Versión de HTML Básico
Andrea Molina Vera
Analítika, Revista de análisis estadístico, (2015), Vol. 9
16
Table 5:
Random assignment of the same sex
By Same Sex
Variable
Controls
sd Treated Sd p-valu
Age
29.482 (3.785) 29.471 (3.788) [0.320]
Age at first birth 19.321 (3.079) 19.321 (3.081) [0.927]
Indigenous
0.081 (0.274) 0.08 (0.272) [0.191]
Urban area
0.608 (0.488) 0.609 (0.488) [0.687]
Years of education 8.853 (4.365) 8.857 (4.358) [0.832]
N
225180
229945
Note: Data source: 2010 Census. The sample is all women. Standard errors are reported in parentheses.
It is important to mention that work measurement are relevant in the sample of all
women since there is an impact only when the definition of work involves characteristics
incompatible with child rearing. On the other hand, for married women there is an impact
using all measurement of work. Perhaps this happens since married women are supported
with their spouses’ revenues.
Table 6:
OLS and 2SLS. Estimates of Labor-Supply Models Using 2010 Census Data
All Women
All Women
OLS
2SLS
2SLS
OLS
2SLS
2SLS
(1)
(2)
(1)
(2)
Instrument for
Same sex Twoboys,
Same sex Twoboys,
More than 2 children
Twogirls
Twogirls
Dependent Variable
Work
-0.1120
-0.0569 -0.0527 -0.1078
-0.0800
-0.0788
[0.0015]*** [0.0393]
[0.0392]
[0.0016]*** [0.0391]** [0.0391]**
Work for pay
-0.1118
-0.0599 -0.0549 -0.1076
-0.0834
-0.0809
[0.0015]*** [
[0.0392]
[0.0391]
[0.0016]*** [0.0390]** [0.0389]**
Work outside home
-0.1084
-0.0847 -0.0815 -0.1036
-0.0934
-0.0921
[0.0015]*** [0.0378]** [0.0377]** [0.0015]*** [0.0372]** [0.0371]**
Paid Work outside the home -0.1080
-0.0893 -0.0860 -0.1031
-0.0988
-0.0974
[0.0015]*** [0.0376]** [0.0375]** [0.0015]*** [0.0370]*** [0.0369]***
Note: * p
<
0.1; ** p
<
0.05; *** p
<
0.01. Robust standard errors in brackets. Covariates: age of women, age
at first birth, indicators for Boy 1st, Boy 2nd and dummies by indigenous and urban status. In (2) the
variable Boy 2nd is excluded.
Also, the OLS estimates and 2SLS are similar for married women but for all women
depend on the definition of work. The results are confirmed by the two boys and two girls’
instrument.
These results are local for the compliers, that is, for women whose fertility decision is
changed for the instrument, which is used for identifying the impact of labor supply when
children move from 2 to 3, but don’t identify other increases in fertility as 0 to 1 child
5
.
5
With the 2001 Census, the first stage is similar to 2010 for both samples. The 2SLS estimates present
10