An Analysis of Ill‐Posed Production Problems Using Maximum Entropy

Abstract Production economics problems are often ill‐posed. This means that the number of parameters to be estimated is greater than the number of observations. In this article we show how to recover flexible cost functions from very limited data sets using a maximum entropy approach. We also argue that there exists a continuum of analysis between mathematical programming and traditional econometric techniques which is based solely upon the available information. The limiting case of a multi‐output cost function recovered using only a single observation of a farmer's allocation decisions can be easily extended to handle more than one observation.