Principles of Econometrics 4e Chapter 2 Solution
Principles of Econometrics 4e Chapter 2 Solution
CHAPTER 2
Exercise Solutions
Chapter 2, Exercise Solutions, Principles of Econometrics, 4e 22
EXERCISE 2.1
(a)
x y x x
( )
2
x x
y y
( )( ) x x y y
0 6 -2 4 3.6 7.2
1 2 -1 1 0.4 0.4
2 3 0 0 0.6 0
3 1 1 1 1.4 1.4
4 0 2 4 2.4 4.8
i
x
=
i
y =
( )
i
x x =
( )
2
i
x x =
( ) y y =
( )( ) x x y y =
10 12 0 10 0 13
2, 2.4 x y = =
(b)
( )( )
( )
2 2
13
1.3
10
x x y y
b
x x
= = =
2
b is the estimated slope of the fitted line.
( )
1 2
2.4 1.3 2 5 b y b x = = =
1
b is the estimated value of ( ) E y when 0 x = ; it is the intercept of the fitted line.
(c)
5
2 2 2 2 2 2
1
0 1 2 3 4 30
i
i
x
=
= + + + + =
5
1
0 6 1 2 2 3 3 1 4 0 11
i i
i
x y
=
= + + + + =
( )
5 5
2
2 2 2
1 1
30 5 2 10
i i
i i
x Nx x x
= =
= = =
( )( )
5 5
1 1
11 5 2 2.4 13
i i i i
i i
x y Nx y x x y y
= =
= = =
(d)
i
x
i
y
i
y
i
e
2
i
e
i i
x e
0 6 5 1 1 0
1 2 3.7 1.7 2.89 1.7
2 3 2.4 0.6 0.36 1.2
3 1 1.1 0.1 0.01 0.3
4 0 0.2 0.2 0.04 0.8
i
x
=
i
y
=
i
y
=
i
e
=
2
i
e
=
i i
x e
=
10 12 12 0 4.3 0
Chapter 2, Exercise Solutions, Principles of Econometrics, 4e 23
Exercise 2.1 (continued)
(e)
(f) See figure above. The fitted line passes through the point of the means, 2, 2.4 x y = = .
(g) Given
1 2
5, 1.3 b b = = and
1 2
y b b x = + , we have ( )
1 2
2.4 5 1.3 2 2.4 y b b x = = + = =
(h) 12 5 2.4
i
y y N y = = = =
(i)
2
2
4.3
1.4333
2 3
i
e
N
o = = =
(j) ( )
( )
2
2 2
1.4333
var 0.14333
10
i
b
x x
o
= = =
0
2
4
6
0 1 2 3 4
x
y Fitted values
Figure xr2.1 Observations and fitted line
Chapter 2, Exercise Solutions, Principles of Econometrics, 4e 24
EXERCISE 2.2
(a) ( )
| $2000 | $2000 | $2000
2 2 2
| $2000 | $2000 | $2000
180 215
180 215
y x y x y x
y x y x y x
X
P X P
= = =
= = =
| |
|
< < = < <
|
o o o
\ .
( )
180 200 215 200
100 100
2 1.5
0.9104
P Z
P Z
| |
= < <
|
\ .
= < <
=
(b) ( )
| $2000 | $2000
2 2
| $2000 | $2000
190
190
y x y x
y x y x
X
P X P
= =
= =
| |
|
> = >
|
o o
\ .
( )
190 200
100
1 1
0.8413
P Z
P Z
| |
= >
|
\ .
= s
=
-2 1.5
0
.
1
.
2
.
3
.
4
f
(
z
)
-5 0 5
z
Figure xr2-2a
-1
0
.
1
.
2
.
3
.
4
f
(
z
)
-5 0 5
z
Figure xr2-2b
Chapter 2, Exercise Solutions, Principles of Econometrics, 4e 25
Exercise 2.2 (continued)
(c) ( )
| $2000 | $2000 | $2000
2 2 2
| $2000 | $2000 | $2000
180 215
180 215
y x y x y x
y x y x y x
X
P X P
= = =
= = =
| |
|
< < = < <
|
o o o
\ .
( )
180 200 215 200
81 81
2.2222 1.6666
0.9391
P Z
P Z
| |
= < <
|
\ .
= < <
=
(d) ( )
| $2000 | $2000
2 2
| $2000 | $2000
190
190
y x y x
y x y x
X
P X P
= =
= =
| |
|
> = >
|
o o
\ .
( )
190 200
81
1 1.1111
0.8667
P Z
P Z
| |
= >
|
\ .
= s
=
Chapter 2, Exercise Solutions, Principles of Econometrics, 4e 26
EXERCISE 2.3
(a) The observations on y and x and the estimated least-squares line are graphed in part (b).
The line drawn for part (a) will depend on each students subjective choice about the
position of the line. For this reason, it has been omitted.
(b) Preliminary calculations yield:
( )( ) ( )
2
21 33 26.5 17.5
5.5 3.5
i i i i i
x y x x y y x x
y x
= = = =
= =
The least squares estimates are:
( )( )
( )
2 2
26.5
1.514286
17.5
x x y y
b
x x
= = =
( )
1 2
5.5 1.514286 3.5 10.8 b y b x = = =
(c) 33 6 5.5
i
y y N = = =
21 6 3.5
i
x x N = = =
The predicted value for y at x x = is
1 2
10.8 1.514286 3.5 5.5 y b b x = + = =
We observe that
1 2
y b b x y = + = . That is, the predicted value at the sample mean x is the
sample mean of the dependent variable y . This implies that the least-squares estimated
line passes through the point ( , ) x y . This point is at the intersection of the two dashed
lines plotted on the graph in part (b) .
2
4
6
8
1
0
1 2 3 4 5 6
x
y Fitted values
Figure xr2.3 Observations and fitted line
Chapter 2, Exercise Solutions, Principles of Econometrics, 4e 27
Exercise 2.3 (Continued)
(d) The values of the least squares residuals, computed from
1 2
i i i i i
e y y y b b x = = , are:
i
x
i
y
i
e
1 10 0.714286
2 8 0.228571
3 5 1.257143
4 5 0.257143
5 2 1.228571
6 3 1.285714
Their sum is 0.
i
e =
(e) ( )
1 0.714286 2 0.228571 3 1.257143 4 0.257143
i i
x e = + + +
( ) 5 1.228571 6 1.285714 + +
= 0
Chapter 2, Exercise Solutions, Principles of Econometrics, 4e 28
EXERCISE 2.4
(a) If
1
0, | = the simple linear regression model becomes
2 i i i
y x e = | +
(b) Graphically, setting
1
0 | = implies the mean of the simple linear regression model
2
( )
i i
E y x = | passes through the origin (0, 0).
(c) To save on subscript notation we set
2
. | = | The sum of squares function becomes
2 2 2 2 2 2 2
1 1
2 2
( ) ( ) ( 2 ) 2
352 2 176 91 352 352 91
N N
i i i i i i i i i i
i i
S y x y x y x y x y x
= =
| = | = | +| = | +|
= |+ | = |+ |
Figure xr2.4(a) Sum of squares for
2
|
The minimum of this function is approximately 12 and occurs at approximately
2
1.95. | =
The significance of this value is that it is the least-squares estimate.
(d) To find the value of | that minimizes ( ) S | we obtain
2
2 2
i i i
dS
x y x
d
= + |
|
Setting this derivative equal to zero, we have
2
i i i
b x x y =
or
2
i i
i
x y
b
x
=
Thus, the least-squares estimate is
2
176
1.9341
91
b = =
which agrees with the approximate value of 1.95 that we obtained geometrically.
10
15
20
25
30
35
40
1.6 1.8 2.0 2.2 2.4
BETA
S
U
M
_
S
Q
Chapter 2, Exercise Solutions, Principles of Econometrics, 4e 29
Exercise 2.4 (Continued)
(e)
Figure xr2.4(b) Fitted regression line and mean
The fitted regression line is plotted in Figure xr2.4 (b). Note that the point ( , ) x y does not
lie on the fitted line in this instance.
(f) The least squares residuals, obtained from
2
i i i
e y b x = are:
1
2.0659 e =
2
2.1319 e =
3
1.1978 e =
4
0.7363 e =
5
0.6703 e =
6
0.6044 e =
Their sum is 3.3846.
i
e =
0.7363 4 0.6703 5 0.6044 6 0 =
0
2
4
6
8
10
12
0 1 2 3 4 5 6
X1
Y
1
* (3.5, 7.333)
Chapter 2, Exercise Solutions, Principles of Econometrics, 4e 30
EXERCISE 2.5
(a) The consultants report implies that the least squares estimates satisfy the following two
equations
1 2
500 10000 b b + =
1 2
750 12000 b b + =
Solving these two equations yields
2 2
2000
250 2000 8
250
b b = = =
1
6000 b =
Therefore, the estimated regression used by the consultant is:
2.04672
2
i
e
N
o = =
it follows that
2
2.04672( 2) 2.04672 49 100.29
i
e N = = =
(b) The standard error for
2
b is
2 2
se( ) var( ) 0.00098 0.031305 b b = = =
Also,
2
2 2
var( )
( )
i
b
x x
o
=
Thus,
( )
( )
2
2
2
2.04672
2088.5
0.00098
var
i
x x
b
o
= = =
(c) The value
2
0.18 b = suggests that a 1% increase in the percentage of males 18 years or
older who are high school graduates will lead to an increase of $180 in the mean income
of males who are 18 years or older.
(d)
1 2
15.187 0.18 69.139 2.742 b y b x = = =
(e) Since ( )
2
2 2
i i
x x x N x =
, we have
( )
2
2 2 2
2088.5 51 69.139 = 245,879
i i
x x x N x = + = +
(f) For Arkansas
1 2
12.274 2.742 0.18 58.3 0.962
i i i i i
e y y y b b x = = = =
Chapter 2, Exercise Solutions, Principles of Econometrics, 4e 33
EXERCISE 2.8
(a) The EZ estimator can be written as
2 1
2 1
2 1 2 1 2 1
1 1
EZ i i
y y
b y y k y
x x x x x x
| | | |
= = =
| |
\ . \ .
where
1
2 1
1
k
x x
,
2
2 1
1
k
x x
=
, and k
3
= k
4
= ... = k
N
= 0
Thus,
EZ
b is a linear estimator.
(b) Taking expectations yields
( ) ( ) ( )
( ) ( )
2 1
2 1
2 1 2 1 2 1
1 2 2 1 2 1
2 1 2 1
2 2 2 1 2 1
2 2
2 1 2 1 2 1 2 1
1 1
1 1
EZ
y y
E b E E y E y
x x x x x x
x x
x x x x
x x x x
x x x x x x x x
(
= =
(
= | +| | +|
| | | |
= = | = |
|
\ .
Thus, b
EZ
is an unbiased estimator.
(c) The variance is given by
( ) ( )
2 2 2
var var( ) var
EZ i i i i i
b k y k e k = = = o
( ) ( ) ( )
2
2
2 2 2
2 1 2 1 2 1
1 1 2
x x x x x x
| |
o
| = o + =
|
\ .
(d) If
( )
2
~ 0,
i
e N o , then
( )
2
2 2
2 1
2
~ ,
EZ
b N
x x
(
o
| (
(
Chapter 2, Exercise Solutions, Principles of Econometrics, 4e 34
Exercise 2.8 (continued)
(e) To convince E.Z. Stuff that var(b
2
) < var(b
EZ
), we need to show that
( ) ( )
2 2
2 2
2 1
2
i
x x x x
o o
>
or that ( )
( )
2
2
2 1
2
i
x x
x x
>
Consider
( ) ( ) ( ) ( ) ( ) ( )( )
2
2 2 2
2 1 2 1 2 1 2 1
2
2 2 2
x x x x x x x x x x x x x x ( +
= =
Thus, we need to show that
( ) ( ) ( ) ( )( )
2 2 2
2 1 2 1
1
2 2
N
i
i
x x x x x x x x x x
=
> +
or that
( ) ( ) ( )( ) ( )
2 2 2
1 2 2 1
3
2 2 0
N
i
i
x x x x x x x x x x
=
+ + + >
or that
( ) ( ) ( )
2 2
1 2
3
2 0.
N
i
i
x x x x x x
=
+ ( + >
This last inequality clearly holds. Thus,
EZ
b is not as good as the least squares estimator.
Rather than prove the result directly, as we have done above, we could also refer Professor
E.Z. Stuff to the Gauss Markov theorem.
Chapter 2, Exercise Solutions, Principles of Econometrics, 4e 35
EXERCISE 2.9
(a) Plots of the occupancy rates for the motel and its competitors for the 25-month period are
given in the following figure.
The repair period comprises those months between the two vertical lines. The graphical
evidence suggests that the damaged motel had the higher occupancy rate before and after
the repair period. During the repair period, the damaged motel and the competitors had
similar occupancy rates.
(b) A plot of MOTEL_PCT against COMP_PCT yields:
There appears to be a positive relationship the two variables. Such a relationship may exist
as both the damaged motel and the competitor(s) face the same demand for motel rooms.
That is, competitor occupancy rates reflect overall demand in the market for motel rooms.
30
40
50
60
70
80
90
100
0 2 4 6 8 10 12 14 16 18 20 22 24 26
month, 1=march 2003,.., 25=march 2005
percentage motel occupancy
percentage competitors occupancy
Figure xr2.9a Occupancy Rates
40
50
60
70
80
90
100
40 50 60 70 80
percentage competitors occupancy
p
e
r
c
e
n
t
a
g
e
m
o
t
e
l
o
c
c
u
p
a
n
c
y
Figure xr2.9b Observations on occupancy
Chapter 2, Exercise Solutions, Principles of Econometrics, 4e 36
Exercise 2.9 (continued)
(c) The estimated regression is
= | +| =
| =
-30
-20
-10
0
10
20
30
0 4 8 12 16 20 24 28
month, 1=march 2003,.., 25=march 2005
r
e
s
i
d
u
a
l
s
Repair period
Chapter 2, Exercise Solutions, Principles of Econometrics, 4e 37
Exercise 2.9(f) (continued)
The expected occupancy rate for the damaged motel is
1 2
| +| during the repair period; it
is
1
| outside of the repair period. Thus
2
| is the difference between the expected
occupancy rates for the damaged motel during the repair and non-repair periods.
The estimated regression is:
j
b = |
1.3189 0.8993 1.2614 1.1882 0.8978 0.4140
The stocks Microsoft, General Motors and IBM are aggressive with Microsoft being the
most aggressive with a beta value of
2
1.3189 b = . General Electric, Disney and Exxon-
Mobil are defensive with Exxon-Mobil being the most defensive with a beta value of
2
0.4140 b = .
(c)
Firm Microsoft
General
Electric
General
Motors
IBM Disney
Exxon-
Mobil
b
1
=
j
o
0.0061 0.0012 0.0116 0.0059 0.0011 0.0079
All estimates of the
j
o are close to zero and are therefore consistent with finance theory.
The fitted regression line and data scatter for Microsoft are plotted in Figure xr2.10.
Fig. xr2.10 Scatter plot of Microsoft and market rate
(d) The estimates for
j
| given 0
j
o = are as follows.
Firm Microsoft
General
Electric
General
Motors
IBM Disney
Exxon-
Mobil
j
|
1.3185 0.8993 1.2622 1.1878 0.8979 0.4134
The restriction o
j
= 0 has led to small changes in the
j
| ; it has not changed the aggressive
or defensive nature of the stock.
-.4
-.3
-.2
-.1
.0
.1
.2
.3
.4
.5
-.20 -.15 -.10 -.05 .00 .05 .10
MKT-RISKFREE
M
S
F
T
-
R
I
S
K
F
R
E
E
Chapter 2, Exercise Solutions, Principles of Econometrics, 4e 39
EXERCISE 2.11
(a)
Figure xr2.11(a) Price against square feet for houses of traditional style
(b) The estimated equation for traditional style houses is:
2
68710 0.012063 PRICE SQFT = +
The marginal effect on price of an additional square foot is:
( )
( ) slope 2 0.012063
d PRICE
SQFT
dSQFT
= =
For a home with 2000 square feet of living space, the marginal effect is:
( )
( )( ) 2 0.012063 2000 48.25
d PRICE
dSQFT
= =
That is, an additional square foot of living space for a traditional home of 2000 square feet
is expected to increase its price by $48.25.
To obtain the elasticity, we first need to compute an estimate of the expected price when
2000 SQFT = :
( )
2
68710 0.0120632 2000 116963 PRICE = + =
Then, the elasticity of price with respect to living space for a traditional home with 2000
square feet of living space is:
( )
( )( )
2000
slope 2 0.0120632 2000 0.825
116963
d PRICE
SQFT SQFT
PRICE dSQFT PRICE
| |
c = = = =
|
\ .
That is, for a 2000 square foot house, we estimate that a 1% increase in house size will
increase price by 0.825%.
0
200000
400000
600000
800000
1000000
0 2000 4000 6000 8000
total square feet
sale price, dollars Fitted values
tangent
Figure xr2.11c Observations and quadratic fitted line
Chapter 2, Exercise Solutions, Principles of Econometrics, 4e 41
Exercise 2.11 (continued)
(d) Residual plots:
The magnitude of the residuals tends to increase as housing size increases suggesting that
SR3 ( )
2
var | e x = o [homoskedasticity] could be violated. The larger residuals for larger
houses imply the spread or variance of the errors is larger as SQFT increases. Or, in other
words, there is not a constant variance of the error term for all house sizes.
(e) SSE of linear model, (b):
2 12
1.37 10
i
SSE e = =
SSE of quadratic model, (c):
2 12
1.23 10
i
SSE e = =
The quadratic model has a lower SSE. A lower SSE, or sum of squared residuals, indicates
a lower value for the squared distance between a regression line and data points, indicating
a line that better fits the data.
-200000
0
200000
400000
600000
R
e
s
i
d
u
a
l
s
0 2000 4000 6000 8000
total square feet
Figure xr2.11d Residuals from linear relation
-400000
-200000
0
200000
400000
R
e
s
i
d
u
a
l
s
0 2000 4000 6000 8000
total square feet
Figure xr2.11d Residuals from quadratic relation
Chapter 2, Exercise Solutions, Principles of Econometrics, 4e 42
Exercise 2.11 (continued)
(f) The estimated equation for traditional style houses is:
( )
2
57728 212.611 SPRICE LIVAREA = +
The marginal effect of an additional 100 square feet is:
( )
( ) slope 2 212.611
d SPRICE
LIVAREA
dLIVAREA
= =
For a home with 1500 square feet of living space, the marginal effect is:
( )
( )( ) 2 212.611 15 6378.33
d SPRICE
dLIVAREA
= =
That is, adding 100 square feet of living space to a house of 1500 square feet is estimated
to increase its expected price by approximately $6378.
(d)
The quadratic model appears to fit the data better; it is better at capturing the
proportionally higher prices for large houses.
SSE of linear model, (b):
2 12
2.23 10
i
SSE e = =
SSE of quadratic model, (c):
2 12
2.03 10
i
SSE e = =
The SSE of the quadratic model is smaller, indicating that it is a better fit.
(e) The estimated equation for houses that are on large lots in the sample is:
2
113279 193.83 SPRICE LIVAREA = +
The estimated equation for houses that are on small lots in the sample is:
2
62172 186.86 SPRICE LIVAREA = +
0
200000
400000
600000
800000
10 20 30 40 50
living area, hundreds of square feet
selling price of home, dollars Fitted values
Fitted values
Figure xr2.12d Linear and quadratic fitted lines
Chapter 2, Exercise Solutions, Principles of Econometrics, 4e 45
Exercise 2.12(e) (continued)
The intercept can be interpreted as the expected price of the land the selling price for a
house with no living area. The coefficient of LIVAREA has to be interpreted in the context
of the marginal effect of an extra 100 square feet of living area, which is
2
2 LIVAREA | .
Thus, we estimate that the mean price of large lots is $113,279 and the mean price of small
lots is $62,172. The marginal effect of living area on price is $387.66LIVAREA for
houses on large lots and $373.72LIVAREA for houses on small lots.
(f) The following figure contains the scatter diagram of PRICE and AGE as well as the
estimated equation which is
( ) 2008
2008
51.053 0.877982 51.053 0.877982 0.220 51.246 VOTE GROWTH = + = + =
This suggests that the incumbent party will maintain the majority vote in 2008. However,
the actual vote share for the incumbent party for 2008 was 46.60, which is a long way
short of the prediction; the incumbent party did not maintain the majority vote.
(d) The figure below shows a plot of VOTE against INFLATION. There appears to be a
negative association between the two variables.
The estimated equation (plotted in the figure below) is:
2
6.08283 0.073489 WAGE EDUC = +
The marginal effect is therefore:
( )
( ) slope 2 0.073489
d WAGE
EDUC
dEDUC
= =
For a person with 12 years of education, the estimated marginal effect of an additional
year of education on expected wage is:
( )
( )( ) slope 2 0.073489 12 1.7637
d WAGE
dEDUC
= = =
That is, an additional year of education for a person with 12 years of education is expected
to increase wage by $1.76.
-30
-20
-10
0
10
20
30
40
50
60
0 4 8 12 16 20 24
EDUC
R
E
S
I
D
Chapter 2, Exercise Solutions, Principles of Econometrics, 4e 52
Exercise 2.15(e) (continued)
For a person with 14 years of education, the marginal effect of an additional year of
education is:
( )
( )( ) slope 2 0.073489 14 2.0577
d WAGE
dEDUC
= = =
An additional year of education for a person with 14 years of education is expected to
increase wage by $2.06.
The linear model in (b) suggested that an additional year of education is expected to
increase wage by $1.98 regardless of the number of years of education attained. That is,
the rate of change is constant. The quadratic model suggests that the effect of an additional
year of education on wage increases with the level of education already attained.
(f)
Figure xr2.15(f) Quadratic and linear equations for wage on education
The quadratic model appears to fit the data slightly better than the linear equation.
(g) The histogram of ln(WAGE) in the figure below is more symmetrical and bell-shaped than
the histogram of WAGE given in part (a).
Figure xr2.15(g) Histogram for ln(WAGE)
0
.
2
.
4
.
6
.
8
D
e
n
s
i
t
y
1 2 3 4 5
lwage
Chapter 2, Exercise Solutions, Principles of Econometrics, 4e 53
Exercise 2.15 (continued)
(h) The estimated log-linear model is
( )
( )
12
exp 1.60944 0.090408 12 14.796
EDUC
WAGE
=
= + =
( )
14
exp 1.60944 0.090408 14 17.728
EDUC
WAGE
=
= + =
Evaluating the marginal effects at these values we have
2
1.3377 12
1.6028 14
EDUC
dWAGE
b WAGE
EDUC dEDUC
=
= =
=
For the linear relationship the marginal effect of education was estimated to be $1.98. For
the quadratic relationship the corresponding marginal effect estimates are $1.76 and $2.06.
The marginal effects from the log-linear model are lower.
A comparison of the fitted lines for the linear and log-linear model appears in the figure
below.
Figure xr12.15(h) Observations with linear and log-linear fitted lines