Chapter 2 Dynamically Complete
Lesson: Is our model dynamically complete? Use theory/literature, but test, test, test.Set Time Series
/Users/Sam/Desktop/Econ 645/Data/Wooldridge
time variable: year, 1913 to 1984
delta: 1 yearWe’ll estimate a Finite Distributed Lag model for the first difference of general fertility onto the first difference of personal exemptions. If our model is dynamically complete, then no additional lags of gfr or pe are needed.
Our First-difference model with one lag of pe
Source | SS df MS Number of obs = 70
-------------+---------------------------------- F(2, 67) = 1.20
Model | 43.7985353 2 21.8992676 Prob > F = 0.3063
Residual | 1218.19557 67 18.1820235 R-squared = 0.0347
-------------+---------------------------------- Adj R-squared = 0.0059
Total | 1261.99411 69 18.2897697 Root MSE = 4.264
------------------------------------------------------------------------------
D.gfr | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
pe |
D1. | -.0456497 .0294689 -1.55 0.126 -.1044699 .0131704
|
pe_1 |
D1. | .0134149 .0295326 0.45 0.651 -.0455324 .0723623
|
_cons | -.8372962 .5117337 -1.64 0.106 -1.858721 .1841286
------------------------------------------------------------------------------Is this dynamically complete? Let’s add a second lag of pe
Source | SS df MS Number of obs = 69
-------------+---------------------------------- F(3, 65) = 6.56
Model | 293.259859 3 97.7532864 Prob > F = 0.0006
Residual | 968.199959 65 14.895384 R-squared = 0.2325
-------------+---------------------------------- Adj R-squared = 0.1971
Total | 1261.45982 68 18.5508797 Root MSE = 3.8595
------------------------------------------------------------------------------
D.gfr | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
pe |
D1. | -.0362021 .0267737 -1.35 0.181 -.089673 .0172687
|
pe_1 |
D1. | -.0139706 .0275539 -0.51 0.614 -.0689997 .0410584
|
pe_2 |
D1. | .1099896 .0268797 4.09 0.000 .0563071 .1636721
|
_cons | -.9636787 .4677599 -2.06 0.043 -1.89786 -.0294976
------------------------------------------------------------------------------Add lag of d.gfr and see it wasn’t dynamically complete
Source | SS df MS Number of obs = 69
-------------+---------------------------------- F(4, 64) = 7.46
Model | 401.286162 4 100.32154 Prob > F = 0.0001
Residual | 860.173657 64 13.4402134 R-squared = 0.3181
-------------+---------------------------------- Adj R-squared = 0.2755
Total | 1261.45982 68 18.5508797 Root MSE = 3.6661
------------------------------------------------------------------------------
D.gfr | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
gfr_1 |
D1. | .3002422 .1059034 2.84 0.006 .0886758 .5118086
|
pe |
D1. | -.0454721 .0256417 -1.77 0.081 -.0966972 .005753
|
pe_1 |
D1. | .002064 .0267776 0.08 0.939 -.0514303 .0555584
|
pe_2 |
D1. | .1051346 .0255904 4.11 0.000 .054012 .1562572
|
_cons | -.7021594 .4537988 -1.55 0.127 -1.608727 .2044079
------------------------------------------------------------------------------