Many manpower planning models assume age to be normally distributed. I
n this paper we have examined the suitability of beta distribution for
age. Beta distribution has a few properties, such as finite positive
intercepts in the x-axis and adjustable skewness, that make it highly
desirable as a distribution for age in a manpower planning model. In t
his paper we have developed an approach for dynamically computing the
retirement probability, and hence the retirement rate, when the age of
manpower in a pool follows beta-distribution. We have made a comparat
ive evaluation of the two distributions in the framework of a simple m
anpower planning system dynamics model, and we have shown that various
manpower-related variables assume more reasonable values for the beta
-distributed age compared to the normal-distributed age. Thereafter we
have applied the approach to a two-stage hierarchical manpower planni
ng model in an R & D laboratory and have studied the behavior of the m
anpower-related variables under two recruitment policies. We have demo
nstrated in this paper that assuming beta-distributed age presents no
difficulty while modelling for time-scale promotion, a combination of
time-scale and merit promotion, and recruitment of persons at differen
t age groups. (C) 1998 Elsevier Science Inc. All rights reserved.