St. Lucia Technical Assistance Report-Rebasing GDP Mission
Author:
International Monetary Fund. Statistics Dept.
Search for other papers by International Monetary Fund. Statistics Dept. in
Current site
Google Scholar
PubMed
Close

A technical assistance mission was undertaken by the Real Sector Statistics Advisor in the Caribbean Regional Technical Assistance Centre to Saint Lucia to provide advice to the Central Statistics Office (CSO) on compiling rebased gross domestic product estimates. The CSO is responding to the needs of the Ministry of Finance for more robust and timely national accounts statistics. All the Gross domestic product by economic activity (GDP-P) compilation workbooks have now been redeveloped and revised current and constant 2018 price quarterly and annual estimates have been compiled up to Q3 2019. The incorporation of revised data on tourist expenditure for 2000 onward have also resulted in revisions to the GDP-P current rice estimates and real growth rates. The revised annual and quarterly GDP-P estimates were assessed, and several methodological improvements were implemented. Improvements were made to the constant price estimates by reviewing and replacing weaker volume indicators. Training on the methodological changes and compiling the rebased estimates has been provided. The training on methodological improvements included the use of the more representative employment indicators and various price indices discussed above; back-casting and linking techniques for the current price estimates and linking the constant 2006 price series with the constant 2018 price series.

Abstract

A technical assistance mission was undertaken by the Real Sector Statistics Advisor in the Caribbean Regional Technical Assistance Centre to Saint Lucia to provide advice to the Central Statistics Office (CSO) on compiling rebased gross domestic product estimates. The CSO is responding to the needs of the Ministry of Finance for more robust and timely national accounts statistics. All the Gross domestic product by economic activity (GDP-P) compilation workbooks have now been redeveloped and revised current and constant 2018 price quarterly and annual estimates have been compiled up to Q3 2019. The incorporation of revised data on tourist expenditure for 2000 onward have also resulted in revisions to the GDP-P current rice estimates and real growth rates. The revised annual and quarterly GDP-P estimates were assessed, and several methodological improvements were implemented. Improvements were made to the constant price estimates by reviewing and replacing weaker volume indicators. Training on the methodological changes and compiling the rebased estimates has been provided. The training on methodological improvements included the use of the more representative employment indicators and various price indices discussed above; back-casting and linking techniques for the current price estimates and linking the constant 2006 price series with the constant 2018 price series.

Summary of Mission Outcomes and Priority Recommendations

1. A technical assistance (TA) mission was undertaken by the Real Sector Statistics Advisor in the Caribbean Regional Technical Assistance Centre (CARTAC) to Saint Lucia during January 27-February 7, 2020, to provide advice to the Central Statistics Office (CSO) on compiling rebased GDP estimates. The 2006 base year for the GDP estimates is outdated and does not reflect the current structure of the economy. There is scope to improve the data sources and GDP compilation methodology and to implement the relevant System of National Accounts 2008 (2008 SNA) recommendations.

2. The CSO is responding to the needs of the Ministry of Finance (MOF) for more robust and timely national accounts statistics. In addition to producing annual and quarterly current and constant price GDP by economic activity (GDP-P), the CSO also compiles annual current price GDP by expenditure estimates. The supply and use tables (SUT) for 2016 have been compiled, with the benchmark estimates, based on the International Standand Industrial Classification of All Economic Activities (ISIC) Revision 4, have been incorporated into the redeveloped compilation system. Implementation of the relevant 2008 SNA changes, benchmark estimates, and improved statistical techniques should result in more robust debt and fiscal revenue to GDP ratios, real GDP growth rates, and internationally comparable GDP estimates.

3. The SUT benchmark estimates were reviewed and further improvements were incorporated. The revised GDP level for 2016 is 9.3 percent higher than the latest published estimate. Improved coverage was the main contributor to the increased GDP level, partly offset by higher intermediate consumption (IC) to output ratios. Improved coverage of activities of non-financial corporations and credit unions contributed 57 percent, informal sector activities and production for own consumption contributed 34 percent, extrabudgetary government units contributed 12 percent, and non-profit institutions serving households (NPISHs) contributed 2 percent. Partly offsetting these increases were improved coverage of IC of financial intermediation services indirectly measured (FISIM) for credit unions (-4 percent), and product subsidies (-2 percent).

4. Revised current and constant 2018 price quarterly and annual GDP-P estimates have been compiled up to Q3 2019. The revised benchmark SUT estimates and the latest input data for Q3 2019 were incorporated into the redeveloped compilation workbooks.

5. These revised annual and quarterly GDP-P estimates at current and constant prices were reviewed during the mission and methodological improvements were implemented. Improvements were made in recompiling the current price estimates back to 2000 for most industries, although a few could only be recompiled back to 2006 due to data constraints. In most instances, where the 2006 base year IC to output ratios were considered robust, a wedge was used to adjust the IC to output ratios for 2007 to 2015 to ensure value added was not over or underestimated, and reflected only the improved coverage of the relevant industries output. For other industries, like fishing and business services, the 2006 IC to output ratios were adjusted to ensure the value added was realistic but still consistent with the improved coverage of the activities. Incorporation of recently revised travel services credits data from 2018 back to 2000 has also resulted in further revisions in the current price estimates.

6. The constant price estimates were improved for a number of industries. This was achieved by replacing weaker volume indicators with more robust employment data from the National Insurance Corporation (NIC) for a few manufacturing and several services industries, and reweighting and replacing component price indices for composite IC price indices for some industries, such as other crops, quarrying and sea transport. For years prior to 2016, the constant price estimates have been linked and back cast while maintaining the real growth rates from the previous 2006 base year series as far as possible.

7. Training on the methodological changes and compiling the revised GDP-P estimates was provided to the CSO staff. The training on methodological improvements included the use of more representative indicators and prices, wedging and linking techniques.

8. The mission and the Director and national accounts staff of the CSO met with the Minister and Permanent Secretary of Economic Development. The changes to the benchmark SUT estimates, quality of the revised GDP-P estimates, and the timetable for the release of the rebased estimates were explained. The annual and quarterly GDP-P estimates are to be updated to Q4 2019 and shared with key data users by February 21, 2020. The estimates will then be disseminated by April 3, 2020. The mission advised the Director of the need to ensure that the national accountants are not given any other duties for the next two weeks.

The action plan to finalize and release the final GDP-P estimates was agreed with the CSO Director. The following priority recommendations were identified:

Table 1.

Priority Recommendations

article image

Compiling the Sut and Rebasing of GDP

A. Update of the SUT Benchmark Estimates

9. Further improvements were made to the benchmark estimates and the SUT was rebalanced. As a result of better industry coding of the NIC employment data, it was possible to calculate more robust estimates of production per worker. This revealed under or overestimation of output and value added for a few industries in the SUT. Further review of the value added tax (VAT) system data and household final consumption expenditure (HFCE) data led to recalculating the production estimates for: electricity, gas and water industry and retail and wholesale trade industry, as a result of reclassifying sale of gas in cylinders; domestic services activity and the other administrative and support services industry due to reclassification of gandening, house cleaning and security services; and to the air transport industry due to the treatment of branch operations of foreign airlines. In addition, older data revisions for the construction industry, and more recent data revisions for accommodation, food and beverage services, taxis, and tour operators were incorporated. Further adjustments were then made to the intermediate demand for the SUT products and IC estimates by industry. The trade margins and taxes on products were then readjusted and the SUT was rebalanced.

10. The revised GDP level for 2016 is 9.3 percent higher than the latest published estimate. The increase in the GDP level is essentially due to better coverage of economic activities, contributing 137 percent of the revision. Partly offsetting these increases were higher IC to output ratios contributing -38 percent of the change. Data revisions contributed 0.8 percent.

Table 2.

Reasons for the Change in the GDP Level

article image

11. Improvements in the source data were the main reason for the increased coverage. Improved coverage of activities of non-financial corporations and credit unions contributed 57 percent, informal sector activities and production for own consumption contributed 34 percent, extrabudgetary government units contributed 12 percent, and NPISHs contributed 2 percent. Partly offsetting these increases were improved coverage of the IC of FISIM for credit unions (-4 percent), and product subsidies (-2 percent).

12. The summary SUT estimates are included in Appendix I and more detailed information at the industry level on the contributors to the change in the GDP level is provided in Appendix II.

B. Compilation of the Revised GDP Estimates

13. All of the GDP-P compilation workbooks have now been redeveloped and revised current and constant 2018 price quarterly and annual estimates have been compiled up to Q3 2019. The redeveloped compilation workbooks include improved production estimates classified according to ISIC Revision 4. The rebalanced SUT estimates, relevant 2008 SNA changes, and improved indicators and compilation techniques have been incorporated. In terms of 2008 SNA changes, the imputed bank services charge method has been replaced by the FISIM method and FISIM is being allocated by industry and sector; and the consumption of fixed capital (CFC) is being compiled for Government and added to the gross value added (GVA) for public administration. In addition to incorporating the Q3 2019 indicator data and price indices, data revisions to tourist expenditure for 2000 onwands have been incorporated, and the value and volume indicators for telecommunications were updated for 2018.

14. The incorporation of revised data on tourist expenditure for 2000 onwands have also resulted in revisions to the GDP-P current rice estimates and real growth rates. The revised travel services credits for 2000 to 2014 were received following the CARTAC balance of payments mission during January 27–31, 2020. The updated data for 2015 onwands on the expenditure of stayover tourists, cruise ship passengers, excursionists and yachties by type of expenditure was provided by the Eastern Caribbean Central Bank (ECCB). As a result, the production estimates for accommodation, food and beverage services, taxis and tour operators were revised. The GDP-P estimates should now be fully aligned with the travel credits estimates in Saint Lucia’s balance of payments.

15. The revised annual and quarterly GDP-P estimates were assessed and a number of methodological improvements were implemented. As separate output, IC and GVA estimates are compiled it has been possible to make improvements in the current price estimates back to 2000 for most industries, although a few could only be recompiled back to 2006 due to data constraints. For industries where the 2006 base year estimates of IC to output ratios were based on the 2002 SUT and considered robust, a wedge was used to adjust the IC to output ratios for 2007 to 2015 to ensure the GVA was not over or underestimated, and reflected only the improved coverage of the relevant industries output. Examples include other food manufacturing, paints and fabricated metal manufacturing, and wholesale and retail trade industries. For a number of manufacturing industries, such as, dairy, bakery, other food and beverages manufacturing; paper and printing; chemicals and paints; plastics; fabricated metals, furniture, other manufacturing; publishing; and insurance services, the 2006 estimates of output had good coverage. In these cases, the output was wedged so that the increased coverage identified in 2016 was reflected in the 2007 to 2015 estimates but not for 2006 and prior years. For other industries, like fishing and business services, where the 2006 IC to output ratios were not based on survey data, the ratios were adjusted upwands before a wedge was used to ensure the value added was realistic but still consistent with the improved coverage of the activities.

16. Improvements were made to the constant price estimates by reviewing and replacing weaker volume indicators. The revised NIC employment data classified to ISIC Revision 4 was used for the wearing apparel, rubber and plastics, fabricated metals and other manufacturing, replacing weaker volume indicators from the industrial production indices. The NIC employment data has now also been used for travel agents, professional services, other administrative and support services, and other personal and social services and several services industries, replacing less suitable indicators like population projections.

17. The IC and output price indices were also improved. The composite IC price index weights were updated for all industries using the SUT benchmark data, although these price indices are used mainly to reflate quarterly constant price estimates to derive value indicators that are benchmarked to annual current price IC estimates to compile the quarterly current price IC estimates. The component price indices for composite IC price indices for some industries, such as other crops, quarrying and sea transport, were changed based on the use of products breakdown from the SUT. The output price indices for education and other services was also changed. The adjusted consumer price index (CPI) for education was replaced with a weighted price index (using the SUT data), comprising the price indices for the three different levels of education. For arts, entertainment and recreation, and for other services, the adjusted non-food CPI has been replaced with the adjusted services CPI.

18. For years prior to 2016, the constant price estimates have been linked and back cast by reversing the real growth rates from the previous 2006 base year series. The 2006 series workbooks for accommodation, food and beverage services, taxis and tour operators have now been updated to include the revisions to the travel services estimates; and the growth rates from the revised constant price estimates for 2006 onwands have been reversed and used to back cast the constant 2018 price series from the 2016 link year. More detailed information on the changes in the GDP-P methodology at the industry level is provided in Appendix III, while the revised industry GVA and GDP-P estimates are included in Appendix IV.

19. Revised annual GDP-P current and constant 2006 price estimates for 2006 to 2018 have been compiled for the ECCB. The estimates of GDP and GVA by industry are classified according to ISIC Revision 3.1. This additional task was undertaken during the mission in order to ensure that the ECCB had a consistent set of estimates to add to the estimates of other member states in order to continue to calculate total GDP-P in current and constant 2006 prices for the Eastern Caribbean Currency Union. The first step was to update the current price annual estimates for 2006 to 2018. Next the previously released constant price growth rates were applied to the revised 2006 base year estimates to derive the updated constant 2006 price estimates for 2006 to 2016. From Q1 2017 onwands, the real growth rates from the 2018 GDP series are applied to the 2006 GDP series at the elementary and aggregated industry levels. This means that the sum of the detailed industries’ constant 2006 price estimates will not equal the equivalent period’s aggregated industries or total GDP-P constant price estimates for Q1 2017 onwands, as they are not additive. This arrangement was agreed between the Director of the CSO and the Director of Statistics at the ECCB.

20. Training on the methodological changes and compiling the rebased estimates has been provided. The training on methodological improvements included the use of the more representative employment indicators and various price indices discussed above; back-casting and linking techniques for the current price estimates, and linking the constant 2006 price series with the constant 2018 price series.

21. The national accounts compilers need to update the annual and quarterly GDP-P estimates to Q4 2019 and produce the initial GDP-P estimates for 2019 by February 21, 2020. These revised GDP-P estimates will then be shared with the Finance Department, Eastern Caribbean Central Bank, and the IMF mission team, and are expected to be released by April 3, 2020.

Recommended actions:

  • Complete checking the formulae in the compilation workbooks and switch to the XLPBM add-in in the shared drive.

  • Incorporate updated indicator data and price indices for Q4 2019.

  • Compile the GDP-P estimates for Q4 2019 and initial 2019 estimates.

  • Share the updated estimates with the ECCB, MOF and IMF mission team.

  • Release the revised current and rebased constant 2018 price annual and quarterly GDP-P estimates.

Detailed Technical Assessment and Recommendations

article image

A. Officials Met During the Mission

article image

Appendix I. Summary Supply and Use Tables

article image
article image

Appendix II. Contributors to the Increase in the GDP Level

article image

Appendix III. GDP-P Compilation Methodology

article image
article image

Appendix IV. Revised Annual GDP-P Estimates

Table 1

Gross Domestic Product by Economic Activity at Current Prices

(ECS Millions)

article image
Table 1

Gross Domestic Product by Economic Activity at Current Prices

(ECS Millions)

article image
Table 2

Gross Domestic Product By Economic Activity at Constant 2018 Prices

(EC$ Millions)

article image
Table 2

Gross Domestic Product By Economic Activity at Constant 2018 Prices

(EC$ Millions)

article image
Table 5

Gross Domestic Product by Economic Activity Current Price Growth Rates

article image
Table 5

Gross Domestic Product by Economic Activity Current Price Growth Rates

article image
Table 6

Gross Domestic Product by Economic Activity Constant Price Growth Rates

article image
Table 6

Gross Domestic Product by Economic Activity Constant Price Growth Rates

article image
Table 7

Gross Domestic Product by Economic Activity Current Price Contribution to GDP

article image
Table 7

Gross Domestic Product by Economic Activity Current Price Contribution to GDP

article image
Table 8

Gross Domestic Product by Economic Activity Constant Price Contribution to GDP

article image
Table 8

Gross Domestic Product by Economic Activity Constant Price Contribution to GDP

article image
  • Collapse
  • Expand