I am not able to get weakivtest to work with Driscoll and Kraay (1998) standard errors. For example, if I take the following simple code from the ivreg2 help file, which shows how to compute Driscoll--Kraay standard errors:
webuse grunfeld
tsset
ivreg2 invest (mvalue=kstock), dkraay(2) small
and then try to run weakivtest, I get the following error:
weakivtest
(obs=200)
option cluster incorrectly specified
r(198);
Can anyone help to explain how this should be done? In other words, what is the proper way to compute the Montiel Olea and Pflueger (2013) robust first-stage F statistic after an IV regression, using Driscoll and Kraay (1998) standard errors?
webuse grunfeld
tsset
ivreg2 invest (mvalue=kstock), dkraay(2) small
and then try to run weakivtest, I get the following error:
weakivtest
(obs=200)
option cluster incorrectly specified
r(198);
Can anyone help to explain how this should be done? In other words, what is the proper way to compute the Montiel Olea and Pflueger (2013) robust first-stage F statistic after an IV regression, using Driscoll and Kraay (1998) standard errors?
