minor comments

Post Reply
jdavalos
Posts: 14
Joined: Sun Jun 10, 2018 8:35 am
Contact:

minor comments

Post by jdavalos » Tue Jun 12, 2018 5:39 am

+ perform some robustness checks by using one instrument at a time. Do not forget to cluster your standard errors at the poverty rate regional level of aggregation.

+ Acknowledge the potential caveats with the poverty rate. Poor regions are likely to attract lower ability households, which in turn would imply a relationship with the probability of being mismatched. This is an argument of the previous comment.

+ You may want to refer to your skills mismatch indicators as subjective skills mismatch ones through the paper. Remember that this is a perception and not a real objective classification. Referees will point this.

jdavalos
Posts: 14
Joined: Sun Jun 10, 2018 8:35 am
Contact:

Re: minor comments

Post by jdavalos » Tue Jun 12, 2018 5:43 am

jdavalos wrote:
Tue Jun 12, 2018 5:39 am
+ perform some robustness checks by using one instrument at a time. Do not forget to cluster your standard errors at the poverty rate regional level of aggregation.

+ Acknowledge the potential caveats with the poverty rate. Poor regions are likely to attract lower ability households, which in turn would imply a relationship with the probability of being mismatched. This is an argument of the previous comment.

+ You may want to refer to your skills mismatch indicators as subjective skills mismatch ones through the paper. Remember that this is a perception and not a real objective classification. Referees will point this.

+ The literature distinguishes between educational mismatch and skills mismatch. The latter concept is a broader human capital related-one, while the former refers the level of education only. Look carefully to the survey question and the literature and explain how should one understand your subjective indicator.

Bilal Nabeel Falah
Posts: 3
Joined: Tue Jun 05, 2018 8:39 am
Contact:

Re: minor comments

Post by Bilal Nabeel Falah » Tue Jun 12, 2018 5:48 am

Research Motivation:

The researchers auto explain what makes Zambia an interesting case to study? Is there any thing special about internship or maybe share of engaging undergrads?

Methodology:

The authors intend to use ordered probit to address the effect of pre-graduation work on educational mismatch.

The dependent variable takes 0 if the outcome is undereducated youth, 1 for matched, and 2 over-educated youth. Are these outcomes typically ordered? Is over-educated outcome better than matched?. Why not use multinomial logit model!

The authors rightly acknowledge the concerns of self-selection into pre-graduation work and sample selection bias.
To address the self-selection concern, the authors proposed to use an instrumental variable approach. The instrument is specified as
1) poverty status of the neighborhood, proxied by the enumeration area, as an instrument for
pre-graduation work. Does the instrument satisfy the exclusion condition? Economic conditions of the neighborhood my determine whether an individual will get a job that fits his education level.

2) is the number of children that a youth? What is the identification assumption?

However, the authors specified the pre-graduation work in model (2) as participation
in mandatory internships (𝑁𝑇). SO, if it is mandatory, then a potential way is to understand the process through which students select universities / fields with such programs. Is it based on test score? Or distance to university? This may help come up with an appropriate instrument that meet the orthogonality condition or even use different impact evaluation tool.

Post Reply

Return to “Occupational preferences and labour market skills mismatch among youth in Zambia – PMMA 20280”

Who is online

Users browsing this forum: No registered users and 1 guest