Re:Re: st: about residuals and coefficients

[Christopher Baum <kit.baum@bc.edu>]

<>
On Sep 5, 2013, at 2:33 AM, David wrote:

> The article by Filoso performs a valuable service by calling attention
> to the "Frisch-Waugh-Lovell theorem." I wish more people, especially
> authors of textbooks, were aware of that property of  regression.
> 
> I do not understand, though, why that result should be called a
> theorem or why it should be attributed to Frisch, Waugh, and Lovell.
> A 1907 paper by Yule contains a more-general result.

We use the term FWL theorem in the ivreg2 documentation and papers describing that software (Baum, Schaffer, Stillman, Stata J 2003, 2007). Greene's widely used Econometric Analysis (5th ed., the one handy at this moment) has a box on p. 27 
"Theorem 3.3 Frisch-Waugh Theorem", which undoubtedly contributes to the widespread use of this terminology. Greene's bibliography includes the Frisch and Waugh article in Econometrica vol. 1 (!!), 1933.  Hayashi's popular textbook also contains a mention of the Frishc-Waugh theorem.

Yule may have made this point first, but then Hal White did not invent Huber-White-Sandwich standard errors, nor did Lars
Hansen invent GMM. Economists are fond of naming things after the author of the article they read on the subject rather than its inventor. 

Kit

Kit Baum   |   Boston College Economics & DIW Berlin   |   http://ideas.repec.org/e/pba1.html
                             An Introduction to Stata Programming  |   http://www.stata-press.com/books/isp.html
  An Introduction to Modern Econometrics Using Stata  |   http://www.stata-press.com/books/imeus.html
                                                                                                   | http://www.crup.com.cn/Item/111779.aspx	


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