Effects of autism spectrum disorder on parental employment in the United States: evidence from the National Health Interview Survey

ABSTRACT

Using results of the US National Health Interview Survey (US-NHIS) for the years 1998–2013 we investigated the impact of having a child with autism spectrum disorder (ASD) on parents’ work behavior. After controlling for numerous background characteristics, we found that having a child with ASD lowered the number of hours of market work per week and the number of months of market work in the previous year for mothers. For fathers, having a child with ASD also reduced the number of hours of market work per week and the number of months of market work in the previous year. However, the magnitude of the effects were smaller for fathers than for mothers, and more sensitive depending on whether the estimates were derived from a linear regression model or propensity score models. Some evidence was also found that the impact of having a child with ASD on parents’ market work depended on whether or not the child with ASD also had an intellectual disability, the parent’s education level, immigration status, and parent race/ethnicity.

RÉSUMÉ

Cette étude utilise les données américaines de l’enquête nationale sur la santé (US-NHIS) pour analyser l’effet de la présence d’un enfant atteint du trouble du spectre autistique (TSA) sur les décisions des ménages concernant le marché du travail. Après la prise en compte de plusieurs caractéristiques de chaque parent, les résultats indiquent une baisse des heures travaillées par semaine et du nombre de mois employé durant l'année qui précède la période de l'enquête chez la mere. Le déclin des heures hebdomadaires et des mois travaillés est aussi présent chez le père mais les effets sont moins forts et dépendent de la méthode d’estimation employée, régression linéaire ou appariement des coefficients de propension. L’impact de la présence d’un enfant diagnostiqué du TSA sur les décisions d’emploi des parents dépend également de leur niveau d’éducation, leur appartenance ethnique et de la présence d’un handicap mental chez l’enfant autiste.

KEYWORDS:

• Autism spectrum disorders
• parental employment
• national health interview survey
• multinomial logit model

MOTS-CLÉS:

• trouble du spectre autistique
• décision d’emploi du ménage
• enquête nationale sur la santé
• modèle logit multinomial

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes on contributors

Brian McCall received his Ph.D. from Princeton University and is a full professor in the School of Education, Department of Economics, and School of Public Policy at the University of Michigan, Ann Arbor. Current research interests include the effect of tuition subsidies on community college attendance, the impact of family wealth on college choice, and the effect of unemployment insurance rules on unemployment durations and re-employment wages.

Elizabeth Starr received her Ph.D. at the University of Alberta and is a full professor in the Faculty of Education, University of Windsor, Canada. Current research interests include the education of children with ASD, and the impact of having a child with ASD on the family. She was a member of the Ontario Ministry of Education’s Autism Reference Group, which has helped to guide current provincial policy for educating this population.

Notes

1 The other conditions are intellectual disability, other developmental delay, ADHD/ADD, Down syndrome, cerebral palsy, muscular dystrophy, cystic fibrosis, sickle cell anemia, diabetes, arthritis, congenital heart disease, and other heart disease.

2 For families with only one child, that child is always in the random sample.

3 These questions are asked about every child in the Family Core questionnaire. Functional limitations include: vision problem, hearing problem, speech problem, asthma or breathing problem, birth defect, injury, intellectual disability or mental retardation, bone/joint/muscle problem, epilepsy or seizure problem, mental/emotional/ behavioral problem, learning disability, hyperactive/attention deficit disorder, other development problem/cerebral palsy, or other impairment.

4 Based on multiple comparison tests, there was some evidence that the education level of mothers could be collapsed into two categories (BA degree or more, less than a BA degree). The empirical findings reported below however, did not depend on whether two or three categories were used to describe a mother’s education.

5 For these children the response was ‘yes’ to the question ‘are [you/any family members] LIMITED IN ANY WAY in any activities because of physical, mental or emotional problems?’ For the sample of mothers with children with ASD, 831 of 1045 (79.5%) answered ‘yes’ to this question, and for the sample of fathers with children with ASD 579 of 740 (78.2) answered ‘yes’ to this question. The sample was further limited to the 2002–2013 period since the question about the duration of the functional limitation was first asked in 2002.

6 Other estimates using this sample produced qualitatively similar results. For the sake of brevity they are not reported but are available upon request.

7 Other estimates using this sample produced qualitatively similar results. For the sake of brevity they are not reported but are available upon request.

8 When separate linear regression models were estimated by survey year the null hypothesis that the effect of having a reference child with ASD was constant across survey years continued to be rejected at the 5% significance level for mothers’ months worked in the previous calendar year (p-value = .017) but not for fathers’ months worked in the previous year (p-value = .060). For hours worked the null hypothesis continued to not be rejected at the 5% significance level fathers (p-value = .070) but now was rejected for mothers (p-value = .008).