[Jake Ed]

I was able to fix the issue. The correct command is:

bootstrap,cluster(newid) idcluster(Names1) reps(10):xtqreg mentions max careerage xcount mentionslag_1 i.year, i(Names1) quantile(.05)

[Jake Ed]

Many thanks! Just figured it out. I have another question about using year and group dummies.


I understand that the following command allows me include year dummies and group (Names1) fixed effects.

bootstrap,cluster(newid) idcluster(Names1) reps(10):xtqreg y max careerage xcount ylag1 i.year, i(Names1) quantile(.05)

However, if I include both year and group dummies, then the output only estimates the year dummies. The output for the group dummies says "omitted".

bootstrap,cluster(newid) idcluster(Names1) reps(10):xtqreg y max careerage xcount ylag1 i.year, i.Names1, quantile(.05)

So it seems, I cannot include both year and group dummies?

Thanks, Jake

[Joao Santos Silva]

Dear Jake Ed,

The command automatically included the group dummies; that is way they are omitted if tou include them a second time.

Best wishes,

Joao

[Aleksandra Djordjevic]

Dear Joao,

Thank you very much for your xtqreg command. I am currently doing my phd thesis and I have two questions about this command.

1. My data has N=132 and T= 11. I want to confirm, is this T "large" enough?
2. Is there possibility to include exporter_time and importer_time fe (I am running gravity equation)? Is the command suitable for that?

Thank you very much for your help in advance!

Best wishes,

Aleksandra

[Joao Santos Silva]

Dear Aleksandra Djordjevic,

T=11 is really short it you may still get reasonable results. Adding additional fixed effects would be asking to much.

Best wishes,

Joao

[sedki zn]

Dear Joao Santos Silva

i have a data with 781 observations ( T=11) and as i see that is a short period to use the xtqreg command. actually i'm working on te relationship between firm performance and risk taking. ( fixed effect). can i use the command ivqreg2 ??

best regards

[Joao Santos Silva]

Dear sedki zn,

I am afraid that ivqreg2 does not really work for panel data. Your options are either xtqreg (as I said above, T=11 is really short but you may still get sensible results) or to use correlated fixed effects (based on qreg2).

Best wishes,

Joao

[sedki zn]

Dear Joao Santos Silva

thank you for your simple and soft explantation.
just in case of the presence of endogeneity issue, how can i deal with it after using the qreg2 command ?

Kind regards
SEDKI

[Joao Santos Silva]

Dear sedki zn,

I am not aware of a method that allows you do deal with FE and endogeneity in QR.

Best wishes,

Joao

[Joao Santos Silva]

Dear All,

With the usual thanks to Kit Baum, an updated version of xtqreg is now available in SSC. This version allows the user to save the estimated fixed effects and predicted values.

Please do let me know if you find any problems with the new version.

Best wishes,

Joao

[Santosh Dash]

Dear Joao Santos Silva Can we run quantile regression with missing observations in some independent variables?

[Joao Santos Silva]

Dear Santosh Dash,

xtqreg works like any other Stata command: if you have missing values of some variables, those observations will be dropped from the estimation sample.

Best wishes,

Joao

[Santosh Dash]

Thank you for clarifying.

[Mohamed Mahjoub Elheddad]

Dear Mohamed Mahjoub Elheddad,

I am not sure if I understand your question, but you can download the code from SCC.

Best wishes,

Joao

Thank you very much Joao Santos Silva. The code is really helpful.

[Ainhoa Oses]

Dear all,

I would like to estimate a quantile multi-level model with fixed effects, and I do need to include probability weights as I'm using survey data. Given that -xtqreg- does not support weights, would anyone know an alternative to this that does support the inclusion of pweights, please?

Many thanks,

Ainhoa