Multi-scenario simulations on the interactions of jobs-housing based on agent-based model

Jobs and housing are two major elements in urban spatial structure. The relationship between jobs and housing affects urban planning and management, and has become one of the hot topics in urban study. Most of the current researches focus on the analyses of urban spatial structure and the empirical study of urban theory based on macro-models. These models, which are traditional "top-down" approaches, have limitations in reflecting the individual behaviors. In fact, the spatial relationship of jobs-housing is the outcome of population individual behaviors including jobs selecting and residential decisions. Hence, the spatial organization of jobs-housing is difficult to simulate with those macro-models. Studies indicated that agent-based models which are "bottom-up" approaches, can offer a way to simulate the complex problems that involve individual behaviors. This paper proposes an agent-based model (Labor Market Based Model of Residential Location - LMBMRL) to simulate the interactions between jobs and housing. In this model, labor economics theory is incorporated to define the influence of labor market on job searching behaviors of population individuals. An agent-based approach is used to simulate the job searching behaviors and the residential location decisions of individuals. The proposed model is applied to the center region of Dongguan, an emerging and renowned manufacturing metropolis in the Pearl River Delta region, China. The simulated results reflect the effect of labor market on residential decisions. Quantitative analysis is carried out to assess the influence of housing cost and commuting cost on jobs-housing balance. It is found that as the housing cost rises, the relationship of jobs-housing will become less balanced. When the commuting cost is reduced, the relationship of jobs-housing will be much more imbalanced. Moreover, it has revealed the jobs-housing characteristics of population in different economic sectors through multi-scenario simulations. The study results are significant for understandings of jobs-housing organization, which can provide important guidance for urban planning and management.