You are on page 1of 6

November 7, 2017

ECON 301: ECONOMETRICS I


Assignment 5 Answer Key
Part I: End-of-Chapter 6 Questions
3. (i) 0.0003/(2*0.000000007)= 21,428.57; remember, this is sales in millions of dollars.

(ii) Probably. Its t-statistic is about –1.89 whereas the one sided critical value for t29 is –1.70.

^
(iii) rdintens=2.613+0.30 salesbil−0.0070 salesbil
2

(0.429) (0.14) (0.037)

(iv) The equation in part (iii) is easier to read because it contains fewer zeros to the right of the
decimal. Of course the interpretation of the two equations is identical once the different scales
are accounted for.

C1. (i)
. reg lprice ldist if year==1981

Source | SS df MS Number of obs = 142


-------------+---------------------------------- F(1, 140) = 30.79
Model | 3.86427772 1 3.86427772 Prob > F = 0.0000
Residual | 17.573039 140 .125521707 R-squared = 0.1803
-------------+---------------------------------- Adj R-squared = 0.1744
Total | 21.4373167 141 .152037707 Root MSE = .35429

------------------------------------------------------------------------------
lprice | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
ldist | .3648756 .0657612 5.55 0.000 .234862 .4948891
_cons | 8.047154 .646241 12.45 0.000 6.769501 9.324807
------------------------------------------------------------------------------
I expect 0: all other relevant factors equal, it is better to have a home farther away from the
incinerator. The estimated equation implies a 1% increase in distance from the incinerator is
associated with a predicted price that is about .37% higher.

(ii)
. reg lprice ldist lintst larea lland rooms baths age if year==1981

Source | SS df MS Number of obs = 142


-------------+---------------------------------- F(7, 134) = 56.68
Model | 16.0253134 7 2.28933049 Prob > F = 0.0000
Residual | 5.41200327 134 .040388084 R-squared = 0.7475
-------------+---------------------------------- Adj R-squared = 0.7344
Total | 21.4373167 141 .152037707 Root MSE = .20097

------------------------------------------------------------------------------
lprice | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
ldist | .0553918 .0576207 0.96 0.338 -.0585718 .1693554
lintst | -.0390352 .0516612 -0.76 0.451 -.1412121 .0631417
larea | .3192937 .0764177 4.18 0.000 .1681529 .4704345
lland | .0768256 .0395044 1.94 0.054 -.0013071 .1549584
rooms | .0425276 .028251 1.51 0.135 -.0133478 .0984031
baths | .1669252 .0419441 3.98 0.000 .0839671 .2498833
age | -.0035673 .0010588 -3.37 0.001 -.0056614 -.0014732
_cons | 7.59232 .6417091 11.83 0.000 6.323131 8.861509
------------------------------------------------------------------------------
_cons | .2843595 .1041904 2.73 0.007 .0796755 .4890435
------------------------------------------------------------------------------

The coefficient on log(dist) becomes about .055 (se .058). The effect is much smaller now, and
statistically insignificant. This is because we have explicitly controlled for several other factors
that determine the quality of a home (such as its size and number of baths) and its location
(distance to the interstate). This is consistent with the hypothesis that the incinerator was located
near less desirable homes to begin with.

(iii)
. reg lprice ldist lintst larea lland rooms baths age lintstsq if year==1981

Source | SS df MS Number of obs = 142


-------------+---------------------------------- F(8, 133) = 58.11
Model | 16.668547 8 2.08356838 Prob > F = 0.0000
Residual | 4.76876969 133 .035855411 R-squared = 0.7775
-------------+---------------------------------- Adj R-squared = 0.7642
Total | 21.4373167 141 .152037707 Root MSE = .18936

------------------------------------------------------------------------------
lprice | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
ldist | .1852369 .0623485 2.97 0.004 .061914 .3085598
lintst | 2.072806 .5009726 4.14 0.000 1.081901 3.06371
larea | .359345 .0726203 4.95 0.000 .2157049 .502985
lland | .0913769 .0373799 2.44 0.016 .0174409 .1653128
rooms | .0381091 .026639 1.43 0.155 -.0145817 .0908
baths | .14955 .0397327 3.76 0.000 .0709602 .2281399
age | -.0029265 .001009 -2.90 0.004 -.0049224 -.0009307
lintstsq | -.1193199 .0281712 -4.24 0.000 -.1750414 -.0635983
_cons | -3.31713 2.645718 -1.25 0.212 -8.550259 1.915998
------------------------------------------------------------------------------

The coefficient on log(dist) is now very statistically significant, with a t statistic of about three.
The coefficients on log(inst) and [log(inst)]2 are both very statistically significant, each with t
statistics above four in absolute value. Just adding [log(inst)]2 has had a very big effect on the
coefficient important for policy purposes. This means that distance from the incinerator and
distance from the interstate are correlated in some nonlinear way that also affects housing price.

We can find the value of log(inst) where the effect on log(price) actually becomes negative:
2.073/[2(.1193)] = 8.69. When we exponentiate this, we obtain about 5,943 feet from the
interstate. Therefore, it is best to have your home away from the interstate for distances less than
just over a mile (Note: A mile is equal to 5280 feet). After that, moving farther away from the
interstate lowers predicted house price.
(iv)
. gen ldistsq=ldist*ldist

. reg lprice ldist lintst larea lland rooms baths age lintstsq ldistsq if year==1981

Source | SS df MS Number of obs = 142


-------------+---------------------------------- F(9, 132) = 51.32
Model | 16.6725098 9 1.85250109 Prob > F = 0.0000
Residual | 4.7648069 132 .036097022 R-squared = 0.7777
-------------+---------------------------------- Adj R-squared = 0.7626
Total | 21.4373167 141 .152037707 Root MSE = .18999

------------------------------------------------------------------------------
lprice | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
ldist | .8709897 2.070622 0.42 0.675 -3.224905 4.966885
lintst | 1.934032 .6542841 2.96 0.004 .639793 3.22827
larea | .3553214 .0738695 4.81 0.000 .2092002 .5014426
lland | .0878552 .0389826 2.25 0.026 .0107437 .1649666
rooms | .0380582 .026729 1.42 0.157 -.0148144 .0909309
baths | .1507075 .0400191 3.77 0.000 .0715456 .2298693
age | -.0028693 .0010271 -2.79 0.006 -.0049009 -.0008376
lintstsq | -.11073 .0383546 -2.89 0.005 -.1865993 -.0348607
ldistsq | -.0364946 .1101445 -0.33 0.741 -.2543714 .1813822
_cons | -5.920924 8.29479 -0.71 0.477 -22.32884 10.48699
------------------------------------------------------------------------------

It is not significant, so no need to add this complication to the model.

∂ log ⁡(wage)
C3. (i) = β1 + β 3 exper
∂ educ

(ii) H0: β3 = 0. H1: β3 > 0.

(iii)
. gen educexper=educ*exper

. reg lwage educ exper educexper

Source | SS df MS Number of obs = 935


-------------+---------------------------------- F(3, 931) = 48.41
Model | 22.3529774 3 7.45099246 Prob > F = 0.0000
Residual | 143.303317 931 .153924078 R-squared = 0.1349
-------------+---------------------------------- Adj R-squared = 0.1321
Total | 165.656294 934 .177362199 Root MSE = .39233

------------------------------------------------------------------------------
lwage | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
educ | .0440498 .0173911 2.53 0.011 .0099195 .0781801
exper | -.0214959 .0199783 -1.08 0.282 -.0607036 .0177118
educexper | .003203 .0015292 2.09 0.036 .000202 .006204
_cons | 5.949455 .2408264 24.70 0.000 5.476829 6.42208
--------------------------------------------------------------------------------------
The t-statistic is given by 2.09, which is associated with a p-value of less than which gives a p-
value .02 for the one sided alternative. We reject the null in favor of the alternative at any
significance level over 2 percent.
(iv)
. gen educexper10=educ*(exper-10)

. reg lwage educ exper educexper10

Source | SS df MS Number of obs = 935


-------------+---------------------------------- F(3, 931) = 48.41
Model | 22.3529774 3 7.45099246 Prob > F = 0.0000
Residual | 143.303317 931 .153924078 R-squared = 0.1349
-------------+---------------------------------- Adj R-squared = 0.1321
Total | 165.656294 934 .177362199 Root MSE = .39233

------------------------------------------------------------------------------
lwage | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
educ | .0760795 .0066151 11.50 0.000 .0630974 .0890617
exper | -.0214959 .0199783 -1.08 0.282 -.0607036 .0177118
educexper10 | .003203 .0015292 2.09 0.036 .000202 .006204
_cons | 5.949455 .2408264 24.70 0.000 5.476829 6.42208
------------------------------------------------------------------------------

It is given by (.0630974,.0890617).

C5. (i)
. reg lprice llotsize lsqrft bdrms

Source | SS df MS Number of obs = 88


-------------+---------------------------------- F(3, 84) = 50.42
Model | 5.15504028 3 1.71834676 Prob > F = 0.0000
Residual | 2.86256324 84 .034078134 R-squared = 0.6430
-------------+---------------------------------- Adj R-squared = 0.6302
Total | 8.01760352 87 .092156362 Root MSE = .1846

------------------------------------------------------------------------------
lprice | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
llotsize | .1679667 .0382812 4.39 0.000 .0918404 .244093
lsqrft | .7002324 .0928652 7.54 0.000 .5155597 .8849051
bdrms | .0369584 .0275313 1.34 0.183 -.0177906 .0917074
_cons | -1.297042 .6512836 -1.99 0.050 -2.592191 -.001893
------------------------------------------------------------------------------

(ii) lprice = 5.61 + .168 * log(20,000) + .700 * log(2,500) + .037(4) = 12.90.

To predict price, we use the equation ^ price= α^0 exp ⁡(lprice) where α^0 is the slope onexp ⁡( ^
lprice)
^
from the regression price on exp ⁡( lprice) (without an intercept).

. predict lpricehat
(option xb assumed; fitted values)

. gen explpricehat=exp(lpricehat)

. reg price explpricehat, no constant


option no not allowed
r(198);

. reg price explpricehat, noconstant

Source | SS df MS Number of obs = 88


-------------+---------------------------------- F(1, 87) = 2942.52
Model | 8256630.76 1 8256630.76 Prob > F = 0.0000
Residual | 244119.863 87 2805.97544 R-squared = 0.9713
-------------+---------------------------------- Adj R-squared = 0.9710
Total | 8500750.63 88 96599.4389 Root MSE = 52.971

------------------------------------------------------------------------------
price | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
explpricehat | 1.022913 .0188573 54.24 0.000 .9854316 1.060393
------------------------------------------------------------------------------

Therefore, for the values of the independent variables given above, price = (1.023)exp(12.90) =
$409,519.

If we forget to multiply by α^0 the predicted price would be about $400,312.

(iii)
. reg price lotsize sqrft bdrms

Source | SS df MS Number of obs = 88


-------------+---------------------------------- F(3, 84) = 57.46
Model | 617130.701 3 205710.234 Prob > F = 0.0000
Residual | 300723.805 84 3580.0453 R-squared = 0.6724
-------------+---------------------------------- Adj R-squared = 0.6607
Total | 917854.506 87 10550.0518 Root MSE = 59.833

------------------------------------------------------------------------------
price | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
lotsize | .0020677 .0006421 3.22 0.002 .0007908 .0033446
sqrft | .1227782 .0132374 9.28 0.000 .0964541 .1491022
bdrms | 13.85252 9.010145 1.54 0.128 -4.065141 31.77018
_cons | -21.77031 29.47504 -0.74 0.462 -80.38466 36.84405
------------------------------------------------------------------------------

. corr price explpricehat


(obs=88)

| price explpr~t
-------------+------------------
price | 1.0000
explpricehat | 0.8589 1.0000

When we run the regression with all variables in levels, the R-squared is about .672. On the other
hand, the square of the correlation between price and exp ⁡( ^ lprice) from part (ii) Is given by
2
0.859 = 0.738, which is the comparable goodness-of-fit measure for the model with log(price) as
the dependent variable. Therefore, for predicting price, the log model is notably better.
Part II – Other Questions:
O1. (i) You can think about taking logs of both sides, i.e. logE (Y | X )=log β 0 + β 1 log x .
(ii) Now, you can think about regressing log(y) on log(x); however, the resulting estimates will
not be unbiased of β 0 and β 1. The reason is the following:logE (Y | X ) ≠ E (log (Y )∨ X ).
O2. FED should not be concerned about the fact that interest rates are more significant than GDP
in a money demand regression because we do not make decisions based on which variable is
more statistically significant. If the FED is concerned about the impact of the policy, it should
focus on economic significance.

O3. Take the second model


Y = β0 + β 1 x1 + β 2 x 2 + β 3 (x 1−x 1)(x 2−x 2)+u .
Rearranging the terms, we get
Y = ( β 0+ β3 x 1 . x 2) +(β 1−β 3 x 2 )x 1 +(β 2−β 3 x 1) x 2 + β 3 x 1 x2 +u .
Therefore,
α 0=β 0 + β 3 x 1 . x 2, α 1=β 1−β 3 x 2, α 2=β 2−β 3 x 1, and α 3=β 3.

As these models are equivalent, there is no difference between them statistically.

Extra: Observe that β 1=α 1+ β3 x 2 and β 2=α 2+ β3 x 1. Therefore, the estimates of β 1 and β 2 would
provide the marginal effects w.r.t. x 1 and x 2 at the average, respectively. Therefore, it might be
easier to interpret model 2.

You might also like