Manpower planning models and their estimation

Manpower planning models are an essential tool for the modern manager. Such models may describe the constituent components of the manpower system using, for example, survival analysis techniques similar to those found in the medical literature. However, it is often more useful to look at the manpower system as a whole using a variety of methods such as the Markov, renewal theory and semi-Markov models. Estimation of the parameters of such models, for sampling frames appropriate to manpower data, forms an important part of the methodology. The paper concentrates on the computational aspects of the problem. In particular, it focuses on the use of a Markov-chain formulation which implements the mathematically intractable semi-Markov approach by means of a non-parametric estimation procedure.