Vanuatu
Technical Assistance Report-Feasibility Study on the Introduction of Quarterly National Accounts
Author:
International Monetary Fund. Statistics Dept.
Search for other papers by International Monetary Fund. Statistics Dept. in
Current site
Google Scholar
PubMedClose
Publication Date:
09 Feb 2021
eISBN:
9781513568836
ISBN:
9781513568836
Language:
English
Keywords:
QNA data; Na data; seasonal adjustment software; higher-frequency data; software technology
Download PDF (1.7 MB)

The Vanuatu National Statistics Office (VNSO) requested a mission by the IMF’s Pacific Financial Technical Assistance Centre (PFTAC) for conducting a feasibility study on the implementation of a quarterly national accounts (QNA) in Vanuatu. This study outlines the staff, organizational and data requirements for Vanuatu to implement a quarterly national accounts program.

Abstract

The Vanuatu National Statistics Office (VNSO) requested a mission by the IMF’s Pacific Financial Technical Assistance Centre (PFTAC) for conducting a feasibility study on the implementation of a quarterly national accounts (QNA) in Vanuatu. This study outlines the staff, organizational and data requirements for Vanuatu to implement a quarterly national accounts program.

Executive Summary

The Vanuatu National Statistics Office (VNSO) requested a mission by the IMF’s Pacific Financial Technical Assistance Centre (PFTAC) for conducting a feasibility study on the implementation of a quarterly national accounts (QNA) in Vanuatu. This study outlines the staff, organizational and data requirements for Vanuatu to implement a quarterly national accounts program.

Vanuatu’s economic policy makers, the Ministry of Finance and Economic Management (MFEM) and the Reserve Bank (RBV), expressed demand for a timelier, high frequency publication of national accounts figures. Timelier information about most recent economic developments is essential for effective, evidence-based economic policymaking. Such data also provide the private sector, foreign investors, rating agencies, and the public in general with important inputs in their decision-making, while informing economic analysis and International Monetary Fund (IMF) surveillance. Typically, NA data are published with a considerable lag (called recognition lag) having a potential negative impact on the effectiveness of a countries’ economic policy decisions.

The introduction of a QNA program in Vanuatu targets a publication date of four months after the end of a quarter for Gross Domestic Product (GDP), by the production and expenditure approaches. While this report shows both approaches are feasible, and could be produced on a seasonally adjusted basis, it could be more practical in a case of limited staff to develop one approach initially and extend to the other approach later.

The main findings of this study include:

Human resources

Vanuatu will require one full time equivalent employee trained in annual well as quarterly national accounts compilation to produce quarterly expenditure based and production-based GDP. It is recommended that Vanuatu establish a quarterly national accounts teams so as to not concentrate the knowledge, skills and responsibility in one person. If Vanuatu would like to move quickly with the implantation of this program it is advisable that Vanuatu re-assign individuals currently working on the Annual National Accounts Program to this work, given they are already trained in national accounts.

Organization

A reorganization of tasks is proposed in Figure 2 in the study.

Information technology infrastructure

One Excel macro and two software packages are recommended for the compilation of QNA figures which are available as freeware and can be installed on normal PCs or laptops.

Data requirements

QNA require economic data on a quarterly frequency shortly after the end of the reference period. Several of the indicators that the VNSO use in their ANA program are available at a quarterly frequency. However, certain data are only available with a significant lag (up to eight months following the reference period). Adding one month for compilation means that QNA could be published only nine months following the reference period. For reaching the target of four months, it is important that VNSO work with the data providers to improve the timeliness in the receipt of data. Specifically, VNSO should target improved timeliness for the following data:

  • BoP data

  • VAT data

  • Banking statistics

  • National Provident Fund data on the number of employees and the wage bill by International Standard Industrial Classification, fourth revision (ISIC Rev. 4) industries.

As a result of the recent and ongoing impact of the novel coronavirus (COVID-19) and Cyclone Harold on statistical compilation activities, the original actions and/or target dates may need to be revised. The IMF and PFTAC fully recognize the uncertainties and difficulties faced at this time and stand ready to provide remote advice and assistance wherever possible to support statistical continuity.

To support the VNSO in meeting its current requirements, the mission recommends a detail action plan with the following near-term priority recommendations:

Table 1.

Priority Recommendations to the Vanuatu National Statistics Office

article image

Feasibilty Study on the Introduction of Quarterly National Accounts in Vanuatu

A. Introduction

1. Following demands from users for more timely, higher-frequency data, the TDoS requested the IMF’s PFTAC to conduct a feasibility study on the implementation of QNA program. This study outlines the staff, organizational and data requirements for the TDoS to implement a quarterly national accounts program. The IMF’s Quarterly National Accounts Manual provides extensive guidance on setting up a QNA system.1

2. In order to identify users’ needs for timeliness and the grade of detail of published QNA data, PFTAC met with economists from the MFEM and the RBV. Both institutions expressed their interest in a target-date for publication of four-month following the end of a quarter. They delegated the evaluation of findings and suggestions made in this study to Vanuatu’s Macroeconomic Committee. After this discussion based on the draft version of the study, feedback will be given to PFTAC.

B. Human Resources Requirements

3. QNA require staff trained in statistical econometric methods, apart from knowledge about the general system of annual and quarterly NA. Part of the methods used for developing QNA that are consistent with annual data – as implemented in specialized software packages – are based on time series regression techniques. The additional provision of seasonal and working day adjusted data, increases the usefulness of QNA data for economic analysis substantially. The methods for adjustment are based on time series statistics as well, which makes a further training necessary. This training can be provided by PFTAC and are included in its QNA training courses.

4. According to the experience of other countries, the human resources permanently necessary for compiling QNA for the production and expenditure side of GDP require up to one person per year trained in the skills mentioned above. It is recommended that VNSO establish a quarterly national accounts teams so as to not concentrate the knowledge, skills and responsibility in one person. If VNSO would like to move quickly with the implantation of this program it is advisable that VNSO re-assign individuals currently working on the Annual National Accounts Program to this work given they are already trained in national accounts.

C. Organization

5. Currently, the ANA work is split between the Principal Statistician for Economics and the two related Senior Statisticians. At present, the administrative structure of the VNSO is organized hierarchically as given in Figure 1. According to Vanuatu’s long-term strategy for developing statistics, staff numbers have been increased from 31 to 52 persons, in 2018, not including extra staff for the introduction of a QNA program.

Figure 1.
Figure 1.

Current Organizational Structure of the VNSO

Citation: IMF Staff Country Reports 2021, 037; 10.5089/9781513568836.002.A001

6. As stated in section 2, a further position is necessary for compiling QNA. In order to ensure the continuity of the QNA program it is recommended that the same team compile both the QNA and ANA. This will improve the economic consistency between ANA and QNA and will facilitate the implementation of the QNA program. The proposed new structure alongside this relocation of tasks is given in Figure 2.

Figure 2.
Figure 2.

Proposed New Organizational Structure of the Economics Section

Citation: IMF Staff Country Reports 2021, 037; 10.5089/9781513568836.002.A001

7. Apart from sharing the workload of QNA compilation, this further position is recommended in order to reap the benefits of specialized knowledge. The specialization of the staff is reorganized according to industries and/or institutions as given in Figure 2. All statisticians should compile the number of employees and the wage bill of the industries assigned to them. This way, the economic consistency of results can be improved.

D. Information Technology Infrastructure

8. Where the quarterly sources or methods differ from the annual ones, the data should be checked to make sure they show compatible annual trends. Very different trends suggest that the indicator is unsuitable and should be replaced. Small differences can be eliminated by mathematical techniques. PFTAC is willing to support VNSO in the implementation, by providing software, documentation, operational support etc.

9. The most straightforward solution is an Excel macro available from the IMF Statistics Department (XLPBM). It does not perform other functions such as seasonal adjustment, but it can be used within an existing spreadsheet system without additional software.

10. EcoTrim was developed by the statistical office of the European Union EUROSTAT. Although this software does not make use of the latest software technologies and requires some data conversion for input and output of data, it offers alternative approaches and works reliably. As it is not downloadable from the internet anymore, PFTAC can provide the files for installation on a desktop or laptop computer.

11. The third solution is a separately downloadable plug-in for temporal disaggregation and benchmarking of the seasonal adjustment software JDemetra+. It has been developed in cooperation between EUROSTAT and the Bank of Belgium and includes state-of-the-art methods for seasonal adjustment which are recommended internationally. Documentation and online course material are available in the internet and a discussion forum as well as regular updates guarantee its functionality. This software package can read various formats of data, including Excel files and has a strong interactive graphical interface. However, the plug-in for temporal disaggregation still has some bugs, so a combination of both software package is recommended. Both software packages are available as freeware and can be installed on normal PCs or laptops. The training necessary to use these software packages is included in the QNA training course which takes one week.

E. Data Requirements

12. The most crucial element for NA compilation is the availability of high quality, high frequency and timely source data. In general, the primary source of information for a quarterly GDP program are administrative records of the government. When administrative records (such as taxation records or employment records) are not available the statistical offices may have to conduct a business survey. While annual surveys increase substantially the quality of ANA results as they are tailormade for NA purposes, they are often too detailed use on a quarterly basis. In this case the content of the annual questionnaire should be adjusted to only collect the information needed for the quarterly national accounts program.

13. The compilation of QNA estimates requires less information than an ANA program but the information needs to be available on a quarterly basis shortly after the end of the reference period. Currently, some important source data VNSO requires for a QNA program are only available with a delay of six months after the end of a period. VNSO needs to work with authorities to improve the timeliness in the receipt of this source data.

14. While a significant amount of high frequency quarterly data exists, there remain a number of data gaps. Fortunately, in many cases these data gaps are not significant and pertain to those industries / components of GDP that do not contribute substantially to total GDP. In cases where the contribution to total GDP is not significant the quarterly movements can be estimated using methods which distribute annual figures over quarters without any indicator. The recommended approach minimizes the quarter growth rates under the restriction that resulting quarterly estimates sum up to their ANA totals. This minimum growth rate criteria guarantees that such components – when added up with other components which are based on economic indicators – have minimal impact on the aggregate.

15. However, this method does not provide any inherent extrapolation for periods beyond the last annual benchmark. Hence, a forecast without an indicator must be compiled which decreases the informational content and quality of the output for the most recent quarter(s). This approach should only be applied for small components. In order to care for this fact in the underlying study, weights of industries and demand categories according to their share in GDP are taken into consideration in the analysis of data requirements.

16. The results of this indicator analysis on an annual as well as on a quarterly basis are given in the tables in the annex. For the production side of ANA, the indicators used for deriving output in nominal or volume terms as well as the price data source for deriving the missing other one, are given. The section of the tables treating the quarterly data analysis shows the indicators proposed for compiling QNA. These can be either the same data sources as used for ANA or some other available indicator on a quarterly frequency which possesses a theoretical connection with the aggregate. Whether each indicator is suitable for compiling QNA has to be decided on how well it tracks the ANA results. Quarterly indicators can be assessed by simple ratios or charts, and measures are also available in the recommended software. If such a test fails, this indicator should be discarded, and some alternative has to be found. Where a quarterly indicator with a theoretical relation to the ANA series could not be found, a question mark or a possible alternative is given it the tables. If this concerns an aggregate contributing more than one percent to GDP, this is regarded as significant and highlighted in the tables in red color.

17. While a significant amount of data exists, the VNSO needs to work with authorities to receive the data in a timely manner. Priority should be given to the following data sources:

  • The RBV stopped providing BoP data on a quarterly as well as on an annual basis in 2018. The last data published data cover the third quarter of 2018. This does not only impede the compilation of ANA but also a future introduction of a QNA program. VNSO needs to work with RBV to access the data on a timelier basis.

  • Currently, there are no published banking statistics data available. The last published data cover the third quarter of 2018. This does not only impede the compilation of ANA but also a future introduction of a QNA program. VNSO needs to work with the banking sector to acquire these data on a quarterly basis at an acceptable timeliness.

  • Due to technical problems the National Provident Fund, Vanuatu stopped data production in 2016. Currently, a new system is implemented to publish the number of employees and the wage bill by ISIC Rev. 4 industries. These data should be on a quarterly basis, accessible three months after the end of quarter.

  • Electricity and Water Supply: NA use business reports of the single provider which are only available annually. A quarterly available nominal indicator can be derived from the company’s VAT data records, but it is recommended to get real output indicators like GWh or millions of liters, additionally.

  • Wholesale trade of fuel: NA currently use business reports of companies which are only available annually. A quarterly available nominal indicator can be derived from the company’s VAT data records, but it is recommended to get real output indicators like sold liters, additionally.

  • Telecommunication: NA use business reports of the companies which are only available for annuals. Quarterly available nominal indicators can be derived from the companies’ VAT data records, but it is recommended to get real output indicators like telephone call minutes, number of SMS and Gigabyte traffic, additionally.

  • Taxes less subsidies on production: They form part of GDP and are derived from government budgets which are only available on an annual basis. As they account for more than 10% of Vanuatu’s GDP, quarterly data are indispensable.

18. For GDP(E) components, no severe gaps of quarterly data were identified. Government consumption as well as the final consumption of Non-Profit Institutions Serving Households (NPISH) are derived from output measures of public administration, education and health service and the output of NPISH. While this is not exactly the approach recommended by the 2008 SNA, this method can be applied the same way on a quarterly basis. The two important components of Gross Fixed Capital Formation (construction investment and equipment) are based on construction material imports and equipment imports from the BoP. These BoP data are currently not available which poses a problem for ANA and QNA as well as. Foremost, this problem impedes the calculation of the external sector accounts as NA exports and imports share the same definitions as BoP.

Annex I. List of Components and Indicators

Table 2.

Production Side of GDP

article image
article image
note: frequency 1=annaul, 3=quarterly, 12=monthly 100.0%
Table 3.

Expenditure Side of GDP

article image
note: frequency 1=annual, 3=quarterly, 12=monthly

Same Series

Other IMF Content

Other Publishers

  • Collapse
  • Expand
Cover IMF Staff Country Reports
Vanuatu: Technical Assistance Report-Feasibility Study on the Introduction of Quarterly National Accounts
Author:
International Monetary Fund. Statistics Dept.
Volume/Issue:
Volume 2021: Issue 037
Publisher:
International Monetary Fund
ISBN:
9781513568836
ISSN:
1934-7685
Pages:
16
DOI:
https://doi.org/10.5089/9781513568836.002
Figure 1.

Current Organizational Structure of the VNSO
Figure 2.

Proposed New Organizational Structure of the Economics Section