Table 2.

OLS regression of outcomes on gender.

Specification models	Dependent variables
	Lost job permanently	Lost job temporarily	Log of expected income reduction	Increased weekly expenses	Increased savings
Model 1	0.010*	−0.005	0.339***	−0.055*	0.065**
	(0.006)	(0.011)	(0.129)	(0.030)	(0.027)
Model 2	0.012**	0.002	0.461***	−0.063**	0.062**
	(0.006)	(0.011)	(0.128)	(0.030)	(0.027)
Model 3	0.012**	0.003	0.453***	−0.066**	0.060**
	(0.006)	(0.011)	(0.129)	(0.030)	(0.027)
Model 4	0.013**	0.007	0.448***	−0.065**	0.053*
	(0.006)	(0.011)	(0.129)	(0.030)	(0.027)

Note: This table reports the estimated coefficient on the female variable for different specification models that differ in the number of control variables. Model 1 only controls for country dummy variables. Model 2 adds to Model 1 demographic characteristics (age group, living alone, and urban areas). Model 3 adds to Model 2 income quintile of respondents. Model 4 adds to Model 3 the geographic region fixed-effects. The full regression results of Model 4 are reported in Table A.4 in Appendix A.

Robust standard errors in parentheses

*** p < 0.01, ** p < 0.05, * p < 0.1