Date
kineat upemag model
Exansion!
PRE
uqrunin indupeodive Aange
(mCuhemuat
stondat di|cd vaui abes’
¥Y-7
sten tandordedv°akesi
tondadued
nclce0
Can be
hur ngeatt ust hare Lineas pados
modals.
ineas model,
fnchonad fmns
Aemi -dog- madel!.
hlcipocal models
model dne sthen nomal pobabiity plet
doer mot o
and
raget bina)
eiduals hae io
prndtie pa ing
’
Quny ftend ine)
Connol eyponunhal eqpomi model i
1:
EXAMPLE 6.3
Table 6.3 presents data on total personal consumption expenditure (PCEXP), expenditure on
Expenditure durable goods (EXPDUR), expenditure on nonduratble goods (EXPNONDUR), and expendi
on Durable ture on services (EXPSERVICES), all measured in 2000 billions of dollars.l3
Suppose we wish to find the elasticity of expenditure on durable goods with respect
Goods in to total personal consumption expenditure. Plotting the log of expenditure on durable
Relation to goods against the log of total personal consumption expenditure, you will see that
Total Persona! the relationship bebween the two variables is linear. Hence, the double-log model may be
appropriate. The regression results are as follows:
Consumption
Expenditure n EXDUR,= -5417 + 1.6266 In PCEX,
se = (0.7161) (0.0800) (6.5.5)
l=(-10.5309)* (20.3152)* '=0.9695
where indicates that the p value is extremely small.
TABLE 6.3 EXPSERVICES EXPDUR EXPNONDUR PCEXP
Total Personal Year or quarter
Expenditure and 2003-1 4,143.3 971.4 2,072.5 7,184.9
Categories 2003-1| 4,161.3 1,009.8 2,084.2 7,249.3
(Bilions of chained 2003-II| 4.190.7 1,049.6 2,123.0 7,352.9
|2000| dollars; 2003-IV 4,220.2 1,051.4 2,132.5 7,394.3
quarterly data at 2004-! 4,268.2 1,067.0 2,155.3 7,479.8
seasonally adjusted 2004-1| 4,308.4 1,071.4 2,164.3 7,534.4
annual rates) 2004-I|| 4,341.5 1,093.9 2,184.0 7,607.1
2004-IV 4,377.4 1,110.3 2,213.1 7,687.1
Sources: Depurtiment of
Lomherce, Bureau 2005-I 4,395.3 1,116.8 2,241.5 7,739.4
Econowic Report 2005-11 4,420.0 1,150.8 2,268.4 7.819.8
of the Prestdenu, 2007. 2005-II 4.454.5 1,175.9 2,287.6 7.895.3
Table B-17.p. 347. 2005-V 4.476.7 1,137.9 2,309.6 7,910.2
2006-1 4.494.5 1,190.5 2,342.8 8,003.8
2006-!1 4,535.4 1,190.3 2,351.1 8,055.0
2006-111 4,566.6 1,208.8 2,360.1 8.111.2
Note Sce Table B-2 for dta for tokal personal consumpion cpenditures for 1959-1989,
EXPSERVICES =apenditure on services, bilions of 2000 dollars
EXPNO
EXPDUR = etpertiture an durable goots, bilhons of 20 dollans.
KEXP loal nersonal consmotign unenditure hallions of 2000 dollurs
EXAMPLE 6.3 As these results show, the elasticity of EXPDUR with respect to PCEX is about 1.63, sug
esting tthat if total personal expenditure goes up by 1 percent, on average, trthe
gesting expendi
(Continued)
ture on durable goods goes up by about 1.63 percent. Thus, expenditure on durable goods
1S very
respon ochanges in personal consumption expenditure. This is one reason why
producers o durable goods keep a keen eye on changes in personal income and personal
consumption expenditure. In Exercise 6.18, the reader is asked to carry out a exer
cise for nondurable qoods expenditure
sLater plo dal f the dat
ine
be alfo
Lin hodel eowTH RATE M0DEL
and in t.
uh
tr
neaneo she lorstant
helahe Aveo abfolute chang n alue
4semi elhat
X00 att at
ot valee. tp in athal tane take
EXAMPLE 6.4 To illustrate the growth model (6.6.6), consider the data on expenditure on services given
The Rate of in Table 6.3. The regression results over time (t) are as follows:
Growth InEXS, 8.3226+ 0.00705t
Expenditure on se (0.0016) (0.00018) P=0.9919 (6.6.8)
Services t= (5201.625)* (39.1667)
Note: EXS stands for expenditure on services and denotes that the p value is extremely
small.
The interpretation of Equation 6.6.8 is that over the quarterly period 2003-1 to 2006-11,
expenditures on services increased at the (quarterly) rate of 0.705 percent. Roughly, this is
equal to an annual growth rate of 2.82 percent. Since 8.3226 = log of EXS at the begin
ning of the study period, by taking its antilog we obtain 4115.96 (billion dollars) as the
beginning value of EXS (i.e., the value at the beginning of 2003). The regression line
obtained in Eq. (6.6.8) is sketched in Figure 6.4.
FIGURE 6.4 8.44
F 8.42
3 8.40
.38
S.36
8.34
0 2 4 6 8 10 12 14 16
Time
DOMS Page No.
Date
Iastantane rus Compoud Rati
instantanenus
onpound bar penid Sine)
Caseo
-do ue the lingaL.
bima
ushich abtlutA
danuwald end
choia Hwaqdio and ino dependo mnedsAnere
nodul AESDLUTE change in or a chng
alaie changtin
the
chang
epandita modol.
EXAMPLE 6.5 As an illustration of the lin-og model, let us revisit our example on food expenditure in
India, Example 3.2. There we fitted a linear-in-variables model as a first approximation.
But if we plot the data we obtain the plot in Figure 6.S. As this figure suggests, food
expenditure increases more slowBy as total expenditure increases, perhaps giving credence
to Engel's law. The results of fitting the lin-Hog model to the data are as follows:
FoodExp = -1283.912 + 257.2700 In TotalExp,
(-4.3848)* (5.6625)" r²=0.3769 (6.6.14)
Note- denotes an extrermeBy srmall p vaue.
(Continued)
EXAMPLE 6.5 FIGURE 6.5
(Continued) 700
600
500
400
300
200
100
300 400 500 600 700 800 900
Total expenditure (Rs.)
Interpreted in the manner described earlier, the slope coefficient of about 257 means
that an increase in the total food expenditure ofl percent, on average, leads to about
2.57 rupees increase in the expenditure on food of the 55 farmilies included in the sample.
(Note: We have divided the estimated slope coefficient by 100.)
Before proceeding further, note that if you want to compute the elasticity coefficient
for the log-lin or lin-Hog models, you can do so from the definition of the elasticity coeffi
cient given before, namely,
Y
Elasticity =X
As a matter of fact, once the functional form of a model is known, one can compute elas
ticities by applying the preceding definition. (Table 6.6, given later, summarizes the elas
ticity coefficients for the various models.)
It may be noted that sometimes logarithmic transformation is used to reduce
heteroscedasticity as well as skewness. (See Chapter 11) A common feature of many
economic variables, is that they are positively skewed (e.g., size distribution of firms
distribution of income or wealth) and they are heteroscedastic. A logarithmic transforma
tion of such variables reduces both skewness and heteroscedasticity. That is why labor
economists often use the logarithms of wages in the regression of wages on, say, schooling,
as measured by years of education.
niide =-1283.9)2 + 257.2100 er Tbaltx;
RS7 270D
=5127
mode dx
inea
kin
A