Choosing a period total fertility index using a simulation model

Daniel Devolder, Universitat Autònoma de Barcelona
Marta Serra, Universitat Autònoma de Barcelona

In the last fifteen years, there has been a flurry of proposals made for accurately measuring total period fertility, in order to improve the results given by the classical Total Fertility Rate (TFR). The main argument is that the TFR gives a biased value for the mean number of children women or men have, based on births that are observed in the current year. These proposals are either new ways to calculate a total period fertility level or procedures that intend to correct the biases existing total indexes have. The problem is that these additions to the demographer's toolset are not accompanied by the definition of criteria, which could allow one to choose the best fertility measure to use, depending on the kind of data available for each country. Our objective is to apply a macro demographic simulation model in order to help analysts to choose the best or the less imperfect period fertility measure, depending on the data at hand and on a study of the evolution in time of the structure of age fertility rates. The results of the simulation model will be comparisons of the fertility levels obtained with the battery of available period indexes or correction methods, with the 'true' level entered as an input, in various scenarios of changing tempo, from simple to more complex one.