Elsevier

Social Science & Medicine

Volume 146, December 2015, Pages 75-84
Social Science & Medicine

Does a college education reduce depressive symptoms in American young adults?

https://doi.org/10.1016/j.socscimed.2015.09.029Get rights and content

Highlights

  • Whether a college education is protective over depressive symptomology is unclear.

  • We estimate this association using a monozygotic discordant-twin design.

  • We find an inverse association net of shared and non-shared controls.

Abstract

Higher levels of educational attainment are consistently associated with better mental health. Whether this association represents an effect of education on mental health, however, is less clear as omitted variable bias remains a pressing concern with education potentially serving as a proxy for unobserved factors including family background and genetics. To combat this threat and come closer to a causal estimate of the effect of education on depressive symptoms, this study uses data on 231 monozygotic twin pairs from The National Longitudinal Study of Adolescent to Adult Health and employs a twin-pair difference-in-difference design to account for both unobserved shared factors between twin pairs (e.g. home, school, and neighborhood environment throughout childhood) and a number of observed non-shared but theoretically relevant factors (e.g. cognitive ability, personality characteristics, adolescent health). We find an inverse association between possessing a college degree and depressive symptoms in both conventional and difference-in-difference models. Results of this study also highlight the potentially overlooked role of personality characteristics in the education and mental health literature.

Introduction

Recent estimates suggest depression constitutes a larger disease burden than any other condition in high-income countries (Mathers et al., 2008). In the United States alone, depression leads to 1445 fewer healthy years of life per 100,000 people (World Health Organization, 2009), as over 25% of U.S. adults experienced symptoms of depression in the last two weeks and nearly 30% will experience depression in their lifetime (Wittayanukorn et al., 2014, Kessler et al., 2012). This problem shows few signs of abating as levels of self-reported depressive symptomology are increasing over time (Twenge, 2015).

Higher education may be one way to protect against symptoms of depression as a wide body of research finds an inverse association between education and depressive symptoms (e.g. Lorant et al., 2003, Miech et al., 2005, Mirowsky and Ross, 2003, Quesnel-Vallée and Taylor, 2012, Ross and Van Willigen, 1997). However, despite the noted association between educational attainment and depressive symptoms, it is unclear as to whether and to what extent higher education produces better mental health.

The goal of this study is to assess the relationship between educational attainment and depressive symptoms. The occurrence of identical twins in society represents a natural occurring phenomenon that can be exploited for causal inference, particularly in addressing spuriousness due to social and genetic endowments (Kohler et al., 2011). Analyzing a sample of identical twins from The National Longitudinal Study of Adolescent to Adult Health (Add Health), a U.S. study that followed middle and high school students from 1994 to 2009, we are able to account for a wider array of potential confounding factors than any work on this topic to date.

This study is of particular importance for at least four reasons. First, prominent conceptual models (Mirowsky and Ross, 2003, Pearlin et al., 2005) assume that education has a causal influence on mental health and our study therefore provides a rigorous test of this assumption. Second, in light of the growing enthusiasm for education as a policy lever to improve population health (House et al., 2009, Montez, 2015), our study will inform discussions on whether college enrollment promotion policies are plausible mechanisms for improving population mental health.

Third, we directly account for the “big-five” personality characteristics, a neglected topic in the research on educational attainment and mental health. Personality characteristics, or dispositional patterns of thoughts, feelings, and behaviors, are strongly linked with both educational attainment (e.g., Shanahan et al., 2014) and mental health (e.g., Kendler and Myers, 2010). Given such strong associations, it is surprising that, to our knowledge, no work on education and mental health has directly accounted for personality as a potential form of spuriousness.

Finally, we extend the study of educational effects beyond high school to college. Past research has investigated the mental health benefits to education by exploiting exogenous variation in the mandatory years of education (Sironi, 2012, Crespo et al., 2014), however, as no state mandates post-secondary education, such designs are unable to assess the effect of college education. This limitation is noteworthy as 31% and 87% of young adults now obtain a bachelor's degree or at attend some college classes, respectively (Ryan and Siebens, 2012).

Section snippets

Background

A full body of research has documented that those with higher levels of education tend to have fewer depressive symptoms (e.g. Mirowsky and Ross, 2003, Miech et al., 2005, Quesnel-Vallée and Taylor, 2012). Education is believed to benefit mental health because it decreases the number and severity of stressors individuals experience and equips the possessor with the resources needed to successfully deal with stressors when they do arise. For instance, education is thought to confer economic,

Data

We use data from Add Health which used a school-based stratified cluster design to construct a nationally representative sample of adolescents in grades 7–12 in the 1994–95 school year and interviewed the adolescents and a parent in an in-home Wave I interview (N = 20,745). Original respondents have been followed over time with three follow-up in-home interviews: 1996 (Wave II), 2001–02 (Wave III), and most recently in 2008–09 (Wave IV). For more information on the study design and

Depressive symptoms

Depressive symptomology was measured by a 10 question subset of the Center for Epidemiologic Studies Depression Scale (CES-D) at Wave IV (Radloff, 1977). Consistent with past research (Bauldry, 2015), a psychometric confirmatory factor analysis of the 10 CES-D items available at Wave IV for our sample indicated that a subset of 4 items identified as suitable indicators produced a better fitting model than using all 10 indicators (results available on request). The items begin with the stem

Results

Descriptive statistics of our monozygotic twin sample stratified by discordancy on educational attainment are shown in Table 1. In the analytic sample (N = 462 individuals from 231 pairs), the average level of depressive symptoms is 2.26 (SD = 2.32). Twin pairs discordant on education had approximately the same average level of depressive symptoms as twin pairs that were not discordant on education. Twin pairs did not differ in the average level of covariates by education discordancy. In the

Conclusion

This study tested for an association of college education on depressive symptoms using a discordant twin-pair difference-in-difference approach. Consistent with previous work (e.g., Pearlin et al., 2005, Mirowsky and Ross, 2003), we found those with a college degree had lower levels of depressive symptoms than those without a college education. Having a college degree was inversely associated with depressive symptoms net of all factors shared by twin pairs, such as an early environment, as well

References (59)

  • S. Wittayanukorn et al.

    Prevalence of depressive symptoms and predictors of treatment among US adults from 2005 to 2010

    Gen. Hosp. Psychiatry

    (2014)
  • V. Amin et al.

    Causal inferences: identical twins help and clarity about necessary assumptions is critical

    Soc. Sci. Med.

    (2015)
  • S. Bauldry

    Variation in the protective effect of higher education against depression

    Soc. Ment. Health

    (2015)
  • P. Bracke et al.

    Education, mental health, and education–labor market misfit

    J. Health Soc. Behav.

    (2014)
  • A. Caspi et al.

    Personality development across the life course: the argument for change and continuity

    Psychol. Inq.

    (2001)
  • G. Conti et al.

    The education-health gradient

    Am. Econ. Rev.

    (2010)
  • M.B. Donnellan et al.

    The mini-IPIP scales: tiny-yet-effective measures of the big five factors of personality

    Psychol. Assess.

    (2006)
  • T. Fujiwara et al.

    Is education causally related to better health? A twin fixed-effect study in the USA

    Int. J. Epidemiol.

    (2009)
  • M. Gangl

    Causal inference in sociological research

    Annu. Rev. Sociol.

    (2010)
  • J. Hagan et al.

    S/he's a rebel: toward a sequential stress theory of delinquency and gendered pathways to disadvantage in emerging adulthood

    Soc. Forces

    (2003)
  • K.M. Harris et al.

    The national longitudinal study of adolescent health (Add health ) twin data

    Twin. Res. Hum. Genet.

    (2006)
  • K.M. Harris et al.

    The National Longitudinal Study of Adolescent Health: Research Design

    (2009)
  • S.A. Haas et al.

    Health and the educational attainment of adolescents: evidence from the NLSY97

    J. Health Soc. Behav.

    (2008)
  • J.S. House et al.

    The health effects of social and economic policy: the promise and challenge for research and policy

  • M. Hout

    Social and economic returns to college education in the United States

    Annu. Rev. Sociol.

    (2012)
  • M.I. Jackson

    Understanding links between adolescent health and educational attainment

    Demography

    (2009)
  • I. Kawachi et al.

    Money, schooling, and health: mechanisms and causal evidence

    Ann. Ny. Acad. Sci.

    (2010)
  • B. Kaymak

    Ability bias and the rising education premium in the United States: a cohort-based analysis

    J. Hum. Cap.

    (2009)
  • K.S. Kendler et al.

    The genetic and environmental relationship between major depression and the five-factor model of personality

    Psychol. Med.

    (2010)
  • Cited by (0)

    View full text