| 5/23/2005 | Frits Bos (Netherlands Bureau for Economic Policy Analysis) | I am in favour of including capital services for government owned assets. The present treatment ignores the substantial social cost/opportunity cost of government investments. This is in particular problematic for capital-intensive production by the government, e.g. the provision of the services of roads, rails, dikes and satellites for military purposes. In terms of national accounts statistics, this implies that GDP, the value added and final consumption of the government are underestimated. Furthermore, the price change of value added and final consumption by the government does not take into account substantial changes in social cost/opportunity cost. Finally, the breakdown into final consumption expenditure by COFOG-function is misleading, in particular for capital-intensive functions like economic policy (which includes infrastructure) and defense.
Estimates about the Netherlands can illustrate these issues. Net fixed capital stock of the Dutch government is about 60% GDP in 2003. A real rate of return of 4% suggests an increase of about 2,4% of Dutch GDP and an increase of government value added from 13% to 15,5%. Big changes in the real rate of return, e.g. to 2% or 6%, should be regarded as important changes in the price of government production and be included as such in government value added and final consumption. Final consumption expenditure on COFOG-function is now about 1,5% GDP and on economic policy (which includes capital consumption on infrastructure) about 3% GDP. Including a rate of return on the assets used for defense and economic policy would substantially increase their size. This would also give a much more balanced picture of the composition of Dutch government final consumption expenditure, e.g. of the size of expenditure on infrastructure in comparison with expenditure on education, health care and social protection.
The purpose of including capital services for government owned assets is therefore to give a much better picture of the size and composition of value added and final consumption expenditure by the government; in this picture the role of opportunity costs is made explicit. Providing a much more balanced picture of the government in the National accounts is very important for policy and policy analysis.
The purpose of such an imputation is in many respects similar to that of other imputations, e.g. imputation for services of owner-occupied dwellings or for including all kinds of income in kind. By its very nature, the estimate of this imputation is a bit crude. It should be regarded as crude correction resulting in a much more balanced picture at a very high level of aggregation. It is not intended to compete with the standards for a full cost-benefit analysis of individual projects.
Such a crude imputation should be based on a transparent and simple method and a relevant rate of return. The rate of return could be the current long term interest rate deflated by the price GDP. Highly indebted governments pay higher interest rates than governments with low debt-ratios. This should not be regarded as a problem for imputing capital services on government assets: a major purpose of the imputation is to show the important and meaningful differences internationally (and intertemporally) in the price of government output due to different opportunity costs!
A link with capital consumption as defined by SNA93 seems preferrable, because simple, practical, consistent and meaningful: all countries should already make estimates on capital consumption on fixed assets of the government and therefore on net stock of fixed assets of the government. The real rate of return should therefore be calculated on the basis of this net fixed capital stock. |
