How to apply the restriction c(2) = c (3)?

[Samster]

Greetings. The Wald test results accept the null hypothesis of c(2) = c(3). How do I run a regression imposing said restrictions on the coefficients ?

The following options work in that the coefficients (in the regressions with the restrictions) are indeed equal to each other. However, there are differences in the coefficients depending on which of the following is used.

(c(2) > c(3))*X1

or (c(2) >= c(3))*X1

or (c(2) > c(3))*X1 + (c(3) > c(2))*X2

Thanks for your time.

[EViews Gareth]

What on earth are you trying to do?

If you want to impose the restriction that c(2)=c(3), then just use c(2) for both regressors.

Code: Select all

y=c(1)+c(2)*x+c(2)*z


[Samster]

Thank you very much man.

However, surely your response could have been more professional. "The what on earth" this non-expert Eviews user, non-expert Econometrician, was doing actually yielded coefficients that were identical up to 4 decimal places (see attachment). However, the approach seemed "unstable" so I sought clarity. Indeed, I could not use the equal sign so I tried something else. Next time I will simply bother you, instead of trying to think outside the box.

Do have a pleasant day/evening.

eviews.pdf

(106.4 KiB) Downloaded 257 times

[EViews Gareth]

Criticism accepted

[EViews Glenn]

An even easier one is to modify the regressor to enforce the restriction

Code: Select all

y c z (x1+x2)