skip to primary navigationskip to content

Joseph Ferreira: Abstract and background reading

Title image Joseph Ferreira

SHORT ABSTACT: Using a relatively new agent-based microsimulation platform, SimMobility, we examine the likely impacts of substantial accessibility improvements on housing market dynamics.  Many metropolitan areas are considering pilot projects that might make urban neighborhoods both more sustainable and accessible through combinations of autonomous vehicle technologies and urban design.  Such programs have the potential to reduce car ownership, and emissions, but risk undermining much of the sustainability benefit as a result of induced gentrification in the local housing market.  Any investment in accessibility improvement should be matched with appropriate housing market policies and programs.


Modeling land use and transportation interactions (LUTI) has a long history and everyone agrees that emerging urban mobility technologies (such as autonomous vehicles) could change mobility enough to have substantial impact on land use patterns.  However, there is limited agreement on the direction and scale of these effects.  Because it will take many years, and much planning, for autonomous vehicles to replace the current fleet, a number of metropolitan areas are planning pilot programs within selected urban neighborhoods to test various implementations of the new technologies.  If the initial experience with the new technologies is perceived to be problematic, rules and regulations will follow quickly and could substantially constrain future implementation paths.  We need only look at the recent haphazard introduction of dockless bikes and e-scooters to see how early experience with new technologies can constrain regulatory paths.

Computational general equilibrium models (CGE) can tell us a lot about long term equilibrium conditions but, with the new technologies, the urban activity patterns could change fairly quickly, and much faster than significant land use changes.  Agent-based LUTI models simulate more of the land use and transportation dynamics, but they tend to balance supply and demand for land use at an aggregate level with yearly adjustments.  The land use component of SimMobility models housing market dynamics using behavioral models and a daily bid-auction approach that (1) accounts explicitly for accessibility and other amenities when valuing housing at building scale and (2) differentiates household willingness-to-pay depending upon household characteristics, including vehicle ownership.  

Using SimMobility, we examine the housing market dynamics of Singapore for the first few years of a hypothetical ‘car-lite’ pilot program that substantially improves the accessibility of residents who live and/or work within a designated study area and do not own private vehicles.  Several scenarios are compared with baseline conditions.  The scenarios differ depending upon the extent to which buyers and sellers value the anticipated improvement in accessibility.   The results indicate that such programs can indeed reduce car-ownership.  However, depending upon market conditions and buyer/seller behavior, housing relocation choices can largely undermine the gains from reduced car ownership through gentrification effects.  Coupling the ‘car-lite’ initiative with appropriate housing development and regulation can reduce car ownership while mitigating much of the adverse gentrification.



Yi Zhu, Mi Diao, Joseph Ferreira Jr., P. Christopher Zegras, “An integrated microsimulation approach to land-use and mobility modeling,” J. of Transport and Land Use, Vol. 11 No. 1 (2018) pp. 633-659

Basu, R. and Ferreira, J. (2020). “A LUTI microsimulation framework to evaluate long-term impacts of automated mobility on the choice of housing-mobility bundles,” Environment and Planning B: Urban Analytics and City Science.