Undervaluation of housework in national accounts

Undervaluation of contribution of women to development
One of the oldest conundrums in the theory of national income statistics is the question: if a man marries his housekeeper, is it correct to show a decline in national income. The answer given is, in a sense, a test of the respondent's concept of what national income statistics are intended to exhibit. The answer 'yes' implies that, because the marriage signifies the disappearance of a marketed activity from the sphere of measured production, national income, or GNP, being regarded as essentially the sum of market activities, correctly falls. The answer 'no' implies a recognition that the woman's activities as a housekeeper may be expected more or less to continue as before: there is no change (necessarily) in the total flow of goods and services.
Working women in the UK are estimated to have lost a total of $27 in overtime in 1987 because of home responsibilities. This is changing, however, as men assist in housework and childraising. Married men in the USA now do an impressive nine minutes more housework each day than they did twenty years ago.
The household is a kind of enterprise, involving decisions on current and capital outlays, on choices between paid work and work at home and between work and leisure, etc, on the basis of a mixture of market prices and shadow prices for the various alternatives. Indeed, for most households, important decisions are in fact influenced to some extent, explicitly or implicitly, by this kind of economic rationalization.
The social and institutional as well as the economic constraints on economically rational planning of a family economy, together with the non-economic variables that may quite rightly dominate decision-making, render this type of approach inappropriate and unrealistic. Although aspects of household behaviour may be subjected to tests of economic rationality by the social scientist, the formal incorporation of the household as an enterprise in national accounting does not seem at all necessary.
Aggravated by 
(F) Fuzzy exceptional problems