PSM, exclusion restrictions and bounding treatment effect

Luca Tiberti
Posts: 33
Joined: Thu Jun 07, 2018 6:31 am

PSM, exclusion restrictions and bounding treatment effect

Post by Luca Tiberti » Thu Sep 05, 2019 4:49 pm

Dear team,
Many thanks for your presentation. Here below a few comments which may hopefully help to further improve the study:
- Would it be interesting to see the capacity of diversifying by different constraints (e.g., by type of land tenure)
- The third methodological approach can be interesting but it can only capture some correlations and not causal effects. In other terms, it would be a rather descriptive econometric approach
- The PSM approach relies on the (strict) condition of “independence assumption” (CIA). To my understanding, your data would not allow obeying this assumption. If so, I suggest to drop the PSM analysis, and focus just on the ESR model
- I do not see the exclusion restriction variables in the selection model of the ESR. In such a case, the identification relies on the non-linearity only, which might not be enough. I would strongly suggest you to identify one restriction variable.
- Also, to validate the results (and the validity of your approach), I suggest to try to estimate set of bounds of your key treatment coefficient. A good reference is the test developed by Oster (2019) – see the full reference below
- Outcome: do you observe total individual monthly income? If so, I suggest to distinguish between incomes earned by women and men. Alternative related outcomes are the hours worked (in farm and non-farm activities).

Oster, E. (2019). Unobservable selection and coefficient stability: Theory and evidence. Journal of Business & Economic Statistics, 37(2), 187-204.


Return to “PMMA-20418 - Women's Employment Issue in Rural Senegal: What Can We Learn About the Shift From Farm Activities to Diversification Strategies?”

Who is online

Users browsing this forum: No registered users and 1 guest