Effects of Psychiatric Disorders on Labor Market Outcomes: A Latent Variable Approach Using Multiple Clinical Indicators

Abstract

Standard-Nutzungsbedingungen:Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden.Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen.Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. In this paper, we estimate the effect of psychiatric disorders on labor market outcomes using a structural equation model with a latent index for mental illness, an approach that acknowledges the continuous nature of psychiatric disability. We also address the potential endogeneity of mental illness using covariance instruments as suggested in Lewbel (2012), thus not requiring questionable exclusion restrictions for identification. Data come from the US National Comorbidity Survey -Replication (NCS-R) and the National Latino and Asian American Study (NLAAS). We find that depression and generalized anxiety disorder detract from the employment and labor force participation of males and females; however, we do not find evidence of adverse effects of panic attack or social phobia on any work outcomes of either males or females. After addressing the potential endogeneity of mental illness, we continue to find that mental illness adversely affects employment and labor force participation for both males and females, but the effect on weeks worked and days missed at work are significant for males only. Using our structural model we assess the policy implications of some of the recommendations in the Affordable Care Act, relating to expansion of benefits for mental health and substance use disorder benefits. We find potential gains in employment for 3.2 million individuals and reduction in workplace cost of absenteeism of $18.9 billion due to improved mental health of individuals who are in most need of treatment. Terms of use: Documents inJEL-Code: I100.

The Bilateral Relationship between Depressive Symptoms and Employment Status

“…While some studies report larger effects for women (Chatterji et al 2007;Frijters et al 2014), others find similar (Ojeda et al 2010;Chatterji et al 2011) or even larger effects for men (Banerjee et al 2017).…”

“…Each can lower labor market attachment and reduce employment options. Empirical evidence from representative samples indicates, for example, that poor mental health is associated with participation (Chatterji et al 2011;Banerjee et al 2017) and employment rates (Ettner et al 1997;Alexandre and French 2001;Chatterji et al 2007;Ojeda et al 2010;Frijters et al 2014;Banerjee et al 2017) that are between 10 and 30 percentage points lower.…”

“…Some researchers find that mental health problems reduce the number of hours and weeks worked in a year (Ettner et al 1997;Alexandre and French 2001;Banerjee et al 2017), while others find no effect (Chatterji et al 2007;Chatterji et al 2011;Peng et al 2013). Further, mental health problems, which interfere with work capacity and productivity, may lower earrings.…”

“…A range of IVs have been employed including: i) early onset of mental health issues (Ettner et al 1997;Chatterji et al 2007;Chatterji et al 2011;Banerjee et al 2017); ii) parental mental health (Ettner et al 1997;Chatterji et al 2011); iii) social support (Alexandre and French 2001;Ojeda et al 2010); iv) religiosity (Alexandre and French 2001;Ojeda et al 2 For reviews of the literature regarding the role of health in the labor market more generally see Currie and Madrian (1999), Chirikos (1993), and Barnay (2016). 3 There is limited evidence on the impact of mental health on unemployment.…”

“…Many of these IVs are difficult to defend theoretically, raising concerns about whether the exclusion restrictions necessary to achieve causal identification actually hold (e.g. Ettner et al 1997;Chatterji et al 2007;Ojeda et al 2010;Banerjee et al 2017).…”