Patterns and determinants of fertility in Melbourne

Abstract

Period fertility has fluctuated markedly in Australia over the last fifty years. The last peak occurred in 1971, and in that year the Department of Demography of the Australian national University carried out in Melbourne the first major fertility survey ever conducted in Australia. This research is based on data from the Melbourne Family Survey, which has proved to be an excellent vehicle for the analysis of many aspects of Australian family formation. The thesis first establishes the physiological setting in which the childbearing of the Melbourne respondents took place, and then goes on to discuss patterns of contraceptive use both within marriage cohorts and within different social groups. This is followed by analyses of cohort fertility patterns and the timing of childbearing. The effect on current fertility of age at marriage is examined both through cumulative marital fertility rates and through multivariate analyses, the latter simultaneously providing indications of social characteristics that influence current family size. A simple model based on the birth intervals of non-contraceptors is next developed, and applied to examine the effects of both volitional and non-volitional fecundity depressants on reproductive performance; contraceptive effectiveness; and parity-specific patterns of family formation. While average completed family size was found to vary little between marriage cohorts and, within marriage cohorts, to vary litter between women with different social characteristics, the tempo of childbearing was found to have been far from stable in the recent past. Changes in family building patterns can be summarized as a compression of the childbearing span of women married in the 1950s and, among women married after 1960, a return to the longer spacing patterns which characterized women who married during the Depression and the Second World War. Such changes have been largely responsible for fluctuations in period marital fertility rates in the recent past: only if the tendency to postpone childbearing is accompanied by some dramatic attitudinal changes (of which we have no evidence) can we expect the average completed family size of the most recently married women in the sample to decline significantly.