Age profile estimation for family and fertility events: the MAPLE (method for age profile longitudinal estimation).

Focusing on fertility behaviors and transitions between various living arrangements, in this paper we deal with the methodological background of the age profiles estimation based on longitudinal survey data. The method, that starts with the specification of the required micro-level data, includes a data processing routine that leads to the estimation of regression models. These models permits, for each transition, to evaluate the smoothed age profile and the relative risks given by time-fixed and time-varying covariates. The age profile for the relevant transitions in the field of families and fertility, will represent data to be included as input in the MicMac model. The major advantage of the method developed here is flexibility: it can be applied to every setting where micro-level data on transitions are available from a large-scale representative survey (e.g., Fertility and Family Survey; Generations and Gender) and for different kind of transitions. Moreover, based on regression model the method permits to evaluate confidence intervals and to test hypotheses. The whole method is written in R software and it could be easily recalled as a R function in order to be applied to real data.