Literature on Stata's Extended Regression Models (ERMs): is there any?

[Bruno Hoepers]

I'm using extended probit (eprobit) and extended linear regression (eregress) in my dissertation. I'd like to reference some literature about these models.

However, the only references for ERMs I've found are Stata references.

I want to know exactly how ERMs handle endogeneity (if they do) and how they differ from IV models.

Thanks.

[William Lisowski]

Among the Stata references you found, did you review the Stata Extended Regression Models Reference Manual PDF included with your Stata installation and accessible from Stata's Help menu? A search for endogeneity in the text seems to find many hits.

[John Moran]

Although the Stata Manuals are extensive, I found the talks by David Drukker (https://www.stata.com/meeting/baltim...17_Drukker.pdf) and Chuck Huber (https://www.stata.com/meeting/oceania19/) were excellent in outlining the working of ERMs.

A standard literature search (Web of Science) did not yield results.

john moran

[Bruno Hoepers]

Thank you William and John for the suggestions.

The Stata manual on ERM is very thorough on how to execute the commands. However, it is less so concerning the theory and logic behind each model.

For instance, I don't understand the difference between a model such as xteregress and xtivreg2. Both handle endogeneity and both use instruments. But how xteregress differs from xtivreg2?

I hope I was more clear now.

Finally, I will contact David Drukker and Chuck Huber for more information on ERMs.

[William Lisowski]

I'm not surprised that Stata documentation does not comment on user-written commands like xtivreg2 (available from SSC).

If you cannot find a direct comparison of ivreg2 and xteregress, you might follow the guidance in the Intro sections (which I assume you have read) of the ERM manual to read Stata's documentation for its ivregress command to better your understanding of xteregress, and then you could look at the author's documentation of xtivreg2 and ivreg2 (on which xtivreg2 is based) to see how xtivreg2 differs from ivregress and thus from xteregress.

[sandra xu]

Hi Bruno,
Can I please ask if you found some literature on ERMs? I am using this method as well and I need literature.
Thanks

[Sebastian Kripfganz]

You can find references in the pdf documentation for eregress in the Section "Methods and formulas".