Mean Age at Childbearing

The period mean age is computed from the age-specific fertility rates observed in a single calendar year, so it summarizes the timing of childbearing across all reproductive ages in that year, drawing on many different birth cohorts of women. The cohort mean age follows one real birth cohort of women through their entire reproductive lives and averages the ages at which they actually had children. During periods of changing timing the two can diverge substantially: if childbearing is being postponed, the period mean can lag or move erratically while the cohort mean climbs smoothly, which is why the distinction matters for interpreting trends.

When fertility rates are reported in five-year groups, the true average age of mothers within, say, the 25-29 group is approximately the midpoint 27.5, not the lower bound 25. Weighting by lower bounds would systematically place every birth at the youngest edge of its interval, dragging the computed mean downward by roughly half the interval width — about 2.5 years for five-year groups. Using midpoints removes most of this bias; finer single-year data removes it almost entirely.

When the mean age at childbearing rises — childbearing is postponed — births that would have occurred in a given year are pushed into later years, temporarily depressing the period total fertility rate even if completed cohort fertility is unchanged. Conversely a falling mean age inflates the period TFR. Bongaarts and Feeney used the year-on-year change in the mean age as the core quantity in their tempo adjustment, which removes this timing distortion to recover a quantum measure of period fertility.