EViews.com

Hi gurus,

Is there any option or plan to adopt the use of super-addfactors in the sense like this:
- I want to impose a 1% increase in GDP by add factoring
- GDP is sum of C, I, G, NX, so the imposing 1% increase in GDP means imposing 1% increase in all of its components.

I think it's a tricky one if the model has layers of identities (i.e. C is another identity), however an option like that would be nice.

Thanks,

T

I'm afraid there's no inbuilt support for an operation like that.

I do see what you're getting at though - you would like the add factor adjustments to end up in equations that contain errors rather than on identities (which don't).

There's quite a few details that would have to be sorted out though. For example - what happens if the identity is non-linear (in which case increasing each input proportionally would not increase the output by the same proportion)? Your question about layers of identities is also a good one.

For any particular identity it shouldn't be too hard to do this yourself 'by hand'.

1) exclude C, I G, NX
2) bump up each of their values by 1%
3) solve the model
4) reinclude C, I, G and NX one by one generating add factors as you go.

But for the moment you'd have to do this on an ad hoc basis for the particular identities/variables.