🏙️ Urban Gentrification — Rent, Migration & Displacement

In this simulation, new high-income arrivals preferentially choose vacant cells near already-affluent or high-rent neighbours rather than choosing locations at random. This mirrors real housing markets, where amenities, safety perceptions, and property values cluster spatially: an area that becomes attractive to one affluent household becomes more attractive to the next. As more affluent residents cluster in an area, the local rent level (driven by the income mix of neighbouring cells) rises further, which in turn increases the incentive for developers and subsequent affluent movers to invest there, reinforcing the initial spatial pattern rather than diffusing it evenly across the city.

A cell occupied by a low- or mid-income resident becomes vacant once its local rent exceeds that resident's fixed budget threshold; the model then attempts to relocate that resident to a cheaper cell elsewhere on the grid, or marks them as displaced if none is found. Real-world displacement is more nuanced — it includes direct eviction, rent increases at lease renewal, conversion of rental units to condominiums, and "exclusionary displacement" where departing low-income residents are simply not replaced by others of similar income when they eventually move for unrelated reasons — but the simplified budget-threshold mechanism captures the core economic logic identified in housing-affordability research.

Cities have experimented with rent stabilisation and rent control to cap how quickly rents can rise for existing tenants, inclusionary zoning that requires new developments to include a share of below-market-rate units, community land trusts that remove land from speculative markets permanently, right-to-return policies for displaced residents, and property tax relief for long-term homeowners in appreciating neighbourhoods. Evidence on effectiveness is mixed: measures that constrain rent growth can reduce displacement but may also discourage the very investment and housing supply growth that could otherwise ease affordability pressure city-wide.

Research findings are more mixed than popular discourse often suggests. Some studies find gentrifying neighbourhoods see improved public services, reduced crime, and increased property values that benefit existing homeowners, while renters — who hold no equity and face the sharpest rent increases — bear a disproportionate share of the costs. The distributional effects depend heavily on local tenure patterns (renter versus owner share), the strength of tenant protections, and whether new housing supply is added fast enough to absorb rising demand without primarily displacing incumbents.

Both models are cellular-automaton-style agent-based models where individually rational relocation decisions produce a large-scale spatial pattern that no single agent intended. Schelling's model shows how mild same-type preferences produce sharp residential segregation; this gentrification model shows how income-based location preferences combined with a rent feedback mechanism produce spatial sorting by income and displacement over time. Both illustrate the broader point in urban economics and sociology that emergent city-scale patterns often cannot be predicted just by looking at individual household preferences in isolation.

Urban economists commonly combine decennial census tract data, American Community Survey estimates, property assessment and sale-price records, eviction filing databases, and increasingly, address-level migration data derived from credit-bureau or tax records to track neighbourhood income composition and displacement over time. Studies such as the Urban Displacement Project at UC Berkeley classify census tracts into stages (susceptible, early, ongoing, or advanced gentrification, or continued exclusion) using these combined indicators, closely paralleling the discrete states this simulation's cells pass through.