The Origins of Happiness

Notes to Pages 158–165

are found to be strongly predictive of children’s educational outcomes
that include completion of high school and attainment of a college
degree (Kim and Sherraden [2011]). Including family assets into the
estimation of children’s educational outcomes also reduces the size of
the income effect and, in some cases, even renders it statistically insig-
nificantly different from zero.

12. Acemoglu and Pischke (2001).


13. Yeung, Linver, and Brooks- Gunn (2002).


14. Waldfogel, Han, and Brooks- Gunn (2002).


15. See Online Annex 10 on financial problems.

Chapter 11. Working Parents

1. This whole chapter is about work outside the home— referred
   to for convenience as “work.”


2. Questionnaires at months 21, 33, 47, 61, 73, 97, 110, 122, and


3. The surveys at these ages ask whether the mother is currently
   working or not. Unfortunately, for those who are not working we do
   not know whether they are out of the labor force or are unemployed.
   This means the analysis must follow the somewhat anachronistic logic
   of working mothers and unemployed fathers.


4. It doesn’t. For explanation of decomposition, see Online An-
   nex 3b.


5. McMunn et al. (2010). Mental health is measured by total SDQ.
   The statement in the article relates to two- parent families.


6. Powdthavee and Vernoit (2013).


7. For a brief summary, see Layard and Dunn (2009), 20– 21. In the
   UK see Sammons et al. (2014). In the United States see also Cooksey,
   Joshi, and Verropoulou (2009) and Berger, Hill, and Waldfogel (2005).


8. This is also true at 11.


9. In future work we shall look at the effects when we do not con-
   trol for income.


10. Using the BCS, see P. Gregg et al. (2005), and using EPE, Sylva
    et al. (2004), and Sammons et al. (2014). By contrast for intellectual
    development up to age 5, Ermisch and Francesconi (2013), using BHPS
    linked data on parents and children, found a negative and statistically
    significant maternal employment effect on intellectual problems up
    to age five. One possible explanation is that they were able to take into
    account both the unobserved heterogeneity that is common across sib-
    lings and the endogeneity of mother’s choice to return to employment.