Definition
Late fertility refers to births to women aged 40 and above; latest-late fertility refers to births to women aged 45 and above. These categories mark the biological and behavioral frontier of human reproductive timing, where physiological constraints on fecundability are severe and where assisted reproductive technology (ART) offers only limited compensatory capacity. The distinction between very late (40+) and latest-late (45+) fertility is analytically useful because the physiological wall at approximately ages 44–45 constitutes a near-absolute barrier that has remained stable across decades despite social change and technological advance.
Key Ideas
- U-shape in Sweden 1900–2001: Late fertility peaked at ≈115.5 births per 1,000 women aged 40+ around 1900, fell to a trough of 8.6 per 1,000 in 1977 (linked to lowest-low fertility and broad postponement), and partially recovered to 26.8 per 1,000 by 2001 — still far below historical levels.
- Behavioral wall at 44–45: First-birth occurrence-exposure rates (rates among childless women) are essentially flat above age 44 across 30 birth cohorts (1925–1955). This wall reflects physiological constraints rather than absence of demand, since demand for late births appears to have increased substantially over this period.
- Absolute probabilities are tiny: The probability of a first birth between ages 40 and 45 for a childless woman was 7.48% in Sweden in 2002 (up from 3.26% in 1970); at age 45+, only 0.23%. These figures include ART effects.
- ART's limited compensatory role: In vitro fertilization (IVF) with own eggs yields <5% live birth rate per cycle in the early 40s, adding only ≈7 percentage points of cumulative first-birth probability at age 40 (Leridon 2004). Egg donation maintains 40–50% success rates, demonstrating that egg quality — not uterine capacity — is the primary biological barrier.
- Physiological constraints are stable: Female sterility reaches 24–46% by age 40 (varying by source). Fecundability (the monthly probability of conception with unprotected intercourse) declines from the late 20s and is substantially reduced by the late 30s, with near-zero rates by the mid-40s. These constraints have not changed despite social transformations.
- Compositional factors in the raw recovery: The aggregate increase in 40–44 fertility is partly attributable to a parity shift (the share of first and second births among all 40–44 births rose from ≈37% to >50% in Sweden) and changes in the age structure of women 15–49, not solely behavioral postponement.
- Rectangularization evidence is weak: The interquartile range (IQR) of the first-birth age schedule (a rectangularization indicator) has been stable at ≈6 years in the period series; the cohort IQR peaked at 7.5 years for the 1956 cohort and declined only 10.5% by the 1961 cohort — insufficient to conclude rectangularization is occurring.
How It Works
Occurrence-exposure rates (rates of the first kind) measure first-birth behavior without parity contamination: the denominator is childless women rather than all women at a given age, eliminating the mechanical effect of parity-composition shifts. This methodology allows direct observation of whether individual childless women are increasingly bearing children at ages 40–45 across cohorts, independent of the changing parity structure of the older-age female population.
The demographic mechanism behind the recovery in aggregate late births involves three channels: (1) genuine behavioral postponement among women who remain childless into their 40s and then attempt a first birth; (2) compositional shifts in the parity structure of older mothers; and (3) cohort size and age-structure effects. The occurrence-exposure analysis isolates channel (1).
Why It Matters
Late fertility trends test the limits of the broader fertility postponement phenomenon associated with the Second Demographic Transition. If postponement continues without compensatory late recovery, total cohort fertility falls below replacement — contributing to lowest-low fertility. The behavioral wall at 44–45 suggests that postponement has a biological stopping point regardless of social or technological change. The limited effectiveness of ART at late ages (with own eggs) challenges policy assumptions that technology can fully offset postponement-induced fertility declines. The distinction between ART with own eggs (bounded by egg quality) and ART with egg donation (bypasses the wall) is central to assessing future fertility ceilings.
Open Questions
- How much of the behavioral increase at ages 40–44 reflects genuine deliberate postponement vs. residual fertility among women who did not actively intend to postpone?
- Will the 44–45 behavioral wall shift as ART with egg donation becomes more accessible and socially normalized, bypassing the egg-quality constraint?
- Does the Swedish pattern generalize to lowest-low fertility countries in Southern and Eastern Europe, where the late-fertility recovery has not occurred?
- Is the cohort IQR trajectory (10.5% decline over 5 cohorts) evidence of incipient rectangularization, or statistical noise given the short observation window?
Related
Sources