EViews.com

Hello everyone,
I am currently trying to get a VAR model to work, however when I try it I get the error message "Insufficient number of observations to estimate ## coefficients per equation in VAR." Can anyone please tell me why this is happening? The file is attached, so if anyone does have some spare time to help me I would appreciate it.

A little bit more information about what I am trying to do:
I am trying to construct a VAR model, with the change_in_GINI being the dependent variable and all the rest being independent. The lag should be 4. I am currently using EViews 8.


Thank you very much in advance!

You have 14 observations. Each variable in the VAR uses one parameter per lag. If you have four lags and four variables, for example, that means 16 parameters.

Thank you for your quick reply. What do you recommend I do? The lag should be 4. Shall I gather more data or eliminate some of the variables? One more thing, when I run the VAR (with 1 lag) I only get 14 observation, why is this? I thought since I have 14 years of data and 7 variables, I have to have 98 observations not 14.

Thank you again!

Thank you for your quick reply. What do you recommend I do? The lag should be 4. Shall I gather more data or eliminate some of the variables? One more thing, when I run the VAR (with 1 lag) I only get 14 observation, why is this? I thought since I have 14 years of data and 7 variables, I have to have 98 observations not 14.

Thank you again!

Gather more data.

A VAR is just a collection of regressions. So you have 14 observations in each regression.