Why t statistics and p values are missing in the regression?

[Fred Lee]

Why t statistics and p values are missing in the regression?
If I drop all the clusters, Stata can calculate the t statistics and p values, however, should I drop all clusters? In other models, I didn't drop clusters and it can calculate, I want to keep consistent for the cluster settings.

[Carlo Lazzaro]

Fred:
1) footnotes explain part of the issue and 13 clusters are definitely not enough;
2) posting screenshots does not help (see the FAQ. Thanks);
3) I'm not clear with your regression code.

[Fred Lee]

Fred:
1) footnotes explain part of the issue and 13 clusters are definitely not enough;
2) posting screenshots does not help (see the FAQ. Thanks);
3) I'm not clear with your regression code.

Thanks Carlo, here is the code:

reghdfe dv $ctrl c.A##i.B ,vce(cluster  expID  firmID  regisAddrA ) absorb(ind  panelID regisAddrA)

[Carlo Lazzaro]

Fred:
the second footnote is pivotal in this respect.
Trying to collect so many fixed effects may be unfeasible and, even if possible, impossible to disseminate.
That said, my advice is to start it all over again with a more parsimonious model (I would start out from -xtreg,fe-) with default standard errors.

[Fred Lee]

Thanks for your advice, Carlo.

[Carlo Lazzaro]

Fred:
please find two links to the same reference about clustered standard errors:
1)

http://cameron.econ.ucdavis.edu/research/Cameron_Miller_JHR_2015_February.pdf
2) http://jhr.uwpress.org/content/50/2/317.short

[Fred Lee]

Fred:
please find two links to the same reference about clustered standard errors:
1)

http://cameron.econ.ucdavis.edu/research/Cameron_Miller_JHR_2015_February.pdf
2) http://jhr.uwpress.org/content/50/2/317.short

Thanks a ton!