Converting

[drewesparon]

Does anyone know how i can convert this dataset in eviews so that i can run an SUR regression in which the system has nine equations, corresponding to 1969–84, 1985–89, 1990–94, 1995–99, and 2000–2004, 2005-2009. The dependent variables are average growth rates of per capita real GDP. the estimation should look like this

GDPPCG74 = C(9)+C(10)*LNIGDP74+C(11)*OPENNESS74 +C(12)*GKF74 +C(13)*GGFCE74 +C(14)*FRT74 +C(15)*CWC74 +C(16)*CREDITS74

GDPPCG79 = C(17)+C(18)*LNIGDP79+C(19)*OPENNESS79 +C(20)*GKF79 +C(21)*GGFCE79 +C(22)*FRT79 +C(23)*CWC79 +C(24)*CREDITS79

GDPPCG84 = C(25)+C(26)*LNIGDP84+C(27)*OPENNESS84 +C(28)*GKF84 +C(29)*GGFCE84 +C(30)*FRT84 +C(31)*CWC84 +C(32)*CREDITS84

GDPPCG89 = C(33)+C(34)*LNIGDP89+C(35)*OPENNESS89 +C(36)*GKF89 +C(37)*GGFCE89 +C(38)*FRT89 +C(39)*CWC89 +C(40)*CREDITS89

GDPPCG94 = C(41)+C(42)*LNIGDP94+C(43)*OPENNESS94 +C(44)*GKF94 +C(45)*GGFCE94 +C(46)*FRT94 +C(47)*CWC94 +C(48)*CREDITS94

GDPPCG99 = C(49)+C(50)*LNIGDP99+C(51)*OPENNESS99 +C(52)*GKF99 +C(53)*GGFCE99 +C(54)*FRT99 +C(55)*CWC99 +C(56)*CREDITS99

im using eviews version 9.5

[drewesparon]

Will someone please help me! Im really stuck on this..

[drewesparon]

this forum is useless..

[EViews Gareth]

You asked a question on a weekend...

[drewesparon]

thanks, sorry for that. I really need some help though

[startz]

Drop-and-drag the Excel file onto EViews, and walk through the dialog to tell EViews that you have a dated panel. Then setup a System and estimate by SUR. You can use the smpl command to control which observations are used for a particular estimate.

[drewesparon]

i tried that and this is what im getting.. in the attatchment

[startz]

Look at the picture you posted. It says the cross section identifier is year country. You want year to be the date identifier.

[drewesparon]

Thank you for your reply but as i mentioned i want to run an SUR regression in which the system has nine equations, corresponding to 1969–84, 1985–89, 1990–94, 1995–99, and 2000–2004, 2005-2009. The dependent variables are average growth rates of per capita real GDP. The way the data is structured in the way you mentioned I will not be able to specify the model in

GDPPCG69 = C(1)+C(2)*LNIGDP74+C(3)*OPENNESS74 +C(4)*GKF74 +C(5)*GGFCE74 +C(6)*FRT74 +C(7)*CWC74 +C(8)*CREDITS74

GDPPCG74 = C(9)+C(10)*LNIGDP74+C(11)*OPENNESS74 +C(12)*GKF74 +C(13)*GGFCE74 +C(14)*FRT74 +C(15)*CWC74 +C(16)*CREDITS74

GDPPCG79 = C(17)+C(18)*LNIGDP79+C(19)*OPENNESS79 +C(20)*GKF79 +C(21)*GGFCE79 +C(22)*FRT79 +C(23)*CWC79 +C(24)*CREDITS79

GDPPCG84 = C(25)+C(26)*LNIGDP84+C(27)*OPENNESS84 +C(28)*GKF84 +C(29)*GGFCE84 +C(30)*FRT84 +C(31)*CWC84 +C(32)*CREDITS84

GDPPCG89 = C(33)+C(34)*LNIGDP89+C(35)*OPENNESS89 +C(36)*GKF89 +C(37)*GGFCE89 +C(38)*FRT89 +C(39)*CWC89 +C(40)*CREDITS89

GDPPCG94 = C(41)+C(42)*LNIGDP94+C(43)*OPENNESS94 +C(44)*GKF94 +C(45)*GGFCE94 +C(46)*FRT94 +C(47)*CWC94 +C(48)*CREDITS94

GDPPCG99 = C(49)+C(50)*LNIGDP99+C(51)*OPENNESS99 +C(52)*GKF99 +C(53)*GGFCE99 +C(54)*FRT99 +C(55)*CWC99 +C(56)*CREDITS99

GDPPCG04 = C(57)+C(58)*LNIGDP04+C(59)*OPENNESS04 +C(60)*GKF04 +C(61)*GGFCE04 +C(62)*FRT04 +C(63)*CWC04 +C(64)*CREDITS04

GDPPCG09 = C(65)+C(66)*LNIGDP09+C(67)*OPENNESS09 +C(68)*GKF09 +C(69)*GGFCE09 +C(70)*FRT09 +C(71)*CWC09 +C(72)*CREDITS09

I need to specify the model in a way the GDPPCG is the dependent variable in each equation. I hope it is clear what im trying to ask. thanks

[startz]

It's not really clear. SUR requires that the errors in the equations are correlated between observations in the different equations. How does that work if one equation is for GDPPCG69 and another is for GDPPCG74 ?

[EViews Gareth]

There's also the issue of more observations in some equations than others?

[drewesparon]

The GDPPCG69 is the GDP per capita growth for all 44 countries in the first equation
The GDPPCG74 is the GDP per capita growth for all 44 countries in the second equation.. so on an so forth.. So each equation should have 44 observation. I thought that SUR requires different dependent variables in each equation.

[startz]

That should work. I believe what you want is a workfile with the 9 variables (plus others) and a total of 44 observations. That may take a bit of work, but I suspect what you need to do is to create the time averaged variables in the workfile you have. Then copy them to a new page which will have only 44 observations per variable.

[drewesparon]

Ok thanks but im not sure how to go about doing that :(

[drewesparon]

Is it possible to help me create this?