Measuring Fertility Associated With Ex-nuptial Conceptions: A Further Response to Betts

Abstract

This paper is part of an ongoing debate over the proper definition and of ex-nuptial conceptions prompted by publication in volume 17 of the same journal of Ex-nuptially conceived births: a note on measurement by K. Betts. The 2 main objections to the original paper are still valid: 1) the proposal to combine ex-nuptial and bridal birth rates breaks an elementary rule for the construction of demographic rates namely that the denominator should consist of persons at risk of experiencing the events which comprise the numerator. Unmarried women cannot have bridal births. Carmichaels proposed ex-nuptial conception rate divisible into rates of ex-nuptial conception resulting in ex-nuptial and nuptial confinements is superior on this count because it allows ex-nuptial conceptions to be related to a risk population of unmarried women. 2) The ex-nuptial fertility rate is subject to biases which make its comparison across maternal age groups problematic. Age-specific ex-nuptial birth rates appear to have considerably more value for studying behavioral trends within age groups than between age groups because ex-nuptial births can occur to women who are legally married and because of the tendency of women living in de facto unions to declare themselves as or be reclassified in the census as married. Both circumstances mean that the denominator of ex-nuptial fertility rates excludes some contributors and potential contributors to the numerator thus biasing the index upward. The upward bias is more severe for age groups in which women living in de facto unions form a high proportion of those not living in marital unions. It is clear for New Zealand at least that the proportion of ex-nuptial children born into de facto marriages in recent years has been extremely variable both between maternal age groups and through time.