On March 1, 2026, New York State will begin enforcing new work requirements for recipients of the Supplemental Nutrition Assistance Program. The flyer from the Office of Temporary and Disability Assistance announces this plainly—bold text, clean design, a checklist of tools arranged neatly around the headline: SNAP ABAWD Work Requirements Are Changing. The policy applies to a category of people the federal government calls “Able-Bodied Adults Without Dependents.” If you are between 18 and 64, do not live with a child under 14, and have not been certified as physically or mentally unfit for employment, you must document 80 hours per month of work, job training, or volunteering to continue receiving food assistance. Fail to meet the requirement for three months in a three-year period, and you lose your benefits. The visual grammar is bureaucratic efficiency. The message is administrative: new rules, new deadlines, comply or lose your food.
But there is nothing neutral about the category that determines who these rules apply to. “Able-Bodied Adults Without Dependents” is not a medical finding. It is not a clinical assessment. It is a policy presumption embedded in the structure of the program since welfare reform in 1996, expanded dramatically by the One Big Beautiful Bill Act signed into law in July 2025. The term does not describe a population. It constructs one. And the construction does real work: it decides who eats.
The Binary That Doesn’t Exist
The ABAWD framework operates on a simple binary: you are either able-bodied or disabled. If disabled, you are exempt. If able-bodied, you work or you lose your food assistance. The logic seems intuitive. The problem is that the binary it relies on does not describe how disability actually works.
Most disability is not a fixed state. It is episodic, fluctuating, context-dependent, and frequently invisible. A person with a chronic pain condition may be able to work twenty hours one week and unable to get out of bed the next. A person managing severe depression may hold a job for months and then lose the capacity to function for weeks. A person with an autoimmune disorder may appear entirely healthy in a caseworker’s office while their body is in a flare that makes sustained employment impossible. None of these experiences fit cleanly into “able-bodied” or “disabled.” They exist in the space the binary refuses to acknowledge.
The policy does not assess capacity. It assumes it. Every person who has not been certified as unfit for work is, by default, able-bodied. The category is not earned through evaluation. It is assigned through omission. If you have not proven you belong in the “disabled” box, the system places you in the “able-bodied” box and starts the clock: three months to document 80 hours of work activity, or your food benefits end.
This inversion matters. In clinical practice, the question is what a person can do, given their specific circumstances. In the ABAWD framework, the question is whether a person has successfully demonstrated that they cannot. The burden of proof sits entirely on the person who needs food.
The Exemption as Proof Ritual
The system’s response to this critique is predictable: there are exemptions. Disabled people are not subject to the ABAWD work requirements. People who are physically or mentally unfit for employment can be exempted. The protections exist.
Look at what the exemption actually requires. New York’s guidance to eligibility workers, issued in a 2020 informational directive, instructs that when “an individual’s unfitness for work is obvious to the eligibility worker, the district should exempt the individual from ABAWD work requirements without requiring a statement or verification from an appropriate medical professional.” When unfitness is not obvious, the individual must produce a medical statement from a physician, psychologist, social worker, or other medical personnel certifying their inability to work 80 hours per month.
That word—obvious—is doing enormous structural work. It means the default is able-bodied. The system presumes capacity unless disability is visible and legible to a caseworker in a brief interaction. Chronic illness is not obvious. Episodic mental health conditions are not obvious. Pain is not obvious. Fatigue is not obvious. Cognitive impairment that does not present as incoherence is not obvious. The entire landscape of disability that does not perform itself for an observer gets sorted, by default, into the work requirement.
For those whose conditions are not “obvious,” the exemption demands medical documentation. This means a person who may be struggling to manage daily functioning must navigate the healthcare system—scheduling appointments, obtaining specific paperwork, delivering it to the correct office within specific deadlines—in order to prove they qualify for food. For people without stable housing, reliable transportation, consistent phone access, or an established relationship with a medical provider, this is not a safeguard. It is an obstacle course that serves as its own screen: those who can navigate bureaucracy get to eat; those who cannot are classified as noncompliant.
Consider a 59-year-old with degenerative disc disease who can stock shelves some weeks but not others. Her condition is not “obvious” in a caseworker meeting—she is upright, she is verbal, she arrived at the office on time. She does not have a current doctor because her last provider stopped accepting Medicaid. She cannot produce the medical statement the system requires to exempt her. So the system classifies her as able-bodied, starts the clock, and waits for her to either document 80 hours of work or lose her food assistance. The machinery processes her body. It does not see it.
The exemption does not fix the binary. It enforces it. To escape the consequences of being classified as able-bodied, you must prove you are disabled. The categories remain rigid. The burden remains on the individual. And the system’s design ensures that the people most likely to need the exemption—those with unstable health, limited resources, and the least capacity to manage paperwork—are the ones least likely to obtain it.
A 2020 study published in JAMA Network Open by researchers at George Washington University found that SNAP work requirements caused disabled adults to lose benefits—even though they were ostensibly exempt. The study was the first published research to demonstrate this spillover effect. Disabled people lost food assistance not because the policy targeted them, but because the exemption process failed to protect them. The documentation burden, the caseworker discretion, the administrative complexity—these mechanisms do not distinguish between people who are able-bodied and people whose disabilities are simply not legible to the system.
The Expansion
The One Big Beautiful Bill Act did not merely maintain the ABAWD framework. It expanded it in every direction.
The age ceiling for work requirements was raised from 54 to 64, sweeping in older adults who had previously been exempt. Parents with children aged 14 and older are now subject to the requirements—a shift from the prior threshold of 18. Exemptions for veterans, people experiencing homelessness, and former foster youth were eliminated. And the threshold for state-level waivers—which had allowed states to suspend work requirements in areas with weak labor markets—was tightened to require unemployment rates above 10 percent, a level so extreme that it effectively eliminates waiver eligibility for most jurisdictions in most economic conditions.
The scale is not subtle. The Center on Budget and Policy Priorities estimates that roughly 11 million people are at risk of losing food assistance entirely or seeing household benefits cut—including more than 4 million children and over half a million older adults and adults with disabilities. The Congressional Budget Office scored the SNAP provisions as $186 billion in cuts through 2034, the largest reduction in food assistance in American history.
Each of these expansions has a specific relationship to disability that is never named in the legislation.
Raising the age ceiling to 64 is not a neutral broadening of the workforce expectation. Adults between 55 and 64 are far more likely to have chronic health conditions, physical limitations, and age-related disability that falls below the threshold of formal disability certification but profoundly affects the capacity for sustained employment. They are the population most likely to be caught in the gap between “able-bodied” and “disabled”—too impaired for the work requirement, not impaired enough for the exemption.
Eliminating the exemption for people experiencing homelessness removes protections from a population with extraordinarily high rates of disability, mental health conditions, and substance use disorders—conditions that the exemption process is structurally ill-equipped to capture. Removing the veterans’ exemption does the same for a population with elevated rates of PTSD, traumatic brain injury, and service-connected disabilities that may not meet the “obviously unfit” standard.
The pattern is consistent. Every expansion widens the population that is presumed able-bodied. Every eliminated exemption narrows the exits. The binary does not change. It simply absorbs more people into the category that triggers loss of food.
The Documentation Trap
The stated purpose of SNAP work requirements is to increase employment among low-income adults. If the policy accomplished this, the debate would be different. It does not.
The Hamilton Project and Brookings Institution conducted a comprehensive review of the research literature in 2025 and concluded that the best evidence shows work requirements do not increase employment. A rigorous study using linked administrative data from Virginia, published in the American Economic Journal: Economic Policy, found that work requirements produced no effects on employment while reducing SNAP participation by 53 percent among adults subject to the policy. Homeless adults were disproportionately screened out.
People do not lose SNAP because they refuse to work. They lose it because they cannot document 80 hours of activity to a bureaucratic system that was not designed to accommodate the conditions of their labor. Gig workers whose hours fluctuate weekly. Part-time employees whose schedules change without notice. Day laborers paid in cash. People in rural areas where the nearest qualifying work program is an hour’s drive away and there is no public transit. People without reliable internet access who cannot log hours through a state portal. People whose paperwork is delayed, lost, or filed incorrectly.
The administrative burden is not a side effect. It is the mechanism. When the Center on Budget and Policy Priorities warns that people lose benefits not because they are not working but because they cannot meet the reporting requirements, they are describing a system that functions as designed. The compliance apparatus does not measure work. It measures the capacity to navigate bureaucracy. And it treats the inability to navigate bureaucracy as evidence that a person does not deserve food.
This is the fraud narrative engine. A person who was working 25 hours a week but could not produce the right documentation in the right format by the right deadline is classified as noncompliant. The system records a “failure to meet work requirements.” The political narrative converts this into evidence that the program is full of people who do not want to work. The data generated by the administrative apparatus confirms the assumption the policy was built on. The category produces the outcome it claims to describe.
What “Able-Bodied” Obscures
The term “able-bodied” performs a specific ideological function. It naturalizes a division that is political. It makes the category seem like a description of physical reality rather than a policy choice about who qualifies for help.
Consider what the term actually claims. To be “able-bodied” is to possess a body that can work—specifically, that can work 80 hours per month in activities that the state recognizes and that the applicant can document. But capacity for work is not a property of a body. It is a relationship between a body, an environment, a labor market, a set of available supports, and a documentation infrastructure. A person with the same body may be “able” to work in a city with public transit, accessible workplaces, and available part-time employment, and “unable” in a rural county with no bus routes, no qualifying work programs, and a single oversubscribed job training provider. The body has not changed. The conditions have.
Disability studies has understood this for decades. Disability is not a property of individual bodies but a product of the interaction between bodies and environments. The ABAWD framework ignores this entirely. It locates capacity in the individual, treats environmental conditions as irrelevant, and then punishes people for failing to meet a standard that the environment makes impossible.
This is the same logic that produces the emergency exception described in disaster governance: institutions designed for an imagined average body, processing deviation as individual limitation rather than design failure. The shelter that cannot accommodate dietary needs. The rebuilding process that suspends accessibility requirements. The food assistance program that presumes capacity and punishes those who cannot prove otherwise. The mechanism is consistent. The institution designs for a norm, defines everyone who falls outside it as deficient, and frames the consequences as personal rather than structural.
The Quiet Violence of the Category
There is a particular cruelty in naming a policy category after the condition of a person’s body. “Able-Bodied Adults Without Dependents” does not say “people who have not been certified as disabled.” It does not say “people subject to work requirements.” It says your body is able. It makes a claim about you—about your physical reality, your capacity, your fitness—and then uses that claim as the basis for withdrawing food if you do not perform in accordance with it.
For people living in the gap—the millions of Americans whose bodies are not fully “able” but not certifiably “disabled,” whose conditions fluctuate, whose impairments do not show, whose access to medical documentation is limited, whose interaction with a caseworker does not happen to reveal the pain or fatigue or cognitive difficulty that shapes their daily existence—the category is not a description. It is an erasure. It says: the state has looked at you and determined that your body works. Now prove it, or stop eating.
The Question
New York’s flyer announces the deadline clearly. March 1, 2026. New rules. New requirements. The visual language is tidy. The institutional grammar is orderly. The message is administrative: here is what you must do to continue receiving food.
What the flyer does not say, because the policy does not require it, is this: the category that will determine whether you eat was not designed to account for your body. It was designed to sort you into a binary that does not reflect how capacity, health, disability, or labor actually work. The exemption process was not designed to find you. It was designed to find the people whose disabilities are visible, documented, and legible to a caseworker in a brief encounter. If your condition is not obvious—if your pain does not show, if your illness cycles, if your impairment is cognitive or psychiatric or chronic rather than acute—the system has already classified you. You are able-bodied. You have three months.
The question is not whether this policy will increase employment. The evidence has answered that. The question is whether a society will continue to let a policy category—a term written into a statute in 1996 and expanded in 2025—decide who is able and who is not, who deserves food and who does not, based on a fiction about bodies that the category itself produces.
The body is not the problem. The category is.