 Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.

st: RE: Re: IVREG2 and Multi-way Clustering

 From "Schaffer, Mark E" To Subject st: RE: Re: IVREG2 and Multi-way Clustering Date Sun, 29 Jan 2012 22:21:47 -0000

Jessie,

> -----Original Message-----
> From: owner-statalist@hsphsun2.harvard.edu
> [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Jessie C
> Sent: 29 January 2012 21:42
> To: statalist@hsphsun2.harvard.edu
> Subject: st: Re: IVREG2 and Multi-way Clustering
>
> Mark,
>
> Thank you very much for your response and for taking the time
> to help out.  I greatly appreciate it.
>
> 1. Your example makes sense of the 2 X 2.  z_1 has 2
> clusters.  z_2 has 2 clusters.  The union of z_1 and z_2 is 4
> clusters but also 4 observations.
>
> I don't quite understand then the 2-way clustering formula in
> terms of 1-way clusters.
>
> Cameron, Gelbach, and Miller say:
>
> 1. OLS regression of y on X with variance matrix estimate
> computed using clustering on g in the set of {1, 2, ...G}; 2.
> OLS regression of y on X with variance matrix estimate
> computed using clustering on h  in the set of {1, 2, ...H};
> 3. OLS regression of y on X with variance matrix estimate
> computed using clustering on (g, h)  in the set of {(1, 1),
> ..., (G, H)};
>
> Given these three components, V[beta] is computed as the sum
> of the ...first and second components, minus the third component.
>
> I thought that would correspond with:
> i. reg y x, cluster(g)
> ii. reg y x, cluster(h)
> iii. reg y x, cluster(i) where egen i = group(g h) and the
> standard error is se(i) + se(ii) - se(iii) or
> sqrt(se(i)^2 + se(ii)^2 - se(iii)^2)
>
> I tried an example in Stata and it worked out using the sqrt formula.
> Not sure if that's just a coincidence.

A coincidence, I think.  To do it properly in the K-regressor case you
need to work with the meat of the cluster-robust sandwich formulation.
In the slides you point to below, see slides 10 and 22.  The meat of the
sandwich in the 2-way cluster case is B_hat on slide 22.

> 2. If I'm understanding correctly, the correction to ivreg2
> is an adjustment based on finite samples?

The finite-sample adjustment we use in -ivreg2- is one of those
recommended by Cameron-Gelbach-Miller; see their paper.

> I am wondering whether you have any response to using the
> t-statistic of G-L as suggested in:
> http://www.stata.com/meeting/mexico11/materials/cameron.pdf
> Section 4, slide 17

No, sorry.

--Mark

>
> Thank you so much for everything!!
>
> On Sun, Jan 29, 2012 at 11:54 AM, Jessie C
> <jessiecoh@gmail.com> wrote:
> > I have 2 questions about two-way clustering using ivreg2.
> >
> > 1. What would be the 1-way cluster equivalent of the
> following 2-way clustering?
> >
> > ivreg2 y x, cluster(z_1, z_2)
> >
> > I thought it would be
> > a. ivreg2 y x, cluster(z_1)
> > b. ivreg2 y x, cluster(z_2)
> > c. ivreg2 y x, cluster(z) where egen z = group(z_1 z_2) and the
> > two-way standard error on x would be the standard error in a + the
> > standard error in b - the standard error in c or would it be
> > sqrt(se(a)^2 + se(b)^2 - se(c)^2)
> >
> > Note. That is not the full model, but I thought would be
> illustrative
> > to get at the main issue.
> >
>
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/
>

--
Heriot-Watt University is a Scottish charity
registered under charity number SC000278.

Heriot-Watt University is the Sunday Times
Scottish University of the Year 2011-2012

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/