Ivreghdfe Instrumental Variable Regression shows error "last estimates not found"

[Sam Helen]

Dear All,

Have you come across issues when attempting to run the Ivreghdfe command in STATA? I am using the 15 version, and receive the following error message when running the following regression on a panel data set (the subscripts are only for clarification purposes):

ivreghdfe F.Y(it) ( X(it)xPost(t) = Z(i) ), absorb (i.Year) cluster(Clustervar)

Here F.Y is a forwarded dependent variable, X(it)xPost(t) is the interaction of the endogenous variable and an indicator variable for the treatment year. Z(i) is the cross-sectional instrument, and Clustervar is a cross-sectional cluster variable, with 59 observations in each year.

Error: "last estimates not found r(301)". There is no additional STATA output (The error also appears when using the contemporaneous value of Y instead of F.Y).

There are about 30,000 observations, and 6 years of data.

Any advise on how I can get this line to work would be much appreciated, since I have checked all available manuals to no avail.

Thank you,

Sam

[William Lisowski]

Welcome to Statalist, Sam.

I'm not familiar with the ivreghdfe command so I asked Stata for help ivreghdfe and this is what it reported.

help for ivreghdfe not found
------------------------------------------------------------------------------------------------------
search for ivreghdfe                                                             (manual:  [R] search)
------------------------------------------------------------------------------------------------------

Search of official help files, FAQs, Examples, SJs, and STBs

Web resources from Stata and other users

(contacting http://www.stata.com)
no matches
(click here to return to the previous screen)

(end of search)

Please review the Statalist FAQ linked to from the top of the page, as well as from the Advice on Posting link on the page you used to create your post. Note especially sections 9-12 on how to best pose your question.

Section 12.1 is particularly relevant here.

12.1 What to say about your commands and your problem

Say exactly what you typed and exactly what Stata typed (or did) in response. N.B. exactly!

If you are using community-contributed (also known as user-written) commands, explain that and say where they came from: the Stata Journal, SSC, or other archives. This helps (often crucially) in explaining your precise problem, and it alerts readers to commands that may be interesting or useful to them.

The more you help others understand your problem, the more likely others are to be able to help you solve your problem.

[Sam Helen]

Hello William,

Thank you very much for your tip.

In case this is useful, the problem was very simple: I had not installed all the necessary files from the researcher's website, which provides the IVREGHDFE package. It now works like a charm, seamlessly combining IV regressions with absorption of a large number of fixed effects.

[William Lisowski]

Hello Sam -

Thank you for closing the loop on this. That's useful information for anyone else who runs into similar issues and finds this Statalist topic by searching.

I searched for ivreghdfe and found it on github. I note that the readme file was recently updated, perhaps after your original installation, and currently has very detailed instructions for a complete installation. So to anyone who finds this topic because they have a similar problem, the instructions at

https://github.com/sergiocorreia/ivreghdfe

are a good place to start.

[Sam Helen]

Thank you for this, William. Absolutely agreed.