# Problem in observation equations

**URL:** <https://forum.dynare.org/t/problem-in-observation-equations/14049>\
**Category:** General DSGE Modeling\
**Created:** [25 June 2019 11:49 UTC](https://forum.dynare.org/t/problem-in-observation-equations/14049 "2019-06-25T11:49:48Z")\
**Posts on this page:** 7\
**Page:** 1

<div class="post-metadata">

**Author:** ![Saurabh](https://forum.dynare.org/letter_avatar_proxy/v4/letter/s/ee7513/32.png) [@Saurabh](https://forum.dynare.org/u/Saurabh)\
**Post date:** [25 June 2019 11:49 UTC](https://forum.dynare.org/t/problem-in-observation-equations/14049/1 "2019-06-25T11:49:48Z")

</div>

Dear people, I have followed the “Guide to specifying…” in specifying observation equation.

I have a model with unit-root labour augmenting growth (mu) and investment specific technical growth (gz). Let us say, output in the model growth with g (which is some combination of mu and gz).  
Now my observation equation linking logged difference of GDP (without demeaning) with model counterparts looks like:

y\_obs = y\_hat - y\_hat(-1) + g\_hat + g\_ss + measurement\_error ;

But using this gives me non-zero residuals. I think I am doing this right. Please suggest where lies the problem?

---

<div class="post-metadata">

**Author:** ![jpfeifer](https://forum.dynare.org/user_avatar/forum.dynare.org/jpfeifer/32/5044_2.png) [@jpfeifer](https://forum.dynare.org/u/jpfeifer)\
**Post date:** [25 June 2019 19:12 UTC](https://forum.dynare.org/t/problem-in-observation-equations/14049/2 "2019-06-25T19:12:28Z")

</div>

The way you wrote it down, the data has not been demeaned and `y_obs` will have steady state `g_ss`

---

<div class="post-metadata">

**Author:** ![Saurabh](https://forum.dynare.org/letter_avatar_proxy/v4/letter/s/ee7513/32.png) [@Saurabh](https://forum.dynare.org/u/Saurabh)\
**Post date:** [26 June 2019 02:23 UTC](https://forum.dynare.org/t/problem-in-observation-equations/14049/3 "2019-06-26T02:23:18Z")

</div>

Thnx, but I dont understand.

---

<div class="post-metadata">

**Author:** ![Saurabh](https://forum.dynare.org/letter_avatar_proxy/v4/letter/s/ee7513/32.png) [@Saurabh](https://forum.dynare.org/u/Saurabh)\
**Post date:** [26 June 2019 02:55 UTC](https://forum.dynare.org/t/problem-in-observation-equations/14049/4 "2019-06-26T02:55:06Z")

</div>

Attached is the mode-file and datat file for your kind perusal.

[BE.mod](https://forum.dynare.org/uploads/short-url/rQuTFJMfonYCuw3HZ3ljKDN0nvG.mod) (9.0 KB) [final\_data.xlsx](https://forum.dynare.org/uploads/short-url/3CLHwo04xCEogilX9rkoM0VEM9a.xlsx) (27.6 KB)

---

<div class="post-metadata">

**Author:** ![jpfeifer](https://forum.dynare.org/user_avatar/forum.dynare.org/jpfeifer/32/5044_2.png) [@jpfeifer](https://forum.dynare.org/u/jpfeifer)\
**Post date:** [26 June 2019 19:11 UTC](https://forum.dynare.org/t/problem-in-observation-equations/14049/5 "2019-06-26T19:11:34Z")

</div>

I am saying you are missing:

```auto
steady_state_model;
dc=mu_ss + (alpha1/(1-alpha1))*gz_ss ;
di=mu_ss + (alpha1/(1-alpha1))*gz_ss ;
dy=mu_ss + (alpha1/(1-alpha1))*gz_ss ;
end;

```

---

<div class="post-metadata">

**Author:** ![Saurabh](https://forum.dynare.org/letter_avatar_proxy/v4/letter/s/ee7513/32.png) [@Saurabh](https://forum.dynare.org/u/Saurabh)\
**Post date:** [27 June 2019 09:04 UTC](https://forum.dynare.org/t/problem-in-observation-equations/14049/6 "2019-06-27T09:04:25Z")

</div>

Thank you so much Professor!!!

May I ask one more question,!! According to the variance decom, in my model, monetary policy shock (white-noise) has very less impact on the dynamics of nominal interest rate. But it has a large impact on inflation and even on other real variables. What is wrong!! Any suggestion?

---

<div class="post-metadata">

**Author:** ![jpfeifer](https://forum.dynare.org/user_avatar/forum.dynare.org/jpfeifer/32/5044_2.png) [@jpfeifer](https://forum.dynare.org/u/jpfeifer)\
**Post date:** [27 June 2019 10:55 UTC](https://forum.dynare.org/t/problem-in-observation-equations/14049/7 "2019-06-27T10:55:15Z")

</div>

Are you sure there is anything wrong and there is no good economic reason for this?