Employment

In the UK, the Office for National Statistics counts autistic people among disabled people aged 16 to 64 who name autism as their main health condition. On that measure, 21.7% were employed in the six months to June 2020, against 53.6% of all disabled people.¹ A year later the figure was 29.0%.² The government's Buckland Review of autism employment, using the national labor force survey, put it at around 3 in 10, against around 5 in 10 of all disabled people and 8 in 10 non-disabled people.³

In Australia, the national disability survey found that 38.0% of autistic people aged 15 to 64 were in the labor force in 2018, and their unemployment rate was 34.1%, against 10.3% for all people with disability and 4.6% for people without.⁴ In 2022, 50.2% were in the labor force, 42.2% were employed (11.4% full time and 29.8% part time), and the unemployment rate was 18.2%. The 2022 report gives 27.3% employed for 2018.⁵

The US has no current national estimate for autistic adults. The best national data, now old, follow young people who received special education. In 2009, at ages 21 to 25, 53.4% of autistic young adults had ever worked for pay since leaving high school and 33.6% were working at the time. That was the lowest figure of the disability groups compared, though not significantly different from young adults with intellectual disability (62.8% had ever worked).⁶ An earlier wave of the same study, at ages 19 to 23, found 55.1% had held a paid job in their first six years after high school, again the lowest of the four groups.⁷

In Denmark, a register study followed everyone born in 1989 to 1991 who was diagnosed before age 12. At 25, excluding students, 27.2% were employed for more than half the year, against 77.7% of matched peers. The rate was 7.9% for those diagnosed with childhood autism and 33.1% for those diagnosed with Asperger syndrome, and 56.9% received public assistance or a pension.⁸

The claim that 85% of autistic adults are unemployed appears in a Wall Street Journal article of March 2014, attributed to "disability experts" with no study named.⁹ A 2017 article on Moneyish, a Dow Jones site, turned it into "85% of college grads affected by autism are unemployed", again without a source.¹⁰ In 2015 the United Nations' World Autism Awareness Day press note said that more than 80 per cent of autistic adults were "unemployed or underemployed"; the Secretary-General's message that year dropped "or underemployed".^11,12 Neither gave a source.

Other common numbers have a source but a narrow one. The UK National Autistic Society's 2016 figures, 16% in full-time paid work and 32% in any paid work, came from an online survey of 2,080 autistic adults or people answering for them; the sample chose itself.¹³ The US figure of 14% comes from autistic adults who use state developmental-disability services, a group more likely than autistic adults in general to have an intellectual disability.¹⁴ Even one national dataset can give two numbers: the 2009 US data above produced 53.4% "ever worked" in one analysis and 58% in a report from the same research center.^6,15

Samples that recruit autistic adults directly, many diagnosed in adulthood, report much higher employment. In a US online survey of 254 autistic adults, 61.4% were employed.¹⁶ In a German clinic sample of 185 adults, nearly all diagnosed as adults, 68.4% of those who had ever worked had a job.¹⁷

In the UK, autistic employees had the largest pay gap of any disability group in 2021: median pay 33.5% below that of non-disabled employees. Taking account of factors such as occupation, qualifications, age and region shrank the gap to 9.9%. Most of the gap is accounted for by differences in occupation, qualifications and similar factors, which may themselves reflect barriers.¹⁸ In 2023 data the gap was 27.9%, or 8.2% after the same adjustment.¹⁹

Graduates show a gap too. Fifteen months after graduating in 2021/22, 43% of autistic UK first-degree graduates were in full-time work, against 60% of graduates with no known disability, and 12% against 5% were unemployed. Autistic graduates had the lowest full-time rate of any disability group.²⁰

Many autistic people who work are underemployed. Of US autistic young adults in 2009 who had worked, 20.9% worked full time in their current or most recent job, at a mean wage of $8.10 an hour.⁶ In an Australian study of 149 autistic adults aged 25 and over, participants were more likely than the general workforce to work part time, to work reduced hours and to be employed below their qualification level.²¹ In the German clinic sample, 31.3% were overeducated for their job, more than in the general population,¹⁷ and in the UK online survey, 51% of those employed said their skills exceeded what the job needed.¹³

Many do not stay in work over time. Among 73 US autistic adults with average-range IQ followed for 12 years, about two thirds took part in paid work or study at some point, but fewer than a quarter kept it up; the women in this small sample were less likely to.²² Over eight years in the Netherlands Autism Register (2,449 adults, average age 42), the largest group, 1,189 people, was out of work throughout, and 801 were in work throughout.²³ In 144 young US autistic adults with jobs at the start of the COVID-19 pandemic, losing a job or having hours or pay cut was followed by worse depressive symptoms.²⁴

A 2012 systematic review of 18 longitudinal studies found that limited cognitive ability, or intellectual disability, was the only factor that consistently predicted lower work participation; other factors gave mixed results.²⁵ In the Danish registers, the difference by diagnosis was large (7.9% employed for childhood autism, 33.1% for Asperger syndrome).⁸

Family resources matter in US data. Young adults from higher-income households, and those with greater everyday functional ability, were more likely to have worked.⁷ Another analysis of the same data linked employment to household income, parental education, gender, social skills, intellectual disability, finishing high school, career counseling in high school, and whether the school had contacted employers or vocational programs.²⁶ In 161 US adults followed for ten years, independence in work and study declined over time, more so in the women in the sample; more independence in daily living went with more independence at work, and receiving more services went, marginally, with more improvement.²⁷

In the eight-year Dutch study, whose participants volunteered for a research register, people who stayed employed differed from those who stayed unemployed: they had fewer autistic traits, were younger, more often male and better educated, had been diagnosed later and had fewer co-occurring conditions.²³ In the US online survey, people who had told their employer about their diagnosis were more than three times as likely to be employed, but the survey cannot show which came first.¹⁶ How disclosure plays out at work is covered in Disclosure at work.

Recent reviews argue that the setting matters as much as the person. A scoping review of 134 studies found that the 36 that tested employment interventions mostly tried to change autistic people's characteristics, with little attention to context, and concluded that environmental factors play a key role.²⁸ In a survey of 687 people in Australia, Sweden and the US, among them 246 autistic people, respondents named stigma, lack of understanding of autism and communication difficulties as key barriers.²⁹ An integrative review of 22 studies concluded that outcomes are strongly influenced by workplace structures, managerial practices and organizational culture, rather than by individual characteristics alone.³⁰ A synthesis of 34 studies of autistic adults' experience argued that traditional, employer-led practices put the burden of making a job work on the worker rather than the employer.³¹ A scoping review of 14 qualitative studies found that misconceptions and workplace cultures that did not include autistic people led to discriminatory practices and fewer opportunities.³⁶

In these samples, more education is not matched by more employment. Besides the UK graduate figures above,²⁰ a German study of 232 adults without intellectual disability, diagnosed late, found unemployment of 25.2% against 5.2% in the general population, although half held a university-entrance qualification, more than the general population. Spells of unemployment averaged 23 months, and participants most often gave interpersonal problems as the reason jobs ended.³²

Services reach few people. One US analysis estimated that only about 1.1% of working-age autistic adults who might need employment services receive them.³³ Autistic students in US vocational rehabilitation received job search, placement and on-the-job supports less often than autistic young adults who had left school, and those who did receive them were more likely to be employed when they left the service.³⁴ In a small US study of young adults who had just left school, those without intellectual disability were about three times as likely as those with it to have no work, study or day program at all.³⁵

In a correspondence test, researchers send matched fictitious applications to real job ads and count the replies. In the US, 6,016 applications to accounting jobs mentioned spinal cord injury, Asperger's syndrome or no disability. Applications that disclosed a disability drew 26% fewer expressions of employer interest (4.87% against 6.58%), though the gap in interview invitations alone was not statistically significant. The gap for Asperger's was about the same size as for spinal cord injury, and it was concentrated among experienced applicants and small private employers.³⁷ In Belgium, paired applications from young graduates to 768 vacancies, about a third of them testing autism, found that candidates who said they were autistic received about half as many positive responses as non-disabled candidates. Telling employers they could receive a wage subsidy did not significantly change this.³⁸

A systematic review of 69 hiring experiments on disability found that most tested wheelchair users, and that autism and Asperger's had each been tested in only one correspondence study, these two.³⁹ How interviews and first impressions affect autistic applicants is covered in Hiring and interviews.

Systematic reviews describe a small, weak evidence base. A 2012 review found five studies of vocational interventions for young autistic people, all of poor quality.⁴⁰ A 2017 review of 50 studies named small samples, lack of randomization and poor measurement of outcomes.⁴¹ A 2021 review update searched 28 databases and found only three controlled studies that met its criteria, too few for a meta-analysis.⁴² Another 2021 review of randomized trials found three in autistic people (214 participants in all), all testing Project SEARCH plus ASD Supports in people aged 18 to 21, none at low risk of bias, and none in autistic adults over 21.⁴³ A 2026 scoping review of vocational training for autistic adults mapped 26 studies from 2015 to 2025, more than half of them from the US, and described the evidence base as expanding but still heterogeneous in how interventions and outcomes are reported.⁵⁹

Project SEARCH plus ASD Supports places students in their last school year in internships at a host business, with job coaches and behavioral support. In a multisite trial of 156 students with significant support needs, 73.4% of the intervention group were in competitive jobs a year after graduation, against 17% of the control group for whom data were available (the paper's wording implies outcome data were not available for every control, which may bias the comparison); employed participants worked about 21 hours a week at about $9.61 an hour.⁴⁴ An earlier, smaller trial by the same group reported similar results.⁴⁵ An older, non-randomized UK study found that 30 verbally able autistic adults in a London supported employment scheme found more work, at higher levels and better paid, than 20 similar adults outside London over two years.⁴⁶

Individual Placement and Support (IPS), which helps people find ordinary jobs quickly and supports them in work, outperformed control conditions in 10 of 11 trials in people with mental illness.⁴³ For autistic people, the evidence so far is an uncontrolled US pilot in which all five participants got jobs within a year,⁴⁷ and a randomized trial of an autism-adapted version in California, with 60 participants planned, registered in 2025.⁴⁸ The team running that trial has also published a 2026 commentary on how hard it is to deliver IPS through California's developmental-disability service system, where clients are referred first to the state Department of Rehabilitation and funding is split across two agencies; the commentary reports no trial results.⁶⁰

Other approaches have mixed results. In a trial of virtual job-interview training for autistic young people (48 against 23), the trained group showed better interview skills; the university receives royalties from the software's maker, and three authors are its employees and shareholders.⁴⁹ A trial of an autism information tool for 84 employers found no difference between groups in the employers' confidence or attitudes.⁵⁰ A systematic review found no strong evidence for or against an "autism advantage" at work, a claim often used to promote hiring programs.⁵¹

US state vocational rehabilitation agencies closed about 17,750 autistic cases in fiscal 2014. Of the roughly 12,100 who received services under an employment plan, 60% left with a job, similar to people with intellectual disability (55%) and other disabilities (56%), but 80% of the jobs were part time and median weekly earnings were $160.⁵² Over ten years, the number of young autistic adults using these services rose while the share employed, their hours and their wages did not improve.⁵³ In another analysis of the same national records, of six factors tested, the total number of services a person received was the only one linked to employment in every year.⁵⁴

Section 14(c) of the US Fair Labor Standards Act lets employers holding a certificate pay some disabled workers below the minimum wage, mostly in sheltered workshops. The US Commission on Civil Rights recommended in 2020 that Congress repeal it with a phase-out period and fund supported employment.⁵⁵ The Government Accountability Office counted about 296,000 workers paid under 14(c) in 2010 and about 122,000 in 2019, and since August 2019 most had earned less than $3.50 an hour; it did not report figures for autism.⁵⁶ The Department of Labor proposed phasing out the certificates in December 2024 and withdrew the proposal in July 2025, citing doubts about its legal authority; the notice gives about 40,579 workers in 2024.⁵⁷

In the UK, the 2024 Buckland Review made 19 recommendations. It said that how many autistic adults use Access to Work, the government's grant for workplace support, was not known, and that "the evidence in some areas is slim or non-existent".³ In Australia, a study of Disability Employment Services with 24 autistic adults, 7 family members and 46 staff found that about a third of the autistic participants (32%) had never been placed in a job through the service, and about half were not working; its authors said the service's effectiveness for autistic clients had not been examined before.⁵⁸

In Sweden, a survey reported in 2026 of 227 autistic adults, mostly aged 25 to 35 and recruited through psychiatric clinics in the southeast of the country, found that fewer than half were employed and about half had received work-related support. The data were collected in 2020. Many described the support as poorly timed or not matched to autism, including some who were in work.⁶¹

Autistic writers often repeat the high unemployment figures. Megan Griffith writes that "85 percent of autistic folks are unemployed", citing a statistics website.^c1 Eric Garcia calls the figure that "75 to 85 percent of college-educated autistic people are unemployed or underemployed" a "commonly cited statistic", noting there is little accurate data on how many autistic adults there are.^c2 Devon Price, in 2022, called 85% "an oft-cited statistic" and wrote that higher-quality research "puts the number closer to 40 percent";^c3 the study cited is the US online survey in which 61% were employed.¹⁶ In 2025, Price described "the roughly 16 percent of us who do manage to hold down a job", drawing on the National Autistic Society;^c4 in that self-selected survey, 16% were in full-time work and 32% in any paid work.¹³ Sol Smith, citing the Buckland Review, writes that autistic people have the largest pay gap of any disability group and are "the population most likely to be overqualified for their job".^c5 Pete Wharmby, an autistic writer and former teacher, gives "around 22 per cent" of autistic adults in any kind of work, which matches the 2020 UK figure,¹ and adds that the data "is probably pretty flawed" because many undiagnosed autistic people manage to work.^c9

Writers describe work as built for other people. Daniel M. Jones writes: "The entire frame of employment is built around neurotypical individuals." Of jobs that end without warning, he adds: "This is all completely normal for us autistic people."^c6 Price describes keeping a job as costly: those who hold one "often do so at the expense of our social lives and emotional well-being", and "masking in order to maintain a job is psychologically perilous".^c4 Wharmby calls the workplace "naturally hostile to autistic people", not by design but because autistic needs "are barely even known about, let alone acknowledged".^c9 How writers describe masking is covered in Masking.

Older writers describe whole working lives shaped this way. In Alex Ratcliffe's collection Our Autistic Lives, Sanana, a nurse diagnosed at 68, recalls 15 to 20 jobs over 50 years, moving on each time co-workers "were starting to see through my mask".^c10 JoBo was sacked after 11 years in electronics "due to my inability to interface well with others", and then took small jobs such as newspaper delivery to stay independent.^c11 Daniel Burford describes the same arc at each employer: "Somewhere between one and two years in, I'd burn out, or be managed out, or hit a ceiling nobody would explain." He writes that he came to see "the flat, jagged, unpromoted career" not as personal failure but as "the standard shape".^c12

Several writers question whether employment should be the goal. Price writes that "the ability to participate in the labor market in no way determines a person's worthiness".^c4 Garcia describes US disability benefits as dividing the "deserving" and "undeserving" poor, and writes that paying autistic people below the minimum wage "sends a message that they can cosplay as workers".^c7 Price argues that workplace neurodiversity groups cannot win changes management opposes, and urges autistic workers to organize in unions instead.^c8 Orion Kelly, an Australian autistic writer, stresses instead what a job can give: "a greater sense of purpose and value", a better quality of life and stronger connections.^c13 Vometia, in Ratcliffe's collection, objects that "We're constantly told that a person's entire worth is based on whether they're in a job".^c14 Jules Edwards, an autistic parent and advocate, writes that US disability benefits leave people "expected to be productive but we may not profit", and that "Subminimum wage and free labor are unacceptable."^c15

endever star, a nonspeaking autistic person whom Eric Garcia met at a retreat for autistic people, told Garcia that "there will always be some of us who simply cannot hold jobs". Star warns: "We need to consciously avoid saying things like 'autistic people can work too if we just have the right supports!'"^c7 Writing to Garcia by email, star argued that people who have never worked, or have not worked "enough", receive only Supplemental Security Income, which leaves them "in inescapable poverty".^c7