The objective of the paper is to construct a general framework which will help to increase the usefulness and reduce the cost of collecting data for public decision‐making about development in low‐income countries. Data required about the rural sector are usually considerable and relatively costly to obtain; this area receives particular attention. The search for useful principles proceeds from the economics of information via Karl Popper's principle of error reduction and the use of information cybernetics in public decision‐making to the design of more cost‐effective models of development processes and the significance of alternative hierarchical administrative structures for the utility obtained from primary data. These components are combined into a unified logical framework. An integrated approach to management information is identified as a desirable adjunct for its application in practice.