Sam Bostaph's quotation from George Reisman that the statement is "dead   
wrong if..."  misses the point that Pigou and subsequent text book writers,   
including myself in several different text books, were making. This is that   
since GDP is meant to measure market transactions, non-market transaction,   
such as housewife's labour, are not included.  
  
Currently StatsCan has added a number of satellite accounts to the main   
National Accounts to cover many non-market transactions in which we are   
interested. But for many macro issues, such as studying inflation,   
unemployment, employment and exchange rates, it is precisely market   
transitions that are of interest. That is why most of these non-market   
transactions are kept to satellite accounts. (Of course there are some   
inconsistencies when imputed costs such as rent for owner occupied housing   
are included in the main accounts.)  But in the National Accounts Advisory   
Committee of StatsCan we considered this non market issue at great length   
and consulted with the departments who use the accounts. The consensus was   
that the fewer non-market transaction included in the main accounts the   
better -- no matter how important they might be for other purposes, and how   
important it was to develop satellite accounts to measure them.  
  
Richard G. Lipsey