Table 4.

Estimated relationship between earnings or education and personality with and without controlling for childhood health and maltreatment indicators (Reported: marginal effects)

	Dependent Variable
	Log Earnings (1a)	Log Earnings (1b)	% Diff	College degree (2a)	College degree (2b)	% Diff
Extraversion	0.050*** (0.014)	0.048*** (0.014)	−4	−0.018*** (0.004)	−0.020*** (0.004)	8.7
Neuroticism	−0.086*** (0.017)	−0.074*** (0.016)	−14	−0.042*** (0.004)	−0.031*** (0.004)	26**
Agreeableness	0.014 (0.019)	0.009 (0.018)	−35.7	0.042*** (0.005)	0.042*** (0.005)	1.7
Conscientious.	0.042*** (0.015)	0.023 (0.015)	−45.2**	0.022*** (0.004)	0.013*** (0.004)	38.4**
Openness	−0.026 (0.017)	−0.029* (0.016)	11.5	0.030*** (0.004)	0.031*** (0.004)	1.4
Family Background and Cognitive Ability	Yes	Yes		Yes	Yes
Childhood health and maltreatment	No	Yes		No	Yes
Observations	8,195	8,195		10,693	10,693
R-squared	0.106	0.134		0.2174	0.2404

Note: All models control for the full set of family background variables and cognitive ability as in Fletcher (2013) for earnings and in Lundberg (2013) for the probability to obtain a college degree. Outcomes and personality are measured in Wave 4. Models (1a) and (1b) are estimated with ordinary least squares. Models (2a) and (b) are estimated with a linear probability model.

^***p<0.01,

^**p<0.05,

^*p<0.1.