Vegard Skirbekk and Thomas Spoorenberg* advocate the combined use of fertility and mortality information, or net reproduction, when discussing fertility levels and trends, both in developing and in developed countries.
For decades, demographers have relied primarily on the total fertility rate (TFR) to understand when and how fertility transitions occur across the globe. This measure, representing the average number of children a woman would bear under current fertility conditions, has become the gold standard for tracking demographic change. However, a critical component has been largely overlooked: mortality.
In a recent paper (Skirbekk and Spoorenberg 2025), we argue that this narrow focus on gross fertility measures provides an incomplete and potentially misleading picture of demographic change. Our comprehensive analysis of 236 countries from 1950 to 2023 reveals that when mortality is incorporated through net reproduction measures, our understanding of global fertility changes dramatically.
The missing piece: mortality’s role in reproduction
We advocate for the systematic use of the 2-Sex Net Reproductive Rate (2SNRR) (Keilman et al 2014) which accounts for both fertility and mortality by measuring how many children actually survive to reproductive age, and which can be approximated as TFR multiplied by the probability of survival to mean age of childbearing. Unlike the TFR, which assumes that all births contribute equally to future generations regardless of survival prospects, the 2SNRR reflects the biological reality that only children who reach adulthood can reproduce. This means that in high-mortality contexts, the effective reproductive contribution of births is substantially lower than raw birth numbers indicate.
The implications of this methodological shift are profound. Globally, fertility peaked in 1963 at 4.1 “effective” children per woman using the 2SNRR framework, compared to 5.3 using TFR measures. While both indicators show the same timing for the onset of decline at the global level, the magnitude of historical fertility levels appears significantly lower when mortality is considered.
These differences become even more pronounced at regional levels. In Middle Africa, TFR data suggest that fertility peaked in 1987 and has since declined, while 2SNRR measurements indicate that fertility has yet to decline at all in this region. Similar discrepancies appear across Southern Asia and parts of Northern and Eastern Africa, where 2SNRR-based fertility transitions occur later, peak at lower levels, and decline more gradually than TFR-based assessments suggest.
The laggard countries: where fertility transitions haven’t begun
Perhaps most striking is the identification of seven countries where, according to 2SNRR measures, fertility transitions have yet to begin as of 2023: Angola, Central African Republic, Chad, Democratic Republic of Congo, Mali, Mozambique, and Somalia. These countries, representing 3% of the world’s population, challenge the conventional wisdom that fertility decline is universal and inevitable (figure 1).
For these nations, declining mortality rates have resulted in stable or even increasing net reproduction, despite changes in gross fertility rates. In other words, improvements in survival have offset declines in birth rates, meaning these populations have not yet reached the point where reproduction begins to fall towards replacement levels.
Decomposing the effects: mortality versus fertility
In our scientific paper, through counterfactual analysis, we show how mortality and fertility improvements have contributed differently to changes in net reproduction since 1950. Globally, if fertility had remained constant at 1950 levels, declining mortality alone would have caused an increase in the 2SNRR by 1.23 births per woman through 2023 (figure 2). However, fertility declines have more than counterbalanced these mortality improvements, resulting in an overall decline in global 2SNRR from 3.36 to 2.13 births per woman.
The timing of when fertility becomes more important than mortality for reproductive outcomes varies significantly across regions. Western and Eastern Europe experienced this shift earlier due to their earlier demographic transitions, with TFR and 2SNRR converging by the early 1980s as mortality ceased to be a major factor.
By contrast, in Central Asia, Northern Africa, and Latin America, this convergence occurred more gradually through the late 20th and early 21st centuries. While TFR levels reached maximums of 7 or more children per woman in regions like Central America and Eastern Africa, 2SNRR peaked at levels only slightly above 5 children per woman. Middle Africa and Western Africa remain the only regions where mortality improvements still outweigh fertility effects in driving changes to net reproduction.
Policy implications
The choice between TFR and 2SNRR measures has significant implications for understanding population growth dynamics. Because 2SNRR-based fertility transitions typically occur later than TFR-based transitions, populations often grow larger before fertility begins declining. For example, taking 1950 as a starting point, Angola’s population increased by less than 1.5 times before reaching its TFR peak, but by more than 5 times when considering the 2SNRR peak. These differences have profound implications for resource planning, infrastructure development, and economic policy.
Our research also provides new perspectives on contemporary low fertility concerns in developed countries. Current fertility levels in high-income countries, typically ranging between 1.2 and 1.8 births per woman, represent only a 25% reduction below replacement fertility when viewed through the 2SNRR framework.
This suggests that current low fertility represents a less dramatic departure from historical norms than TFR-based analyses indicate. This perspective could help shift policy discussions away from “fatalistic concerns” about both high and low fertility, recognizing that current levels may not be as extreme as commonly perceived.
Even in low-mortality settings, mortality considerations remain important. Some developed countries experience excess premature adult mortality among specific population groups, and recent reversals in survival gains in countries like the United States demonstrate that mortality cannot be ignored even in affluent societies.
A call for methodological integration
This research represents more than a technical adjustment to demographic measurement. It calls for fundamental reconsideration of how we conceptualize and study fertility transitions. By bringing mortality back into the analytical framework, demographers can develop more nuanced understandings of reproductive change that better reflect the biological and social realities of human reproduction.
For policymakers and researchers working on population issues, these insights underscore the importance of considering both components of demographic change. Only by accounting for mortality alongside fertility can we develop a comprehensive understanding of how human reproduction actually evolves across different contexts and time periods.
*The views expressed herein are those of the author and do not necessarily reflect the views of the United Nations.
References
Keilman, Nico, Krzysztof Tymicki and Vegard Skirbekk. 2014. “Measures for human reproduction should be linked to both men and women.” International Journal of Population Research 2014, 908385, 10 pages. https://doi.org/10.1155/2014/908385Skirbekk
Skirbekk, Vegard and Thomas Spoorenberg (2025), “Bringing mortality back into our understanding of fertility change: Revisiting the onset of fertility transitions using net reproduction measures”, Journal of Population Research, 42, 45. https://doi.org/10.1007/s12546-025-09394-x