Skip to main content


Project Type
UTRC Research Initiative
Project Dates
12/31/2007 - 08/31/2009
Principal Investigators
Project Status

The complex activity and travel patterns in the New York Metropolitan Region are currently modeled by the “Best Practice Model” (BPM), maintained by New York Metropolitan Transportation Council. Though representing the state of practice, it does not include a land use model that can create feedback loops between urban system and regional traffic patterns. A key component of a land use model is households’ residential location choice model. Currently, households’ residential locations are assumed fixed in the BPM model, which will be detrimental in the years ahead, especially amid the recent movement on “smart growth” and “mixed-use neighborhoods”. The basic idea behind this recent surge of interest in land reform is the hope to change people’s travel behavior by making activity opportunities more accessible to them. To achieve this, the very first step is to understand how people respond to density-driven or mixed-use neighborhoods. The lack of a land use component in the BPM model nullifies planners’ ability to examine these new policy scenarios.

The objective of this proposed research is to develop a theoretically sound and empirically tested model for the residential relocation decision comprising features noted above. This research will form an important basis for the regional BPM model when it is improved to include the land use component. The specific aims of this research include:

  • To develop a theoretical framework for a joint choice of residential mobility (whether to move) and residential location (where to move) possessing features noted above;
  • To empirically test this theoretical model. In particular, we are interested in answering the following questions:
    • How are various factors (e.g., commute time, accessibility to activities, and closeness to social networks) valued in a residential relocation decision?
    • How does the threshold value affect the probability of moving/staying?
    • Everything else being equal, will people tend to select new locations that are closer to their current locations?
    • How are people’s preferences in residential selection process influenced by the current location?

In this study, we will propose to develop a theoretically sound and empirically tested residential relocation model, called MGEV model. Compared to existing models, the proposed model has four unique features. First, the model must recognize a wide variety of factors that come into people’s housing decisions, for example accessibility to activities and closeness to one’s social networks, for example. Second, the model realizes that there is a psychological threshold value associated with the current location. The value of this psychological threshold may vary by households with unique characteristics. Understanding this threshold value is critical for a MPO agency to keep up with this recent movement of land use reforms. Third, the model must acknowledge the role that the current location plays in residential selection process, by allowing correlation between the current location and other alternative ones. Last, the developed model models two sequentially connected decisions together: a mobility decision (whether to move) and a residential relocation decision (where to move).

In order to test the model, this study also proposes an empirical data collection on households’ residential relocation choices in New York City. A sample of 500 households in the area will be sampled via mail-out and mail-back questionnaires. A variety of potential factors that influence the relocation choice will be examined, including socio-economic and demographic factors, land use variables, attitudinal information, and social network related variables etc. In addition to the collected dataset, we will supplement with other data sources, including census datasets, New York City land use datasets, and transportation network and volume information.

Though the collected sample size is relatively small, the proposed study will form an important theoretical and practical basis the next generation of New York Metropolitan Area’s Best Practice Model, as it prepares itself to incorporate a crucial land use component. In addition, the results of the proposed project will provide important policy insights on the effectiveness of the newest movement across the country today, which is density-driven and mixed-used neighborhoods.