Page 1 of 1
Posted: Sat Jun 01, 2019 9:26 am
attached you will find my comments.
Posted: Sat Jun 01, 2019 4:11 pm
Please below my comments. Thank you.
Examining the Impact of Early Childbearing Labor Force Participation, Earnings and Education in Brazil
Vaqar Ahmed, Ronelle Burger, Mitzie Conchada, Ana L Kassouf, Agnes Zabsonré,
- Is it possible to show/report statistics on teenage women´s labor force participation in the country ?
Pathways of impact mechanisms :
- The paper says : « if a teenage mother spends a large number of hours taking care of her baby or spends time working into the labor market, she will not have time to study and will probably drop out of school. Less human capital formation will result in less labor opportunities, low level jobs and lower wage rates. »
So the impact may go these two way ? :
1 : teenage mother/early pregnancy ----more hours spent for taking care of baby ---less hours for studying ----drop out of school (less education)
2 : teenage mother/early pregnancy ----more hours spent for working in labor market ---less hours for studying ----drop out of school (less education)
Which channel fit the best with your results ?
- What is the sample (size) ?
Please see the definition of the treatment variable : childbearing = 0 for what sample ?
- « Table 1 has the description of the variables used in the estimations as well as the mean and standard-deviation for the sample of women who had baby before the age of 20 (childbearing = 1) or after 19 years old (childbearing = 0) ».
- Figure 4 : Percentage of women from 20 to 49 years old working when had a baby before the age of 20 (1) or after 19 years old (0).
- Some data have been collected at the time of the survey or at the time these women had baby ?
- The paper says that : « …., there is a discussion in the literature about the difficulties in isolating the true effect of teenage childbearing, due to the fact that most teenage mothers are born in deprived families and are not well educated. Low socioeconomic conditions faced by young mothers confound the pure effect of a baby born [Hoffman (2011)] ».
I think that socioeconomic conditions at the time the women had baby should be controlled for in the regression instead at the time of survey ?
- How do the paper calculate the education variable ? When women get pregnant/when dropped out school or at the time of survey?
- I think the outcome and treatment variables are not in the same temporal reference. Between the two time periods many things could happen ?
- How do the paper include health status in the model (see table 1 : « Levels of health status from very good (1) to very bad (5) ».
I think categories should be detailed)
- Figure 6. One could say also that : those the first occurence of menstration is earlier are more likely to be mature and more likely to enter labor market anyhow
- For Figure 8, the papers says : There are women who had their first occurrence of menstruation at 8 years old, going up to 23 years old. The largest percentage occurred at 12 and 13 years old, but at 11, 14 and 15 there is also a large number of women.
Is it possible to restrict the sample to those women to see whether the results change ?
- The paper says : « more independent women with higher economic autonomy and that are more career oriented will probably postpone pregnancy or even choose to not have children and at the same time will choose to participate more actively in the labor market ».
What is the correlation between the two equations of biprobit. Is it negative ? The coefficient is not reported.
- The paper estimate the Effect of childbearing on women´s years of education.
Then education is endogenous in the wage/work equation ?
- The IV diagnostic tests are not reported in the regressions tables.