## wp18234

## Source details

**Canonical URL:** [wp18234](https://www.imf.org/-/media/files/publications/wp/2018/wp18234.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2018/wp18234.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2018/wp18234.pdf.json)

---

### Introduction and study scope
- Study period: 2000-2015.
- Sample: 136 countries (71 EMs and 65 LICs).
- Dataset highlights:
  - 55 episodes of tax revenue mobilization in which revenue administration and tax policy reforms played a crucial role; of these, 29 are observed in LICs, 18 in EMs, and 8 in resource-rich (RR) economies.
  - For identified episodes in the dataset, tax-to-GDP ratios increased by at least an average of 0.5 percent of GDP per year over a minimum of three years.
- Government levels used: general government (GG) tax revenue where available; otherwise central government (CG) tax revenue used as proxy.
- Data sources: World Revenue Longitudinal Database (WoRLD), World Economic Outlook (WEO), IMF staff estimates, IMF country reports, internal IMF technical assistance reports, and external country-case sources.
- Cyclical adjustment: tax revenue ratio adjusted using an output elasticity of one (cyclically adjusted revenue-to-trend GDP ratio equivalent to tax revenue-to-GDP).

### Key findings (overview)
- Complementarity of policy and administration:
  - Many countries that observed large revenue increases pursued revenue administration and tax policy reforms in parallel; only a few pursued either approach alone.
- Base broadening and compliance:
  - Broadening the tax base for both direct and indirect taxes via administration and tax policy—especially through improvements in compliance and by reducing exemptions and/or eliminating tax holidays—were frequently used measures.
- Indirect taxation emphasis:
  - Introduction, simplification, and efficiency improvements of VAT often led to significant revenue gains.
  - Excise taxation was also used frequently.
  - Property taxes played a limited role during identified episodes in both LICs and EMs.
- Sustainability of revenue gains:
  - About one third of increases in the tax-to-GDP ratios were sustained (they did not fall for at least three years after the increase), and another one third fell moderately (less than one third of the initial gain was lost).
  - Sustainability usually coincides with tax administration reforms in key compliance areas: risk-based audits, registration, filing, payment, and reporting.
- Political economy factors:
  - High-level political commitment and broad stakeholder buy-in were crucial for success; deeper social dialogue enhanced likelihood of implementation and sustainability.
- Causality caveat:
  - Causality cannot be inferred from this stocktaking exercise; while reforms likely contributed to observed revenue increases, causality cannot be reliably established given data limitations and lack of micro-evidence or natural control groups.
- Caveat on revenue quality:
  - Not all revenue-raising reforms imply better tax systems (examples: raising equity concerns or obtaining revenue via failing to give full and timely VAT refunds).

### Tax revenue developments and composition (selected statistics)
- Aggregate trends:
  - Average (unweighted) tax revenue increased by 3.5 percent of GDP between 2000 and 2015, measuring 16.4 percent in 2015.
  - EMs: increase of 0.9 percent of GDP between 2000 and 2015, reaching an average 17.3 percent in 2015.
  - Advanced economies (AEs): average tax-to-GDP ratio measured 25.5 percent in 2015.
- Resource-rich (RR) vs non-RR:
  - Since 2000, average annual increase was 0.2 percent for non-RR countries and 0.3 percent for RR countries.
- Direct and indirect tax details:
  - Average income tax (PIT and CIT combined) as a share of GDP:
    - EMs: 6.3 percent in 2015, up from 4.5 percent in 2000.
    - LICs: 5.1 percent in 2015, up from 3.3 percent in 2000.
  - In LICs, personal income tax (PIT) roughly doubled between 2000 and 2015; CIT increased only modestly.
- VAT adoption and role:
  - Since 2000, 23 (out of 70) LICs and 20 (out of 83) EMs introduced VAT (examples cited in source).
  - VAT revenue accounted for approximately one third of all tax revenue in both LICs and EMs in 2015.
  - Countries with VAT recorded higher tax-to-GDP ratios than those without.
  - Dataset footnote: In the dataset used, 21 of 65 LICs and 17 of 71 EMs introduced VAT since 2000.
- VAT collection efficiency (C-efficiency):
  - LICs: improved from 0.25 in 2000 to 0.36 in 2015.
  - EMs: improved from 0.47 in 2000 to 0.57 in 2015.
  - Progress slowed since the global financial crisis; slight decline in EMs; remains low in LICs.
- Excise and property taxes:
  - Excise tax collection in LICs rose from 1.7 percent of GDP in 2000 to 2.1 percent of GDP in 2015.
  - Property tax revenue remained below one percent of GDP for both LICs and EMs but increased steadily in EMs.

### Empirical strategy and episode identification
- Episode definition:
  - Episodes are defined as 3-year compounded windows of an average increase of 0.5 percent of GDP per year or more (equivalent to 1.5 percent within three years).
  - The first episode is triggered by the first 3-year window meeting the 0.5 percent per year threshold.
  - Successive overlapping three-year windows that continue to meet the criterion are compounded into the same episode; the end-point is the last overlapping window fitting the criterion.
- Identification results:
  - From 2000-2015 sample of 136 countries, 172 episodes were identified that fit the 3-year, 0.5 percent per year criterion.
  - Episode breakdown: 61/30 are observed in non-RR/RR EMs, and 65/16 in non-RR/RR LICs.
  - The paper limits in-depth analysis to a subset of episodes meeting stricter criteria, resulting in 107 countries subject to in-depth analysis.
- Episode qualifiers and in-depth sample:
  - In-depth sample selection combined: (i) Countries increasing tax-to-GDP ratios by at least 0.5 percent each year for at least 3 consecutive years (48 cases); (ii) Countries with beyond average tax increases relative to episode length (72 cases); 19 cases overlap; adding 6 robustness cases yields 107 countries.
  - Revenue administration and tax policy measures coincided with observed tax increases in 55 out of 107 studied cases.
- Robustness check (“Top Revenue Performers”):
  - Countries’ tax-to-GDP ratios related to logged GDP per capita and split by income groups (LICs and EMs, excluding RR).
  - Countries categorized into quartiles by residual percentiles relative to a 2015 fitted trend: “High tax performers” (above the 85th percentile); “Good tax performers” (50th–85th); “Low tax performers” (15th–50th); “Poor tax performers” (below the 15th).
  - Between 2009 and 2015, 18 out of 79 countries registered an improvement in tax performance.
  - Of the 18 improved cases, all except six (Albania, Bahamas, Jamaica, Philippines, Senegal, and Turkey) were also identified under the main identification strategy.
  - Six countries not in the original sample were added to the in-depth analysis; Albania was excluded from further analysis for lack of evidence linking revenue increase to tax policy or RA reforms.

### Episode characteristics and sustainability (selected results)
- Duration of Episodes:
  - Average episode duration is 3.5 years in RR countries and 3.4 years in non-RR countries.
  - Median episode duration is 3 years across all countries.
  - 16 percent of episodes lasted only one year.
  - 23 percent of identified episodes lasted longer than 5 years.
  - Episodes can be shorter than three years if the entire 1.5 percent increase (or more) occurred within one or two years.
- Initial Tax Ratios and starting points:
  - A “tipping point” of tax revenue is noted around 13 percent of GDP; studies find countries that achieve a tax-to-GDP rate of at least 12.75 experience sustained economic growth.
  - For LICs and RR countries most large increases began while the tax-to-GDP/non-commodity revenue-to-GDP ratio was below 15 percent.
  - Case examples:
    - Afghanistan increased its tax-to-GDP ratio by 3.3 percent within two years, from an initial ratio of 5.6 percent.
    - Liberia increased its tax-to-GDP ratio from 11.4 percent to above 16 percent within two years and sustained the increase.
- Sustainability of Revenue Gains:
  - Revenue increases were sustained or increased in slightly less than one third of the cases.
  - Revenue fell by up to one third of the initial tax revenue increase in close to one third of the cases.
  - In more than one third of the cases the gains were not sustained, especially in RR countries.
  - Major explanations for declines post-episode include macroeconomic shocks (internal or external, above all the global financial crisis) and/or revenue-reducing policy measures.
    - Example: Moldova’s tax revenue fell by about 1.3 percent in 2008 following a deep recession caused by the global financial crisis.
    - Policy-driven declines example: Antigua and Barbuda, and Dominica experienced temporary tax relief and reductions in personal income tax rates during the GFC.

### Tax policy measures observed (detailed findings)
- Share and types of measures:
  - Changes to indirect taxation make up more than 56 percent of all discretionary tax policy changes.
  - Changes to income taxes account for another 26 percent.
  - Most country-episodes used multiple tax instruments; “0” tax policy instruments indicate RA was the primary reform-driver.
- Rate changes and exemptions:
  - Rate changes are most common: 85 out of 187 observations (excluding subsidies), driven by those to excise taxation.
  - Reductions in exemptions (especially to VAT, CIT, and other goods and services taxes) are the second most frequent tool.
  - Threshold changes are not common.
  - Rate increases are observed primarily for indirect taxes.
  - Rate changes to PIT and CIT were primarily negative: 7 out of 10 rate changes were negative for both PIT and CIT during the identified episodes.
  - Half of the identified 8 threshold adjustments were increases.
- VAT and general goods and services taxes:
  - Introduction of VAT led to substantial increases in tax revenue in some cases.
  - In 48 percent of episodes, changes to VAT (introduction, rate changes, thresholds or exemptions) were part of the strategy.
  - 29 percent of episodes used changes to other general goods and services taxation.
- Excise taxation:
  - Excise increases frequently applied on fuels, tobacco, alcoholic and non-alcoholic drinks, cars, telecom sector.
- Less frequent instruments:
  - Property taxation observed in only 3 percent of cases.
  - Reduction of subsidies observed in 4 percent of observations.
- Illustrative country examples (from source table excerpts):
  - Afghanistan, Burkina Faso, Burundi, Cabo Verde, Cambodia, Liberia, Maldives among selected episodes with reported initial revenue and revenue increase figures (exact figures preserved in source tables).

### Revenue administration (RA) measures observed
- Overall prevalence:
  - RA measures were common in almost all (90 percent) of the episodes, regardless of income group or tax policy mix.
- Audit and verification:
  - 89 percent of episodes that involved an improvement in RA targeted changes in audits and verification programs.
  - Focus on administering audits, combatting tax evasion, reducing corruption.
- Management, governance & HR:
  - 77 percent of episodes used measures such as hiring qualified staff, strategic planning, training, strengthening tax legislation, and empowering revenue collection agencies.
  - Establishing semi-autonomous revenue authorities sometimes coincided with tax-to-GDP increases; success depends on capacity and staffing.
- IT systems:
  - In more than half of the episodes, IT system improvements supported core RA processes.
  - Measures included IT strategic plans, computerization, and use of third-party information.
  - Caution: IT improvements need to accompany administrative advances to be fully beneficial.
- Registration, filing, enforcement, and customs:
  - Improvements in registration and filing, timely filing/payment enforcement, strengthened customs clearance, and voluntary compliance initiatives supported revenue increases.
- Management of payment obligations and LTUs:
  - Policies to manage and recover tax arrears contributed positively.
  - Large Taxpayer Units (LTUs) implemented in some countries to manage entities that can account for as much as 80 percent of all tax revenue.
- Multiple RA instruments:
  - In more than 30 percent of country-episodes, 4-5 elements of RA were addressed in parallel.
  - Examples: Paraguay, Georgia (multiple RA elements implemented jointly).
- Sustainability caveat:
  - Initial gains from tax policy measures can be eroded without sustained RA reforms (examples: Antigua and Barbuda; The Gambia).

### Role of IMF programs and external factors
- 58 percent of the analyzed episodes were accompanied by IMF programs.
- IMF programs tend to have a positive effect on tax revenue, especially where revenue conditionality applies, and most strongly for LICs where revenue ratios are below the group average.
- Stronger institutions (less perceived corruption) strengthen the positive impact of revenue conditionality.
- The in-depth analysis excludes episodes primarily driven by exogenous economic factors or external shocks (example: Lesotho and Swaziland episodes linked to SACU revenues were excluded).

### Conclusions and policy-relevant implications
- Sustained and large revenue gains are usually accompanied by a comprehensive strategy linking tax policy reforms and revenue administration reforms.
- Most frequently observed tax policy instruments:
  - Rate increases in indirect taxation (excises and GST).
  - Elimination or reduction of exemptions and tax holidays.
  - Simplification (reducing number of taxes or introducing broad-based VAT) in some episodes.
  - Limited role for property taxes due to legal and RA capacity constraints.
- Key RA measures associated with mobilization:
  - Risk-based audits.
  - Management/governance/HR strategies.
  - IT system enhancements.
- Policy recommendations (derived from findings):
  - Combine revenue administration reforms with tax policy changes to achieve larger and more sustainable revenue gains.
  - Prioritize base broadening (reducing exemptions, eliminating tax holidays).
  - Strengthen compliance-enhancing administration measures (risk-based audits, improved registration, filing, payment, reporting).
  - Implement VAT introduction and VAT reforms where appropriate, with adequate administrative capacity to realize C-efficiency gains.
  - Secure high-level political buy-in and sustained social dialogue to increase likelihood of successful implementation and sustainability of reforms.

### Suggestions for future research (as proposed)
- Analyze episodes of tax reform failures.
- Assess the relative importance of RA versus tax policy changes and measures within categories.
- Explore differences in revenue mobilization and sustainability between IMF program and non-program countries, and the impact of IMF and other donors’ technical assistance.
- Analyze drivers of sustainability such as international assistance or growth acceleration.

*Source: wp18234 (References and accompanying chapter content provided in the source PDF).*

### REFERENCES .............................................................................................................

### wp18234 - REFERENCES

### Introduction and study scope
- Study period: 2000-2015.
- Sample: 136 countries (71 EMs and 65 LICs).
- Dataset highlights:
  - Provides a comprehensive analysis of efforts to increase tax revenues in LICs and EMs.
  - Covers 55 episodes of tax revenue mobilization during which revenue administration and tax policy reforms played a crucial role; of these, 29 are observed in LICs, 18 in EMs, and 8 in resource-rich (RR) economies.
  - For identified episodes in the dataset, tax-to-GDP ratios increased by at least an average of 0.5 percent of GDP per year over a minimum of three years.
- Government levels used: general government (GG) tax revenue where available; otherwise central government (CG) tax revenue used as proxy.
- Data sources: World Revenue Longitudinal Database (WoRLD), World Economic Outlook (WEO), IMF staff estimates, IMF country reports, internal IMF technical assistance reports, and external country-case sources.

### Key findings
- Complementarity of policy and administration:
  - Many countries that observed large revenue increases pursued revenue administration and tax policy reforms in parallel; only a few pursued either approach alone.
- Base broadening and compliance:
  - Broadening the tax base for both direct and indirect taxes via administration and tax policy—especially through improvements in compliance and by reducing exemptions and/or eliminating tax holidays—were frequently used measures.
- Indirect taxation emphasis:
  - Many countries focused on indirect taxation. Introduction, simplification, and efficiency improvements of VAT often led to significant revenue gains.
  - Excise taxation was also used frequently.
  - Property taxes played a limited role during identified episodes in both LICs and EMs.
- Sustainability of revenue gains:
  - About one third of increases in the tax-to-GDP ratios were sustained (they did not fall for at least three years after the increase), and another one third fell moderately (less than one third of the initial gain was lost).
  - Sustainability usually coincides with tax administration reforms in key compliance areas: risk-based audits, registration, filing, payment, and reporting.
- Political economy factors:
  - High-level political commitment and broad stakeholder buy-in were crucial for success; deeper social dialogue enhanced likelihood of implementation and sustainability.
- Causality caveat:
  - Causality cannot be inferred from this stocktaking exercise; while reforms likely contributed to observed revenue increases, causality cannot be reliably established given data limitations and lack of micro-evidence or natural control groups.
- Caveat on revenue quality:
  - Not all revenue-raising reforms imply better tax systems (examples: raising equity concerns or obtaining revenue via failing to give full and timely VAT refunds).

### Tax revenue developments and composition (selected statistics)
- Aggregate trends:
  - Average (unweighted) tax revenue increased by 3.5 percent of GDP between 2000 and 2015, measuring 16.4 percent in 2015.
  - EMs: increase of 0.9 percent of GDP between 2000 and 2015, reaching an average 17.3 percent in 2015.
  - Advanced economies (AEs): average tax-to-GDP ratio measured 25.5 percent in 2015.
- Resource-rich (RR) vs non-RR:
  - Since 2000, average annual increase was 0.2 percent for non-RR countries and 0.3 percent for RR countries.
- Direct and indirect tax details:
  - Average income tax (PIT and CIT combined) as a share of GDP:
    - EMs: 6.3 percent in 2015, up from 4.5 percent in 2000.
    - LICs: 5.1 percent in 2015, up from 3.3 percent in 2000.
  - In LICs, personal income tax (PIT) roughly doubled between 2000 and 2015; CIT increased only modestly.
- VAT adoption and role:
  - Since 2000, 23 (out of 70) LICs and 20 (out of 83) EMs introduced VAT (examples cited: Seychelles (2012), The Gambia (2013), Malaysia (2015), Suriname (2016)).
  - VAT revenue accounted for approximately one third of all tax revenue in both LICs and EMs in 2015.
  - Countries with VAT recorded higher tax-to-GDP ratios than those without.
  - Note (dataset footnote): In the dataset used, 21 of 65 LICs and 17 of 71 EMs introduced VAT since 2000.
- VAT collection efficiency (C-efficiency):
  - LICs: improved from 0.25 in 2000 to 0.36 in 2015.
  - EMs: improved from 0.47 in 2000 to 0.57 in 2015.
  - Progress slowed since the global financial crisis; slight decline in EMs; remains low in LICs.
- Excise and property taxes:
  - Excise tax collection in LICs rose from 1.7 percent of GDP in 2000 to 2.1 percent of GDP in 2015.
  - Property tax revenue remained below one percent of GDP for both LICs and EMs but increased steadily in EMs.

### Data description and measurement notes
- Countries covered: 136 countries (71 EMs, 65 LICs); of these, 31 are fragile states (FS) per IMF (2017) definition.
- Tax revenue categories: PIT; CIT; goods and services taxes (VAT, excise taxes, general/other GST); trade tax; property tax; others.
- For RR countries, non-commodity revenue is computed by subtracting total commodity revenue and grants from GG revenue (non-commodity revenue sourced from WEO).
- Cyclical adjustment:
  - The tax revenue ratio is adjusted for the cyclical element by applying an output elasticity to tax revenue of one (cyclically adjusted revenue-to-trend GDP ratio used as baseline).
  - With elasticity of one, the cyclically adjusted tax revenue-to-trend GDP ratio is equivalent to tax revenue-to-GDP.

### Empirical strategy and episode identification
- Episode definition:
  - Episodes are defined as 3-year compounded windows of an average increase of 0.5 percent of GDP per year or more (equivalent to 1.5 percent within three years).
  - The first episode is triggered by the first 3-year window meeting the 0.5 percent per year threshold.
  - Successive overlapping three-year windows that continue to meet the criterion are compounded into the same episode; the end-point is the last overlapping window fitting the criterion.
  - The 0.5 percent of GDP per year threshold is often set as a target in IMF-supported programs and recommended for building stronger institutions and supporting growth.
- Identification results:
  - From 2000-2015 sample of 136 countries, 172 episodes were identified that fit the 3-year, 0.5 percent per year criterion.
  - Episode breakdown: 61/30 are observed in non-RR/RR EMs, and 65/16 in non-RR/RR LICs.
  - Subsequent analysis narrows the 172 episodes to those with more tangible tax revenue mobilization results using two additional episode qualifiers (detailed in the paper).

### Policy-relevant implications (derived from findings)
- Combining revenue administration reforms with tax policy changes tends to be associated with larger and more sustainable revenue gains.
- Prioritize base broadening (reducing exemptions, eliminating tax holidays) and compliance-enhancing administration measures (risk-based audits, improved registration, filing, payment, reporting).
- VAT introduction and reforms (simplification, efficiency gains) are effective tools for robust revenue mobilization in many LICs and EMs, but require administrative capacity to realize potential C-efficiency gains.
- Political buy-in at the highest level and sustained social dialogue increase the likelihood of successful implementation and sustainability of reforms.

*Source: wp18234 (References and accompanying chapter content provided in the source PDF).*

### 0.44 percent of GDP in RR countries.

### wp18234 - 0.44 percent of GDP in RR countries

### Overview
- The analysis identifies 172 episodes of large tax revenue increases across countries using WoRLD, WEO, and IMF staff estimates.
- Episodes are defined as 3-year compounded windows with an average increase of 0.5 percent per year or more (1.5 percent within three years).
- The paper limits in-depth analysis to a subset of episodes meeting stricter criteria, resulting in 107 countries subject to in-depth analysis.

### Duration of Episodes
- Average episode duration is 3.5 years in RR countries and 3.4 years in non-RR countries.
- Median episode duration is 3 years across all countries.
- 16 percent of episodes lasted only one year.
- 23 percent of identified episodes lasted longer than 5 years.
- Episodes can be shorter than three years if the entire 1.5 percent increase (or more) occurred within one or two years.

### Initial Tax Ratios and Starting Points
- Episodes occurred with both low and high initial tax-to-GDP ratios.
- A “tipping point” of tax revenue is noted around 13 percent of GDP (Gaspar et al. (2016a) and Gaspar et al. (2016b)); the studies find that countries which achieve a tax-to-GDP rate of at least 12.75 experience sustained economic growth.
- For LICs and RR countries, most large increases began while the tax-to-GDP/non-commodity revenue-to-GDP ratio was below 15 percent.
- Examples from cases cited:
  - Afghanistan increased its tax-to-GDP ratio by 3.3 percent within two years, from an initial ratio of 5.6 percent.
  - Liberia increased its tax-to-GDP ratio from 11.4 percent to above 16 percent within two years and sustained the increase.
- Revenue mobilization can be achieved from a wide range of initial revenue ratios, though it is more frequently observed from a lower starting point.

### Sustainability of Revenue Gains
- Revenue increases were sustained or increased in slightly less than one third of the cases.
- Revenue fell by up to one third of the initial tax revenue increase in close to one third of the cases.
- In more than one third of the cases the gains were not sustained, especially in RR countries.
- Major explanations for declines post-episode include macroeconomic shocks (internal or external, above all the global financial crisis) and/or revenue-reducing policy measures.
  - Example: Moldova’s tax revenue fell by about 1.3 percent in 2008 following a deep recession caused by the global financial crisis.
  - Examples of policy-driven declines: Antigua and Barbuda, and Dominica experienced temporary tax relief and reductions in personal income tax rates during the GFC.

### Robustness Check: “Top Revenue Performers” by Income Level
- Second strategy relates countries’ tax-to-GDP ratios to logged GDP per capita, splitting by income groups (LICs and EMs, excluding RR economies).
- Countries categorized into four groups based on residual percentiles relative to a 2015 fitted trend:
  - “High tax performers” – above the 85th percentile.
  - “Good tax performers” – between the 50th and 85th percentiles.
  - “Low tax performers” – between the 15th and 50th percentiles.
  - “Poor tax performers” – below the 15th percentile.
- Between 2009 and 2015, 18 out of 79 countries registered an improvement in tax performance.
  - The Maldives jumped two categories, from poor to good.
- Of the 18 improved cases, all except six (Albania, Bahamas, Jamaica, Philippines, Senegal, and Turkey) were also identified under the main identification strategy.
- Six countries not in the original sample were added to the in-depth analysis to capture all countries that improved their tax performance.
- Albania was excluded from further analysis because country documents did not provide evidence that the tax revenue increase resulted from tax policy or revenue administration reforms.

### Tax Policy and Revenue Administration Reforms Behind Mobilization
- In-depth country-by-country analysis cannot infer causality; other factors (sectoral composition, foreign aid, external debt, political-institutional indicators) are not controlled for.
- The in-depth sample selection:
  - (i) Countries increasing tax-to-GDP ratios by at least 0.5 percent each year for at least 3 consecutive years (48 cases).
  - (ii) Countries with beyond average tax increases relative to episode length (72 cases).
  - 19 cases overlap between the two methods; adding 6 robustness cases yields 107 countries for in-depth analysis.
- Revenue administration and tax policy measures coincided with observed tax increases in 55 out of 107 studied cases.
- The descriptive dataset corresponding to episodes and measures is referenced for further details (dataset title provided in source text).

### Common Measures and Findings
- Successful revenue mobilization is most frequently supported by combinations of:
  - Rationalization of exemptions.
  - Implementation of or adjustments to a broad-based VAT and other goods and services taxes.
  - Rate changes to or introduction of excises on selected goods.
  - Strong revenue administration management and governance arrangements.
  - Improvements in core tax administration functions.
- Revenue administration measures underpinned mobilization efforts in almost all episodes and country cases, regardless of income group.
  - In some cases, administrative measures were the main reason for the increase (examples: Cambodia, Republic of Congo, Sierra Leone).
  - In a few cases no significant administrative measures could be identified (examples: St. Vincent & the Grenadines, Antigua & Barbuda, Barbados, Turkey, Ecuador).

### Role of IMF Programs and External Factors
- 58 percent of the analyzed episodes were accompanied by IMF programs.
- IMF programs tend to have a positive effect on tax revenue, especially where revenue conditionality applies, and most strongly for LICs where revenue ratios are below the group average (as noted in Crivelli and Gupta (2014)).
- Stronger institutions (less perceived corruption) strengthen the positive impact of revenue conditionality.
- The in-depth analysis excludes episodes primarily driven by exogenous economic factors or external shocks (example: Lesotho and Swaziland episodes linked to SACU revenues were excluded).

### Sample Table Highlights (from Table 1 excerpts)
- Selected country episode examples with initial revenue and increase:
  - Afghanistan 2009-2010: Revenue in T-1 = 5.6; Revenue Increase = 3.3; Persisted = FS.
  - Burkina Faso 2009-2013: Revenue in T-1 = 11.8; Revenue Increase = 4.4; Persisted = FM.
  - Burundi 2009-2011: Revenue in T-1 = 11; Revenue Increase = 2.0; Persisted = FS.
  - Cabo Verde 2004-2007: Revenue in T-1 = 17.8; Revenue Increase = 3.4; Persisted = FS.
  - Cambodia 2012-2015: Revenue in T-1 = 10.3; Revenue Increase = 4.0; Persisted = -.
  - Liberia 2006-2007: Revenue in T-1 = 11.4; Revenue Increase = 4.9; Persisted = S.
  - Maldives 2011-2015: Revenue in T-1 = 10.0; Revenue Increase = 14.6; Persisted = -.
- Notes on persistence categories:
  - S - Sustained (revenue increased or stayed the same for at least three years after the revenue period ended; two years for episodes ending in 2015).
  - FM - Fell moderately (revenue fell by less than one third of the overall increase).
  - FS - Fell strongly (revenue fell by more than one third of the overall increase).

*Source: wp18234 - 0.44 percent of GDP in RR countries (IMF staff estimates, WoRLD, WEO).*

### 6.9 FS●   ●●-

### 6.9 FS●   ●●-

### A. Key Findings – Tax Policy Measures
- Large tax revenue increases often coincide with higher collection of revenue from indirect taxation.
- Changes to indirect taxation make up more than 56 percent of all discretionary tax policy changes.
- Changes to income taxes account for another 26 percent.
- Countries rely on a wide range of tax policy instruments; most country-episodes used multiple tax instruments ("0" tax policy instruments indicate revenue administration was the primary reform-driver).
  - Examples: Belize, 2013-2014 (GST and excise taxes); Kyrgyz Republic, 2011-12 (CIT, VAT, excises, and other GST); Seychelles, 2008-11 (PIT, CIT, VAT, excises, other GST, trade taxes and subsidies).
- Tax policy measures categorized as: tax introductions ("new taxes"), rate changes (positive or negative), threshold changes (increases or reductions) and changes to exemptions (base broadening).
  - Rate changes are most common: 85 out of 187 observations (excluding subsidies), driven by those to excise taxation.
  - Reductions in exemptions (especially to VAT, CIT, and other goods and services taxes) are the second most frequent tool.
  - Threshold changes are not common.
- Rate changes observed:
  - Rate increases are observed primarily for indirect taxes.
  - Rate changes to PIT and CIT were primarily negative: 7 out of 10 rate changes were negative for both PIT and CIT during the identified episodes.
  - Half of the identified 8 threshold adjustments were increases.
- Broadening the tax base and reducing the number of tax rates was frequently part of revenue mobilization strategies.
  - Examples: Georgia reduced the number of taxes to be paid from 21 to 7; Morocco reduced the number of VAT special regimes; Jamaica introduced VAT to residential electricity consumption; Seychelles converted a 7 percent GST into a 12 percent GST.
- Curbing tax exemptions was prominent: in more than half of the country episodes various exemptions were eliminated (frequently related to VAT, general GST, and CIT).
  - Specific examples: reducing statutory and discretionary exemptions (Guyana, Solomon Islands); ending tax holidays (Burkina Faso); ending CIT exemptions (Mauritania); eliminating VAT exemptions (Uganda).
- VAT and other general goods and services taxation:
  - Introduction of VAT led to substantial increases in tax revenue in some cases (Bahamas, Bosnia and Herzegovina, Dominica, The Gambia, Guyana).
  - In 48 percent of episodes, changes to VAT (introduction, rate changes, thresholds or exemptions) were part of the strategy.
  - 29 percent of episodes used changes to other general goods and services taxation.
- Excise taxation increases were frequently applied (often on fuels, tobacco, alcoholic and non-alcoholic drinks, cars, telecom sector) and used for immediate revenue needs or to balance shortfalls.
- Taxing sector-specific domestic economic rents (tourism, telecom, oil) was less frequent but helpful in some countries (Maldives, Malawi, Rwanda).
- Property taxation and reduction of subsidies were least frequently observed:
  - Property taxation observed in only 3 percent of cases.
  - Reduction of subsidies observed in 4 percent of observations.

### B. Key Findings – Revenue Administration Measures
- Revenue administration (RA) measures were common in almost all (90 percent) of the episodes, regardless of income group or tax policy mix.
- The paper does not quantify RA impact nor establish causal links, but notes international evidence that improved tax administration helps reduce the tax gap.
- Audit and verification improvements:
  - 89 percent of episodes that involved an improvement in RA targeted changes in audits and verification programs.
  - Focus: administering audits, combatting tax evasion, reducing corruption.
  - Examples: Georgia and Guinea coincided with intensified tax and customs audits.
- Management, governance, and human resources:
  - 77 percent of the episodes used measures such as hiring qualified staff, strategic planning, training, strengthening tax legislation, and empowering revenue collection agencies.
  - Establishing semi-autonomous revenue authorities sometimes coincided with tax-to-GDP increases (example: Jamaica), but success depends on capacity and staffing.
- Information Technology (IT) improvements:
  - In more than half of the episodes, IT system improvements supported core RA processes (examples: Burkina Faso, Tonga, Guinea-Bissau, Uruguay).
  - Measures included IT strategic plans, computerization of tax and customs administration, and use of third-party information.
  - Caution: IT improvements need to accompany administrative advances to be fully beneficial.
- Registration, filing, and enforcement:
  - Improvements in registration and filing supported revenue increases (examples: Central African Republic, Georgia, Guinea-Bissau).
  - Emphasis on ensuring timely filing, payment, and enforcing sanctions.
- Customs and voluntary compliance:
  - Strengthened customs clearance and voluntary compliance supported revenue (examples: Tonga, Guinea-Bissau).
- Management of payment obligations and segmentation:
  - Policies to limit, manage, and recover tax arrears contributed positively (Burundi, Comoros, Dominica).
  - Large Taxpayer Units (LTUs) implemented in some countries (Cabo Verde, Uruguay) to manage entities that can account for as much as 80 percent of all tax revenue.
- Use of multiple RA instruments:
  - In more than 30 percent of country-episodes, 4-5 elements of RA were addressed in parallel.
  - Examples: Paraguay (streamlined auditing and collection, improved customs efficiency, anti-corruption measures); Georgia (mass-replacement and training of staff, strengthened legal procedures, upgraded IT, simplified tax returns, improved audits).
- Sustainability caveat:
  - Initial gains from tax policy measures can be eroded without sustained RA reforms (examples: Antigua and Barbuda; The Gambia).

### VI. CONCLUSIONS
- The study assesses large tax revenue increases in LICs and EMs, identifying reform episodes and describing tax policy and RA measures without inferring causality.
- Key common observations:
  - Sustained and large revenue gains are usually accompanied by a comprehensive strategy linking tax policy reforms and revenue administration reforms.
  - Rate increases in indirect taxation and broadening the tax base for both direct and indirect taxation were the most frequently observed tax policy instruments.
    - Rate increases in excises and GST were most common, followed by elimination of exemptions and tax holidays.
    - Simplification (reducing number of taxes or introducing broad-based VAT) occurred in some episodes.
    - Property taxes played a limited role due to legal loopholes and weak RA capacity at local level.
  - Capacity-enhancing RA measures most frequently observed: risk-based audits, management/governance/HR strategies, and IT system enhancements.
- Suggestions for future research using the dataset:
  - Analyze episodes of tax reform failures.
  - Inform discussion on the relative importance of RA versus tax policy changes and measures within categories.
  - Explore differences in revenue mobilization and sustainability between IMF program and non-program countries, and the impact of IMF and other donors’ technical assistance.
  - Analyze drivers of sustainability such as international assistance or growth acceleration.

*Source: wp18234 - 6.9 FS●   ●●-*

### REFERENCES

### REFERENCES

### Major bibliographic themes
- Tax revenue mobilization and administration: multiple working papers and technical notes by Akitoby et al. (2018a, 2018b, 2018c, 2018d), Crandall (2010), Crivelli and Gupta (2014), Kidd and Crandall (2006), Russell (2010), Norregaard (2013).
- VAT and trade-related tax research: Keen (2013), Keen and Ligthart (2002), Keen and Lockwood (2010), Ufier (VAT adoption study).
- State capacity, governance, and fiscal capacity: Besley and Persson (2009, 2010), Gaspar, Jaramillo, and Wingender (2016a, 2016b), IMF policy notes on governance and capacity (2016, 2017, 2018).
- Corruption, compliance, and tax performance: Baum et al. (2017), Carrillo, Pomeranz, and Singhal (2017), von Haldenwang and Ivanyna (2012), Tanzi (1992).
- Regional and empirical studies: Hausmann, Pritchett and Rodrik (2004); Baunsgaard and Keen (2010); Clist and Morrissey (2011); International Monetary Fund (2011, 2015); OECD (2008, 2017).

### Selected working papers and IMF outputs (representative entries)
- Akitoby, Bernardin, Anja Baum, Svetlana Cerovic, and Jiro Honda, 2018a, “Tax Revenue Mobilization in Emerging Market Countries: Lessons from Country Cases,” IMF Working Paper, forthcoming, (Washington: International Monetary Fund).
- Akitoby, Bernardin, Jiro Honda, Hiroaki Miyamoto, Keyra Primus, and Mouhamadou Sy, 2018b, “Case Studies in Tax Revenue Mobilization in Low-Income Countries,” IMF Working Paper, forthcoming, (Washington: International Monetary Fund).
- Gaspar, Vitor, Laura Jaramillo and Philippe Wingender, 2016a, “Political Institutions, State Building, and Tax Capacity: Crossing the Tipping Point,” IMF Working Paper, WP/16/233, (Washington: International Monetary Fund).
- Keen, Michael, 2013, “The Anatomy of the VAT,” IMF Working Paper, WP/13/111, (Washington: International Monetary Fund).

### Empirical and thematic coverage in the references
- Revenue mobilization strategies in LICs and EMs, including country case studies, datasets on large tax revenue mobilizations, and evaluations of IMF program conditionality on revenue reform.
- Revenue administration reforms: performance measurement, taxpayer compliance programs, revenue authorities evaluation, IT and large taxpayer unit roles.
- Interaction of trade liberalization and tax revenue, VAT adoption effects, and the role of non-tax revenue and resource dependence.

### Appendix I — Country Classification (as reported)
- Note: Country classification is reported only for those countries for which tax revenue data are available.
- Low Income Country (LIC): 65Emerging Market (EM): 71Resource Rich: 35Fragile: 31
- Examples (as presented in source lists): AfghanistanMarshall IslandsAlbaniaLebanonAlgeriaAfghanistan; BangladeshMauritaniaAlgeriaMacedonia, FYRAngolaAngola; BeninMicronesiaAngolaMalaysiaBahrainBosnia & Herzegovina; (full country lists follow in source).

### Appendix II — Summary Statistics (selected table extracts, as presented)
- Table AII.1. Summary Statistics, as a percent of GDP, 2000-2015
  - Source: WoRLD.
  - Non-Resource Rich
    - Emerging Markets
      - Tax Revenue73718.718.65.67.039.3
      - PIT6042.52.11.80.08.3
      - CIT6042.92.51.70.09.9
      - G&S6486.46.53.50.019.3
      - Trade6512.71.63.30.024.3
    - LIC
      - Tax Revenue77315.014.46.91.758.1
      - PIT3772.11.61.60.08.7
      - CIT3782.01.71.30.09.0
      - G&S4304.43.92.90.014.5
      - Trade4564.42.84.90.039.9
  - Resource-Rich
    - Emerging Markets
      - Non-Commodity Revenue34112.612.47.20.931.5
      - PIT2011.71.41.40.19.2
      - CIT2785.03.75.00.125.5
      - G&S2824.64.53.50.025.9
      - Trade2911.81.12.20.012.6
    - LIC
      - Non-Commodity Revenue18412.212.34.04.224.9
      - PIT842.01.71.30.15.8
      - CIT852.01.71.30.17.3
      - G&S923.53.11.90.28.1
      - Trade952.92.12.10.27.9

- Table AII.2. Tax Revenue Average by Region (EMs and LICs), 2000-2015, as a percent of GDP
  - Notes: RR (resource rich) countries use noncommodity revenue less grants. Non-RR (non-resource rich) countries use tax revenue. Sources: WEO and WoRLD.
  - RRNon-RRRR
    - Non-RR
      - RRNon-RRRRNon-RRRRNon-RR
      - Tax Revenue11.915.814.514.926.422.48.416.118.517.4
      - PIT2.12.62.01.73.53.41.52.31.41.8
      - CIT4.52.34.33.14.62.23.72.74.52.4
      - G&S3.24.84.73.66.09.44.06.65.45.0
      - Trade3.25.61.62.76.21.11.41.71.14.2
    - Regions indicated: WHD AFR APD EUR MCD (as in source table layout).

### Appendix III — Method of Identification of Episodes (criteria and examples)
- Note: “Above average increase” and “3 years or more subsequent increases” are identified under Strategy 1; “von Haldenwang and Ivanyna” corresponds to the analyzed episodes under Strategy 2.
- Examples of identified episodes (selection from source tables):
  - Afghanistan 2009-2010 √
  - Burkina Faso 2009-2013 √ √
  - Cambodia 2012-2015 √ √ √
  - Gambia, The 2011-2015 √ (-) √ (-) √ (-) √ √
  - Maldives 2011-2015 √ √ √
  - Rwanda 2010-2014 √ √ (-) √ (-) √ √ √ √, √ (-)
  - Many other country-period episodes listed in source tables for Low Income Countries, Low Income Resource-Rich Economies, and Emerging Markets.

### Appendix IV — Tax Policy Measures (categories and coding)
- Tax policy measure categories coded per country-period include:
  - Property, General GST, PIT, CIT, VAT, Excise, Trade Taxes.
- Example entries (as presented):
  - Afghanistan 2009-2010: Property: √
  - Burkina Faso 2009-2013: Property/General GST/PIT/CIT/VAT/Excise/Trade Taxes: √ √ √ √ √ (multiple ticks in source)
  - Multiple country-period rows indicate presence of New Rate, Threshold, and Exemptions flags (√ and √(-) symbols as recorded in source tables).
- Source: IMF staff estimates (as indicated in source).

### Appendix V — Revenue Administration Measures (categories and examples)
- Revenue administration categories (aligned with IMF’s Results-Based Management (RBM) Framework for Revenue Administration):
  - Management, governance & HR
  - Large taxpayers' office & segmentation
  - IT system
  - Registration & filing
  - Audit & verification program
  - Management of payment obligations
  - Improving compliance
  - Customs clearance
- Example coded entries (selection as presented):
  - Afghanistan 2009-2010: Management, governance & HR √; Large taxpayers' office & segmentation √; IT system √; Registration & filing √
  - Burkina Faso 2009-2013: all eight categories marked √√√√√√√√
  - Nepal 2007-2010: Management, governance & HR √√√√√√√√
  - Guyana 2004-2010: Management, governance & HR √√√√√
  - Tables continue with country-period rows for Emerging Markets, Low Income Countries, Low Income Resource-Rich Economies, and Emerging Markets Resource-Rich Economies.
- Note: The revenue administration measures were categorized in line with IMF’s Results-Based Management (RBM) Framework for Revenue Administration. Source: IMF staff estimates.

*Source: wp18234 - REFERENCES (pdf content unit).*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2018/wp18234.pdf_