Background: The diffusion of cohabitation and, presumably, of childbearing within cohabitation, inspires interest in measuring the respective contribution of childbearing within marriage and within cohabitation to overall fertility. However, there is no consensus on a proper way to do so.

Objective: Contribute to the development of tools for assessing the relative importance of marriage and cohabitation to overall fertility by developing period measures closely related to age-specific fertility rates and the total fertility rate.

Methods: We introduce two measures: 1) the contribution of the conjugal state (living alone, living in a cohabiting union, being married) to age-specific fertility rates (CASFR) and 2) the contribution of the conjugal state to the TFR (CTFR). These measures are similar in construction to the marital (legitimate) fertility rates and marital (legitimate) TFR, but they are weighted by the proportion of women living alone, cohabiting, or being married at each age, so that their sum is the overall TFR. Taken together, they represent the fertility of the average woman of a synthetic cohort who moves across the various conjugal states (living alone, cohabiting, being married) over her life course. They provide "realistic" estimates of completed fertility within each conjugal state.

Conclusions: CASFRs provide a description of the fertility, over her life course, of a synthetic woman who would have spent her reproductive years living alone, cohabiting, and being married as the average woman of the synthetic cohort. CTFR provides a decomposition of the cumulative fertility of this synthetic woman. Over her life course, she would have had exactly the number of children computed using the overall TFR, but CTFR details the proportion of these children she would have had while living alone, while cohabiting, and while being married.

Comments: Despite being defined as a conditional ASR weighted by the age-specific proportion of women living alone, cohabiting, or being married, computing the CASFR does not require that one know the population distribution of women by conjugal status, as this quantity cancels out. As a consequence, CASFR and CTFR may be computed without this information, as the parity-specific ASFR may be computed without knowing the population distribution of women by parity.