Page 1 of 1

A change in endogenous variables in general equilibrium

Posted: Thu Dec 07, 2017 7:05 am
by veronna
Hi,

Primarily, developing a CGE model is intended to examine the policy change using the changes in coefficients and exogenous variables.
Regarding to this matter, when I intended to examine a change in any endogenous variable (for example, a change in domestic good price),
I find myself lost in proceeding with the analysis.
Thus, may I know what can I do if I want to get to know the change impacts in endogenous variables?
Or, it could be done by creating a new parameter (to represent the related endogenous variable intended to change) before the solution?

Thank you very much in advance.

Re: A change in endogenous variables in general equilibrium

Posted: Thu Dec 07, 2017 2:47 pm
by Renger
Hi
You are on the right track. First, run the benchmark that replicates your data (usually, all prices are equal to 1). This is also a good test to see if you have errors in the model). Then do a counterfactual and compare the results.
parameter results(*,i) Results for the sectors
* PD0 is the benchmark price
results("Price change PD",i) = 100 * (PD.L(i) / PD0(i) - 1);
etc.

Re: A change in endogenous variables in general equilibrium

Posted: Sun Dec 10, 2017 4:53 pm
by veronna
Thanks Renger once again!

So, if I try to solve a simulation in a LOOP (as following), the endogenous variable should be fixed with the new parameter (for simulation) in order to make it as an exogenous variable.
Am I doing right?

LOOP(SIM,

PD.FX('AGR') = PAGRINCR(SIM);
SOLVE "name of model" USING MCP;

.
.
QARAP(A,SIM) = QA.L(A);
QDRAP(C,SIM) = QD.L(C);
QERAP(C,SIM) = QE.L(C);
QMRAP(C,SIM) = QM.L(C);
.
.
);

Thank you very much in advance.

Re: A change in endogenous variables in general equilibrium

Posted: Mon Dec 11, 2017 10:06 am
by Renger
Hi Veronna

No, this is not the way: you can easily change parameters in the model and rerun the model. If you fix a variable, you, in fact, drop the equation from the model, and this is not what you want (you want that even with the fixed variable, the complementary equation should be fine). If you fix a variable, you should free another variable or introduce one.
I always think of it this way: If I fix a variable, something else has to adjust to make sure that I can keep the variable at this level. For example, if I want to have the government income to be constant, I fix this variable, but in order to keep it at the fixed level, I have to introduce a tax/subsidy or transfer that is received/paid by the government to keep its income at the given level.
If you fix one price and leave everything else as before, the model doesn't have an "instrument" to keep the price at that level (all other variables are linked to the other equations).

Hope this helps
Cheers
Renger

Re: A change in endogenous variables in general equilibrium

Posted: Thu Jan 11, 2018 2:46 pm
by simonalia
In that case the reverse swing was the best option seriously