From: Matt Berkley Sent: 21 June 2003 19:34 To: king@harvard.edu Subject: Poverty Dear Gary I thought you might be interested in a couple of recent events in what I am doing. Ravi Kanbur of Cornell has published a paper on premature mortality and poverty measurement; this paper is now being considered by the UN Statistical Division. It's pretty much the same as what I told him in 2001 about what I had worked out. Also, I seem now to have a more comprehensive overview of the assumptions in economists' measurements of poverty (and welfare) than anyone else has explained. Something which economists shy away from talking about is the fact that they don't take the cost of living into account in their studies of ""poverty"". They use the overall inflation rate in a country. I say more about this in the letter below. Having now understood this missing link - and taken in the implications - I now feel that I know what I'm talking about. The letter is a short summary of what I say about development/welfare economics, leaving out two major points: 1) the economist's definition of utility (a person's level of welfare at one time, aggregated using cross-sectional statistics) is not the same as the philosopher's (greatest good for the greatest number, which can only be a longitudinal statistic). The reason is that the economist's leaves out the duration of happiness/welfare/utility. Bentham left this in. The consequence is really that all measures of ""income inequality"" underestimate it. Poor people live less time, so that the distribution to people during the period is more unequal than the annual figures suggest. The distribution to people is a longitudinal statistic. It is not knowable from population averages.