Section ThreeForecasters combine Markov analysis with demand scenarios,
by beginning with a specification of the desired future
distribution of employees in various job categories, typically in
higher-level positions. By working backward, the forecaster
then determines the magnitude of the transition probabilities
that will be needed to create the flow of employees from the
existing distribution into the desired future distribution.
Promotion rates and termination rates are examples of human
resource policies that would be adjusted to obtain the desired
flow of human resources through the company to achieve the
desired future distribution of employees. Transition rates also
can be changed to reflect changes in such policies. By running
the models with different transition rates, corresponding to
policy changes, the impact on human resource flows between
jobs and hierarchical levels can be determined.^66
Another example of an application of Markov analysis is
provided by the experiences of the Weyerhaeuser Company. In
this example, the model was first developed during a growth
phase of the company. The model was used to forecast the
number of employees in a specialty, on a corporate-wide basis,
that would be available from the company’s internal labor
supply. In the event that internal supplies were forecasted to
fall short of demand, accelerated programs were instituted to
provide the training needed to qualify current employees to
take on the responsibilities for the positions in which shortages