Subj:     Flaws and alternative

Date:     03/02/03



Dear Alex

Thanks very much.   I've got various people who are very interested in what I'm

doing, including the professor of moral philosophy in Oxford (ex-economist, John

Broome) and a professor of political science in Harvard (Gary King) who's an

expert in the kinds of things I look at.     Some of them are fundamental

questions of what is an adequate structure for welfare economics.     I could

publish them in academic journals but that could take a couple of years.  


I've realised something else.   Economists make the mistake of thinking that a

1% rise in a person's consumption expenditure means they consumed 1% more.  

That is not something we can assume, even where the expenditure is adjusted for



Here's a very simplistic example.   If a government makes basic food much

cheaper, the overall inflation rate in a country will fall, but not by that

much.   Poor people's money is now worth much more.    But the

inflation-adjusted "average income in the poorest fifth" doesn't change.    The

economist will falsely conclude that the poor aren't much better off.    And

that subsidising food doesn't work.    Making basic goods more expensive for the

poor will be falsely taken by the economist to show no great adverse effect on

them.   That's because they still spend the same amount of (inflation-adjusted)

money, but the economist hasn't realised that inflation for the poor was worse.   


Of course the real world is more complex:   many poor people produce their own

food, the price affects their income, etc.   But it shows the principle.   The

inflation level is affected (kept constant, or changed) by things that have

nothing to do with poor people.   This is basically the same as the problem

publicised by Reddy and Pogge about PPP, but the point is that it applies to all

national cost-of-living indices, not just to international ones.    Adam Smith

wrote about this in the 1770s.    It's another thing that's obvious once we

think about it.     Charge poor people more for water, and their consumption

expenditure is the same!    They haven't lost out!     Economists make this

stuff out to be very complicated and difficult, but that sort of thing is

clearly barking mad. 


I think that just about completes my understanding of the flaws of "Growth is

good for the poor".    I can now say much more than anyone else has about why

its conclusions are invalid.   By the way, the authors seem to have dropped

completely from the latest draft their previous claim (which I seem to remember

was quoted in the Economist) that there was a proportional relationship between

poorest-quintile average rises and rises in average income in poor countries. 

 I'll have another look to check.    That paper, which is full of mistakes, is

used as justification for the World Bank claiming in its training programmes

that growth is a vital component of "poverty reduction".      As I say in

various writings, the term "poverty reduction" is often meaningless.  Its use as

a policy instrument may do more harm than good to poor people.     


I agree completely with your point about the World Bank winning arguments

upstream.     They win them with journalists, with the experts disagreeing in

private.    Meghnad Desai of the LSE told me ""Growth is good for the poor" is

crap".    Frances Stewart, director of development studies in Oxford, told me

"Growth isn't good for the poor...... yes, I think demography has a lot to do

with it [the statistical results]".   Justin Forsyth, head of policy at Oxfam,

told me Amartya Sen said one of the authors was a "third-rate economist".   

Matthew Sudders, DFID statistician, told me that he and his colleagues didn't

use the World Bank income data because the quality was so bad.     Robert

Johnston, Chief of Statistical Services at the UN, told me "growth isn't good

for the poor  -   look at your country, or mine [the USA]".    He indicated to

me that the World Bank were selecting the research results to give the

conclusion they wanted.    The nutrition data are also bad quality.    I asked

him what data were more reliable, and he said the child mortality data.    That

makes me more sure that the sane way to measure the progress of hungry people is

to see how long they last.