Subj: Global poverty Date: 01/02/03 To: carolinelucas@greenmeps.org.uk Dear Ms Lucas I am a constituent of yours who is writing a book on the measurement of economic welfare. In particular, I analyse the major World Bank documents claiming success for their policies in helping poor people. I have done this more comprehensively than any economist, while drawing on observations by individual economists on the methods. I have consulted something over 1000 policy, research and theoretical documents, and dozens of experts. In the process, I have come to realise that there are several fundamental assumptions commonly made in economics which no-one seems to have pointed out. Here is one: if people on below-mean income (that's most of us) live longer, other things being equal, GNP per capita will fall. Another is: economists treat the needs of children the same as those of adults, in their conclusions about welfare gains and losses. But the ratio of children to adults varies across countries, social classes and times. A change in the average does not tell us the average gain or loss to individuals. Economists say that they are concerned with aggregate social welfare. But in reality it is not the total they look at , but the average. There are also various technical problems with the World Bank's analyses. For instance, the "dollar-a-day" line is not a poverty line, but a consumption line. Standard "best practice" at the World Bank is to count people living in families where per capita consumption is less than, in theory, what a dollar a day could buy in the USA (not, by the way, a real dollar's worth in a poor country but several times less). So a family with more children is automatically counted as "poorer" even if they are not. As time goes by and people have fewer children, the proportion of people under this line may fall. But the point is that adults need to eat more. Therefore, the fall may mean that some poor people rose above a minimum standard of living, or it may not. Also, Amartya Sen has pointed out that the fall in the proportion does not tell us that the poor on average became better off, because the majority may have been further impoverished even if some cross the line upwards. A clever government will help only those nearest the line. In any case, the proportion will fall faster if poor people die earlier. The Millennium Goal of "halving poverty" ignores this. The World Bank has failed to measure the changing depth of poverty, and so we do not know whether those under the consumption line are poorer or better off than at the start of the period of the Goals in 1990. There is something very disturbing about the World Bank's promotion of a document claiming "Growth is good for the poor". Two heads of department in development studies at the LSE have told me that they don't like it; Meghnad Desai told me it was "crap". The head of development studies at Oxford told me "Growth isn't good for the poor....yes, I think demography has a lot to do with [the statistical results]". The head of policy at Oxfam told me that Amartya Sen described one of the authors as a "third-rate economist". Robert Johnston, the Chief of Statistical Services at the UN said to me "Growth isn't good for the poor - look at your country, or mine [the US]". He also questioned the fact that the World Bank do not release research showing growth fails to reduce poverty. I asked him what he meant, and he said that he thought they wanted to show a particular result. I asked the head of the World Development Indicators project at the World Bank about the problem that the figures would look "better" somewhere like Uganda if poor people died of AIDS. He said that after the plague the people were better off because the density of people on the land was less; that he was looking at the macro level and the long term; and that "we could find out how many people died [rather than becoming better off] but I don't know if it would be a good idea to say it". When I told some of this to Robert Johnston at the UN, he laughed and said "Larry Summers tried to do that with AIDS when he was at the Bank". The technical problems involved in estimating consumption levels among poor people using money are vast: you have to collect data from many regions, cities, towns, villages; standardise the questions globally; and then (something which the World Bank does not do) find out what prices were paid by how many poor people in each region of a country; and then try to convert the monetary values internationally even though people eat different food. But there is a further problem. The notion that the "income share" of the poorest 20% tells you about changing consumption levels is a fallacy, because the prices paid by the poor may have risen or fallen relative to the rate of inflation for the whole economy. I seem to be the first person to point this out. Consumption expenditure - which is measured in "income" studies of poor people - is of course only one component of economic welfare. Environmentalists point out correctly that this excludes both natural resources and non-monetary economic activities. But it also excludes changes in land ownership, and even savings made by poor people in better times. The list of fallacious inferences from the World Bank - and even from its critics, who make the same assumptions - is quite long. I have just sent Oxfam a list of some of them. The general failures of "welfare economics" to a) measure average outcome, rather than the change in average income in the geographical area which is subject to significant influence from demographic change; b) distinguish between total welfare and average welfare c) distinguish between declared income gains and real economic gains including inheritance, gifts, assets, debt reduction, undeclared income and so on d) recognise that adults need to eat more than children are in addition to its failure to take into account other factors mentioned by environmentalists. Greater life length among the poor slows statistical rises in average income - whether or not anyone else's income goes down. There is a purely mathematical effect. Especially in the age of AIDS, and in any age where malnutrition affects many people, GNP per capita does not show average outcome. If you would like to know more, please let me know. I can publish these things (and others) as academic papers, but in reality these issues are beyond the competence and knowledge of economists. Yours sincerely