The Inventory Questionnaire Summary Overview Task Force meeting, Lisboa, 9-10 December Wojtek Szewczyk and Shalini Mittal 9 December 2013
Outline of the presentation Introduction: TF vs. EU (sample vs population) Current state of play Inventory of methods Some overview results Inventory of sources Availability of / missing data for the first quarter Way forward
introduction
TF as representative for the EU28 19 : 10 in – TF Bulgaria Czech Republic Finland France Germany Hungary Italy Latvia Lithuania Luxembourg Netherlands Poland Portugal Romania Slovakia Slovenia Sweden Switzerland United Kingdom TF - out Austria Belgium Croatia Cyprus Denmark Estonia Greece Ireland Malta Spain
TF geographical coverage
TF and EU28 GDP (80%)
Inventory of methods
Current State of Play Approach Issues Econometric Methodology
Current - Approach Both Expenditure Approach Production Approach Portugal France Production Approach Hungary Lithuania Poland Romania UK Czech Republic Bulgaria Finland Both Germany Latvia Slovakia Sweden Italy, Netherlands did not specify their approach Luxembourg, Slovenia, switzerland do not publish or calculate Slovenia more interested in quality than in rapidity
Current - Issues Missing Data for Third Month Cost Ineffective Organisational Issues Availability of Short-Term Indicators? No estimates at all Information provided by 19 TF members. France: cost ineffectiveness Netherlands stopped the test runs because of pressure of work due to the SNA 2010 Luxembourg produces data with a lag of 90 days Hungary started producing in 2012 but had to suspend due to High Work Load and Lack of Financing
Current-Econometric Methodology Estimate Third Month of the Quarter ARIMA ARIMAX TRAMO-SEATS Univariate Methods VAR Dynamic Factor Models Multivariate Methods ECM Temporal Disaggregation Guestimation Others Germany-ARIMA
Advancement of T+30 across the TF Experimenting Experienced Finland Czech Rep. Poland Romania Slovakia Hungary Netherlands Portugal France Italy Latvia Bulgaria Germany Lithuania UK Sweden Beginners Experienced: who already had t+45 systems or some system in place and have also moved onto t+30 Experimenting: Who have some system in place already and have run a few trials or are looking at the potential to do so. Scratch: no system in place Luxembourg Slovenia Switzerland
Inventory of sources
Sources Collected questionnaires for 16 countries 169 unique indicators Across 23 different combinations of NACE aggregations And for Taxes and Subventions
Indicators across sectors
Mapping sectors and indicators Use broad, aggregated NACE If an indicator for aggregation of sectors (eg BCDE) is not available, an indicator for a constituent sub-aggregation can be used (eg C) Indicators First appearance of an indicator for a sector marks availability timeline will produce most optimistic picture, but will be refined in practice
Sectors mapping NACE aggregations NACE Final Sectors A B-E BCDE C F B GHI D J B-D K F (primary construction industry) L MN G-I OPQ G 46 RSTU G 47 TAX H 49 SUB L (letting of dwellings) L (real estate services without letting of dwellings) M_N M-N O-Q R-T R-U Subv. Taxes
Used NACE aggregation EU28 GDP shares
Share of countries GDP ‘indicated’ by end-April
Indicators over time Stock of data available at t+25 % of EU28 GDP “indicated” 40% 37% 16% January February March April t+25 time
% of EU28/TF GDP covered by the indicators at t+25 - example for Q1 % of EU28 covered % of TF covered
Indicators for March become available in April…day-by-day
Indicators for March as become available in April (1) % of sectoral GDP
Indicators for March as become available in April (2) % of sectoral GDP
Indicators for March as become available in April (3) % of sectoral GDP
Way forward
Way forward Decision on the approach Verification of the relevant inventories Verification of the past data (pre 2014:1) Transmission of the monthly data (from 2014) Production of the t+30 GDP test estimate
Questions, comments, discussion… Thank you