and the service he enjoys is simply equalled to the fictitious rent. In a 
spirited criticism of this procedure Kopcke-Munnel-Cook (1991) have pointed out 
that the "income" of the house owner is underestimated in this way and that 
anyway part of it should be regarded as investment and not consumption so that 
the procedure of NIPA leads to a substantial underestimate of saving. This 
criticism is well taken but I would prefer to go further to what seems to me the 
root of the evil. The owner-occupier as such is not a business man and while we 
may not be able to avoid fictitious and arbitrary figures in the accounts we 
must at least strive to limit them if possible. I therefore suggest that we 
measure consumption as in other cases by the spending of the consumer which we 
have only to spread over a number of years in this case. The stream of 
consumption is calculated ( like depreciation but with a rather different 
meaning ) by distributing the purchase price of the house over the whole of the 
useful life. We have to introduce then a separate ACCOUNT FOR OWNER-OCCUPIED 
HOUSES which on the left hand side will show the spending of consumers on houses 
in the given period; on the right hand side it will show the stream of "use 
value" measured by depreciation of the whole existing stock of such houses. The 
remaining balance of the account will represent the investment (equal to saving 
) of consumers in housing ( this is independent of the manner of finance - by 
mortgage or otherwise). Quite properly this investment will be the difference 
between the stream of current use and the spending on new houses. The question 
arises now whether we ought to adjust this flow of "use value" to inflation in 
the way which the NIPA does with depreciation. If we want to base our estimate 
of consumption on the actual spending there is no need for such an adjustment. 
As T. Scitovsky ( 1987 ) argued, home-owning is not a business and the owner is 
normally not concerned with keeping his capital intact; if he were he would have 
to consider the appreciation of housing value, too. 
The durable consumption goods might in principle be treated in the same way as 
houses; spending on them would be regarded as investment as is done by the Flow 
of Funds. In the sketch of accounting relations which is represented in the 
tables p. I have chosen a mixed treatment which is unsystematic but may serve 
some practical purposes: I have regarded the durable consumers goods as 
investment only in so far as they are financed by consumers credit, the 
remainder are treated as consumption as before. This is a compromise between the 
old way and a recognition of the similarity which to some extent exists between 
the durables in consumption and those in business. 
I want to go back now to the question of capital gains and consider the case of 
pension funds which is most important in this context because it is here that