The mobility of urban residents has been expanding over time. Kitamura and susilo (2005) have shown that this expansion stems more from structural change (i.e. change in the relationship between travel behavior and demographic factors), than from change in demographic and socio-economic characteristics (for example, attributes of the individuals and households, such as, more women employed, the household size shrinking, and the resident population aging) of urban residents. Urry (2005) went to conjecture that this structural change is due to increasingly prevailing automobility, i.e., conversion of social and economic system and way of life to adapt to the ownership and use of the automobile. In this study, this conjecture is explored by examining automobility characteristics across lifecycle stages and across automobility cohorts over time. The level of automobility is operationally defined in this study in terms of: automobile ownership, total auto travel time, modal split, and the fraction of trip attraction in traditional central city in the study area. The Kyoto-Osaka-Kobe (Keihanshin) metropolitan area of Japan is the study area of this effort. Intra-household interaction has dominant influences on household members' activity and travel, and also it is closely associated with lifecycle stage. Lifecycle stage factor is introduced into the analytical scope of this study and regarded as a main factor through this research. Nine stages of household lifecycle are formulated according to the classification scheme of the family lifecycle stage. The classification scheme utilizes the criteria, which generally are age and marital status of household head, presence and age of children of head, presence of other relatives and non-relatives. On the other side, changing the built environment affects urban residents' travel behavior to a large extent. This study explores how automobility characteristics and travel activity behavior changed across lifecycle stages within different residential areas over time using statistical analyses. The results confirm that the residence area rather than lifecycle stage is a significant explainer for automobile ownership and automobile use. It further suggests that even within each lifecycle stage, change in the automobile use over time is suppressed in commercial and mixed commercial/residential areas. However, the fraction of automobile trips for suburbs, unurbanized areas, and autonomous areas increased over time in the range of 0 to 4 times depending on the lifecycle stage. Younger childless couple stage and all adults' stage are more auto-oriented in suburbs, unurbanized area, and autonomous areas, and this trend becomes stronger as automobility progresses. No significant differences were observed in the numbers of trips for households of the same lifecycle stage across different residential areas, suggesting that similarly active lifestyles exist. The results suggest that household members' age is also a strong explainer for the fraction of auto trips and total auto travel time, through a four variable ANOVA analysis, including lifecycle stage, residence area, time, and age effect. It has been pointed out that the elderly of these days behave differently than the elderly grew up with the automobile and have been using it ever since their habit forming ages. Thus another important factor introduced into this research is automobility cohort which is defined by grouping individuals who turn 20 years old during the time period indicated. Each time period is chosen with respect to the level of automobility. The following five cohorts are developed for the study area and used in the analysis: pre-war (up to 1945), pre-motorization (1946-1960), initial growth (1961-1970), mass-ownership (1971-1980), and multi-car ownership (1980-). Using the repeated cross-sectional data of Kyoto-Osaka-Kobe metropolitan area in 1970, 1980, 1990, and 2000, this study has attempted to offer a possible explanation of the increases in automobility characteristics by examining automobility characteristics of automobility cohorts. In addition, time effects and age effects are introduced into the analysis as in standard cohort analysis. It focused on statistical age-period-cohort analysis using the popular multiple classification APC model. The identifiability problem attendant with the use of APC model was discussed with repeated cross-sectional data. An interesting finding is shown that pre-war and pre-motorization cohorts show little, roughly 6%, increase on the fraction of auto trips and nearly unchanged on total auto travel time over 1970 through 2000, although their household automobile ownership has increased more than 2 times. Initial growth, mass-ownership, and multi-car ownership cohorts show a great growth of automobile ownership, the fraction of auto trips, and auto travel time from 1970 to 2000, but a little surprising result is that mass-ownership cohorts, not multi-car ownership cohorts, show the largest increases to rely on auto use over 1970 through 2000. The above results confirm that each cohort having certain automobility traits that are unique, especially in terms of auto use. The standard age-period-cohort analysis confirms that automobility cohort effect do exist, unfortunately, automobility cohort effect is not an important explainer for automobility characteristics, while time effect plays an important part in automobile ownership choice, and age effect mainly determines automobile use. An attempt at APC-RA model illustrate that residence area rather than time effect have the strongest impact on automobile ownership, and age effect is still a significant explainer for the fraction of auto trips and auto travel time. This result is different with the results of age-period-cohort analysis, which further emphasize that residence area is a significant explainer for household automobile ownership in the Kyoto-Osaka-Kobe metropolitan area from 1970 to 2000. Significant changes in demographic and socio-economic characteristics of urban resident have taken place over the past several decades. Most notable are: aging of the population and resulting increases in retired, non-employed individuals; decreasing household size caused by increasing fractions of single individuals and couples with fewer children; increased labor force participation by women; general increased in income; and increasing auto ownership and auto dependence. The overall effects on travel of these changes are complex and future trends are not immediately obvious, partly because some of the changes have opposite, cancelling effects on travel, and partly because these changes themselves are not independent but closely linked to each other. Prevailing tendencies in travel, however, have been expansion—urban residents' travel has continuously expanded over time in terms of total travel time (or distance), auto use, energy consumption, and the spatial extension of their action space. Will these trends continue into the future? Or will the trend change due to the aging of the urban population? Or are there other factors at work? If so, what are the magnitudes of demographic effects relative to theirs? The focus of this study is on auto travel. The analysis examines how auto travel has changed over time with changing demographics, residential location, and metropolitan structure. Simultaneous equations model systems are developed at the household level, with auto ownership, fraction of auto trips and total auto travel time as its dependent (or endogenous) variables. Their automobility characteristics are characterized and behavioral distinction identified through examination of the models' coefficient estimates. Using the repeated household travel survey results, the stability over time of the simultaneous equation system is statistically examined, and thereby the effects of demographics changes are separated from those of structural change. Using the results, it is shown how much of the change in urban auto travel is due to changes in demographics and how much is due to structural change. The statistical analyses have offered strong evidence that urban residents' auto use have been expanding. The results have further indicated that this expansion has been caused primarily by changes in the structural relationships even mixed changes in demographic factors have had opposite, cancelling effects on auto travel. In addition, the resultant model system is applied in a scenario analysis to forecast possible changes in future auto travel that will follow hypothetical demographic changes in the metropolitan area. To face the coming global energy crisis and air pollution issues, the above results with the findings of this study would suggest that significantly more sustainable behavior for society would be possible with more compact built environments that facilitate non-motorized and public transit travel. Unfortunately, it takes time, money, resources, and the political will to change the built environment and initial steps that educate the public such as voluntary travel behavior change may be necessary first steps on the move to more sustainable travel. As a suggestion for future works, more statistical analysis on interaction effects of three variables or four variables ANOVA analysis, including lifecycle stage, residence area, time, and age effects, need to be considered. Also, the interaction effects of age-period-cohort analysis need us to pay more attention on the future work. The simultaneous equations model system is developed as an attempt to explore how much of the change in urban travel is due to changes in demographics and how much is due to structural change, more endogenous variable could be considered in the future research, such as, residential location, commute distance, and commute trip mode choice. / Kyoto University (京都大学) / 0048 / 新制・課程博士 / 博士(工学) / 甲第14925号 / 工博第3152号 / 新制||工||1473(附属図書館) / 27363 / UT51-2009-M839 / 京都大学大学院工学研究科都市社会工学専攻 / (主査)教授 藤井 聡, 准教授 吉井 稔雄, 准教授 宇野 伸宏 / 学位規則第4条第1項該当