EViews.com

I am attempting to execute first differenced equations that include polynomial lags. I do not want the first differenced equations to contain a constant, as the constant should cancel out of this type of equation. However, I got a lag specification error in my first differenced regression equation each time I attempted to run it. After some troubleshooting, I found that the error dissapeared if I included a constant in the equation.

Is there a way for me to run PDL regessions on first differenced variables without a constant?

For example, I would like to run a regression like this:
eq!X.ls(cov=hac) ps1 pdl(d(ipr),!OL,!D) pdl(d(un),!OL,!D) pdl(d(unionprq),!OL,!D) pdl(d(unipr),!OL,!D) pdl(d(unionipr),!OL,!D) pdl(d(tsaspl),!TSAL,!D) pdl(d(tfp4),!TFPL,!D)

Thanks for your help

I think you've found a bug. It doesn't look to have anything to do with the data being first differenced. It seems the bug happens any time pdl is the only right hand side variable.

Really odd, given that pdl has been in the code for probably 30 years.