Istat s experience in the KLEMS data set

Istat’ s experience in the KLEMS data set Antonella Baldassarini Massimiliano Iommi The First World KLEMS Conference Harvard University August 19 -20, 2010

EU KLEMS experience: general overview (1/2) Two Italian public institutions have been involved in the EU KLEMS project: Ø the Institute for Studies and Economic Analysis (ISAE) as official member of the Consortium Ø the National Statistical Office (Istat) as subcontractor. ISAE is a public research Institute that conducts analyses, research projects and forecasts suited to economic and social policy decisions. Istat is a public Institute that provides official information on social, economic and environmental conditions of the Country, in strict compliance with legal provisions on confidentiality. Boston, 20 August 2010

EU KLEMS experience: general overview (2/2) The database on growth accounts has been implemented by industry (Nace Rev. 1. 1) with a breakdown into contributions from capital (K), labour (L), energy (E), materials (M) and service inputs (S) as follows: q Istat has provided official time series at 30 / 60 industry level (Nace Rev. 1. 1) on basic variables q ISAE has provided the breakdown requested (Nace 60+) by industry, the analytical module on labour input and the backward calculation of time series Boston, 20 August 2010

Basic variables from Istat: output and intermediate inputs (1/2) q Output measures § Official National Accounts series (gross output and gross value added) § Time period since 1970 q Intermediate inputs § Supply and Use Tables at 60 industry/product level § Use tables distinguished between imported and domestic tables at current and previous year prices § Time period since 1992 Since the revision of the ESA 95 Transmission programme (November 2007), all EU countries are required to provide yearly SUTs at current and previous year’s prices. Output and intermediate inputs data needed for KLEMS database are now included in the standard data produced by European NSIs. Boston, 20 August 2010

Basic variables from Istat: output and intermediate inputs (2/2) q Further issue § In Italy, SUTs are compiled as a fully integrated part of national accounts calculation § SUTs are the main tool to balance the estimates from the expenditure approach and the production approach at the level of products § SUTs are the compilation framework for constant price estimates too (but balancing supply and uses at the same time at current and constant prices is not implemented yet) Boston, 20 August 2010

Basic variables from Istat: GFCF (1/2) q Gross fixed capital formation cross-classified by industry (30) and the following assets (9): § machinery and equipment § furniture § hardware § communication equipment § software § road transport equipment § air, sea and rail transport equipment § construction works (distinguished in dwellings and other buildings and structures) § other intangibles and services Disaggregation by asset and by industry provided is higher than what is required by ESA 95 Transmission Program. Boston, 20 August 2010

Basic variables from Istat: GFCF (2/2) q Further issue § According the new assets classification (AN_F 6+), ICT assets and intangible assets can be identified § Adoption of Nace Rev. 2 requires a depth changing industry classification and it offers the opportunity to extend the industry detail of GFCF estimates but: 1) resources needed to produce reliable estimates increase with the level of industry disaggregation required; 2) the backward calculation of long time series to produce capital stock estimates using the PIM model seems difficult Boston, 20 August 2010

Basic variables from Istat: labour input (1/2) q Labour Input measures: § Persons employed Time period since 1970 Breakdown by industry: - 101 industries since 1991 - 32 industries before 1991 § Full-time equivalent units (as a proxy of hours worked) Time period since 1970 Breakdown by industry: - 101 industries since 1991 - 32 industries before 1991 § Hours actually worked Time period since 1980 Breakdown by industry: - 60 industries since 1991 - 32 industries before 1991 Boston, 20 August 2010

Basic variables from Istat: labour input (2/2) q Further issue Labour input measures (persons employed, jobs, full-time equivalent units and hours actually worked) are perfectly integrated within the national accounts framework. Data are consistent: § among them § with output estimates § with compensation of employees data Boston, 20 August 2010

Basic variables from Istat: labour compensation q Compensation of employees: § Wages and salaries Time period since 1970 Breakdown by industry: - 101 industries since 1991 - 32 industries before 1991 § Social contributions Time period since 1970 Breakdown by industry: - 101 industries since 1991 - 32 industries before 1991 Boston, 20 August 2010

ISAE contribution to database release q Providing all basic variables according to the EU KLEMS breakdown by industry (Nace 60+) q Hours worked for the period 1970 -1979 q Labour input module with information on: § Persons employed and hours worked (detailed by gender, age, educational attainment) gathered by the Population Census for years 1971, 1981, 1991, 2001 § Labour compensation (detailed by gender, age and educational attainment), obtained using the Bank of Italy survey on households income 1977 -2004 Boston, 20 August 2010

Istat checking on KLEMS database: capital input q Gross fixed capital formation data are coherent with official data (as expected, differences are only registered for the price index of computing equipment) q From a methodological point of view, Istat calculation on real fixed capital stock follows a different assumption respect to EU KLEMS calculations: Istat uses a linear depreciation and bell shaped retirement distribution while EU KLEMS uses the geometric model. Relevant differences emerge in terms of growth rates q Istat is going to develop specific studies on the average service life-depreciation rate by asset Boston, 20 August 2010

Istat checking on KLEMS database: labour input q Istat calculates data on hours worked not detailed by type (gender, age, educational attainment). Istat is able now to provide the labour input in terms of persons employed and hours actually worked by industry and types (Source of data: LFS on a continuous base since 2004). q In the next future, sources of data on labour input remuneration will be investigated, in particular the EU SILC Survey. The above survey focuses on income and detailed income components are collected mainly at personal level. In addition, information on labour, education, health, social exclusion and housing condition information is obtained. q EU SILC is based on the idea of a common “framework” that is: the harmonised lists of target primary (annual) and secondary (every four years or less frequently) variables to be transmitted to Eurostat; common guidelines and procedures; common concepts (household and income) and classifications aimed at maximising comparability of the information produced. Boston, 20 August 2010

Open issues for the future releases (1/2) üConsistency between KLEMS sources of growth analysis and NSIs’ estimates This issue will be very important, at least for EU countries, when Eurostat releases its own estimates for EU KLEMS. What will be the official data for countries that produce their own measures of productivity (e. g. Italy)? üKLEMS coverage The EU KLEMS database is extended to the whole economy. Is this the best option? Some new in depth analysis on non market services according to the growth accounting theory should be promoted? What about owners occupied housing? In its productivity calculation, Istat excludes the activities of the General Government institutional sector, the letting of own property, the activity of private households with employed persons, the activity of the extra-territorial organizations and bodies) Boston, 20 August 2010

Open issues for the future releases (2/2) üThe level of industry disaggregation There is an obvious trade-off between reliability and industry detail. How far should we be demanding with respect to the industry detail? ü Availability of time series data With the change from Nace Rev. 1. 1. to Nace Rev. 2 classification, the availability of official time-series for gross output and intermediate inputs according to a supply and use framework is not guaranteed before of the year 2008 ü Capital input estimates For international comparability, it should be better to clarify well which measures of capital stock to calculate and for which purpose (capital as a provider of services in production or as a store of wealth) Boston, 20 August 2010

What Istat has learnt by the EU KLEMS experience q The KLEMS database is fundamental to guarantee the international comparability and to promote a different perspective analysis of the economic growth, in particular among NSIs q Analysis of productivity, prices, industry structures, technology, innovation indicators and labour markets is of fundamental relevance, better if developed with the support of the NSIs q EU KLEMS idea and experience should be used for realizing a new database according to common criteria, practical and methodological, well defined among NSIs q In Europe, Eurostat seems to the best institution that can implement the KLEMS database Boston, 20 August 2010