## Model Specification for Excise Gaps

## Source details

**Canonical URL:** [Model Specification for Excise Gaps](https://www.imf.org/-/media/files/publications/tnm/2017/tnm1705.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/tnm/2017/tnm1705.pdf.md)
- [Structured JSON version](/-/media/files/publications/tnm/2017/tnm1705.pdf.json)

---

### Overview and purpose
- The excise gap is defined as the difference between potential revenue and actual revenue for a given excise.
- Top-down excise gap estimates use independent statistical data to estimate total potential collections and compare this figure to actual collections; the difference is the excise gap.
- The excise gap is deconstructed into two main components:
  - the impact of non-compliance (the compliance gap), and
  - the impact of policy measures (the policy gap).
- Why measure the excise gap:
  - Provides a measure of tax revenues lost or foregone through noncompliance, avoidance, and policy decisions.
  - Helps tax administrations monitor voluntary compliance and identify unexplained shortfalls (or over-performance) against forecasts.
  - Top-down estimates provide breakdowns into the compliance gap and the policy gap and estimates of overall noncompliance and changes in noncompliance.
  - Unless specifically excluded, top-down estimates generally include avoidance losses; avoidance reduces tax payable and thus contributes to the observed compliance gap unless separately identified.

### Methodological approaches (focus and alternatives)
- Top-down approach (focus of this TNM):
  - Uses independent statistical data to model the overall tax base and derives total potential collections to compare with actual collections.
  - Produces a static model that measures compliance and policy gaps given current consumption; it does not account for potential price or behavioral effects from changes in compliance or policy.
- Other approaches (sometimes combined with top-down):
  - Bottom-up approach:
    - Auditing a random sample of taxpayers or analyzing compliance risk and intervention results.
    - Useful to identify behaviors creating losses and to test/interpret top-down estimates.
    - Covers only specifically identified sources of the tax gap and can be costly.
  - Econometric techniques:
    - Tools such as frontier or time series analysis to estimate efficiency or revenue losses.
    - Results can be sensitive to selection of determinants and assumptions and are not recommended as the primary method for estimating the tax gap itself.
  - Survey techniques:
    - Household surveys or collections of discarded cigarette packs (as used in Project Star and Project Sun) to estimate market share of untaxed products.

### Design criteria for an effective top-down gap estimation methodology
- 1. Independent Source of Statistics for the Tax Base
  - Rely on good statistics on the size of the tax base derived from sources other than taxpayer records; check that independent sources do not rely on tax records to impute critical missing values or control totals.
- 2. Accurate Statistical Data
  - Statistical data needs to be relatively accurate with detailed documentation on the method of compilation and, ideally, an indication of the estimation error.
- 3. Consistency in Statistical Data
  - When combining different statistical sources, ensure consistency in definitions and scale; verify that survey totals align with national accounts totals and that definitions are stable over time.
- 4. Sufficiently Detailed Statistical Data
  - Data must be sufficiently disaggregated to match detailed definitions in tax rate schedules.
- 5. Comprehensive Statistical Data
  - Data needs to cover the full tax base and be available for more than one period; top-down models are usually more accurate at identifying changes to tax gaps than their precise level.
- 6. Accurate Modeling of the Tax Structure
  - Model the tax structure to adhere closely to statutory description; use statutory rates where possible and capture special treatments such as exemptions, thresholds, credits, etc.

### Analytical framework: four-stage iterative process
- 1. Context:
  - Identify scope and feasibility of the excise gap study based on motivation and data availability.
- 2. Data capture:
  - Identify and capture data required to model the tax base, including statistical data, excise returns and payments data, and legislative and administrative framework.
- 3. Analysis:
  - Model the excise tax base and compare to actual collections; prepare and clean data so it matches model specifications and definitions.
- 4. Reporting:
  - Report estimated level and trend of excise gaps, interpret results with reference to study motivation, and present conclusions and recommendations for further work.
- Iteration and sensitivity testing:
  - Revise model specification or seek alternative sources when data limitations emerge.
  - Use sensitivity testing (alternative assumptions, Monte Carlo) to estimate ranges and error margins.

### Data capture: legislative, administrative, statistical requirements
- Legislation:
  - Establish the legislative framework of the excise for the entire study period; ensure tax rates and coverage are established for all years and account for special rate determinants (e.g., alcohol content).
- Administration data:
  - Use micro-taxpayer returns and payments data to calculate actual excise collections and accrue payments to the tax period; detailed data allow disaggregation into tax bands and commodity types.
  - Discount timing/forestalling effects so collections match consumption periods where relevant.
- Statistical data:
  - Capture statistical data at the most detailed level possible, preferably micro-level, to model coverage of different tax rates and to allow quality checks and cleaning.
  - Consumption or expenditure data (or proxies converted to consumption) are generally used because independent production data are rarely available.
  - Be aware of potential under-reporting in surveys for alcohol and tobacco which could bias gap estimates downwards.
- Confidentiality and security:
  - Replace TINs with ATINs before sharing; encrypt data and avoid identifiers that could re-identify taxpayers.

### Product-specific modelling guidance (tobacco, alcohol, mineral oils)
- Tobacco:
  - Model ready-made cigarettes and loose tobacco separately; convert sticks to weight using an assumed average weight per stick (typically around one gram).
  - Data sources: household consumption/surveys, household expenditure surveys, national accounts, pack-swap/collection surveys.
- Alcohol:
  - Model separately by drink type and tax band; account for specific duty per volume, specific per unit of alcohol, and/or ad-valorem bases.
  - Data sources: household consumption surveys, household expenditure surveys (adjust for on-premise vs off-premise), national accounts.
- Mineral oils (consumption-based modelling preferred):
  - Potential excise (specific-rate) = total distance travelled × fuel efficiency × excise rate (per fuel type).
  - Data sources: household consumption surveys, expenditure surveys, national accounts, transport surveys, administrative data (vehicle registrations, technical controls, road tolls).
  - Where excise is ad-valorem use average price in formula.

### Defining coverage, policy vs compliance gap, and treatment of personal imports/exports
- Define potential tax base explicitly (domestic consumption or production/importation).
- Compliance gap = potential collections under current law − actual collections under current law.
- Policy gap = potential collections under current policy framework − potential collections under a theoretical framework where all relevant transactions taxed at the standard rate.
- Personal imports of tax-free excise goods represent excise tax foregone and should be included in the excise gap as part of the policy gap unless explicitly separated.
- Exports of tax-paid excise goods:
  - Commercial exports typically receive 100 percent excise relief; personal exports may be treated as negative tax gap amounts and excluded from positive gap estimates due to data limitations.
- Positive gaps: excise collected < potential (net revenue losses).
- Negative gaps: excise collected > potential estimated domestic activities (can occur with net exporting of taxed goods or cross-border effects).

### Feasibility, scope and reporting practices
- Feasibility:
  - Critical requirement: independent statistical data with good coverage of the excise tax base.
  - Typical conditions: relatively small number of excise taxpayers; straightforward legislative framework; modelling actual collections and tax base rarely problematic.
  - Main difficulty: finding independent statistical data on consumption sufficiently independent of excise returns.
- Scope:
  - Cover more than one period; aim for at least five years for consistency and trend analysis.
  - Partial analyses remain useful when only partial independent data are available.
- Reporting:
  - Report level and trend, with caveats on uncertainties; trends generally more reliable than levels.
  - Present estimates explicitly as ranges when significant uncertainty exists.
  - Reconcile estimated gaps with existing risk and compliance analysis to inform strategic compliance risk management.

### IFP Mineral Oils case study (Slovakia) — data, model, and key results
- Motivation and high-level context:
  - Revenues from mineral oils excise taxation formed the largest share of overall excise taxes by commodity group in Slovakia.
  - Total revenues from excises were EUR 1,077 million in 2014, of which:
    - EUR 354 million were received from excise taxes on gasoline
    - EUR 716 million from diesel
    - EUR 7 million from other mineral oils
  - Mineral oils excise tax accounted for 53 percent of total excise taxation in 2014.
- Data sources used and suitability summary:
  - Vehicle registration database, technical and emission controls, Skytoll road tolls, taxpayer returns and payments microdata; several sources had limitations and required merging and imputation.
- Model specification for mineral oils:
  - Potential excise (fuel type) = Slovak mileage × average consumption per km × excise rate (fuel type)
  - Tax gap (fuel type) = Potential excise (fuel type) - Actual excise (fuel type)
  - Total tax gap = tax gap (gasoline) + tax gap (diesel)
  - The study’s tax gap includes elements of the policy gap as well as the compliance gap; both legitimate and illegal imports of untaxed oils are treated as being in the gap.
- Data preparation and imputations:
  - Removed entries without required identification details: about 1.2 percent of 8 million entries.
  - Multiple entries in same month for same registration/VIN: only last entry kept.
  - Outlier mileage values over 200,000 km between controls were capped at 200,000km.
  - Negative mileage values (three to five percent of total mileage travelled per year) were treated as typographic errors and sign changed.
  - Assumptions made on vehicle lifespan after last recorded control and imputations from registration database for missing vehicles.
  - Road tolls (Skytoll) provided complete records for foreign owned commercial vehicle mileage in Slovakia; all commercial vehicles assumed to use diesel.
- Treatment of cross-border mileage:
  - Missing data for oils consumption abroad by domestically owned vehicles was offset against missing data for domestic consumption by foreign-owned private vehicles; it was assumed missing values offset one another.
- Headline result and sensitivity:
  - The published estimate represented an upper bound.
  - Headline result: overall tax gap in mineral oils excise tax of 27 percent of potential excise in 2014 with a strongly rising trend since 2011.
  - Results are sensitive to assumptions, in particular:
    - i. The assumed lifespan of vehicles after the last recorded technical and emissions control.
    - ii. The treatment of negative mileage values as typographic errors.
- Selected time-series values (Figure 2.5 Excise Gaps for Mineral Oils 2008–14):
  - 2008:
    - Gasoline Gap 34 (Millions of Euros); Diesel Gap 118; Total Mineral Oils Gap 152
    - Gasoline Gap 27%; Diesel Gap 15%; Total Mineral Oils Gap 12%
    - Gasoline Gap 0.05% of GDP; Diesel Gap 0.17% of GDP; Total Mineral Oils Gap 0.22% of GDP
  - 2009:
    - Gasoline Gap 91; Diesel Gap 194; Total Mineral Oils Gap 285
    - Gasoline Gap 18%; Diesel Gap 25%; Total Mineral Oils Gap 22%
    - Gasoline Gap 0.14% of GDP; Diesel Gap 0.30% of GDP; Total Mineral Oils Gap 0.45% of GDP
  - 2010:
    - Gasoline Gap 69; Diesel Gap 49; Total Mineral Oils Gap 118
    - Gasoline Gap 14%; Diesel Gap 8%; Total Mineral Oils Gap 11%
    - Gasoline Gap 0.10% of GDP; Diesel Gap 0.07% of GDP; Total Mineral Oils Gap 0.17% of GDP
  - 2011:
    - Gasoline Gap 77; Diesel Gap 79; Total Mineral Oils Gap 157
    - Gasoline Gap 16%; Diesel Gap 11%; Total Mineral Oils Gap 13%
    - Gasoline Gap 0.11% of GDP; Diesel Gap 0.11% of GDP; Total Mineral Oils Gap 0.22% of GDP
  - 2012:
    - Gasoline Gap 92; Diesel Gap 194; Total Mineral Oils Gap 285
    - Gasoline Gap 20%; Diesel Gap 23%; Total Mineral Oils Gap 22%
    - Gasoline Gap 0.13% of GDP; Diesel Gap 0.27% of GDP; Total Mineral Oils Gap 0.39% of GDP
  - 2013:
    - Gasoline Gap 92; Diesel Gap 280; Total Mineral Oils Gap 372
    - Gasoline Gap 21%; Diesel Gap 29%; Total Mineral Oils Gap 26%
    - Gasoline Gap 0.12% of GDP; Diesel Gap 0.38% of GDP; Total Mineral Oils Gap 0.50% of GDP
  - 2014:
    - Gasoline Gap 99; Diesel Gap 299; Total Mineral Oils Gap 397
    - Gasoline Gap 22%; Diesel Gap 29%; Total Mineral Oils Gap 27%
    - Gasoline Gap 0.13% of GDP; Diesel Gap 0.39% of GDP; Total Mineral Oils Gap 0.52% of GDP
  - Note: The figures originally presented contained a programming error for 2013 that resulted in over-estimates for 2013 for diesel duty and the overall tax gap. The figure shown uses the corrected figures.

### Interpretation, actions and follow-up from the case study
- Publication impact:
  - The IFP’s published report generated media interest and prompted internal and external discussions on reliability, implications, and measures to close the tax gap.
  - Meetings were held between the IFP, the Financial Administration, and other stakeholders to discuss calculation, reliability, and counter-measures.
- Stakeholder views:
  - Trade representatives and the Financial Administration questioned the estimated level but agreed on the existence of a gap and increasing risks.
  - The Financial Administration considered policy gap elements represented a large, though not quantified, part of the tax gap (legitimate cross-border shopping by households and purchases of diesel by international freight companies).
- Countermeasures and updates:
  - The Financial Administration announced a series of counter measures; additional compliance measures were implemented in 2015 to reduce the compliance gap and address MTIC VAT risks using mineral oils.
  - The IFP updated its estimate in 2016 (IFP, 2016), reviewing and refining methodology and publishing estimates as a range of upper and lower bounds; the 2016 updated estimate was lower than the 2015 estimate, though similar annual changes and trends were found.

### Key methodological recommendations and future work
- Use excise gap estimates to take a strategic view of compliance and policy risks; quantify strategic risks and prioritize resource allocation to minimize revenue inefficiencies.
- Use excise gap analysis as an outcome-based analytical framework for excise policy, risk and compliance analysis, and performance management.
- Update and improve excise gap analysis as more data become available; keep estimates as up to date as possible.
- Consider quicker proxy measurements for ongoing monitoring; example proxy: Effective Tax Rate (ETR) = excise collections divided by the value of sales of the relevant commodity (or reasonable proxy) to indicate market share of taxed goods and likely changes in the gap.
- Report both levels and trends, explicitly present ranges when uncertainty is significant, and reconcile gap estimates with risk/compliance intelligence.

*Source: tnm1705 - 1.   Model Specification for Excise Gaps (excerpt) from the provided PDF content.*

### 1.   Model Specification for Excise Gaps.......................................29

### 1.   Model Specification for Excise Gaps

### Overview
- The excise gap is defined as the difference between potential revenue and actual revenue for a given excise.
- Top-down excise gap estimates use independent statistical data to estimate total potential collections and compare this figure to actual collections; the difference is the excise gap.
- The excise gap is deconstructed into two main components:
  - the impact of non-compliance (the compliance gap), and
  - the impact of policy measures (the policy gap).

### Methodological approaches to estimating excise gaps
- Top-down approach (focus of this TNM):
  - Uses independent statistical data to model the overall tax base and derives total potential collections to compare with actual collections.
  - Produces a static model that measures compliance and policy gaps given current consumption; it does not account for potential price or behavioral effects from changes in compliance or policy.
- Other approaches (sometimes combined with top-down):
  - Bottom-up approach:
    - Auditing a random sample of taxpayers or analyzing compliance risk and intervention results.
    - Useful to identify behaviors creating losses and to test/interpret top-down estimates.
    - Covers only specifically identified sources of the tax gap and can be costly.
  - Econometric techniques:
    - Tools such as frontier or time series analysis to estimate efficiency or revenue losses.
    - Results can be sensitive to selection of determinants and assumptions, produce perverse results, and be difficult to interpret for compliance purposes.
    - Not recommended as primary method for estimating the tax gap itself.
  - Survey techniques:
    - Household surveys or collections of discarded cigarette packs (as used in Project Star and Project Sun) to estimate market share of untaxed products.

### Box 1: IMF RA-GAP Program (summary)
- The Revenue Administration Gap Analysis Program (RA-GAP) is conducted by the IMF Fiscal Affairs Department’s Revenue Administration Divisions (FADR1 and FADR2).
- RA-GAP provides revenue administrations with comprehensive and detailed estimates of the gap between current and potential collections and reviews operational performance in key functions.
- Initially focused on VAT gaps; methodology for VAT gaps is set out in (Hutton, 2017).
- The program is being extended to other taxes, including excises.

### Design criteria for an effective top-down gap estimation methodology
- 1. Independent Source of Statistics for the Tax Base
  - Rely on good statistics on the size of the tax base derived from sources other than taxpayer records; check that independent sources do not rely on tax records to impute critical missing values or control totals.
- 2. Accurate Statistical Data
  - Statistical data needs to be relatively accurate with detailed documentation on the method of compilation and, ideally, an indication of the estimation error.
- 3. Consistency in Statistical Data
  - When combining different statistical sources, ensure consistency in definitions and scale; verify that survey totals align with national accounts totals and that definitions are stable over time.
- 4. Sufficiently Detailed Statistical Data
  - Data must be sufficiently disaggregated to match detailed definitions in tax rate schedules.
- 5. Comprehensive Statistical Data
  - Data needs to cover the full tax base and be available for more than one period; top-down models are usually more accurate at identifying changes to tax gaps than their precise level.
- 6. Accurate Modeling of the Tax Structure
  - Model the tax structure to adhere closely to statutory description; use statutory rates where possible and capture special treatments such as exemptions, thresholds, credits, etc.

### Why measure the excise gap?
- Provides a measure of tax revenues lost or foregone through noncompliance, avoidance, and policy decisions.
- Helps tax administrations monitor voluntary compliance and identify unexplained shortfalls (or over-performance) against forecasts.
- Top-down estimates:
  - Provide breakdowns into the compliance gap (difference between actual and potential collections given current policy) and the policy gap (difference between potential collections under current legislation and under a normative regime taxing all domestic activities and imports at the standard rate).
  - Provide estimates of overall noncompliance and changes in the level of noncompliance; estimates of changes are typically more robust than point estimates of levels.
- Treatment of avoidance:
  - Unless specifically excluded, top-down estimates generally include avoidance losses.
  - Avoidance reduces tax payable and thus contributes to the observed compliance gap unless separately identified and quantified.
  - If avoidance can be identified, an appropriate adjustment to the compliance gap should be considered; appropriate treatment for avoidance is generally a change in policy, not a compliance intervention.

### Analytical framework: four stages in measuring an excise gap
- The estimation of tax gaps is an iterative process involving four stages:
  - 1. Context:
    - Identify scope and feasibility of the excise gap study based on motivation and data availability.
  - 2. Data capture:
    - Identify and capture data required to model the tax base, including statistical data, excise returns and payments data, and legislative and administrative framework.
  - 3. Analysis:
    - Model the excise tax base and compare to actual collections; prepare and clean data so it matches model specifications and definitions.
  - 4. Reporting:
    - Report estimated level and trend of excise gaps, interpret results with reference to study motivation, and present conclusions and recommendations for further work.
- The framework requires iteration; data limitations may necessitate revising the model specification or seeking alternative sources.

### Context and motivation (selected observations)
- Excise taxation typically shows stable collections levels unless undermined by noncompliance or cross-border shopping because many excises are levied on commodities with low own-price elasticity.
- Where identical goods are available taxed and untaxed, consumers may substitute toward untaxed goods, creating relatively high cross-price elasticities and putting collections at risk from noncompliance.
- Substitution can occur via:
  - legitimate purchases in other countries with lower excise rates (cross-border shopping),
  - smuggling into the domestic market without paying tax,
  - non-tax paid counterfeit goods,
  - diversion or other frauds.

*Source: tnm1705 - 1.   Model Specification for Excise Gaps (excerpt) from the provided PDF content.*

### 2. Excise compliance risks may be reported by tax administrations and other government

### 2. Excise compliance risks may be reported by tax administrations and other government agencies

### Sources of reported compliance risks
- Revenue agencies (compliance, risk, intelligence, investigations) should understand and systematically report compliance behaviors in taxpayer populations.
- Other government agencies (police, border agency, health professionals) may report changes in the source and distribution of excise goods.
- Trade representatives, businesses, and external agencies (health lobbyists, pressure groups) may report:
  - concerns about compliant excise taxpayers, cross-border shopping, smuggling, and noncompliance as commercial risks;
  - changing patterns of purchasing and consumption of excise goods to the revenue agency.
- Private-sector studies of noncompliance (trade-sponsored) can provide evidence but require critical review for methodology and agenda bias.

### Feasibility
- Critical requirement: statistical data with good coverage of the excise tax base that is independent of taxpayers’ returns.
- Typical conditions:
  - number of excise taxpayers in any one country is relatively small;
  - legislative framework is straightforward to model;
  - calculating actual excise collections and modelling the tax base is rarely problematic.
- Main difficulty: identifying independent statistical data on consumption of excise goods sufficiently independent of excise returns.

### Scope and coverage
- Study scope should cover more than one period; aim for at least five years to test consistency, reliability, and trends.
- If independent statistical data are available only for part of the tax base, a useful partial analysis can still be executed.
- Level of statistical detail may prevent decomposition by individual tax rate or commodity type; set expectations reasonably.

- IFP CASE STUDY: CONTEXT — Mineral Oils Excise Gap Estimate
  - Motivation: Long-term decline in excise collections, not explained by excise rate changes.
  - Feasibility: No single independent source for production, consumption, or expenditure available. Consumption can be derived from combined data sets on motor vehicle registrations, controls, and road tolls.
  - Scope and Coverage: Potential excise defined as domestic consumption multiplied by the appropriate excise rate. Several years' data available for time series.

### Data capture
- Legislation
  - Establish the legislative framework of the excise for the entire study period.
  - Ensure tax rates and coverage are established for all years; account for rates determined by goods’ content or type (e.g., alcoholic drinks).
- Administration data
  - Use micro-taxpayer returns and payments data, where possible, to calculate actual excise collections and accrue payments to the tax period for which they were made.
  - Detailed data allow disaggregation into tax bands and commodity types and aid understanding of transaction distribution.
  - Where forestalling and timing effects occur (forestalling can involve up to several months of clearances being made in advance), discount such timing effects so collections match consumption periods.
- Statistical data
  - Capture statistical data at the most detailed level possible, preferably micro-level, to model coverage of different tax rates and to allow data quality checks and cleaning.
  - Consumption or expenditure data (or proxies converted to consumption) are generally used to model the potential excise tax base because independent production data are rarely available.
  - Caveat: reported production data are often based on excise declarations and thus not independent; proxy production data may be used but are more suitable for risk assessment and trends than robust central excise gap estimates.
  - NB: potential under-reporting issues for consumption of alcohol and tobacco products could bias gap estimates downwards.

- IFP CASE STUDY: CONTEXT — Data sources used
  - Legislation: Detailed descriptions obtained from Ministry of Finance and Financial Administration.
  - Administration Data: Taxpayers returns and payments microdata obtained from Financial Administration.
  - Statistical Data: Microdata obtained from databases maintained for vehicle registrations, control and road toll payments.

### Analysis
- Data preparation
  - Examine administrative and statistical data in detail and resolve quality issues, including:
    - Missing data
    - Outliers and other anomalies
    - Duplication
    - Consistency (internal and across sources)
    - Typographic errors
    - Coverage
  - Manipulate and re-format data as required; match multiple sources at micro-level where possible; derive variables (e.g., convert alcoholic drink volumes to units of alcohol, convert mileage to average mileage per tax period).
- Model specification
  - Match modelling parameters to legislative framework and administrative processes (tax rates, coverage, historic changes).
  - Ensure model parameters match data definitions; preserve greatest level of disaggregation possible until final results.
- Analysis process
  - Iterative: input data, produce results, and revisit data or model when non-credible results appear.
  - Test results against external data and analysis (expert opinion, risk/compliance specialists); reconcile significant differences through iteration.
  - Use sensitivity testing to assess impact of alternative assumptions and data; options include:
    - Running model with alternative assumptions/data to estimate range of possible results;
    - Statistical techniques (e.g., Monte Carlo analysis) to estimate likely scale and distribution of error margins.
  - Sensitivity testing may lead to alternative assumptions, changes in model specification, or different data sources.

- IFP CASE STUDY: ANALYSIS
  - Data Preparation: Anomalous and incorrect data entries corrected, and incomplete entries removed.
  - Model Specification: Average mileage per year per fuel type and vehicle type derived or inferred from database entries. Actual collections calculated from taxpayer returns and payments data.
  - Analysis: For each fuel type and vehicle type, consumption derived from annual mileage and used to determine potential excise collections. Actual collections compared to potential collections to derive tax gaps per fuel type and overall. Sensitivity of results to data cleaning rules tested to find upper and lower bounds.

### Reporting
- Presentation of results
  - Report the level and trend of the excise gaps, together with caveats on uncertainties and limitations.
  - Report both level and trend; trends are generally more reliable than levels.
  - Where significant uncertainty produces a range of possible results, present estimates explicitly as a range with plausible upper and lower bounds.
  - It is good practice to publish excise gap results to support transparent government decision making and allow external review.
- Interpretation of results
  - Beyond estimating scale, understand risks and taxpayer behavior underlying the gap.
  - Reconcile and map estimated tax gap to existing risk and compliance analysis; use it as a cornerstone for strategic compliance risk management.

- IFP CASE STUDY: REPORTING
  - Presentation of Results: Results were published in a detailed report by the IFP, presented as upper bound estimates.
  - Interpretation of Results: Report generated interest from both the trade and the public. General acceptance of the existence of compliance gaps, but estimated levels questioned.

### Conclusions and future work
- Use excise gap estimates to take a strategic view of compliance and policy risks; quantify strategic risks and prioritize resource allocation to minimize revenue inefficiencies efficiently.
- Use excise gap analysis as an outcome-based analytical framework for excise policy, risk and compliance analysis, and performance management.
- Update and improve excise gap analysis as more data become available; keep estimates as up to date as possible.
- Consider quicker proxy measurements for ongoing monitoring; example proxy: Effective Tax Rate (ETR) = excise collections divided by the value of sales of the relevant commodity (or reasonable proxy) to indicate market share of taxed goods and likely changes in the gap.

*Technical Notes and Manuals 17/05 | 2017 — Section 2*

### Conclusions and

### Conclusions and Future Work

### V. WHAT DATA IS REQUIRED TO MEASURE THE EXCISE GAP?
- RA-GAP analysis for tax gaps relies on two data sets: taxpayer records and statistical data.
- Data should be for at least five full calendar years.
- Statistical data should be collected on a consistent basis to minimize measurement errors.

### A. Specification of Tax Administration Records
- Tax administration data should be detailed micro-level records to match individual payments and refunds to tax periods, identify errors and anomalies, and resolve queries.
- Required taxpayer registry data:
  - ATIN—Anonymized Taxpayer Identification Number (see below);
  - date of registration for excise taxation;
  - date of deregistration for excise (if applicable);
  - periods of inactive status (if any);
  - main sector of activity; and
  - any special filing provisions (e.g., identifying quarterly versus monthly filers).
- Required excise return data:
  - ATIN;
  - excise tax period;
  - filing date;
  - original volumes and/or values from all lines on the return used in calculating the excise due as submitted by the taxpayer; and
  - current assessed volumes and/or values for the same lines.
- Required excise payment and refund data:
  - ATIN;
  - type of transaction (payment or refund);
  - tax period(s) that the payment/refund is for;
  - date of transaction; and
  - amount of the transaction.
- Required customs declarations data (where excise is collected by the customs agency):
  - ATIN of the declarant;
  - date of the importation for excise purposes;
  - customs procedure code providing details on the nature of the transaction;
  - any special codes detailing any special tax applications for the declaration;
  - commodity code for each item on the declaration;
  - landed volume and/or value of each item in local currency; and
  - amount of excise declared.
- Confidentiality and security requirements:
  - Taxpayers’ privacy should be respected; third parties must preserve confidentiality.
  - Data provided to third parties should be encrypted (for example, password protected zip files) and transmitted securely.
  - Tax administration datasets provided to third parties should not include names, addresses, or any other details that could allow individual taxpayers to be identified.
  - Existing taxpayer identification numbers (TIN) should be replaced by ATIN by the tax administration before the data are passed to the third party.
  - Best practice: tax administration converts TINs into ATINs using an algorithm and does not pass the algorithm to the third party so the administration can identify taxpayers from ATINs later while the third party cannot.

### B. Specification of Statistical Data for Excise Tax Bases
- Independent statistical data are required to estimate potential collections for each excise; approaches differ by country and by duty type (specific duties use volumes/weights; ad-valorem uses values).
- General form for specific duties: total amounts of excise goods consumed or produced (including imports) × relevant excise rates.
- Critical requirements:
  - Statistical data must be independent of excise declarations and payments.
  - Be cautious where reported production and imports are derived directly from excise returns or adjusted series—obtain unadjusted data where possible.
  - Time series must account for methodological or definition changes; revise prior results or chain-link series to preserve consistency.
  - Published production data are typically derived from aggregated excise returns and are not appropriate for top-down potential excise estimation (they can be useful in bottom-up approaches).
- Treatment of untaxed products and under-reporting:
  - Estimates of total consumption and expenditure should include domestic purchases, goods that have evaded domestic excise, and legitimate personal imports not charged domestic excise.
  - Survey under-reporting (Box 4): respondents typically under-estimate alcohol and tobacco consumption; any such under-reporting must be accounted for or potential excise will be under-estimated, producing an under-estimate of the excise gap.
  - If reported aggregates have been adjusted using excise declarations, that adjustment must be discounted and an independent assumption used; otherwise analysis risks circularity.
  - Where extant research on under-reporting exists, base assumptions on it; otherwise consider indexing potential excise results back to periods with little or no market share taken by untaxed product or apply a constant adjustment factor to assess changes in the gap even if level uncertainty remains.
  - Where expenditure survey data are used, average prices of both tax-paid and untaxed goods are needed (see Box 5).
- Practical guidance on untaxed prices (Box 5):
  - For specific duty: potential excise = consumption × excise rate.
  - For ad-valorem: potential excise = total consumption × average taxed price × excise rate.
  - Where only expenditure is available: consumption = expenditure ÷ average price (accounting for taxed and untaxed prices).
  - Key consumption equations provided:
    - (1) Taxed expenditure = taxed clearances × average taxed price
    - (2) Total consumption = taxed clearances + (total expenditure - taxed expenditure) ÷ untaxed price)
  - If untaxed price is unknown, it may be assumed (informed by market research or calculated by deducting excise from taxed price).
  - If the tax gap is relatively small (say, less than ten percent of potential collections), or the price difference between taxed and untaxed goods is relatively small, untaxed prices can be ignored for a first approximation; the gap estimate will be biased downwards but the error will likely not be more than about one percent of potential collections.
  - Resulting gap formulae:
    - Specific duty gap = Untaxed consumption × excise rate = (total expenditure - taxed expenditure) ÷ untaxed price) × excise rate
    - Ad-valorem gap = Untaxed expenditure × excise rate = (total expenditure - taxed expenditure) × excise rate

### Product-specific data and modelling guidance
- Tobacco products:
  - Model ready-made cigarettes and loose cigarette tobacco separately due to different excise bases and rates (cigarettes often taxed per stick and/or ad-valorem; loose tobacco taxed by weight).
  - Possible data sources and models:
    - Household consumption (or smoking prevalence) surveys:
      - Potential specific duty = total number of smokers × average consumption × excise rate
      - Potential ad-valorem excise = total consumption × average price × excise rate
    - Household expenditure surveys:
      - Potential specific duty = total expenditure ÷ average price × excise rate
      - Potential ad-valorem excise = total ad valorem excise paid + (untaxed expenditure × taxed price ÷ average untaxed price) × excise rate
    - National accounts: household expenditure on manufactured tobacco products can be used but source data with more granularity is preferred.
    - Pack swap and pack collection surveys: can estimate market share of untaxed goods by examining cigarette packs; useful as an ad-hoc validation tool though relatively expensive and requiring care to avoid selection bias.
    - Other data: published assessments of smoking levels/trends can be used where survey time series are unavailable but should be examined for bias.
  - Note on unit conversions: if tobacco excise is levied on cigarettes on the basis of weight, it should be possible to convert number of sticks consumed to weight using an assumed average weight per stick—typically (and fairly consistently) around one gram of tobacco per stick.
  - Monitor loose tobacco due to substitution risks and smuggling concealability; model other tobacco types (pipe, cigar, raw tobacco) separately where excise rates differ.
- Alcoholic drinks:
  - Excise may be levied as a specific duty on volume, on unit of alcohol, and/or ad-valorem; rules and rates often differ by drink type (beer, wine, spirits), alcohol content bands, and special reduced rates for traditional or micro/artisan producers.
  - Tax bases should be modelled separately by product type and tax band where possible.
  - Possible data sources and models:
    - Household consumption surveys:
      - For specific rate per volume: Potential excise = total number of households × average consumption × excise rate
      - For specific rate per unit of alcohol: Potential excise = total number of households × average consumption × average strength × excise rate
      - For ad-valorem taxation: Potential excise collections = total consumption × average price × excise rate
    - Household expenditure surveys (with caveats):
      - Price differs between home consumption and consumption on commercial premises; estimate separately.
      - Business purchases (for catering, corporate events) should be modelled separately or adjusted for if material.
      - For each product/tax band/purchase type:
        - Specific per volume: Potential specific duty = total expenditure ÷ average price × excise rate
        - Specific per unit alcohol: Potential specific duty = total number of households × average consumption × average strength (percent of alcohol) × excise rate
        - Ad-valorem: Potential ad valorem excise = total ad valorem excise paid + (untaxed expenditure × taxed price ÷ average untaxed price) × excise rate
    - National accounts: GDP breakdowns include household expenditure on beverages; may lack granularity but source data may be useful.

### Notes on methodology and data limitations
- Time series comparability is crucial; where methodologies or definitions change, revise or chain-link series and consult the producing body where possible.
- Published aggregate production and trade data are often derived from excise returns and are not independent for top-down estimation; they can be useful in bottom-up approaches or to identify anomalies (for example, branded exports re-entering the domestic market).
- Where consumers may not disclose purchases of evaded goods in surveys, this could bias gap estimates downward, but there is no substantive evidence that this is a major problem in practice.
- For countries with one or very few large producers/distributors dominating excise markets, anonymized records may still identify large taxpayers; tax administrations must satisfy themselves that third parties will preserve confidentiality.

*Conclusions and Future Work — Technical Notes and Manuals 17/05 | 2017*

### 3. Excise taxes may be levied in different ways and at different rates on the different types of

### 3. Excise taxes may be levied in different ways and at different rates on the different types of

### Overview
- Excise taxation on petroleum products typically yields the majority of revenue from gasoline (private motor vehicles) and diesel (commercial vehicles).
- If potential excise for gasoline and diesel can be modelled, the remainder of the market may be accounted for by assumption or simple adjustment.
- Consumption-based modelling is preferred because production data for excise goods is rarely independent of excise returns and payments.

### Data sources and modelling approaches
- a. Household consumption surveys
  - Household surveys commonly record motor vehicle usage as miles travelled by vehicle type, engine type and size, age, etc., but usually exclude commercial operators.
  - Required auxiliary data: fuel efficiency of private motor vehicles (sources: ministry of transportation, market research, manufacturers/distributors).
  - Adjustments:
    - Exclude travel abroad when potential excise is defined by domestic consumption.
    - Convert distance travelled per householder or per passenger to distance travelled per vehicle; remove double-counting of trips with passengers from more than one household.
    - For distance per passenger, adjust using average number of passengers per vehicle.
  - Good practice: model consumption at the most disaggregated level possible (fuel type, vehicle type, engine size, age).
  - Formulae:
    - For product types or bands where excise is charged at a specific rate per volume:
      - Potential excise = total distance travelled × fuel efficiency (per km or mile) × excise rate
      - Where: Total distance travelled = distance travelled by vehicle × total number of vehicles
    - For product types or bands subject to ad-valorem excise taxation:
      - Potential excise = total distance travelled × fuel efficiency × average price × excise rate

- b. Expenditure surveys
  - Household expenditure surveys typically capture spending on road fuels but not commercial travel; expenditures must be split between fuel types (gasoline vs diesel).
  - Expenditure surveys usually lack details on vehicle type.
  - Formulae:
    - Where excise is charged at a specific rate per volume:
      - Potential specific duty = total expenditure ÷ average price × excise rate
    - Where excise is charged ad-valorem:
      - Potential ad valorem excise = total ad valorem excise paid + (untaxed expenditure × taxed price ÷ average untaxed price) × excise rate
  - Business expenditure surveys can capture commercial expenditure but require more detailed distinctions (e.g., buses vs taxi cabs) and exclusion of travel abroad.

- c. National accounts
  - National accounts include values of final consumption and intermediate consumption of petroleum products by businesses.
  - Such aggregates can be used similarly to expenditure surveys but are unlikely to have sufficient granularity and may be calibrated to excise clearances.

- d. Transport surveys
  - Transport surveys report distances travelled by vehicle type for businesses and householders; distances may be reported as “passenger miles” or “passenger kilometers” and must be converted to vehicle distances.
  - Formula:
    - Total distance travelled by vehicles = total distance travelled by passengers ÷ average passengers per trip
  - Where excise rates vary by fuel type, distances should be disaggregated similarly; if not available, estimate from market research or assumptions (e.g., private vehicles more likely gasoline; almost all large commercial vehicles use diesel).
  - Repeat formulae for potential excise as in (a):
    - Specific-rate: Potential excise = total distance travelled × fuel efficiency × excise rate
    - Ad-valorem: Potential excise = total distance travelled × fuel efficiency × average price × excise rate

- e. Administrative data from transport authorities and road tolls
  - Databases (vehicle registrations, roadworthiness checks, licensing, road tolls) may contain vehicle-specific details (registration number, distance travelled, vehicle/engine type, size, age).
  - Where coverage is sufficient, these data can estimate consumption from total distances travelled.
  - Often multiple sources are required to cover breadth/detail of the excise tax base.
  - Example cited (Slovakia): combining police, transport ministry, and road toll databases using vehicle registration to estimate distances travelled by private and commercial vehicles.

### Mathematical model specification for excise gaps (Appendix 1)
- Definition:
  - Excise gap = Potential excise collections - Actual excise collections
  - Excise gap = policy gap + compliance gap
- Potential excise collections (general form for specific-rate excises):
  - Potential excise collections = [Consumption (dutiable product) x excise rate (product type)]
- Actual excise collections:
  - Actual excise collections = Domestic excise collections + Import excise collections
- Notes:
  - For ad-valorem excise, use the defined taxable value of the goods instead of consumption; ad-valorem excises can be calculated using either tax-inclusive or tax-exclusive prices—set the taxable value accordingly.
  - Consumption used as proxy for production (including imports) because production data are rarely independent of excise returns.
  - Consumption generally defined as domestic consumption (residents and non-residents within the country).

### Defining coverage and treatment of personal imports/exports
- Defining coverage:
  - Potential tax base must be defined (domestic consumption or production/importation).
  - Compliance gaps = potential collections under current law − actual collections under current law.
  - Policy gaps = potential collections under current policy framework − potential collections under a theoretical framework where all relevant transactions taxed at standard rate.
- Personal imports of tax-free excise goods:
  - Represent excise tax foregone or lost through policy choices and should be included in the estimated excise gap as part of the policy gap (tax expenditure).
  - Because consumption-based models include tax-free personal imports unless discounted, a simple comparison between potential excise and actual collections will identify excise foregone as part of the total excise compliance gap unless explicitly separated.
- Exports of tax-paid excise goods:
  - Commercial exports typically receive 100 percent excise relief; travelers’ personal exports may still have excise paid and can be considered additional to domestic consumption.
  - Symmetry and pragmatism suggest treating personal exports as negative tax gap amounts (i.e., windfalls against the intended tax base) and excluding them from positive gap estimates due to data limitations.

### Positive and negative excise gaps
- Positive gaps: excise collected is less than total potential (net revenue losses) — the typical case.
- Negative gaps: excise collected exceeds potential estimated from domestic activities (can occur with relatively low excise rates and net exporting of taxed excise goods).
  - Causes: domestic taxed prices lower than neighboring countries, import restrictions, cross-border shopping, smuggling.
  - Negative gaps may contain both policy gap and compliance gap components.
  - Example noted: Northern Ireland experienced high compliance gap in road fuels duty in the 1990s (up to 50 percent in some areas) due to cross-border shopping and smuggling; later exchange rate changes produced a negative gap.

### Appendix 2 — IFP Mineral Oils Excise Tax Gap Estimate for Slovakia: Context and key statistics
- Motivation and high-level findings:
  - Revenues from mineral oils excise taxation formed the largest share of overall excise taxes by commodity group in Slovakia.
  - Total revenues from excises were EUR 1,077 million in 2014.
  - Of this EUR 1,077 million:
    - EUR 354 million were received from excise taxes on gasoline
    - EUR 716 million from diesel
    - EUR 7 million from other mineral oils
  - Mineral oils excise tax accounted for 53 percent of total excise taxation in 2014.
- Historical trends and indicators:
  - Effective Tax Rate (ETR) for mineral oils calculated as:
    - ETR = (Total collections) / (GDP) (measured at constant prices)
  - ETR declined from 1.6 percent of GDP in Quarter 3 of 2009 to 1.5 percent in Quarter 2 of 2015, with a possible decline to 1.4 percent in Quarter 3 of 2012 followed by stabilization.
  - Excise rate changes: diesel duty decreased in February 2010 from EUR 481.31 per 1,000 liters to EUR 368 per 1,000 liters, explaining the 2010 ETR drop; no other significant changes subsequently.
  - Road transport trends:
    - Number of road vehicles using diesel increased rapidly relative to gasoline.
    - Average consumption per vehicle decreased (likely due to improved car manufacture).
    - Tax returns show increasing diesel consumption and market share since 2010; gasoline consumption shows a significant decline.
  - Cross-border dynamics:
    - Gasoline excise rates and final prices in Slovakia higher than neighboring countries; diesel rates and prices generally competitive (except Ukraine).
    - High gasoline prices create significant risk of cross-border shopping given proximity of population centers to land borders.
- Policy rationale:
  - Estimating the mineral oils tax gap would improve excise administration effectiveness by evaluating measures to minimize fraud and evasion.
  - Tax avoidance/evasion shifts burden to compliant taxpayers and reduces resources for government expenditure; past VAT gap analysis shows benefits of gap analysis for administration.

*Source: tnm1705 - 3. Excise taxes may be levied in different ways and at different rates on the different types of (Technical Notes and Manuals 17/05, 2017).*

### 7. To estimate the excise gap for mineral oils, an estimate of total consumption of mineral oils in

### 7. To estimate the excise gap for mineral oils, an estimate of total consumption of mineral oils in Slovakia is required

### Overview and definitions
- For the purposes of estimating the tax gap for mineral oils, it is assumed that all oils consumed should be taxed (in Slovakia). Amounts of excise not paid on domestic consumption represent the overall tax gap.
- The total gap ideally is split between:
  - Policy gap: mineral oils bought abroad and legitimately imported into Slovakia for the purchasers’ own use (including purchases in cross-border shopping trips).
  - Compliance gap: smuggling of non-Slovak taxed mineral oils into Slovakia for re-sale and excise evaded through fraud or non-declaration.
- The study’s definition of the tax gap treats both legitimate and illegal imports of untaxed oils as being in the gap; it is not possible under the method used to identify the relative shares of policy and compliance gaps.

### Potential data sources for consumption (Table 2.2 summary)
- Vehicle registration database (police)
  - Data: vehicle type, fuel type, date of registration, average consumption per km.
  - Limitations: No mileage data; does not cover foreign registered vehicles.
  - Suitability: Limited––incomplete data. Can be used with other data sources.
- Household expenditure survey (Statistical Office)
  - Data: total expenditure on mineral oils by households.
  - Limitations: Does not include business spending; last complete survey in 2012; 2013 & 2014 data very limited.
  - Suitability: Not suitable––incomplete coverage.
- Transport survey (Statistical Office)
  - Data: questionnaire on transport usage sent to companies (results in passenger kilometers or weight).
  - Limitations: Low response rates; conversion to vehicles requires assumptions; does not include empty bus and freight movements.
  - Suitability: Not suitable––potentially large selection bias.
- National accounts (Statistical Office)
  - Data: aggregate supply and use of mineral oils expressed as values.
  - Limitations: Indexed to excise returns; no split between gasoline and diesel.
  - Suitability: Not suitable––not independent of tax returns, and insufficiently detailed.
- Trade data (Slovak Petrol Association)
  - Data: published estimate based on production reported by association members.
  - Limitations: Danger of commercial bias; likely not independent of excise declarations.
  - Suitability: Not suitable––not independent of tax returns and potential bias.
- Technical and emission controls data (Ministry of Transport, Construction and Regional Development)
  - Data: vehicle registration numbers, engine code (VIN), vehicle type, type of fuel and mileage on date of control, date of control.
  - Limitations: Missing data on average consumption; limited to domestically owned vehicles; missing mileage since most recent control; includes mileage travelled abroad.
  - Suitability: Limited––incomplete data and coverage. Could be used with other data sources.
- Road tolls data (“Skytoll”)
  - Data: detailed Slovak mileage for all foreign freight and buses entering Slovakia (electronic data from 2011).
  - Limitations: Limited to commercial vehicles; missing mileage in Slovakia by foreign owned private vehicles.
  - Suitability: Limited––incomplete data and coverage. Could be used with other data sources.

### Data captures (legislation, administrative, statistical)
- Legislation
  - Excise framework: two main rates for gasoline and diesel, and for each fuel type a special rate for biofuels (introduced in 2011). Biofuels account for only a very small market share.
  - Rate change during study period: reduction in diesel duty from EUR 481.31 per 1,000 liters to EUR 368 per 1,000 liters in February 2010.
- Administrative data
  - Detailed microdata on taxpayer returns and payments were provided by the Slovak Financial Administration; returns record volumes of each fuel and excise paid.
- Statistical data
  - Detailed microdata from vehicle registration and controls databases for 2007 to 2014 and from Skytoll for 2011 to 2014 were obtained.
  - The vehicle registration database was merged with technical and emissions control data to create a single database of mileage by every domestically owned vehicle, plus vehicle age, engine type, control dates, average consumption per kilometer, and fuel type.
  - Road tolls microdata for foreign owned commercial vehicles was obtained from Skytoll.

### Analysis: preparation, modeling, and assumptions
- Data preparation
  - Removed entries without required identification details: about 1.2 percent of 8 million entries.
  - When multiple entries in the same month for same registration and VIN, only the last entry was kept.
  - Outlier mileage values over 200,000 km between controls were capped at 200,000km.
  - Missing data in controls database was inferred from the vehicle registration database where appropriate.
- Model preparation and imputations
  - Annual mileage per vehicle was calculated as the annual average over the period between successive controls.
  - Negative mileage values (three to five percent of total mileage travelled per year) were treated as typographic errors and sign changed.
  - For first control of a vehicle, average mileage was calculated from vehicle age. If vehicle age > registration in Slovakia, only mileage in years the vehicle was in Slovakia was calculated.
  - For vehicles with last control in 2011 or later, mileage after the last control was inferred from previous average mileage with the assumption vehicles remained in use in 2014.
  - New vehicles and other vehicles missing from controls were imputed from the vehicle registration database by comparing totals and inferring averages from similar vehicle types.
  - Road tolls data required no adjustments; Skytoll provided complete records for foreign owned commercial vehicle mileage in Slovakia. All commercial vehicles were assumed to use diesel and average consumption per category inferred from vehicle registration database.
- Treatment of cross-border mileage
  - Missing data for oils consumption abroad by domestically owned vehicles was offset against missing data for domestic consumption by foreign-owned private vehicles.
  - No reliable data was available to subtract mileage travelled abroad by domestically owned vehicles or add mileage in Slovakia by foreign owned vehicles; it was assumed missing values offset one another. Partial justification: fuel prices in Slovakia are similar to those in neighboring countries.
- Tax gap model specification
  - Potential excise (fuel type) = Slovak mileage x average consumption per km x excise rate (fuel type)
  - Tax gap (fuel type) = Potential excise (fuel type) - Actual excise (fuel type)
  - Total tax gap = tax gap (gasoline) + tax gap (diesel)
  - The relative share of consumption for each fuel type was derived from taxpayer returns.
- Definition note
  - The study’s tax gap includes elements of the policy gap as well as the compliance gap; both legitimate and illegal imports of untaxed oils are treated as being in the gap.

### Results and key statistics
- The published estimate represented an upper bound.
- Headline result: overall tax gap in mineral oils excise tax of 27 percent of potential excise in 2014 with a strongly rising trend since 2011.
- Results are sensitive to assumptions, in particular:
  - i. The assumed lifespan of vehicles after the last recorded technical and emissions control.
  - ii. The treatment of negative mileage values as typographic errors.
- Figure 2.5 Excise Gaps for Mineral Oils 2008–14 (selected values)
  - 2008: Gasoline Gap 34 (Millions of Euros); Diesel Gap 118; Total Mineral Oils Gap 152; Gasoline Gap 27%; Diesel Gap 15%; Total Mineral Oils Gap 12%; Gasoline Gap 0.05% of GDP; Diesel Gap 0.17% of GDP; Total Mineral Oils Gap 0.22% of GDP
  - 2009: Gasoline Gap 91; Diesel Gap 194; Total Mineral Oils Gap 285; Gasoline Gap 18%; Diesel Gap 25%; Total Mineral Oils Gap 22%; Gasoline Gap 0.14% of GDP; Diesel Gap 0.30% of GDP; Total Mineral Oils Gap 0.45% of GDP
  - 2010: Gasoline Gap 69; Diesel Gap 49; Total Mineral Oils Gap 118; Gasoline Gap 14%; Diesel Gap 8%; Total Mineral Oils Gap 11%; Gasoline Gap 0.10% of GDP; Diesel Gap 0.07% of GDP; Total Mineral Oils Gap 0.17% of GDP
  - 2011: Gasoline Gap 77; Diesel Gap 79; Total Mineral Oils Gap 157; Gasoline Gap 16%; Diesel Gap 11%; Total Mineral Oils Gap 13%; Gasoline Gap 0.11% of GDP; Diesel Gap 0.11% of GDP; Total Mineral Oils Gap 0.22% of GDP
  - 2012: Gasoline Gap 92; Diesel Gap 194; Total Mineral Oils Gap 285; Gasoline Gap 20%; Diesel Gap 23%; Total Mineral Oils Gap 22%; Gasoline Gap 0.13% of GDP; Diesel Gap 0.27% of GDP; Total Mineral Oils Gap 0.39% of GDP
  - 2013: Gasoline Gap 92; Diesel Gap 280; Total Mineral Oils Gap 372; Gasoline Gap 21%; Diesel Gap 29%; Total Mineral Oils Gap 26%; Gasoline Gap 0.12% of GDP; Diesel Gap 0.38% of GDP; Total Mineral Oils Gap 0.50% of GDP
  - 2014: Gasoline Gap 99; Diesel Gap 299; Total Mineral Oils Gap 397; Gasoline Gap 22%; Diesel Gap 29%; Total Mineral Oils Gap 27%; Gasoline Gap 0.13% of GDP; Diesel Gap 0.39% of GDP; Total Mineral Oils Gap 0.52% of GDP
- Note: The figures originally presented contained a programming error for 2013 that resulted in over-estimates for 2013 for diesel duty and the overall tax gap. The figure shown uses the corrected figures.

### Interpretation, actions taken, and follow-up
- Publication impact
  - The IFP’s published report generated media interest and prompted internal and external discussions on reliability, implications, and measures to close the tax gap.
  - Meetings were held between the IFP, the Financial Administration, and other stakeholders to discuss calculation, reliability, and counter-measures.
- Stakeholder views
  - Trade representatives and the Financial Administration questioned the estimated level but agreed on the existence of a gap and increasing risks.
  - The Financial Administration considered policy gap elements represented a large, though not quantified, part of the tax gap (legitimate cross-border shopping by households and purchases of diesel by international freight companies).
  - Slovakia’s position as a small, open country with major international freight routes contributes to risks.
- Countermeasures and updates
  - In response, the Financial Administration announced a series of counter measures; additional compliance measures were implemented in 2015 to reduce the compliance gap, noting risks from Missing Trade Intra-Community (MTIC) VAT frauds using mineral oils.
  - The IFP updated its estimate in 2016 (IFP, 2016), reviewing and refining methodology and publishing estimates as a range of upper and lower bounds. The 2016 updated estimate was lower than the 2015 estimate, though similar annual changes and trends were found.

*Source: Technical Notes and Manuals 17/05 | 2017. Source: Country authorities.*

### REFERENCES

### REFERENCES

### Reports and studies
- “Asia-14 Illicit Tobacco Indicator”, September 2014, International Tax and Investment Center and Oxford Economics; http://www.oxfordeconomics.com/asia14.
- HMRC, 2012 (1), “Measuring tax gaps 2015 edition methodological annex”, October 2015, HM Revenue & Customs; http://webarchive.nationalarchives.gov.uk/20160612044958/https://www.gov.uk/government/statistics/measuring-tax-gaps.
- HMRC, 2012 (2), “Measuring Tax Gaps 2012 Edition”, October 2012, HM Revenue & Customs; http://webarchive.nationalarchives.gov.uk/20150612044958/https:/www.gov.uk/government/uploads/system/uploads/attachment_data/file/249146/mtg-2012.pdf.
- Hutton, 2017, “The Revenue Administration–Gap Analysis Program: Model and Methodology for Value-Added Tax Gap Estimation”, January 2017, International Monetary Fund Fiscal Affairs Department.
- IFP, 2015, “Danova medzera na dani z mineralnych olejov”, May 2015, Institute for Financial Policy, Ministry of Finance, Bratislava.
- IFP, 2016, “Danová medzera na dani z minerálnych olejov (aktualizácia)”, August 2016, Institute for Financial policy, Ministry of Finance, Bratislava; http://www.finance.gov.sk/Default.aspx?CatID=10192.
- “Measuring tax gaps in Australia, 2014–15, November 2015, Australian Taxation Office; https://www.ato.gov.au/About-ATO/Research-and-statistics/In-detail/Tax-gap/Prior-analysis/Measuring-tax-gaps-in-Australia,-2014-15/.
- “Project SUN: A study of the illicit cigarette market”, June 2016, KPMG, https://home.kpmg.com/uk/en/home/insights/2015/05/project-sun-a-study-of-the-illicit-cigarette-market.html.
- “The development of the tax gap in Sweden 2007-12”, January 2014, Skatteverket; https://www.skatteverket.se/download/18.15532c7b1442f256baeae28/1395223863657/The+development+of+the+tax+gap+in+Sweden+2007-12.pdf.

### Contact information
- TNM/17/05
- International Monetary Fund
- Fiscal Affairs Department
- 700 19th Street NW
- Washington, DC 20431
- USA
- Tel: 1-202-623-8554
- Fax: 1-202-623-6073

*TNM/17/05 International Monetary Fund Fiscal Affairs Department, contact details as listed above.*

---


_Source: https://www.imf.org/-/media/files/publications/tnm/2017/tnm1705.pdf_