| Summary: | 碩士 === 國立成功大學 === 都市計劃研究所 === 82 === In the past, the majority of the studies of automobile growth
has been concentrated on the problem of ownership. The equally
importent issue of their utilization has been ignored. Indeed,
auto choice and use are two different yet highly related pro-
blems. The factors influencing auto ownership and their utili-
zation are different, therefore, the strategies for controlling
them should be different as well. This is part of the reason
why the commonly used strategies of controlling automobile
growth in the pass have not been successful. Auto ownership is
a discrete choice while its use is a conti- nuous choice.
Therefore, auto choice and use is a mixed discrete and
continuous problem. This study tries to develop a disaggre-
gate discrete-continuous model for studying the behavior of
automobile holding and use in Taiwan. Zero consumption is
treated as an alternative in the discrete choice. Due to time
and budget constrain, this study surveyed only the households
of five major metropolitan areas in Taiwan to collect the data
for analysis. The model was estimated using Heckman''s two-stage
process. The auto choice is specified as Logit and Probit
models, and auto use is specified as regression model. Data
analysis and empirical estimation results indicated that the
most significant variable influencing households'' auto
ownership is "% use for work", and the major variable affecting
households'' auto utilization is "transit travel time".
Sensitive analysis also showed that these two variables are
quite sensi- tive in the choice of auto ownership and usage.
|
|---|
