90978 ICP Regional Coordinator Workshop February 3-10, 2006 Overview: The Workshop consisted of three separate meetings. One group met on February 3, 4, and 6 and reviewed matters related to the expenditure weights as estimated in the national accounts. It ensured that there was a procedure in place to detect anomalies. The results of the Group’s discussions are shown in “Expert Meeting on ICP Expenditure Data” which will be posted on the website along with other materials prepared for the meeting. The proceedings of the meeting and the accompanying materials will constitute the basis of guidelines designed to critically review expenditure weights and per capita expenditures. The list of participants appears at the end of the summary. The Data Access Working Group created by the ICP Executive Board met on Feb 6, 2006 and considered matters related to confidentiality of data and access by researchers. Their report will be presented to the ICP Executive Board on March 6, 2006. Regional coordinators have received copies of the draft and are urged to comment on the adequacy and appropriateness of the report. The Regional Coordinator Workshop (February 7-10) reviewed mostly parities on the basis of Quaranta and Dikhanov tables for the ICP regions other than the CIS. The regional coordinators and Global Office staff also reviewed technical issues and the timetable for pricing health, housing, equipment, construction, and government. The work plan for the Ring Comparison was reviewed and coordinators requested to consider it further and confirm it or suggest amendments prior to March 2, 2006. There is a planned conference call for that day involving the regional coordinators and Global Office staff to finish the work plan. Summary of Actions Taken: Need for Meta Data to describe National Accounts. National accounting data are expected to be comparable across countries but at times, and particularly with components that are difficult to estimate, the results may not be consistent from one country to the next. For purposes of the ICP, official estimates of GDP and its published components will not be altered. But where there are known cases of inconsistency, suitable metadata will be collected and made accessible to describe the limitations of comparability. In order to collect the necessary information, the Global Office will start out by using what is recorded on the IMF GDDS. National statistical offices will be contacted and requested to either update the GDDS information, or else if they have not yet supplied it to do so. Final Decisions on inclusion of countries in Global Results: The matter was discussed of how to dispose of country data not deemed to be of the required standard through lack of underlying detail, or because of its anomalous behaviour. At the extreme, the question was asked in terms of all data submitted by a country leading to that country’s potential exclusion from the ICP. The arguments heard suggested that: • In some regions, there will be internal pressure to include as many countries as possible and to deal with data problems by appropriate footnotes; and • In other regions, the inclusion of a country in the regional comparison was seen as an endorsement by the Regional Coordinator that should not be overruled by the Global Office. However, no one was prepared to take away the Global Office’s responsibility to ensure that results across the regions be comparable and mutually satisfactory as to their completeness and professional quality when preparing the global results. Accordingly, it was agreed that the Global Office will make an assessment of data quality and follow the fundamental data quality principles including coverage and scope in addition to reviewing extremes singled out in the Quaranta and Dikhanov diagnostic tables. This review, conducted in conjunction with the regional coordinators, will assess whether a country meets the minimum standards for inclusion. The principle of Coverage in the ICP case means that the prices and GDP represent the country as a whole while scope means that the full GDP was taken into consideration. The regional coordinators and the Global Office will agree on the countries that should be included in the regional report. If a country does not meet the requirements for inclusion in the Global report, the region will need to take this into consideration when preparing its PPPs to ensure that fixity is ensured for the remaining countries when included in the global report. The Global Office will reserve the right to make the final decision regarding inclusion in the global publication. Regional Data Validation: The data from LAC, Africa, Asia, and W. Asia were reviewed using the Quaranta and the Dikhanov tables. The CIS presented Quaranta tables but they were incomplete because the number of observations and corresponding CVs were suppressed. The variability of item PPPs within basic headings was large in most regions. A more detailed inspection showed that some countries submitted average prices based on only 1- 2 observations. It seemed that there were errors caused by the wrong unit of measurement, by the wrong size or simply by data entry errors. The removal of items based on too few observations would improve the consistency of PPPs. There were also cases where the removal of a country would significantly decrease the variability in PPPs in the remaining countries. The following guideline was recommended: • In cases where there are less than 3 observations and a review of the editing does not result in valid additions to that number, delete the item. Equipment and Construction: Data Collection will take place around mid-2006. Regional coordinators said they would not be able to collect prices retrospectively back to mid-2005 as the TAG recommended. Furthermore, they also said that indices required to “move” the 2006 prices to a 2005 base were of insufficient quality. Accordingly, it was agreed the prices would be used “as is” except in cases where inflation exceeds 5 percent in which case some adjustments are in order. There are 34 construction components. Twenty-two are composite components containing costs for materials, labor, and equipment. In addition, there are 12 basic components where unit costs are to be obtained for items such as sand, skilled labor, backhoe rental, etc. The PPPs will be computed based on the 22 composite components; however, the prices are needed for the basic components as a quality check on the overall process and to evaluate the BOCC process. The regional collections for equipment and construction will also be used for the Ring comparison. There was a discussion about how the data validation will be done at the global and regional levels. If necessary, experts will be called on to assist with the data review and validation. Health: The full cost concept will be used with the understanding that health PPPs are based on the combination of expenditures financed by households and by the public sector. The Health guidelines have been updated and are attached. They can also be found on the ICP website. Housing: Within regions, the nature of the comparison will vary depending on the nature of the rental markets and the availability of breakdowns of the national accounts expenditures on dwellings. For all regions except the CIS, there is at least one Ring country whose price comparison will be based on a rental survey. The CIS would be linked to the OECD using Slovenia and Estonia. All Ring Countries will need to fill out the questionnaire on volumes and amenities of housing with estimates of expenditures by renters and owners and type of dwelling (modern vs traditional, urban vs rural) which will soon be sent to the regions. Ring work plan and timetable: Data collection for consumption items will take place in the Ring Country capital or in one of its major cities, and will be one-off in the 2nd quarter. Collection for the other items will be the same as what is being done for the regional comparisons. All communications with the Ring Countries will be conducted through the regional coordinators. Ring Countries are expected to review the data they collected for Ring purposes following the same methods they applied to their regional data collection. The individual observations should be forwarded to the regional coordinator. In turn, the regional coordinator will review the data received and ensure that they are broadly consistent with what was submitted for the regional comparison. The regional coordinator will send on the individual observations to the Global Office where the data will be compared globally. The basic heading ring data will be jointly reviewed by the regional coordinators and the GO by mid-July 2006, implying that the Ring price data will be made available to the GO by end-June. Publication: Preliminary results will be published first for consumption followed by data for the full GDP. The following points will be taken into consideration: • Consumption is defined in ICP Handbook chapter 3 which states that Actual Final Household Consumption includes individual consumption expenditure by households plus individual consumption expenditures by NPISH and Government. • Expenditure shares by basic heading will be based on the most recent available data, and be calibrated to the best available estimate of 2005 GDP. Publication: Tentative Calendar Preliminary Results Final Results Region Consumption Full GDP Full GDP Africa Oct 2006 Dec 2006 Qtr II, 2007 Asia Same Same Same CIS Same Same Same W. Asia Same Same Same EU/OECD No prelim Same Same LAC June 27, 2006 Same Same Global Nov 2006 Qtr I 2007 Same Summary: As a result of the issues covered during the workshop, several documents are either being prepared or have been finalized and will be available on the ICP website. 1. Review Guidelines for Expenditure Weights 2. Health survey guidelines 3. Equipment and Construction survey guidelines 4. Government data collection form 5. Housing form for all Ring countries to complete We wish to express appreciation to the regions who have worked hard to comply with the agreed-upon supply of average prices and diagnostics to the Global Office, thus ensuring that appropriate reviews of data quality take place not only among countries but among regions. The principle of transparency is as important among countries as it is among regions. The CIS region was informed that at least one other region will not allow its data to be used in the global comparison along with the CIS data until such time as the CIS provides the necessary information to subject its data to the same inter-regional review conducted for the other four regions. The Global Office is working with the CIS region to resolve this situation. Statement from LAC Over the last two years, the members of the committee of regional coordinators of the ICP have developed an understanding of how to coordinate their work in order to ensure adherence to and consistency with three basic principles: • Securing harmony in their professional relations through measures designed to promote mutual trust; • Making sure that data problems are detected and dealt with so as to avoid threats to the program’s credibility; and • Striving for transparency by frequent exchanges of data, descriptions of methods and metadata so as to avoid adverse public perceptions about the program’s design and execution At the seventh meeting of the RC’s held in Washington from the 5th to the 10th of February, it became apparent that one of the regions taking part in the program has not abided by the understanding reached, has not done enough to promote the fundamental principles of the program and has ignored repeated entreaties to exchange information with no reservations. We consider that this behavior is detrimental to the objectives of the program and accepting it would constitute a bad precedent for future rounds of the ICP. We have come to the conclusion that unless there is a change of mind we are not in a position to accept as valid the data submitted to the Global Office and offered for critical review by the other RC’s. We are saddened to have reached this conclusion but satisfied that we did so after issuing a sufficient number of cautionary notes designed to avoid a crisis. Without in any way involving other RC’s, nor intending to interfere with the GO’s freedom of action, we are serving notice that we intend to take the following steps: • Requesting from the Bureau of the Statistical Commission of the Americas, in session on 6 March, permission to withdraw Latin America from any global calculation involving a region incapable or unwilling to abide by the understanding that the RCs reached with the GO; and • Informing the United Nations Statistical Commission or else requesting that the GO do so on our behalf that and why we have taken this decision. As a reminder to all participants in the ICP, the program’s minimum requirements stipulate that each regional coordinator shall make available to the Global Office average prices accompanied by a measure of the number of individual prices on which the average is based as well a measure of the dispersion of the prices in question. It is the repeated failure to provide the committee of regional coordinators with these details that has triggered our decision. List of Participants ICP Meetings, 10-17 February 2006, Washington DC Data Access Group Louis Marc Ducharme Chairman of the Data Access Group - Canada Eduardo Pereira Nunes Brazil Addulrhaman Al-Mansouri Kuwait David Fenwick United Kingdom Vasily Kuznetsov Russia Mohamed Abulata Egypt Vishnu Kumar India National Accounts meeting Jacob Ryten Member of the ICP Executive Board Alan Heston Chairman of the ICP Technical Advisory Group Vaskar Saha ICP Regional Coordinator – Asia and the Pacific Heber Camelo ICP Regional Coordinator – Latin America Luc Mbong Mbong ICP regional team – Africa Abdoulaye Adam ICP regional Coordinator – Africa Tarik Alami ICP Regional Coordinator – Western Asia Farah Nsouly ICP regional team – Western Asia David Baran United Kingdom National Statistics Bureau (ONS) Kim Ziechang IMF Statistics Department Derek Blades Consultant Sultan Ahmad Consultant Regional Coordinators Meeting Abdoulaye Adam ICP regional team – Africa Vaskar Saha ICP Regional Coordinator – Asia and the Pacific Youri Ivanov ICP Regional Coordinator – CIS Vasily Kuznetsov ICP Regional Coordinator – CIS Louis Marc Ducharme ICP Regional Coordinator – Latin America Heber Camelo ICP Regional Coordinator – Latin America David Fenwick United Kingdom ONS David Baran United Kingdom ONS Tarik Alami ICP Regional Coordinator – Western Asia Farah Nsouly ICP regional team – Western Asia Jacob Ryten Member of the ICP Executive Board Alan Heston Chairman of the ICP Technical Advisory Group Kim Ziechang IMF Statistics Department Derek Blades Consultant Sultan Ahmad Consultant