動態的な空間的マイクロシミュレーションモデルを用いた社会シミュレーション

―京町家の取壊し分析を事例に―

抄録

 The purpose of this article is to evaluate a social simulation using a dynamic spatial microsimulation model for predicting demolitions of Kyomachiya, which are traditional wooden townhouses and core elements of the historical landscape of Kyoto City, Japan. This model is also applied to estimate the number of Kyomachiya surviving when preservation policies are introduced. The results are summarized as follows:
 (a) Spatially disaggregated synthetic microdata of Kyomachiya residents were constructed by combining multiple existing datasets in a manner whereby the sums of synthetic microdata agree with those of census datasets. Using synthetic microdata allows us to analyze detailed household demographics and the process of Kyomachiya demolitions at small area and individual levels.
 (b) Decision-making units such as individuals, households, and Kyomachiya can be modeled in the same way that they exist, behave, and interact with each other in the real world using object-oriented modeling. Another merit is that re-using and extending classes are possible due to object-oriented architecture.
 (c) The results of simulations show that, during the next 15 years, only 67.3% of Kyomachiya will be preserved and the proportion of the population aged 60 and over will increase from 43.9% to 51.6%. On the other hand, when a comprehensive preservation policy is implemented, those numbers are reduced to 82.5% and 49.2%, respectively. In this manner, a dynamic spatial microsimulation model is useful for understanding the process and cause-and-effect of Kyomachiya demolitions under the status quo. Furthermore, what-if simulations on the basis of Kyomachiya preservation policies help to evaluate which policy is most effective for reducing the number of demolitions.