This note describes the current methodology used for compiling OECD 'main country groupings' (zones) for the Main Economic Indicators (MEI) publication and datasets. The MEI publishes zones for the following main country groupings:
OECD-total: Consists of all OECD countries
OECD-Europe: The 25 European countries that are members of the OECD
EUU: The European Union of 27 countries
G7: The major seven economies of the OECD: United States, Canada, Japan, France, Italy, Germany and United Kingdom
Euro area: The 17countries of the euro area
All or a subset of these zones are available for the following subjects: Industrial Production, Composite Leading Indicators, Confidence Indicators, Passenger Cars, Retail Trade, Consumer Price Indices, Producer Price Indices, Hourly Earnings, Unit Labour Costs, Harmonised Unemployment Rates, Employment, Monetary Aggregates and International Trade. The OECD compiles the series for OECD-total, OECD-Europe and G7 whereas data for the European Union and Euro area are provided by the relevant European authorities.
Country coverage and length of time series
The MEI aims to provide as long a time series as possible for area totals, which requires long time series for component countries. However a special case exists for transition economies (Czech Republic, Estonia, Hungary, Poland, Slovak Republic and Slovenia), where a cutoff has been set at 1990 for these countries contributions to area totals. In fact, for many subjects, data may only be available from 1995 onwards for these countries. Therefore, in principle, area totals series will consist of all component countries back to 1990 (Industrial Production, Composite Leading Indicators, Monetary Aggregates and International Trade, Unit Labour Costs) 1995 (Retail Trade, Consumer Price Indices, Producer Price Indices, Hourly Earnings, Harmonised Unemployment Rates and Employment) or 2003 (Passenger cars). These series and are then linked to a reduced country version of the area total as far back as possible which excludes Poland, Hungary, Czech Republic and Slovak Republic.
In practice, the time series length and country composition of each area total will differ by subject depending on the length of time series available for countries and whether the relevant series (or a close proxy) exists for each country. However, estimations are performed by the secretariat in a small number of cases to ensure the above principles are adhered to as best as possible and to enable long time series to be produced for each subject. Also, for some subjects the series for European Union and Euro area as provided by European authorities are linked to series previously compiled by the OECD for the original European Union or the Euro area. These instances and points where the links occur are detailed in the online metadata.
Weights used for area totals
The weights used to combine country data in forming area totals differ depending on the subject. Industrial Production, Composite Leading Indicators, Retail Trade, Hourly Earnings, Unit Labour Costs, Consumer Price Indices, Producer Price Indices and Monetary Aggregates all use relevant aggregates in the most recently available year (e.g. GDP by industrial sector, Household Private Final Consumption Expenditure, Compensation of Employees, Total GDP) adjusted by Purchasing Power Parities. Other series using weights that are not adjusted by Purchasing Power Parities are Passenger Cars which uses the value in the base year. Other subjects area totals (Harmonised Unemployment Rates, Employment and International Trade) are simply based on the sum of contributing countries data.
Where weights are used, area total series are annually chain-linked Laspeyres indices. The series are chain-linked every five years when new weights are introduced, coinciding with the introduction of a new base year, which is currently 2005. The exceptions are area totals for Industrial Production, Retail Trade, Unit Labour Costs, Consumer Price Indices, Producer Price Indices, Hourly Earnings, and Monetary Aggregates which are annually re-weighted chained Laspeyres indices.