Did the Expansion of Birth Control Clinics Change Fertility and Mortality in the United States?
Bauernschuster, Grimm, and Hajo (2026) examine whether Margaret Sanger’s early birth control clinics reduced fertility and improved maternal and infant health in the United States. They combine newly digitized information on 639 clinics opened before 1940 with full-count U.S. Census data and historical county- and city-level vital statistics. They find that clinic access reduced fertility substantially: the clinics explain about 5.4–6.5 percent of the national fertility decline through 1940. They also reduced stillbirths by about 3–6 percent and infant mortality by roughly 3.5 percent, largely by increasing birth spacing. They find no significant reduction in maternal puerperal deaths.
Why This Article Was Selected for The Policy Scientist
Access to effective contraception is a fundamental determinant of fertility, maternal and infant health, and families’ ability to determine the timing and spacing of births. These issues remain timely as access to reproductive health services continues to vary substantially across jurisdictions. This study is especially valuable because its newly digitized clinic records, full-count Census data, and historical vital statistics permit unusually detailed analysis of an important period in U.S. demographic change. Its staggered-treatment, fixed-effects, event-study, and difference-in-differences designs provide substantially stronger causal evidence than conventional multivariate regression. The findings may also inform settings where demand for effective family planning exceeds access. Forthcoming in American Economic Journal: Economic Policy, an American Economic Association journal devoted to rigorous research on economic policy, the article makes a notable contribution to the family-planning and demographic-transition literature.
Full Citation and Link to Article
Bauernschuster, S., Grimm, M., & Hajo, C. M. (in press). The impact of Margaret Sanger’s birth control clinics on early 20th century U.S. fertility and mortality. American Economic Journal: Economic Policy. American Economic Association.
Central Research Question
The study asks whether the spread of Margaret Sanger’s birth control clinics during the early twentieth century altered fertility and health outcomes in the United States. Beginning with Sanger’s first clinic in Brooklyn in 1916, hundreds of clinics opened across the country, particularly during the 1920s and 1930s. These facilities provided contraceptive counseling and, most importantly, fitted women with diaphragms and instructed them in their proper use. The authors investigate whether access to these services reduced births, increased the spacing between births, and affected stillbirths, infant mortality, and maternal mortality.
The question addresses a distinctive period in the history of American family planning. Modern highly effective contraceptives did not yet exist, information about contraception was often poor, and federal and state laws restricted access to contraceptive devices and information. The clinics therefore represented a substantial change in women’s practical access to effective contraception. The study asks not simply whether fertility was declining during this period, which is well established, but whether the expansion of these clinics causally contributed to that decline and generated additional health effects.
Previous Literature
The study connects several strands of research. One examines the historical decline in U.S. fertility. Jones and Tertilt (2006), Bailey and Hershbein (2018), Aaronson, Lange, and Mazumder (2014), and Ager, Herz, and Brueckner (2020), among others, analyze the economic and social forces contributing to the American demographic transition. Kitchens and Rodgers (2023), for example, estimate that agricultural price changes accounted for part of the fertility decline between 1910 and 1930, while research on the Great Depression finds that severe economic conditions also reduced fertility.
A second literature examines access to contraception and family planning. Goldin and Katz (2002) famously document how access to the birth control pill affected women’s marriage and career decisions. Bailey (2006, 2010, 2012, 2013) examines the effects of contraception and family-planning programs on fertility and women’s economic outcomes. This literature primarily concerns developments after World War II, making the earlier Sanger clinics an important extension into a period when contraceptive technology and institutions were substantially different.
The study also builds on research linking birth spacing to maternal and infant health. Knodel and Hermalin (1984), Conde-Agudelo, Rosas-Bermudez, and Kafury-Goeta (2006), and Molitoris (2017) provide evidence that closely spaced pregnancies are associated with poorer health outcomes. Finally, a substantial historical literature—including Reed (1978), McCann (1994), Hajo (2010), and Engelman (2011)—documents Sanger and the early birth-control movement. That literature has largely been qualitative. The present study provides a quantitative assessment of the clinics’ demographic and health consequences.
Data
The analysis combines several unusually extensive historical data sources. The authors first digitize information on the universe of birth control clinics established in the United States before 1940. Their database identifies 639 clinics across 44 states and records the county and year in which each clinic opened. Because women sometimes traveled across county lines for services, a county is considered exposed when a clinic opened either within that county or in an adjacent county.
The individual-level analysis uses the complete-count U.S. Censuses of 1920, 1930, and 1940 from IPUMS. The principal sample contains more than 45 million married women ages 15 to 39. Fertility is measured primarily as the number of a woman’s own children younger than five living in her household. The Census also provides information on age, race, birthplace, literacy, farm residence, and urban residence.
A second analysis uses annual county-level vital statistics. These data record live births, infant deaths, and stillbirths and allow the authors to calculate crude birth rates, infant mortality rates, and stillbirth rates. Coverage expands over time and includes essentially all U.S. counties by the 1930s. For maternal mortality, the researchers use annual cause-of-death records for more than 300 cities between 1920 and 1937, focusing on puerperal deaths associated with pregnancy and childbirth.
The authors supplement these sources with information on employment, urbanization, religion, economic conditions during the Great Depression, public relief spending, and other county characteristics that could potentially be related both to clinic placement and to fertility.
Methods
The empirical strategy exploits geographic and temporal variation in the expansion of birth control clinics. Clinics opened in different counties at different times, creating a staggered introduction of access to family-planning services.
For the Census analysis, the authors estimate how a woman’s accumulated years of exposure to a clinic are related to fertility. Models include county and Census-year fixed effects, age controls, and numerous socioeconomic covariates. Identification therefore comes from differences in the timing and duration of clinic exposure rather than simple comparisons between places that did and did not eventually receive clinics.
The county-level analysis uses difference-in-differences and event-study designs. Counties receiving clinics are compared with counties that had not yet received them, while county and year fixed effects account for persistent geographic differences and nationwide changes over time. Event studies allow the authors to examine fertility and mortality trends before and after clinic establishment. The absence of meaningful pre-treatment differences provides evidence supporting the required parallel-trends assumption.
Because conventional two-way fixed-effects estimators can produce biased results when treatment begins at different times and treatment effects vary, the authors also apply the Sun and Abraham (2021) interaction-weighted estimator. The principal results remain similar. Numerous robustness tests address urbanization, female employment, state-specific trends, Great Depression severity, social welfare programs, and changes in health infrastructure.
Findings/Size Effects
Access to birth control clinics produced a measurable decline in fertility. In the Census analysis, the average woman was exposed to a clinic for approximately 2.7 years. That level of exposure reduced the number of her children under age five by approximately 1.8 percent. Aggregated nationally, the estimates imply that Sanger clinics accounted for roughly 5.4 percent of the decline in this fertility measure between 1920 and 1940.
The independent vital-statistics analysis produces a similar result. Clinic establishment reduced the crude birth rate by approximately 0.47 births per 1,000 population, or about 2.5 percent relative to the birth rate when the clinic opened. These estimates imply that clinics accounted for approximately 6.5 percent of the overall decline in crude birth rates during the period examined. Thus, two different datasets and two different measures of fertility produce effects of similar magnitude.
The clinics also affected infant and fetal health. Stillbirths per 1,000 population declined by approximately 5.8 percent following clinic establishment. More importantly, stillbirths per 1,000 births declined by about 3.0 percent. Because the denominator accounts for the decline in births themselves, this finding indicates that fewer stillbirths cannot be explained simply by fewer pregnancies.
Infant mortality also fell. Clinics reduced infant deaths per 1,000 births by approximately 2.5 percent. The effect became larger during the years following a clinic’s establishment. These patterns are consistent with increased birth spacing and changes in the composition of pregnancies, including fewer pregnancies carrying comparatively high health risks.
The study finds no statistically significant effect on puerperal mortality. The estimated maternal mortality effect is close to zero, indicating that the measurable health benefits were concentrated among fetuses and infants rather than maternal deaths.
Conclusion
The findings indicate that early birth control clinics contributed meaningfully to the final stages of the long U.S. fertility transition. Although these clinics were not the dominant explanation for declining fertility, they account for approximately 5.4 to 6.5 percent of the national decline measured in the study. Their influence was therefore substantial given that the intervention involved a relatively new and initially limited network of clinics operating before modern contraception became available.
The evidence also shows that expanded access to contraception affected more than the number of births. Reductions in stillbirths and infant mortality suggest that the ability to increase intervals between pregnancies changed the health profile of births as well. The results support a mechanism in which professional contraceptive services allowed women to exercise greater control over the timing and spacing of pregnancies.
More broadly, the study demonstrates that family-planning services can influence demographic behavior even when the available contraceptive technology is comparatively simple. Its findings are most directly applicable to settings in which people desire fewer or more widely spaced births but face substantial barriers to obtaining reliable contraception.
Comments