## itappea ## Source details **Canonical URL:** [itappea](https://www.imf.org/-/media/files/publications/dp/2026/english/itappea.pdf) ## Other formats - [Markdown version](/-/media/files/publications/dp/2026/english/itappea.pdf.md) - [Structured JSON version](/-/media/files/publications/dp/2026/english/itappea.pdf.json) --- ### Executive Summary — Scope, Purpose, and Structure - Presents ISORA 2023 round results (data for FY 2022) and overview of paper structure covering objectives, participation, data availability, high-level findings for 2022, and indices reflecting practices and structural foundations. - Major thematic sections referenced: Introduction; ISORA 2023 Participants; Information Available for Comparative Analysis; High-Level Findings on Tax Administration in 2022; Practices and Structural Foundations; Annexes. - ISORA 2023 collected detailed data from 166 tax administrations worldwide for FY 2022. ### Overview of ISORA and Participation - ISORA is a collaborative project of IMF, ADB, CIAT, IOTA, and OECD to develop a unified standardized database on tax administration covering FYs 2014 to 2023 and over 180 administrations across rounds. - ISORA 2023 participation and coverage: - 166 tax administrations submitted FY 2022 data. - FCSs: 22; SDSs: 31. - MCD participation limited with 11 participants. - Participant FTEs range from fewer than 10 staff to almost 700,000; taxpayers and returns processed vary by five orders of magnitude. ### Information Collected, Data Types, and Availability - Data types collected: - Quantitative (numerical), categorical (including yes/no), and free-text fields. - Annual quantitative data and periodic qualitative data. - Data completeness (FY 2022): - Approximately one-third of quantitative indicators have over 90 percent completeness. - Nearly two-thirds of quantitative indicators exceed 80 percent completeness. - Revenue collections and staff demographics are most complete; payments, arrears, and audits show persistent gaps (roughly one-third of requested data points not provided). - Completeness by thematic group (numerical indicators — percent completeness shown in source): - Revenue: 95 - Resources: 88 - Staff demographics: 94 - LT/HNWI: 86 - Registration: 84 - Filing: 78 - Payment: 66 - Arrears: 67 - Audit: 67 - Tax crimes and disputes: 78 - Incoming service contacts: 65 - Administrations with over 95 percent data availability for FY 2022 include (as listed in source) AEs: Hong Kong; Greece; Ireland; Lithuania; Malta; The Netherlands; Portugal; Slovak Republic; Slovenia; Trinidad and Tobago; Ukraine; EMEs: Albania; Armenia; Barbados; Bulgaria; Eswatini; Kosovo; Maldives; Peru; St. Vincent and the Grenadines; LIDCs: Argentina; Nigeria; Aruba; Sierra Leone; Belize; Uganda; Dominican Republic; Georgia; Malaysia; Mongolia; St. Lucia. - ISORA data and metadata publicly accessible via ISORA data portal (https://isoradata.org); multiple formats and interactive tools available. ### Tax Administration Characteristics and Scope - Institutional arrangements: - Approximately 40 percent of tax administrations operate jointly with customs. - Semiautonomous bodies: AEs 63 percent; EMEs 38 percent; LIDCs 40 percent. - 87 percent collect PIT, CIT, and VAT. - 60 percent collect nontax revenues. - 28 percent responsible for collecting social security contributions. - Revenue collection (median/average by economic group): - AEs: 24 percent of GDP collected by tax administrations. - EMEs: 17 percent of GDP. - LIDCs: 13 percent of GDP. - Tax administrations typically collect around 60 percent of total government revenue across economic groups. - Tailored taxpayer segmentation: - LTO/P operation: about 90 percent of EMEs and LIDCs; 78 percent of AEs. - HNWI programs exist in 28 percent of administrations. - Simplified income tax regimes more prevalent in LIDCs and FCSs (offered in three-quarters of FCSs). ### Resources and Human Capital - Operating expenditure medians (2022, percent of GDP): - All: 0.146 - AEs: 0.154 - EMEs: 0.152 - LIDCs: 0.125 - SDS: 0.186 - FCS: 0.102 - AFR: 0.174 - APD: 0.107 - EUR: 0.168 - MCD: 0.099 - WHD: 0.137 - Salary expenditure (2022, percent of operating expenditure): - All: 70 - AEs: 72 - EMEs: 71 - LIDCs: 66 - SDS: 74 - FCS: 67 - ICT expenditure: higher for AEs; ICT expenditure medians show AEs invest more both relative to GDP and operational costs. - FTEs and staff ratios: - Annex medians: FTEs per 1,000 labor force — All 0.9; AEs 1.6; EMEs 1.1; LIDCs 0.3. - Core tax returns per FTE medians: All 731.5; AEs 1614.3; EMEs 659.2; LIDCs 168.9. ### Core Performance Indicators — Registration, Filing, Payment, Prefilling, and Arrears - Active taxpayers and registration: - Active PIT taxpayers as a percentage of labor force: median for AEs roughly a hundred times greater than the median for LIDCs. - Active taxpayers increased between 2018 and 2022 across most groups and tax types except a marginal decline in PIT for LIDCs. - On-time filing (TADAT-defined) and payments: - Percent of returns received on time (All ISORA participants — medians): - PIT: 65.3 - CIT: 58.5 - VAT: 76.1 - Employer Withholding: 68.1 - Percent of payments received on time (All ISORA participants — medians): - PIT: 69.0 - CIT: 81.0 - VAT: 81.2 - Employer Withholding: 81.5 - AEs report substantially higher on-time filing and payment rates than EMEs or LIDCs (example AEs on-time filing: PIT 91.7; CIT 81.4; VAT 87.5). - Electronic channels: - E-filing availability counts (All ISORA participants): - PIT: 134 of 159 - CIT: 137 of 160 - VAT: 132 of 148 - PAYE: 133 of 155 - E-filing rates (All ISORA participants, percent): PIT 85.6; CIT 88.3; VAT 90.5; PAYE 89.8 - E-payment availability (PIT as percent of participants in group) — 75.5 - Over 60 percent of LIDC administrations require e-filing and e-payment for most taxpayers; LIDCs have rapidly increased e-filing and e-payment take-up with more than half of returns filed electronically and more than three-quarters of payments received electronically in FY 2022 on average for LIDCs. - Prefilling PIT returns: - AEs: about 90 percent prefilling using third-party data in both 2018 and 2023. - EMEs: prefilling increased to about 50 percent. - LIDCs: just over one-seventh (13 percent) use any third-party data for prefilling in FY 2022. - Arrears: - Arrears as percent of GDP: around 2 percent in AEs and LIDCs, around 5 percent in EMEs (summary in Executive Summary). - Arrears as percent of collections (median values from source panels): 2.9, 1.9, 5.6, 1.5, 3.3, 2.4, 6.2, 2.0, 2.9, 1.9, 5.0, 2.3 (preserved exactly as shown in source). - Arrears by tax type median values preserved exactly in source: 37, 36, 16, 23, 11, 13, 25, 22, 7, 12, 3, 5, 10, 9, 13, 11 (panel values). - SOE contribution to arrears (percent medians preserved exactly): 8, 4, 12, 5, 12, 5, 19, 10, 9, 4, 14, 11 (panel sequence). - Collectible arrears panel numbers preserved exactly: 7, 4, 8, 0.3, 0.8, 0.4, 4, 3, 4, 19, 15, 22. ### Audit and Verification - Pre-audit compliance interventions: - Around 80 percent of administrations undertake compliance interventions prior to resource-intensive audits. - Over 80 percent use third-party data to identify inconsistencies; 80 percent conduct risk reviews. - Audit coverage and types: - Over 90 percent undertake comprehensive audits and desk audits; limited-scope and single-issue audits slightly lower. - Verification during filing: over 80 percent of AEs undertake verification during filing; less than half of EMEs and LIDCs do so. - Audit performance metrics (values preserved exactly): - Audits and verifications undertaken (percentage sequence from source): 53 84 71 80 83 92 85 86 86 92 80 - Audit hit rate (percent average sequence): 69, 52, 71, 78, 67, 51, 69, 79 - Assessments raised through audit (percent of GDP, median sequence): 0.51, 0.45, 0.60, 0.31, 0.32, 0.27, 0.33, 0.39 - CIT assessments raised through audit (percent of CIT collected, median sequence): 3.1, 2.4, 3.8, 5.0, 2.4, 1.1, 2.4, 4.3 ### Efficiency Measures and Observations - Three efficiency measures used: - Operational cost as percent of revenue collected (cost of collection). - Active taxpayers per FTE. - Returns received per FTE. - Key findings: - Efficiency highest in AEs, followed by EMEs, then LIDCs. - SDSs: higher operational costs relative to revenue collected than LIDCs, but better on FTE-related measures than LIDCs. - FCSs: lower cost of collection than LIDCs due to low budgets, but lower efficiency on FTE-related measures. - Regional pattern: AFR lowest efficiency; EUR highest efficiency on FTE-related measures. - Numeric indicators preserved exactly from source panels: - Operational Cost as Percent of Revenue Collected (FY 2018 and FY 2022 median sequence): 0.9, 0.7, 0.9, 1.4, 1.4, 2.7, 1.3, 0.7, 0.7, 0.7, 1.0, 0.76, 0.62, 0.77, 1.10, 1.40, 0.82, 1.70, 0.68, 0.63, 0.62, 0.83 - Active Taxpayers per FTE (FY 2018 and FY 2022 median sequences): 512, 1,198, 382, 161, 369, 113, 152, 408, 806, 214, 594 (FY18) and 732, 1,627, 665, 169, 266, 106, 120, 433, 1,518, 458, 743 (FY22). ### Practices and Structural Foundations — Seven FY 2022 Indices - Seven thematic indices (scores range 0 to 100) constructed from ISORA FY 2022 responses synthesizing over 200 survey questions: - Compliance management (CM) - Digitalization (DG) - Governance (GV) - Human resources management and development (HRMD) - Management autonomy (MA) - Public trust (PT) - Taxpayer services (TS) - Distribution highlights (exact values preserved where given): - Public trust mean score: 47 - Management autonomy mean score: 72 - Dispersion widest for HRMD, MA, and PT; narrowest for TS and CM. - Pairwise correlations among indices (preserved exactly): - CM–DG: 0.73; CM–GV: 0.69; CM–HR: 0.58; CM–MA: 0.46; CM–PT: 0.65; CM–TS: 0.69 - DG–GV: 0.68; DG–HR: 0.59; DG–MA: 0.51; DG–PT: 0.65; DG–TS: 0.77 - GV–HR: 0.84; GV–MA: 0.83; GV–PT: 0.73; GV–TS: 0.83 - HR–MA: 0.76; HR–PT: 0.60; HR–TS: 0.68 - MA–PT: 0.48; MA–TS: 0.56 - PT–TS: 0.71 - Sample sizes for indices: - AEs: 37; EMEs: 83; LIDCs: 46; FCSs: 22; SDSs: 31. - Group comparisons: - AEs outperform EMEs and LIDCs across all seven indices; AE averages exceed LIDC/EME averages by at least 10 percentage points for CM and by 18 percentage points for the overall average. - Largest AE–LIDC average score gaps exceed 20 percentage points for PT, HRMD, and DG. - LIDC and EME averages are within 4 percentage points of each other for every index. - Index drivers (components with largest AE–LIDC gaps include): - Availability and management of third-party data (Social Security Agency, property ownership, sales). - Prefilling, data-science/analytics tools, automated verifications. - Publication of annual reports, plans, taxpayer satisfaction surveys, tax gap analyses. - Policies for flexible working arrangements, staff engagement and time-reporting systems. - Specialized audit staff and international tax capabilities; APAs. - Association of index scores with administration size: - Scores against all practice indices increase with FTE size; correlations by economic group preserved exactly for CM, DG, GV, HRMD, PT, MA, TS (see Box 3.1 correlations): - AEs: CM 0.23; DG 0.07; GV 0.42; HRMD 0.54; PT 0.21; MA 0.25; TS 0.42 - LIDCs: CM 0.44; DG 0.45; GV 0.38; HRMD 0.17; PT 0.20; MA 0.22; TS 0.46 - EMEs: CM 0.63; DG 0.81; GV 0.68; HRMD 0.61; PT 0.52; MA 0.64; TS 0.75 ### Associations between Indices, Performance, Resources, and Efficiency - Operational correlations (selected Pearson coefficients preserved exactly from Table 6): - Active PIT taxpayers as percentage of Population: DG 0.41**; Ave 0.31** - CIT returns filed on time as percent of expected returns: DG 0.50**; TS 0.43**; Ave 0.47** - Percent PIT returns e-filed: DG 0.73**; TS 0.69**; Ave 0.66** - Percent electronic payments by number: DG 0.59**; TS 0.56**; Ave 0.60** - Closing stock of arrears at year end as percentage of revenue collected: GV −0.30**; TS −0.30**; Ave −0.29** - CIT arrears as percent of CIT collected: CM −0.38**; GV −0.44**; PT −0.44**; TS −0.46**; Ave −0.45** - Audit hit rate: DG −0.28**; TS −0.26**; Ave −0.23** - Resource and efficiency correlations (selected from Table 7, preserved exactly): - Operating cost as percent of GDP: DG −0.23**; CM −0.22*; TS −0.18*; Ave −0.14* - Operating cost as percent of collections: DG −0.43**; CM −0.33**; TS −0.34**; Ave −0.33** - Percent staff with a degree: TS 0.50**; DG 0.43**; CM 0.41**; Ave 0.44** - Percent staff with 20 or more years of service: DG 0.43**; Ave 0.29** - Log active taxpayers per FTE: DG 0.40**; PT 0.40**; Ave 0.39** - Log returns received per FTE: DG 0.60**; TS 0.54**; Ave 0.55** - Revenue, tax effort, VAT gap, and public administration correlations (selected from Table 8, preserved exactly): - Revenue collected as percent of GDP: DG 0.44**; PT 0.33**; TS 0.33**; Ave 0.35** - Revenue collected as percent of government revenue: DG 0.45**; TS 0.41**; Ave 0.42** - Tax effort (113 administrations): DG 0.36**; HRMD 0.45**; Ave 0.41** - VAT compliance gap (37 administrations): DG −0.32*; Ave −0.21 - Blavatnik Public Administration Index (110 countries): DG 0.59**; TS 0.52**; Ave 0.57** - Government Effectiveness (159 countries): HRMD 0.39**; DG 0.34**; Ave 0.32** ### Key Conclusions and Suggested Areas for Further Analysis - ISORA 2023 provides a rich, globally representative dataset on tax administration performance, practices, and structural foundations for FY 2022. - Principal highlights: - Clear disparities across economic groups (AEs outperform EMEs and LIDCs across indices and performance metrics). - Persistent data availability gaps for critical functions (payments, arrears, audit) despite improvements in some areas. - Strong positive roles of digitalization (DG), taxpayer services (TS), public trust (PT), and governance (GV) in associations with performance, efficiency, and revenue collection. - Administration size (FTEs) is correlated with higher index scores and wider adoption of practices. - Suggested further analyses (as flagged in source): - Investigate high levels of VAT noncompliance (and arrears). - Examine methods for sustaining taxpayer compliance beyond audit. - Explore relationship between resources and tax administration effectiveness, particularly among FCSs. - Identify practices that promote effectiveness in small tax administrations. *Source: Executive Summary and chapters/excerpts from ISORA 2023 departmental paper (ISORA data for FY 2022) as presented in the provided content.* ### Executive Summary................................................................................................vi ### Executive Summary ### Scope and Purpose - Presents an overview of the International Survey on Revenue Administration (ISORA) 2023 round and the structure of the paper covering ISORA 2023 results. - Key organizational topics referenced in the Executive Summary align with the paper’s broader structure, including objectives, participation, data availability, high-level findings for 2022, and indices representing practices and structural foundations. ### Major Thematic Sections Covered in the Paper (as reflected in the Executive Summary) - Introduction - A. Objectives and Nature of International Survey on Revenue Administration (ISORA) - B. The 2023 ISORA Round - C. Purpose and Structure of This Paper - ISORA 2023 Participants - A. Participation by Economic Grouping and Region - B. Participation by Other Commonly Used Grouping - C. ISORA Participants by Size and Scale of Operations - Information Available for Comparative Analysis - A. Information Collected and Its Use in Comparative Analysis - B. Availability of Quantitative Data - C. Accessing and Using ISORA Data - High-Level Findings on Tax Administration in 2022 - A. Broad Characteristics of Tax Administrations - B. Tax Administration Resources - C. Core Tax Administration Performance Indicators - D. Efficiency Measures - Practices and Structural Foundations for Effective Tax Administration - A. The Use of Indices to Reflect Practices and Structural Foundations - B. ISORA Indices Based on FY 2022 Data - C. Scores against the FY 2022 Indices for Groupings of Interest - D. Administrative Practices and Institutional Foundations—Exploring Associations with Performance and Resources - Annexes (listing of annex topics) - Annex 1 through Annex 7 covering participants, quantitative indicators, data availability rates, accessing ISORA data and metadata, selected comparative aggregate statistics, indices using ISORA data, and index structure/examples - Boxes, Figures, and Tables cited as part of the Executive Summary and paper ### Types of Analytic Content Referenced - Comparative analysis across economic groupings, IMF regions, and other groupings. - Quantitative indicator availability and completeness, with graphic summaries (e.g., Figure 7. Completeness of Numerical Inputs by Thematic Grouping of Indicators). - High-level comparisons of institutional arrangements across years (e.g., Figure 9. High-Level Comparison of Institutional Arrangements (2017 and 2022)). - Resource and operational metrics including: - Operational expenditures (Figure 14. Operational Expenditures of Tax Administrations) - Human resources and staff allocation (Figure 15. Tax Administration Human Resources (2018 and 2022); Figure 16. Allocation of Staff by Function (2018 and 2022)) - Measures of active taxpayers, filings, payments, arrears, audit and verification activity (Figures 17–24) - Efficiency measures and their distribution (Figure 25. Tax Administration Efficiency Measures; Figure 26. Distribution of Scores of the FY 2022 Indices) - Construction and use of seven FY 2022 indices to reflect practices and structural foundations, including: - Distribution and average scores against the seven indices for all participants and subgroup comparisons (Figures 26–29) - Correlations among indices and between indices and operational/resource measures (Table 4; Tables 6–8) - Examination of drivers of differences across economic groups (Table 5; Box material referencing index-size relationships) ### Enumerated Illustrative Lists and Visuals Included in the Executive Summary Context - Boxes highlighted include: - Box 1. Consistency between ISORA and TADAT Quantitative Measures - Box 2. Variation across Large Taxpayer Offices/Programs (LTO/Ps) - Box 3. Index Scores and Size of Tax Administrations - Selected figures include (by exact captions): - Figure 1. ISORA’s Interaction with Key Organizational Management Practices of Tax Administration - Figure 2. Participation in ISORA by Economic Group (2014 to 2022) - Figure 3. Participation in ISORA by the IMF Region (2014 to 2022) - Figure 4. ISORA Participation of Tax Administrations in FCS and SDS - Figure 5. Tax Administration FTEs and the Population Served (2022) - Figure 6. Total Active Taxpayers and Returns Received (2022) - (Additional figures 7 through 29 referenced in the paper) - Selected tables include (by exact captions): - Table 1. High-Level Picture of Tax Administration Data Collected through ISORA 2023 - Table 2. Participation in ISORA by the IMF Economic Grouping and the IMF Region (2022) - Table 3. Administrations with Data Availability Rates of over 95 Percent for FY 2022 - (Additional tables 4 through 8 referenced in the paper) ### Structure and Data Access - The Executive Summary situates the reader to the detailed annexes that include participant lists, quantitative indicators, data availability rates, guidance on accessing ISORA data and metadata, selected aggregate statistics, and index construction and examples. - References to metadata and instructions for accessing the ISORA dataset are part of the paper’s annex material. *Executive Summary — itappea - Executive Summary (IMF DP: ISORA 2023)* ### Executive Summary ### Executive Summary ### Overview of ISORA - ISORA is a collaborative project of the International Monetary Fund (IMF), the Asian Development Bank, the Inter-American Center of Tax Administrations, the Intra-European Organisation of Tax Administrations, and the Organisation for Economic Co-operation and Development to develop a unified, standardized database on tax administration to support comparative analysis and capacity development. - The comparative database contains data collected annually starting 2014 and includes data from over 180 tax administrations that participated in at least one round of the survey. ### ISORA 2023 and Participants - ISORA 2023 collected detailed data from 166 tax administrations worldwide for the fiscal year (FY) 2022. - The 2023 round incorporated both annual quantitative data and periodic qualitative data on practices and structures. - Participation coverage: broad across advanced economies (AEs), emerging market economies (EMEs), low-income developing countries (LIDCs), IMF regions (except limited participation in Middle East and Central Asia Department (MCD) with 11 participants), fragile and conflict-affected states (FCSs: 22), and small developing states (SDSs: 31). - Participant administrations vary widely in size (FTEs range from fewer than 10 staff to almost 700,000) and operational scale (numbers of taxpayers and tax returns processed range by five orders of magnitude). ### Information Available - Data Types: - ISORA collects three main types of data: (1) quantitative (numerical), (2) categorical data (including yes/no questions), and (3) free-text fields. - Quantitative data primarily collected annually; qualitative data collected periodically. - Data Quality: - Approximately one-third of quantitative indicators have over 90 percent data completeness for 2022, and nearly two-thirds exceed 80 percent. - Revenue collections and staff demographics are the most complete data; payments, arrears, and audits show persistent gaps (roughly one-third of the data points requested are not provided). - Quantitative data availability varies by economic grouping: AEs generally provide more complete data than EMEs and LIDCs; EMEs provide more complete data on taxpayer registration and segmentation than AEs. - Administrations with over 95 percent data availability span all economic groups and regions and include SDSs. - Accessing ISORA Data: - ISORA data and metadata are publicly accessible through the ISORA data portal (https://isoradata.org), with multiple data formats and interactive tools. Questionnaires, guides, and publications related to ISORA are also available. ### Tax Administration Characteristics - Institutional Arrangements and Scope: - Approximately 40 percent of tax administrations operate jointly with customs. - Semiautonomous bodies: AEs 63 percent; EMEs 38 percent; LIDCs 40 percent. - 87 percent collect the three main taxes: personal income tax (PIT), corporate income tax (CIT), and value-added tax (VAT). - 60 percent collect nontax revenues. - 28 percent are responsible for collecting social security contributions. - Revenue Collection: - Average revenue collected by tax administrations: AEs 24 percent of GDP; EMEs 17 percent of GDP; LIDCs 13 percent of GDP. - Tax administrations typically collect around 60 percent of total government revenue across all economic groups. - Tailored Responses to Taxpayer Segments: - Large taxpayer office/program (LTO/P) operation: about 90 percent of EMEs and LIDCs; 78 percent of AEs. - High-net-wealth individual programs exist in 28 percent of administrations. - Simplified income tax regimes are more prevalent in LIDCs and fragile states. - Financial and Human Resources: - Operational expenditure relative to GDP is about 30 percent higher in AEs than in LIDCs; FCS generally allocate even less than LIDCs. - Salaries constitute about 66 percent to 74 percent of operational costs for all economic groups. - ICT expenditure is higher among AEs. - AEs have roughly 10 times the FTEs per population compared with LIDCs. - Most staff allocated to taxpayer services for all groups; AEs dedicate more staff to audit and verification than LIDCs. ### Tax Administration Performance — Core Functions - Registration: - Active taxpayers as a percent of taxpayers on the register have increased among LIDCs for all taxes except PIT, but still below AEs for PIT and CIT. - Active PIT taxpayers as a percentage of the labor force: median for AEs is a hundred times greater than the median for LIDCs. - Across all taxes except VAT, active taxpayer growth over five years outstripped labor market growth. - On-time filing and payment: - On-time filing rates highest for VAT and PIT; significantly better in AEs than in EMEs and LIDCs. - PIT and CIT registrations have grown in LIDCs while on-time filing rates have declined. - On-time payment rates follow similar patterns; fewer administrations report payment data. - Electronic filing and payment: - More than 60 percent of LIDC tax administrations have made e-filing and e-payment mandatory for most taxpayers. - Average LIDC e-filing and e-payment rates have grown rapidly over 2018–23, approaching similar e-filing levels to EMEs and exceeding them in e-payment. - E-filing and e-payment rates in LIDCs and EMEs still lag AEs, which increased more slowly by about 5 percentage points over the same period. - Prefilling PIT returns: - Prefilling using third-party data remains common in AEs (about 90 percent in both 2018 and 2023). - Among EMEs, prefilling increased to be used by half of administrations. - LIDCs: just over one-seventh use any third-party data for prefilling. - Arrears: - Arrears constitute around 2 percent of GDP in AEs and LIDCs, and around 5 percent of GDP in EMEs. - Arrears as reported are highest for CIT and VAT as a percent of the respective collections. - Arrears for PIT and PAYE are significantly lower for AEs than in EMEs and LIDCs. - Arrears associated with state-owned enterprises constituted 22 percent of the stock of tax arrears among LIDCs. - Audit and verification: - Over 80 percent of administrations undertake compliance interventions prior to audit. - Almost all administrations reported undertaking one or more types of audits. - Most AEs undertake verifications during return filing; less than half of EMEs and LIDCs do so. - AEs are more likely to use third-party data or data analytics in verification before triggering an audit. - Audit hit rates and assessments raised are lowest in AEs; CIT assessments raised through audit are larger than those for other tax types across all economic groups. ### Efficiency - Three efficiency measures: (1) operational cost as a percentage of revenue collected (cost of collection), (2) active taxpayers per FTE, and (3) returns received per FTE — show higher efficiency in AEs, followed by EMEs, with LIDCs the lowest. - SDSs exhibit higher operational costs relative to revenue collected than LIDCs, but perform better than LIDCs on the two FTE-related measures. - FCSs have a lower cost of collection than LIDCs due to low budgets, but lower efficiency on FTE-related measures. - By IMF region: - African Department tax administrations exhibit the lowest efficiency. - European Department (EUR) administrations exhibit the highest efficiencies using the two FTE-related measures, followed by Western Hemisphere Department. - EUR’s cost of collection is close to the smallest cost of collection seen in the MCD region. - All three measures indicate efficiency gains across all economic groups between 2018 and 2022. ### Tax Administration Practices and Structural Foundations - Analysis Using Indices: - Seven thematic indices constructed from ISORA 2023 responses: (1) compliance management (CM), (2) digitalization (DG), (3) governance (GV), (4) human resources management and development (HRMD), (5) management autonomy (MA), (6) public trust (PT), and (7) taxpayer services (TS). - Indices synthesize responses to over 200 survey questions. - Dispersion in index scores widest for HRMD, MA, and PT. Lowest variation for TS and CM. - Strong positive correlations exist among all indices. - Group Comparisons: - AEs outperform EMEs and LIDCs across all seven indices, especially in DG, HRMD, and PT. - Key drivers of differences include greater use of data in core processes, publication of plans and performance, flexible working arrangements, and monitoring of employee activity and engagement in AEs. - FCSs and SDSs score lower than LIDCs on all indices; FCS scores notably low in digitalization and management autonomy. - Within each economic grouping, index scores correlate positively with administration size. - Regional differences are less pronounced than economic group differences: EUR scores highest; African Department, APD, and WHD cluster lower; MCD scores intermediate (limited representativeness due to participation). - Associations with Performance, Resources, and Efficiency: - Indices correlate positively with performance and operational metrics: PIT taxpayers as a percent of population; TADAT on-time filing measures for PIT, CIT, VAT, and employer withholding; e-filing and e-payment rates. - Indices correlate negatively with arrears measures. - Strongest correlations generally for DG, TS, GV, and PT. - Indices correlate positively with percentage of staff with over 10 or 20 years experience and staff with a degree (strongest associations with DG and TS respectively). - Proportion of staff allocated to audit and verification correlates positively with all seven indices (most strongly with DG). - Cost of collection correlates negatively with all indices (stronger correlation than between indices and operating cost as percent of GDP). - Active taxpayers per FTE and returns per FTE correlate positively with all indices (most strongly with DG). - Ratios of revenue collection to GDP and to government revenue correlate positively with all indices (most strongly with DG, TS, and PT). - All indices correlate positively with tax effort from a stochastic frontier model and negatively with the VAT compliance gap using IMF’s RA-GAP methodology. ### Conclusion - ISORA 2023 provides a rich, globally representative data set on tax administration performance, practices, and structural foundations. - Key highlights: - Disparities across economic groups. - Persistent challenges in data availability for critical functions such as audit and arrears management. - Positive role of digitalization, taxpayer services, public trust, and governance in tax administration outcomes. - Suggested areas for further analysis: - High levels of VAT noncompliance (and arrears). - Sustaining taxpayer compliance through measures other than audit. - Relationship between resources and tax administration effectiveness, particularly among FCSs. - Practices that promote effectiveness of small tax administrations. - The thematic indices offer tools for benchmarking and guiding capacity development efforts worldwide. *Source: Executive Summary, ISORA 2023.* ### 1. Introduction ### 1. Introduction ### A. Objectives and Nature of International Survey on Revenue Administration (ISORA) - ISORA provides a worldwide comparative database on revenue administration performance, profile, structure, and practice. - Data by fiscal year (FY) are collected from tax administrations through a questionnaire developed and administered collaboratively by five organizations: the Asian Development Bank (ADB), the Inter-American Center of Tax Administrations (CIAT), the International Monetary Fund (IMF), the Intra-European Organisation of Tax Administrations (IOTA), and the Organisation for Economic Co-operation and Development (OECD). - The IMF’s Fiscal Affairs Department began developing surveys of tax and customs administration practices and performance prior to 2010 to enhance country capacity to monitor revenue administrations and to support capacity development (CD) delivery. - Challenges motivating ISORA included poor data management practices and poor-quality data in many low-income developing countries (LIDCs) and small developing states (SDSs). - Establishment of the ISORA partnership achieved: - elimination of duplication of effort among partner organizations; - reduced compliance burden for tax administrations participating in multiple surveys; - creation of a single, standardized database on tax administration with wider coverage. - A unified database enables: - comparison between and among individual tax administrations and groups by economic characteristics, region, and scale; - identification of trends in practices and performance through FY-level data. - Data have been collected for 10 FYs (2014 to 2023). - Tax administrations participate voluntarily and engage directly in all survey phases: - staff access an online data collection platform developed and administered by the IMF, with built-in consistency and completeness checks; - participating administrations are offered training on ISORA concepts, definitions, and database use; - administrations review data they provided prior to finalization for publication. - The data collection process is linked to organizational management practices: planning, performance monitoring and evaluation, and data management for evidence-based decisions. - Footnote detail: The reference period of the survey is the tax administration’s fiscal year (FY). The FY may differ from the calendar year. Of 159 participants in ISORA 2018, the FY coincided with the calendar year for 119 participants, that is, the FY differed from the calendar year for a quarter of tax administrations completing ISORA. ### B. The 2023 ISORA Round - ISORA 2023 collected data from tax administrations for FY 2022; data were more extensive than data collected for 2018 to 2021. - The 2023 questionnaire included: - annual questions expected to change year over year (for example, resources, tax returns filed, and audits undertaken); - a set of questions covering tax administration practices and structures (periodic questions). - Initial ISORA rounds were biennial (collecting data for two FYs in 2016 and 2018), but this imposed a large response burden. - To improve data quality and reduce response burden, ISORA shifted to: - annual collection of a more streamlined set of data; - supplemented by four- or five-yearly collection of “periodic” data points (generally not expected to change each year, e.g., institutional arrangements, human resources management [HRM] approaches). - Data collections for FY 2018 to 2021 focused on the annual data points only. - ISORA 2023 topics (high-level): - Annually Collected Data: Roughly 90 percent numerical questions - A Revenue roles and segmentation features - B Revenue collected - C Resources: expenditure and people (full-time equivalents) - D Staff demographics - E Taxpayer segments - F Operational metrics - G Taxpayer and other stakeholder interactions - Periodically Collected Data: Roughly 95 percent categorical or yes/no questions - C-P Institutional arrangements, governance and management practices - C-P Human resources management - E-P Segmentation and compliance management - F-P Tax operations - G-P Engagement with stakeholders: taxpayers, information providers, and others - The periodic data points complement largely numerical annual time-series data (for example, staff demographics collected annually may be examined with HRM approaches collected periodically). - The questionnaires and associated guides for all rounds of ISORA are available through the ISORA data portal. ### C. Purpose and Structure of This Paper - The IMF periodically publishes analytical papers highlighting findings from ISORA data. - This paper focuses on ISORA 2023 data provided by all 166 participants in the survey for FY 2022. - Purpose of the paper: 1. provide information on survey participants and data collected to facilitate interpretation and analysis of ISORA 2023 data; 2. share selected findings from the survey; 3. present exploratory analysis of relationships between tax administrations’ practices, structures and resources, operations, and performance. - The underlying aim is to encourage the use and further analysis of the ISORA comparative database. - Use of the ISORA database in analyses: - The IMF generally uses the entire ISORA database for its analyses (the vast majority of ISORA participants are IMF members or members of IMF regional capacity development centers). - ISORA partner organizations generally use data pertaining to their members in their analyses (OECD publications cover Forum on Tax Administration participants; ADB, CIAT, and IOTA publish analyses focused on their membership). - The paper continues themes from previous IMF ISORA publications, including: - performance-related data; - profile data (institutional arrangements, scope, powers); - data on practices and structural foundations (including indices developed for analytical purposes). - Paper structure (chapter overview): - Chapter 1: overview of the survey and this paper. - Chapter 2: information on ISORA participants by economic grouping, IMF region, fragile and conflict-affected states (FCSs), SDSs, and size and scale of operations. - Chapter 3: discusses information available in the survey, including a novel analysis of the availability of quantitative measures. - Chapter 4: highlights from the most recent survey and trends in key performance areas and tax administration characteristics. - Chapter 5: findings on tax administration practices and structural foundations, using indices constructed from ISORA 2023 data. - Throughout the publication, the source of data is ISORA, as published in July 2024. All charts are the work of the authors, based on their analysis. *Source: ISORA, as published in July 2024.* ### 0.1 to 1 ### 0.1 to 1 ### Tax Returns Received in 2022 - ISORA collects the number of returns received by tax type. - Returns received in FY 2022 range widely across administrations and exceed the number of active taxpayers processed. - Survey coverage for taxpayers on register and active taxpayers in the FY includes: personal income tax (PIT), corporate income tax (CIT), value-added tax (VAT), domestic excise, and employers who make returns on taxes withheld on PIT. - Not all administrations provide responses for all tax types they administer, so totals of taxpayers and returns summed across tax types may be understated for some administrations. - The 2023 ISORA survey obtained more accurate and complete figures for returns received than in previous rounds due to a change in the way the question was asked and validated. - The 2023 survey covers returns for four tax types: PIT, CIT, VAT, and employer withholding. - Taxpayer totals shown for 34 AEs, 76 EMEs, and 38 LIDCs, after dropping administrations that did not provide data for two or more taxes they administer. - Return totals shown for 37 AEs, 80 EMEs, and 37 LIDCs, after dropping administrations that did not provide data for two or more taxes they administer. - Taxpayer figures summed across available data for five tax types. Return figures summed across available data for four tax types. ### Information Collected and Its Use in Comparative Analysis - ISORA data categories: - Characteristics of the administration. - Data on practices and structural foundations. - Performance-related data. - Response types collected: - Quantitative—numbers; responses in thousands of local currency units; percentages. - Yes/No—choice between “yes” and “no” or checkbox selections where blank means “no”. - Categorical—choice between two or more options. - Free-text fields—used infrequently and generally not placed in the public domain. - Frequency: - Most quantitative data points are collected annually. - Most data on practices and structural foundations are categorical or yes/no and collected periodically. - Questionnaire administration: - All administrations are required to answer all survey questions applicable to their scope and practices. - A symbol is used to indicate no data available for numerical questions, distinguishing it from missed questions. - Only fully completed surveys are included in the published ISORA data set. - FY 2022 data span both annually and periodically collected data, providing a rich snapshot of tax administration for the period. - Quantitative measures expected to change annually include staff levels, revenue collections, registered taxpayers, and audits completed. - Comparative analysis approach: - Construct dimensionless ratios (for example, taxpayers as a percentage of population or labor force; monetary stocks as percentage of revenue collections or GDP). - Construct thematic indices for many yes/no data points (example: index for digitalization of core operations and services). ### Using Quantitative Indicators in Comparative Analysis - ISORA collects annual quantitative indicators covering: - Revenue collections. - Resources available to the tax administration. - Resource allocation. - Staff demographics. - Large taxpayer offices or programs. - Operational metrics. - Comparisons must account for differences in administration size: staffing levels and taxpayers managed range over five orders of magnitude among contributors to the ISORA FY 2022 database. - Annex 2 presents examples of quantitative indicators available from ISORA data for FY 2022, often used to develop quantitative performance measures and benchmarks. ### Consistency between ISORA and TADAT Quantitative Measures (Box 1) - ISORA data are publicly available; many TADAT reports are not public. Aggregate statistics on TADAT indicator scores and distributions come from a publication reviewing a decade of TADAT assessments (Acedo and others 2024). - ISORA data values can be translated to TADAT scores using the 2019 TADAT Field Guide (TADAT 2019) to create distributions for benchmarking. - On-Time Filing (TADAT Indicator P4–12): - On-time filing rates determined in ISORA for CIT, PIT, employer withholding, and VAT using administrations’ responses on expected returns and returns received on time. - Key points: - On-time filing rates are generally better for PIT and VAT than for CIT and employer withholding, for both ISORA-based TADAT scores and TADAT assessments. - ISORA-derived distributions show a slightly higher proportion of A- and B-scores, consistent with a higher proportion of AEs in ISORA data than in the TADAT assessment population. - E-Filing (TADAT Indicator P4–14): - E-filing rates for CIT, PIT, employer withholding, and VAT calculated using FY 2022 ISORA data on returns received and return channel. - Key points: - The ISORA-based distribution for FY 2022 is close to the published TADAT distribution. - A higher proportion of scores above D in ISORA-derived distributions is consistent with a higher proportion of AEs in ISORA than in the TADAT population. - A second factor shifting ISORA-based distributions toward higher scores is that all ISORA data points are for FY 2022, whereas TADAT distributions include assessments completed before 2019; growth in e-filing is observed in ISORA time-series data and first vs. repeat TADAT assessments. - On-Time Payment by Value (TADAT Indicator P5–17): - On-time payment rates by value for FY 2022 are determined in ISORA for CIT, PIT, employer withholding, and VAT from administrations’ responses on value of payments due and value received on time. - TADAT uses payment data for VAT as a proxy for on-time payment for all core taxes; ISORA-derived distributions use the equivalent of TADAT scoring criteria applied to VAT. - Key points: - The distribution for VAT based on ISORA data for 2022 resembles the published TADAT distribution, with over 50 percent D-values. - The higher proportion of A- and B-scores in ISORA-derived VAT distributions is consistent with a higher proportion of AEs in ISORA data than in the TADAT assessment population. - As in TADAT, on-time payment rates (for non-D-scores) tend to be higher than on-time filing rates in ISORA. ### Availability of Quantitative Data - Completing ISORA requires administrations to provide responses to all applicable questions; numerical questions can be marked with a symbol to indicate data are unavailable (distinct from zero or not applicable). - Some questions may not be applicable (for example, administrations that collect fewer tax types or do not have a large taxpayer office). - It is possible to examine completeness of responses to numerical questions by question and by administration. - Factors affecting data availability include: - Administrations that do not use particular measures or are unable to provide particular measures (example: VAT refunds). - Highly decentralized administrations (example: China and Germany) may find organization-wide statistics challenging to provide because data are managed regionally. - Legal or policy constraints where administrations are not permitted to provide data unless already published elsewhere. - A review of tax administration performance based on TADAT assessments (Acedo and others 2024) points to weaknesses in performance monitoring and poor ratings where numeric indicators are lacking. - Provision of consistent numerical data over time suggests data management practices that support effective performance management. - Examining data availability by indicator is useful to: - Inform users of comparative statistics from ISORA. - Identify topics where measurement can be improved. - Show areas of relative strength and weakness in administrations’ data management systems and track progress over time. - ISORA 2023 includes 150 questions requiring numerical responses; completeness may be expressed as a percentage of total numerical responses and “no-data” responses. - Some administrations may mistakenly use zero instead of the no-data symbol; analysis accounts for “mistaken zeros” and applicability of indicators. *Source: Authors’ compilation; IMF DEPARTMENTAL PAPERS • ISORA 2023.* ### 1. Response Completeness for All ### 1. Response Completeness for All ### Numerical Questions — Availability of Data by Thematic Grouping - Percent completeness (numerical indicators) by thematic group (visual ranges indicated in source): - 95 and above - 90 to 95 - 80 to 90 - 67 to 80 - 50 to 67 - Below 50 - Specific percent completeness values shown in source figure (numerical labels preserved as presented): - Revenue: 95 - Resources: 88 - Staff demographics: 94 - LT/HNWI (large taxpayers/high-net-wealth individual): 86 - Registration: 84 - Filing: 78 - Payment: 66 - Arrears: 67 - Audit: 67 - Tax crimes and disputes: 78 - Incoming service contacts: 65 - Key points: - There is a wide range in data availability by indicator, ranging from 100 percent to less than 50 percent for 3 of 150 numerical indicators. - One-third of the quantitative indicators are 90 percent or more complete, whereas almost two-thirds are over 80 percent complete. - Caution is warranted in interpreting results for indicators with many missing numerical responses; these indicators are discussed in Annex 3. - Questions on revenue collections and staff demographics are generally the most complete. FY 2022 data availability levels in these areas are 22 and 9 percentage points higher, respectively, than for the equivalent sets of indicators collected for 2014 in the first round of ISORA. - The highest proportion of missing data is for questions on payments, arrears, and audit—areas critical for effective tax administration. Improvement has been limited; for arrears, 63 percent of the equivalent arrears data points for 2014 were available, compared with 67 percent for 2022. - Data input required in the survey on arrears and audit since the first two rounds has been streamlined and clarified; limited improvement suggests further investigation is warranted. - Audit and the management of arrears may be contentious in some jurisdictions, so data gaps may reflect reluctance to be transparent or a general lack of availability within administrations. - Note on zeros: Zeros are assumed to be valid responses for the following indicators: all staff demographics apart from staff levels at beginning and end of year, individual taxpayers managed under a large taxpayer office/program, filing channels, electronic payments, arrears associated with state-owned enterprises, criminal cases, disputes, and incoming service channels. ### Data Availability by Tax Administration (Average data availability rate in percent, 2022) - Overall patterns by economic grouping (AEs, EMEs, LIDCs): - Administrations within AEs generally exhibit higher data availability rates than those in EMEs, and EMEs provide more numerical data than LIDCs. - Exceptions: for registration and segmentation, EMEs are more likely to provide numerical inputs than AEs. For registration, this may reflect differences in tax registration approaches and reliance by AEs on administrative records compiled by other public entities. - For all groups, availability of data for revenue collections and staff demographics is high; these areas show the smallest differences between economic groupings. - The largest spread across economic groups is observed for core tax administrative functions: return filing, arrears, audit, cases, and service channel usage. - LIDCs show the greatest relative weakness in data provision on filing, arrears, audit, disputes, incoming service channels, and resources. - Key aggregated numeric indicators (from figure numeric sequence as presented): - Example numeric sequence appearing in figure: 81 95 88 94 88 85 78 66 66 72 78 65 87 96 97 98 87 86 86 68 76 77 89 74 73 92 80 93 84 79 66 59 55 60 62 43 40 50 60 70 80 90 100 (presented in source figure; included here verbatim as displayed). ### Administrations with Data Availability Rates of over 95 Percent for FY 2022 - Listed administrations (grouped by economic classification in source): - AEs: Hong Kong; Greece; Ireland; Lithuania; Malta; The Netherlands; Portugal; Slovak Republic; Slovenia; Trinidad and Tobago; Ukraine - EMEs: Albania; Armenia; Barbados; Bulgaria; Eswatini; Kosovo; Maldives; Peru; St. Vincent and the Grenadines - LIDCs: Argentina; Nigeria; Aruba; Sierra Leone; Belize; Uganda; Dominican Republic; Georgia; Malaysia; Mongolia; St. Lucia - Note: These administrations span all economic groups and regions and include SDS administrations. ### Comparisons between and among Data for Different FYs — Considerations for Time-Series Analysis - ISORA collects time-series data; users often compute averages or medians to compare FYs or discern trends. - Factors that can affect changes in group statistics over time: - ISORA is voluntary; participants differ across rounds. - Some administrations participating in all rounds may not provide data for specific numerical questions in every period. - Reclassification of participants between economic groupings across rounds (for example, from LIDC to EME) can affect group statistics. - Questionnaire changes across rounds (wording, question structure, sequencing, switch from yes/no to checkbox, movement of questions) can impact comparability. - Changes in the operating environment of tax administrations necessitate question changes. - Specific notes on periodic questions: - Many periodic questions asked about FY 2022 were last asked about FY 2017, when the full set of ISORA questions (annual and periodic) was administered. - Periodic questions tend to change less frequently than quantitative measures, but many reasons exist for changes over five years. - Ahead of ISORA 2023 (FY 2022), ISORA partners’ technical working group reviewed all periodic questions previously asked about FY 2017: some were discarded, new ones added, many reworded or restructured to improve consistency and accuracy. - ISORA was still relatively new when FY 2017 data were collected; improvements in regular questions were transferable to periodic questions. - Much training and assistance provided to participants over the past five years resulted in higher quality responses in ISORA 2023. ### Accessing and Using ISORA Data - ISORA data are publicly available on the ISORA portal, https://isoradata.org. - Annex 4 in the source provides guidance on accessing data for individual tax administrations, ISORA questionnaires and guidance to respondents, and publications that have used ISORA data. - Annex 5 presents a selection of comparative aggregate statistics for various groupings of tax administrations, calculated from the published data for individual tax administrations. - Note from source: The original ISORA portal, https://data.rafit.org, was decommissioned in the course of 2025. *Source: Authors’ calculations and analysis using ISORA data as presented in the ISORA 2023 departmental paper.* ### 1. Net Revenue Collected ### 1. Net Revenue Collected ### Net revenue levels (by economic grouping) - AEs: nearly a quarter of GDP collected by tax administrations (presented as “almost a quarter of GDP”). - EMEs: 17 percent of GDP collected by tax administrations. - LIDCs: 13 percent of GDP collected by tax administrations. ### Net revenue as share of total government revenue - Collections by tax administrations average around 60 percent of total government revenue across AEs, EMEs, and LIDCs. - Time series (FY 2018–FY 2022) statistics derived from data for 20 LIDCs, 61 EMEs, and 35 AEs. ### Revenue composition by tax type (2018 vs 2022; percent of net revenue collected) - VAT constitutes the largest contribution to collections by tax administrations among EMEs and LIDCs. - For AEs, the contribution of PIT is slightly larger than that of VAT. - CIT makes a larger contribution to collections in EMEs and LIDCs than in AEs. - SSCs: - Make up about a quarter of collections by tax administrations in AEs. - Make up a little less than a quarter in EMEs. - Few tax administrations in LIDCs collect SSCs; where collected, contribution is around 10 percent of total collections. - Nontax revenue (excluding SSC) is a nonnegligible portion of revenue collected by the tax administration in LIDCs, and to a lesser extent in EMEs. - Note: In panel 3 percentages may add to more than 100 because contributions exclude administrations that do not collect a revenue type. In FY 2018, only two LIDCs provided SSC collection figures. ### Pandemic years (FY 2020–FY 2021) - Although GDP and government revenues fell because of the pandemic, revenue collected by tax administrations as a percent of revenue or of GDP did not drop in FY 2020 or FY 2021. - Revenue collected by administrations, especially those in LIDCs, was less adversely affected in FY 2020 than other government revenue sources, increasing their share of total revenue in overlapping FYs with calendar year 2020. --- ### 2. Tailored Responses to Taxpayer Segments ### Prevalence of specialized units and regimes (2018 vs 2022) - LTO/P (large taxpayer office/program): - Around 90 percent of EMEs and LIDCs have an LTO/P. - 78 percent of AEs have an LTO/P. - Apparent small decline in administrations with an LTO/P between 2018 and 2022 is largely driven by change in participation of SDS in ISORA. - HNWI programs: - More frequently found in AEs. - Slow growth in number of HNWI programs in EMEs and LIDCs, including SDSs and FCSs. - Simplified income tax regimes: - More prevalent in LIDCs than in EMEs or AEs. - Offered in three-quarters of FCSs. ### Practical notes - In tax administrations with few staff, dedicating staff to specific taxpayer segments may not be practical. - Large variation exists in size, function, resourcing, and operating metrics of LTO/Ps; no “typical” characteristics can be discerned. --- ### 3. Variation across Large Taxpayer Offices/Programs (Box 2) ### Contribution and variation - LTO/Ps contribute significantly to revenue collections, especially among LIDCs. - The interquartile range for percent revenue collected through LTO/Ps in LIDCs is 26 percentage points (middle 50 percent range). ### Percent revenue collected through LTO/P by economic group (box figure summary) - Second and third quartiles shown for AEs, EMEs, LIDCs (figure-based summary). ### Percent revenue collected through LTO/P by region (average percent; box figure summary) - Regional averages presented for AFR, APD, EUR, MCD, WHD. ### Functions and inclusion - HNWI program included in 20 percent of LTOs. - Just over one-third of LTO/Ps do not include individual taxpayers at all. - Functional coverage (percent LTO/Ps with function/characteristic): - Audit, investigation, other verification: 97 percent. - Taxpayer services: 88 percent. - Return and payment processing: 76 percent. - Enforced debt collection: 74 percent. - Dispute resolution: 52 percent. - Registration: 43 percent. - All six functions: 28 percent. - HNWI unit/program part of LTO: 20 percent. - Individual taxpayers included in LTO/P: 63 percent. - Individual TPs outnumber corporate TPs: 8 percent. ### Resources and audit coverage (box figure summary) - Percent FTEs in LTO/Ps (by economic group) and audits per 100 large taxpayers, taxpayers per FTE presented in chart form showing wide variation across AEs, EMEs, LIDCs. - No universal benchmarks or “ideal” LTO/P characteristics can be defined given heterogeneity. --- ### 4. Tax Administration Resources ### Operating and ICT expenditure (ISORA data) - Operating expenditure medians (2022, percent of GDP): - All: 0.146 - AEs: 0.154 - EMEs: 0.152 - LIDCs: 0.125 - SDS: 0.186 - FCS: 0.102 - AFR: 0.174 - APD: 0.107 - EUR: 0.168 - MCD: 0.099 - WHD: 0.137 - Operating expenditure 2018–2022 medians for administrations that provided data every year (panel 2): 0.153 (All); 0.159 (AEs); 0.166 (EMEs); 0.154 (LIDCs) with annual points 2018–2022 shown. - Salary expenditure (2022, averages, percent of operating expenditure): - All: 70 - AEs: 72 - EMEs: 71 - LIDCs: 66 - SDS: 74 - FCS: 67 - AFR: 67 - APD: 66 - EUR: 71 - MCD: 72 - WHD: 73 - ICT expenditure (2018–2022, medians, using country average over 2018–2022): - Percent of operating expenditure axis shown (0 to 12 scale) and percent of GDP axis shown; ICT expenditure is higher for AEs. ### Key points on resource allocation - AEs generally allocate a larger operating budget to tax administration relative to GDP than EMEs and LIDCs; the median allocation for AEs is 23 percent higher (measured relative to GDP) than the median allocation to LIDCs. - FCSs generally allocate even less funding to tax administrations than LIDCs. - SDSs allocate more than other groupings (measured relative to GDP). - Operational expenditure shows a slight peak in 2020 and lower levels in 2022 than in 2018 for many groupings; trend suggests potential increased efficiency but warrants deeper exploration. - Salaries dominate operational costs: medians fall in range 66 percent to 74 percent of operational costs; SDS median is 84 percent. - ICT expenditure is higher for tax administrations in AEs, whether measured relative to GDP or operational costs. --- ### 5. Human Resources ### FTEs and staff ratios (population and labor force per FTE) - Population per FTE (medians, interquartile ranges shown in figure): - FY 2018 and FY 2022 depicted with group medians; specific medians noted in figure captions: - FY18: AEs median 1,122; EMEs median 11,440; LIDCs median (charted). - FY22: AEs median 1,224; EMEs median 2,062; LIDCs median 10,728 and other charted values. - Labor force per FTE (figure medians): - FY18 and FY22 values charted; selected numeric medians from figure captions: 1,224; 2,062; 10,728; and other plotted medians such as 2,101, 5,576, 1,012, 4,373, 3,829, 942, 615. - Key points on staffing: - Measured in relation to population, tax administrations in LIDCs have roughly one-tenth of the FTEs in AE tax administrations. - Administrations in EMEs have roughly half the FTEs of AE administrations. - For each staff member (FTE) in an LIDC tax administration (near the median), there are approximately eight times the number of labor force participants than for their AE counterpart. - For each staff member in an EME administration, there are twice the number of labor force participants than for their AE counterpart. - Within groupings, FTE resource ranges are large: ratios at 75th percentile are three to four times ratios at 25th percentile. - Over the five-year period, resources of AEs have not kept pace with population or labor force growth, whereas LIDCs show the opposite trend. --- ### 6. Allocation of Staff by Function ### Functional allocation (2018 vs 2022; percent staff as FTEs) - 2018 (selected group averages): - AEs (2018): Taxpayer services 32.9; Audit 29.7; Debt collections 9.6; Other remainder. - EMEs (2018) and LIDCs (2018) presented similarly in figure. - 2022 (selected group averages with expanded functions): - AEs (2022): Taxpayer services 34.4; Audit 19.0; Debt collections 12.7; Dispute Management 4.9; HRM 2.8; ICT support 5.5; Other 15.1. - EMEs (2022): Taxpayer services 33.9; Audit 32.2; Debt collections 29.0; (figure shows detailed splits). - LIDCs (2022): Taxpayer services 32.6; Audit 18.3; Debt collections 10.8; Dispute Management 2.4; HRM 4.3; ICT support 4.1; Other 28.1. ### Key points on allocation - Average percentage of staff allocated to services, audit, and debt collection changed little between 2018 and 2022 across AEs, EMEs, LIDCs. - AEs and LIDCs allocate close to one-third of staff to taxpayer services. - EME administrations allocated almost equal proportions of staff to taxpayer services and to audit (26 percent each in 2022). - AEs allocate a higher proportion of staff to audit and verification (29 percent in 2022) than LIDCs (19 percent in 2022). - “Other” functions proportion fell for all groupings from 2018 to 2022 after adding three functions in 2022; “Other” remains highest among LIDCs. - Breakdown of the three categories introduced in 2022 (collectively 12–13 percent of staff across groups): - ICT support: AEs 5.5 percent; EMEs 4.8 percent; LIDCs 4.1 percent. - Dispute management: AEs 4.9 percent; LIDCs 2.4 percent. - HRM: LIDCs 4.1 percent; EMEs 3.2 percent; AEs 2.8 percent. --- ### 7. Core Tax Administration Performance Indicators — Registration ### Definition and measurement - ISORA defines active taxpayers as: “...those for whom a tax consequence arises during the fiscal year (tax liability or tax refunds without any obligation to file) or that for any other purpose file a tax return).” - From FY 2021 onward ISORA added: “Are most employees that have tax deducted from their salary by the employer required to register for personal income tax?” — enabling separate analysis of taxpayer numbers for administrations answering “yes” or “no.” - Many administrations report identical figures for taxpayers on register and active taxpayers, implying no inactive taxpayers on register or difficulty in distinguishing active status. - ISORA analyzes ratios of taxpayers to labor force for PIT, CIT, employers that withhold tax from employees, VAT, and domestic excise to compare active taxpayer population over time and across economies. *Italic source attribution: Authors’ calculations using government revenue and GDP from the IMF’s World Economic Outlook database; and revenue collected by tax administrations from ISORA.* ### 1. Administrations Reporting Equal Figures for ### 1. Administrations Reporting Equal Figures for ### Active taxpayers equal to taxpayers on register - Administrations within AEs are more likely to provide identical figures for taxpayers on register and active taxpayers than either EMEs or LIDCs for all tax types. - Within AEs, tax administrations are more likely to have access to other reliable public administration registers that provide information on actual and potential taxpayers. - Domestic excise registration figures provided by ISORA participants are most likely to be equal; that is, all taxpayers on the domestic excise register are active. - Between 2018 and 2022, there was a notable decline in the number of LIDC tax administrations that did not distinguish between inactive and active taxpayers across all five tax types. - For most countries, there are far fewer domestic excise taxpayers than there are CIT or VAT taxpayers, so maintaining the domestic excise register should be easier than for larger registers. ### Active Taxpayers as Percent of Taxpayers on Register (Average, in percent) - Data presented by tax type and economic group for 2018 and 2022 (figures shown in source panels). - Note: Administrations with active taxpayers equal to registered taxpayers are excluded from consideration in the statistics presented in panel 2; AEs = advanced economies; CIT = corporate income tax; EMEs = emerging market economies; Employers = employers who withhold tax and other contributions from their employees; LIDCs = low-income developing countries; PIT = personal income tax; VAT = value-added tax. ### Key aggregate observation on active taxpayers - Overall, the percentage of active taxpayers increased between 2018 and 2022. - This is true for most tax types when disaggregated by economic group, except: - a marginal decline in PIT for LIDCs, - declines in active employers, VAT, and domestic excise for EMEs, - a small decline among AE administrations for VAT. --- ### Active PIT taxpayers in relation to labor force - Large range in the ratio of active PIT taxpayers to the labor force (interquartile distance, median). - Median ratio for AEs roughly five times as large as that of EMEs and a hundred times that of LIDCs. - In most AEs, active taxpayers exceed the number of people in the labor force. - Within LIDCs, active taxpayers typically make up less than 1 percent of the labor force. - Active taxpayers grew appreciably faster than the labor force between 2018 and 2022, especially within EMEs and LIDCs. ### PIT registration requirements for employees (2022) - In all three economic groups, in most countries, most employees are not required to register for PIT. - Among LIDCs, most employees are required to register as taxpayers in one-third of all countries. - Median active-taxpayer-to-labor force ratio is lower for administrations that do not require most employees to register (true for each economic group). - Large difference in LIDCs between administrations that do/do not require most employees to register; smaller difference among AE administrations. - Possible reasons: larger proportion of retirees in active PIT population in AEs; incentives for salary earners with lower incomes to register due to tax credits and/or other social benefits administered by the tax administration. ### Active CIT, VAT, employers, and domestic excise taxpayers (median relations to labor force) - CIT and VAT active taxpayer numbers are lower than those for PIT across all economic groups. - In AEs, at the median, there are approximately 10 active PIT taxpayers for each active CIT or VAT taxpayer. - LIDC figures roughly one-tenth of EME figures; EME figures 3 to 5 times smaller than those of AEs. - Active employers (who withhold taxes) measured relative to the labor force: numbers in AEs exceed those in EMEs, which in turn exceed those in LIDCs. - Domestic excise taxpayer numbers are lower by a factor of roughly 100 than CIT or VAT taxpayers. - Taxpayer numbers have outstripped growth in the labor market in most tax administrations across the five-year period, with the exception of VAT. --- ### On-Time Filing - ISORA uses the TADAT definition for on-time filing: number of on-time returns/returns expected from active taxpayers. - FY 2022 questionnaire change requires an FY figure for total returns received, enabling comparison of total returns received on time to total returns received. - Observed on-time filing (TADAT-defined) 2018–2022 and percent returns received on time (available for FY 2022 only) for CIT, PIT, VAT, and employers. Key findings on on-time filing: - On-time filing rate is generally higher for VAT than for the other tax types, particularly among LIDCs. - Across all taxes, on-time filing rates reported by AEs are significantly higher than those of EMEs or LIDCs; difference greatest for PIT. - On-time filing rates change year over year but are most stable among AEs. - Downward trends in on-time filing rates among LIDCs for CIT and PIT are visible between 2018 and 2022. - Registered taxpayers in LIDCs grew significantly over the five-year period. - Many LIDCs use active taxpayers to estimate expected returns, increasing expected returns. - Numbers of returns and on-time returns have not grown as much, implying higher nonfiling rates among EMEs and LIDCs. - Some administrations report receiving more returns, or even more on-time returns, than they expected, complicating characterization of the expected/received relationship. ### On-Time Payment - ISORA collected data to determine on-time payment rates using TADAT definition since survey start. - FY 2022 questionnaire change includes payments due, payments received, and payments received on time, enabling direct comparison for each tax type. - TADAT-defined on-time payment rate by value is value of on-time payments received/value of payments due; rates exceeding 150 percent were dropped. - Fewer administrations provide data enabling determination of on-time payment rates than on-time filing measures. Key findings on on-time payment: - On-time payment rates reported by AEs are higher than those of EMEs or LIDCs, by as much as 20 percentage points over rates for LIDCs. - On-time payment rates show a lot of annual variability but are more stable among AEs. - More often than not, payments received are less than payments due, leading to higher averages for on-time payments as a percentage of payments received. --- ### Electronic Filing (E-Filing) and Electronic Payment (E-Payment) - ISORA includes time series on return channels for PIT, CIT, VAT, employers, and percentage of payments received electronically. - FY 2022 periodic component probed mandatory nature of e-filing and e-payment. Key findings: - Over 60 percent of tax administrations in LIDCs require e-filing and e-payments from most taxpayers (and active employers) across all tax types, exceeding proportions in AEs and EMEs. - More administrations required e-filing and e-payment by most taxpayers in 2022 than in 2017 (despite changes in question wording). - E-filing rates for all taxes in AEs are higher than in EMEs and LIDCs, but rapid growth in offering and uptake of e-channels in LIDCs: - More than half the returns filed in LIDCs are now filed electronically, even for PIT. - E-payment rates mirror e-filing patterns: - E-payment rates are highest among AEs at close to 90 percent. - Rapid growth in e-payment within LIDCs; more than three-quarters of payments in FY 2022 were received electronically by tax administrations. - Note: Average rates include administrations that do not offer e-services (rates of zero included). --- ### Prefilling of PIT Returns - ISORA probes prefilling of PIT returns by administration, including income types and deductions pre-populated from third-party data. - Figures requested on numbers of electronic returns pre-populated using third-party data for all main tax types, though some administrations have difficulty providing separate figures. Key findings on prefilling (percent of administrations that prefill): - Almost 90 percent of AEs have been using third-party data to populate PIT returns for over five years. - Approximately 50 percent of EMEs prefill PIT returns (FY 2022). - 13 percent of LIDCs prefilled PIT returns in FY 2022. - Number of administrations in AEs reporting they prefill PIT returns did not change over five years, but many added new data sources during the period. - Appreciably more EMEs prefilled PIT returns in FY 2022 than five years earlier. - Growth in prefilling among LIDCs is less clear despite increased e-filing adoption and uptake. - Effective prefilling requires high confidence in completeness and accuracy of third-party data and ability to associate third-party data with correct taxpayer; these factors may slow adoption in LIDCs. - FY2018 LIDC prefilling figures are less reliable than FY2022; some LIDCs reporting prefilling in FY2018 may have misinterpreted the question. --- ### Arrears - Management and reduction of tax arrears is a priority; common performance measure is ratio of stock of total arrears at year-end to total collections for the FY. - ISORA collects data to calculate this indicator and figures on collectible arrears. - Questions on arrears owed by state-owned enterprises (SOEs) were added to ISORA to capture collection challenges from SOEs identified in EMEs and LIDCs. - Figure 23 in the source presents statistics on arrears by economic group (details in source figure). *Source: Authors’ calculations. IMF DEPARTMENTAL PAPERS • ISORA 2023 (figures and notes as presented in the provided content).* ### 1. Arrears as Percent of GDP ### 1. Arrears as Percent of GDP ### Arrears overview and key findings - The stock of arrears reported by most tax administrations in ISORA constitutes 3 or more percentage points of GDP (panel 1). - Measured relative to either GDP or annual revenue collections, the stock of arrears among EMEs is generally higher than that of either AEs or LIDCs. - Levels of arrears are similar for AEs and LIDCs when measured against GDP, but because LIDCs generally collect lower revenue as a percent of GDP, their arrears are larger in comparison with annual collections. - Most tax administrations experienced an increase in arrears in FY 2020 (when COVID-19 had the greatest impact on economies), but arrears have fallen again in subsequent years. ### Numeric indicators shown in the source (preserved exactly) - Arrears as Percent of Collection (Median, in percent): 2.9, 1.9, 5.6, 1.5, 3.3, 2.4, 6.2, 2.0, 2.9, 1.9, 5.0, 2.3 - Categorical labels present in charts: All AEs EMEs LIDCs and years: 2018 2020 2022 ### Arrears by tax type - EMEs exhibit the highest arrears ratios across core tax types, with the highest ratios seen for CIT. - VAT arrears are the second highest for all tax types, despite VAT often being collected monthly and being relatively “young” compared with income tax. - LIDCs record higher levels of PAYE arrears than PIT arrears, suggesting some administrations may have reported PIT arrears that exclude PAYE arrears; however, PAYE arrears are significantly lower than PIT arrears among both AEs and EMEs. ### Numeric indicators for tax-type arrears (preserved exactly) - Arrears by Tax Type (Median, in percent) values shown in source: 37, 36, 16, 23, 11, 13, 25, 22, 7, 12, 3, 5, 10, 9, 13, 11 - Chart labels: 2018 2022 and series labels: "CIT arrears as percentage of CIT collected", "PIT arrears as percentage of PIT collected", "PAYE arrears as percentage of PIT collected", "VAT arrears as percentage of VAT collected", with groups All AEs EMEs LIDCs ### Collectible arrears and SOE contribution - Broad picture of collectible arrears resembles that of the total stock, with EMEs reporting the highest levels of collectible arrears and a peak in FY 2020; LIDCs show an exception with an increase in collectible arrears. - The categorization of debt as “collectible” or “noncollectible” is not standardized across time or across administrations, complicating interpretation. - Contribution of SOEs to the debt stock varies by group: - Among AEs, arrears associated with SOEs constitute less than 1 percent of the debt stock, and many AEs do not specifically track SOE debt. - Over half of tax administrations in EMEs and LIDCs report arrears figures for SOEs. - Administrations in EMEs report a median of 3 percent to 4 percent of debt associated with SOEs. - For LIDCs the figures are significant, with SOE debt constituting around one-fifth or more of debt for half of the LIDCs that provide these data. ### Numeric indicators for SOE and collectible arrears (preserved exactly) - Arrears Owed by SOEs (Percent) values shown: 8, 4, 12, 5, 12, 5, 19, 10, 9, 4, 14, 11 - Collectible arrears panel numbers: 7, 4, 8, 0.3, 0.8, 0.4, 4, 3, 4, 19, 15, 22 - Chart labels: All AEs EMEs LIDCs and years: 2018 2020 2022 (for SOE panel) and 2018 2022 (for collectible arrears) --- ### Audit and Verification Metrics ### Summary of audit and verification practice trends - In FY 2022, around 80 percent of tax administrations made compliance interventions prior to triggering resource-intensive audits. - There is a sharp difference in verification during filing: over 80 percent of AEs undertake verification during filing, whereas less than half of EMEs and LIDCs reported doing so in FY 2022. - Over 80 percent of administrations use third-party data to identify inconsistencies and request further information; 80 percent conduct risk reviews. - Use of data analytics to identify anomalies is well developed among AEs and lags among EMEs and LIDCs. - Over 90 percent undertake comprehensive audits and desk audits; limited-scope and single-issue audits are lower (86 percent and 85 percent respectively), largely because fewer LIDCs report undertaking them. - Audit hit rates remain highest among LIDCs. Hit rates among AEs and EMEs fell slightly between FY 2018 and FY 2022. - Assessments raised through audit as a percent of GDP and of collections have fallen for all economic groups; the median assessment ratio for AEs in FY 2022 is less than half its FY 2018 value. - CIT assessments raised through audit are larger than assessments for other tax types for all groups. - Among AEs and LIDCs, PIT audit assessments are higher than PAYE assessments; LIDCs report the opposite in ISORA data. - Assessments raised on VAT are very small compared with VAT compliance-gap estimates of 10 percent to 50 percent of collection. ### Numeric indicators preserved exactly from charts and panels - Audits and verifications undertaken (Percentage of administrations that undertake type) (values in source sequence): 53 84 71 80 83 92 85 86 86 92 80 - (Chart labels show types: compliance checks, identify inconsistencies through third-party data crossmatching, identify anomalies through data analytics to prompt taxpayer disclosure, risk reviews, request further information; and types of audit: Desk, Single issue, Limited scope, Comprehensive, Avoidance and evasion investigations; timing: During filing, After filing but before audit) - Audit Hit Rate (In percent, average): 69, 52, 71, 78, 67, 51, 69, 79 (labels correspond to 2018 and 2022 across All AEs EMEs LIDCs) - Assessments Raised through Audit (In percent of GDP, median): 0.51, 0.45, 0.60, 0.31, 0.32, 0.27, 0.33, 0.39 (2018 and 2022 across groups) - Assessments Raised through Audit (In percent of collections, median): chart panels present values (see source figures) - CIT Assessments Raised through Audit (In percent of CIT collected, median): 3.1, 2.4, 3.8, 5.0, 2.4, 1.1, 2.4, 4.3 (2018 and 2022 across All AEs EMEs LIDCs) - Time series values shown in source for broader sequence: 7.0 5.1 5.5 5.1 4.2 (presented in a panel in the source) - PIT and PAYE Assessments Raised through Audit (In percent of PIT collected, median): 2.2, 2.4, 1.7, 2.0, 1.8 (series across years/groups shown in source) - VAT Assessments Raised through Audit (In percent of VAT collected, median): 2.2? (values and series shown in source charts) --- ### Efficiency Measures ### Measures used - Revenue collected in relation to operational costs (cost of collection). - Active taxpayers in relation to FTEs (sum of active taxpayers across PIT, CIT, VAT, plus employers who withhold). - Returns received in relation to FTEs (FY 2022 only; returns from PIT, CIT, VAT, and employers). ### Key findings - All three measures indicate higher tax administration efficiency in AEs than in EMEs, and higher efficiency in EMEs than in LIDCs. - Differences in cost of collection arise primarily from differences in revenue collected as a percent of GDP and operating budgets as a percent of GDP. - In a typical AE there are about 100 times the number of PIT taxpayers relative to a typical LIDC with a similarly sized labor force; administrations in AEs have roughly eight times the number of staff that an LIDC with a similar labor force would have. - Small developing states (SDS) show lower efficiencies than EMEs on all three measures and a higher cost of collection even than LIDCs. - Fragile and conflict-affected states (FCS) show a mixed picture, with declining operating expenditure and low returns per FTE, suggesting need for deeper investigation. - Regionally, sub-Saharan Africa (AFR) exhibits the highest median cost of collection; EUR shows higher efficiency on taxpayers per FTE and returns per FTE. ### Numeric indicators preserved exactly - Operational Cost as Percent of Revenue Collected (FY 2018 and FY 2022) (Percent, median) values shown in source: 0.9, 0.7, 0.9, 1.4, 1.4, 2.7, 1.3, 0.7, 0.7, 0.7, 1.0, 0.76, 0.62, 0.77, 1.10, 1.40, 0.82, 1.70, 0.68, 0.63, 0.62, 0.83 - Active Taxpayers per FTE (FY 2018 and FY 2022) (Number of active taxpayers, median) values shown: 512, 1,198, 382, 161, 369, 113, 152, 408, 806, 214, 594 (FY18 series) and 732, 1,627, 665, 169, 266, 106, 120, 433, 1,518, 458, 743 (FY22 series) - Returns Received per FTE (FY 2022) (Number of returns, median) labels and values are presented in the source charts (see figures) ### Additional observations - Comparison between FY 2018 and FY 2022 suggests an increase in tax administration efficiency for most administrations: operating expenditure (relative to GDP) has fallen for most AEs and LIDCs; FTEs grew relative to labor market size in most LIDCs but shrank in most EMEs and AEs. - Growth in active taxpayers per FTE among EMEs and LIDCs may in part reflect improvements in data management and register quality. - Tracking returns per FTE alongside active taxpayers per FTE over coming years may yield valuable insights into efficiency and effectiveness. *Source: Authors’ calculations. Note: AEs = advanced economies; CIT = corporate income tax; EMEs = emerging market economies; FY = fiscal year; LIDCs = low-income developing countries; PAYE = pay as you earn; PIT = personal income tax; SOEs = state-owned enterprises; VAT = value-added tax.* ### 5. Practices and Structural Foundations ### 5. Practices and Structural Foundations for Effective Tax Administration ### A. The use of indices to reflect practices and structural foundations - ISORA 2023 included many questions on practices and structural foundations, resources, and some performance measures, enabling analysis of relationships among practices, performance, and operational metrics. - Indices were developed to distill many ISORA data points into thematic measures suitable for comparison over time and across jurisdictions. - Indices are compiled from responses to survey questions (mostly “yes/no”), with each index having several weighted components based on one or more ISORA questions. - Higher numbers of positive responses yield higher index scores; each index reflects the degree of “good” practice or the extent of “strong” structural foundations. - The ISORA dataset was not specifically designed to produce indices; indices are produced as by-products of ISORA responses to facilitate comparative analysis. - Previous IMF work used similar indices; changes in ISORA questions over time complicate direct comparisons across vintages. ### B. ISORA indices based on FY 2022 data - Seven indices developed (scores range 0 to 100): - Compliance management (CM) — extent of sound compliance practices and institutional features, including a risk-based approach. - Digitalization (DG) — extent to which practices and systems support the use of digital technologies in business processes. - Governance (GV) — extent to which laws, authorities, procedures, and structures provide a solid framework for governance. - Human resources management and development (HRMD) — extent to which modern HR practices and development policies support tax administration. - Management autonomy (MA) — extent to which laws, regulations, and policies provide autonomy for strategy, resource allocation, and operations. - Public trust (PT) — extent of information placed in the public domain as part of accountability and transparency. - Taxpayer services (TS) — extent of taxpayer services available to support voluntary compliance. - Weighting approach: - Components (questions or groups of subquestions) are weighted to reflect relative contribution to an index. - Weightings were reviewed and adjusted for FY 2022 indices to ensure consistent cross-index relativities. - Some ISORA responses contribute to more than one index when practices are pertinent to multiple domains. - Distribution highlights (scores and dispersion): - Public trust mean score: 47. - Management autonomy mean score: 72. - Three indices show the widest variation in scores: HRMD, MA, and PT. - Taxpayer services and compliance management show the most limited variation in scores. - For most indices, the mean score lies between 3 and 6 percentage points below the median, except compliance management where mean and median are very close. - The fourth quartile has the widest range, reflecting wide adoption heterogeneity among lower-capacity administrations. - Correlations among indices (pairwise correlation coefficients; all correlations positive): - CM–DG: 0.73 - CM–GV: 0.69 - CM–HR: 0.58 - CM–MA: 0.46 - CM–PT: 0.65 - CM–TS: 0.69 - DG–GV: 0.68 - DG–HR: 0.59 - DG–MA: 0.51 - DG–PT: 0.65 - DG–TS: 0.77 - GV–HR: 0.84 - GV–MA: 0.83 - GV–PT: 0.73 - GV–TS: 0.83 - HR–MA: 0.76 - HR–PT: 0.60 - HR–TS: 0.68 - MA–PT: 0.48 - MA–TS: 0.56 - PT–TS: 0.71 - Observations on correlations: - Governance correlates strongly with all other indices (governance index includes key components of other indices). - Strong association between taxpayer services and digitalization. - Associations with management autonomy are positive but relatively weaker; correlation MA–PT is one of the smaller observed correlations. ### C. Scores against the FY 2022 indices for groupings of interest - Sample sizes: - Indices calculated for 37 AEs, 83 EMEs, and 46 LIDCs. - FCS and SDS sample counts: 22 FCSs and 31 SDSs (seven countries classified as both FCSs and SDSs). - Key comparative points by economic group: - AE averages are significantly higher for all seven indices, exceeding the higher of the LIDC or EME average by at least 10 percentage points for CM, and for the overall average by 18 percentage points. - The largest differences between average LIDC and AE scores exceed 20 percentage points for public trust, human resources management and development, and digitalization. - LIDC and EME averages are close for all indices, separated by less than 4 percentage points in every case. - Principal component analysis: first principal component explains 72 percent of the variance; correlation between simple average of seven index scores and first principal component is 0.999. - Drivers of AE versus LIDC differences (components with largest response gaps): - Compliance management: - Availability of data from the Social Security Agency - Availability of data on property ownership and sales - Specialized audit staff for international tax issues - Entering into advance pricing arrangements - Facilitating taxpayer access to third-party data before filing - Digitalization: - Prefilling income and expense items in PIT tax returns - Data science/analytics tools in place - Systems for managing data from the Social Security Agency - Systems for managing data on property ownership and sales - Facilitating taxpayer access to third-party data before filing - Automated verification of returns - Governance: - Administration has authority to terminate employment - Administration can design its own internal structure - Conducts individual taxpayer satisfaction surveys - Conducts business taxpayer satisfaction surveys - HRMD: - Policy for flexible working arrangements in place, and associated flexible working arrangements - Surveying staff on diversity and inclusion - Sharing results of staff engagement surveys - Administration has authority to terminate employment (also relevant for governance) - Time-reporting system - Management autonomy: - Administration has authority to terminate employment (also governance) - Administration can design its own internal structure (also governance) - Public trust: - Administration publishes an annual report - Administration publishes an annual plan or operational business plan - Administration publishes results of taxpayer satisfaction surveys - Administration publishes tax gap analyses - Taxpayer services: - Special provisions for taxpayers with disabilities - Whole-of-government identification systems in place - Web services and apps available in languages other than official languages - Virtual assistants (chatbots) in place - Interpretations: - Digitalization gaps largely reflect access to and management of third-party data and analytic capacity, limiting LIDCs’ ability to facilitate voluntary compliance and detect inaccurate declarations. - Public trust gaps reflect lower rates of publishing annual reports, plans, taxpayer satisfaction survey results, and tax gap analyses, reducing public visibility into administration performance in LIDCs. - HRMD gaps (flexible working arrangements, staff engagement surveys, authority to terminate employment, time-reporting systems) may reduce staff morale, increase turnover, and impair productivity in LIDCs. - Areas where LIDCs lead or match AEs include having an integrity strategy, producing annual business/operational plans and annual reports (though less likely to publish them), and having a formal document of taxpayer rights. - Fragile and conflict-affected states (FCSs) and small developing states (SDSs): - Across all indices, average scores for FCSs are lower than average scores of LIDCs. - FCS average scores are lowest relative to all ISORA participants for management autonomy, HRMD, and digitalization. - SDS average scores are lower than FCS averages for six indices, and 1 percentage point higher for management autonomy (53 versus 52). - Caution: small administration scale limits range of services and compliance responses; comparisons with larger administrations require care. *Source: Authors’ analysis of ISORA FY 2022 data as presented in the chapter "5. Practices and Structural Foundations" of the ISORA 2023 departmental paper.* ### Box 3. Index Scores and Size of Tax Administrations ### Box 3. Index Scores and Size of Tax Administrations ### Index Scores and Administration Size - Practice indices reflect a range of sound approaches and successful strategies for effective tax administration. Constraints in human and financial resources limit the ability of smaller tax administrations to develop and offer the spectrum of services and approaches larger counterparts may be able to provide. - Scores against all practice indices increase for tax administrations grouped by size, as measured in FTEs. - For some indices (for example, compliance management and taxpayer services) scores increase steadily with size; other indices (for example, public trust) hint at “size threshold” effects associated with an effort “overhead.” - Administration size effects are observable even within economic groups for most indices. - Box Figure 3.1 (panel notes): - Scatterplots indicate relationships such as taxpayer service index score against Log FTEs for AEs, EMEs, and LIDCs. - Reported R² values visible in the figure: R² = 0.18; R² = 0.57; R² = 0.21. ### Correlations of Index Scores against Log FTEs (Box Table 3.1) - Correlation coefficients by Economic Group (CM = compliance management; DG = digitalization; GV = governance; HRMD = human resources management and development; PT = public trust; MA = management autonomy; TS = taxpayer services): - AEs: CM 0.23; DG 0.07; GV 0.42; HRMD 0.54; PT 0.21; MA 0.25; TS 0.42 - LIDCs: CM 0.44; DG 0.45; GV 0.38; HRMD 0.17; PT 0.20; MA 0.22; TS 0.46 - EMEs: CM 0.63; DG 0.81; GV 0.68; HRMD 0.61; PT 0.52; MA 0.64; TS 0.75 ### Average Index Scores by Economic Group and SDS Classification (Box Figure 3.2) - The average index scores by economic group and SDS classification reveal a strong pattern across all seven indices (Compliance management; Digitalization; Governance; HR management and development; Management autonomy; Public trust; Taxpayer services). - Note: Five LIDCs are also SDS; 41 LIDCs are not. 26 EMEs are SDS; 57 are not. AEs are not classified as SDS, although four AEs that completed ISORA have populations below 1.5 million. ### The Regional Picture (Figure 29) - All IMF regions include a mix of economic groups and jurisdictions; performance against indices varies strongly by economic group and administration size. - Regional patterns (Figure 29, panel 1): - EUR (dominated by AEs) has the highest scores of the five regions. - AFR, APD, and WHD average scores are close and fall 10 to 20 percentage points below EUR scores. - MCD’s aggregate scores generally fall below EUR’s, but above AFR/APD/WHD; caution advised due to markedly lower ISORA participation from MCD, especially Middle East and North Africa subregion. - Within regions, group scores follow global patterns: in EUR and APD AEs score higher than EMEs; where FCS and SDS groups exist, their averages fall below LIDCs. Regional variation exists (for example, EMEs in Europe score higher than EMEs elsewhere). - Figure 29 panel data: index score ranges shown from 0 to 90 across the seven indices for All ISORA participants and by region (Africa, Asia Pacific, Europe, Middle East and Central Asia, Western Hemisphere). ### Associations with Operational Metrics - Purpose: explore correlations between the seven practice indices (and their average) and selected operational metrics associated with registration, filing, payment, arrears, and audit and verification using 2022 data. - Table 6: Correlations (Pearson correlation coefficients) of metrics with indices (columns: CM, DG, GV, HRMD, MA, PT, TS, Ave). *p < .05; **p < .01. - Active PIT taxpayers as percentage of Population: CM 0.16; DG 0.41**; GV 0.20*; HRMD 0.33**; MA 0.25**; PT 0.24**; TS 0.21**; Ave 0.31** - CIT returns filed on time as percent of expected returns: CM 0.40**; DG 0.50**; GV 0.39**; HRMD 0.32**; MA 0.29**; PT 0.48**; TS 0.43**; Ave 0.47** - PAYE returns filed on time as percent of expected returns: CM 0.31**; DG 0.43**; GV 0.34**; HRMD 0.29**; MA 0.24**; PT 0.35**; TS 0.42**; Ave 0.40** - VAT returns filed on time as percent of expected returns: CM 0.21*; DG 0.45**; GV 0.25**; HRMD 0.25**; MA 0.20*; PT 0.25**; TS 0.34**; Ave 0.34** - PIT returns filed on time as percent of expected returns: CM 0.19*; DG 0.36**; GV 0.19*; HRMD 0.19*; MA 0.10; PT 0.26**; TS 0.30**; Ave 0.26** - Percent CIT returns e-filed: CM 0.53**; DG 0.68**; GV 0.62**; HRMD 0.52**; MA 0.44**; PT 0.57**; TS 0.71**; Ave 0.67** - Percent PIT returns e-filed: CM 0.61**; DG 0.73**; GV 0.56**; HRMD 0.45**; MA 0.42**; PT 0.58**; TS 0.69**; Ave 0.66** - Percent PAYE returns e-filed: CM 0.60**; DG 0.73**; GV 0.66**; HRMD 0.57**; MA 0.51**; PT 0.60**; TS 0.72**; Ave 0.72** - Percent VAT returns e-filed: CM 0.47**; DG 0.64**; GV 0.64**; HRMD 0.55**; MA 0.52**; PT 0.56**; TS 0.65**; Ave 0.69** - Percent electronic payments by number: CM 0.40**; DG 0.59**; GV 0.55**; HRMD 0.55**; MA 0.47**; PT 0.48**; TS 0.56**; Ave 0.60** - Percent electronic payments by value: CM 0.43**; DG 0.54**; GV 0.59**; HRMD 0.51**; MA 0.51**; PT 0.47**; TS 0.58**; Ave 0.61** - Closing stock of arrears at year end as percentage of revenue collected: CM −0.22*; DG −0.21*; GV −0.30**; HRMD −0.24**; MA −0.23**; PT −0.24**; TS −0.30**; Ave −0.29** - CIT arrears as percentage of CIT collected: CM −0.38**; DG −0.32**; GV −0.44**; HRMD −0.36**; MA −0.28**; PT −0.44**; TS −0.46**; Ave −0.45** - PIT arrears as percentage of PIT collected: CM −0.30**; DG −0.30**; GV −0.41**; HRMD −0.35**; MA −0.29**; PT −0.43**; TS −0.39**; Ave −0.42** - PAYE arrears as percentage of PIT collected: CM −0.01; DG −0.14; GV −0.06; HRMD −0.02; MA −0.02; PT −0.18; TS −0.12; Ave −0.09 - VAT arrears as percentage of VAT collected: CM −0.15; DG −0.11; GV −0.16; HRMD −0.10; MA −0.02; PT −0.15; TS −0.20; Ave −0.15 - Audit hit rate: CM −0.10; DG −0.28**; GV −0.20*; HRMD −0.18*; MA −0.11; PT −0.18*; TS −0.26**; Ave −0.23** - Additional assessments raised through audits and verifications as percentage of collections: CM −0.15; DG −0.20*; GV −0.17; HRMD 0.02; MA −0.09; PT −0.13; TS −0.22**; Ave −0.15 - Key insights from operational correlations: - Signs of correlations align with expectations: positive where good practice should raise a metric (for example, on-time filing) and negative where good practice should lower a metric (for example, arrears). - Strength of correlations varies by index; stronger correlations generally seen for digitalization, taxpayer services, public trust, and governance. - Registration: All indices correlate reasonably strongly with active PIT taxpayers to population, strongest with digitalization. - On-time filing: Positive correlations across four core taxes, strongest for CIT (digitalization 0.50; public trust 0.49). - On-time payment: Correlations weaker than for on-time filing; fewer administrations provide payment figures. - E-filing and e-payment: E-filing rates for the four tax types have the strongest correlations with practice indices—around 0.7 for digitalization and taxpayer services; over 0.5 for governance and public trust. E-payment correlations also strong (coefficients over 0.5 for several indices). - Arrears: Negative correlations as expected; stronger for CIT and PIT arrears. Correlations with collectible arrears as percent of collections are lower. - Audit and verification: Correlations generally negative (−0.10 to −0.28), noting ambiguity in interpreting audit hit rate. Strongest correlation of assessments raised is with taxpayer services (negative), consistent with better taxpayer services increasing voluntary compliance and lowering assessments from interventions. - Dispute resolution correlations are of expected sign but small and not highly significant (internal review cases per taxpayer negative; percent of appellate cases found in favor of administration positive). ### Associations with Resources and Efficiency Measures - Purpose: examine associations between resource levels/allocation and practice indices, and relations between good practices and efficiency. - Table 7: Correlations with selected resource-related variables and efficiency measures (CM, DG, GV, HRMD, MA, PT, TS, Ave). *p < .05; **p < .01. - Operating cost as percent of GDP: CM −0.22*; DG −0.23**; GV −0.09; HRMD −0.07; MA −0.01; PT −0.10; TS −0.18*; Ave −0.14* - Percentage staff allocated to audit, investigation and other verification: CM 0.19*; DG 0.22**; GV 0.19*; HRMD 0.17*; MA 0.15*; PT 0.18*; TS 0.17*; Ave 0.21* - Percent staff with 20 or more years of service: CM 0.21*; DG 0.43**; GV 0.20*; HRMD 0.23**; MA 0.14; PT 0.23**; TS 0.25**; Ave 0.29** - Percent of staff with 10 or more years of service: CM 0.11; DG 0.31**; GV 0.11; HRMD 0.12; MA 0.07; PT 0.08; TS 0.12; Ave 0.16 - Percent staff with a degree: CM 0.41**; DG 0.43**; GV 0.40**; HRMD 0.28**; MA 0.28**; PT 0.33**; TS 0.50**; Ave 0.44** - Operating cost as percent of collections: CM −0.33**; DG −0.43**; GV −0.25**; HRMD −0.21*; MA −0.15; PT −0.26**; TS −0.34**; Ave −0.33** - Log active taxpayers per FTE (correlations based on 109 administrations providing active taxpayer figures for all core tax types): CM 0.26**; DG 0.40**; GV 0.35**; HRMD 0.31**; MA 0.23*; PT 0.40**; TS 0.35**; Ave 0.39** - Log returns received per FTE: CM 0.40**; DG 0.60**; GV 0.49**; HRMD 0.44**; MA 0.37**; PT 0.47**; TS 0.54**; Ave 0.55** - Key insights from resource/efficiency correlations: - Better practice indices tend to be associated with lower operating costs as percent of collections (notably digitalization −0.43**) and lower operating costs as percent of GDP (negative coefficients for several indices). - Higher shares of staff with degrees and longer-tenured staff correlate positively with many indices (percent staff with a degree: Ave 0.44**; percent staff with 20+ years: Ave 0.29**). - Digitalization and taxpayer services show strong positive correlations with measures of productivity per FTE (Log returns received per FTE: DG 0.60**; TS 0.54**; Ave 0.55**). - Allocation of staff to audit, investigation and verification shows small positive correlations with practice indices (Ave 0.21*). *Source: Authors’ calculations.* ### 2. These correlations are for the 118 administrations that provide the number of returns received for all the core tax t ### itappea - 2. These correlations are for the 118 administrations that provide the number of returns received for all the core tax types that they administer ### Correlations with financial resources and ICT - Correlations of financial resource measures with the practice indices are generally not large. - Operating expenditure (as percent of GDP): - Shows a pattern of negative correlations with four practice indices (while essentially zero for governance, HRMD, and management autonomy), indicating that the extent of good practices reported in ISORA is largely independent of funding levels measured relative to GDP. - Small administrations have higher levels of funding relative to GDP and generally have lower scores against the practice indices. - Using operating expenditure as percent of GDP, there is limited difference between the operating budgets of AEs and LIDCs. - Salary costs as percentage of operating expenditure: correlations with the indices are not significant. - ICT operating expenditure (as percent of operating expenditure, for FY 2022): correlations with the indices are not significant, including for digitalization. - ICT spending is volatile; using a five-year average (FY 2018 to FY 2022) for ICT expenditure both as percent of operating expenditure and as percent of GDP yields negligible correlation coefficients. - Implication: Investment in ICT may require a longer time horizon to influence practice indices, or the manner in which ICT funding is applied matters more than the quantum of spending. ### Human resources, staff allocation, and staff characteristics - Human resources levels: - Correlation coefficients between practice indices and population per FTE, labor force per FTE, and the logs of both variables do not exceed 0.2. - Conclusion: the level of resourcing (FTEs relative to population or labor force) does not show a strong relationship with extent of good practice. - Staff allocation: - Small but universally positive correlations exist between the seven indices and percent staff allocated to audit and verification; this may reflect a higher proportion of staff in audit and verification among AE administrations and warrants deeper scrutiny. - Correlations with staff allocated to all other functions are negligible, including percent of staff in headquarter roles. - Footnote: Values provided by ISORA respondents for the percentage of staff performing headquarters functions vary between 0.1 percent and 100 percent. - Staff characteristics: - Percentage of staff with 20 or more years of experience: - Strongly positively correlated with the digitalization index. - Correlations here are stronger than corresponding relationships using percentage of staff with 10 or more years of experience. - Percentage of staff with a degree: - Strongly positively associated with the taxpayer services index. ### Efficiency measures and digitalization - Correlations of efficiency measures (Figure 25) with the seven practice indices have expected signs: - Cost of collection: negative correlations with the practice indices. - Core tax returns received per FTE and core active taxpayers per FTE: positive correlations with the practice indices. - Magnitude ranking of correlation coefficients (largest to smaller): digitalization, followed by taxpayer services, public trust, and governance. - Causality is unclear: greater digitalization may be driven by the requirement to increase efficiency, or digitalization may itself drive greater efficiency. ### Associations between the indices and revenue collection, tax administration effectiveness, and public administration strength - Revenue collection measures assessed: - Revenue collected by the tax administration as percent of GDP, and revenue collected as percent of government revenue. - These measures are influenced by tax policy choices and which revenues governments assign to central tax administrations. - Tax administration effectiveness measures: - Tax effort (ratio of actual collections to potential collections), calculated through a stochastic frontier model. - VAT compliance gap (difference between actual VAT collections and VAT that would be collected under existing policy with full compliance). - Public administration and governance measures: - Blavatnik Index of Public Administration (2024) used as a measure of public administration strength. - World Bank Governance Indicators: Government Effectiveness and Regulatory Quality. - Correlations reported in Table 8 (selected highlights preserved exactly as in source): - Revenue collected as percent of GDP: - CM 0.25**; DG 0.44**; GV 0.27**; HRMD 0.22**; MA 0.22**; PT 0.33**; TS 0.33**; Ave 0.35** - Revenue collected as percent of government revenue: - CM 0.40**; DG 0.45**; GV 0.38***; HRMD 0.26**; MA 0.27**; PT 0.38**; TS 0.41**; Ave 0.42** - Tax effort (113 administrations): - CM 0 .15; DG 0.36**; GV 0.30**; HRMD 0.45**; MA 0.33**; PT 0.29**; TS 0.26**; Ave 0.41** - VAT compliance gap (37 administrations): - CM −0.10; DG −0.32*; GV −0.19; HRMD −0.07; MA −0.13; PT −0.27; TS −0.12; Ave −0.21 - Data availability rate: - CM 0.29**; DG 0.27**; GV 0.23**; HRMD 0.26**; MA 0.22**; PT 0.29**; TS 0.24**; Ave 0.30** - Blavatnik Public Administration Index (110 countries): - CM 0.41**; DG 0.59**; GV 0.35**; HRMD 0.47**; MA 0.21*; PT 0.43**; TS 0.52**; Ave 0.57** - Government Effectiveness (159 countries): - CM 0.20*; DG 0.34**; GV 0.23**; HRMD 0.39**; MA 0.22**; PT 0.22**; TS 0.24**; Ave 0.32** - Regulatory Quality (159 countries): - CM 0.23**; DG 0.38**; GV 0.27**; HRMD 0.41**; MA 0.26**; PT 0.27**; TS 0.28**; Ave 0.36** - Note: Ave = average; CM = compliance management; DG = digitalization; GV = governance; HRMD = human resources management and development; MA = management autonomy; PT = public trust; RA-GAP = estimates of the VAT gap using the IMF’s Revenue Administration Gap Analysis Program methodology; TS = taxpayer services; VAT = value-added tax. *p < .05; **p < .01. - Key interpretive points: - Positive correlations are observed between all seven practice indices and macrocollection variables despite other influencing factors (economic factors, tax policy). - The strongest correlations against both revenue measures are with the digitalization index, followed by taxpayer services; correlations are significant against all seven indices. - Tax effort correlates positively with the indices, with the strongest association against human resources management and development, followed by digitalization. - All indices correlate negatively with the VAT compliance gap; only the correlations with digitalization and public trust are significant given limited sample size for the VAT gap. - All correlations with the Blavatnik Index are positive, strongest for digitalization (0.59), taxpayer services (0.52), and HRMD (0.47); all four Blavatnik subindices correlate positively with the seven ISORA indices. - Government Effectiveness and Regulatory Quality show positive correlations with the tax administration practice indices; the strongest correlations are with human resources management and development and with digitalization. *Source: ISORA 2023 analysis and authors’ calculations as presented in the provided content.* ### Annex Table 2.1 Comparative Quantitative Measures of Tax Administration Performance (Continued) ### Annex Table 2.1 Comparative Quantitative Measures of Tax Administration Performance (Continued) ### Annex 3. Data Availability Rates - Completeness levels for numerical questions with lowest completeness in each thematic grouping are presented; lower availability affects completeness of comparative quantitative indicators derived from responses. - Annex Table 3.1: Numerical Questions with Lower Completeness Levels in ISORA 2023 — Thematic grouping, questions, lowest completeness level (percent), and notes: - Revenue collections - VAT collected on imports; VAT refunds — 86 - Resources - ICT expenditures — 75 - Note: In some cases, ICT expenditure is not part of the budget allocation to the tax administration. - Segmentation - Revenue administered under HNWI program as percent of total revenue collected — 58 - Note: Some administrations indicate these figures are not available as their HNWI program is not yet fully developed. - Registration - Taxpayers (domestic manufacturers/producers) registered for excise; Active domestic excise taxpayers — 73 - Filing - Returns expected from employers that withhold tax from employees; Returns received from employers; Returns received on time from employers; Return channels used by employers — 73 - Payment - Value of expected PAYE payments; On-time payments for PIT — 53 - Arrears - Arrears associated with state-owned enterprises; Arrears associated with state-owned enterprises considered noncollectible; Arrears associated with employers (65) — 44 - Note: Arrears data points requested have been streamlined and clarified in the past five years, but improvements in data completion levels have been limited. It is unclear whether a reluctance to be transparent about these figures is a factor limiting provision of data, or whether they are unavailable within the administration. - Audit and verification - Additional assessments raised from audits of employers; Additional assessments raised from PIT (62) — 48 - Note: Data inputs for audit and verification have been streamlined and clarified in past five years, but improvement in completion levels has been limited. It is unclear whether a reluctance to be transparent about these figures is a factor limiting provision of the data, or whether they are unavailable within the administration. - Disputes - Number of cases resolved by a higher appellate court; Number of cases resolved by a higher appellate court in favor of the administration — 73 - Note: Fewer administrations are able to provide data about cases outside the administration. - Incoming service channels - Number of incoming service contacts by mail/post — 50 ### Annex 4. Accessing International Survey on Revenue Administration (ISORA) Data and Metadata - ISORA data publication timing and access: - ISORA data are published on the ISORA data portal annually after data collection and data quality review are completed, generally in June of the year after collection. - Getting to the portal: - Type in isoradata.org in your browser. - ISORA data can also be found from the data.imf.org portal. Use the search box to search for “ISORA.” - Accessing ISORA data (formats and tools): - ISORA data available as preformatted tables, customized tables, PowerBI tables, full dataset downloads in CSV or SDMX. - The Latest Review, Derived, and Periodic Tables: - Preformatted tables cover countries for FY2018 and later years; Review Tables display reported data with some revisions; Derived Tables have comparative indicators calculated using reported data. The most recent set of Periodic Tables is for FY 2022. - Links may change; find via ISORA portal search with terms ‘ISORA review tables’, ‘ISORA derived tables’, ‘ISORA periodic tables’. - Data Explorer (https://data.imf.org/en/Data-Explorer?datasetUrn=ISORA:ISORA_LATEST_DATA_PUB(4.0.0)): - Access via “View Data” on ISORA portal home page. iData’s Data Explorer customizes tables and selects needed data; default shows latest data set. - Selection Panel allows country, indicator, and time period selection. - With “Full List” checked under “Display All/By Table”, multiple countries can be selected for comparison. - Additional details via “Set up columns”; selected data downloadable in CSV or SDMX; whole data set download option available. - PowerBI Dashboard (https://data.imf.org/en/dashboards/isora%20data%20dashboard): - Use to access ISORA data by country with customized and thematic tables. - Option 1: Navigate to Latest Data and drill through to country data tables. - Option 2: Navigate directly to customized or thematic tables and use “Filters” to select countries, table, and indicators. - ISORA Forms, Guides, and Publications: - Documents page is a repository of ISORA-related documents including questionnaires, guides to complete questionnaires, and various publications using ISORA data. - Find the page by clicking “Documents Catalog” and selecting ISORA under data sets. ### Annex 5. Selected Comparative Aggregate Statistics - Overview: - Aggregate statistics by economic and regional groups and IMF regional capacity development centers are tabulated for selected indicators used in Chapters 3 to 5. All statistics drawn from published ISORA responses for FY22, except where indicated otherwise. - Note: Statistics for METAC are not provided as only one METAC member completed ISORA 2023. - Annex Table 5.1. ISORA 2023 Participation, Revenue Collection, and Resource Ratios — Key group medians/averages: - All ISORA participants: - ISORA Completion — 166 - GDP per Capita (1000s USD) — 7.9 - General Government Revenue as Percent of GDP — 30.4 - Revenue Collected to GDP — 17. 3 - Resource Ratios: - FTEs per 1,000 Labor Force — 0.9 - Core Tax Returns per FTE — 731.5 - Operating Budget as Percent of Revenue Collected — 0.8 - Salary as Percent of Operating Budget — 71.7 - ICT Expenditure as Percent of Total Budget — 5.8 - Economic groups (selected): - Advanced economies: - ISORA Completion — 37 - GDP per Capita (1000s USD) — 49.1 - General Government Revenue as Percent of GDP — 4 0 .1 - Revenue Collected to GDP — 24.0 - FTEs per 1,000 Labor Force — 1.6 - Core Tax Returns per FTE — 1614 . 3 - Operating Budget as Percent of Revenue Collected — 0.6 - Salary as Percent of Operating Budget — 72.2 - ICT Expenditure as Percent of Total Budget — 9.0 - Emerging market economies: - ISORA Completion — 83 - GDP per Capita (1000s USD) — 8.2 - General Government Revenue as Percent of GDP — 3 0 .1 - Revenue Collected to GDP — 16 .1 - FTEs per 1,000 Labor Force — 1.1 - Core Tax Returns per FTE — 659. 2 - Operating Budget as Percent of Revenue Collected — 0.8 - Salary as Percent of Operating Budget — 73.4 - ICT Expenditure as Percent of Total Budget — 5.3 - Low-income developing countries: - ISORA Completion — 46 - GDP per Capita (1000s USD) — 1.6 - General Government Revenue as Percent of GDP — 23.2 - Revenue Collected to GDP — 13. 2 - FTEs per 1,000 Labor Force — 0.3 - Core Tax Returns per FTE — 168.9 - Operating Budget as Percent of Revenue Collected — 1.1 - Salary as Percent of Operating Budget — 63.5 - ICT Expenditure as Percent of Total Budget — 4.8 - Regional groups (selected): - Africa: - ISORA Completion — 40 - GDP per Capita (1000s USD) — 1.5 - General Government Revenue as Percent of GDP — 20.9 - Revenue Collected to GDP — 11. 6 - FTEs per 1,000 Labor Force — 0.3 - Core Tax Returns per FTE — 120.4 - Operating Budget as Percent of Revenue Collected — 1.7 - Salary as Percent of Operating Budget — 69. 5 - ICT Expenditure as Percent of Total Budget — 4.2 - Europe: - ISORA Completion — 43 - GDP per Capita (1000s USD) — 30.2 - General Government Revenue as Percent of GDP — 40.9 - Revenue Collected to GDP — 25.3 - FTEs per 1,000 Labor Force — 1.7 - Core Tax Returns per FTE — 1526 .1 - Operating Budget as Percent of Revenue Collected — 0.6 - Salary as Percent of Operating Budget — 73.2 - ICT Expenditure as Percent of Total Budget — 7. 8 - Annex Table 5.2. FTE and Selected Staff Metrics — Selected averages (percent): - All ISORA participants — FTEs by Function: - Taxpayer Services, Returns, Payment — 29. 3 - Audit — 24.9 - Arrears — 11. 5 - Staff by Age: - 34 and under — 26 .1 - 35–54 — 52.3 - 55 and over — 21.5 - Staff by Gender: - Female — 47.7 - Male — 52.3 - Female Executives (percent of executives) — 40.4 - Economic groups (examples): - Advanced economies — Taxpayer Services, Returns, Payment — 32.2; Audit — 29.0; Arrears — 10.5; 34 and under — 18.2; 35–54 — 51.3; 55 and over — 30.5; Female — 60.5; Male — 39.5; Female Executives — 46.6 - Emerging market economies — Taxpayer Services, Returns, Payment — 26.2; Audit — 26.1; Arrears — 12.3; 34 and under — 28.0; 35–54 — 55.7; 55 and over — 16.3; Female — 56.6; Male — 43.4; Female Executives — 44.9 - Annex Table 5.3. Large Taxpayer Office/Program and HNWI Unit/Program — Selected indicators: - All ISORA participants: - Has LTO/P? (as percent of participants in group) — 8 6 .1 - FTEs as Percent of Total FTEs — 6.8 - Revenue as Percent of Total Revenue — 53.8 - Number of Audits as Percent of Total Audits — 14.6 - Additional Assessments Raised through LTO/P as Percentage of Total Value of Additional Assessments Raised from Audits — 46.4 - Has HNWI Unit/Program? (as percent of participants in group) — 28.3; Part of LTO/P? — 29 of 47 - Regional groups (examples): - Africa — Has LTO/P? — 95.0; FTEs as Percent of Total FTEs — 8.5; Revenue as Percent of Total Revenue — 65.4; Number of Audits as Percent of Total Audits — 28.2; Additional Assessments Raised through LTO/P — 59.0; Has HNWI Unit/Program? — 30.0; Part of LTO/P? — 7 of 12 - Europe — Has LTO/P? — 86.0; FTEs as Percent of Total FTEs — 4.0; Revenue as Percent of Total Revenue — 40.9; Number of Audits as Percent of Total Audits — 4.8; Additional Assessments Raised through LTO/P — 25.4; Has HNWI Unit/Program? — 30.2; Part of LTO/P? — 8 of 13 - Annex Table 5.4. On-Time Filing, On-Time Payment, and Registration — Selected averages/medians: - All ISORA participants — Percent of Returns Received on Time: - PIT — 65.3 - CIT — 58.5 - VAT — 76 .1 - Employer Withholding — 6 8 .1 - Percent of Payments Received on Time: - PIT — 69.0 - CIT — 81.0 - VAT — 81.2 - Employer Withholding — 81.5 - Active Taxpayers (percent) — medians: - PIT to Labor Force — 33.2 - CIT to Labor Force — 4.3 - VAT to Labor Force — 3.5 - Economic group highlights: - Advanced economies — Percent of Returns Received on Time: PIT 91.7; CIT 81.4; VAT 87. 5; Employer Withholding 88.9. Percent of Payments Received on Time: PIT 83.3; CIT 8 8 .1; VAT 8 6 .1; Employer Withholding 9 0 .1. Active Taxpayers (percent): PIT to Labor Force 122.8; CIT to Labor Force 10.7; VAT to Labor Force 15.9 - Low-income developing countries — Percent of Returns Received on Time: PIT 47. 4; CIT 45.9; VAT 62.7; Employer Withholding 54.4. Percent of Payments Received on Time: PIT 6 3 .1; CIT 74.0; VAT 72.0; Employer Withholding 65.2. Active Taxpayers (percent): PIT to Labor Force 0.9; CIT to Labor Force 0.3; VAT to Labor Force 0.2 - Annex Table 5.5. Electronic Filing and Electronic Payment — Availability and take-up: - All ISORA participants — Number Where E-Filing Is Available (counts): - PIT — 134 of 159 - CIT — 137 of 160 - VAT — 132 of 148 - PAYE — 133 of 155 - E-Filing Rate (percent): PIT 85.6; CIT 88.3; VAT 90.5; PAYE 89. 8 - E-Payment Availability (as percent of participants in group) — PIT 75.5 - Economic groups (examples): - Advanced economies — Number Where E-Filing Is Available: PIT 37 of 37; CIT 37 of 37; VAT 35 of 35; PAYE 34 of 34. E-Filing Rate (percent): PIT 86.6; CIT 93.0; VAT 9 7.7; PAYE 91.9. E-Payment (percent, by value) — 87. 8 - Fragile and conflict-affected states — Number Where E-Filing Is Available: PIT 14 of 23; CIT 13 of 22; VAT 11 of 19; PAYE 12 of 22. E-Filing Rate (percent): PIT 44.4; CIT 49.0; VAT 45.6; PAYE 82.6. E-Payment (percent, by value) — 56.4 - Regional and RCDC examples: - Europe — E-Filing Rate (percent): PIT 8 6 .1; CIT 9 7.7; VAT 98.6; PAYE 90.7. E-Payment (percent, by value) — 96.2 - Economies not in RCDCs — E-Filing Rate (percent): PIT 87. 8; CIT 95.3; VAT 98.2; PAYE 91.8. E-Payment (percent, by value) — 88.0 *Source: CIAT, IMF, IOTA, and OECD; International Survey on Revenue Administration; data tables for FY2018 to FY2022; thematic tables cited in source content.* ### Annex Table 5.5. Electronic Filing and Electronic Payment (Continued) ### Annex Table 5.5. Electronic Filing and Electronic Payment (Continued) ### Arrears and Audit (Annex Table 5.6) - Indicator definitions: - “Arrears as Percent of Revenue Collected” = median of the group. - “Audit Assessments Raised As Percent of Revenue Collected” = average. - All ISORA participants: - Arrears As Percent of Revenue Collected: 18.3 - Audit Assessments Raised As Percent of Revenue Collected: 4.2 - Economic groups: - Advanced economies: Arrears 8.0; Audit assessments 2.3 - Emerging market economies: Arrears 31.3; Audit assessments 3.5 - Low-income developing countries: Arrears 14.7; Audit assessments 7.5 - Special groups: - Fragile and conflict-affected states: Arrears 34.3; Audit assessments 4.1 - Small developing states: Arrears 57.8; Audit assessments 2.6 - Regional groups: - Africa: Arrears 27.3; Audit assessments 7.5 - Asia Pacific: Arrears 12.9; Audit assessments 3.1 - Europe: Arrears 10.2; Audit assessments 2.6 - Middle East and Central Asia: Arrears 14.9; Audit assessments 4.6 - Western Hemisphere: Arrears 47.7; Audit assessments 4.2 - RCDC groupings and selected entries: - AFC: Arrears 41.3; Audit assessments 15.8 - AFE: Arrears 31.1; Audit assessments 16.5 - AFS: Arrears 29.0; Audit assessments 4.7 - AFW: Arrears 13.8; Audit assessments 6.6 - AFW2: Arrears 11.0; Audit assessments 4.0 - CC AMTAC: Arrears 12.0; Audit assessments 3.6 - PF TAC: Arrears 19.0; Audit assessments 1.2 - SAR T TAC: Arrears 65.9; Audit assessments 6.7 - C AP TAC- DR3 4 .1: Arrears 12.6 (entry formatting preserved as in source) - C AR TAC: Arrears 78.2; Audit assessments 1.7 - Economies not in RCDCs: Arrears 9.8; Audit assessments 3.0 - Sources: CIAT, IMF, IOTA, and OECD; International Survey on Revenue Administration, https://data.rafit.org; data tables for FY2018 to FY2022; Thematic Table D.41 and D.46. - Notes: ISORA = International Survey on Revenue Administration; RCDCs = regional capacity development centers. ### The Seven Indices and Data Availability Rate (Annex Table 5.7) - Indices and Data Availability Rate for All ISORA participants: - Data Availability: 61.8 - Compliance Management: 56.7 - Digitalization: 73.0 - Governance: 63.0 - HR Management and Development: 71.6 - Management Autonomy: 46.7 - Public Trust: 62.9 - Taxpayer Services: 81.0 - Selected economic group scores: - Advanced economies: Data Availability 70.1; Compliance Management 75.4; Digitalization 83.0; Governance 81.0; HR Management and Development 84.2; Management Autonomy 63.6; Public Trust 74.3; Taxpayer Services 86.5 - Emerging market economies: Data Availability 59.2; Compliance Management 52.7; Digitalization 69.6; Governance 58.6; HR Management and Development 68.7; Management Autonomy 41.5; Public Trust 58.9; Taxpayer Services 83.0 - Low-income developing countries: Data Availability 59.8; Compliance Management 49.1; Digitalization 71.3; Governance 56.5; HR Management and Development 66.7; Management Autonomy 42.6; Public Trust 61.1; Taxpayer Services 73.0 - Selected regional group scores: - Africa: Data Availability 60.2; Compliance Management 48.2; Digitalization 71.0; Governance 57.0; HR Management and Development 69.5; Management Autonomy 43.6; Public Trust 58.9; Taxpayer Services 71.6 - Europe: Data Availability 67.8; Compliance Management 71.2; Digitalization 81.8; Governance 74.5; HR Management and Development 80.8; Management Autonomy 61.4; Public Trust 72.0; Taxpayer Services 85.6 - Western Hemisphere: Data Availability 58.9; Compliance Management 54.2; Digitalization 66.8; Governance 56.5; HR Management and Development 66.1; Management Autonomy 39.6; Public Trust 57.8; Taxpayer Services 88.4 - RCDC entries (selected): - AFC: Data Availability 56.5; Compliance Management 34.8; Digitalization 52.5; Governance 32.0; HR Management and Development 48.1; Management Autonomy 25.9; Public Trust 41.1; Taxpayer Services 51.9 - AFE: Data Availability 79.0; Compliance Management 63.8; Digitalization 89.3; Governance 85.5; HR Management and Development 89.8; Management Autonomy 51.4; Public Trust 84.9; Taxpayer Services 82.2 - AFW2: Data Availability 66.8; Compliance Management 57.7; Digitalization 83.5; Governance 77.7; HR Management and Development 84.6; Management Autonomy 58.0; Public Trust 64.4; Taxpayer Services 87.9 - Economies not in RCDCs: Data Availability 69.1; Compliance Management 71.4; Digitalization 82.4; Governance 77.0; HR Management and Development 81.1; Management Autonomy 58.7; Public Trust 73.9; Taxpayer Services 83.6 - Note: All indicator values presented in the table are average values for the respective groups. HR = human resources; ISORA = International Survey on Revenue Administration; RCDCs = regional capacity development centers. ### Annex 6 — Use of ISORA Data in Indices (Annex Table 6.1) - Summary of indices constructed from ISORA data (selected entries): - IMF: Crandall, Gavin, and Masters (2019) - ISORA 2016; 39 categorical and yes/no data points for FY 2015. - Four independent indices, equally weighted: meeting performance standards; management and HR autonomy; public accountability; service orientation. - Purpose: Comparison of groups of tax administrations (by income, institutional arrangements, small states, fragile states). - OECD (2019) - ISORA 2018; 15 yes/no data points for FY 2017. - Innovation index, equally weighted. - Purpose: Innovation level benchmark for 58 administrations for FY 2017. - IMF: Chang and others (2020) - ISORA 2016 and 2018; 66 yes/no data points for FY 2014 to FY 2017. - OSI composed of seven equally weighted subindices (compliance risk management; use of third-party data; degree of digitalization; service orientation; public accountability; autonomy; LTO/P). - Purpose: Identify relationship between OSI and revenue collected, controlling for tax policy and macro factors. - IMF: Crandall, Gavin, and Masters (2021) - ISORA 2018; Seven indices composed of weighted elements (management and human resources autonomy; public accountability; general management; human resources management; service orientation; compliance risk-management foundations; degree of digitalization). - Purpose: Compare groups of tax administrations; examine correlations with performance and efficiency measures. - CIAT: Díaz de Sarralde Miguez and Morán (2022) - ISORA 2020; 9 categorical and 7 yes/no data points; 14 quantitative indicators. - INDITEC with four equally weighted subindices (technological innovation; compliance improvement; operational digitalization; resources and budget). - Purpose: Benchmarking tax administrations internationally. - OECD, Tax Administration (2024) - ISORA 2023; high-level description: weighted averages to create indices for control/oversight features and people management. - Purpose: Provide high-level picture for FY22; distribution of scores presented; individual scores not provided. - CIAT: Garcimartín and Díaz de Sarralde Miguez (2025) - ISORA 2023; 156 variables mostly yes/no; seven quantitative indicators mapped to three-point scale. - IFIAT compiled from five dimensions using principal component analysis weights (management autonomy; governance: control and planning; personnel management; taxpayer relations; available resources). - Purpose: Measure institutional strength; group statistics and individual scores tabulated; correlation with performance indicators calculated. - IMF: Atsebi, Gueorguiev, and Nose (2025) - ISORA 2023 and earlier vintages; 193 categorical and yes/no data points. - OSI compiled from nine subindices weighted by a Delphi approach: compliance risk management; use of third-party data; degree of digitalization; service orientation; public accountability; autonomy; LTO and HNWI; tax enforcement; human resources management and development. - Purpose: Develop granular OSI to estimate revenue gains from tax administration reforms across fiscal years. - IMF (2025) - ISORA 2023; 239 categorical or yes/no data points. - Seven indices compiled from weighted components: compliance management; digitalization; governance; human resources management and development; management autonomy; public trust; taxpayer services. Simple unweighted average (normalized 1-to-100) also calculated. - Purpose: Comparative analysis of groups of tax administrations; explore relationship between indices and resourcing and performance. - Notes and abbreviations provided in source: CIAT; FTE; FY; HR; HNWI; IFIAT; IMF; INDITEC; ISORA; LTO; LTO/P; OECD; OSI. ### Annex 7 — Structure of the Seven FY2022 Indices and Examples of Responses - Purpose: Provide index components and weightings used in Chapter 5; references to ISORA 2023 questionnaire questions. - Compliance Management (Annex Table 7.1) — selected components, questions, and weights: - Formal strategy and approach - Does the administration have a formal compliance risk-management strategy? EPP6 a) Weight 3 - Is there a formal approach for assessing and prioritizing key compliance risks for two or more of the following? EPP6 b) Weight 3 - i) Return filing EPP6 b) i) (component within above) - ii) Payment processing EPP6 b) i) - iii) Collection enforcement EPP6 b) i) - iv) Verification/audit EPP6 b) i) - v) Taxpayer services EPP6 b) i) - Does the administration’s general approach to compliance include: EPP11 a) (i) Providing a proportionate response to noncompliance EPP11 a) i) Weight 0.5; (ii) Providing opportunities for taxpayers to correct mistakes or omissions in tax returns voluntarily EPP11 a) ii) Weight 0.5; (iii) Ensuring taxpayers benefit if they make early disclosure of errors or omissions and if they fully cooperate with the administration EPP11 a) iii) Weight 0.5; (iv) Reminding and encouraging taxpayers to fulfill all tax obligations EPP11 a) iv) Weight 0.5 - Does the administration make key compliance risks public regularly? EPP6 c) Weight 2 - Does the administration make reports of outcomes in addressing compliance risks public regularly? EPP6 e) Weight 2 - Are tax gaps produced for any tax type? EPP10 Weight 2 - Data used in managing compliance - Are electronic invoices used to monitor compliance? GQ4 a) Weight 1 - Are there computer-based systems for managing third-party data? GPP11 components (customs data, stock exchanges, Social Security Agency, online vendors, utilities, property ownership and sales) each Weight 0.5 - Does your administration undertake fully automated compliance checks based on data matching or analysis? FPP4 a) Weight 1 - Are the following payments to individuals reported to the administration? GPP10 components (wage and salary 1; dividends 0.5; interest 0.5; rents 0.5; specified business income 0.5; royalties, patents 0.5; proceeds of sales/purchase of shares 0.5; proceeds of sales/purchases of real estate 0.5; other income 0.5) - Are automated verifications undertaken? FQ8 b) Weight 1 - Segmentation and structure - Does the administration have a large taxpayer office/program? AQ3 a) Weight 2 - Does the administration have an office/program to manage HNWI? AQ3 b) Weight 2 - Does the administration have specialized audit staff for international tax issues? EPP9 a) Weight 1 - Does the administration enter into APAs? EPP9 b) Weight 1 - Interventions and audits - Formal framework for compliance interventions? EPP11 b) Weight 2 - Pre-filing interventions EPP12 (facilitating taxpayer access to third-party data EPP12 a) 0.5; targeting campaigns EPP12 b) 0.5; reminding of filing deadlines EPP12 c) 0.5; providing information on predetermined compliance interventions EPP12 d) 0.5) - Post-filing but pre-audit interventions EPP13 (identifying inconsistencies EPP13 a) 1; identifying anomalies EPP13 b) 1; risk reviews EPP13 c) 1; requesting further information EPP13 d) 1) - Types of audits after filing EPP15 a): desk audits 1; single-issue audits 1; limited-scope audits 1; comprehensive audits 1; avoidance and evasion investigations 1 - Indicators to measure effectiveness of enforcement EPP15 b): audit coverage EPP15 b) i) 1; value of additional assessments EPP15 b) ii) 1; percentage of audits yielding positive result EPP15 b) iii) 1 - Does the administration measure effectiveness of interventions? EPP14 a) Weight 1 - Standards for auditor productivity? EPP15 c) Weight 1 - Are random audits conducted? FPP5 a) Weight 1 - Penalties and sanctions - Administrative sanctions for nondisclosure? FPP7 a) Weight 1 - Common administrative penalty framework across major tax types? FPP7 b) i) Weight 0.5 - Do penalties consider taxpayers’ culpability? FPP7 b) ii) Weight 0.5 - Administration empowered to make public details of taxpayers subject to administrative penalties? FPP7 b) iv) Weight 0.5 - Weighted maximum score: 52.5 - Notes: CIT = corporate income tax; HNWI = high-net-wealth individual; PIT = personal income tax; VAT = value-added tax. - Digitalization (Annex Table 7.2) — selected components, questions, and weights: - Innovation and advanced technologies (GQ7) - Distributed ledger technology/blockchain GQ7 a) Weight 0.3 - AI, including machine learning GQ7 b) Weight 0.3 - Cloud computing GQ7 c) Weight 0.3 - Data science/analytics tools GQ7 d) Weight 0.3 - RPA GQ7 e) Weight 0.3 - APIs GQ7 f) Weight 0.3 - Whole-of-government identification systems GQ7 g) Weight 0.3 - Digital identification technology (biometrics, voice identification) GQ7 h) Weight 0.3 - Virtual assistants (chatbots) GQ7 i) Weight 0.3 - Data analytics and automation - Identify inconsistencies via crossmatching of third-party sources EPP13 a) Weight 2 - Identify anomalies or outliers through data analytics EPP13 b) Weight 2 - Fully automated compliance checks FPP4 a) Weight 2 (and if yes, automatic communication to taxpayer FPP4 b) Weight 1) - Use predictive modeling to initiate verification/audit FPP6 b) Weight 2 - Automated verifications FQ8 b) Weight 2 - Registration, filing, and payment - Is an e-registration channel available? GQ1 i) Weight 1 - Are e-filing channels used for any tax types? FQ2 d) ii)–iv) Weight 3 - Does the administration provide percentage of electronic payments in number and value? FQ5 Weight 3 - Does the administration prefill any income or expense items in tax returns? GQ32 - Digital services (GPP3) - Tax calculators, requesting extensions, payment arrangements, secure communication, filing objections, uploading data files, taxpayer portal, view third-party info: each GPP3 subcomponent Weight 0.3 - Mobile applications to access taxpayer account information GPP3 i) Weight 0.3 - Special services for digitally disadvantaged taxpayers GPP2 f) Weight 1 - Third-party data management - Computer-based systems for managing third-party data GPP11 components (customs, stock exchanges, Social Security Agency, online vendors, utilities, property) each Weight 0.5 - Receive data from devices that register transactions? GQ5 a) Weight 2 - Facilitate taxpayer access to third-party data before filing? EPP12 a) Weight 1 - Weighted maximum score: 33 - Note: AI = Artificial intelligence; APIs = Application programming interfaces; RPA = Robotic process automation. *Italic source attribution: Annex Table 5.5. Electronic Filing and Electronic Payment (Continued); source content from provided IMF PDF chapter/section excerpts.* ### Annex Table 7.2. Digitalization (Continued) ### Annex Table 7.2. Digitalization (Continued) ### Governance - Weighted maximum score: 77 - Components, question identifiers, and weights: - Produce its strategic plan? CPP5 a) 3 - Produce its annual business/operations plans? CPP5 b) 3 - Produce a formal set of service delivery standards? CPP5 d) 3 - Produce its annual report? CPP5 c) 3 - Formal compliance risk-management strategy? EPP6 a) 3 - Make key compliance risks public regularly? EPP6 c) 2 - Make reports of outcomes in addressing compliance risks public regularly? EPP6 e) 2 - Formal taxpayer service and assistance strategy? GPP6 a) 3 - Publish results of taxpayer satisfaction surveys for: - Individual taxpayer? GPP7 a) 2 - Business taxpayer? GPP7 b) 2 - Document that formally sets out taxpayer rights? GPP4 a) 3 - Mechanism for managing taxpayer complaints: - Internal? GPP5 a) 2 - External? GPP5 b) 2 - Are rulings provided to taxpayers? GPP9 a) 3 - Publish organizational chart? CPP7 a) 2 - Formal diversity policy included in HRM approach? DPP6 h) 3 - Authority regarding staff: - Make appointments of new staff? DPP1 b) 3 - Decide on promotion of existing staff? DPP1 c) 3 - Decide skills and qualifications required for appointment or promotion? DPP1 d) 2 - Terminate employment? DPP1 g) 3 - Apply disciplinary sanctions? DPP1 h) 3 - HR strategy? DPP6 a) 3 - Public service–wide code of conduct that applies to staff? CPP4 c) 2 - Administration’s own code of conduct? CPP4 d) (weight not given) - Determine its own internal structure? CPP3 a) 3 - Exercise discretion over the operating budget? CPP3 b) 3 - Exercise discretion over the capital budget? CPP3 c) 2 - Establish performance standards? CPP3 d) 3 - Internal audit (assurance) function? CPP4 b) 3 - Integrity strategy? CPP4 e) 3 ### Human Resources Management and Development - Weighted maximum score: 50 - Formal strategies and policies (components, identifiers, weights): - HR strategy/multiyear work force plan? DPP6 a) 3 - Training strategy in place? DPP6 b) 2 - Specific leadership and talent management programs? DPP6 f) i) 2 - Diversity policy? DPP6 h) 3 - Staffing plan? DPP6 d) 2 - Recruitment plan? DPP6 d) i) 2 - Policies for flexible working arrangements? DPP6 e) 1 - Time-reporting system? DPP6 g) 1 - Formal program(s) to support new staff? DPP6 c) 1 - Formal induction program for new staff? DPP6 c) i) 1 - Formal mentoring program for new staff? DPP6 c) ii) 1 - Formal training cycle process? DPP6 b) i) 1 - Specific training plan? DPP6 b) ii) 1 - Assess current and future capability needs? DPP7 1 - Formal plan to address gaps in staff capability? DPP7 iv) 1 - Knowledge transfer personalized (one-to-one)? DPP6 f) ii) 1 - Knowledge transfer documented (products: documents or videos)? DPP6 f) iii) 1 - Either a public service–wide code of conduct that applies to staff (CPP4 c) or its own code of conduct (CPP4 d) — presence of either counts 1 - Flexible work options: - Work flexible working hours? DPP6 e) i) 1 - Work from home/elsewhere occasionally? DPP6 e) ii) 1 - Work from home/elsewhere regularly? DPP6 e) iii) 1 - Executives work part-time? DPP6 e) iv) 1 - Staff engagement: - Periodically survey staff on diversity and inclusion? DPP5 b) 1 - Periodically survey staff (must include all elements): - Assess staff engagement DPP5 a) i) - Share results with staff DPP5 a) ii) - Engage staff in action plans DPP5 a) iii) — combined weight 2 - Authorities of the administration (weights): - Make appointments of new staff? DPP1 b) 3 - Decide on promotion of existing staff? DPP1 c) 3 - Decide skills and qualifications required for appointment or promotion? DPP1 d) 2 - Terminate employment? DPP1 g) 3 - Apply disciplinary sanctions? DPP1 h) 3 - Performance management: - Staff performance management system that includes: - Individual plans DPP3 a) - Specific objectives DPP3 b) - Annual evaluation DPP3 c) — combined weight 2 - Is performance linked to pay? DPP4 1 - More pay for good performance DPP4 a) - Either reduced pay for poor performance DPP4 b) or denial of increment for poor performance DPP4 c) ### Management Autonomy - Weighted maximum score: 27 - Components and weights: - Determine (design) its own internal structure? CPP3 a) 3 - Exercise discretion over the operating budget? CPP3 b) 3 - Exercise discretion over the capital budget? CPP3 c) 2 - Establish performance standards? CPP3 d) 2 - Determine work requirements? DPP1 a) 2 - Make appointments of new staff? DPP1 b) 3 - Decide on promotion of existing staff? DPP1 c) 3 - Decide skills and qualifications required for appointment or promotion? DPP1 d) 2 - Place staff within a salary range? DPP1 f) 1 - Terminate employment? DPP1 g) 3 - Apply disciplinary sanctions? DPP1 h) 2 - Are pay scales at the administration independent of civil/public sector pay scales? DPP2 1 ### Public Trust - Weighted maximum score: 27 - Publication and disclosure items and weights: - Publish its strategic plan? CPP5 a) 3 - Publish its annual business/operations plans? CPP5 b) 3 - Publish its annual report? CPP5 c) 3 - Publish a formal set of service delivery standards? CPP5 d) 3 - Make key compliance risks public regularly? EPP6 c) 2 - Make reports of outcomes in addressing compliance risks public regularly? EPP6 e) 2 - Publish results of taxpayer satisfaction surveys for: - Individuals? GPP7 a) 0.5 - Businesses? GPP7 b) 0.5 - Document that formally sets out taxpayer rights? GPP4 a) 3 - Specific mechanism for managing taxpayer complaints: - Internal? GPP5 a) 1 - External? GPP5 b) 1 - Publish periodic estimates of the tax gap? EPP10 1 - Publish a formal strategy to reduce compliance burden? GPP13 b) vi) 2 - Publish its organizational chart? CPP7 a) 2 ### Taxpayer Services - Weighted maximum score: 42 - Strategy and rights: - Formal set of service delivery standards? CPP5 d) 2 - Formal taxpayer service and assistance strategy? GPP6 a) 3 - Taxpayer rights set out in a formal document? GPP4 a) 3 - Complaints and rulings: - Internal mechanism for managing taxpayer complaints? GPP5 a) 2 - External mechanism for managing taxpayer complaints? GPP5 b) (weight not given) - Rulings provided to taxpayers? GPP9 a) 3 - Taxpayer compliance burden: - Administration evaluates taxpayer compliance burden? GPP13 a) 1 - Perceptions of compliance burdens measured? GPP13 b) ii) 0.5 - Compliance burden monitored for different taxpayer segments? GPP13 b) iv) 0.5 - Formal strategy to reduce compliance burdens? GPP13 b) v) 1 - Education and accommodation (weights): - Educational services to children, youth, and students? GPP8 a) i) 1 - Educational services to new businesses? GPP8 a) ii) 1 - Taxpayer education or awareness campaigns targeting a particular gender? GPP8 b) 1 - Free tax services (including accounting and benefit information) for lower-income individuals? GPP8 a) iii) 1 - Free tax services for new businesses? GPP8 a) iv) 1 - Special provision for taxpayers with disabilities? GPP2 a) 1 - Special services for digitally disadvantaged taxpayers? GPP2 f) 1 - Telephone services in languages other than official languages? GPP2 d) 0.5 - Web or app services in languages other than official languages? GPP2 e) 0.5 - Technology in service: - Virtual assistants (for example, chatbots)? GP7 i) 1 - Whole-of-government identification systems? GP7 g) 1 - Use social media to distribute information? GPP2 b) 0.5 - Use social media interactively? GPP2 c) 0.5 - Online services (each weight 0.3 unless stated): - Tax calculators GPP3 a) 0.3 - Requesting extensions of deadlines GPP3 b) 0.3 - Asking for tax payment arrangements GPP3 c) 0.3 - Secure communication with taxpayers GPP3 d) 0.3 - Taxpayer portal providing a “whole-of-taxpayer” view GPP3 g) 0.3 - Ability to view taxpayer information captured from third parties GPP3 h) 0.3 - Mobile applications (including commercial arrangements) GPP3 i) 0.3 - Filing tax-related objections GPP3 e) 0.3 - Uploading data files onto the tax administration’s system GPP3 f) 0.3 - Analysis of service: - Taxpayer satisfaction surveys conducted for: - Individuals? GPP7 a) 1 - Businesses? GPP7 b) 1 - Intermediaries? GPP7 c) 0.5 - Taxpayer satisfaction analysis based on gender? GPP7 e) 0.5 - Use of service channel usage statistics to: - Encourage service adoption GPP1 a) 0.5 - Create new or improved services GPP1 b) 0.5 - Anticipate service demand to adjust staff allocation GPP1 c) 0.5 - Reduce ICT service disruption GPP1 d) 0.5 - Administration keeps track of incoming service contacts? GQ2 a) 1 - Registration, filing, and payment: - Offer more than one registration channel? GQ1 1 - Provide some level of electronic filing (across tax types)? FQ2 d) 2 - Provide some level of electronic payments (across tax types)? FQ5 2 - Offer some level of income/expense prefilling? GQ3 1 ### Index Elements Exhibiting the Largest Differences in Response by AEs and LIDCs (select items and percentage-point differences) - Offering some level of income/expense prefilling TS8: AE-LIDC Percentage Point Difference 70.2 (AE Score Rate 81.13, EME Score Rate 6 .110, LIDC Score Rate 10.97) [source table values preserved as presented] - Does the administration enter into APAs? CM9: AE-LIDC Percentage Point Difference 63.6 (AE Score Rate 91.94, EME Score Rate 41.02, LIDC Score Rate 28.36) - Does the administration prefill any income or expense items in tax returns? DG8: AE-LIDC Percentage Point Difference 61.5 (AE Score Rate 81.1, EME Score Rate 41.0, LIDC Score Rate 19.6) - Technologies implemented: data science/analytics tools DG8: AE-LIDC Percentage Point Difference 60.4 (AE Score Rate 86.5, EME Score Rate 31.3, LIDC Score Rate 26.1) - Have policies for flexible working arrangements? HRMD9: AE-LIDC Percentage Point Difference 60.3 (AE Score Rate 97.3, EME Score Rate 51.8, LIDC Score Rate 37.0) - Systems for importing, storing, and managing third-party data: Social Security Agency CM, DG8: AE-LIDC Percentage Point Difference 56.1 (AE Score Rate 86.5, EME Score Rate 39.8, LIDC Score Rate 30.45) - Special provision made for taxpayers with disabilities? TS9: AE-LIDC Percentage Point Difference 54.9 (AE Score Rate 91.9, EME Score Rate 54.2, LIDC Score Rate 37.0) - Specialized audit staff for international tax issues? CM8: AE-LIDC Percentage Point Difference 47.9 (AE Score Rate 89.2, EME Score Rate 48.2, LIDC Score Rate 41.3) - Facilitate taxpayer access to third-party data already collected before filing? CM, DG7: AE-LIDC Percentage Point Difference 45.8 (AE Score Rate 78.4, EME Score Rate 45.8, LIDC Score Rate 32.6) - Technologies implemented: whole-of-government identification systems TS7: AE-LIDC Percentage Point Difference 45.2 (AE Score Rate 75.7, EME Score Rate 30.13, LIDC Score Rate 30.4) - Services (on the web or app) provided in languages other than the official languages? TS8: AE-LIDC Percentage Point Difference 44.1 (AE Score Rate 81.1, EME Score Rate 41.0, LIDC Score Rate 37.0) - Systems for importing, storing, and managing third-party data: property ownership and sales CM, DG7: AE-LIDC Percentage Point Difference 43.6 (AE Score Rate 78.4, EME Score Rate 54.2, LIDC Score Rate 34.8) - Annual report—produced (published) PT8: AE-LIDC Percentage Point Difference 43.5 (AE Score Rate 89.2, EME Score Rate 54.2, LIDC Score Rate 45.7) - Automated verifications undertaken? CM, DG8: AE-LIDC Percentage Point Difference 43.0 (AE Score Rate 86.5, EME Score Rate 43.4, LIDC Score Rate 43.5) - Technologies implemented: virtual assistants (for example, chatbots) TS6: AE-LIDC Percentage Point Difference 42.6 (AE Score Rate 62.2, EME Score Rate 25.3, LIDC Score Rate 19.6) - E-registration channel available? DG9: AE-LIDC Percentage Point Difference 42.4 (AE Score Rate 94.66, EME Score Rate 66.3, LIDC Score Rate 52.2) - Online services: tax calculators TS, DG8: AE-LIDC Percentage Point Difference 40.8 (AE Score Rate 86.5, EME Score Rate 62.7, LIDC Score Rate 45.7) - Educational services to new businesses TS5: AE-LIDC Percentage Point Difference -10.6 (AE Score Rate 56.8, EME Score Rate 77.1, LIDC Score Rate 67.4) — negative AE-LIDC difference indicates higher LIDC rate relative to AE - Selected items with large negative AE-LIDC differences: - Payments to individuals reported to the administration: proceeds of sales/purchases of real estate CM4: AE-LIDC Percentage Point Difference -11.1 (AE Score Rate 43.2, EME Score Rate 36.15, LIDC Score Rate 54.3) - Payments to individuals reported: proceeds of sales/purchase of shares CM4: AE-LIDC Percentage Point Difference -12.7 (AE Score Rate 45.9, EME Score Rate 34.95, LIDC Score Rate 58.7) - Administration has a large taxpayer office/program? CM7: AE-LIDC Percentage Point Difference -12.9 (AE Score Rate 78.4, EME Score Rate 86.7, LIDC Score Rate 91.3) - Payments to individuals reported: other income CM4: AE-LIDC Percentage Point Difference -13.8 (AE Score Rate 40.5, EME Score Rate 47.05, LIDC Score Rate 54.3) - Systems for importing, storing, and managing third-party data: data from utilities CM, DG16: AE-LIDC Percentage Point Difference -14.2 (AE Score Rate 28.9, EME Score Rate 30.4, LIDC Score Rate 43.6) - Free tax services (including accounting and benefit information) for lower-income individuals TS3: AE-LIDC Percentage Point Difference -16.5 (AE Score Rate 37.8, EME Score Rate 32.5, LIDC Score Rate 54.3) - Pay scales at the administration not tied to civil/public sector pay scales MA2: AE-LIDC Percentage Point Difference -19.1 (AE Score Rate 13.5, EME Score Rate 20.5, LIDC Score Rate 32.6) - Taxpayer education or awareness campaigns targeting a particular gender TS2: AE-LIDC Percentage Point Difference -23.4 (AE Score Rate 2.7, EME Score Rate 10.8, LIDC Score Rate 26.1) - Free tax services for new businesses TS3: AE-LIDC Percentage Point Difference -30.6 (AE Score Rate 32.4, EME Score Rate 44.6, LIDC Score Rate 63.0) - Electronic invoices used to monitor compliance? CM16: AE-LIDC Percentage Point Difference -33.8 (AE Score Rate 16.2, EME Score Rate 47.0, LIDC Score Rate 50.0) - Payments to individuals reported: rents CM4: AE-LIDC Percentage Point Difference -39.3 (AE Score Rate 32.4, EME Score Rate 51.8, LIDC Score Rate 71.7) *Source: Authors.* --- _Source: https://www.imf.org/-/media/files/publications/dp/2026/english/itappea.pdf_