The variance of direct expansions from a common area sampling design

Many countries use a sample design in their agricultural surveys that is assumed, incorrectly, to be equivalent to stratified simple random sampling without replacement. This article discusses how the variance of a direct expansion based on that design should be estimated.