Item 4.3.4 of the agenda
B1 - National accounts methodology,
statistics for own resourcesLuxembourg, November 1999
Capital stock accounts
Anton Steurer
Eurostat B1
Joint meeting of the
Working Party ‘ECONOMIC ACCOUNTS FOR THE ENVIRONMENT’
and the Sub-Group ’Environmental Expenditure Statistics’
Meeting of 6 and 7 December 1999
BECH building – Room ‘Quételet’
1
Capital stock accounts and models
In early 1999, Eurostat visited the Statistical Offices of Germany and the Netherlands to discuss the methods and systems in place to estimate the consumption of fixed capital (‚depreciation‘) of environmental capital.
The two systems, while different in the details and adjusted to differences in availability of basic data, were surprisingly similar in the principles applied and the main functions these systems are designed for. In essence,
-both systems provide data on the environmental capital stock in place using the perpetual inventory method based on time series of investment data, lifetime assumptions and price indices
-both systems have a similar level of detail in terms of number of industries, environmental domains, types of capital distinguished etc.
-both systems are used to estimate the consumption of fixed capital
-both systems are used, to a varying degree, to estimate the operating expenditure of the equipment
-in both cases an initial capital stock for a year in the 1970s was established, either by survey or by estimation
Therefore, the experience available in the Netherlands and in Germany could form the basis for a generalised description of models and systems to calculate the stock of environmental protection capital in place, the consumption of this fixed capital as well as the use of these systems for calculating (some parts of) operating expenditure associated with the capital stock. These systems could also qualify as ad hoc solutions for other countries.
In this context the German experience with building up an environmental capital stock system for the new Länder (former GDR) is of particular relevance. Such systems can generate estimates of quite acceptable quality that can be improved progressively as more and more detailed primary data from surveys become available.
A summary of the systems in place in the Netherlands and in Germany is provided below.
A summary of the Dutch method
Principles of the Dutch method have been described in E. J. Dietz: Costs and financing of environmental control by the Public Sector – Methodology and results and in J. W. Pouwelse: Assessing the current costs of facilities for environmental control – Experience in the Netherlands (see Eurostat 1994: Contributions of Member States and EFTA Countries to the SERIEE System 1992 and 1993, contributions B1 and C3). Since then, improvements have been made in several respects, in particular for the public sector.
What is of relevance in a capital stock modelling context is that:
The consumption of fixed capital for the public sector is based on the data for depreciation as can be found in the public budgets (behind these data there may be different attitudes towards depreciation so that data are not strictly calculated according to the same methods
Data on the stock of fixed capital is available for industry. Surveys of the stock of capital were made in 1979 and 1984. These surveys produced data on the capital stock in place and on the age of that stock. Later, the stock data were derived from the annual investment data plus assumptions on the lifetime.
Data on operating expenditure are annually observed for a number of big companies (representing about 20% of total current expenditure). For the other enterprises the data are surveyed once in the lifetime of an investment. This is done by sending, in year t+1, a questionnaire on current expenditure to those units that reported new investment in the year t. These initial data are then used to estimate the operating expenditure during the whole lifetime of the equipment by applying price indices and an ageing factor to account for increased maintenance (based on engineering studies).
The capital stock data are available for each year since 1968 in a breakdown by 6 environmental domains (waste, waste water, air, noise, soil and nature protection), broken down by two types of capital (machinery and buildings), by end-of-pipe and integrated investment and for 14 industries.
Summary of the German Stock of Fixed Capital Account (SFCA)
In 1986, the national accounts division of the German Federal Statistical Office established an environmental satellite account for environmental expenditure. This satellite account produces data on capital stock (gross and net) and consumption of fixed capital in current and constant prices. No basic statistics on business current environmental expenditure were available so that the model was also used to produce estimates of the current expenditure linked to the capital stock in place (i.e. operating expenditure). Therefore, the SFCA allows calculation of the (non-market) output of ancillary environmental protection activities. Recently, the responsibility for maintaining and updating the capital stock model and for the annual publication of the results was transferred from the national accounts division to the environmental accounts section (Gruppe Umweltökonomische Gesamtrechnungen – UGR).
The main recurrent inputs to the capital stock model are environmental investment in the mining, manufacturing and energy supply industries (based on an annual survey) and environmental investment of government units (based on budget analysis and public finance statistics). Further, assumptions on the lifetime of equipment by type (machinery, vehicles, and infrastructure) with mortality functions (probability distribution of earlier or delayed retirement around the average lifetime) and various price indices are needed for the calculations in current and constant prices.
When the SFCA was set up in 1986, a detailed estimate of the capital stock in place at the beginning of 1975 had to be made. This was done by establishing very long times series of investment: from 1960 for enterprises and from 1950 for government. Additionally an initial capital stock for government in 1950 (for the sewage system) was taken into account. However, only the data from 1975 are considered reliable and are published.. The time series before 1975 and the initial capital stock were based on all available information (estimates of private institutes, finance statistics, etc.).
The model is principally using the perpetual inventory method to calculate capital stocks and consumption of fixed capital. The estimate of operating expenditure is made with the help of a matrix of coefficients that is multiplied with the matrix of the gross capital stocks. Coefficients are based on expert assessment, case studies and engineering data and provide for an estimate of the operating expenditure associated with different kinds of equipment.
The SFCA generates data for each year beginning with 1975 by 4 environmental domains (air, water, waste, noise) and by 2 institutional sectors (corporations and government) and for corporations in a breakdown of 20 industries.
The data generated are:
- Gross capital stock for environmental protection in current and constant prices
- Net capital stock for environmental protection in current and constant prices
- Consumption of fixed capital in current and constant prices
- Current (operating) expenditure linked to the capital stock in place for the mining, manufacturing and energy supply industries
From this can be derived: expenditure (investment plus operating expenditure) and costs (consumption of fixed capital plus operating expenditure) as well as the non-market (ancillary) output.
Consumption of fixed capital for public sewage networks etc. is also calculated (at the time the model was set up the national accounts did not calculate consumption of fixed capital for government-owned infrastructures but this has changed with the ESA 1995).
Recently, the environmental accounts section of the German Federal Statistical Office has working on integrating the New Länder into the capital stock model. For details please see Document ACCT-EXP/99/4.2.4 (Implementation of SERIEE in Germany - Reporting year 1995). The integration of the New Länder posed several challenges. For example, much of the existing stock lost its value after German re-unification – e.g. all landfills had to be upgraded, big polluters had to upgrade their equipment - mainly sewage networks were still of value.
The next step will be various revisions of the SFCA based on new sets of basic data that will become available within the next 1-2 years. These new basic data include:
- Since the 1996 statistical year the environmental expenditure questionnaire to businesses includes questions on current expenditure in a detailed breakdown (e.g. separate data for operating expenditure by type – wages and salaries, energy, maintenance, other operating expenditure, payments for external services etc.). These new data will allow consolidation with the estimates of operating expenditure produced by the SFCA.
- A survey on the ‚Composition of environmental investment 1994-1997‘ is under way. Results of the survey will allow a revision of the capital stock with respect to its composition (groups of goods and services the capital goods are made of – this is relevant for the application of price indices in the calculation of current and constant prices).
- A survey on ‚Producers of goods and services for environmental protection‘ (Environment industry) is currently ongoing – this will allow further refinements of the SFCA
- The improvements of the business register will allow to better quantify the specialised producers of environmental protection services (NACE Rev.1 class 90.00)
1