In the late 1980s I received an impressive missive from the Crown Prince of Bhutan inviting me to a conference, which would seek to provide measures of a country's level of happiness to replace the traditional measures of GDP as the sign of a country's progress. Thinking this as something of a joke I politely declined.

But, in the 1990s, many economists and politicians embraced this "happiness economics". In the UK, Tony Blair's Policy Unit took it up, and David Cameron, the Tory leader, has proclaimed that "gross national well being" should replace GDP as the policy goal. And the Labour peer Richard Layard has written an influential book Happiness: Lessons from a New Science. What are we to make of all this?

The first point to note is that this "new" economics is another attack on capitalism. It echoes a refrain of the romantic critique of capitalism: capitalism breeds unhappiness. Many, from Marx to Etzioni, have seen this leading to immanent trends for capitalism's self-destruction. But with the spread of globalising capitalism it is the theories of immanence and not capitalism that seem to destroy themselves.

The second is that, ever since Bentham sought a happiness meter to measure the pleasures and pain, which would allow his Utilitarian calculus to be put to use for public policy, many economists of a dirigiste disposition have sought to create a quantitatively informed Brave New World.

But lacking a happiness meter they had to fall back on basing "happiness" on the satisfaction of individual desires, as revealed in actual choices in the market place.

Hating the recent victory of the market over the plan and the great material prosperity it has engendered, the new dirigisme is seeking to replace the quantitative measures of this prosperity provided by GDP, by purportedly "hard" measures of subjective well-being, provided by the cross-country and cross-cultural data from the World Values Survey.

There are serious methodological problems with the data and the inferences that some happiness researchers have drawn, which Helen Johns and Paul Omerod (Happiness, Economics and Public Policy, Institute of Economic Affairs, London) show, are far from robust.

Many variables like longevity, crime and unemployment, which would commonly be expected to increase a country's well-being, do not seem to affect the happiness indices. What the happiness research finds is that, on its measures of reported subjective happiness there is a moderate positive correlation with per capita income (in PPP $), which is strongest with countries with a per capita GNP up to $10,000 (in 1995).

"There are no rich countries where people's happiness, on average is low. But, for the rich countries, it does not seem that higher per capita income has any marked effect on happiness" (B Frey and A Stuzer: Happiness and Economics, Princeton).

This is hardly surprising, because real income is likely to be only one element in a person's happiness, and as economic theory postulates diminishing marginal utility from increased income (consumption) we would expect rich countries to have reached what Frank Ramsey postulated as the "Bliss" level of utility.

But, as the data refer to "reported" rather than "actual" happiness, even this inference is insecure. Happiness economics has also tried to econometrically estimate the determinants of "happiness", and to argue that the standard microeconomics of utility theory needs to be revised. These byways need not detain us.

More serious are the implications various dirigistes have drawn from this line of research. Some have argued that the estimated "happiness functions" provide an approximation to the social welfare function - which hitherto had been a merely theoretical construct of modern welfare economics - yielding the optimal values of its determinants which public policy is then expected to implement.

But this raises Arrow's famous Impossibility Theorem, which showed that under a set of "reasonable assumptions" it was not possible to aggregate individual preferences into a social welfare function which consistently ranked outcomes: something which a dictator would obviously be able to do, and a role these social planners hope to fill!

The other implications are more disturbing. The rat race for social status is deemed to increase unhappiness requiring heavy taxation of the winners. As working too hard is a source of unhappiness, work should be taxed more heavily.

So that people substitute leisure for work at the margin - something the French government has imposed by fiat with its statutory reduction of the working week with no apparent improvement in happiness to offset the obvious losses in economic output.

These complaints about the psychologically dysfunctional effects of capitalism confuse questions about how best to make a living (material beliefs), with those concerning, in Plato's words, "how one should live" (cosmological beliefs).

Capitalism provides a new and highly productive way of making a living. Whilst the social ills adduced to capitalism, like the decline in family values and companionship, have less to do with capitalism than individualism.

It is the demoralisation of their societies rather than the instrument of their prosperity - capitalism - which is to be blamed for the social ills cited by the "happiness researchers".