Italy’s Fertility Collapse Is No Longer About Delay
For decades, Italy's extremely low fertility has been explained away as postponement. Women were having children later and later; the mean age at childbirth kept rising; and because of that mechanical shift in timing, the period total fertility rate (TFR) appeared to collapse. Yet completed family size, measured by cohort fertility, did not fall as dramatically. The implication was reassuring: much of the decline was a distortion created by delay rather than a fundamental change in how many children women ultimately had.
That interpretation held for the 1980s and 1990s. It no longer does.
Using official Italian data downloaded directly from demo.istat.it, including period fertility by citizenship, cohort fertility, and mean age at childbirth, we can disentangle timing effects from genuine reductions in family size. Once that distinction is made, the recent decline looks fundamentally different from earlier episodes. The drop since 2010 is not primarily about births being shifted to later ages. It reflects a deeper structural change in completed fertility itself.
Period versus cohort: what actually changed?
There are two conceptually different ways to think about fertility. Period TFR asks how many children a woman would have if she experienced the age-specific birth rates of a given year throughout her life. Cohort fertility instead measures how many children women born in a particular year actually end up having by the end of their reproductive lives.
In Italy, cohort fertility declined gradually from around 2.3 children for women born in the early 1930s to roughly 1.4 for those born in the mid-1970s. That long decline was driven overwhelmingly by the disappearance of third and higher-order births. Italy did not suddenly become a society of childlessness; rather, it became a society in which large families ceased to be the norm.
To understand what actually changed inside cohorts, it helps to break total fertility down by birth order. When we do that, the mechanism of Italy’s long-run decline becomes much clearer. The early transition was not driven primarily by rising childlessness. It was driven by the disappearance of third and higher-order births. Large families collapsed first. First births declined only modestly, and second births eroded gradually. The shift from a three-child norm to a one- or two-child norm explains most of the structural decline in completed fertility.
Period TFR, however, fell more abruptly in the late 1970s and 1980s. The reason was postponement. As women delayed first births and the mean age at childbirth rose, the annual TFR was temporarily depressed. Once those postponed births occurred at older ages, part of the decline was recuperated in cohort terms. That was the classic “tempo distortion.”
The central question today is whether the recent fall from 1.44 in 2010 to 1.18 in 2024 can be explained in the same way.
Correcting for postponement
To answer that, I applied the standard Bongaarts–Feeney adjustment, which corrects period TFR for annual changes in the mean age at childbirth. If postponement is driving the decline, the tempo-adjusted TFR should fall less than the observed TFR. Removing timing effects should shrink the apparent drop.
It does not.
Between 2010 and 2024, observed TFR fell by about 0.26 points. After correcting for changes in the mean age at childbirth, the decline is slightly larger rather than smaller. This is the opposite of what we would expect if delay were the main mechanism.
In other words, the recent drop cannot be attributed primarily to births being shifted into the future. It reflects fewer births occurring at all.
This is a departure from the 1980s and 1990s, when postponement explained a substantial portion of the collapse.
Italian and foreign mothers: convergence downward
Aggregate fertility in Italy combines two distinct groups: births to Italian mothers and births to foreign mothers. For years, immigration partly masked the depth of native fertility decline.
In 2010, TFR among Italian mothers was roughly 1.33, while among foreign mothers it was about 2.31. By 2024, those figures had fallen to around 1.11 and 1.81, respectively.
Foreign mothers still have higher fertility, but the downward trend is clear for both groups. When the same tempo adjustment is applied separately to Italian and foreign mothers, the conclusion remains the same: the post-2010 decline is not mainly a timing effect. It represents a genuine reduction in fertility levels.
Moreover, the fertility of foreign mothers is converging downward toward Italian norms. Immigration is no longer providing the demographic offset it once did.
What will the 1990–2000 cohorts complete at?
To understand the longer-term implications, I reconstructed synthetic age-specific fertility schedules using observed TFR and mean age at childbirth by citizenship. From these, I estimated implied completed fertility for cohorts born between 1990 and 2000 under two scenarios: one in which fertility stabilizes at 2024 levels, and one in which recent trends continue.
For Italian mothers, the implied completed fertility is approximately:
1990 cohort: 1.23–1.25
1995 cohort: around 1.14–1.15
2000 cohort: roughly 1.02–1.11
If recent trends persist, the 2000 cohort could approach one child per woman.
For foreign mothers, implied completed fertility remains higher but is also declining:
1990 cohort: about 2.2
2000 cohort: around 1.8
While these are synthetic projections rather than observed completed fertility, the trajectory is difficult to ignore. Younger cohorts do not appear to be heading back toward 1.4 or 1.5. They are drifting lower.
A structural regime shift
Italy's fertility decline has moved through distinct phases. The first stage involved the collapse of large families, as third births disappeared. The second stage was characterized by widespread postponement, which amplified the period decline without fully collapsing cohort fertility. The third stage, now emerging, is defined by weaker entry into parenthood and reduced progression to second births. In this phase, fertility declines are no longer primarily about timing; they reflect a lower completed family size.
That pattern is characteristic of countries that have stabilized at fertility levels around 1.0–1.2.
If cohorts born around 2000 ultimately complete at close to one child per woman, Italy will not merely be experiencing a cyclical trough. It will have moved into a new demographic equilibrium.
The earlier reassurance that births were simply coming later no longer fits the data.
The decline is real.
I read that the decline, generally, is also explained by the fact that about 50% of women above the age of 30 who don’t have children at that age never end up having children. Do you think that is true for Italy? And what does that say about the average family size of those who actually end up having children?