3
$\begingroup$

I'm currently solving a new Keynesian model with government spending. It's the basic model version, with constant returns to scale on the only production factor work.

Everything has worked out fine in terms of solving the model with Dynare, the only thing that I don't understand is the output regarding P(Price level, not inflation) and W(Wages, not in real terms).

After the gov. spending shock hits, all variables move back into their steady state i.e. the impulse response functions return to zero as time goes on. This also is the case for >real

The only variables that do not return to their steady state are P and W(nominal). They both spike upwards as the shock hits and then return to a state even higher than that, as the shock fades out.

Why is this the case, what is the economic intuition behind it?

$\endgroup$

1 Answer 1

2
$\begingroup$

Try to give a look at what happens to inflation's IRF. If it stays positive for the whole horizon of the IRF then simply prices have increased over time at the inflation rate. I guess that any non-degenerate price level (nominal!) is compatible with such model structure, as its system is written down in growth rates, as that's what loglinearised variables are.

$\endgroup$
2
  • $\begingroup$ Yes of course! I have not thought of that, inflation does go up. And even though it goes back to zero, since it went up, prices must go up as well. Thanks a lot. Now I'm thinking about the wage level. Since real wages must go back to zero, wages have to move up as well, to compensate for higher prices. Can you interpret it like this? Or is there more to it? $\endgroup$
    – steven
    Commented Apr 4, 2018 at 17:20
  • $\begingroup$ Does your model features sticky wages as well as sticky prices? $\endgroup$
    – Ceschi
    Commented Apr 5, 2018 at 22:03

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.