country VS individual clustered standard errors

[sophie maene]

Dear Statlist,

I am estimating whether elderly age in a healthier way because of the LTC system they are in. I do this by grouping European countries in different LTCsystem groups. I regress health (grip strength) on factors determining health and an interaction of age with the LTC system categorical variable. I estimate the model using random effects.

However, I am stuck on whether I should use individual clustered standard errors or country clustered standard errors.

Reasoning for;
- individual clustered standard errors: allow for same individual to have correlated health shocks over time. This individual-level robust estimator is consistent in the presence of heteroskedasticity and autocorrelation (Cameron & Trivedi, 2009)
- country clustered standard errors: all the individuals in the same country may experience the same health shocks in a specific point in time
I would greatly appreciate some advice on this matter.

Thank you!

[Carlo Lazzaro]

Sophie:
welcome to this forum.
As you report, standard errors are usually clustered on -panelid-.
You do not say how many clusters at individdual level (ie, how many panels) your dataset is composed of; anyway, the number of panels should be pretty high to make the clustered standard errors work appropriately, provided that you have detected/suspect heteroskedasticity and/or autocorrelation.
Surely you have even less clusters at country level: hence clustering on this variable might hamper more than help your analysis.

[sophie maene]

Thank you very much for the help!

[Sapna Goel]

Hello everyone


I am trying to understand the effect of male-out migration on labor force participation of left behind women.

Specifically, I want to know if (say) any male member migrated in a given month, then how does it change the monthly days spent by a women in farm activities.
So my dependent variable is : number of days spent on farm in a month

And my independent variable is household level monthly migration status. It takes value 1 it any member of the household migrated in that month , 0 otherwise.


I have data with monthly frequency for 5 years. I plan to use both month level and year level fixed effects.

Now, the point is that i also want to know how the effect of migration may vary across women from different caste. Caste is a categorical variable and is time-invariant.
So I want to interact caste with household level migration status.

Now my query is : can I still use fixed effects model?

Or if I decide to use hybrid model, then we divide the total effect into (with-in) and (in-between) effect.

But caste doesn't have any with-in effect. So , then do I need to divide total effect into between and within effect?