level of government  revenue. Section  VI presents

### wpiea2020094-print-pdf - section V assesses its effects on both volatility and level of government  revenue. Section  VI presents

### Data and tax revenue composition
- Sample coverage: 127 countries over the period 2000-15; composed of 47 advanced economies (AEs), 31 Emerging Market Economies (EMEs), and 49 low-income developing countries (LIDCs). Regional counts: 25 SSA, 2 NA, 7 SA, 19 LAC, 14 MENA (of which 5 GCC), 21 EAP, 39 ECA.
- Data source: IMF’s Government Financial Statistics (GFS) (GFSM 2014 definitions and classifications).
- Tax measurement and grouping:
  - Taxes expressed in percent of GDP.
  - Two blocks: (i) direct taxes (income, profits, property, capital gains) and (ii) indirect taxes (goods and services, international trade and transactions, other taxes).
- Full-sample average tax shares (percent of GDP):
  - Total taxes: 19.0
  - Direct taxes (39% of total): 2.9 (corporate tax 2.9? — see table breakdown)
  - Corporate taxes: 2.9
  - Personal taxes: 4.3
  - Property taxes: 0.5
  - Indirect taxes (61% of total): 8.5
  - Tax on goods and services: 8.5 (44% of total)
  - Taxes on international trade: 2.5 (15% of total)
  - Other taxes: 0.3 (2% of total)
- Cross-group observations (selected exact figures from Table 2):
  - AEs (High income: OECD) total tax revenue: 25.8 percent of GDP.
  - Non-resource-rich countries: 20.2 percent of GDP.
  - Resource-rich countries: 14.9 percent of GDP.
  - Fragile states: 11.6 percent of GDP.
  - Small states: 20.4 percent of GDP; Non-small states: 18.8 percent of GDP.
- Data handling: Missing GFS observations filled using IMF’s Worldwide Revenue Database with consistency checks; gaps not filled if they cause substantial discrepancies between totals and subcomponents.

### Construction of the Revenue Diversification Index (RDI)
- Methodological choice: Theil index (Theil, 1972) preferred to HHI for stability and robustness to outliers; Theil index allows within/between decomposition and is additive across groups.
- Formula used (Theil index):
  - T = (1/n) ∑_{i=1}^n (Tax_i / μ) × log(Tax_i / μ)
  - T refers to Theil index; Tax_i to a tax subcomponent; μ = (1/n) ∑ Tax_i.
- RDI specifics:
  - Constructed on six tax categories (tier-3 disaggregation).
  - Theil-based RDI range: 0 (perfect diversification) to 1.8 (reliance on one type of tax only).
  - Robustness: HHI-based RDI also computed (Annex 6) and correlated with Theil-based RDI (correlation coefficient 0.98).

### Stylized facts and RDI summary statistics
- Full-sample average RDI: 0.51.
- Extremes:
  - Lowest RDI: Japan 0.05 (2000-15 average; table shows 0.053, 0.055, 0.057 across periods).
  - Highest RDI: Kingdom of Bahrain 1.39 (noted as highest in text).
- Country ranking excerpts (selected exact RDI values from Table 3):
  - Top performers (2010-2015): Japan 0.06, France 0.15, United Kingdom 0.17.
  - Bottom performers (2010-2015): Bolivia 1.34, Kuwait 1.34, Anguilla 1.32.
- Regional and group patterns:
  - NA and EU: lowest RDI (below full-sample average).
  - GCC, LAC, SSA: highest RDI (least diversified).
  - Income-level pattern: OECD countries most diversified; high-income non-OECD and low-income countries record highest RDI.
  - Correlation between log real GDP per capita and RDI: Correlation = -0.43 (Figure shows negative relationship).
- Dynamics over time:
  - Some AEs and EMEs diversified between 2000-04 and 2010-15 (e.g., Austria, Denmark, France, Germany, Japan, Morocco, South Africa).
  - Some countries became more concentrated (e.g., Kuwait, Bahrain, Sri Lanka).
  - Some LIDCs diversified (e.g., Algeria, Côte d'Ivoire, Ghana, Kenya, Mauritius, Uganda).

### Correlations between RDI and macroeconomic indicators (preliminary)
- Key correlations (2000-15):
  - RDI and total tax revenue: Correlation = -0.39 (higher concentration associated with lower tax revenue).
  - RDI and tax revenue volatility: Correlation = 0.29 (higher concentration associated with greater volatility).
  - RDI and GDP growth volatility: Correlation = 0.13.
  - RDI and spending volatility: Correlation = 0.28.
  - RDI and income inequality (GINI): Correlation = 0.37.
  - RDI and exports concentration: Correlation = 0.46.
  - RDI and tax compliance gap: Correlation = 0.54.
  - RDI and tax collection efficiency: Correlation = -0.34.
- Interpretation: More concentrated tax sources tend to coincide with lower revenue, greater volatility (tax, growth, spending), higher income inequality, weaker tax collection efficiency, and larger tax compliance gaps. RDI positively correlated with export concentration (economic diversification link).

### Drivers of tax revenue diversification — econometric findings
- Empirical approach:
  - Dynamic panel regressions using system-GMM on full sample (2000-15).
  - Specification includes lagged RDI and lagged covariates; diagnostics reported (AR(2), Hansen).
  - RDI shows strong persistence: lagged RDI coefficient suggests up to 60-74 percent of current RDI predicted by lagged value (text: "Up to 60-74 percent").
- Structural and macroeconomic drivers (selected exact results and signs):
  - Per capita real GDP: significant non-linear (inverted U-shaped) relationship with RDI — coefficient on log real GDP_pc negative and on squared term positive. Average per capita GDP threshold level reported as $ 4222.
  - Export concentration index: positive and significant — higher export concentration associated with higher tax revenue concentration.
  - Natural resource rents: positive and statistically significant — evidence of a "natural resources curse" on tax diversification.
  - Net ODA per capita: coefficient positive and statistically significant — stronger aid dependency associated with less diversification incentives.
  - Informal sector share: positive — larger informal sectors hinder diversification.
  - Financial development: coefficient significantly negative — deeper financial systems associated with greater diversification (facilitate formalization).
  - Trade openness: coefficients mostly negative (insignificant in many specifications) with one specification showing significance at conventional levels.
  - IMF program dummy: negative and statistically significant in one specification — IMF programs may help diversify tax sources.
- Heterogeneities:
  - LIDCs and EMEs have more room to diversify vs AEs (column findings).
  - Regions with less diversified tax sources include South Asia, LAC, and MENA.
  - Resource-rich dummy: resource-rich countries exhibit less diversified tax structures (positive and significant RR dummy coefficient 0.095** in Table 5).
- Political and institutional drivers (selected exact findings):
  - Democracy (Polity_2): negative and significant — deeper democracy fosters diversification.
  - Political polarization and government fractionalization: negative coefficients (statistically significant) — more polarized/fractionalized systems linked to greater diversification.
  - Political stability: negative and significant — greater stability conducive to diversification.
  - Largest government party orientation: negative and significant — more socialist-oriented governments more prone to diversify taxation sources (variable coded: Right=1, Center=2, Left=3).
  - Institutional quality: quality of bureaucracy, rule of law, government effectiveness, voice and accountability, and control of corruption — negative and significant coefficients indicate stronger institutions support greater diversification.

### Impacts of tax revenue diversification (Section V)
- Empirical strategy: GMM estimations of RDI’s effect on (i) tax-to-GDP ratio (tax revenue) and (ii) tax revenue volatility (standard deviation over 3-year rolling window). Specification includes lagged dependent variable and controls.
- Impact on tax revenue level:
  - Finding: Higher RDI score (higher concentration) associated with lower tax revenue.
  - Magnitude: "A 10 percent increase in the RDI score can yield additional tax revenue of up to 0.2-0.4 percentage points of GDP." (text presents this as result of reversing concentration into diversification).
  - Table 7 regression results: RDI (-1) coefficients are negative and statistically significant across specifications (examples: -0.036***, -0.025***, -0.028***, etc.) with lagged tax revenue persistence high (e.g., Tax revenue (-1) coefficients 0.866*** to 0.966***).
- Impact on tax revenue volatility:
  - Finding: Higher tax revenue concentration (higher RDI) associated with greater revenue volatility; conversely, improving diversification reduces volatility.
  - Magnitude: "A one-point improvement in tax revenue diversification is associated with a reduction in tax revenue volatility of up to 0.5-2.8 points." (text gives range 0.5-2.8).
  - Table 8 regression results: RDI (-1) coefficients positive and statistically significant in volatility regressions (examples: 2.422***, 2.801***, 1.644***, etc.) with revenue volatility persistence also high (Revenue volatility (-1) coefficients e.g., 0.694***, 0.657***).
- Robustness: Standard diagnostic tests for instrument validity generally passed (AR(2), Hansen), with some exceptions noted in footnotes for specific columns.

### Key policy-relevant findings and implications
- Tax revenue diversification is associated with:
  - Higher tax-to-GDP ratios (better revenue collection) — a 10 percent increase in RDI score can translate into up to 0.2-0.4 percentage points of GDP additional tax revenue.
  - Lower tax revenue volatility — a one-point improvement in RDI linked to a reduction in tax revenue volatility up to 0.5-2.8 points.
- Determinants amenable to policy:
  - Strengthening institutions (quality of bureaucracy, rule of law, government effectiveness, control of corruption) can foster diversification.
  - Formalization of the economy and financial development support broader tax bases.
  - Reducing aid dependency and managing natural resource reliance can increase incentives to diversify taxation.
  - Political arrangements (democracy depth, polarization, stability) matter for taxation portfolio choices.
- Practical guidance (inference from empirical results in text):
  - Policies that improve tax administration capacity, deepen financial systems, and broaden the tax base (formalization) are likely to increase diversification and revenue resilience.
  - For resource-rich countries, reforms to reduce over-reliance on resource rents (e.g., diversify economic activity, broaden non-resource tax base) are important for reducing concentration and volatility.
  - IMF-supported programs and policy conditionality may help diversify tax structures where implemented.

### Conclusions and avenues for future research
- Contributions:
  - New national-level RDI for 128 countries (2000-15) based on Theil index; first paper to create such an index at the national level (previous work focused on U.S. states).
  - Demonstrated systematic differences across income groups, regions, fragile/resource-rich status, and over time.
- Main takeaways:
  - AEs exhibit more diversified tax structures than EMEs and LIDCs; resource-rich and fragile states have the most concentrated tax sources.
  - Macroeconomic structure, political and institutional conditions significantly shape tax revenue diversification.
  - Tax revenue diversification materially improves revenue collection and reduces revenue volatility, strengthening fiscal resilience — a critical consideration in contexts such as the coronavirus pandemic.
- Suggested future research directions highlighted in the source:
  - Investigate causal links and transmission channels between per capita GDP and tax revenue diversification.
  - Explore the influence of tax revenue diversification on income inequality.
  - Study how diversification affects policymakers’ leeway for implementing countercyclical fiscal policies.

*Italic: Source — wpiea2020094-print-pdf (sections V and VI and supporting sections) contained in the supplied content.*

### References

### References (wpiea2020094-print-pdf)

### Bibliographic entries
- Full list of bibliographic references cited in the paper, including working papers, journal articles, books, IMF publications, World Bank publications, datasets, and indices. Examples (as listed):
  - Akanbi, O. and Akitoby, B. (2018). “Institutions and Tax Capacity: Is There Any Causality?”, IMF WP, forthcoming.
  - Akitoby, B., Baum, A., Hackney, C., Harrison, O., Primus, K., and Salins, V. (2018). “Tax Revenue Mobilization Episodes in Emerging Markets and Low-Income Countries: Lessons from a New Dataset”, IMF WP 18/234.
  - Aldasoro, I. and Seiferling, M. (2014). “Vertical Fiscal Imbalances and the Accumulation of Government Debt”, IMF WP 14/209.
  - Alesina, A.and Perotti, R. (1995). “Fiscal Expansions and Fiscal Adjustments in OECD Countries”, NBER WP No. 5214 (Cambridge, Massachusetts: NBER).
  - Araujo, J. D., David, A. C., van Hombeeck, C., & Papageorgiou, C. (2017). “Non-FDI Capital Inflows in Low-Income Countries: Catching the Wave?“ IMF Economic Review, 65(2), 426–465.
  - Auerbach, A.J. (1985). “The Theory of Excess Burden and Optimal Taxation.” In Handbook of Public Economics, vol. 1, edited by Auerbach, A. and Feldstein, M. 69–127. Amsterdam: North-Holland, 1985.
  - Bleaney, Y. M., Gemmel, N., and Greenaway, D. (1995). “Tax Revenue Instability, with Particular Reference to Sub-Saharan Africa”, The Journal of Development Studies, 31, 883–902.
  - Budina, N.T., Kinda, T., Schaechter, A. and Weber, A. (2012). “Fiscal Rules in Response to the Crisis-Toward the 'Next-Generation' Rules: A New Dataset.” IMF WP No. 12/187.
  - Cadot, O., Carrère, C., and Strauss-Kahn, V. (2011). “Export diversification: What's behind the hump?”, Review of Economics and Statistics, 93(2), 590–605.
  - Carroll, D. (2005). “Are State Governments Prepared for Fiscal Crises? A Look at Revenue Diversification during the 1990s”, Public Finance Review, 33(5), 603–633.
  - Furceri, D. (2007). “Is Government Expenditure Volatility Harmful for Growth? A Cross-Country Analysis”, Fiscal Studies, 28(1), 103–120.
  - Gaspar, V., Jaramillo, L. and Wingender, P. (2016). “Political Institutions, State Building and Tax Capacity: Crossing the Tipping Point”, IMF WP 16/233.
  - Gaspar, V., and Selassie, A. (2017). “Taxes, Debt and Development: A One-Percent Rule to Raise Revenues in Africa”, IMF Blog 2017/12/05.
  - Gentry, W.M., and Ladd, H.F. (1994). “State Tax Structure and Multiple Policy Objectives”. National Tax Journal, 47 (4), 747–772.
  - Groves, H. M., and Kahn, C.H. (1952). “The Stability of State and Local Tax Yields”. The American Economic Review”, 42 (1), 87–102.
  - Gygli, S., Haelg, F., Potrafke, N. and Sturm, J. (2019): The KOF Globalisation Index – Revisited, Review of International Organizations, 14(3), 543-574
  - Hettich, W., and Winer, S.L. (1984). “A Positive Model of Tax Structure.” Journal of Public Economics, 23, 67–87.
  - Hirschman, A. O. (1964). “The paternity of an index”. The American Economic Review, 54(5), 761–762.
  - International Monetary Fund (2016). “Diversifying Government Revenue in the GCC: Next Steps”, GCC, Annual Meeting of Ministers of Finance and Central Bank Governors, Riyadh, October.
  - International Monetary Fund (2009), “Fiscal Rules—Anchoring Expectations for Sustainable Public Finances”. Washington: International Monetary Fund.
  - Loayza, N., Rancière, R., Servén L, and Ventura, J. (2007). "Macroeconomic volatility and welfare in developing countries: an introduction”, World Bank Economic Review, 21, 343-357.
  - Kilby, N. (2014). “Does Tax Revenue Diversification Help States Weather Economic Downturns? Evidence from the Great Recession”, MPA/MPP Capstone Projects. 12.
  - Medina, L., Jonelis, A. and Cangul, M. (2017). “The Informal Economy in Sub-Saharan Africa: Size and Determinants”, IMF WP 17/156.
  - Marshall, Monty G. and Gurr, Ted Robert 2018. Political Regime Characteristics and Transitions, 1800-2017, Center for Systemic Peace.
  - Papageorgiou, C., and Spatafora, N. (2012). Economic Diversification in LICs: Stylized Facts and Macroeconomic Implications, IMF Staff Discussion Note No. SDN/12/13 (Washington, DC).
  - Roberts, T. (2014). When bigger is better: A critique of the Herfindahl-Hirschman Index's use to evaluate mergers in network industries. Pace L. Rev., 34, 894–946.
  - Schunk, D. and Porka, S. (2005). “State-Local Revenue Diversification, Stability, and Growth: Time Series Evidence”, The Review of Regional Studies, 35 (3), 246–65.
  - Suyderhound, J. (1994). “State-Local Revenue Diversification, Balance, and Fiscal Performance.” Public Finance Review, 22(2):168–194.
  - Svirydzenka, K. 2016. Introducing a New Broad-based Index of Financial Development. IMF Working Paper 16/5, Washington DC
  - Theil, H. (1972). Statistical decomposition analysis; with applications in the social and administrative sciences, 4, HA33, T4.
  - Thornton, J. (2014). “Does foreign aid reduce tax revenue? Further evidence”, Applied Economics 46(4), 359–373.
  - White, C.F. (1983). “Trade-off in Growth and Stability in State Taxes”, National Tax Journal 36, 103–114.
  - Wilford, Walton T. (1965). “State Tax Stability Criteria and the Revenue-Income Elasticity Coefficient Reconsidered.” National Tax Journal, 18, 304–312.
  - Williams, W. V., Anderson, R.M., Froehle, D.O., and Lamb, K.L. (1973). "The Stability, Growth and Stabilizing Influence of State Taxes." National. Tax Journal 36, 267–273.
  - World Bank (2014). Introduction to poverty analysis. World Bank WP No 90288 (Washington, DC: World Bank Group).

### Annex 1 — Sample and Country Groups
- Presents country-level classifications with the following columns: Country; Income group; Region; Small country; Fragile country; Resource rich country.
- Income groups defined in the annex: HIC: High Income Country; UMIC: Upper Middle Income Country; LMIC: Lower Middle Income Country; LIC: Low Income Country.
- Regions defined in the annex: ECA: Europe and Central Asia; EAP: East Asia and Pacific; SA: South Asia; LAC: Latin America; SSA: Sub-Saharan Africa; NA: North America; MENA: Middle East and North Africa.
- The annex lists each country and the values for the columns above (examples shown in the table include Algeria UMIC MENA No No Yes; Kuwait HIC MENA No No Yes; Japan HIC EAP No No No; United States HIC NA No No No, etc.).

### Annex 2 — Data Sources and Variable Descriptions
- Lists variables and their exact descriptions and data sources. Key variables (as listed):
  - Real GDP_pc: Real GDP per capita — IMF's World Economic Outlook (WEO) database
  - Real GDP_pc_squared: Square of real GDP per capita — Authors’ calculations
  - Financial development: Index of financial development — Svirydzenka (2016)
  - Trade openness: Sum of imports and exports over GDP — World Bank's World Development Indicators
  - Exports concentration index: Theil index of exports concentration — IMF datasets
  - Natural resource rents: Natural resource rents in percentage of GDP — World Bank's World Development Indicators
  - Net ODA received_pc: Net Official Development Assistance received per capita — World Bank's World Development Indicators
  - Informal share: Share of the informal sector in the economy (percentage) — Medina, Jonelis and Cangul (2017)
  - Inflation rate / Informality: Consumer price index growth rate (in percentage) — IMF's World Economic Outlook (WEO) database
  - De jure globalization index: It measures the extent of investment restrictions, capital account oppenness and international investment agreements. — Gygli et al. (2019)
  - GDP growth: Rate of real GDP growth — IMF's World Economic Outlook (WEO) database
  - Human capital index: Human capital index, based on years of schooling and returns to education — Penn World Tables 9.1
  - IMF program dummy: Binary variable taking the value of 1 if the country has an IMF program and 0 otherwise — IMF datasets
  - Democracy: Degree of democracy. The polity 2 score ranges from -10 to +10, with higher value representing more democracy. — Marshall and Gurr (2018)
  - Political polarization: It measures the maximum polarization between the executive party and the four principle parties of the legislature. — Database of Political Institutions
  - Government fractionalization: It measures the probability that two deputies picked at random from among the government parties will be of different parties. — Database of Political Institutions
  - Political/Government stability: It measures the likelihood that the government will be destabilized or overthrown by unconstitutional or violent means. — World Bank's Worldwide Governance Indicators
  - Largest gov. party orient.: It measures the largest party orientation with respect to economic policy — Database of Political Institutions
  - Quality of bureaucracy: It measures the institutional strength and quality of the bureaucracy — International Country Risk Guide (ICRG)
  - Rule of law; Government effectiveness; Voice and accountability; Control of corruption — World Bank's Worldwide Governance Indicators
  - Agriculture VA: Agriculture valued added (in percentage of GDP) — World Bank's World Development Indicators
  - Growth volatility: Standard deviation of GDP growth (using rolling window method) — Authors’ calculations
  - Regulatory quality; Political risk; Internal conflicts; Fiscal rules (Dummy: 1 if numerical fiscal rule in place, 0 otherwise) — IMF Fiscal Rules Dataset; various indicators from World Bank, ICRG, Database of Political Institutions
  - Foreign direct investment (FDI), net inflows: Direct investment equity flows — World Bank's Worldwide Governance Indicators
  - Public debt (% GDP): General government total debt, percent of fiscal year GDP
  - Overall fiscal balance: Overall fiscal balance percentage of GDP — IMF's World Economic Outlook (WEO) database
  - Exchange rate: Official exchange rate (LCU per US$, period average) — World Bank's Worldwide Governance Indicators; International Country Risk Guide (ICRG)

### Annex 3 — Countries with Filled up Missing Observations
- Lists countries and specific Year(s) for which missing observations were filled. Examples from the list (exact entries):
  - Armenia, Republic of 2003
  - Brazil 2000-2005
  - Burundi 2010 & 2015
  - Cabo Verde 2010-2015
  - Canada 2010 & 2015
  - China, P.R: Mainland 2000-2004
  - Congo, Republic of 2000-2003
  - Costa Rica 2000-2001
  - Croatia 2000-
  - Egypt 2000-2001
  - Georgia 2000-2002
  - Honduras 2000-2002
  - Indonesia 2000 & 2007
  - Jamaica 2000-2002
  - Korea, Republic of 2000 & 2006
  - Lesotho 2000-2002
  - Mauritius 2000-2001
  - Moldova 2000-2001
  - Seychelles 2000-2004
  - Turkey 2000 & 2007

### Annex 4 — Full RDI-based Country Ranking (by subperiod)
- Provides country rankings by RDI for three subperiods: 2000-2004, 2005-2009, 2010-2015. Each ranking lists Rank, Country, and RDI value (exactly as reported). Selected top entries (exact values preserved):
  - 2000-2004 top 10:
    - Rank 1 Japan 0.053
    - Rank 2 France 0.160
    - Rank 3 United Kingdom 0.167
    - Rank 4 United States 0.188
    - Rank 5 South Africa 0.192
    - Rank 6 Switzerland 0.193
    - Rank 7 Norway 0.216
    - Rank 8 Australia 0.229
    - Rank 9 Israel 0.230
    - Rank 10 Spain 0.243
  - 2005-2009 top 10:
    - Rank 1 Japan 0.055
    - Rank 2 United Kingdom 0.140
    - Rank 3 France 0.149
    - Rank 4 United States 0.156
    - Rank 5 Switzerland 0.182
    - Rank 6 South Africa 0.187
    - Rank 7 Israel 0.192
    - Rank 8 Australia 0.193
    - Rank 9 Norway 0.204
    - Rank 10 Spain 0.209
  - 2010-2015 top 10:
    - Rank 1 Japan 0.057
    - Rank 2 France 0.153
    - Rank 3 United Kingdom 0.172
    - Rank 4 United States 0.176
    - Rank 5 South Africa 0.178
    - Rank 6 Switzerland 0.181
    - Rank 7 Singapore 0.186
    - Rank 8 Australia 0.193
    - Rank 9 Norway 0.207
    - Rank 10 Korea, Republic of 0.217
- The annex continues with full rankings through lower ranks and includes entries with RDI values for many countries across the three subperiods (examples include entries with RDI values such as Luxembourg 0.245, Ireland 0.289, India 0.448, Ghana 0.595, Sierra Leone 0.595, up to values exceeding 1.0 for some countries in certain subperiods).

*References and annex material as provided in the source PDF.*

### Annex 5. Alternative estimates

### Annex 5. Alternative estimates

### Accounting for shift in value added across sectors — regression specification
- Dependent variable: Revenue diversification index (RDI)
- Control: VA share of services
- Four baseline specifications reported (columns (1) to (4))

### Main coefficient estimates (by variable)
- RDI (t-1)
  - (1): 0.611***
  - (2): 0.478***
  - (3): 0.459***
  - (4): 0.453***
  - Standard errors: (0.011), (0.024), (0.022), (0.023)
- Log real GDP_pc (t-1)
  - (1): -0.6170***
  - (2): -0.8073***
  - (3): -0.7172***
  - (4): -0.7804***
  - Standard errors: (0.080), (0.123), (0.128), (0.135)
- Log real GDP_pc_squared (t-1)
  - (1): 0.038***
  - (2): 0.051***
  - (3): 0.0442***
  - (4): 0.0487***
  - Standard errors: (0.005), (0.008), (0.008), (0.009)
- Financial development (t-1)
  - (1): -0.3958***
  - (2): -0.6439***
  - (3): -0.5385***
  - (4): -0.6052***
  - Standard errors: (0.108), (0.133), (0.138), (0.139)
- Trade openness (t-1)
  - (1): -0.0318*
  - (2): -0.0212
  - (3): -0.0257
  - (4): -0.0116
  - Standard errors: (0.019), (0.026), (0.026), (0.028)
- Export concentration index (t-1)
  - (1): 0.0252***
  - (2): 0.0294***
  - (3): 0.0339***
  - (4): 0.0302***
  - Standard errors: (0.007), (0.009), (0.009), (0.009)
- VA Services / VA Agri.
  - (1): -0.0001***
  - (2): -0.0001***
  - (3): 0.000
  - (4): 0.000
- VA Services / VA Manuf.
  - (3): 0.0023**
  - (4): 0.0020**
  - Standard errors: (0.001), (0.001)
- Constant
  - (1): 2.631***
  - (2): 3.486***
  - (3): 3.137***
  - (4): 3.361***
  - Standard errors: (0.346), (0.518), (0.540), (0.564)

### Sample, controls, and diagnostics
- Nb. of observations:
  - (1): 1141
  - (2): 1089
  - (3): 1074
  - (4): 1074
- Countries:
  - (1): 97
  - (2): 96
  - (3): 95
  - (4): 95
- AR(1):
  - (1): 0.08
  - (2): 0.04
  - (3): 0.05
  - (4): 0.06
- AR(2) p-value:
  - (1): 0.3
  - (2): 0.14
  - (3): 0.25
  - (4): 0.25
- Hansen OID (p-value):
  - (1): 0.07
  - (2): 0.11
  - (3): 0.2
  - (4): 0.22
- Nb. of instruments:
  - (1): 70
  - (2): 71
  - (3): 71
  - (4): 72
- Year fixed effects: Yes (in all specifications)
- Region fixed effects: Yes (in all specifications)

*Source: Annex 5. Alternative estimates, wpiea2020094-print-pdf*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2020/english/wpiea2020094-print-pdf.pdf_