Strategic Human Resource Management

Section Three

Markov Analysis

In the past, some researchers observed that companies tended
to have greater expertise and placed greater emphasis on
forecasting the demand for human resources than for their
supply. However, the application of Markov analysis to human
resource forecasting changed the situation by pro viding a
practical and versatile technique for forecasting internal supply.
As such, the techniques can serve the strategic purpose of
evaluating the availability of human resources required for
different strategies. Markov models have an advantage of being
relatively simple to understand, although they can be
quantitatively sophisticated.^59

In setting up Markov models (see Figure 3-1), the
forecaster must account for all possible moves or flows of
employees in an organization. Such moves include moves into
the organization, moves from one job to another, and exit
moves. Moves between jobs can be upward moves in
hierarchical level as well as moves across functions. Essentially,
Markov models begin with distributions of the number of
employees in various job categories at a starting point in time.
These distributions are then transformed by a transition
probability matrix into a forecasted distribution of employees
across these same job categories one period later. The
transition probabilities in each row of the matrix must total