[Joao Santos Silva]

Dear Kyriaki Tselika,

The command has an option to save the fitted values; if you subtract those from the dependent variable you get the residuals.

Best wishes,

Joao

[Kyriaki Tselika]

Dear Joao Santos Silva ,

Thank you so much for the answer.

Best regards,
Kyriaki

[OWUSU ANSAH]

Hello, I am using xtqreg, for growth studies. Now my director of studies want me to show the list of countries which falls under each quantile of GDP. I have found that FDI only have positive impact on countries in the higher quantiles of gdp, so now how do i show the list of these countries. Please a command to do this will be highly appreciated.
Currently I have divided the gdp variable into quantiles and have applied bysort gdppc_quan: tab country
but these countries are appearing in more than one quantile.

[Joao Santos Silva]

Dear OWUSU ANSAH

You can get the predicted values and see which countries have a dependent variable below the estimated quantile. Keep in mind that with panel a country can be below the quantile in some years but not in others.

Best wishes,

Joao

[OWUSU ANSAH]

Dear OWUSU ANSAH

You can get the predicted values and see which countries have a dependent variable below the estimated quantile. Keep in mind that with panel a country can be below the quantile in some years but not in others.

Best wishes,

Joao

Thankyou Joao for the reply. Please can you suggest a command to do this please? thankyou

[Joao Santos Silva]

Please open the help file for the command and check the available options.

Best wishes,

Joao

[Jeffery Xu]

Dear Joao Santos Silva,

I have two questions for using xtqreg with JK correction and clustered standard errors.

1. I have a panel for 18 years (=9 waves: data collected every two years) but many individuals (2/3 of the sample) are observed for only three or less than three waves (because losses to follow up). In this case, is it appropriate to use xtqreg here?
2. For the stata code posted here: https://jmcss.som.surrey.ac.uk/MM-QR-JK.do. Can I use this code to obtain the clustered standard errors if my panels are not nested within clusters? That is, I want to cluster the standard errors at the county level, but many Individuals will move in/out of the county and therefore they are not nested within a particular cluster. When I try to apply the aforementioned code in my analysis, I obtain the error message "repeated time values within panel, the most likely cause for this error is misspecifying the cluster(), idcluster(), or group option. Do you have any comments or suggestions on how to deal with this problem?

Many thanks!

[Joao Santos Silva]

Dear Jeffery Xu,

I am afraid that with such short panel your only option if to use the so called correlated random effects approach as it

Jason Abrevaya & Christian M Dahl (2008). "The Effects of Birth Inputs on Birthweight," Journal of Business & Economic Statistics, 26:4, 379-397, DOI: 10.1198/073500107000000269

As for the error, assuming that you wrote the command correctly, I suggest that you do not use the time variable in xtset.

Best wishes,

Joao

[OWUSU ANSAH]

Hello Guys,
Can anybody help me the code to run the Galvao dynamic quantile regression for fixed effects? I know this approached included the lag of the dependent variable. I have so amny sources but can't find anythiing useful. I understand there is no code for it but I can see Joao Santos Silva used this approached as a robust check in his paper. Please if anybody can help me i will appreciate it alot. Thankyou for your help.

[Raymond Zhang]

@Joao Santos Silva Dear Pro Joao,I have sent you an email.Can you send me the codes of -xtqreg- in your paper?

best
Raymond

[FernandoRios]

Hi Raymond,
you can get the codes for -xtqreg- by installing it directly from ssc. Just type -ssc install xtqreg-.
Best

[sedki zn]

Dear Mohamed Mahjoub Elheddad,

There are 2 commands related to that paper: xtqreg allows for fixed effects and therefore deals with endogeneity only to the extent that the inclusion of the fixed effects can solve the endogeneity; ivqreg2 allows the estimation using instruments, but does not allow for fixed effects.

Best wishes,

Joao

Dear Joao Santos Silva
here you mention that ivqreg2 does not allow for fixed effect and in general QR does not allow for random effect. so when we can use this ivqreg2 ?

best regards

[Joao Santos Silva]

Dear sedki zn,

You can use ivqreg2 if you have endogenous variables (and instruments), but no fixed effects.

Best wishes,

Joao

[sedki zn]

Dear Joao Santos Silva

thank you for you quick response as always. ))

yes i got this.
but for a panel data we need to run our model with FE or RE based on hausman test ..
so if i got you correctly, if we find FE is appropriated we can not use ivqreg2?? ( if RE is appropriated, we can not even apply a QR )

so here in both case we can not use the ivqreg2 , only if we are not talking about the same meaning of "fixed effect'

i hope you get what i'm trying to say.

Best regards
sedki

[Joao Santos Silva]

Dear sedki zn,

The command ivqreg2 is not designed to be used with panel data; we can use it with panels if we do not include the fixed effects.

Best wishes,

Joao