Twenty-two states, the District of Columbia and six local governments asked a federal court Monday to block a new immigration rule before it takes effect September 18, opening a high-stakes dispute over how the government decides whether someone seeking a green card is likely to become a “public charge.” The two lawsuits, filed in the Southern District of New York, argue that the Department of Homeland Security has replaced a defined standard with broad discretion that will deter legally eligible families from using food, health and housing assistance. The administration says the change restores congressional intent and gives officers the flexibility to make individualized decisions.

The litigation is consequential well beyond the number of immigration applications that may ultimately be denied. In its own final rule, DHS estimated that people in households containing noncitizens could forgo or leave public programs in ways that reduce federal and state benefit transfers by about $13.05 billion annually. The agency also acknowledged possible effects on hospitals, grocery stores, landlords and local economies. That makes the case a test not only of immigration authority, but of whether administrative discretion can be expanded when the government predicts substantial indirect effects on people who may never face a public-charge finding.

No judge had ruled on the requests by Monday night. The litigation therefore does not itself suspend the rule, and applicants filing on or after Friday remain subject to it unless a court intervenes. Reporting by Reuters and the Associated Press described parallel complaints: one led by New York, California and Illinois on behalf of states, and another led by New York City with Chicago, San Francisco, Santa Clara County, Seattle and King County, Washington.

What the rule changes

Federal immigration law has long made inadmissible an applicant who is “likely at any time to become a public charge.” Congress directs officers to consider at least age, health, family status, assets, resources, financial status, education and skills. The dispute concerns how DHS interprets that language. A 2022 regulation limited the benefit component principally to cash assistance for income maintenance and long-term institutional care paid by the government. The new rule rescinds most of that framework and moves away from what DHS calls a bright-line “primary dependence” standard.

Under the replacement policy, officers may consider an applicant’s receipt of any means-tested public benefit after the effective date, together with the statutory factors, case-specific circumstances and relevant data about self-sufficiency. DHS says the inquiry must remain individualized and based on the totality of the circumstances. Benefit use is not an automatic denial, and the rule does not establish a fixed dollar amount or duration that independently decides a case. Benefits received before September 18 are to be assessed under the narrower 2022 standard.

The administration’s accompanying USCIS guidance identifies categories officers may examine, including cash aid, food assistance, public or assisted housing, postsecondary financial aid and government-funded health coverage. Earned benefits such as Social Security retirement, Medicare, unemployment insurance, government pensions and veterans benefits are not treated as means-tested assistance. The policy also does not erase statutory exemptions for refugees and several humanitarian categories.

A critical distinction is how the rule treats a family member’s benefits. DHS says assistance received solely by a child or another relative is generally not attributed to the applicant. However, the facts that made a household eligible—such as the applicant’s income, assets or role as the family’s source of support—may still enter the broader financial analysis. That nuance is important because opponents describe a risk to mixed-status families, while the government insists officers are evaluating the applicant rather than automatically counting a citizen child’s enrollment as the parent’s receipt.

The state coalition alleges that DHS exceeded the meaning Congress gave “public charge” and violated the Administrative Procedure Act by adopting an arbitrary and insufficiently explained policy. In announcing the case, New York’s attorney general said the states seek an order declaring the rule unlawful and vacating it. California’s filing notice similarly argues that the rule departs from a historical focus on primary dependence and leaves families without a dependable list of programs or thresholds.

The complaints also attack the rule’s connection between benefit receipt and future dependence. A person may use short-term assistance during pregnancy, a job transition or a medical crisis without being primarily reliant on government over time. The states argue that allowing any means-tested benefit to carry unspecified weight lets an officer treat temporary help as evidence of a much broader condition. Whether that discretion is a faithful implementation of Congress’s instruction or an impermissible expansion is likely to shape the court’s review.

The administration’s rationale

DHS says the 2022 regulation was too restrictive and prevented officers from considering pertinent information. In the rulemaking record, the department argues that Congress deliberately required a forward-looking judgment and listed minimum factors rather than an exhaustive formula. From that perspective, detailed regulatory definitions can become a straitjacket: two applicants with similar benefit histories may have different health, income, sponsorship and employment prospects that warrant different conclusions.

But flexibility creates a tradeoff. Clear thresholds can produce underinclusive decisions, while open-ended standards can reduce consistency between officers and make behavior harder for applicants to plan. A KFF analysis of the implementation guidance said key concepts remain undefined and warned that applicants may struggle to know how much weight a particular benefit will receive. The court will have to assess whether subregulatory guidance supplies enough structure for the discretion DHS reclaimed.

Formal denials and broader effects

The government’s own historical figures complicate claims on both sides. DHS reported an average of roughly 68,072 annual adjustment-of-status denials in recent years, but only about 65 were recorded as public-charge denials—less than one-tenth of 1 percent of all adjustment denials. The agency further said that, under the totality analysis used during the 2019 rule’s effective period, the few identified denials or notices of intent to deny were reopened, rescinded or ultimately approved. Past enforcement therefore offers little basis for estimating how many applications the new standard will change.

Formal denials, however, are not the only measurable outcome. DHS projects reduced benefit participation by noncitizens and citizens in mixed-status households, even though many of those people are not themselves subject to the rule. The department characterizes those reductions as lower transfer payments, while also recognizing downstream costs for health providers, food retailers, agricultural producers and housing participants. The states describe the same mechanism as a chilling effect that harms eligible residents and public systems.

Evidence from the first Trump administration explains why both sides focus on deterrence. A Migration Policy analysis of Census Bureau data found that from 2016 to 2019 participation in Medicaid, food and cash-assistance programs fell about twice as fast among noncitizens as among citizens. Among low-income noncitizens, SNAP and cash-assistance participation each fell 37 percent and Medicaid participation fell 20 percent. The researchers cautioned that improving economic conditions reduced participation generally, but concluded that the sharper noncitizen decline was consistent with policy-related fear and confusion.

Current projections are similarly uncertain. KFF modeled, rather than predicted, that if 10 percent to 30 percent of Medicaid or Children’s Health Insurance Program enrollees in households with a noncitizen disenrolled, 1.4 million to 4.1 million people could lose coverage, including 560,000 to 1.7 million citizen children. Those scenarios do not establish what will occur. They show why small changes in household decisions could exceed the direct immigration caseload affected by an adverse determination.

What the court must decide

The immediate question is whether the challengers can obtain relief before the September 18 start date. To justify an injunction, they must generally show likely success on the merits, irreparable harm, a favorable balance of equities and consistency with the public interest. The administration can counter that its reading fits the statute, that the rule preserves case-by-case review and that projected disenrollment reflects choices rather than a legal command to leave benefits.

The deeper issue is where discretion ends and notice begins. Congress plainly authorized a prospective judgment about dependence, and DHS plainly predicts that its implementation will alter conduct far beyond applicants who receive final denials. The court’s task is not to choose a preferred immigration policy. It is to decide whether DHS lawfully explained the change, reasonably accounted for foreseeable effects and gave officers and applicants a workable standard.

Until a judge acts, the practical dividing line remains Friday. Applications filed before September 18 are governed by the existing framework; those filed on or after that date face the new one. For families, benefit administrators and immigration lawyers, that four-day gap between the lawsuits and implementation leaves little time for certainty. The rule’s ultimate reach will depend first on emergency court orders, then on the merits of a dispute likely to extend well beyond its effective date.