A Mathematical Model for the Growing Sector of the Poultry Industry in England and Wales

This paper deals with the problem of forecasting the population of growing pullets in England and Wales. The term Growing Pullet is used for a hen from birth until it is moved to the laying flock. Quarterly data is available on the number of growing pullets alive, while data is published every month on the number of hens hatched.Previous work in the field of forecasting agricultural populations has been based on Cobweb Theorem considerations and relies on past prices and expectations of future prices. However, given the available data and non-price relationships that take place in the growing sector, it is possible to build a non-price model.The approach followed is to start with a naive regression model estimated using ordinary least squares. The model does not conform to prior expectations of regression coefficients and has to be re-estimated using Quadratic Programming. This provides new ideas for a further model that incorporates adaptive expectations and that is updated using a Box and Jenkins time series approach.