Abortion statistics

1. National totals and the recent trend: what the big numbers mean

Recent aggregated reporting places the United States at about 1,037,000 abortions in 2023 (a figure cited by Brookings) and Guttmacher’s work reports an abortion rate of approximately 15.4–15.9 abortions per 1,000 women aged 15–44 in 2023–2024, reflecting a modest increase from earlier years and the highest annual totals in recent history, driven in part by changing access after Dobbs and expanded medication abortion provision ^{[2] [1]}. These headline numbers are useful for tracking macro trends, but they mask important measurement challenges: Guttmacher’s estimates come from facility surveys and modelling, Brookings synthesizes multiple sources, and both organizations note that these counts likely understate total abortion incidence because they can miss abortions obtained via mail-order pills or through networks and some services provided across state lines ^{[1] [2]}.

2. State-by-state variation: legal context and reporting gaps

Abortion rates and counts vary dramatically by state, with some states and jurisdictions showing rates above 20 per 1,000 and others near zero in official reporting; for example, pre‑Dobbs analyses highlighted Washington, D.C. at 24 per 1,000 and large state differences such as Florida’s high reported totals ^{[4] [5]}. However, the CDC’s surveillance is voluntary and several large states (including California, Maryland, and New Jersey at times) have not consistently reported, producing holes in the national picture and complicating direct state-to-state comparisons ^{[3] [4]}. In addition, legal changes after the Dobbs ruling shifted where abortions occur: states with bans see far fewer in‑state procedures but may have out‑of‑state residents traveling for care or turning to mailed medication, which official in‑state counts often do not capture ^{[3] [1]}.

3. Race, ethnicity and demographic patterns: data and limitations

Available analyses show clear disparities: CDC data from jurisdictions that reported racial and ethnic information found abortion rates in 2021 were about 28.6 per 1,000 for non‑Hispanic Black women, 12.3 for Hispanic women, and 6.4 for non‑Hispanic White women, illustrating concentrated impacts across groups ^[6]. But these figures are drawn from a subset of reporting areas and exclude some jurisdictions, and both the CDC and other researchers caution that denominators, reporting completeness, and demographic classification differences limit how precisely these rates can be compared or attributed solely to policy differences ^{[3] [6]}. Guttmacher and other organizations also note that counts do not fully capture self‑managed or mail-order medication abortions, which could shift demographic patterns if different groups use those pathways at different rates ^[1].

4. Safety, complications, and mortality: what surveillance shows

Public health surveillance consistently finds that legal induced abortion is a low‑mortality procedure: CDC analyses report case-fatality rates below 1 death per 100,000 abortions for multi‑year periods, underscoring low fatality in settings where care is provided by licensed clinicians ^[3]. The CDC and clinical reviews also note that about 2% of abortions involve some type of complication such as bleeding or infection, most of which are considered minor; however, surveillance systems track clinical complications unevenly and do not capture self‑managed care outcomes well, leaving gaps in our understanding of safety outside clinical settings ^{[3] [6]}.

5. Why different sources give different answers: methods, politics, and incentives

Differences among CDC, Guttmacher, state health departments, and third‑party compilations stem from methodology: CDC data are voluntarily reported by states and exclude some large jurisdictions, Guttmacher uses facility surveys and modeling that attempt to estimate unreported care, and private or advocacy‑linked sources may focus on different subsets ^{[3] [1] [7]}. These methodological choices reflect both technical limitations and implicit agendas: public agencies emphasize clinical reporting and surveillance consistency, researchers aim to estimate total incidence (including out‑of‑state care), and some private compilers may have incentives to highlight particular trends; readers must weigh each source’s scope and potential biases when interpreting headline numbers ^{[3] [1] [7]}.

6. What readers should watch next: data gaps and evolving measures

Key items to monitor are how surveillance adapts to the rise of medication abortions and self‑management (which many systems currently undercount), whether states restore or change reporting practices (affecting CDC completeness), and how researchers update demographic analyses as more recent post‑Dobbs years of data are compiled ^{[1] [3] [4]}. Until reporting becomes more uniform or new methodologies explicitly measure mail‑order and self‑managed care, national and state statistics will remain estimates with important blind spots, and analysts should triangulate Guttmacher, CDC, and state reports rather than rely on a single series ^{[1] [3]}.