Table 5

Effects of Own-Race Bias for Hiring Managers vs. New Managers

	Quits		Dismissals		Promotions
	Hazard Ratio	χ² (Pr > χ²)	Hazard Ratio	_χ2 (Pr > χ²)	Hazard Ratio	_χ2 (Pr > χ²)
White employee ^white manager^hiring manager	0.957 (0.030)	3.42^{^‡}	0.997 (0.073)	4.15^*	0.803 (0.158)	0.00
White employee^white manager^new manager	0.858^** (0.049)	(0.064)	1.319^* (0.178)	(0.042)	0.801 (0.168)	(0.986)
Black employee^black manager^hiring manager	1.013 (0.048)	0.20	0.784^* (0.075)	2.16	1.734^{^‡} (0.563)	0.04
Black employee^black manager^new manager	1.057 (0.096)	(0.653)	0.988 (0.152)	(0.141)	1.892 (0.736)	(0.840)
Hispanic employee^Hispanic manager^hiring manager	0.935 (0.044)	0.42	0.857 (0.095)	0.14	1.222 (0.433)	0.61
Hispanic employee^Hispanic manager^new manager	0.955 (0.051)	(0.517)	0.791 (0.155)	(0.706)	1.644 (0.552)	(0.436)
Asian employee^Asian manager^hiring manager	0.946 (0.076)	2.45	0.817 (0.147)	0.00	0.756 (0.805)	0.10
Asian employee^Asian manager^new manager	0.890 (0.076)	(0.118)	0.836 (0.337)	(0.956)	1.116 (0.786)	(0.748)

Note: Hazard ratios from Cox proportional hazard models with control variables as in Table 3, plus all interactions of manager race and employee race indicators with an indicator that the manager is new. Robust standard errors in parentheses, adjusted for clustering on employee.
↵‡ Significant at 10 percent;
↵* significant at 5 percent;
↵** significant at 1 percent (based on test that the hazard ratio is different from one). Second column for each set of results reports Wald test of equality for each pair of hazard ratios.