Section 14(c): the subminimum wage for workers with disabilities

This article is about people with intellectual, developmental and other disabilities, many working a few hours a week in nonprofit workshops, and their families. People on every side of it are trying to protect someone they care about, and we have tried to write about it that way.

It also matters to the wider debate: Section 14(c) is where Congress wrote down, in plain words, that a uniform wage floor can shut some people out of work.

How 14(c) works

Section 14(c) of the Fair Labor Standards Act directs the Secretary of Labor, "to the extent necessary to prevent curtailment of opportunities for employment," to provide for certificates allowing employers to pay workers "whose earning or productive capacity is impaired by age, physical or mental deficiency, or injury" at rates below the federal minimum wage. It dates to the FLSA's enactment in 1938.

According to the Labor Department's Wage and Hour Division (WHD), a certificate holder must pay a "commensurate wage," based on the worker's individual productivity compared with experienced workers without disabilities doing essentially the same work in the area. Productivity must be re-measured at least every six months for hourly workers, and a worker or guardian can petition for review by a DOL administrative law judge.

There is, however, no floor beneath the commensurate wage. Congress removed the earlier minimum percentages in 1986. In 2014, the Workforce Innovation and Opportunity Act added Section 511 to the Rehabilitation Act, effective 2016. It requires workers aged 24 or younger to go through vocational rehabilitation and transition steps before they can be paid a subminimum wage, and requires all subminimum-wage workers to receive regular career counseling and referrals.

A program in steep decline

These figures come from the Labor Department's December 2024 proposed rule, which drew on GAO estimates and WHD certificate data.

• 2001: the GAO estimated about 424,000 workers were paid subminimum wages by 5,612 employers holding certificates.
• May 1, 2024: WHD counted 801 employers with an issued or pending certificate, and employers with issued certificates reported paying about 40,579 workers below the minimum. That is roughly a 90 percent fall in workers and an 86 percent fall in employers.
• November 2024: a GAO report put the number at "nearly 40,000," with active certificates in 37 states, down from every state but Vermont in 2014.

According to WHD's 2024 data, about 93 percent of certificate holders were community rehabilitation programs, the nonprofit providers often called sheltered workshops, and only 30 were for-profit businesses. About 91 percent of workers had an intellectual or developmental disability listed as their primary disability. The mean of workers' average hourly earnings was $4.08 and the median was $3.46, for a mean of 11.45 hours a week. Nearly half were paid under $3.50 an hour and about 10 percent $1.00 or less, while about 16 percent earned at least $7.25 (certificates valid October 2023 to March 2024). Very few employers are entering the program: over 97 percent of applications each year are renewals.

States that have ended it

States cannot change the federal program, but they can set their own wage rules. According to the GAO, as of January 2025, 16 states had enacted legislation eliminating the use of 14(c) certificates: Alaska, California, Colorado, Delaware, Hawaii, Illinois, Maine, Maryland, Nevada, New Hampshire, Oregon, Rhode Island, South Carolina, Tennessee, Virginia and Washington. Illinois's Dignity in Pay Act, the most recent, ends the practice on December 31, 2029, and creates a transition program for current certificate holders.

The Labor Department's 2024 proposal counted another nine states that limit subminimum pay without banning it. Kansas, for example, sets a floor of 85 percent of the state minimum. Vermont has no statute, but it stopped funding new sheltered-workshop placements in 2000 and has had no active certificate holders for years. We could not confirm any later enactments from official sources.

The 2020 Civil Rights Commission report

On September 17, 2020, the U.S. Commission on Civil Rights released Subminimum Wages: Impacts on the Civil Rights of People with Disabilities, a 349-page report. The Commission majority found "persistent failures in regulation and oversight" by the Departments of Labor and Justice, and concluded that the program was inconsistent with the civil rights protections people with disabilities are entitled to. It found that people with intellectual and developmental disabilities paid under 14(c) "are not categorically different in level of disability" from people with similar disabilities in competitive integrated employment.

It did not call for an overnight ban. It asked Congress to repeal Section 14(c) with a planned phaseout period, to expand funding for supported employment services, to build capacity in states that are transitioning, and to keep collecting data on how former 14(c) workers fare. Two commissioners dissented from the report's conclusions.

DOL's 2024 rule and its withdrawal

On December 4, 2024, the Labor Department published a proposed rule (89 FR 96466) that would have stopped issuing new 14(c) certificates to employers applying after a final rule took effect, and let existing certificate holders continue for up to three years. DOL cited the program's decline, state laws and the USCCR report, and reported that its own analysis found "no statistical evidence" that employment or labor force participation of people with cognitive disabilities differed in states that had stopped subminimum pay. Comments closed January 17, 2025.

On July 7, 2025, the Department withdrew the proposal (90 FR 29817). It had received over 17,000 comment submissions. The withdrawal turned mainly on legal authority: because the statute says the Secretary "shall" provide for certificates, DOL concluded that it has a mandatory duty and lacks authority "to tear down what Congress has mandated." It also argued that the tens of thousands of workers still in the program suggest "a nonzero population for whom section 14(c) remains necessary."

Current status. The withdrawal ended that rulemaking; any future change would need a new proposed rule. As of October 11, 2026, the Federal Register shows no new 14(c) rulemaking since July 2025. Certificates therefore remain available under DOL's existing regulations at 29 CFR part 525, last substantively revised in 1989.

The Raise the Wage Act

The main bill that would end 14(c) is the Raise the Wage Act of 2025 (H.R. 2743, sponsored by Rep. Bobby Scott, and S. 1332, sponsored by Sen. Bernie Sanders), introduced April 8, 2025. Besides raising the federal minimum to $17, Section 6 would:

• stop DOL from issuing a 14(c) certificate to any employer that did not already hold one when the bill is enacted;
• raise the minimum pay for existing 14(c) workers to $5.00 an hour at the start, then $7.50, $10.00, $12.50 and $15.50 in the following years, reaching the full federal minimum after five years;
• end 14(c) authority entirely after that last step, with DOL offering transition assistance in the meantime.

Only the introduced versions of the two bills have been printed, and we found no record of a committee markup or floor vote. Our pros and cons covers the rest of the bill, and its tipped-wage phase-out is in our tipped minimum wage article.

The debate

The case for ending 14(c). Disability-rights advocates, self-advocates and the USCCR majority see subminimum pay as a civil rights problem, not just a wage problem. Their goal is competitive integrated employment: jobs in ordinary workplaces, at the regular wage, alongside coworkers without disabilities. They argue that 14(c) has been a default path from school into segregated workshops with low expectations, that productivity testing can be demeaning, and that oversight has been weak. The USCCR heard from Melwood, a nonprofit that moved its employees to at least the minimum wage in 2013 and reported higher morale and productivity. Advocates also cite DOL's finding that the employment of people with cognitive disabilities did not measurably drop in states that ended subminimum pay.

The case for keeping it, or slowing down. Some families, guardians and providers, especially of adults with the most significant disabilities, fear that ending 14(c) will mean losing a program, not gaining a job. In the comments DOL summarized in 2025, many said that people with significant disabilities "would face unemployment, underemployment, or loss of ancillary services." For some, a workshop also means structure, friendships and somewhere to go each day.

What the transition data show. A GAO study published April 29, 2025, followed former 14(c) workers in Colorado and Oregon. It supports parts of both views:

• Former workers who found jobs after leaving 14(c) earned far more per hour. In Oregon, average hourly pay rose from $4.01 in 14(c) work to $13.93 in 2019, and in Colorado from $4.33 to $12.74.
• But weekly hours fell, from 16.0 to about 10.8 in Oregon, and only a minority reached competitive integrated employment. In Oregon, 18 percent were in competitive or group employment with Medicaid support as of September 2023, and 54 percent could no longer be tracked in Medicaid data.
• Interviewed former workers and caregivers described more positive than negative experiences, but named problems: finding jobs matching their interests, job coach quality, and how earnings affect SSI and Medicaid.

The GAO cautions that results from two states cannot be generalized, and that the status of people who left the Medicaid data is unknown. The honest reading is that higher pay per hour is real, and so is the risk that some people end up with fewer hours, or none.

Where we come down

We are not arguing that people with disabilities should be paid less. We are asking which tool delivers more work, more income and more dignity.

Start with what 14(c) concedes. Congress wrote a wage floor and, in the same law, a release valve "to prevent curtailment of opportunities for employment." In 2025 the Labor Department described state bans as tradeoffs "between the minimum wage and employment opportunities." That is Milton Friedman's argument: a legal minimum does not make anyone more productive, so people whose output an employer values below the floor risk not being hired.

But the 14(c) remedy is a poor one. It puts the whole cost of the gap on the worker, in the form of a paycheck that can be a few dollars or cents an hour. It ties work mostly to segregated workshops, under oversight the USCCR found weak. We also cannot claim the price-out effect is large here: DOL found no measurable employment drop in states that ended subminimum pay, though the GAO's tracking problems mean that question is not settled.

A ban alone does not fix the underlying problem either. If an employer still values a worker's output at $4 an hour, a ban removes the $4 job; it does not create a $15 one. The better tool has the public, not the worker, pay the difference:

• Wage subsidies and EITC-style earnings support. Let a public subsidy cover the gap between a full wage and what the employer can justify. It is the logic of minimum wage vs. EITC: reward the hire instead of raising its cost.
• Supported employment. Fund job coaches, job development and customized employment, as the USCCR recommended. Coach quality and job matching were exactly the problems former workers raised with the GAO.
• Fix the benefit cliffs. If more earnings put SSI or Medicaid at risk, people will rationally work fewer hours.
• Phase and measure. Any repeal should follow the USCCR's advice: a planned phase-out, transition support, and tracking of every former worker, including those who stop working.

The question is not whether workers with disabilities deserve a full wage. It is who should pay the difference when an employer will not: the worker, through a subminimum paycheck, or the public, through a subsidy that keeps the job open.

A ban without that support risks trading low-paid jobs for no jobs, as families fear. Keeping 14(c) leaves the cost on the people least able to bear it. For the research on wage floors and employment more generally, see our research tracker.