Parametric models of fertility (cont.)

Problems with the Brass fertility model, and extensions

The Brass polynomial, although having the advantage of being mathematically simple does not quite capture the full range of fertility variation observed in human populations - in particular, when it fits the build up at young ages, it may not portray the decline at older ages accurately, and vice-versa. Another weakness of this model is that the single-year detail for the initial year or two is not a good representation of fertility, as it tends to increase too rapidly in the early ages.

This polynomial subsequently has been generalised by Gage (2001), who allowed the width of the fertile period – constrained to be a constant 33 years by Brass – to vary, and hence be replaced by a variable, w, transforming the Brass fertility model into a three-parameter model.