Published Research
The academic research foundations behind Birthgap·org
The academic research foundations behind Birthgap·org
The core research of Stephen J. Shaw has been peer-reviewed and published in Scientific Reports (Nature Portfolio). In this section, you can find:
- an accessible overview of the paper itself, which introduces the Microdemographic Framework (MDF),
- key implications of the framework for demographic research,
- a detailed case study from Japan on The Fertility Shock,
- and notes on continuing research in structural demography.
Access the 2025 Scientific Reports paper
On a Microdemographic Framework for Analyzing Contemporary Fertility Dynamics, by Stephen J. Shaw, published in Scientific Reports (Nature Portfolio).
Read the paper →Introducing the Microdemographic Framework
The peer-reviewed research developed by Stephen J. Shaw is built around a structural approach to fertility measurement known as the Microdemographic Framework (MDF). Rather than treating fertility as an average number of children per woman—the traditional approach—the MDF focuses on how parenthood actually occurs across the life course, using observed demographic outcomes for family size and childlessness. The paper examined these patterns across more than 30 nations and regions, drawing on data from over 300 million mothers, with trends in some cases extending back to the 1970s. This novel yet intuitive framework was formally introduced in Scientific Reports as a powerful tool for policymakers and researchers seeking to better understand low-fertility outcomes.
See how the MDF differs from conventional fertility measures
Much of the conventional discussion around fertility relies on the Total Fertility Rate (TFR). While TFR is a useful summary statistic, it compresses several distinct demographic processes into a single number—making it difficult to identify what is actually driving changes in birth rates.
The Microdemographic Framework reveals that fertility can be meaningfully decomposed into two independent components:
- Total Maternal Rate (TMR) — the lifetime probability that a woman becomes a mother at least once
- Children per Mother (CPM) — the average number of children born to women who do become mothers
A central empirical finding is that TMR and CPM are statistically independent across countries and over time. In other words, falling birth rates are frequently driven not by parents choosing smaller families, but by fewer people becoming parents at all.
By separating entry into parenthood (TMR) from completed family size (CPM), the MDF greatly reduces information loss, brings the role of unplanned childlessness into clearer focus, and offers a more accurate structural picture of demographic change. The statistical independence of these two components challenges many conventional assumptions and research methods in fertility studies.
Read the full Scientific Reports paper or explore a summary of its impact on demographic measurement below.
The MDF’s Implications for Demographic Research
The Microdemographic Framework (MDF) extends beyond a mathematical decomposition. The formal demonstration that TMR and CPM are statistically independent across countries and time calls into question the conventional reliance on the Total Fertility Rate (TFR) as a singular explanatory measure.
Because TFR is the product of two independent components (TFR = TMR × CPM), similar aggregate fertility rates can emerge from structurally different underlying processes. As illustrated in the CPM distribution chart, Children per Mother remains remarkably stable across decades in advanced economies. In contrast, variation in aggregate births often reflects shifts in entry into motherhood (TMR) rather than changes in desired family size among those who become parents.
See the statistical evidence and recommendations
The costs of aggregation become clear when looking at model performance. As the histogram of Adjusted R² gains shows, switching from TFR to separate modeling of TMR and CPM produces an average improvement of 0.385 — a substantial and consistent boost across specifications. The gains cluster around a positive mean, with confidence intervals confirming the effect is systematic, not random. In short, models built on the MDF's independent components deliver materially stronger explanatory power.
Quantitatively, reliance on TFR alone is therefore found to be associated with substantial information compression, obscuring nearly half of the structural variation captured when the components are modeled independently. In practical terms, this means that structural signals—particularly shock-like declines in entry into parenthood—may be partially concealed when fertility is treated as a single aggregated outcome.
Empirically, the study showed that the average TFR (to one decimal place, as is typical in demographic reporting) was associated with an average of 13 unique combinations of TMR and CPM. For example, the TFR for 1980 was identical to that in 2016, at 1.82, yet the underlying rates of motherhood and family size had shifted in opposite directions—in effect canceling each other out. This danger can only be revealed by examining TMR and CPM separately, reinforcing the Microdemographic Framework's role as an essential diagnostic tool for interpreting TFR trends.
As a result, researchers are encouraged to use both TMR and CPM as explanatory or response variables in modeling, given the likelihood of stronger statistical performance and greater structural clarity. This approach increases the probability of isolating family size dynamics (CPM) from changes in the incidence of motherhood (TMR). A systematic re-examination of existing literature using this decomposition is also recommended, as prior findings based solely on TFR may conceal distinct and separable demographic mechanisms.
Case Study: Japan’s 1974–1975 Fertility Shock
One of the clearest demonstrations of the Microdemographic Framework (MDF) comes from Japan in the mid-1970s, as covered in the published paper. Using annual data across all 47 prefectures, the study identifies a sharp and highly synchronized decline in entry into motherhood between 1974 and 1975, while those with children already did not delay having subsequent children. This dramatic difference in behaviour forms a natural experiment to enable better understanding of contemporary fertility decline. There is evidence the same pattern unfolded in other nations too at the same moment in time.
These examples align with the Global Oil Shock of late 1973 – Japan was the world's largest oil importer at the time, and the crisis dominated the media for weeks, with one newspaper on November 14th 1973, Mainichi Shinbun, devoting its first nine pages to the crisis.
Read the full analysis from the Scientific Reports paper
The key signal appears in the Total Maternal Rate (TMR)—the lifetime probability that a woman becomes a mother at least once. In this single year-pair, every prefecture shows a decline in TMR, with the vast majority experiencing a material drop. No adjacent year-pair in the post-1960 period exhibits the same level of coordinated directional change.
The timing aligns closely with the 1973 Oil Crisis, a major economic shock in Japan. Importantly, the demographic pattern was detected first, and the historical context was investigated afterward. This sequence supports the interpretation that broad uncertainty and economic disruption can trigger synchronized postponement of parenthood at scale.
The rise in the Total Childlessness Rate (TCR) during this period reinforces the interpretation that the disruption operated primarily through delayed or foregone entry into parenthood. TCR captures multiple pathways to childlessness—including voluntary decisions, medical or biological factors, and circumstantial outcomes. However, the speed and synchronicity of the shift during a period of economic uncertainty suggest that structural and circumstantial pressures were likely central, consistent with elevated levels of what has been described as Unplanned Childlessness. This interpretation aligns with broader demographic evidence, including findings by Beaujouan et al., who show that much of 20th-century European childlessness reflected structural conditions rather than deliberate long-term preference shifts.
Continuing Research
Building on the published MDF paper, ongoing work is now focused on further structural decomposition—this time of the Total Maternal Rate (TMR) itself. The goal is to identify persistent patterns in the timing and incidence of first births across cohorts and populations. This emerging direction is being developed under the umbrella term Structural Demography.
Learn more about the broader conceptual framework and intersections
The term Structural Demography is not new. It has appeared most prominently in the work of Peter Turchin and colleagues under the label Structural-Demographic Theory, where demographic structure is used to explain long-run patterns of social and political instability. In the present context, the term is applied more narrowly to fertility and life-course dynamics, focusing on enduring population structures—such as timing distributions and cohort configurations—that may shape reproductive outcomes across societies.
This orientation also intersects with strands of evolutionary demography (Wachter & Bulatao, 2003), a field that examines reproductive timing, life-history trade-offs, and population-level patterns through an evolutionary lens. While the present work remains grounded in observed demographic outcomes rather than genetic mechanisms, the recurrence of highly structured timing distributions and coordinated shifts in entry into parenthood is consistent with deeper system-level regularities. In this sense, Structural Demography provides a bridge between formal demographic measurement and broader evolutionary approaches to understanding how reproductive behavior organizes across cohorts.
Although structural approaches have traditionally been less central than proximate socioeconomic explanations in fertility research, the recurrence of similar low-fertility patterns across diverse cultural and institutional contexts suggests that deeper population-level regularities may be operating. A structural perspective therefore complements conventional social-scientific analysis by examining systemic dynamics that persist across time and geography.
Together with the Microdemographic Framework, this direction seeks to develop a more causally coherent, systems-oriented account of contemporary fertility—situated at the intersection of formal demography, life history theory, and evolutionary approaches. Further findings will be published in due course.
Access Microdemographic Data
Explore the Birthgap Barometer
Access interactive charts for over 40 nations, with trends often going back decades, showing rates of motherhood, children per mother, average age of motherhood, and birthrate trajectories.
An intuitive introduction to fertility measurement
Alongside the formal research, Stephen J. Shaw has written a short story, Measured Futures, which introduces the Microdemographic Framework and explains the limitations of conventional fertility statistics.
Read the story →Support Independent Research on Global Birthrates
Donations help fund ongoing research, educational materials, public events, and open public discussion on demographic change. You may choose a one-time gift or a monthly contribution via our partner Donorbox. We appreciate all support that makes this site and our work possible.