In 1955 Kuznets constructed some fictional examples showing that among the reasons why measured inequality could initially rise in developing countries, was that workers migrating owards higher incomes could not possibly move all at once. This is not so disturbing: in fact, Fields suggests using a different measure, showing a dicrease in inequality in this case. Finally, the association between initial inequaklity and subsequent development depends crucially on the choice of countries included in the sample.