Sex differences in employment and supports for adults with autism spectrum disorder

Abstract

This study explored sex differences in employment, reasons for unemployment, benefits, and supports among a large, international sample of adults with autism spectrum disorder. The sample included 443 adults with autism spectrum disorder (60% female; 74% residing in the United States) who consented to be part of an autism research registry and completed an Internet survey. Outcome variables included current employment status, number of hours working, number of jobs in the past 5 years, reasons for unemployment, as well as the number of benefits received and the amount of financial support currently being received from families of origin. Using multiple regression models, we found that males and females were working at similar rates. Females were more likely than males to say that their unemployment was a result of choosing to withdraw from the labor market. Similar percentages of males and females reported receiving some form of benefits or family support, but of those receiving benefits/family support, males received more than females. These results are consistent with other studies finding subtle, but potentially important sex differences in life-course outcomes of individuals with autism spectrum disorder.

Keywords

adults autism spectrum disorders benefits sex differences vocational/labor force participation

Over 2.6 million adults in the United States have been diagnosed with autism spectrum disorder (ASD; Buescher, Cidav, Knapp, & Mandell, 2014), and more than 500,000 of them are women (Baio et al., 2018). The prevalence of ASD diagnoses among children has risen dramatically over the past two decades, with recent estimates suggesting that 1 in every 59 children has an ASD (Baio et al., 2018). As these children age, we anticipate a tidal wave of individuals diagnosed with ASD, many of them women, entering adulthood. Currently, the United States spends approximately US$175 billion per year supporting adults with ASD (Buescher et al., 2014), and that number will very likely rise as more and more youth with an ASD diagnosis enter the adult service system.

Adults with ASD face many vulnerabilities. Employment and post-secondary education (PSE) are difficult to obtain for these individuals (Shattuck et al., 2012; Taylor & Seltzer, 2011, 2012) and are rarely maintained over time (Taylor, Henninger, & Mailick, 2015; Taylor & Mailick, 2014). Studies of adult outcome that take employment, living situation, and friendships/intimate relationships into account overwhelmingly find that the majority of adults with ASD have poor or very poor outcomes (Henninger & Taylor, 2013).

In the general population, women are disadvantaged relative to men in many of these same ways (Hess, Milli, Hayes, & Hegewisch, 2015). Although women now graduate from college at higher rates than men, they are less likely to be employed. Earnings for females average 78 cents to every dollar earned by similarly employed males. Women are more likely than men to have family incomes that place them below the federal poverty line (Hess et al., 2015).

Women with ASD may be doubly vulnerable in these areas, as they are at risk due to both their sex and their disability status. However, our knowledge of specific risk mechanisms for women with ASD is very limited. Since ASD affects 2 to 5 times as many males as females (Lai, Lombardo, Auyeung, Chakrabarti, & Baron-Cohen, 2015), almost all of the information about outcomes for adults with ASD has come from samples that are primarily or exclusively male. Furthermore, because most studies that have included women have used small samples, results are seldom stratified by sex, making it almost impossible to identify the unique challenges these women face or what is required to address such challenges.

A small number of studies, however, provide insight into the unique needs that women with ASD might face in the workplace or in PSE. Specifically, women with ASD seem to obtain post-secondary vocational or educational placements at the same rate as men, but have greater difficulty maintaining those positions over time. In a study of 161 adults with ASD (45 women), 80% of whom had an intellectual disability (ID), Taylor and Mailick (2014) examined how independence in vocational/educational activities changed over 10 years. Although sex did not predict vocational independence at the first time point of the study, it was the only factor that significantly predicted change in vocational independence over time, with females declining at a rate 15 times greater than that of males (Taylor and Mailick, 2014). Similar findings were reported in a sample of 73 adults with ASD who did not have ID (15 women; Taylor et al., 2015). Men and women were equally likely to be working in the community or attending PSE at a single point in time. However, there was a significant sex difference in patterns of PSE/employment over time. Specifically, none of the women in this sample were consistently engaged in employment or PSE across the study, compared to 31% of the men. Note that in both studies, there were no differences between men and women in their autism symptoms, functional abilities, or family functioning. Furthermore, none of the women were married or had children (which could account for different employment patterns). Thus, sex differences in vocational activities could not be attributed to these factors.

The goal of this study is to explore potential factors that might help explain sex differences in maintaining post-secondary employment and educational activities for adults with ASD. Using data from the Interactive Autism Network, we examined sex differences among these adults in (1) employment and reasons for unemployment and (2) receipt of benefits and family financial support. These analyses will provide support for future hypotheses that can be tested to understand sex differences in life course outcomes for individuals with ASD and inform our understanding of the unique supports that may be needed by females with ASD.

Method

Sample and methods

The Interactive Autism Network (IAN: Johns Hopkins Medicine Institutional Review Board (IRB) NA_00002750; principal investigator (PI): Dr. Paul H. Lipkin), is an international, US-based, online research database and registry for individuals with ASD and their families. IAN’s revised “Independent Adult with ASD Questionnaire” was launched on 5 January 2017. The survey was made available to self-consenting, independent adults with ASD who had joined IAN. Eligible participants received email notifications about the survey and were also informed about the survey through IAN eNewsletters. A total of 529 independent adults with ASD completed the self-report online survey through 18 December 2017. Participants who had completed the survey were selected for the current analyses if (1) they indicated that they had exited high school (n = 518), (2) their sex at birth matched their current gender identity (n = 451), and (3) they reported receiving a diagnosis of ASD from a professional or they had self-reported autism symptoms on the Autism Spectrum Quotient (AQ; Baron-Cohen, Wheelwright, Skinner, Martin, & Clubley, 2001) that indicated a strong likelihood of ASD (score of 32 or higher; n = 443). We included those without a formal diagnosis but with elevated ASD symptoms (n = 10), as formal ASD diagnoses can be difficult to obtain in adulthood (Baron-Cohen et al., 2001). Validation of the AQ found that a cut-off of 32 identified 80% of those with ASD, with a 2% false-positive rate (Baron-Cohen et al., 2001).

Present analyses included 443 adults with ASD: 176 males and 267 females (60.3% female). Demographic information for the sample is presented in Table 1. The sample was primarily White; 3.6% were Asian, 1.8% were Black, 6.1% reported multiple races, and 3.8% reported another race or chose not to report on race. The majority of the sample was non-Hispanic. Most adults lived in the United States, with the next most frequent countries of residence being the United Kingdom (5.9%), Canada (4.1%), Australia (3.2%), and the Netherlands (2.9%). Nearly all (97.7%, n = 433) had received an ASD diagnosis from a professional, with most of those diagnosed as adults (aged 18 years or older). Most adults reported that they had been diagnosed with or treated by a professional for a co-occurring psychiatric disorder at some point in their lifetime. Information on education, marital status, and parenting status are presented in Table 1.

Table 1.

Demographics and background variables by sex.

Demographic/background variable	Total sample, M (SD) or %	Male, M (SD) or %	Female, M (SD) or %	t-test or chi-square for sex differences
Current age (years)	40.40 (13.96)	41.70 (15.06)	39.54 (13.15)	1.60
Race (White)	85.2%	85.2%	85.2%	0.00
Hispanic	5.0%	4.0%	5.7%	0.63
Currently living in the United States	74.3%	71.0%	76.4%	1.61
Diagnosed with ASD before the age of 18 years	25.9%	34.1%	20.5%	9.87**
Earned bachelor’s degree	51.2%	46.8%	54.0%	2.16
Currently married	40.0%	35.2%	43.1%	2.72
Currently living with family of origin	25.3%	33.3%	20.0%	9.88*
Have children	40.2%	38.1%	41.6%	1.90
Currently living with children	24.8%	19.0%	28.7%	5.31*
Ever diagnosed with a co-occurring psychiatric disorder	88.7%	80.7%	94.0%	18.68**

p < 0.05; **p < 0.01.

Independent adult with ASD questionnaire key questions

Employment/PSE

Adults were asked whether they currently participated in paid work activities (0 = no; 1 = yes), as well as whether they were currently attending school and what type of school they were attending. For analysis, all types of PSE programs (vocational or trade school, college, graduate or professional school, other) were combined into an indicator of “PSE participation” (0 = not currently participating in PSE; 1 = currently participating in PSE). Information on PSE and work participation was combined into an overall category indicating current participation in work or PSE (0 = not currently participating in paid work or PSE; 1 = currently participating in paid work or PSE). Adults reported on the number of jobs they had in the past 5 years (0 = none; 1 = 1–3; 2 = 4–6; 3 = 7–9; 4 = 10–12; 5 = 13–15; 6 = 16 or more). Those who were working for pay were asked the number of hours they were working (1 = 1–9 h; 2 = 10–19 h; 3 = 20–29 h; 4 = 30–39 h; 5 = 40 or more hours) and whether they would like to work more hours than they currently are able to get (0 = no; 1 = yes, would like to work more hours).

Those who were not working for pay were asked to report on why they were not working. Adults could endorse any of the following reasons: unemployed—want to work but cannot find work, do unpaid/volunteer work instead, have tried to work but faced discrimination or other difficulties with employers because of their ASD, do not wish to work at present, not able to work because it would interfere with federal or state benefits, not able to work because the workplace would be too challenging because of their ASD, and other. Each reason was coded as 0 (unendorsed) or 1 (endorsed).

Family financial support

The degree of family financial support was assessed through four questions. Respondents were asked “How much, if any do you rely on your extended family (such as parents and siblings) for…”: housing, transportation, health care, and school/education costs. For each area, adults could respond on a scale of 0 (my family does not provide any support for me in this area) to 4 (my family provides all of my support in this area). We added scores in each domain to generate a total amount of family financial support (possible range of 0–16, higher scores indicated more family financial support).

Benefits

Respondents were asked whether they received any federal or state benefits (0 = no; 1 = yes). For those who reported receiving benefits, nine specific benefits were queried, to which participants could response with 0 (not receiving benefit/service) or 1 (receiving benefit/service): Social Security Disability Income (SSDI), Supplemental Security Income (SSI), state disability programs that only use state and/or local funds, Medicaid (for health insurance), Medicare, Medicaid Home- and Community-Based Services (HCBS) waiver or Developmental Disability waiver, employment assistance or on-the-job support (sometimes called Vocational Rehabilitation), Section 8 Housing, and transportation for people with disabilities. For analysis, we calculated a number of total benefits received (possible range of 0–9).

Demographic information

Adults with ASD reported their age at the time of data collection as well as age at first ASD diagnosis. Given the small percentage of sample members diagnosed with ASD during childhood, we collapsed age at diagnosis into a dichotomous variable indicating childhood versus adult diagnosis (0 = diagnosed prior to the age of 18 years; 1 = diagnosed at the age of 18 years or older). Race and ethnicity information was collected, and collapsed for analysis into White/non-White and Hispanic/non-Hispanic. Information on education was collected and coded into 0 = did not attain a bachelor’s degree and 1 = attained a bachelor’s degree. Information on current marital and parental status was collected and coded into whether the adult was currently married or in a long-term relationship (yes/no) and whether he or she was a parent to a biological, adopted, or stepchild (0 = no children; 1 = children). Data on current living arrangement were collected and coded as the following dichotomous variables (yes/no for each): living with family of origin (parents/siblings/other relatives) and living with children.

Adults were asked whether they had ever been diagnosed by a professional for any of the following psychiatric conditions: attention-deficit hyperactivity disorder or attention-deficit disorder (ADHD/ADD), anxiety, obsessive–compulsive disorder (OCD), depression/major depressive disorder, bipolar disorder/manic depressive disorder, oppositional defiant disorder (ODD), and schizophrenia. Conditions were collapsed into an indicator of whether the adult had any co-occurring psychiatric disorder (0 = no psychiatric disorder; 1 = any psychiatric disorder).

Data analysis

Independent t-tests (continuous variables) and chi-squares (categorical variables) were used to test for sex differences in demographic/background variables including whether adults were living in the United States (yes/no), age, race (White/non-White), ethnicity (Hispanic/non-Hispanic), age at diagnosis (above vs below 18 years), bachelor’s degree (yes/no), currently married (yes/no), currently living with family of origin (yes/no), has children (yes/no), currently living with children (yes/no), and any diagnosis of a co-occurring psychiatric disorder (yes/no).

We next conducted preliminary analyses to determine whether we should combine data collected from those adults living in the United States (about two-thirds of the sample) with those not living in the United States. Information on these analyses and the results are presented in the Supplemental Material. We observed an overall pattern of similarities between the subgroups, and thus, we analyzed the whole sample.

Regression analyses were used to test for sex differences in the percent of adults with ASD currently working or in PSE, the number of jobs in the past 5 years, and (for those working) the number of hours working in a regular week and the percent who would like to work more hours. For those who were not working, regression analyses were used to test for sex differences in each of the seven reasons for unemployment. Binary logistic regression was used for dichotomous outcomes, and ordinal logistic regression was used for ordinal outcomes. In all analyses, the age of the adult with ASD, whether he or she lived in the United States (yes/no), and race/ethnicity (White, non-Hispanic vs others) were statistically controlled. Three participants (0.7% of the sample) were missing data on race and/or ethnicity. Following well-established guidelines (Harrell, 2001), we substituted the most frequent category (i.e. White, non-Hispanic) for those missing cases.

For family financial support, over one-half of participants had a score of “0” (indicating no support). Thus, we calculated two variables for analysis: (1) a dichotomous indicator of whether the adult was receiving any financial support (0 = no support, 1 = any financial support) and (2) for those who were receiving family financial support, the amount of support they were receiving. We observed the same pattern for the number of benefits (i.e. a large percentage who was receiving no benefits) and calculated the same two indicators for that construct (i.e. dichotomous indicator of any vs no benefits, number of benefits for those receiving benefits). Regression models were used to examine the effects of sex on these outcomes, with binary logistic regression as the model for dichotomous outcomes and ordinal logistic regression for ordinal outcomes. The same variables were statistically controlled in all models (age, living in the United States vs living elsewhere, White, non-Hispanic vs other). Note that to examine sex differences in benefits, we only included those individuals who lived in the United States (as most of the benefits questions were US-centric; n = 125 males and 204 females).

Results

Demographic/background variables

Sex differences in demographic and background variables are presented in Table 1. There were no statistically significant sex differences in the age, race, or ethnicity of the adults with ASD, whether they were living in the United States or the percentage who had completed a bachelor’s degree or were currently married. Females were more likely than males to be diagnosed at age 18 or older, with 80% of women and two-thirds of men diagnosed as adults. Males were more likely to be living with their family of origin. Although males and females were similarly likely to have children, females were more likely than males to be living with children. Lifetime rates of co-occurring psychiatric disorders were extremely high for both sexes, but significantly higher for females, who nearly all (94%) reported being diagnosed with a co-occurring disorder.

Employment and reasons for unemployment

Results from the regression models examining the relations between sex and employment variables are presented in Table 2. Approximately two-thirds of both males and females were employed or in PSE at the time of data collection (67.6% of males vs 66.2% of females; unadjusted percentages); the effect of sex on current employment or PSE participation was not statistically significant. Males and females did not significantly differ in the number of jobs held in the past 5 years (unadjusted medians corresponding to 1–3 jobs for both males and females). Of the 102 males and 135 females who were employed, there were no sex differences in the number of hours currently working, but there was a significant sex difference in the likelihood of reporting that they would like to work more hours; the odds of a female endorsing this response were about one-half that of males.

Table 2.

Results of regression models examining relations between sex and employment variables.

Employment variable	B	SE B	Wald chi-square	Odds ratio	95% CI for the odds ratio
Working or in PSE (d)	−0.17	0.22	0.61	0.85	0.55–1.29
Number of jobs in the past 5 years	−0.26	0.20	1.73	0.77	0.53–1.13
For those employed
Number of hours currently working	0.07	0.24	0.07	1.07	0.66–1.72
Would like to work more hours (d)	−0.66	0.28	5.40*	0.52	0.30–0.90
For those unemployed (all d)
Want to find work but cannot	−0.71	0.35	4.08*	0.49	0.25–0.98
Do unpaid/volunteer work instead	−0.12	0.37	0.10	0.89	0.43–1.83
Face discrimination/other difficulties with employers	0.50	0.38	0.02	1.05	0.50–2.21
Do not wish to work at present	1.13	0.40	8.22**	3.10	1.43–6.73
Work would interfere with benefits	−0.11	0.39	0.08	0.89	0.42–1.93
Workplace would be too challenging	−0.05	0.30	0.03	0.95	0.53–1.72
Other	0.17	0.42	0.17	1.19	0.53–2.67

d: dichotomous outcome; SE: standard error; CI: confidence interval.

All regression models control for age, whether the adult lives in the United States and whether he or she is White, non-Hispanic. Binary logistic regression models were used for dichotomous outcomes and ordinal regression models were used for ordinal outcomes.

p < 0.05; **p < 0.01.

Reasons for unemployment were asked of those not currently working; results from the regression models are presented in Table 2, and raw percentages of participants endorsing each response stratified by sex are presented in Table 3. The most common reason for not working reported by both sexes, at nearly 50% of those unemployed, was that the workplace was too challenging for them because of their ASD. The odds of unemployed females reporting that they would like to find work but could not were about one-half that of males; this sex difference was statistically significant. Alternatively, the odds of females reporting that they were not working because they do not wish to work at present were over 3 times that of males (see Table 3). The remaining reasons for unemployment did not significantly differ by sex.

Table 3.

Percentages of those unemployed who endorsed each reason for unemployment (unadjusted).

Reason for unemployment	Male (%)	Female (%)
Want to find work but cannot	30.6	19.1
Do unpaid/volunteer work instead	22.2	19.8
Face discrimination/other difficulties with employers	19.4	20.6
Do not wish to work at present	13.9	34.4
Interfere with benefits	18.1	18.3
Workplace would be too challenging	45.8	45.8
Other	19.4	17.6

Accounting for potential effects of child-rearing

Compared to most other studies of employment in adults with ASD (e.g. Farley et al., 2009; Howlin, Moss, Savage, & Rutter, 2013; Taylor & Seltzer, 2012), the current sample was more likely to be married and have children. It may be that the higher percentage of females than males choosing to withdraw from the workforce could be accounted for by women who stay home to take care of children. To test this hypothesis, we excluded adults who had children living at home and tested whether sex differences in choosing to withdraw from the workforce remained among those without children in the home. Within this sub-group, the odds of unemployed females saying that the reason for their unemployment was because they chose not to work were 2.3 times that of males, a difference that was marginally statistically significant: Wald χ² (1, N = 158) = 3.65, p = 0.056, and odds ratio = 2.30 (95% confidence interval (CI): 0.98–5.39).

Family financial support and benefits

Results of regression models examining sex differences in support and benefits are presented in Table 4. Males and females were similarly likely to be receiving any financial support from their families of origin; unadjusted percentages were 47.1% of males and 43.0% of females who were receiving any support. Of those receiving support, however, females received significantly less family financial support than males. In terms of benefits, males and females living in the United States were similarly likely to be receiving any state or federal benefits, with unadjusted percentages at about one-third for both sexes (36.7% of men vs 33.8% of women). However, of those getting benefits, females received significantly fewer benefits than males.

Table 4.

Results of regression models examining relations between sex and support/benefit outcomes.

Support/benefit variable	B	SE B	Wald chi-square	Odds ratio	95% CI for the odds ratio
Receiving financial support from families (d)	−0.38	0.23	2.81	0.68	0.44–1.07
Amount of support (for those receiving support)	−0.70	0.26	7.41**	0.50	0.30–0.82
Receiving benefits (d)	−0.12	0.24	0.26	0.88	0.55–1.42
Number of benefits (for those receiving benefits)	−0.81	0.36	5.18*	0.45	0.22–0.89

p < 0.05; **p < 0.01.

d: dichotomous outcome; SE: standard error; CI: confidence interval.

Accounting for potential effects of living with families of origin

Residential financial support is a component of the family financial support measure, and males were more likely than females to be living with their families of origin (see Table 1). Thus, we examined whether the financial support inherent in living with parents could explain why males reported more family financial support than females. To do this, we reran the ordinal regression models testing for sex differences in family financial support, but just for those males and females receiving support who were living with their families of origin. Among adults living with parents, females received significantly less family financial support than did males, Wald χ² (1, N = 105) = 7.28, p < 0.01, odds ratio = 0.39 (95% CI: 0.20–0.77).

Accounting for potential effects of age of diagnosis

Finally, we examined whether differences between males and females in the number of benefits received could be explained by the greater proportion of males (vs females) diagnosed prior to the age of 18 years (see Table 1), since those diagnosed in childhood might have better access to services in adulthood. Thus, we reran our ordinal regression model examining the number of benefits (for those receiving benefits) for just those males and females who were diagnosed prior to the age of 18 years. Among those diagnosed in childhood, females received significantly fewer benefits than did males, Wald χ² (1, N = 37) = 6.54, p < 0.05, and odds ratio = 0.17 (95% CI: 0.04–0.66).

Discussion

This study adds to the literature by suggesting subtle yet potentially important differences between men and women with ASD in employment, formal services, and family financial support. The pattern of findings in this sample is consistent with findings from other studies examining sex differences in employment. In the Taylor and Mailick studies (Taylor et al., 2015; Taylor & Mailick, 2014), sex differences were not observed when looking at rates of employment/PSE or independence in vocation at a single point in time, but only when digging beneath the surface to examine patterns of employment/PSE. Similarly, in the present analyses, there were no overall sex differences in rates of employment/PSE participation, but instead there were differences in attitudes toward employment (thoughts about hours working, reasons for unemployment). There were also no overall sex differences in the percentages of those receiving any benefits or family financial support, but instead in the amount of assistance received (for those receiving support/benefits). Thus, it appears that sex differences for adults with ASD might not be observed when examining broad indicators of adult outcomes, but instead when delving deeper to examine outcomes in a more fine-grained manner.

This similar pattern of findings across studies of nuanced sex differences in outcomes is all the more striking given the vast differences in sample characteristics. The studies by Taylor, Mailick, and colleagues (Taylor et al., 2015; Taylor & Mailick, 2014) were drawn from a sample of adolescents and adults with ASD recruited in 1998, who all had received ASD diagnoses as children before the widening of the autism diagnostic criteria in the Diagnostic and Statistical Manual of Mental Disorders (4th ed., DSM-IV; American Psychiatric Association., 1994). About 75% of individuals with ASD in that sample had a co-occurring ID (though a subsample without ID was drawn for the 2015 paper). Furthermore, parents were the primary reporters for the Taylor and colleagues studies, removing the necessity of the person with ASD having fluent communication for study participation. The IAN sample utilized in this study was very different—adults in this sample were all capable of completing an Internet survey. Most of the sample members were diagnosed with ASD as adults, and they were much more likely than individuals with ASD in the Taylor, Mailick, and colleagues’ studies to be working, married, raising children, and have completed a bachelor’s degree. Yet, across these two samples of adults with ASD with very distinct characteristics, there were no sex differences when examining gross indicators of employment/support but rather significant differences emerged when probing more fine-grained variables (e.g. reasons for unemployment, amount of support for those getting support).

When exploring reasons for unemployment, we found that females were more likely than males to indicate that they are choosing to withdraw from the labor market. Childrearing and domestic responsibilities are—by far—the most common reason why women in the general population leave the labor force (e.g. Garcia-Manglano, 2015; Krueger, 2017; Spain & Bianchi, 1996). Accordingly, some of the women with ASD in this study may be withdrawing from the labor market to stay home and care for children. However, this pattern of sex differences held to some extent among adults with ASD who did not have children living at home. That is, women with ASD who did not have children were more likely than men to say that they did not wish to work at present. This stands in contrast to gendered patterns of labor force participation in the general population: using data from population-based surveys, Krueger (2017) found that after removing women who withdrew from the labor force because of home responsibilities (the vast majority being child-related), the percentages of men and women who were not in the labor force were similar. Thus, there may be something unique about the experiences of women with ASD that is leading them to withdraw from the labor market at a higher rate—even when they do not have children living at home. This greater likelihood of choosing to withdraw from the labor market for women with ASD (vs men) may help explain findings from earlier studies suggesting that these women experience greater post-secondary vocational and educational instability relative to men.

Future research should more carefully examine the employment experiences of women with ASD compared to typically developing women, to better understand the ways in which sex differences in ASD reflect typical sex patterns and the ways in which they are different. One hypothesis is that women with ASD are doubly vulnerable in the workplace, both by virtue of being female and having an ASD. Sociologists commonly write about the gendered nature of many workplaces, in which females (vs males) are more likely to experience discrimination, inflexibility, and have limited opportunities for advancement (e.g. Cahusac & Kanji, 2014; Zimmerman & Clark, 2016). Added to that is the risk factor of having an ASD; numerous studies have showed that adults with this diagnosis experience high rates of unemployment and underemployment (e.g. Shattuck et al., 2012; Taylor & Seltzer, 2011, 2012)). Combining these two factors might make the workplace more difficult for women with ASD compared to men with ASD and women in the general population. Another potential explanation borrowed from the typically developing literature relates to societal gender norms (Zimmerman & Clark, 2016); it might be that women with ASD are less likely than men with ASD to derive identity from their employment and work-based achievements and thus might more easily disengage from the labor force. Future research should continue to examine the perspectives of adults with ASD around their employment and unemployment experiences, to understand how their perspectives might relate to stability in these activities over time. It will also be important to understand to what extent these sex differences are specific to ASD versus shared with the general population.

It is important to note that for both males and females in this sample, the most common reason for not working was the belief that the workplace would be too challenging for them because of their ASD. Nearly 50% of those not working endorsed this idea, which is particularly surprising given that this is a relatively capable sample of adults with ASD, with more independent adult outcomes than many other samples (Gotham et al., 2015; Henninger & Taylor, 2013; Howlin et al., 2013). These findings corroborate more recent qualitative analyses of the self-reported difficulties of adults with ASD in the workplace (e.g. Sosnowy, Silverman, & Shattuck, 2018; Tint & Weiss, 2018) and point to the need to better understand how to make workplaces more accommodating and accessible to these adults.

Our findings related to support and benefits suggest that women with ASD are equally likely as men to receive some form of support, but that once they “get their foot in the door” in terms of benefits and supports, they tend to receive less overall support than do males. In our analyses, this finding was consistent both for formal benefits and services (e.g. Medicaid) and for informal financial support from parents and other family members. Because most samples of adults with ASD include a small number of females, almost nothing is known about the characteristics of women with ASD versus men who receive support. Our findings point to three potential hypotheses that could be tested in future research: (1) it may be that females are getting less support than males because they have fewer support needs; (2) it may be that females have the very same support needs as males, but are less likely to have those needs recognized by family or service providers and thus get less support; or (3) it may be that females need different supports than males, and those types of supports are less/unavailable. To understand how to better support females with ASD, further research is needed to examine how their support needs differ from and are similar to the support needs of males.

Finally, it is worth nothing that adults with ASD who lived in and out of the United States seemed to have similar life circumstances in terms of education, marital and parenting status, employment experiences, and family support (see Supplemental Materials). Because the majority of IAN participants were from the United States, we collapsed across non-US countries. Future research should examine how life circumstances of adults with ASD might differ across non-US countries.

As in any study, this study has a number of limitations that must be taken into account. Self-reported diagnoses of ASD by IAN sample members were not corroborated through gold-standard autism diagnostic procedures. This limitation is addressed, at least in part, by the inclusion of participants in these analyses who reported an ASD diagnosis by a medical or educational professional and by using the AQ (Baron-Cohen et al., 2001) as an indicator of likely ASD. Generalizability of the sample must be considered, as results are only potentially generalizable to those who have the capabilities and access to be registered in the IAN registry and complete an Internet survey. The significant proportion of females in this sample, relative to the proportion of females with ASD in epidemiological studies (Lai et al., 2015), suggests further limits to generalizability. Note that the IAN adult survey was not designed to examine sex differences and women were not over-recruited—the relatively large sample of women reflects those who signed up for the IAN registry and completed the survey. It is also consistent with other Internet surveys of adults with ASD (e.g. McConachie et al., 2018; McDonald, 2017).

Though the information on employment and job history in IAN was more detailed than many other large-scale Internet surveys of individuals with ASD—going beyond present work status to include information on the number of jobs, duration of employment, and reasons for not working—more extensive information on employment and workplace participation is needed to more fully understand how the employment experiences of females with ASD differ from their male counterparts. Future research should include information on the types of jobs that females and males hold, the problems that they encounter in the workplace, and the reasons for job termination. These “deep dive” variables will be helpful in understanding the unique experiences of women with ASD in the workplace, ultimately informing interventions and programs to meet their specific needs.

In an attempt to keep the IAN survey accessible to as many adults with ASD as possible, detailed information on functioning (e.g. IQ, adaptive behavior, psychiatric symptoms, family functioning) that would inform generalizability are not available. Though generalizability concerns exist, as noted above, patterns of findings from this study were consistent with other studies that included very different samples of adults with ASD with different methodologies. These limitations are offset by the significant strength of the IAN sample, which is multi-national and large, allowing for the investigation of more nuanced questions than would be possible in a smaller dataset.

In conclusion, findings from the present analyses point to some specific areas of risk that might be unique to females with ASD (vs males). It may be that females are more likely to choose to withdraw from the labor market; future research should examine whether this finding can be replicated in other samples and, if so, reasons for choosing to withdraw and consequences of withdrawal for employment stability, earnings, and other life course outcomes. Women who are receiving benefits and supports might be “falling through the cracks,” ultimately receiving less support than their male counterparts. Further research with large samples is needed to examine whether autism and related characteristics of women with ASD have a greater or lesser impact than males on their life course outcomes. Ultimately, we expect that developing a more careful understanding of the experiences of women versus men with ASD across the lifespan will lead to the development and implementation of tailored supports to meet their unique needs.

Supplemental Material

AUT827417_Lay_Abstract – Supplemental material for Sex differences in employment and supports for adults with autism spectrum disorder

Supplemental material, AUT827417_Lay_Abstract for Sex differences in employment and supports for adults with autism spectrum disorder by Julie Lounds Taylor, Leann Smith DaWalt, Alison R Marvin, J Kiely Law and Paul Lipkin in Autism

Supplemental Material

AUT827417_Supplemental_material – Supplemental material for Sex differences in employment and supports for adults with autism spectrum disorder

Supplemental material, AUT827417_Supplemental_material for Sex differences in employment and supports for adults with autism spectrum disorder by Julie Lounds Taylor, Leann Smith DaWalt, Alison R Marvin, J Kiely Law and Paul Lipkin in Autism

Footnotes

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the National Institute of Mental Health (R03 MH112783, PIs: DaWalt & Taylor) with core support from the National Institute of Child Health and Human Development (U54 HD083211, PI: Neul; U54 HD090256, PI: Messing). The Interactive Autism Network is a partnership project between the Kennedy Krieger Institute and the Simons Foundation and receives funding from PCORI as a Patient-Powered Research Network in PCORnet.

Supplemental material

Supplemental material for this article is available online.

References

American Psychiatric Association. (1994). Diagnostic criteria from DSM-IV. Washington, DC: Author.

Baio

Wiggins

Christensen

Maenner

M. J.

Daniels

Warren

. . . Dowling

N. F.

(2018). Prevalence of autism spectrum disorder among children aged 8 years—Autism and developmental disabilities monitoring network, 11 sites, United States, 2014. Morbidity and Mortality Weekly Report: Surveillance Summaries, 67, 1–23.

Baron-Cohen

Wheelwright

Skinner

Martin

Clubley

(2001). The autism-spectrum quotient (AQ): Evidence from Asperger syndrome/high-functioning autism, males and females, scientists and mathematicians. Journal of Autism and Developmental Disorders, 31, 5–17. doi:10.1023/A:1005653411471

Buescher

A. S.

Cidav

Knapp

Mandell

D. S.

(2014). Costs of autism spectrum disorders in the United Kingdom and the United States. JAMA Pediatrics, 168, 721–728. doi:10.1001/jamapediatrics.2014.210

Cahusac

Kanji

(2014). Giving up: How gendered organizational cultures push mothers out. Gender, Work & Organization, 21, 57–70. doi:10.1111/gwao.12011

Farley

M. A.

McMahon

W. M.

Fombonne

Jenson

W. R.

Miller

Gardner

. . . Coon

(2009). Twenty-year outcome for individuals with autism and average or near-average cognitive abilities. Autism Research, 2, 109–118. doi:10.1002/aur.69

Garcia-Manglano

(2015). Opting out and leaning in: The life course employment profiles of early baby boom women in the United States. Demography, 52, 1961–1993. doi:10.1007/s13524-015-0438-6

Gotham

Marvin

A. R.

Taylor

J. L.

Warren

Anderson

C. M.

Law

P. A.

. . . Lipkin

P. H.

(2015). Characterizing the daily life, needs, and priorities of adults with autism spectrum disorder from Interactive Autism Network data. Autism, 19, 794–804. doi:10.1177/1362361313481860

Harrell

F. E.

(2001). Regression modeling strategies: With applications to linear models, logistic regression, and survival analysis. New York, NY: Springer.

10.

Henninger

N. A.

Taylor

J. L.

(2013). Outcomes in adults with autism spectrum disorders: A historical perspective. Autism, 17, 103–116. doi:10.1177/1362361312441266

11.

Hess

Milli

Hayes

Hegewisch

(2015). The status of women in the states: 2015. Washington, DC: Institute for Women’s Policy Research.

12.

Howlin

Moss

Savage

Rutter

(2013). Social outcomes in mid- to later adulthood among individuals diagnosed with autism and average nonverbal IQ as children. Journal of the American Academy of Child and Adolescent Psychiatry, 52, 572–581. e1. doi:10.1016/j.jaac.2013.02.017

13.

Krueger

A. B.

(2017, September 7). Where have all the workers gone? An inquiry into the decline of the U.S. labor force participation rate. Brookings Papers on Economic Activity. Retrieved from https://www.brookings.edu/bpea-articles/where-have-all-the-workers-gone-an-inquiry-into-the-decline-of-the-u-s-labor-force-participation-rate/

14.

Lai

M. C.

Lombardo

M. V.

Auyeung

Chakrabarti

Baron-Cohen

(2015). Sex/gender differences and autism: Setting the scene for future research. Journal of the American Academy of Child and Adolescent Psychiatry, 54, 11–24. doi:10.1016/j.jaac.2014.10.003

15.

McConachie

Mason

Parr

J. R.

Garland

Wilson

Rodgers

(2018). Enhancing the validity of a quality of life measure for autistic people. Journal of Autism and Developmental Disorders, 48, 1596–1611. doi:10.1007/s10803-017-3402-z

16.

McDonald

T. A. M.

(2017). Discriminative and criterion validity of the autism spectrum identity scale (ASIS). Journal of Autism and Developmental Disorders, 47, 3018–3028. doi:10.1007/s10803-017-3221-2

17.

Shattuck

P. T.

Narendorf

S. C.

Cooper

Sterzing

P. R.

Wagner

Taylor

J. L.

(2012). Postsecondary education and employment among youth with an autism spectrum disorder. Pediatrics, 129, 1042–1049. doi:10.1542/peds.2011-2864

18.

Sosnowy

Silverman

Shattuck

(2018). Parents’ and young adults’ perspectives on transition outcomes for young adults with autism. Autism, 22, 29–39. doi:10.1177/1362361317699585

19.

Spain

Bianchi

S. M.

(1996). Balancing act: Motherhood, marriage, and employment among American women. New York, NY: Russell Sage Foundation.

20.

Taylor

J. L.

Henninger

N. A.

Mailick

M. R.

(2015). Longitudinal patterns of employment and postsecondary education for adults with autism and average-range IQ. Autism, 19, 785–793. doi:10.1177/1362361315585643

21.

Taylor

J. L.

Mailick

M. R.

(2014). A longitudinal examination of 10-year change in vocational and educational activities for adults with autism spectrum disorders. Developmental Psychology, 50, 699–708. doi:10.1037/a0034297

22.

Taylor

J. L.

Seltzer

M. M.

(2011). Employment and post-secondary education activities for young adults with autism spectrum disorders during the transition to adulthood. Journal of Autism and Developmental Disorders, 41, 566–574. doi:10.1007/s10803-010-1070-3

23.

Taylor

J. L.

Seltzer

M. M.

(2012). Developing a vocational index for adults with autism spectrum disorders. Journal of Autism and Developmental Disorders, 42, 2669–2679. doi:10.1007/s10803-012-1524-x

24.

Tint

Weiss

J. A.

(2018). A qualitative study of the service experiences of women with autism spectrum disorder. Autism, 22, 928–937. doi:10.1177/1362361317702561

25.

Zimmerman

L. M.

Clark

M. A.

(2016). Opting-out and opting-in: A review and agenda for future research. Career Development International, 21, 603–633. doi:10.1108/Cdi-10-2015-0137.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.50 MB

0.29 MB