Faculty of social studies department of economics supply response of tobacco output to price changes in



Download 1.11 Mb.
View original pdf
Page16/26
Date13.07.2023
Size1.11 Mb.
#61703
1   ...   12   13   14   15   16   17   18   19   ...   26
taku dissertation
4.2 Unit Root Test Results
Unit root tests using the Augmented Dickey-Fuller test were conducted for all variables except of the sales quota dummy and time trend variables. As shown in Table 2, lagged price for maize
(lnPM
t-1
), population of tobacco farmers (lnATG
t
) and lagged tobacco prices (lnPT
t-1
) were found stationary in levels that is were I. Total tobacco output (lnTOUT
t
) and lagged tobacco output (lnTTOUT
t-1
) were stationary after first differencing that is they were I(1).
Table 2: Results for unit root test at level
Variable
ADF test
Critical value
Critical value
Critical value
Decision statistic at at at not stationary ln(TOUT
t-1
)
0.33
-2.63
-1.95
-1.61
not stationary ln(PT
t-1
)
-2.95**
-3.63
-2.95
-2.61
I(0)
ln(PM
t-1
)
-5.49*
-4.25
-3.54
-3.21
I(0)
ln(ATG
t
)
-3.31*
-4.24
-3.54
-3.20
I(0)
*(**) [***] means stationary at 1 % (5 %) [10 %] level, Ii means order of integration



Table 3: Results for unit root test at first difference
Variable
ADF
test
Critical
Critical value
Critical value
Decision statistic value at at at 10%
dln(TOUT
t
)
-6.52*
-2.63
-1.95
-1.61
I(1)
dln(TOUT
t-1
)
-5.83*
-2.63
-1.95
-1.61
I(1)
*(**) [***] means stationary at 1 % (5 %) [10 %] level, d means first difference, Ii meansi
order of integration
18


4.3 Model Diagnostic Tests Results
Table 4: Correlation Matrix Results
Variable dln(TOUT
t-1
)
ln(PT
t-1
)
ln(MP
t-1
)
ln(ATG
t
)
SQ
t-1
T
dln(TOUT
t-1
)
1
ln(PT
t-1
)
0.27 1
ln(MP
t-1
)
-0.01 0.54 1
ln(ATG
t
)
-0.08 0.12 0.17 1
SQ
t-1
-0.03 0.04
-0.16
-0.15 T 0.07 0.58 0.17
-0.55 Table 4 presents the correlation matrix showing the degree of linear association between a pair of regressors. The correlation coefficients are less than 0.8 implying that each variable’s effect on the dependent variable can easily be isolated. The model does not suffer from serious problems of multicollinearity, therefore the OLS approach can be used for estimation.

Download 1.11 Mb.

Share with your friends:
1   ...   12   13   14   15   16   17   18   19   ...   26




The database is protected by copyright ©ininet.org 2024
send message

    Main page