Which federal agencies publish annual abortion statistics and how do their methods differ?
Executive summary
Two federal actors publish or have published national abortion statistics: the Centers for Disease Control and Prevention (CDC), via its long‑running Abortion Surveillance System, and non‑government research groups (not federal agencies) such as the Guttmacher Institute produce national estimates used by policymakers and media (CDC remains the sole governmental source historically) [1] [2]. The CDC’s surveillance is a voluntary, aggregate, state‑based reporting system with a typical two‑year release lag and incomplete coverage; Guttmacher’s independent monthly and annual censuses contact providers directly and use statistical models to fill gaps, producing higher and more timely counts [3] [4] [5].
1. CDC: the official but voluntary federal surveillance system
The CDC has compiled the Abortion Surveillance report since 1969 by asking central health agencies in 50 states, DC and New York City to send aggregate counts and characteristics of legal induced abortions; reporting is voluntary, the agency publishes analyses after receiving state data, and reports typically appear on a two‑year delay [3] [6] [2]. The 2022 MMWR showed 613,383 abortions reported from 48 reporting areas and documented method, gestational age and demographic breakdowns — but it excluded data from some large jurisdictions and therefore undercounts the true national total [2] [3]. CDC’s system emphasizes standardized templates and public‑health uses such as trend monitoring and contraceptive program evaluation [6] [3].
2. How CDC’s methods shape its strengths and limits
CDC relies on state vital‑statistics or health‑department aggregates rather than a provider census; it issues model forms and technical guidance but cannot compel states to submit consistent fields, and states vary in what they collect [3] [7]. That design yields consistent, comparable tables on variables like method and gestational age when jurisdictions comply, but it produces incomplete national coverage (several states have historically not reported) and time lags that blunt its usefulness for near‑real‑time policy and program analysis [3] [7] [8]. Recent reporting and staffing disruptions at CDC have introduced further uncertainty about whether and when future annual surveillance reports will continue [8] [9].
3. Guttmacher Institute: a non‑federal census with modeling for timeliness
The Guttmacher Institute is not a federal agency, but its work is central to national abortion counts. Guttmacher contacts every known provider and runs the Monthly Abortion Provision Study to estimate clinician‑provided abortions monthly, combining provider surveys with historical caseload data and statistical models to fill nonresponse and deliver more timely, fuller totals [4] [5] [10]. In 2024 Guttmacher estimated 1,038,100 clinician‑provided abortions in states without total bans; its monthly model produced preliminary 2025 half‑year figures for states without bans [4] [5].
4. Why Guttmacher and CDC report different totals
Differences stem from source frames and methods: CDC’s numbers are aggregate state reports that may omit non‑reporting jurisdictions and exclude some telehealth or out‑of‑state counting conventions; Guttmacher conducts a provider census and imputes missing data to approximate the full universe of clinician‑provided abortions, so its totals are consistently higher and more complete but rely on modeling assumptions [3] [4] [10]. Independent reviewers note CDC historically captured only about two‑thirds of the abortions counted by Guttmacher in 2020, illustrating systematic divergence due to method and coverage [11].
5. Tradeoffs for policymakers and journalists
CDC’s strengths: standardized public‑health tables on method, gestation and demographics, and the imprimatur of a federal surveillance system — valuable for long‑range trend analysis and public‑health planning [6] [2]. CDC’s limits: voluntary reporting, excluded jurisdictions, multi‑year lag and, as of 2025, uncertainty about continuity of the published series [3] [8] [9]. Guttmacher’s strengths: near‑real‑time monthly estimates, provider‑level reach and fuller national totals useful for timely policy debates; its limits: reliance on survey response and modeled imputations for nonrespondents and providers not in the sampling frame [5] [4] [10].
6. Data completeness, privacy and political context
State reporting laws vary and can include detailed patient identifiers; after Dobbs, concerns about surveillance and criminalization prompted some states and providers to withhold or rethink reporting, and federal staffing changes have further clouded CDC’s ability to continue the series — all of which affect completeness and the politics of using the data [11] [2] [8]. The Government Accountability Office and reporting outlets also flagged that agency realignments and staffing cuts create real uncertainty about future federal abortion data collection [9] [12].
7. Bottom line for users of these statistics
Use CDC when you need standardized, historically comparable federal surveillance tables and method‑by‑age‑by‑gestation detail, but account for excluded jurisdictions and lag [2] [3]. Use Guttmacher (and complementary provider or #WeCount reports) for timelier, fuller national and state estimates that include telehealth and provider counts — but disclose that those numbers are model‑based estimates that fill gaps in provider responses [4] [5] [13]. Available sources do not mention a second active federal agency publishing annual national abortion totals beyond CDC; independent organizations fill the information vacuum when federal reporting is incomplete [1] [4].