GMM first differences - lag instrumental variables

[cdbooth]

Hi there, I am using Eviews 9 to run a GMM calculating first differences between two time periods: 2000 and 2005. I am using a number of lag variables as instruments and I want to know how I should enter them correctly in a panel data excel file so that eviews will identify them correctly. I have attached my panel data.

As you can see the lagged variables I want to use as instruments I have left as blank columns: Yt-1, Yt-2, Yt-3, NHWt-1, NHWt-2, NHWt-3. I have two years 2000 and 2005. These lags are for the 3 years prior to 2000: 1999, 1998, 1997. So do I fill in the same values of these variables for both 2000 and 2005 rows?

My identifier (PID) is for households. So for each household, for 2000 and 2005 would I put the same values. i.e to illustrate:

PID Year Yt-1(1999) Yt-2 (1998) etc...

PID 1 2000 8 7
PID 1 2005 8 7

PID 2 2000 5 4
PID 2 2005 5 4

Here for both 2000 and 2005 I have the values of the lags which correspond to 1999, 1998... Would eviews take first differences between the value for 2005 and 2000 if I was using it as an instrument. Or would it simply take the value that I have entered for the 2000 row.

I hope you understand my issue it is very difficult to explain. I essentially want to know how to enter these lags for 1999, 1998, 1997 so I can use them as instruments in my GMM regression.

Many thanks and I hope you can get back to me as soon as possible