United Nations Economic Commission for Europe Statistical Division Production and dissemination of short-term economic statistics: the need for long timeseries, fixed base indices, etc Artur Andrysiak Economic Statistics Section, UNECE
UNECE Statistical Division Slide 2October 2008 Overview STES introduction Key indicators Users and uses Purpose of STES Characteristics Importance of fixed base indices, long timeseries and metadata
UNECE Statistical Division Slide 3October 2008 Short-term economic statistics Short-term economic statistics (STES) are important for economic analysis. They help users understand which phase of economic cycle they are in at present and what will be the probable development in the near future.
UNECE Statistical Division Slide 4October 2008 STES vs Structural statistics STESStructural statistics Usually produced using smaller sample Monthly or quarterly indicators Usually take form of indices but also include absolute values Produced to strict timetable Required as soon as possible Often preliminary estimate are subject to revisions Usually produced using larger sample Refer to whole reference year Show change from year to next year Can be used to judge accuracy of sub-annual data Can provide benchmarks for analyzing infrequent or one-off surveys
UNECE Statistical Division Slide 5October 2008 Some of the key STES indicators Consumer Prices Index, Producer Prices Index, Industrial Production Index Retail Turnover Wages and Salaries No of persons employed, employment rate, unemployment rate Building and construction Services producing sector
UNECE Statistical Division Slide 6October 2008 Some of the key STES indicators STS in the EU STS-Regulation includes Production Turnover (domestic and non-domestic) New orders Number of persons employed Hours worked Gross wages and salaries Producer price indices, goods Export and import price indices Construction costs Labour costs Producer price indices, services
UNECE Statistical Division Slide 7October 2008 Users of STES Internal users External users Government Business community Academics Other users: international users, the media, the general public
UNECE Statistical Division Slide 8October 2008 The purpose of STS For compilation of various components of national accounts and other statistics Early indicators, for validation and checking Decision-making Monetary policy and other policy deliberations Economic analysis Forecasting
UNECE Statistical Division Slide 9October 2008 Key characteristics of good STS Relevant – there should be a demand for the statistics Reliable and of sufficient statistical quality Timely - Produced to strict timetable and required as soon as possible by users Compiled and published in accord with international recommendations Publication of appropriate documentation (metadata) STS is often subject to revisions – a revision policy should be established
UNECE Statistical Division Slide 10October 2008 Overview of short-term economic statistics available from the websites of the SEECIS NSOs
UNECE Statistical Division Slide 11October 2008 Dissemination – good practices Countries should be publishing fixed base indices or/and absolute values for discrete periods It is not an acceptable practice to only publish the movements between the periods Cumulative, year-to-date, and same period previous year statistics should only be published as additional information Where possible countries should publish seasonally adjusted series Appropriate documentation (metadata)
UNECE Statistical Division Slide 12October 2008 Number of countries publishing fixed base indices/absolute values and year-to-date indices/cumulative values
UNECE Statistical Division Slide 13October 2008 Year-to-date statistics Publishing year-to-date indices or movements comparing year-to-date results is a legitimate way of presenting statistics. It allows for analysis of the corresponding periods. It permits users to draw some conclusions (without having to seasonally adjust the estimates) about performance this year as compared with same period in previous year/s. but
UNECE Statistical Division Slide 14October 2008 Year-to-date/cumulative statistics - important points to noted Year-to-date indices or movements comparing year-to-date result with same period previous year can not be used as substitutes for fixed base indices or values for individual months/quarters. They should only be provided as additional information complementing fixed base indices and values for individual months/quarters statistics. Comparing year-to-date results with same period previous year allow users to draw some valid conclusions without seasonally adjusting the statistics. However it should be recognised that in terms of seasonality there are still significant differences between the two periods. Year-to-date indices or movements comparing year-to-date result with same period previous year should not be used to derive the fixed base indices and values for individual months/quarters statistics unless complete and up-to-date timeseries are available. Year-to-date statistics are not suitable for international comparison and analysis
UNECE Statistical Division Slide 15October 2008
UNECE Statistical Division Slide 16October 2008 Importance of reliable long time series To allow users to perform their own analysis To allow international comparisons A minimum of 3-5 years of data required for seasonal analysis For analysis and appraisal of business cycles in both a historical perspective and for forecasting purposes Ideally complete timeseries should be available to users Revisions should be reflected in the timeseries
UNECE Statistical Division Slide 17October 2008 Documentation - metadata Accurate metadata is essential for any international comparison and analysis of statistics. Metadata should assist the user in: searching for statistical information interpreting its content after obtaining statistics, it should help in the post-processing and drawing accurate conclusions As much as is needed to ensure correct interpretation and use Prevent misunderstanding or misuse Use clear and consistent language in the explanations
UNECE Statistical Division Slide 18October 2008 Number of countries publishing metadata and timeseries of more then 6 observations
UNECE Statistical Division Slide 19October 2008
UNECE Statistical Division Slide 20October 2008 Examples of good dissemination practices rylist/ rylist/ (Estonia – overall website, timeseries, fixed base indices, release calendars) (Ukraine – long time series, advance calendars and metadata) simination/national_summary/ (long time series) simination/national_summary/
UNECE Statistical Division Slide 21October 2008 Questions? THANK YOU