WP 1 Interindustry Accounts 2 nd EUKLEMS Consortium

WP 1 Inter-industry Accounts 2 nd EUKLEMS Consortium Meeting, 9 -11 June 2005, Helsinki This project is funded by the European Commission, Research Directorate General as part of the 6 th Framework Programme, Priority 8, "Policy Support and Anticipating Scientific and Technological Needs".

MFP growth • • Need value shares and volume growth rates of gross output, intermediate inputs (domestic and import), labour and capital services at industry level Collect data in consistent framework: Supply and Use tables (SUT). Why? • • Consistency between input and output, and various prices Data useful for wider applications (tax reforms, role of trade and transport, macro-economic modelling, IO-analysis etc. )

EUK industries • • Why SUT and not Input-output tables? • • • IO are derived from SUT available from 1995 onwards Easier to deflate and balance EUK industry list • • Starting point A 60 of Eurostat Additional breakdown for • High tech sectors • Future revisions • Imputed rent

Required Supply and Use tables 1. 2. 3. 4. USE table for domestic output at basic prices USE table for imports at basic (c. i. f. ) prices SUPPLY table at basic prices VALUATION matrices a. b. c. d. Use-side trade margin table Use-side transport margin table Use-side non-deductible VAT on commodities Use-side other taxes net of subsidies on domestic commodities e. Use-side other taxes net of subsidies on imports

Transformation to basic prices

Availability of SUTs A. Long series of constant SUT: Denmark, Finland, Netherlands and France. B. Short series of constant price SUT for 1990 s and IO tables for earlier years: Germany, Luxembourg, Sweden C. Long-term series of current price SUT and IO and reasonable deflators: Italy, Austria, UK, Spain D. without reasonable (service) deflators: Belgium , Ireland E. Countries with no SUT or IO before 1995: Eastern European countries, Greece, Portugal Standard approach B-D: derive consistent series of current price SUT and deflate

Derivation of time series of current price SUTs (1) • • • Our default option is to generate long-term series of ESA 95 NA compatible SUTs with FISIM distributed and EUK list of industries Steps to be taken 1. Start with benchmark tables and latest NA output and value added series. 2. Make ESA 95 compatible - software, transport, government depreciation reallocation - industrial classification NACE 1 (EUK) - FISIM 3. Interpolation techniques for non-benchmark years Alternative: make series of ESA 79/NACE 0 and link to ESA 95 series

Current price SUTs (2) • • Reallocation of FISIM: Why and How? • Proportional to output (Eurostat) • Proportional to financial service input • Proportional to outstanding loans Estimation of valuation matrices • Often only total margins, total net tax etc. Blow up to total matrix using reference year, or output matrix structure. IO-tables can be transformed into SUT, avoid negatives (see Beutel 2005) Industry breakdowns • Industry classification and level of detail does not need to be constant across countries and/or over time • Keep as much detail as possible! • Trade-off in detail and resources

Current price SUTs (3) Interpolating benchmarks RAS-techniques (biproportional or more advanced) • • • More research needed into appropriate methods For projection and retrojection: • • • Eurostat method (Beutel 2002): based on value added and final demand totals from National Accounts Kratena (2005) Others?

Constant price SUTs • • Basics: • • Default: use chained Paasche price indices and Laspeyres volume indices Deflate at lowest level possible Deflate intermediate use at basic price Need to deflate margins and net taxes as well Assumption 1: Basic prices are identical in all uses. Assumption 2: industry outputs at both current and constant prices have to be equal to National Accounts values Starting point: (implicit) gross output deflator (basic price) from NA • • All products from industry j have same basic price, so basic price of product i differs by producing industry Calculate domestic basic price of i using Supply table Deflate row in Use Table with this basic price Hence SUT rows are balanced

Austria (Kurt Kratena) • • • SUT/IOT for 1995/97/99 -01 (ESA 95), 1990 SUT ESA 79 -> ESA 95 base with Stat Aus How to backdate from earlier table on basis of NA series, 1976 -1990? • • • How to transform NA data into row and column total of II matrix. Best initial guess of elements of II matrix, using unpublished NA data on intermediate inputs by broad product groups followed by biprop RAS

Austria • • Aim: II matrix at basic prices Row total: II= (Y+M)-(C+I+G+E) by product at basic price • • Ex and Im from Int. trade stat Y from NA (by industry) transformed with S matrix to product C, G, I from NA at purchase price and different classification. Need bridge matrices to convert to basic price and EUK. Bridge matrices change over time and this change is estimated with help of tables after 1990. Column total: II = Y – VA by industry at basic price from Nat Acc Initial guess for elements of II: use time series on input data by 8 broad categories for 1976 -1990. Estimate bridge matrix to disaggregate to EUK on post 1990 data. RAS 1976 -1990.

Finland (Toivola and Pasanen) • • • 1995 -200 x: SUT series. Annual NA series 1974 -1995 conform ESA 95 NOT using old IO-tables Reconstructed SUT at current and constant price for 1974 -1995. Basic data from NA: GO and II at basic price by industry; Im/Ex/C/G/I by product; net tax by product; trade and trans margins (NOT by product)

Finland • For reference years (1975/80/85/90) • • 32 industries S at basic price: secondary production (7% in 1995 fixed) U at purchase price: final use from NA, II breakdown based on 1995 U at basic price: distribute net tax, some assumptions about margins, distributed based on shares in purchase price output Manual balancing of S and U at basic price by product: shift II and secondary production RAS to eliminate remaining discrepancies Linear intrapolation for years inbetween Deflation: • previous year prices, use export and domestic produce prices (? ), volume of margins same as volume of basic price output

Way forward • • SUT framework is overarching framework Keep details when possible General guidelines, each country is different Important to document basic data sources and methodologies used (share) For most countries: might wait for FISIM adjustment of NA in coming months (if!), but older tables can be put in shape Bilateral consultation with other members Workshop?