2011 International Conference on Alternative Energy in Developing Countries and Emerging Economies
- 190 -
confidence level. The positive interactions between the
two characteristics and socio-economics (PE*Number car
and CN*Num family) were at 95% confidence interval.
There was a higher level of pollution emissions
interaction with number of cars in the family which
showed that the respondent would choose a biodiesel
fuel. Also, the higher the level of cetane number
interaction with the number of families, the more likely it
was that the respondent would select a biodiesel fuel
scenario (Table II).
TABLE
II
R
ESULTS FOR BINARY LOGIT MODEL FOR BIODIESEL FUEL
Attribute
Basic Conditional
Logit Model
Conditional Logit Model
With Interaction
Coefficient Std.
error
Sig.
Coefficient
Std.
error
Sig.
ASC
Fuel cost (FC)
Operation cost (OC)
Cetane number (CN)
Pollution emissions (PE)
Socio-economic
variables
Gender
3
Number in family
Income of family
Hour trip
Number of cars
Expenditure on fuel
FC*Num family
CN*Num family
FC* Number car
PE*Number car
FC* Expenditure fuel
PE* Expenditure fuel
13.421
-.745
-.371
.265
-.256
.845
.043
.035
.022
.017
.000
**
.000
**
.000
**
.000
**
.000
**
13.307
-.756
-.375
.367
-.284
.184
.307
-.058
.045
-1.819
.002
-.015
.022
.091
.031
.000
.000
1.693
.086
.036
.045
.029
.059
.163
.033
.024
.951
.001
.008
.009
.048
.015
.000
.000
.000
**
.000
**
.000
**
.000
**
.000
**
.002
**
.059
.079
.063
.056
.068
.064
.011
*
.060
.040
*
.082
.079
-2 Log likelihood
Significant
Pseudo R- Square
Number of observation
Sample size
7145.97
.000**
.115
634
6340
7105.02
.000**
.123
634
6340
** Significant at 1% level, *Significant at 5% level
Table III shows the implicit prices or marginal
willingness to pay (MWTP) values for each of the
biodiesel fuel attributes with the respective 95%
confidence intervals. These were all positive, implying
that respondents had a positive WTP for increases in the
quality or quantity of each attribute. It was also indicated
that the Bangkok residents had the most WTP for OC,
CN and PE respectively by using both analysis methods.
It was slightly different at about £0.0002 between CN and
PE with the basic CL method, and around £0.0001
difference between OC and CN with the interaction
m
ethod. In general, respondents’ average WTP level with
3
Male code is 1 and Female code is 2.
the CL model interactions was higher than respondents’
average WTP with the basic CL model (see Table III).
TABLE
III
E
STIMATES OF WTP FOR BIODIESEL FUEL IN BAHT
(
B
)
AND
£
4
PER
RESPONDENT USING PAIRED COMPARISON
Attribute
Mean WTP Baht(B) and £
Order
Basic CL model & CL model with interactions
Operation Costs
(OC)
0.50B (£0.0068) &0.50B (£0.0068)
1
Cetane Number
(CN)
0.36B (£0.0049) &0.49B (£0.0067)
2
Pollution
Emissions (PE)
0.34B (£0.0047) &0.38B (£0.0052)
3
The implicit prices are shown
in Table III above; there
was no evidence of estimates of compensating surplus.
To estimate overall WTP for biodiesel fuel, it is
necessary to include the ASC, which captures the
systematic
but
unobserved
information
about
respondents’ choices. In order to estimate th
e
respondents’ WTP for alternative biodiesel fuel
scenarios, the study concentrated on possible pollution
emission options.
The estimates of WTP for the pollution emission
scenarios are described in Table IV. These are marginal
estimates showing WTP for a change from the current
situation
5
. These estimated consumer surplus using the
CL model with interactions, both social and economic.
The variables were all set to the sample averages.
From Table IV, the decrease in percentage of pollution
emission generated higher benefits in both models. Recall
that the CL model with interactions had a better fit than
the basic model, and hence the former was the preferred
model as a result of preferences which varied across
individuals in accordance with their socio-economic
background.
The model parameters can also be used to estimate
value differences between profiles. For instance, it is
able to evaluate the overall fuel characteristics between
an alternative scenario and the current situation.
Estimates of willingness to pay were
calculated for five
alternative scenarios in choice sets in terms of the PC
survey.
The compensating surplus (CS) welfare value using
the interaction logit model indicates that the value
attached to scenario 5 was £0.2477 (18.1155 Baht). That
is, the average willingness to pay found in the Bangkok
survey residents for biodiesel fuel characteristics
(scenario 5) compared to diesel fuel (current situation)
was £0.2477 (Table V) . In the paper, the CS value is
used with the logit model with interaction because it has a
better fit than the basic model.
4
Average exchange rate 73.1304 Baht was £1 in 2005 (source:
.
5
No changed for pollution emission.
