Matt Berkley To: charlotte.denny@guardian.co.uk G8 and poverty Dear Ms Denny I have just sent a letter to the Guardian in response to your article on the G8. In case it is not printed, I thought that you might like to have a copy. These are matters on which there is far more detail available. I have references for my sources, and suggestions for more methodologically sound monitoring of Millennium Goals. My own training is not in economics, but in philosophy, history, and experimental psychology. I have read hundreds of academic papers on poverty and official papers. Some senior economists are taking notice of what I say. Prior to sending the letter I sent a short document discussing theoretical issues in welfare economics to Larry Elliott. I think there may be a news story here which is not of minor interest. Yours sincerely, Matt Berkley 33 Howard St Oxford .................................................................................................................................................. Sir, Your story “UN fears Iraq will dominate summit” (May 30), states: ”Rapid economic growth in China has lifted 150m out of poverty over the past decade, and the UN believes the overall target of halving the share of the world's population living on less than $1 a day will now be met by 2015”. Here are some fundamental problems with those statistics. 1. The goal of “halving poverty” is fundamentally flawed. The “share of the world’s population” in poverty does not only depend on the number rising out of poverty. It also depends on how long poor people live. If they live longer, the proportion falls more slowly - at least until they start trusting that their babies will survive, and so reducing the birth rate. In countries where poor people die earlier, the proportion falls faster. In the age of AIDS, this is more than a theoretical possibility. The target is therefore conceptually flawed. It also fails to take any account of the benefit of longer life in itself, or the cost to a person of dying. All of the UN global statistics on poverty come from the World Bank. No statements from the UN on the number of people who have “risen out of poverty” are reliable, since they have not taken survival rates into account. This is despite the facts that: a) the man in charge of the World Bank’s global poverty counts, Martin Ravallion, wrote of the life-length flaw in 1996; b) Professor Angus Deaton of Princeton warned the Bank’s Chief Economist of the flaw in 2000, and c) I warned the head of the World Development Indicators project at the World Bank, Eric Swanson, of the flaw in 2001. I also warned one of Clare Short’s senior economists of the flaw in 2000, and herself and several of her staff in 2001 and 2002. 2. These statistics have not taken into account the facts that a) more of today’s population are adults, and b) adults need more food than children. A dollar line for both adults and children makes no sense. Where birth rates fall, this reduces total food requirement but increases average food requirement. FAO statistics on hunger make some allowance for this, and are superior to the World Bank statistics in this regard. The World Bank method overestimates the number of people “rising out of poverty” - according to their own definition of poverty, which is inadequate daily consumption. To treat adults as needing the same size of meal as children is not a competent way to measure poverty. Martin Ravallion made that point himself in an academic paper in 1995. 3. Economists do not use, or have, price data on poor people’s goods (see “Poor but pedicured”, May 6). To look at income is to look at less than half of the equation for economic welfare. The number of people under a dollar line tells you nothing about whether they could buy more or less food with their money. All statements that x number of poor people “rose out of poverty” on the basis of income data with no price data are irrational. The same applies to all large-scale studies of “poverty” by economists. It also applies to their claims about the progress of poor people based on trends in “income inequality”. The reality is simply this: you cannot tell whether poor people did better or worse by counting their money, unless you know about prices. Adam Smith knew this in 1776. In 1996 Martin Ravallion wrote that “best practice” in poverty research includes price data. The inflation flaw, like the life-length flaw and the children flaw, severely damages the credibility of economists’ claims as to outcomes for poor people. The trend in average income is not the average percentage gain: the figure for average income also depends on demographic change. So does the figure for annual “income inequality”. If poor people live longer, inequality of annual income will be higher among the people still alive. That is a better result for them, not as an economist might think, a worse result. 4. These statistics are based on household surveys. Statisticians and many economists are aware that these are grossly unreliable. One problem is that the data are very sparse. That is not surprising - these are poor countries, with few resources for consistent and thorough data collection. Another problem is that different surveys use different survey questions. Another is that surveys use varying methods for setting a money value on the food which people eat. The unreliability has been shown by, among others, the economists Tony Atkinson of Oxford, and James Galbraith of Texas. Angus Deaton has also warned the IMF and the World Bank of the reliability problem. Martin Ravallion has himself said that rich people and very poor people do not cooperate with surveys. This makes the survey data even more unreliable. Why then have the World Bank published statements on the basis of methods which this man - described to me by the head of the World Indicators Project as their “poverty guru” - has himself described as defective in several major respects? More to the point, why should you or I take any notice whatsoever of these claims about the “number who rose out of poverty” in China, or “the number on less than $1 a day”? Without prices, these figures signify nothing. Even with information on life length, statements about poor people doing better or worse would depend on how much you value a year of life. The present Labour government made a claim in 2001 as to the number of children in the UK which it had “raised out of poverty”. Its figure was based on a hypothetical estimate of the number who would have been in poverty under the Tories, not the actual number. Journalists might take a more sceptical view of statements from governments about the welfare of people many of whom cannot even read, let alone protest. To escape from extreme poverty is to increase your food consumption to the level where your chances of survival become tolerable. Extreme poverty is vulnerability. Poverty is not low consumption. The relationship between low consumption and poverty depends on the number of extra items needed at a particular time and place. Poverty is not low income, which in addition depends on prices. For economists to adjust for all these factors would be massively wasteful. If they achieved it somehow, then what would they have achieved? They would then have measured food consumption, duplicating the work of the FAO. There is no need whatsoever for an extra category of “extremely poor” including people who are hungry and people who are not. The World Bank statistics on “poverty” are at best a very expensive waste of taxpayers’ money, and at worst a grossly misleading basis on which to devise policies. The cheap, easy and reliable way to estimate the progress of hungry people is to measure how long they live. You can judge governments’ seriousness in helping poor people by the attention which they give to improving that. Matt Berkley Oxford