How to construct an error-correction model?

[monica88]

Hello,

I have an issue with my error-correction model (I use Eviews 7) as I am new to econometrics. I want to estimate the effect of various wealth effects on consumption through OLS in levels, first differences and in error-correction form.

Here is my model in levels: C_t = β_0 + (β_1*I_t) + (β_2*N_t) + (β_3*F_t) + (β_4*H_t) + Fixed Effects + ε_t
Here is my model in first differences: ∆C_t = β_0 + (β_1*∆I_t) + (β_2*∆N_t) + (β_3*∆F_t) + (β_4*∆H_t) + Fixed Effects+ ε_t
(C = consumption, I = income, F = financial wealth, N = net financial wealth, H = housing wealth).

How do I construct an error-correction model? (how would it look like and how do I do it in Eviews?).

I hope someone can help me! This is for my dissertation and I am soon handing in my second draft.

Thank you in advance :)

[monica88]

Would it look like this?

∆C_t = a_0 + (a_1*∆I_t) + (a_2*∆N_t) + (a_3*∆F_t) + (a_4*∆H_t) - (a5*U_t-1) + Fixed Effects+ ε_t

where U_t-1 = C_t-1 - B_0 - (B_1*I_t-1) - (B_2*N_t-1) - (B_3*F_t-1) - (B_4*H_t-1)

I hope someone out there can help me :) EviewsGlenn?

[EViews Glenn]

Doing a quick google search yields (among other things) this

http://www.google.com/url?sa=t&rct=j&q= ... 2593,d.b2I

and this

http://www.youtube.com/watch?v=PWOoPxa4Rw8

Note that the fixed effects vanish from your first-differenced model.