Environmental Accounting
Environmental Accounts in Physical Units
Aldo Femia
A contribution to the development of the European EW-MFA Standard Tables and Compilation Guide
Roma, 18 gennaio 2005
Contents
1.Overview______
2.Standard tables
3.Inputs for the Compilation Guide
3.1Construction Minerals
3.2Used agricultural by-products
3.3NAMEA breakdown of DMI
3.4Residence principle and fuel consumption
3.5Packaging Materials imported and exported with products
3.6Material stock changes
3.7Unused Domestic Extraction
References
APPENDIX A – PROPOSED STANDARD TABLES______
APPENDIX B – DE-NAMEA, Italy, 2001______
APPENDIX C – CODES OF 4-DIGIT NACE HOMOGENOEUOS ACTIVITIES PRODUCING INVESTMENT GOODS OR CONSUMER DURABLES
Aldo Femia Istat – Environmental Accounts in Physical Units
A contribution to the development of the European EW-MFA Standard Tables and Compilation Guide
1.Overview
This document presents the Italian proposals for the finalisation of the Standard Tables and provides some inputs for the drafting of the Compilation Guide to be agreed upon by the European Task Force on Economy-wide Material Flow Accounting and derived Indicators, building on the documents circulated to the TF members after their meeting in Luxembourg on 8-9 November 2004. It is also meant to respond to the requests made to the TF members, and to Italy in particular, during the TF meeting mentioned above (see the meetings’ Minutes).
2.Standard tables
Appendix A reproduces the worksheets included in an excel file sent along with the present document. The 6 tables included are a revised version of the Standard Tables sent out by Eurostat (document of 14 December 2004).
Table 1 is on Domestic Extraction. DE is the most basic component of all country-level MFAs, and therefore we think it should be reported first. The table we propose is based on table 2 of Eurostat’s document, but a split by kind of material is also introduced. Assuming that the activities are classified into homogeneous production branches, a many-to-one relationship can be established between the extracting activities and the kinds of extracted materials. Not-applicable cells have been blackened accordingly, as a guidance for compilation and to prevent cross-classification errors. It is important to stress that this table contains all the data for the NAMEA breakdown of DE only if NAMEA activities are homogeneous production branches.
This table can be made more general and applicable to a classification of industries by main activity (non-homogeneous branches) by removing the restrictions of the black cells and adding the appropriate additional rows, according to the possible occurrence of material extraction as secondary production in activities other than the ones already included in the table.
Table 2 concerns Imports, classified according to the kind of material they are derived from. Due to the presence of products in which two or more kinds of materials are mixed, no simple and reliable method is available to allocate all imports to the categories applicable to Domestic Extraction. Therefore, the residual category “compound products” has been introduced. It must be highlighted that the other categories are defined as “products from”, which means e.g. that plastic products would be classified as “products from fossil fuels” and not as “other” under the “compound products”.
Table 3 concerns Exports, dealt with in the same manner as Imports in the previous table.
In table 4, the most simple material flow indicators based on domestic extraction and physical trade – namely DMI, DMC and PTB - are automatically derived by the use of formulas, by linking to the data of the former sheets.
Table 5, concerning the allocation of Imports to importing economic activities and final use activities, is almost identical to table 3 of Eurostat’s document of 14 Dec. 2004[1]. With respect to the latter, two headings have been introduced in order to highlight the distinction between intermediate and final uses. It can be pointed out that this distinction, as well as the allocation of intermediate Imports to the branches directly using them, requires the application of national accounting expertise, and table 5 should therefore not be regarded as a necessary part of a basic EW_MFA exercise.
Table 6 corresponds to table 1 of Eurostat’s document: as these indicators are derived from the data in tables 1-3, we think it advisable to have this table at the end of the set.. The “resource productivity” have been redefined as the inverse of those featuring under this heading, which we renamed “resource intensity indicators”. For completeness, in the “per capita indicators” referring to Imports and Exports we have added the indicator “compound products imports per capita”. The indicators referring to DMI and DMC per capita by kind of material (rows with a grey background in table 6) are doubtful in meaning, therefore we recommend not to have them in the standard tables, but only have “DMI per capita” and “DMC per capita”. In the excel file, formulas have been introduced in order to avoid the introduction of computation errors by the compilers.
3.Inputs for the Compilation Guide
3.1Construction Minerals[2]
Two different estimation procedures have been put in place at Istat, that exploit information from various sources, trying to integrate them in order to have a complete coverage of extraction activities concerning these materials. We will describe here mainly the procedure concerning materials extracted from quarries, which gives on average 95% of total construction minerals’ input from domestic extraction in Italy. The remaining 5% is soil from excavation activities which is reused in construction.
The estimation procedure adopted for the construction of the time series builds on the results of a regression model and integrates these results with the better knowledge of the phenomenon provided by the PRODCOM survey, whose results are available for the latest years of the period for which the estimate has been constructed.
The regression model describes the data on quantities of construction materials extracted, aggregated at the province level, that have been reported through the years by the producers in the survey on the production of quarries and peat fields[3].
The “regression” step of the estimation procedure adopted has consisted in the reconstruction of the missing data (concerning the province/year couples for which no data were reported), made by extending the statistical regularities that could be observed in the available data with reference to the relationship between the quantities of extracted materials that had been declared and relevant supply and demand variables (number of employees of the quarries, value of public works realised, buildings’ growth in cubic meters)[4].
The following step of the estimation procedure has consisted in the comparison of the results with those of the PRODCOM survey, which were available for 1997 and 1998, and had in turn been adapted in order to cover extraction sites with less than three employees, not covered by the PRODCOM survey. This comparison resulted in the discovery of a substantial “hole” in the estimate based on administrative data, which however were the only available source for the years before 1997. The corrected PRODCOM data have therefore been adopted for the last two years of the time series, and a further correction has been made to the results of the regression model for the previous years, rescaling everything to make it consistent with 1997 and 1998. Therefore, the level of the series is dictated by the PRODCOM results, while its profile is determined by the results of the regression model.
As far as reused soil from excavation is concerned, we used the information given for 1997 by the waste statistics published by the Italian Environment Protection Agency (ANPA, now APAT). According to our calculations, based on the data provided by this source, the percentage of reused waste soil resulted to be 28,4%. The total quantity of excavated soil, on which the estimate of the reused quantity is based, was in turn derived by a procedure which is described in Femia et al. 2001 (§2.2.3.4, concerning unused flows).
3.2Used agricultural by-products[5]
These comprise straw from grains production used as fodder, as bed for animals and for other minor production purposes, and beetroot leaves used as fodder. The grains concerned are wheat, rye, barley, oats and rice.
There are two steps - to which two distinct sets of coefficients correspond – to be considered: 1) estimation of the quantities of the materials concerned; 2) estimation of the quantities of these materials which are used and therefore enter DE, as percentages of the total quantities.
1)Like unused biomasses, the total quantities have been calculated by the application of technical coefficients to harvested plant production, as no data on actual weight of these flows are directly available. The estimation has followed a two-steps procedure, with the use of two groups of coefficients. The coefficients of the first group transform in dry weight the used materials removed (the quantities included in DMI), which are reported in harvest statistics in terms of total weight at the time of harvest. The coefficients in the second group allow to calculate, starting from harvested products' dry matter thus calculated, the connected flows of unused materials, again in terms of dry matter. The multiplication of the two coefficients gives the following results for straw in dry weight per unit of grain fresh weight: from wheat, 0,67; from rye, 1,65; from barley, 0,63; from oats, 0,54; from rice, 1,22. The choice of accounting these flows in terms of dry matter was imposed by the fact that the data available in literature usually concern dry matter contents of used and unused parts of the plants when they are ready to be harvested. Standardisation to 15% water content has not yet been implemented, but is straightforward.
2)When the time series of Italian EW-MFA indicators was first estimated a quite raw hypothesis had been adopted: following Bringezu and Schuetz (2001), we assumed that half of the straw produced had been used, while the rest was considered part of the unused flows of agriculture. Later, national accountants in charge of the estimates of agricultural activities provided the information that all of the straw produced is included in domestic production, as it has a market, and therefore we now consider it all as used. As for beetroot leaves, they have been considered all used from the beginning, even though they don’t have a market.
3.3NAMEA breakdown of DMI
The matter is very different depending on whether Domestic Extraction or Imports is concerned.
Also, the problems are quite different according to the way NAMEA activities are defined, whether in terms of homogeneous production branches or not. As Italy does not have any specific experience on allocation of DMI to non-homogeneous production branches we will limit ourselves to the simple case of homogeneous activities.
In the homogeneous activities case DE is very easily allocated to “extracting” activities, by simply looking at the NACE classification for determining for each product of extraction the activity that provides it as main product. The DE-NAMEA table in Appendix B provides an example of the results thus obtained for Italy. As pointed out in § 2, this table contains the same data of our proposed standard table
As far as imports are concerned, their allocation to domestic activities should answer the question “which are the activities that demand material inputs from the rest of the world?”. This question would not be answered by a table allocating Imports to producing activities, which would not be very difficult to construct (at least for the case of homogeneous branches). The allocation of imports to importing (final and intermediate activities) is a much more complicated business, which in Italy has been addressed so far only at an experimental level a) in connection to the use of the data in an Input-Output analytical framework and b) in exercises for the construction of a PIOT.
In the first case, the allocation of the Imports, initially classified by main producing activity, has been carried out on the basis of the monetary Input-Output table for Imports. The precision of this method depends on the degree of disaggregation of the table.
In the second case, we obtained a physical Use table of Imports by kind of good and importing industry (not by homogeneous branch, though). The text in the following box, extracted from Femia et al. (2004) explains how this exercise has been carried out.
The construction of the matrix imports’ by product (CN8) and by economic activity of the importing enterprise
(Author: Carmela Pascucci, Istat)
This section describes the sources of information used and the procedure followed in order to obtain the matrix of the quantities of imported products by kind of product and main economic activity of the importing enterprise (from now on “imports by main activity”), as well as the results obtained.
The sources of information used
To get the “imports by main activity” matrix it is necessary to combine information present in two different registers: the Enterprises Business Register and the Foreign Trade Operators Business Register.
The main problem to be taken into account with reference to the linking of these two registers is the different definition of the statistical unit and of the field of interest of the two registers.
The enterprise is the basic statistical unit of the Enterprises Business Register, in which it is identified on the basis of its fiscal code. The field of interest of the Enterprise Business Register is given by the enterprises belonging to a subset of economic activities, which does not include agriculture and fishing activities and the Public Administration.
The field of interest of the Trade Register is given by commodity trade. The Trade register refers to the foreign trade operator as basic statistical unit, which is identified by the VAT code. A single economic operator can have different VAT codes within the fiscal year according to changes of location of its business headquarter. It is clear that this may cause both an overestimation of the number of foreign trade operators and problems in linking the two Registers.
The Trade Register derives indirectly from the micro data of the foreign trade survey. From a technical point of view this procedure is very easy to carry out. Indeed it is sufficient to sort all the records by the variable “VAT code” and to aggregate information on micro data in order to obtain a single record for each VAT code. Nevertheless this presents some statistic problems. In fact, for some records the VAT code information can be incomplete, so the value and the quantity of the import flows of the operators included in the Trade Register is often lower, for a given NC8 code, than the total imports flow of that code. It is necessary, also, to consider that the origin of the data is a statistical survey whose reference unit is the commodity flow and therefore, all variables are referred to the imported commodities. Any commodity-related variable must be used carefully, according to the new statistical unit used in the Trade Register (the foreign trade operator). In particular, the code of the product-related economic activity (CPA) refers to the imported good (i.e. says which product is taken into account) and not necessarily reflects the main economic activity of the importing enterprise, that can be found in the Enterprises Business Register.
The linking key adopted and the information extracted
The fiscal code has been chosen as key to link the two different business register, because this information is more reliable than the VAT code. In general, a one to one relationship occurs between the VAT and the fiscal code, except for the case of a change of location of the business headquarter of enterprises, as said above. Before linking the information of the two registers, the information of different VAT codes within the Trade Register that correspond to a single fiscal code in the Business Register must be added up.
A register’s Tributary file joining together the fiscal code and the VAT code has been used to link the two registers.
The information of the two registers has therefore been linked using the fiscal code and merged to obtain a matrix of the quantities of imported products by main economic activities of the importing enterprises.
The information taken from the Foreign Trade Operator Business Register is:
- products imported by every economic operator according to the more disaggregated classification available, the Combined Nomenclature (CN8), in 1997;
- imports value of every operator for each item of CN8 in 1997;
- imports quantity of every operator for each item of CN8 in 1997.
The information taken from the Enterprises Business Register is only the enterprise main economic activity according to the Level 5 of the General Industrial Classification of Economic Activities (NACE Rev.1).
Coverage analysis
It is important to analyse the coverage of the final results in terms of trade values and quantities, as there are reasons to expect that a consistent number of records in both of the two registers do not match. Indeed, the two following different causes of loss of coverage can be highlighted:
- the incomplete VAT codes present in the Trade register;
- the different fields of interest of the two registers.
The coverage ratio of the results of the linkage in terms of values and quantities on national imports is high. In fact, in terms of values the coverage ratio of the data processing is 98,2 per cent on national imports values, in terms of quantity the rate is 98,1 per cent.
3.4Residence principle and fuel consumption[6]
The main discrepancies between material flows taking place on a nation’s territory (“domestic” flows) and residents’ material flows (“national” flows) – i.e. between DMI as it is currently computed and the same aggregate as it should be in order to comply with the residence principle of National Accounting – are due to the difference between the purchase of fuels for transportation processes by resident units abroad and the purchase of fuels by non-resident units. The consumption of these fuels gives a parallel discrepancy between “domestic” and “national” air emissions.