## In response to a request from the Statistics Department (SD) in the Ministry of Sustainable Development (MSD) in Saint Kitts and Nevis: Technical Assistance Mission June 5–16, 2023

## Source details

**Canonical URL:** [In response to a request from the Statistics Department (SD) in the Ministry of Sustainable Development (MSD) in Saint Kitts and Nevis: Technical Assistance Mission June 5–16, 2023](https://www.imf.org/-/media/files/publications/tar/2024/english/tarea2024063-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/tar/2024/english/tarea2024063-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/tar/2024/english/tarea2024063-print-pdf.pdf.json)

---

### Mission scope and objectives
- Mission period: June 5–16, 2023.
- Objectives:
  - Re-reference the constant price (KP) annual and quarterly estimates of GDP to 2018 prices and develop long run time series for GDP from 1977, based on a simplified compilation procedure.
  - Undertake analysis of revisions arising from the re-referencing.
  - Undertake a real-time study of the impact of benchmarking quarterly estimates to annual estimates of GDP.
  - Provide advice on improving the estimates of GDP by expenditure components (GDP-E).
  - Outline plans for the development of Supply and Use Tables (SUT) for 2024 and how these can be used to rebase GDP.

### Re-referencing the GDP series and historical reconstruction
- Re-referencing completed to 2018 prices and a long-run chainlinked constant price series prepared for 1977–2022.
- ISIC: Adopted International Standard Industrial Classification revision 4 (updated from revision 3.1); publication tables are presented using ISIC revision 4.
- Chain linking:
  - Chainlinked series were constructed by linking the 2006-based and 2018-based constant series.
  - Link year chosen: 2012.
  - Note: Chain linking causes loss of additivity in years before the link year.
- Current price backcasting: Levels of GVA for elementary activities backcast using previously published growth rates to produce aggregate series.
- Constant price chain linking: All levels backcast based on previously published growth rates; preserves previous base years before the link year.
- Developed historical series in current and constant prices covering 1977–2022.

### Methodological improvements introduced
- Final Intermediation Services Indirectly Measured (FISIM):
  - FISIM consumed by businesses allocated to economic activities.
  - ECCB data used to estimate total FISIM consumed by each economic sector (government, households, businesses, rest of the world).
  - Result: No longer necessary to show a separate adjustment line for FISIM in the GDP-P publication.
  - FISIM consumed by owner-occupied dwelling services estimated and treated as intermediate consumption for those implicit services.
- Taxes on products in constant prices:
  - Previous SD method: Deflate current price estimates using deflator for total GVA (incorporates some price changes).
  - Recommended (short-cut) method used: Extrapolate base year level of taxes on products in current prices by an indicator of volume change, here the volume of GVA.
  - This methodological change contributes to significant revisions in estimated changes in volume of taxes.
- Other improvements:
  - Re-referenced series prepared for both annual and quarterly series, and for Saint Kitts and Nevis separately and the Federal aggregate.
  - Business processing documentation prepared (Appendix I referenced).

### Data sources, compilation practice, and constraints
- Annual GDP-P current price estimates largely derived from SD’s National Accounts Statistics (NAS) and supplemented by administrative sources (social security, tax authority) due to NAS response rates below 30 percent.
- Quarterly GDP-P estimates:
  - Published around 60 days after the reference quarter (e.g., first quarter estimates published in May/June).
  - Quarterly indicators used for each economic activity separately for Saint Kitts and for Nevis.
  - Benchmarking aligns quarterly series with annual estimates around two years after the reference period, using Denton’s proportional method implemented with IMF XLPBM software.
- Supply and Use Tables (SUT):
  - SD planned SUT for 2018 in 2020; COVID-19 delayed data collection.
  - SUT development mapped out; realistically may be undertaken in 2025 in respect of 2024 (resources required mapped out).

### Analysis of revisions arising from re-referencing and benchmarking — key quantified findings (annual growth, 2022)
- Current price series:
  - Published estimate of growth for 2022: 20.2 percent.
  - Revised figure for 2022: 15.3 percent.
  - Total revision: -4.9 percentage points.
  - Manufacturing activity contribution to revision: -2.5 percentage points (new annual data replaced quarterly indicator data).
  - Hotels activity contribution: -0.8 percentage points (from more recent annual data).
- Constant price series:
  - Published estimate for 2022: 13.0 percent.
  - Revised figure for 2022: 8.5 percent.
  - Manufacturing activity contribution to revision: -2.7 points.
  - Change in methodology for estimating volume of taxes contributed: -2.2 percentage point.

### Real-time study, forecasting, and short-term indicators
- Real-time revisions study undertaken to assess impact of benchmarking quarterly estimates to annual estimates (see Section IV and Appendix II).
- Current SD short-term forecast practice:
  - SD provides an internal forecast of the GDP-P series using a simple moving average of the equivalent quarters in the previous four years.
- Proposed improvement:
  - Holt-Winters (triple exponential smoothing) approach proposed and can be implemented using a simple Excel function.
  - Caveat: Holt-Winters does not account for multiplicative moving seasonality and can produce forecasts that lag underlying trend; may not perform well with significant seasonal or cyclical variations where ARIMA models could produce smaller forecast errors.
  - Despite known biases, Holt-Winters expected to provide more robust forecasts compared to the moving average approach.

### Institutional engagement and publication
- Mission met with representatives from the Eastern Caribbean Central Bank (ECCB), which coordinates GDP compilation across the eight ECCU member countries; opportunity to explain compilation basis and revisions.
- CARTAC advisor discussed outcomes and recommendations with the Director of Statistics; SD to publish re-referenced series for GDP in July 2023.

### Priority recommendations (summary)
- Target Date: July 2023
  - Finalize the compilation of the re-referenced estimates of GDP for quarters up to and including quarter four of 2022 and publish the results together with a note describing the updated series.
  - Responsible Institution: SD
- Target Date: December 2023
  - Review potential quarterly indicators and annual estimates for Manufacturing, Communications, and Hotels to identify if more reliable data may be available, including data for the VAT system for the Inland Revenue Department.
  - Responsible Institution: SD
- Target Date: September 2024
  - Continue to work with the IRD to access the data on sales from the tax systems which may provide more comprehensive estimates of the level of output in each period.
  - Responsible Institution: SD

### Operations, documentation, and next steps
- Business processing documentation prepared to support simplified updating of annual, quarterly, and island-level series.
- Action plan and detailed technical assessment contain further actions/milestones.
- Proposed timeline for SUT development: resources mapped; realistic target 2025 to produce SUT for 2024.

### Section IV — Real Time Study of Revisions Arising from Benchmarking: purpose, methodology, and metrics
- Study objective: Isolate revisions arising purely from the benchmarking procedure when quarterly indicator series are replaced by annual benchmark data.
- Study period and data:
  - Used data from 2012–2022.
  - First estimates based on quarterly indicator data computed starting from 2015.
  - ‘Second estimates’ based on annual data computed when available; resulting revisions stored.
  - Revisions aggregated for all 44 ‘elementary’ components of total GDP; average of absolute contributions to the total revision calculated for the years 2015–2022.
- Metrics computed:
  - Root mean square error as a percentage of the average absolute revision.
  - Average revision (bias).

### Key findings — contributions to benchmarking-induced revisions (average taken over 2015–2022)
- Constant 2018 prices — Average Absolute Percentage Point Contributions to the Total Revision to GDP:
  - Manufacturing: Average absolute contribution 1.35; % of total absolute revision 18%
  - Communications: Average absolute contribution 0.96; % of total absolute revision 13%
  - Hotels: Average absolute contribution 0.75; % of total absolute revision 10%
  - Rented Dwellings: Average absolute contribution 0.75; % of total absolute revision 10%
  - Taxes on Products: Average absolute contribution 0.72; % of total absolute revision 10%
- Current prices — Average Absolute Percentage Point Contributions to the Total Revision to GDP:
  - Manufacturing: Average absolute contribution 2.39; % of total abs revision 28%
  - Hotels: Average absolute contribution 1.70; % of total abs revision 20%
  - Rented Dwellings: Average absolute contribution 0.85; % of total abs revision 10%
  - Global FISIM: Average absolute contribution 0.57; % of total abs revision 7%
  - Private Education: Average absolute contribution 0.52; % of total abs revision 6%
- Interpretation:
  - Manufacturing was the largest single cause of revisions:
    - 18 percent of the total absolute revision to the constant price series.
    - 28 percent of the total absolute revision to the current price series.
  - Hotels activity is also a prominent cause of revisions.

### Case study — Manufacturing (data limitations and indicator performance)
- Quarterly indicators used:
  - Current price indicator: wages and salaries of employees (reported to the Social Security Board).
  - Volume indicator: numbers of weeks worked (reported to the Social Security Board).
- Annual series basis:
  - Data from around seven companies for value and physical quantities of production (electronic goods, beverages, cement).
- Empirical comparison:
  - Correlation coefficient between the ‘first estimate’ (indicator-based) and the ‘latest estimate’ (annual-based) of annual percentage change in GVA for manufacturing in constant prices: -0.51.
    - Interpretation: Very low (negative) correlation; the inverse of the indicator series correlates reasonably well with the annual series.
- Data and conceptual limitations:
  - Quarterly current price series ignores changes in operating surpluses.
  - Volume indicator (number of weeks worked) does not account for productivity and changes in hours worked.
  - Annual series relies on a very limited sample and physical quantity measures that depend on stability of product mix over long periods.
- Conclusion: Both quarterly indicators and annual series have weaknesses; alternative data sources should be considered to better reflect actual changes in manufacturing GVA.

### Recommended actions (Section IV)
- The SD should:
  - Review potential quarterly indicators and annual estimates for Manufacturing, Communications, and Hotels to identify if more reliable data may be available, including data for the VAT system for the Inland Revenue Department (IRD).
  - Continue to work with the IRD to access the data on sales from the tax systems which may provide more comprehensive estimates of the level of output in each period.

### Workbooks: GDP_SKN_Annual_Data.xlsb and GDP_SKN_Final.xlsb — purpose and key procedures
- GDP_SKN_Annual_Data.xlsb — purpose:
  - Produces estimates of annual GDP in both current and constant prices for Saint Kitts and Nevis combined (SKN) based on the selected base year.
  - Used to produce the annual benchmarks for the published tables.
- GDP_SKN_Annual_Data.xlsb — key structure and calculations:
  - Group 1: 27 worksheets for economic activities plus two for taxes and subsidies; use indicator indexes to extrapolate base year levels; GVA = Output − IC.
  - Group 2: Benchmark Data (BM, BM_Add) contains benchmark levels of Output and IC; change cell B2 in BM to re-reference base year.
  - Group 4: FISIM calculations allocate FISIM to each activity based on GVA proportions.
  - Single extrapolation method: KP IC = KP output × base year IC/O ratio.
  - Updating steps: Update GVA workbooks, copy formulas, update Latest_Yr worksheet.
- GDP_SKN_Final.xlsb — purpose:
  - Calculates the annual and quarterly GDP for SKN and separately for Saint Kitts (SK) and Nevis (Nev).
  - Allows selection of latest quarter and units for final tables.
- GDP_SKN_Final.xlsb — key structure and calculations:
  - Group 1: Indicator Data links to GVA system and price indicators.
  - Group 2: 44 worksheets for each economic activity; use CP GVA series for SK as indicator and deflate using elementary price index to derive KP GVA; benchmarking implemented with IMF ‘XLPBM’ function as ‘=Bench()’.
  - Group 3: Aggregates into CP and KP annual and quarterly series for SKN, SK, and Nevis.
  - Group 6: Final publication tables for quarterly and annual series in CP and KP, and implicit price deflators.
  - Group 7: Contributions Analysis worksheets for quality assurance.
  - Group 8: Revisions Analysis worksheets to establish reasons for revisions in KP to growth in GDP.
- Updating both workbooks: ensure individual activity workbooks updated, open each linked workbook and check links via Data->Edit Links->Check Status.

### Notes on updating the GDP-P system
- GDP-P built on top of existing GVA system; no changes to calculations in existing GVA workbooks required though some additional calculations were added to worksheets.
- Example: [ISIC A011 - Crops.xlsb] includes calculation of the constant price series for total crop output based on base year in BM worksheet.

### Appendix II — Real-time revisions studies: method and example metrics (exact figures preserved)
- Example annual growth rates table (selected entries shown exactly):
  - 2015 2% 11% 11% 11% 11% 11% 11% 11%
  - 2016 -7% -7% -16% -16% -16% -16% -16% -16%
  - 2017 -5% -3% -4% -4% -4% -4% -4%
  - 2018 -8% -7% -1% -1% -1% -1%
  - 2019 -8% -10% -12% -12% -12%
  - 2020 -8% -7% -3% -3%
  - 2021 -9% -11% -8%
  - 2022 -9% -11% -10%
- Example 1: 1-year ahead ‘forecasts’ summary (exact figures)
  - Rows:
    - 2015 2% 11% 9.0% 0.80%
    - 2016 -7% -16% -9.3% 0.87%
    - 2017 -3% -4% -0.7% 0.00%
    - 2018 -7% -1% 6.2% 0.38%
    - 2019 -10% -12% -1.4% 0.02%
    - 2020 -7% -3% 4.5% 0.20%
    - 2021 -11% -8% 3.4% 0.11%
  - Summary metrics:
    - Bias 1.7%
    - RMSE 5.8%
    - Avg abs growth 7.8%
    - RMSE/ avg abs growth 75%
- Example 2: 2-year ahead ‘forecasts’ summary (exact figures)
  - Rows:
    - 2016 -7% -16% -9% 0.76%
    - 2017 -5% -4% 1% 0.01%
    - 2018 -8% -1% 6% 0.40%
    - 2019 -8% -12% -4% 0.13%
    - 2020 -8% -3% 5% 0.26%
    - 2021 -9% -8% 1% 0.01%
  - Summary metrics:
    - Bias 0.2%
    - RMSE 5.1%
    - Avg abs growth 6.2% 3.7%
    - RMSE/ avg abs growth 83%

### Appendix III — Data sources for Supply and Use Tables (selected)
- Domestic Output: Data from NAS survey, and VAT/CIT tax system, plus existing estimates based on GDP-P.
- Imports of goods: Eurotrace data based on HS 8-digit headings.
- Imports of services: Balance of Payments Statistics (coded to EBOPS).
- Taxes on products:
  - VAT: Total non-dedicational VAT, based on government revenue statement.
  - Trade taxes: Customs data.
  - Excise taxes: Government revenues statement.
- Final expenditure of Households: Household Budget Surrey (2019/19) price and volume updated.
- Exports of goods: Eurotrace data based on HS 8-digit headings.
- Exports of services: Balance of Payments Statistics (coded to EBOPS).

### Appendix IV — Proposed dimensions of Supply and Use Tables for 2024 (selected entries)
- Product-industry mapping examples (codes and descriptions preserved):
  - No. 1 — Code A010 — Product: Crops; Industry No. 1 — Code 010 — Growing of crops (011, 012 and 013)
  - No. 9 — No. 9 — Code C011 — Product: Meat and meat products; Industry No. 10 — Code 101-102 — Processing and preserving of meat and fish
  - No. 69 — Code H523 — Product: Seaport services; Industry No. 53 — Code 5223 — Service activities incidental to air transportation
  - No. 87 — Code L681 — Product: Imputed rents /OOD; Industry No. 72 — Code 6811 — Owner-occupied dwellings
  - No. 127 — Code T970 — Product: Activities of households as employers of domestic personnel; Industry No. 105 — Code 97 — Activities of households as employers of domestic personnel

*IMF Technical Assistance Report – Saint Kitts & Nevis Improving Estimates of GDP*

### 1. In response to a request from the Statistics Department (SD) in the Ministry of Sustainable

### 1. In response to a request from the Statistics Department (SD) in the Ministry of Sustainable Development (MSD) in Saint Kitts and Nevis: Technical Assistance Mission June 5–16, 2023

### Mission scope and objectives
- Mission period: June 5–16, 2023.
- Objectives:
  - Re-reference the constant price (KP) annual and quarterly estimates of GDP to 2018 prices and develop long run time series for GDP from 1977, based on a simplified compilation procedure.
  - Undertake analysis of revisions arising from the re-referencing.
  - Undertake a real-time study of the impact of benchmarking quarterly estimates to annual estimates of GDP.
  - Provide advice on improving the estimates of GDP by expenditure components (GDP-E).
  - Outline plans for the development of Supply and Use Tables (SUT) for 2024 and how these can be used to rebase GDP.

### Re-referencing the GDP series and historical reconstruction
- Re-referencing completed to 2018 prices and a long-run chainlinked constant price series prepared for 1977–2022.
- ISIC: Adopted International Standard Industrial Classification revision 4 (updated from revision 3.1); publication tables are presented using ISIC revision 4.
- Chain linking:
  - Chainlinked series were constructed by linking the 2006-based and 2018-based constant series.
  - Link year chosen: 2012.
  - Note: Chain linking causes loss of additivity in years before the link year.
- Current price backcasting: Levels of GVA for elementary activities backcast using previously published growth rates to produce aggregate series.
- Constant price chain linking: All levels backcast based on previously published growth rates; preserves previous base years before the link year.

### Methodological improvements introduced
- Final Intermediation Services Indirectly Measured (FISIM):
  - FISIM consumed by businesses allocated to economic activities.
  - ECCB data used to estimate total FISIM consumed by each economic sector (government, households, businesses, rest of the world).
  - Result: No longer necessary to show a separate adjustment line for FISIM in the GDP-P publication.
  - FISIM consumed by owner-occupied dwelling services estimated and treated as intermediate consumption for those implicit services.
- Taxes on products in constant prices:
  - Previous SD method: Deflate current price estimates using deflator for total GVA (incorporates some price changes).
  - Recommended (short-cut) method used: Extrapolate base year level of taxes on products in current prices by an indicator of volume change, here the volume of GVA.
  - This methodological change contributes to significant revisions in estimated changes in volume of taxes.
- Other improvements:
  - Re-referenced series prepared for both annual and quarterly series, and for Saint Kitts and Nevis separately and the Federal aggregate.
  - Business processing documentation prepared (Appendix I referenced).
  - Developed historical series in current and constant prices covering 1977–2022.

### Data sources, compilation practice, and constraints
- Annual GDP-P current price estimates largely derived from SD’s National Accounts Statistics (NAS) and supplemented by administrative sources (social security, tax authority) due to NAS response rates below 30 percent.
- Quarterly GDP-P estimates:
  - Published around 60 days after the reference quarter (e.g., first quarter estimates published in May/June).
  - Quarterly indicators used for each economic activity separately for Saint Kitts and for Nevis.
  - Benchmarking aligns quarterly series with annual estimates around two years after the reference period, using Denton’s proportional method implemented with IMF XLPBM software.
- Supply and Use Tables (SUT):
  - SD planned SUT for 2018 in 2020; COVID-19 delayed data collection.
  - SUT development mapped out; realistically may be undertaken in 2025 in respect of 2024 (resources required mapped out).

### Analysis of revisions arising from re-referencing and benchmarking
- Revisions analysis undertaken for both current price and constant price series; charts provided for 2013–2022 comparisons (referred to as Chart 1.A, Chart 1.B, Chart 1.C).
- Key quantified findings (annual growth, 2022):
  - Current price series:
    - Published estimate of growth for 2022: 20.2 percent.
    - Revised figure for 2022: 15.3 percent.
    - Total revision: -4.9 percentage points.
    - Manufacturing activity contribution to revision: -2.5 percentage points (new annual data replaced quarterly indicator data).
    - Hotels activity contribution: -0.8 percentage points (from more recent annual data).
  - Constant price series:
    - Published estimate for 2022: 13.0 percent.
    - Revised figure for 2022: 8.5 percent.
    - Manufacturing activity contribution to revision: -2.7 points.
    - Change in methodology for estimating volume of taxes contributed: -2.2 percentage point.
  - Chart 1.C compares annual percentage change in taxes on products in constant prices (published versus re-referenced series, 2013–2018).

### Real-time study, forecasting, and short-term indicators
- Real-time revisions study undertaken to assess impact of benchmarking quarterly estimates to annual estimates (see Section IV and Appendix II).
- Current SD short-term forecast practice:
  - SD provides an internal forecast of the GDP-P series using a simple moving average of the equivalent quarters in the previous four years.
- Proposed improvement:
  - Holt-Winters (triple exponential smoothing) approach proposed and can be implemented using a simple Excel function.
  - Caveat noted: Holt-Winters does not account for multiplicative moving seasonality and can produce forecasts that lag underlying trend; may not perform well with significant seasonal or cyclical variations where ARIMA models could produce smaller forecast errors.
  - Despite known biases, Holt-Winters expected to provide more robust forecasts compared to the moving average approach.

### Institutional engagement and coordination
- Mission met with representatives from the Eastern Caribbean Central Bank (ECCB), which coordinates GDP compilation across the eight ECCU member countries; opportunity to explain compilation basis and revisions.
- CARTAC advisor discussed outcomes and recommendations with the Director of Statistics; SD to publish re-referenced series for GDP in July 2023.

### Priority recommendations (summary of Table 1)
- Target Date: July 2023
  - Priority Recommendation: Finalize the compilation of the re-referenced estimates of GDP for quarters up to and including quarter four of 2022 and publish the results together with a note describing the updated series.
  - Responsible Institution: SD
- Target Date: December 2023
  - Priority Recommendation: Review potential quarterly indicators and annual estimates for Manufacturing, Communications, and Hotels to identify if more reliable data may be available, including data for the VAT system for the Inland Revenue Department.
  - Responsible Institution: SD
- Target Date: September 2024
  - Priority Recommendation: Continue to work with the IRD to access the data on sales from the tax systems which may provide more comprehensive estimates of the level of output in each period.
  - Responsible Institution: SD

### Operations, documentation, and next steps
- Business processing documentation prepared to support simplified updating of annual, quarterly, and island-level series.
- Action plan and detailed technical assessment contain further actions/milestones (referenced under Detailed Technical Assessment and Recommendations).
- Proposed timeline for SUT development: resources mapped; realistic target 2025 to produce SUT for 2024.

*IMF | Technical Assistance Report – Saint Kitts & Nevis Improving Estimates of GDP*

### Section IV. A Real Time Study of Revisions Arising from

### Section IV. A Real Time Study of Revisions Arising from Benchmarking

### Purpose and Methodology
- Study objective: Isolate revisions arising purely from the benchmarking procedure when quarterly indicator series are replaced by annual benchmark data.
- Study period and data:
  - Used data from 2012–2022.
  - First estimates based on quarterly indicator data computed starting from 2015.
  - ‘Second estimates’ based on annual data computed when available; resulting revisions stored.
  - Revisions aggregated for all 44 ‘elementary’ components of total GDP; average of absolute contributions to the total revision calculated for the years 2015–2022.
- Metrics computed:
  - Root mean square error as a percentage of the average absolute revision.
  - Average revision (bias).
  - These metrics used to assess whether first estimates (quarterly indicators) are useful early estimates of later (annual-based) estimates.

### Key Findings — Contributions to Revisions
- Overall finding: Benchmarking to annual data produces non-trivial revisions to annual percentage changes in GDP; certain activities drive a large share of total absolute revision.
- Study design note: The study isolates benchmarking-induced revisions only and does not account for revisions due to changes in actual underlying data.

- Constant 2018 prices — Average Absolute Percentage Point Contributions to the Total Revision to GDP (Average Taken Over 2015–2022):
  - Manufacturing: Average absolute contribution 1.35; % of total absolute revision 18%
  - Communications: Average absolute contribution 0.96; % of total absolute revision 13%
  - Hotels: Average absolute contribution 0.75; % of total absolute revision 10%
  - Rented Dwellings: Average absolute contribution 0.75; % of total absolute revision 10%
  - Taxes on Products: Average absolute contribution 0.72; % of total absolute revision 10%

- Current prices — Average Absolute Percentage Point Contributions to the Total Revision to GDP (Average Taken Over 2015–2022):
  - Manufacturing: Average absolute contribution 2.39; % of total abs revision 28%
  - Hotels: Average absolute contribution 1.70; % of total abs revision 20%
  - Rented Dwellings: Average absolute contribution 0.85; % of total abs revision 10%
  - Global FISIM: Average absolute contribution 0.57; % of total abs revision 7%
  - Private Education: Average absolute contribution 0.52; % of total abs revision 6%

- Interpretation:
  - Manufacturing was the largest single cause of revisions:
    - 18 percent of the total absolute revision to the constant price series.
    - 28 percent of the total absolute revision to the current price series.
  - Hotels activity is also a prominent cause of revisions.

### Case Study — Manufacturing
- Quarterly indicators used:
  - Current price indicator: wages and salaries of employees (reported to the Social Security Board).
  - Volume indicator: numbers of weeks worked (reported to the Social Security Board).
- Annual series basis:
  - Data from around seven companies for value and physical quantities of production (electronic goods, beverages, cement).
- Empirical comparison:
  - Chart 2 compares the ‘first estimate’ (indicator-based) and the ‘latest estimate’ (annual-based) of the annual percentage change in GVA for manufacturing in constant prices.
  - Correlation coefficient between the two series: -0.51.
    - Interpretation: Very low (negative) correlation; the inverse of the indicator series correlates reasonably well with the annual series.
- Data and conceptual limitations noted:
  - Quarterly current price series ignores changes in operating surpluses.
  - Volume indicator (number of weeks worked) does not account for productivity and changes in hours worked.
  - Annual series relies on a very limited sample and physical quantity measures that depend on stability of product mix over long periods.
- Conclusion: Both quarterly indicators and annual series have weaknesses; alternative data sources should be considered to better reflect actual changes in manufacturing GVA.

### Recommended Actions (Section IV)
- The SD should:
  - Review potential quarterly indicators and annual estimates for Manufacturing, Communications, and Hotels to identify if more reliable data may be available, including data for the VAT system for the Inland Revenue Department (IRD).
  - Continue to work with the IRD to access the data on sales from the tax systems which may provide more comprehensive estimates of the level of output in each period.

*Source: Section IV. A Real Time Study of Revisions Arising from Benchmarking, IMF Technical Assistance Report – Saint Kitts & Nevis Improving Estimates of GDP*

### 1. Workbook GDP_SKN_Annual_Data.xlsb

### 1. Workbook GDP_SKN_Annual_Data.xlsb

### Purpose and overview
- Produces estimates of annual GDP in both current and constant prices for Saint Kitts and Nevis combined (SKN) based on the selected base year.
- Used purely to produce the annual estimates needed as benchmarks for the published tables (as produced in the workbook GDP_SKN_Final.xlsb).
- The Instructions worksheet allows the compiler to select the reference year for GDP (the year used in the worksheet BM, see later).

### Workbook structure (7 groups) and purpose
- Group 1: Economic Activities
  - 27 worksheets, one for each economic activity, plus two for taxes and subsidies (T_S_Tax and T_S_Subs).
  - Each worksheet contains:
    - Links (light blue) to ‘indicator’ data from the GVA system.
    - Links to the BM worksheets (yellow) to the benchmark levels of Output, IC and GVA for each sub-activity. Note: these levels exclude the allocation of FISIM to economic activities, which is added later.
  - Broad procedure used in most activity worksheets:
    - Calculate indexes (light green) based on the indicator data. Indexes are referenced to the base year for the series (cell B2 in BM worksheet). Changing this value re-references all calculations.
    - Use the index (divided by 100) to extrapolate the base year level of output in current prices (CP), and for output in constant prices (KP) and similarly for CP intermediate consumption (IC). KP IC is derived by multiplying the KP output by the base year IC/O ratio (method known as ‘single extrapolation’).
    - GVA is then calculated in CP and KP as the difference between output and IC.

- Group 2: Benchmark Data
  - Worksheets: BM, BM_Add.
  - Contain benchmark levels of Output and IC (before and after inclusion of FISIM). BM_Add contains additional detailed information for the benchmark year as required.
  - Values in blue in BM are linked to the live GDP system and are the current estimate of Output and IC. Additional data entered into columns F and G get added to the published levels and the results in columns H:K set the levels of activity for each of the 51-component series.

- Group 3: GDP Excluding FISIM Allocation
  - Worksheets: Summary, GDP_KP_X, GDP_CP_X, O_KP, O_CP.
  - Show GDP before allocation for FISIM to IC of each economic activity. Working worksheets used in allocation of FISIM.

- Group 4: FISIM Calculations
  - Worksheets: FISIM_KP, FISIM_CP, FISIM_Sectors.
  - Calculate amount of FISIM to allocate to each activity. Allocation is based on the proportion of GVA in each activity.

- Group 5: Detailed Results
  - Worksheets: GDP_KP, GDP_CP, IC_KP, IC_CP.
  - Contain final detailed annual series in CP and KP needed for benchmarking of the final series (see GDP_SKN_Final.xlsb).
  - IC in IC_CP and IC_KP includes the allocation of FISIM from group 4.

- Group 6: Transposed Results
  - Worksheets: IC_KP_tran, IC_CP_tran, O_KP_tran, O_CP_tran.
  - Transpositions of results needed in workbook GDP_Q_Summary.xlsb.

- Group 7: Latest Year
  - Worksheet: Latest_Yr.
  - Indicates the latest year for which complete annual data have been entered in Activity worksheets. Compiler must specify which activities have been fully updated.

### Key calculation notes
- Index re-referencing: change cell B2 in BM to re-reference all calculations to a different base year (only when required).
- Single extrapolation method: KP IC = KP output × base year IC/O ratio.
- GVA calculations: GVA = Output − IC in both CP and KP (after IC includes allocated FISIM).

### Updating the GDP_SKN_Annual_Data.xlsb workbook (steps)
- Step 1: Update the GVA workbooks (use “Data->Edit links” in Excel and ‘Check status’). If issues, use ‘Change source’.
- Step 2: Update GDP_SKN_Annual_Data.xlsb by copying the formula ‘to the right’ for all ISIC worksheets in Groups 1, 2, 3, 4, 5, 9 and 10.
- Step 3: Update the Latest_Yr worksheet to specify the latest complete year for each activity.

---

### 2. Workbook GDP_SKN_Final.xlsb

### Purpose and overview
- Calculates the annual and quarterly GDP for SKN (and separately for Saint Kitts (SK) and Nevis (Nev)).
- Instructions worksheet allows selection of the latest quarter for results and units to be shown on final tables (EC$s or 000s EC$s).

### Workbook structure (8 groups) and purpose
- Group 1: Indicator Data
  - Worksheet: Data.
  - Links to each of the workbooks in the existing GDP system and takes total GVA in CP for each of SK and Nev and the price indicator for each (based on the GVA deflator).
  - Note: GVA deflators calculated in each ‘Quarterly Summary’ worksheet in the ISIC workbooks (example: Livestock deflators in columns AT:AU of ‘Quarterly Summary’).

- Group 2: Results for Economic Activities
  - 44 worksheets, one for each economic activity (including taxes and an empty worksheet for subsidies, since there are no subsidies after 2005).
  - Method overview (common to each worksheet):
    - Use CP GVA series for SK as indicator (column F) and the Price index for GVA for SK as elementary price index (column H).
    - Re-reference Price index to selected base year (in BM) in column I. Use it to deflate CP GVA indicator to derive KP GVA indicator for SKN (column H). Same procedure for Nevis in columns K:N. Columns C:D calculate indicators for SKN.
    - Columns AB:AG calculate summary annual indicator series for SK and for Nevis; total indicator for SKN in columns AH:AI. Ratio of SK to total SKN (CP and KP) shown in columns AJ:AK.
    - Actual annual SKN GVA (CP and KP) shown in columns AP and AT based on GDP_SKN_Annual_Data.xlsb (that workbook must be updated first).
    - Columns AX and AY: annual series for SKN extrapolated for all quarters in years after latest annual results based on sum of annual indicators (AH and AI) — bench‑marking step for later quarters.
    - SK and Nevis annual series calculated by applying proportion of SK to SKN (column AK) to total SKN series (column AX); Nevis derived as difference in columns BE:BF.
    - Benchmarking uses IMF ‘XLPBM’ function implemented as ‘=Bench ()’ in columns R:S for SKN and U:V for SK; Nevis in X:Y as difference.

- Group 3: Aggregate GDP Series
  - Worksheets: GDP_CP_A, GDP_KP_A, GDP_CP_Q, GDP_KP_Q.
  - Aggregate Group 2 results into time series in CP and KP, annual and quarterly, for SKN, SK, and Nevis.

- Group 4: The Aggregation Structure
  - Worksheet: Agg_Stru.
  - Controls aggregation across levels of GDP series and presentation in final tables.
  - Columns K:L: most detailed (‘elementary’) level corresponding to 44 worksheets in Group 2 and correlated to ISIC Rev 4 in columns I:J.
  - Columns A:B: higher-level aggregates in ISIC Rev 4.
  - Columns P:R: set level of detail shown in final tables. Tax and Subs must be left alone; other rows can be adapted.
  - Column R indicates if series is derived from addition of other rows (=1) or if aggregate (= blank).
  - Columns W:X and AD:AF provide alternative presentations that can be copied into P:R.
  - Examples provided for modifying presentation (e.g., replacing detailed rows 55 and 56 with aggregate row I and marking with ‘1’ in column R).

- Group 5: The Benchmark Levels
  - Worksheet: BM (linked from GDP_SKN_Annual_Data.xlsb).
  - Included for reference only; not used in calculations.

- Group 6: The Results (final publication tables)
  - Subgroup KP and CP including implicit price deflator (DEF).
  - Worksheets (final publication tables for quarterly series):
    - DEF_Q_SKN
    - CP_Q_SKN
    - CP_Q_SK
    - CP_Q_N
    - CP_A_SKN
    - CP_A_SK
    - CP_A_N
    - KP_A_SKN
    - KP_A_SK
    - KP_A_N
    - KP_Q_SKN
    - KP_Q_SK
    - KP_Q_N

- Group 7: Contributions Analysis
  - Worksheets: Cont_CP_A_SKN, Cont_KP_A_SKN.
  - Show contributions to annual growth in total GDP (in CP and KP), in percentage points, only for SKN combined and annual series.
  - Use for quality assurance to identify ISIC headings contributing most to annual percentage changes; basis for supplementary tables.

- Group 8: Revisions Analysis
  - Worksheets: Prev, Rev_CP, Rev_KP.
  - Developed to establish reasons for revisions in KP to growth in GDP; not required for regular annual calculation.
  - Prev contains results as published in May 2023 (based on ISIC Rev 3.1).
  - Revs_CP and Revs_KP present analysis of revisions arising from rebasing, comparing published series with rebased series (all based on ISIC 3.1).

### Updating the GDP_SKN_Final.xlsb workbook (steps)
- Update individual activity workbooks (as per Annual_Data step 1) and ensure GDP_SKN_Annual_Data.xlsb is fully updated.
- Calculations are automatic thereafter.
- To ensure links to individual workbooks are updated, open each of the 44 workbooks and check links via Data->Edit Links->Check Status.

---

### C. Updating the GDP-P System (notes)
- GDP-P is built on-top of the existing GVA system to allow regular data capture to continue.
- No changes to calculations in existing GVA workbooks required, though some additional calculations were added to worksheets (example: [ISIC A011 - Crops.xlsb] includes calculation of the constant price series for total crop output based on base year in BM worksheet; see row 397 in [Production & Prices] worksheet).
- Steps to update summarized above for both workbooks.

---

### Appendix II. Real Time Revisions Studies — method and metrics

### Purpose and definition
- Simulates revisions that arise from benchmarking by producing series of estimates for each period (’maturities’) and comparing successive estimates to show evolution and resulting revisions.
- Real-time study analyzes revisions purely from benchmarking quarterly to annual series; it does not simulate revisions from changes to underlying data sources.

### Data required
- Annual estimates of the level of constant price output for the activity.
- Annual sums of the quarterly constant price indicator data for the activity.
- Analysis can also be undertaken based on current price data.

### Step-by-step guide
- Step 1: Populate framework table for each year with actual annual estimates (shown in yellow). Extrapolate last annual estimate in each column by percentage change in annual (sum of quarters) based on indicator data (shown in green).
  - Each column represents a point in time (e.g., column ‘2016’ is when annual data for 2016 were available but none for 2017 and 2018).
- Step 2: Calculate annual rates of growth based on the table.
  - Example annual growth rates table (selected entries shown exactly as source):
    - 2015 2% 11% 11% 11% 11% 11% 11% 11%
    - 2016 -7% -7% -16% -16% -16% -16% -16% -16%
    - 2017 -5% -3% -4% -4% -4% -4% -4%
    - 2018 -8% -7% -1% -1% -1% -1%
    - 2019 -8% -10% -12% -12% -12%
    - 2020 -8% -7% -3% -3%
    - 2021 -9% -11% -8%
    - 2022 -9% -11% -10%
- Step 3: Calculate summary metrics using aligned maturities (example layout for 1-year ahead ‘forecasts’ and 2-year ahead ‘forecasts’).

### Example 1: 1-year ahead ‘forecasts’ summary (exact figures)
- Table columns: M1, Latest, Revision, sq revision.
- Rows:
  - 2015 2% 11% 9.0% 0.80%
  - 2016 -7% -16% -9.3% 0.87%
  - 2017 -3% -4% -0.7% 0.00%
  - 2018 -7% -1% 6.2% 0.38%
  - 2019 -10% -12% -1.4% 0.02%
  - 2020 -7% -3% 4.5% 0.20%
  - 2021 -11% -8% 3.4% 0.11%
- Summary metrics:
  - Bias 1.7%
  - RMSE 5.8%
  - Avg abs growth 7.8%
  - RMSE/ avg abs growth 75%

- Definitions:
  - Revisions = Latest − M1 estimates of growth.
  - RMSE is square root of average of squared revisions.
  - Average absolute growth is average of absolute values of the ‘Latest’ column.
  - RMSE / avg abs growth provides metric for size of revisions compared to growth rates; rule of thumb: value <10 percent signifies revisions relatively small compared to growth rates.
  - Bias is average of revisions; bias close to zero desirable.

### Example 2: 2-year ahead ‘forecasts’ summary (exact figures)
- Rows:
  - 2016 -7% -16% -9% 0.76%
  - 2017 -5% -4% 1% 0.01%
  - 2018 -8% -1% 6% 0.40%
  - 2019 -8% -12% -4% 0.13%
  - 2020 -8% -3% 5% 0.26%
  - 2021 -9% -8% 1% 0.01%
- Summary metrics:
  - Bias 0.2%
  - RMSE 5.1%
  - Avg abs growth 6.2% 3.7%
  - RMSE/ avg abs growth 83%

*Source: IMF Technical Assistance Report – Saint Kitts & Nevis Improving Estimates of GDP (workbook descriptions and Appendix II).*

### Appendix III. Data Sources for Supply and Use Tables

### Appendix III. Data Sources for Supply and Use Tables

### Data sources by SUT component
- Domestic Output
  - Data from NAS survey, and VAT/CIT tax system, plus existing estimates based on GDP-P.
- Trade/transport margins
  - Data from NAS survey for broad product groups
- Imports of goods
  - Eurotrace data based on HS 8-digit headings
- Imports of services
  - Balance of Payments Statistics (coded to EBOPS)
- Taxes on products
  - VAT: Total non-dedicational VAT, based on government revenue statement
  - Trade taxes: Customs data
  - Excise taxes: Government revenues statement
- Subsidies on products
  - Government expenditure account
- Intermediate demand (by product)
  - Broad percentages of intermediate consumption by activity derived of the NAS survey allocated to products. May also borrow from SUTs for nearby countries (GRE, VCT, BAR, ATG, and LCA)
- Gross fixed capital formation
  - (No explicit single-source line provided in the excerpt)
- Changes in inventories
  - May be possible to Ise molded elates based on imports.
- Acquisition less disposal of valuables
  - Eurotrace data based on HS 8-digit headings
- Final expenditure of General Government
  - Government accounts
- Final expenditure of Non-profit organizations serving households
  - Data from NAS survey, and VAT/CIT tax system, plus existing estimates based on GDP-P.
- Final expenditure of Households
  - Household Budget Surrey (2019/19) price and volume updated
- Exports of goods
  - Eurotrace data based on HS 8-digit headings
- Exports of services
  - Balance of Payments Statistics (coded to EBOPS)

*IMF | Technical Assistance Report – Saint Kitts & Nevis Improving Estimates of GDP | 33*

### Appendix IV. Proposed Dimensions of Supply and Use Tables for 2024

### Potential products and industries for Supply and Use Tables for Saint Kitts and Nevis
- 1. No. 1 — Code A010 — Product: Crops
  - Industry No. 1 — Code 010 — Growing of crops (011, 012 and 013)
- 2. No. 2 — Code A014 — Product: Live animals and animal products (excluding meat)
  - Industry No. 2 — Code 014 — Raising of livestock (including hunting and trapping)
- 3. No. 3 — Code A021 — Product: Logs and other forestry products
  - Industry No. 3 — Code 02 — Forestry and logging
- 4. No. 4 — Code A031 — Product: Fish and other fishing products
  - Industry No. 4 — Code 03 — Fishing
- 5. No. 5 — Code B061 — Product: Crude petroleum
  - Industry No. 5 — Code 061 — Extraction of crude petroleum
- 6. No. 6 — Code B062 — Product: Natural gas
  - Industry No. 6 — Code 062 — Extraction of natural gas
- 7. No. 7 — Code B081 — Product: Quarrying products
  - Industry No. 7 — Code 081 — Quarrying
- 8. No. 8 — Code B089 — Product: Other mining
  - Industry No. 8 — Code 089 — Other mining
- 9. No. 9 — Code C011 — Product: Meat and meat products
  - Industry No. 10 — Code 101-102 — Processing and preserving of meat and fish
- 10. No. 10 — Code C020 — Product: Processed fish
- 11. No. 11 — Code C030 — Product: Processed fruit and vegetables
  - Industry No. 11 — Code 103-104 — Processing of fruit and vegetables, and edible oils
- 12. No. 12 — Code C040 — Product: Edible oils and fats
- 13. No. 13 — Code C050 — Product: Dairy products
  - Industry No. 12 — Code 105 — Manufacture of dairy products
- 14. No. 14 — Code C068 — Product: Grain mill and starch products and animal feed
  - Industry No. 13 — Code 106&108 — Manufacture of grain mill products and animal feeds
- 15. No. 15 — Code C071 — Product: Bakery products
  - Industry No. 14 — Code 1071 — Manufacture of bakery products
- 16. No. 16 — Code C079 — Product: Other food products n.e.c.
  - Industry No. 15 — Code 1079 — Manufacture of other food products
- 17. No. 17 — Code C111 — Product: Alcoholic beverages
  - Industry No. 16 — Code 1101-03 — Manufacture of alcoholic beverages
- 18. No. 18 — Code C114 — Product: Non-alcoholic beverages
  - Industry No. 17 — Code 1104 — Manufacture of non-alcoholic beverages
- 19. No. 19 — Code C120 — Product: Tobacco products
  - Industry No. 18 — Code 12 — Manufacture of tobacco products
- 20. No. 20 — Code C130 — Product: Textiles, wearing apparel and leather products
  - Industry No. 19 — Code 13-15 — Manufacture of textiles, wearing apparel and leather products
- 21. No. 21 — Code C160 — Product: Cork, wood and straw products
  - Industry No. 20 — Code 16 — Manufacture of products of cork, wood and straw
- 22. No. 22 — Code C170 — Product: Paper and paper products
  - Industry No. 21 — Code 17 — Manufacture of paper and paper products
- 23. No. 23 — Code C180 — Product: Printed and recorded media
  - Industry No. 22 — Code 18 — Printing and reproduction of recorded media
- 24. No. 24 — Code C191 — Product: Liquified natural gas
  - Industry No. 23 — Code 19 — Manufacture of coke and refined petroleum products
- 25. No. 25 — Code C192 — Product: Other petroleum products
- 26. No. 26 — Code C201 — Product: Petrochemicals
  - Industry No. 24 — Code 201 — Manufacture of Petrochemicals
- 27. No. 27 — Code C202 — Product: Fertilizer, nitrogen, and agrochemical products
  - Industry No. 25 — Code 202 — Manufacture of other chemical products
- 28. No. 28 — Code C203 — Product: Paint, varnishes and similar coatings, printing ink
- 29. No. 29 — Code C204 — Product: Soap and detergents, cleaning and polishing preparations, perfumes, and toilet preparations
- 30. No. 30 — Code C205 — Product: Other chemical products
- 31. No. 31 — Code C210 — Product: Pharmaceuticals, medicinal and botanical products
  - Industry No. 26 — Code 21 — Manufacture of pharmaceuticals, medicinal chemical, and botanical products
- 32. No. 32 — Code C220 — Product: Rubber and plastic products
  - Industry No. 27 — Code 22 — Manufacture of rubber and plastics products
- 33. No. 33 — Code C231 — Product: Glass and glass products
  - Industry No. 28 — Code 2310 — Manufacture of glass and glass products
- 34. No. 34 — Code C232 — Product: Bricks, tiles, and other ceramic products
  - Industry No. 29 — Code 2391 — Manufacture of refractory products
- 35. No. 35 — Code C233 — Product: Cement, lime, and plaster
  - Industry No. 30 — Code 2394 — Manufacture of cement, lime, and plaster
- 36. No. 36 — Code C234 — Product: Articles of concrete, cement, and plaster
  - Industry No. 31 — Code 2399 — Manufacture of other non-metallic mineral products
- 37. No. 37 — Code C235 — Product: Other non-metallic products n.e.c.
- 38. No. 38 — Code C240 — Product: Basic metals
  - Industry No. 32 — Code 24-25 — Manufacture of basic and fabricated metals
- 39. No. 39 — Code C250 — Product: Fabricated and other metal products
- 40. No. 40 — Code C260 — Product: Computer, electronic and optical products
  - Industry No. 33 — Code 26-28 — Manufacture of machinery and equipment
- 41. No. 41 — Code C270 — Product: Electrical equipment
- 42. No. 42 — Code C280 — Product: Machinery and equipment n.e.c.
- 43. No. 43 — Code C291 — Product: Motor vehicles/cycles
- 44. No. 44 — Code C292 — Product: Motor vehicle/cycle supplies and parts
  - Industry No. 34 — Code 29-30 — Manufacture of transport equipment
- 45. No. 45 — Code C301 — Product: Boats and other transport equipment and supplies/parts
- 46. No. 46 — Code C310 — Product: Furniture
  - Industry No. 35 — Code 31 — Manufacture of furniture
- 47. No. 47 — Code C321 — Product: Other manufacturing
  - Industry No. 36 — Code 32 — Other manufacturing
- 48. No. 48 — Code C330 — Product: Repair and installation of machinery and equipment
  - Industry No. 37 — Code 33 — Repair and installation of machinery and equipment
- 49. No. 49 — Code D351 — Product: Electricity supply
  - Industry No. 38 — Code 351 — Electric power generation, transmission, and distribution
- 50. No. 50 — Code E361 — Product: Water supply via pipeline
  - Industry No. 39 — Code 352 — Distribution of gaseous fuels through mains
- 51. No. 51 — Code E362 — Product: Water supply trucked
  - Industry No. 40 — Code 36-37 — Water collection, treatment and supply, and sewerage
- 52. No. 52 — Code E370 — Product: Sewerage
- 53. No. 53 — Code E380 — Product: Waste collection, treatment, and disposal activities; materials recovery
  - Industry No. 41 — Code 38-39 — Waste collection, treatment, and disposal activities; materials recovery
- 54. No. 54 — Code F410 — Product: Construction
  - Industry No. 42 — Code 41 — Construction
- 55. No. 55 — Code G451 — Product: Sale of motor vehicles and motorcycles
  - Industry No. 43 — Code 45 — Sales and repair of motor vehicles and motorcycles
- 56. No. 56 — Code G452 — Product: Repairs of motor vehicles and motorcycles
  - Industry No. 44 — Code 46 — Wholesale trade
- 57. No. 57 — Code G460 — Product: Wholesale and retail trade
  - Industry No. 45 — Code 471 — Other retail trade
- 58. No. 58 — Code H491 — Product: Passenger transport by road
  - Industry No. 46 — Code 473 — Distribution of refined petroleum products
- 59. No. 59 — Code H492 — Product: Freight transport by road
  - Industry No. 47 — Code 4922 — Passenger transport by road
- 60. No. 60 — Code H501 — Product: Water passenger transport
  - Industry No. 48 — Code 4923 — Freight transport by road
- 61. No. 61 — Code H502 — Product: Water freight transport
  - Industry No. 49 — Code 50 — Water transport
- 62. No. 62 — Code H511 — Product: Air passenger transport
  - Industry No. 50 — Code 51 — Air transport
- 63. No. 63 — Code H512 — Product: Air freight transport
  - Industry No. 51 — Code 521 — Warehousing and storage
- 64. No. 64 — Code H521 — Product: Warehousing and storage
  - Industry No. 52 — Code 5222 — Service activities incidental to water transportation
- 65. No. 65 — Code H523 — Product: Seaport services
  - Industry No. 53 — Code 5223 — Service activities incidental to air transportation
- 66. No. 66 — Code H524 — Product: Airport services
  - Industry No. 54 — Code 5224 — Other transport support services
- 67. No. 67 — Code H525 — Product: Cargo handling and other transport support services
  - Industry No. 55 — Code 531 — Postal activities
- 68. No. 68 — Code H531 — Product: Postal services
  - Industry No. 56 — Code 532 — Courier activities
- 69. No. 69 — Code H532 — Product: Courier services
  - Industry No. 57 — Code 551 — Hotels and resorts
- 70. No. 70 — Code I55 — Product: Accommodation services
  - Industry No. 58 — Code 552 — Other accommodation services
- 71. No. 71 — Code I56 — Product: Food and beverage services
  - Industry No. 59 — Code 561 — Food catering services (predominantly)
- 72. No. 72 — Code J58 — Product: Books, newspapers and other published products and software
  - Industry No. 60 — Code 562 — Beverage catering services (predominantly)
- 73. No. 73 — Code J590 — Product: Audio-visual production and distribution
  - Industry No. 61 — Code 58 — Publishing activities
- 74. No. 74 — Code J601 — Product: Radio and TV broadcasting
  - Industry No. 62 — Code 59 — Audio-visual production and distribution
- 75. No. 75 — Code J602 — Product: Subscriber cable TV services
  - Industry No. 63 — Code 60 — Broadcasting and programming activities
- 76. No. 76 — Code J61 — Product: Telecommunication and internet services
  - Industry No. 64 — Code 61 — Telecommunications
- 77. No. 77 — Code J62 — Product: Computer services
  - Industry No. 65 — Code 62 — Computer programming, consultancy, and related activities
- 78. No. 78 — Code J63 — Product: Information service activities
  - Industry No. 66 — Code 63 — Information services
- 79. No. 79 — Code K641 — Product: Monetary authority and financial supervision
  - Industry No. 67 — Code 6411 — Central Bank of Trinidad and Tobago
- 80. No. 80 — Code K642 — Product: FISIM on loans and deposits
  - Industry No. 68 — Code 6412 — Banks (FISIM)
- 81. No. 81 — Code K643 — Product: Explicit financial charges and fees
  - Industry No. 69 — Code 6419 — Banks (non-FISIM)
- 82. No. 82 — Code K651 — Product: Life Insurance services
  - Industry No. 70 — Code 65 — Insurance services, pension funds
- 83. No. 83 — Code K652 — Product: Non-life Insurance and pension management services
- 84. No. 84 — Code K653 — Product: Reinsurance services
- 85. No. 85 — Code K661 — Product: Other financial services
  - Industry No. 71 — Code 66 — Other financial and insurance services
- 86. No. 86 — Code K662 — Product: Insurance agents and brokers
- 87. No. 87 — Code L681 — Product: Imputed rents /OOD
  - Industry No. 72 — Code 6811 — Owner-occupied dwellings
- 88. No. 88 — Code L682 — Product: Actual rents on rented dwellings
  - Industry No. 73 — Code 6812 — Rented dwellings
- 89. No. 89 — Code L683 — Product: Actual rents on commercial property
  - Industry No. 74 — Code 6813 — Commercial property rentals
- 90. No. 90 — Code L684 — Product: Real estate agents and property managers
  - Industry No. 75 — Code 6820 — Real estate activities on a fee or contract basis
- 91. No. 91 — Code M691 — Product: Legal services
  - Industry No. 76 — Code 69 — Legal and accounting services
- 92. No. 92 — Code M692 — Product: Accounting and auditing services
- 93. No. 93 — Code M700 — Product: Business and management consultancies
  - Industry No. 77 — Code 70 — Activities of head offices; management consultancy activities
- 94. No. 94 — Code M710 — Product: Architecture, engineering and technical
  - Industry No. 78 — Code 71 — Architecture and engineering activities; technical testing and analysis
- 95. No. 95 — Code M730 — Product: Advertising services and market research
  - Industry No. 79 — Code 73 — Advertising and market research
- 96. No. 96 — Code M724 — Product: Other professional, scientific, research and technical services
  - Industry No. 80 — Code 72&74 — Other professional, scientific, research and technical activities
- 97. No. 97 — Code M750 — Product: Veterinary services
  - Industry No. 81 — Code 75 — Veterinary activities
- 98. No. 98 — Code N771 — Product: Renting of motor vehicles
  - Industry No. 82 — Code 77 — Renting and leasing of motor vehicles, machinery and equipment, and personal and households
- 99. No. 99 — Code N772 — Product: Renting of machinery and equipment, personal and household goods
- 100. No. 100 — Code N791 — Product: Travel agent services
  - Industry No. 83 — Code 791 — Travel agency services
- 101. No. 101 — Code N792 — Product: Tour operator services
  - Industry No. 84 — Code 792 — Tour operator and other reservation service activities
- 102. No. 102 — Code N800 — Product: Security and investigation activities
  - Industry No. 85 — Code 80 — Security and investigation activities
- 103. No. 103 — Code N810 — Product: Building cleaning and landscaping activities
  - Industry No. 86 — Code 81 — Services to buildings and landscape activities
- 104. No. 104 — Code N820 — Product: Other business and employment support services
  - Industry No. 87 — Code 78&82 — Office administrative and support activities
- 105. No. 105 — Code O84 — Product: Public administration and defense; social security
  - Industry No. 88 — Code 84 — Public administration and defense; social security
- 106. No. 106 — Code P851 — Product: Pre-primary/primary education Private sector
  - Industry No. 89 — Code 851 — Pre-primary/primary education
- 107. No. 107 — Code P852 — Product: Pre-primary/primary education public sector
- 108. No. 108 — Code P853 — Product: Secondary general, technical, and vocational education Private
  - Industry No. 90 — Code 852 — Secondary general, technical, and vocational education
- 109. No. 109 — Code P854 — Product: Secondary general, technical, and vocational education Public
- 110. No. 110 — Code P855 — Product: Tertiary education Private
  - Industry No. 91 — Code 853 — Higher education
- 111. No. 111 — Code P856 — Product: Tertiary education Public
- 112. No. 112 — Code P857 — Product: Private tutoring and other education n.e.c.
  - Industry No. 92 — Code 854 — Other education
- 113. No. 113 — Code Q861 — Product: Human health services Private
  - Industry No. 93 — Code 86 — Human health services
- 114. No. 114 — Code Q862 — Product: Human health services Public
- 115. No. 115 — Code Q871 — Product: Social work activities Private
  - Industry No. 94 — Code 878 — Social work activities
- 116. No. 116 — Code Q872 — Product: Social work activities Public
- 117. No. 117 — Code R900 — Product: Creative arts and entertainment
  - Industry No. 95 — Code 90 — Creative, arts and entertainment activities
- 118. No. 118 — Code R910 — Product: Libraries, museums, historical sites
  - Industry No. 96 — Code 91 — Libraries, archives, museums and other cultural activities
- 119. No. 119 — Code R920 — Product: Gambling and betting
  - Industry No. 97 — Code 92 — Gambling and betting activities
- 120. No. 120 — Code R930 — Product: Sports clubs, gyms and other activities
  - Industry No. 98 — Code 93 — Sports activities and amusement and recreation activities
- 121. No. 121 — Code S940 — Product: Membership organizations
  - Industry No. 99 — Code 94 — Activities of membership organizations
- 122. No. 122 — Code S950 — Product: Repair of other personal and household goods
  - Industry No. 100 — Code 95 — Repair of computers and personal and household goods
- 123. No. 123 — Code S961 — Product: Washing and (dry-) cleaning of textile and fur products
  - Industry No. 101 — Code 9601 — Washing and (dry-) cleaning of textile and fur products
- 124. No. 124 — Code S962 — Product: Hairdressing and other beauty treatment
  - Industry No. 102 — Code 9602 — Hairdressing and other beauty treatment
- 125. No. 125 — Code S963 — Product: Funeral and related activities
  - Industry No. 103 — Code 9603 — Funeral and related activities
- 126. No. 126 — Code S964 — Product: Other personal service activities n.e.c.
  - Industry No. 104 — Code 9609 — Other personal service activities n.e.c.
- 127. No. 127 — Code T970 — Product: Activities of households as employers of domestic personnel
  - Industry No. 105 — Code 97 — Activities of households as employers of domestic personnel

*IMF | Technical Assistance Report – Saint Kitts & Nevis Improving Estimates of GDP | 34–37*

---


_Source: https://www.imf.org/-/media/files/publications/tar/2024/english/tarea2024063-print-pdf.pdf_