Table 8: analysis of the importance of variables taken into account in the stated choice experiment
(MNL-estimation without considering possible heterogeneity effects)
MNL
|
Attributes______________________________________ |
Coefficient |
T-value |
P-value |
|
bedrooms |
0.2641 |
10.863 |
0.0000 |
|
monthly cost |
-0.0027 |
-6.416 |
0.0000 |
|
big city |
-0.4939 |
-14.599 |
0.0000 |
|
small town |
0.2842 |
9.229 |
0.0000 |
|
medium sized city (= - big city - small town) |
0.2097 |
- |
- |
|
travel costs |
-0.2914 |
-18.562 |
0.0000 |
|
travel time |
-0.0122 |
-4.961 |
0.0000 |
|
adjusted ρ2 |
0.1490 | ||
|
-2LogLikelihood |
-2990.6 |
By comparing the coefficients, the importance of the different variables in residential location
decisions can be assessed. From the viewpoint of studying the importance of a road pricing
policy on location choices, the comparison of trip related factors (i.e. travel time and travel
costs) on one hand and location based variables on the other hand is especially interesting.
These comparisons are presented in table 9. The table indicates how much extra travel time or
travel costs respondents seem to accept in order to attain a certain location benefit, overall
without being off better or worse (no disutility).
Table 9: location benefits compensated by trip costs and travel time (no disutility)
|
Compensation trip components | |||
|
Travel cost |
Travel time | ||
|
Location |
Save 1 euro on housing cost/day |
0Γ |
9“ |
|
1 bedroom extra |
1.8" |
43^ | |
|
Not living in a big city |
3Γ |
8^Γ | |
|
Living in a small town |
20" |
47“ | |
To be able to compare the influence of monthly housing costs on one hand and daily travel
cost and travel time on the other hand, the coefficient of monthly housing costs in table 6 has
been converted into costs per day. This makes comparison between the housing cost
component and trip related factors easier. Table 9 shows that respondents on average want to
pay 0.4 euro of travel cost per day extra (or accept an extra travel time of 9 minutes per day)
16