Manpower Planning Models in use in the Civil Service Department
| DOI | https://doi.org/10.1108/eb055287 |
| Published date | 01 March 1975 |
| Date | 01 March 1975 |
| Pages | 23-35 |
| Author | A.F. Forbes,R.W. Morgan,J.A. Rowntree |
| Subject Matter | HR & organizational behaviour |
Manpower Planning Models
in
use
in the Civil Service Department
A. F. Forbes, Institute of Manpower Studies
R. W. Morgan, University of Cambridge
and J. A. Rowntree, Civil Service Department, London
Abstract
This paper describes the mathematical models used by the
Civil Service Department for manpower planning supply
work, their inter-relationship, and the problems to which
they are suited. Mathematical detail is normally not
included, the reader being referred to other publications.
Introduction
Civil Servants have a great variety of specializations and
qualifications and the Civil Service has the objective of
providing a career which matches the ability of each one
while at the same time filling every vacancy with a suitably
experienced person. To help plan this extremely complex
system the statisticians of the Civil Service Department
have taken an active interest in creating and developing
mathematical descriptions of the system and have been
helped in this task by research groups at the Universities of
Kent and Cambridge.
For management purposes the Civil Service is divided into a
number of subgroups, each having a hierarchical grade
structure. As a typical example of such a subgroup we shall
use the 'Statistical Research Group' which possesses many
of the features which are often to be found in a Civil
Service hierarchy. The Statistical Research Group (which
we shall subsequently call SR) is fictional but based on a
real subgroup. Data from SR is of two types.
'Stock' data tell us for a particular point in time the precise
numbers of people in various categories. We usually divide
up the employees by age and by grade and we show this
breakdown for SR in Figure 1 using five year age-groups.
'Flow' data on recruitment, retirement, promotion and
wastage are available to us only in the form of information
on past flows. An important part of our work concerns the
problem of how best to use such past data to estimate flows
in the future. We shall not deal with this problem in this
paper, preferring to concentrate on the models themselves.
However, in what follows the reader should remember that
in the form in which it is input to the model, flow
information consists of estimates and not of facts.
We identify three motives for the creation of our models.
The first motive is insight. Without necessarily being con-
cerned with detail we want to be able to perceive quickly
how the system works, its current state, and where the
potential problems lie. Our second objective, which requires
a more detailed approach, is to be able to forecast what the
future holds under a variety of different assumptions.
Age
60-64
55-59
50-54
45-49
40-44
35-39
30-34
25-29
20-24
Total
Assistant
31
37
68
Main
3
12
15
18
12
33
27
13
133
Chief
7
17
13
8
3
48
Director
2
9
9
20
Total
12
38
37
26
15
33
27
44
37
269
Figure 1: Statistical Research Group (SR): grade-age distribution of
current stocks.
Thirdly we want our models to help in decision-making.
Given that a particular future is desirable what action
concerning recruitment, retirement and promotion should
be taken in the meantime? We shall describe four
approaches each of which is useful in some measure for
meeting these objectives. These are Stationary Population
Models, Markov Models, Renewal Models, and the 'Camel'
Model. The particular model chosen will depend on the
question of most immediate interest. For example the
Stationary Population Models are good for obtaining insight
To continue reading
Request your trial