turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

Find a Community

- Home
- /
- Analytics
- /
- Forecasting
- /
- Fixed Effects Panel Data Regression

Topic Options

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

10-01-2014 11:55 AM

Hello;

I have a panel data set, some of the variables change over time, while some others don't. For example,

dependent variable is the number of publications by professors in each year which changes over time.

independent variable is the amount of research grants that a professor has per year, which also changes over time.

the other independent variable is the type university with which the professor is affiliated with (teaching vs research school), which does not change in my data set.

When I run a panel regression with fixed effect, I do not get any estimates for the university type, the university type effects are only only estimated in the random effects model. However, I know that in addition to the university type, there are some characteristics that pertain to professors and do not change over time which I have not been able to measure and thus theoretically I need to run a fixed effects model.

Is there any way to keep the university type as a variable which does not change over time and still run a fixed effects model?

Thanks

Accepted Solutions

Solution

10-03-2014
11:05 AM

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

Posted in reply to niam

10-03-2014 11:05 AM

Hi Niam,

Thanks for your question. You seem to have two distinct wants for your model. The first, and your primary research question seems to be whether funding influences publications. If that is the case, then your FE estimate with the university type dropped is effectively controlling for the university type and all other idiosyncratic confounding effects specific to the professor. So your estimates, assuming the modeling specified correctly, should be consistent.

If you are interested in the marginal effect of type and you want to still control for some idiosyncratic professor effect, you will need to relax some of the assumptions on the way that university type interacts with professor specific effects. If willing to do this, you can employ another estimator known as a Hausman-Taylor estimator. Currently we do not have super convenient access to this estimator in ETS (but look for it soon ). In the meantime, if you look at this document, page 210 http://eco.cueb.edu.cn/upload/2012/9/616432590.pdf

then you will see how to estimate this in SAS. It is a multistep process. You can also find IML code here.

Copyright C 2012 by R. Carter Hill and Randall C. Campbell

"Using SAS for Econometrics"

by R. Carter Hill and Randall C. Campbell (2012)

John Wiley and Sons, Inc

Chapter 15

Again, the key assumption for this model to make sense is that there is no correlation between the individual professor characteristics and the university type (which seems a little implausible to me though this is your call)

Best of luck-Ken

All Replies

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

Posted in reply to niam

10-02-2014 06:19 PM

Solution

10-03-2014
11:05 AM

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

Posted in reply to niam

10-03-2014 11:05 AM

Hi Niam,

Thanks for your question. You seem to have two distinct wants for your model. The first, and your primary research question seems to be whether funding influences publications. If that is the case, then your FE estimate with the university type dropped is effectively controlling for the university type and all other idiosyncratic confounding effects specific to the professor. So your estimates, assuming the modeling specified correctly, should be consistent.

If you are interested in the marginal effect of type and you want to still control for some idiosyncratic professor effect, you will need to relax some of the assumptions on the way that university type interacts with professor specific effects. If willing to do this, you can employ another estimator known as a Hausman-Taylor estimator. Currently we do not have super convenient access to this estimator in ETS (but look for it soon ). In the meantime, if you look at this document, page 210 http://eco.cueb.edu.cn/upload/2012/9/616432590.pdf

then you will see how to estimate this in SAS. It is a multistep process. You can also find IML code here.

Copyright C 2012 by R. Carter Hill and Randall C. Campbell

"Using SAS for Econometrics"

by R. Carter Hill and Randall C. Campbell (2012)

John Wiley and Sons, Inc

Chapter 15

Again, the key assumption for this model to make sense is that there is no correlation between the individual professor characteristics and the university type (which seems a little implausible to me though this is your call)

Best of luck-Ken

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

Posted in reply to ets_kps

10-05-2014 10:59 AM

Thank you very much. Your references address my question.

Best