GDP alternatives

[-1=e^ipi]

It's not about 'being more likely to spend money', I never said that. It's about the utility of an additional dollar.

Utility maximizing individuals will tend to spend their initial income on essentials and then later income on things that give less and less marginal utility.

[bush_cheney2004]

Utility maximizing individuals will tend to spend their initial income on essentials and then later income on things that give less and less marginal utility.

Not always......marginal utility does not apply to many aspects of economic activity (education, durable goods, books, some commodities, etc.)

GDP aggregates economic activity without these limitations, but was never intended to be the only metric.

[-1=e^ipi]

Okay, looked into the definitions of both HDI and IHDI.

Both indices are ultimately dependent on arbitrary choices of minimum and maximum values of life expectancy, GNI per capita and schooling. This results in a very questionable ranking of outcomes. For example, a society of people with a life expectancy of 20 years that earn $1000000000000000000000000000000000000000000000000000000000000000 per year and have 20 years of schooling will have a lower HDI than a society of people with a life expectancy of 21 years that earn $101 per year and have 1 year of schooling.

Another issue is the choice of how they weight health, income and education, which is also arbitrary. They use Health^1/3 * Education^1/3 * Income^1/3, but they could have just as easily picked Health^1/4 * Education^1/4 * Income^1/2, or alternatively, they could have added another factor, such as Freedom, in which case they could have picked Health^1/4 * Education^1/4 * Income^1/4 * Freedom^1/4. The choice of parameters doesn't have any empirical basis.

I think the relative weighting of Health to Income is probably reasonable (since they are taking the logarithm of income) since for higher HDI values it arguably is what you would get for the average person's von Neumann-Morgenstern utility function (assuming a coefficient of relative risk aversion of 1, which is not that unreasonable). However, the weighting of education with respect to the other 2 parameters is rather questionable. For example, if you have 2 societies with equal life expectancy, one consisting of gender studies graduates with 18 years of schooling and a GNI per capita of $8255 per year and a second society of high school graduates with 12 years of schooling and a GNI per capita of $75000, they would count as having an equal HDI. Personally, I'd prefer the second society.

One more thing is that both HDI and and IHDI are based upon a coefficient of relative risk aversion of 1, which is arguably slightly below what the best empirical evidence suggests it is. This means that HDI and IHDI might not be 'socialist' enough.

The expected value of the von Neumann-Morgensterm utility of an individual in society (which is the same thing as the Hillian Social Welfare of society) seems like a better measure of societal well being that has a stronger theoretical (and empirical once you measure the appropriate parameters) measure of well being.

[Michael Hardner]

Moved topic to 'Business and Economy'