This paper seeks to examine the use of the cohort component method as a method of deriving replacement demand for manpower forecasting.
The paper explains the principal concepts of replacement demand and how replacement demand has been estimated in two alternative ways: the cohort component method and using longitudinal data on individuals. The paper focuses on one of these ways, the cohort component method, and illustrates how this method can fail to capture all the relevant flows driving replacement demand. It also compares the method to the alternative approach based on individual data and discusses US and Irish results using both methods.
The cohort component method is found to underestimate replacement demand significantly in many occupations.
Research estimates of replacement demand should be based on individual longitudinal data rather than the cohort component method.
Many countries undertake some form of occupational employment forecasting including, in many cases, making estimates of replacement demand. This paper should help to clarify the appropriate choice of methodology for estimating replacement demand.
