Transcript D6.1
INTERNATIONAL SEMINAR ON TIMELINESS, METHODOLOGY AND COMPARABILITY OF RAPID ESTIMATES OF ECONOMIC TRENDS Session 6 Summary REVISIONS IN QUARTERLY GDP OF OECD COUNTRIES: AN UPDATE AC188-S62a OECD Revisions Analysis • • • examines the revisions histories of eighteen OECD member countries for the first estimates of constant prices, seasonally adjusted quarter-on-quarter GDP growth rates methodology can be used to test hypothesis that an increase in timeliness for official statistics leads to an increase in revisions and a decrease in quality uses mean absolute revision and relative mean absolute revision to take into account the average size of growth in a given period for a given country as high growth numbers result in higher absolute revisions OECD Revisions Analysis • well established and documented methodology • uses extensive data holdings of OECD • tools available free from the OECD - links to programs included OECD Revisions Analysis Typical use is for ongoing series such as quarterly GDP, comparing first estimate compared to three years later but methodology could be useful to see if quality drops e.g. with the introduction of more timely estimates. Indicators for long-term developments Eurostat AC188-S63 Structural indicators • focus on growth and jobs • 79 indicators arranged within 6 themes • • very high visibility of these indicators (APR, most consulted tables on the Website) can compare countries and country to EU Sustainable development indicators Measuring progress is integral part of the EU strategy • use 130 indicators covering 10 themes and 11 headline indicators Main output: Indicators on the Website Biennial monitoring report of the EU SDS PEEIs EuroIND and the Euro-indicators/PEEIs tables are structured in eight main groups: • Balance of payments • Business and consumer surveys • Consumer prices • External trade • Industry, commerce and services • Labour market • Monetary and financial indicators • National accounts Aggregated and national data on web with access/ graphing etc tools. Dissemination Datasets appear to have developed separately, but in general: • readily available on web • linked metadata available • • • aggregated and national data available for a selection of periods standard outputs plus user selectable datasets sophisticated tables/ graphs/ cross classifications available World Bank Country At A Glance Tables on Recent Economic and Financial Indicators AC188-S64 • • • monitoring the rapidly evolving global financial crisis requires current or near current, high frequency data lots of such data available, but in different places and only for high-income and large developing economies to meet emerging needs, the World Bank has developed Country At A Glance - Recent Economic and Financial Indicators (AAG-REFI) AAG-REFI includes indicators covering the external sector, financial sector, fiscal sector, and real sector that help to measure the vulnerability of countries to shocks arising from a sudden drop in private capital flows, volatility in primary commodity prices, and the emergence of stress in financial markets AAG-REFI • available for twenty two developing countries • one or two month lag in most data • readily accessible through standard presentations World Bank High Frequency Data Sources GEM website – macroeconomic, financial and commodity data for high income and emerging countries JEDH / QEDS – external debt data Lots more – see paper Some messages: Dissemination and communication Most recommendations apply whether we are in a crisis or not i.e. we should probably be doing all this anyway, but the crisis highlights the issues. data hard to access despite best efforts of international organisations priority is to make visible existing data lots of data out there – if more collected/ disseminated, there is a risk of plethora of datasets is there too much data? need to guide users to the most relevant data in the sea of data however, what is relevant may change quickly make metadata visible, but don't swamp the data if it needs lots of caveats, is it fit to publish? coordination is important if data is to be approved/ disseminated quickly and more than one agency is involved talk to users - then talk to them again International organisations duplicate data collection and holding - data often old however, there are some gains in this area e.g. http//:principalglobalindicators.org www.jedh.org International Organisations make an enormous contribution in setting standards e.g. SNA • • • • How to present data: user friendly - electronic straight to spreadsheet (not necessarily fancy and complex) focus on month on month/ quarter on quarter rather than annual changes seasonally adjusted and trend – upfront and visible (but down to what level of detail?) focus on price/ volume measures prices contain best market signals packaged presentations like StatCan story are useful. publish methodology/ metadata educate users Who disseminates? only talking about data which is disseminated (what is status of data “for internal use only”?) NSO - collected data (including leading indicators) who should do forecasts? there is a role for commercial data providers a rigid framework/ list of indicators may be of limited use, because we don't know what is coming next, so needs may change quickly Bits: GDP(A) There are well established methodologies to test quality of rapid estimates e.g. OECD tools Centres of excellence? e.g. OECD for revisions analysis Don Drummond: "I would rather have no data than wrong data."