Summary

  • The Guardian reports a wave of US maternity ward closures attributed to federal Medicaid cuts, a mechanism the reporting details while the causal attribution to individual closures remains without authoritative adjudication.
  • Protect Our Care counted 52 US maternity ward closures since the HR1 spending bill was announced in July 2025, with a further eight wards having announced their upcoming closure.
  • March of Dimes identified 31 at-risk rural hospitals with labor and delivery services whose possible closure would leave 96 counties without a labor and delivery unit.
  • Benjamin Sommers said the Medicaid changes add financial strain to the providers and insurance program that pregnant women and low-income parents rely on.
  • Rita Hamad said the disruption of the CDC’s pregnancy surveillance system removed the only dataset of its kind, leaving recent maternal and infant health trends unmeasurable.

Maternity wards across the United States face a wave of closures as the operative provisions of the HR1 spending bill take effect in the new year. Protect Our Care, a nonprofit, has counted 52 US maternity ward closures since the bill was announced in July 2025, with a further eight wards having announced their upcoming closure. The reporting attributes the wave to the scale of the reductions: Vaishu Jawahar, the organization’s director of policy and programs, said the cuts force hospitals into “hard decisions” and that maternity wards “are extremely expensive to run.” Medicaid pays for about 40% of births in the United States, and the US maternal mortality rate of 17.9 deaths per 100,000 live births is among the highest of high-income nations, according to the latest figures from the Centers for Disease Control and Prevention. Experts expect more closures of maternity units in the run-up to and after the changes take effect. The 52-closure figure is the best available tally, but its causal status is documented rather than adjudicated: the two named hospitals cited the federal policy changes as a reason, and the reporting identifies no body producing a closure count with stated attribution criteria.

The tally and the population at stake

The reporting’s factual spine is the count, the payer share, and the outcome rate. Medicaid pays for about 40% of births in the United States, and the Centers for Disease Control and Prevention reports a US maternal mortality rate of 17.9 deaths per 100,000 live births, among the highest of high-income nations. March of Dimes has identified 31 at-risk rural hospitals with labor and delivery services; their possible closure would leave 96 counties without a labor and delivery unit — a conditional the group states as such, not a forecast. Experts quoted in the reporting expect more closures of maternity units in the run-up to and after the federal changes take effect in the new year.

The mechanism, documented

The bill’s operative provisions, which take effect in the new year, require many adults covered through Medicaid to meet new work requirements, subject them to more frequent eligibility checks, and provide less federal funding for states’ Medicaid programs.

Vaishu Jawahar, Protect Our Care’s director of policy and programs, attributes the closures to the scale of the reductions: “When you’re cutting Medicaid by a trillion dollars, you’re cutting this money from hospitals. So these hospitals are going into the red and being forced to make hard decisions, which usually results in maternity wards being closed as they are extremely expensive to run.” The reporting supplies no baseline, annual figure, or time horizon against which the trillion-dollar figure can be read.

Two of the permanently closed wards — Centra Southside Community Hospital in Virginia and St Mary’s Sacred Heart Hospital in rural northeast Georgia — “cited upcoming federal policy changes to Medicaid as a reason.” Those stated reasons carry the causal link that the count alone does not: the tally records closures since the July 2025 announcement rather than attributing each closure to the policy, and the reporting does not establish how many of the 52 would have occurred absent the Medicaid changes. A timing caveat bears on the attribution. Closures have been counted from the announcement, while the operative provisions do not take effect until the new year, and labor and delivery units are typically closed only after prolonged declines in delivery volume and clinical staffing — so some share of the 52 closures reflects decisions that predate the announcement.

The Guardian contacted the Department of Health and Human Services for comment; the article reports no response. The account records no competing federal explanation of the closures, and likewise no federal account of the PRAMS actions. The reporting documents the fiscal mechanism — reduced federal funding and tightened eligibility — but no party realizing an offsetting benefit: the closures represent withdrawal of the contested resource, not a transfer to a documented beneficiary.

How this is being framed

The same provisions are scored in two frames that do not share a metric. In a fiscal frame, the changes are measured as adjustments to programme cost and hospital solvency. In an access frame, they run against Medicaid’s role as the payer for about 40% of births in the United States, and the maternal-mortality rate of 17.9 deaths per 100,000 live births is the outcome that frame foregrounds. The planning theorists Horst Rittel and Melvin Webber gave the name “wicked” to problems of this structure, in which no test settles whether a policy is working because the competing accounts are scoring different things. Whether any test could settle this policy’s success is not stated in the reporting; the characterization follows from the frames’ lack of a shared metric rather than from an explicit statement.

The evidentiary reliance on advocacy counting is partly a consequence of the suspension of official counting. Where the government instrument has been withdrawn, the nonprofit tally is the remaining instrument, and treating it as inherently suspect would deepen the evidence gap rather than close it. The combined effect is that the consequences of the policy change are documented principally by organizations that oppose it, and no body is producing an authoritative closure count with stated attribution criteria — so whether specific closures are caused by the Medicaid changes or merely coincide with them has no authoritative adjudication.

What happens next

Rural exposure. Erin Jones, senior director of legislative and strategic counsel at March of Dimes, said the consequences for maternity care would be particularly devastating within rural communities. “Losing these services means fewer places to deliver safely and fewer options for pregnancies requiring specialized care,” she said. Jones said reduced funding and restrictions on how states finance Medicaid will put additional pressure on state budgets and hospitals, “increasing the risk that communities lose essential maternity services and specialized providers close to home.” Beyond the eight announced closures, experts expect more wards to close in the run-up to and after the changes take effect.

The enrollment failure inside the eligibility system. Benjamin Sommers, professor of healthcare economics at the Harvard TH Chan School of Public Health and former deputy assistant secretary for health policy under Joe Biden, said the changes would “absolutely” affect women who are pregnant and specialist pregnancy organizations. “It’s not true that this law in any way protects Medicaid for pregnant mothers, for low-income parents. It adds financial strain to the providers and the insurance program that they rely on,” he said. The mechanism he describes operates inside the eligibility system rather than in hospital finances: although the new work requirements are supposed to exempt pregnant women, Sommers said his research has shown that many women are not enrolled in the Medicaid pregnancy pathway, meaning the state might not know they are pregnant. “Someone could be enrolled through Medicaid expansion, they could get pregnant and they could mistakenly think – because they heard it on the news that this law doesn’t affect pregnant people – [they’re] fine, and it turns out the state doesn’t know they’re pregnant and they lose coverage,” he said. The exemption’s operating assumption is that a state can identify a beneficiary’s pregnancy status; that is the boundary at which the design can fail, and any corrective turns on whether pregnancy status can be established through a channel other than the pregnancy pathway.

The measurement sequel. The CDC’s Pregnancy Risk Assessment Monitoring System (PRAMS) has for more than four decades collected data on women’s experiences before, during, and after pregnancy. The entirety of the system’s team was placed on administrative leave and its data portal shut down, according to an upcoming report by academics and former CDC staff described to the Guardian, co-authored by Harvard social epidemiologist Rita Hamad. Hamad said the disruption removed the only system of its kind. “There are no other datasets like it. Teams like mine used to have access to this data before the portal was shut down, and so now there’s no way to study the more recent trends in maternal and infant health,” she said. “I think we’re going to see those geographical and maternal disparities widening. Or we’re not going to see it because we don’t have the data, but that’s what is going to happen.” The reporting records no replacement for the dataset; restoring the ability to measure the policy’s effect would require rebuilding the capability that was removed.

Who is documenting, and who is not

The three instruments that would establish the size, cause, and downstream effect of the wave are all held outside government: the closure tally by Protect Our Care, the at-risk hospital inventory by March of Dimes, and the pregnancy-surveillance account by academics and former CDC staff. Documentation functions in the account are otherwise active — Hamad, Sommers, and former CDC staff are recording what is happening, and the two nonprofits are the only parties maintaining closure inventories. Networks that span hospitals and state health officials exist, but the reporting shows none of them organizing in the affected areas.

In the vocabulary of third-party roles catalogued by the conflict theorist William Ury for a surrounding community, Protect Our Care, by counting closures, and March of Dimes, by publishing the at-risk hospital list, function as witnesses — third parties who make a conflict visible without intervening in it, whose visibility can itself restrain escalation. Protect Our Care’s work also carries an equalizing function in Ury’s sense, supplying information and voice to a population — pregnant women and low-income parents — that is not itself a party to the funding decision. In Ury’s taxonomy, no actor in the reporting is documented performing a provider role; Medicaid is the contested resource whose withdrawal the closures represent, not a role-holder. Ury’s resolution roles — mediator, arbiter, referee, healer — and the bridge-building role are not evidenced in the reporting. The reporting does not set out the taxonomy’s definitions; this mapping is interpretive.

Three functions are unfilled. No independent party — a state auditor, the Government Accountability Office, or a university center without an advocacy mandate — is producing a closure count with stated attribution criteria. No state health department or Medicaid agency is documenting how many pregnant enrollees are being dropped in eligibility redeterminations, the failure Sommers describes. And no neighboring hospitals, midwifery practices, community health centers, or transportation systems are reported as preparing for the influx from a 96-county footprint that would lose its labor and delivery units.

Two escalation signals appear in the account: a public symbolic marker in a photograph of people protesting against Medicaid cuts in Washington DC on 26 June 2025, and the exit of hospitals from a service line, which withdraws capacity rather than contesting it. The enrollment-communication gap Sommers describes — beneficiaries who “mistakenly think” they are unaffected — is the kind of deficit a teaching role would address, though no such role is documented. Third-side capacity in the reporting is concentrated in witnessing and amplification rather than resolution; the funding decision is made between other parties, and the affected population is not seated at it.

What would settle the question

The structural constraint on the dispute is that one party sets the funding terms unilaterally, so the availability of negotiable ground between the parties is narrow. The dispute compounds across three layers: a contested causal claim resists settlement when the metric is itself disputed; a disputed metric resists resolution when the dataset that would adjudicate it has been withdrawn; and a population without resolution roles has fewer levers to contest either.

The requirements that would settle the question are specific. An attribution standard would have to be applied to closures by a party with no stake in the bill’s defense or repeal. Pregnancy-surveillance data would have to be restored or independently replaced. And state-level documentation would have to record how many pregnant enrollees lose coverage in eligibility redeterminations. None of the three currently exists. Whether the closures prove to be the policy’s direct effect or a deterioration it accelerated, the reporting records no instrument that would distinguish the two. Until one of the three instruments exists, the 52-closure figure remains both the best available tally and permanently contestable — a condition the reporting documents as a product of policy, not of reporting.

Evidentiary gaps

None of the reporting’s central quantitative and dated claims carries an independent verification in the article: the 52-and-eight tally; the July 2025 announcement date; the 31-hospitals and 96-counties figures; the 40-percent share of US births paid by Medicaid; the 17.9 maternal-mortality rate and its series vintage; the placement of the PRAMS team on administrative leave; and the two hospitals’ stated reasons. Each is carried on the attribution of the reporting and of the sources it quotes, with the hedges recorded above. The PRAMS account rests on an upcoming report described to the Guardian and the statement of one co-author. The trillion-dollar figure stands unanchored: the reporting supplies no official score against which to read it. The interpretive characterizations in this analysis — the role mapping drawn from William Ury’s taxonomy, and the framing of the policy question in the terms Rittel and Webber called “wicked” — rest on inference from the reporting, which does not supply the taxonomy’s definitions or state that no test can settle the policy’s success.

Analytical techniques used in this piece

This analysis applies the methods below. Each links to a short, plain-English explainer you can read and reuse.

Red-Team Assessment
Models a capable adversary probing a plan for the seams they would exploit.
The Third Side
Takes the vantage of the surrounding community that has a stake in resolving a conflict (Ury).
Wicked Problems
Treats a problem as wicked — no stopping rule, no clean test of success, every attempt consequential.