Scribd
Upload
22 views

Econometrics Assignment Week 4

This document summarizes key points from an econometrics course: - It defines endogenous and exogenous variables and identifies which are which in the given models. - It explains that equations 1-3 are economic and econometric models respectively, with equation 1 not needing to be estimated. - It provides details on the identification of equations 2 and 3, finding they are both exactly identified. - It reports output from estimating equations 2 and 3 separately via OLS, including parameter estimates and significance tests. - It performs a test for autocorrelation and finds evidence against the null hypothesis of no autocorrelation. - It explains the potential for introducing autocorrelation through a variable

Uploaded by

jantien De Groot
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
22 views

Econometrics Assignment Week 4

This document summarizes key points from an econometrics course: - It defines endogenous and exogenous variables and identifies which are which in the given models. - It explains that equations 1-3 are economic and econometric models respectively, with equation 1 not needing to be estimated. - It provides details on the identification of equations 2 and 3, finding they are both exactly identified. - It reports output from estimating equations 2 and 3 separately via OLS, including parameter estimates and significance tests. - It performs a test for autocorrelation and finds evidence against the null hypothesis of no autocorrelation. - It explains the potential for introducing autocorrelation through a variable

Uploaded by

jantien De Groot
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 6

Econometrics week 4

Jantien de Groot & Annerijn Leegte

a.
Endogenous: Yt, Ct, It, these variables are explained by the model
Exogenous: Ct-1, Rt, these variables are not explained by the model, but from outside the model

b.

Equation 1 is an economic model because there is no residuals and parameters. Whereas 2


and 3 are econometric models because there are residuals and parameters. So, equation 1 is
not to be estimated.
c.
Equation (2): g = 2 (Y, Ct), k = 1 (Rt), k = g – 1 so equation 2 is exactly identified.
Equation (3): g = 2 (It, Yt), k = 1 (Ct-1), k = g – 1 so equation 3 is exactly identified.

d.

c: dependent variable

_constant:
Sign = 𝛼1
Size = 46432.98
Significance= P = 0.000 so _constant is significant and not equal to 0
Meaning = If y and Ct-1 are 0. Than c is 46432.98.
Yt :
Sign = 𝛼2
Size = 0.2236
Significance = P = 0.00, so Y is significant and not equal to 0
Meaning = If Y increases by 1 unit, c increases with 0.223613. Holding C t-1 constant.

Ct-1:
Sign = 𝛼3
Size = 0.3607
Significance = P = 0.004, so L is significant and not equal to 0
Meaning = If Ct-1 increases with 1 unit, c increases with 0.3606744. Holding Yt constant.

Equation (3)

i: dependent variable

_constant:
Sign = ß1
Size = -10114.58
Significance =P = 0.687, _constant is not significant, and could be equal to 0.
Meaning = If Y and R are 0. I is -10114.58

Yt:
Sign = ß2
Size = 0.1818939
Significance = P = 0.000, so Yt is significant and not equal to 0.
Meaning = If Yt increases with 1 unit, I increases with 0.1818939. Holding R constant.

R:
Sign = ß3
Size = 1276.766
Significance = P = 0.559, so R is not significant and could be equal to 0.
Meaning = If R increases with 1 unit, I increases with 1276.766. Holding Yt constant.
e.
H 0 : ρ=0 , H a : ρ ≠ 0

Command: estat bgodfrey, nomiss 0


TS = 7.771
Critical value: display invchi2 (1,.95) = 3.8414588
Conclusion: TS > critical value so reject H0, there is first-order autocorrelation.

f.
if Ct-1 is used as a variable we can introduce autocorrelation in this model.

g.
H 0 :cov ( Y t , u2 t ) =0 , H 1 :cov ( Y t ,u2 t ) ≠ 0

Step 1: reduced model: reg y lc r


Predict vhat, residuals
Step 2: Original model with saved residuals: reg i y r vhat
Step 3: the t-test of the residuals differs significantly from 0(t=0.03). Therefore the H0 is rejected.
Estimation using OLS would give biased and inconsistent estimates.
C4

without command:

With command, fist stage:

With command second stage:

h.
OLS:
2SLS:

The estimates of the OLS and 2SLS vary from each other, especially for r for OLS: 1276.766
and for 2SLS: -582.9555. This indicates that there is simultaneity bias in this equation. So
2SLS is more efficient.
i.

Command: reg3 (two: c y lc) (three: i y r), inst (r, lc)

There is a gain in efficiency, because the parameter estimates are almost the same. The
standard errors are smaller for the system estimates. For y it decreased from 0.3547 to
0.3336 and for r is decreased from 2417.305 to 2273.572. The p-values are also lower.
For r it decreased from 0.812 to 0.798.

Efficiency can be expected, because there are endogenous variables in the model with
cov(x1,ε)≠0. Also the two equations are related through residuals. If there is a shock in
the economy it will have an effect for the residuals from the household consumption and
the investment by companies.

You might also like