On Mon, 29 Apr 1996 gehrig@banyan.doc.gov wrote:
> I have read about this measure [GPI], and in concept I recall it was an
> interesting idea. However, the problem with using indexes such as
> this is that the biases of the constructors will heavily influence the
> outcome. For instance, how do you put a definite quantitiative value
> on a child dying of malnutrition? ...
Is there an unbiased index that can be constructed? I don't
accept using an index number that we know does not apply because the index
numbers that are constructed in an effort to apply are biased. As a
welfare measure GDP is biased, deflated GDP is biased twice, both in
accepting market values as welfare measures and then by the choice of
price index. And we still know that in real terms, middle quintile GDP
has grown at a very slow rate in the US since the last middle income
growth surge in the later 1970's, bottom quintile incomes have declined,
and only the top two quintiles have grown strongly.
The philosophy of the GPI is to not attempt to construct a
complete 'social welfare' index, but to construct a more narrowly defined
'economic welfare' index that uses market values, but does not add
together spending on 'bads' as well as spending on 'goods'. If we don't
like the way its constructed, we can construct our own. We don't get
away from exercising critical judgement simply by saying 'add all market
values of newly produced goods and services': that *is* a critical
judgement, and it should be defended as such, not on the grounds that if
it is a mistake, at least it is a common mistake.
We can see the direction of the GDP bias by the simpler World
Bank index that gives consumption and government expenditure less defense
spending. That's available in the Penn World tables (check the www
server on the CSF site that hosts this list at http://csf.colorado.edu,
there is a pointer to the Penn World tables in the PKT archives).
Subtract spending on prisons, you get an even greater discrepency between
GDP growth trends a MPWI (my personal welfare index) growth trends.
Subtract any value from GDP that you think should not be counted as a
welfare improvement, and justify the values that you have retained as
welfare measures, and if there is a discrepency with what the GPI says,
you have a disagreement between your critical judgement and that of the
GPI buidlers. But we can no longer use GDP on the excuse that there is
nothing else available. And, after all, that was why the GPI was
constucted and published, warts and all.
Virtually,
Bruce R. McFarling, Newcastle, NSW
ecbm@cc.newcastle.edu.au