URBAN PLANNING & MOBILITY · DEVELOPMENT
A French developer was preparing the program for a new mixed-use neighborhood on 24 hectares of industrial brownfield in the metropolitan core : 2 800 homes, divided into 30 % social housing, 20 % intermediate housing and 50 % market-rate housing, with public facilities, retail space and tertiary activities.
Programming can create social mix on day 1. The question was different : what mechanisms will ensure that different populations are still living in the same neighborhood ten years later ?
THE PROBLEM
The political ambition was to create a neighborhood with lasting social mix, against the classic mechanism of social specialization in new neighborhoods. The known risk was drift from programmed social mix toward specialization at 10 years.
Several comparable projects in other French metropolitan areas had seen their initial social mix erode : either through progressive impoverishment as market-rate households left, or through progressive gentrification displacing lower-income households. These trajectories are not inevitable ; they are documented.
The mechanisms involved had been identified — perceived quality of public facilities, school, perceived safety, retail dynamics — but their interactions were not understood. The project team used our system to test 6 urban-programming configurations, with population trajectories projected over 120 months.
WHAT IS THE SAME FOR EVERYONE
WHAT DIFFERS FOR EACH GROUP
A MECHANISM : SOCIAL MIX IS A FLOW
Conceptual diagram. It does not represent any projected value : only the year-0 distribution, set by the program, is known.
A program can impose 30 / 20 / 50 at delivery. It does not decide what happens next : the following years are made up of individual residential decisions.
THE INSUFFICIENT VARIABLE
The programmed distribution solves the first-day question. It says nothing about the mechanisms that, over ten years, retain or drive away households. These mechanisms are known individually. Their interactions were what remained to be established.
WHAT THE PROGRAM PRODUCES
Social mix on day 1. The delivered neighborhood meets the political objectives: 30 % social, 20 % intermediate, 50 % market-rate.
WHAT IT DOES NOT PRODUCE
Duration. The drift mechanisms documented in comparable developments do not concern the initial distribution.
WHAT IS AT STAKE
Perceived quality of public facilities, school, perceived safety, retail dynamics, residential trajectories, neighborhood perception.
WHAT REMAINED TO BE ESTABLISHED
Their interactions were not understood. The challenge was to determine which ones are genuinely structuring over time.
WHAT IS TESTED
The six configurations combine the dimensions below. Several neighborhood configurations produce several population trajectories — it was this difference in trajectory, not the state at delivery, that was the object of the simulation.
4 200 synthetic households were interviewed individually on the 6 urban-programming configurations, with dynamic follow-ups on residential trade-off criteria. Trajectories were projected over 120 months, modeling population, schooling, retail-activity and perceived-safety dynamics.
SIMULATED POPULATION
The reconstructed population covers 1,4 million residents and workers in the metropolitan area, calibrated on public INSEE data, metropolitan surveys on residential trajectories and qualitative studies of social-mix dynamics in new European neighborhoods.
It is structured into 18 profiles crossing household structure, income level, residential situation, residential trade-off criteria and sensitivity to the quality of public facilities.
FAMILIES WITH CHILDREN
residential decisions are driven largely by schooling
HIGH-MOBILITY MARKET-RATE HOUSEHOLDS
greatest ability to choose, therefore the earliest departures
SOCIAL-HOUSING HOUSEHOLDS, MORE STABLE
constrained residential mobility, longer settlement horizon
WORKERS SENSITIVE TO PROXIMITY
trade-off centered on access to jobs and services in the metropolitan core
HOUSEHOLDS DECIDING ON THE BASIS OF SCHOOL
perceived quality of the primary school outweighs other criteria
HOUSEHOLDS SENSITIVE TO SERVICES AND SHOPS
the local retail offer weighs on the decision to stay
HOUSEHOLDS SENSITIVE TO PERCEIVED SAFETY
the local safety arrangement matters more than the facility itself
INTERMEDIATE HOUSEHOLDS
the most finely balanced trade-off between housing cost and facility quality
These configurations make part of the population’s heterogeneity visible. The simulation concerns synthetic individuals, not a few persona types : the population is not reduced to lower-income households versus affluent households, but to distinct residential trade-off logics.
REACTIONS
STRICT SIZING ERODES SOCIAL MIX
Our system identified a structural dynamic: the initial social mix of a new neighborhood erodes when public facilities — school, sports facilities, green spaces, safety arrangements — are sized strictly for the expected program. Market-rate households, the first able to choose where they live, leave the neighborhood as soon as the perceived quality of facilities deteriorates.
A HIGHER STANDARD REVERSES THE DYNAMIC
Sizing public facilities to the standard of the metropolitan area’s more affluent neighborhoods, regardless of the neighborhood’s social program, reverses this dynamic: market-rate households stay, social mix holds. Equal sizing to the expected population is therefore not equal trajectory.
ONE VARIABLE OUTWEIGHS THE OTHER TWENTY-ONE
Among the 22 variables tested on the 120-month population trajectory, one structurally dominates: perceived quality of the neighborhood’s primary school. The other variables are secondary. This result applies to this project and this simulated population; it is not stated as a general law of new neighborhoods.
MIXED RETAIL PROGRAMMING IS MORE STABLE
Prestige programming initially attracts market-rate households, but creates a retail exodus at 5 years when lower-income households cannot afford it and it loses its local customer base. Mixed programming — accessible neighborhood shops, dining at varied price points, popular cultural offerings — produces a stable retail trajectory at 120 months and supports residential social mix.
Equal sizing is not necessarily equal trajectory. A facility calibrated exactly to the expected population is compliant ; it is not necessarily stabilizing.
22 VARIABLES, ONE DOMINANT VARIABLE
Among the 22 variables tested, one structurally dominates the 10-year population trajectory : perceived quality of the neighborhood’s primary school. This result applies to this project and this simulated population.
It says nothing about the actual quality of the schools concerned : the modeled variable is a perception, and its effect is residential behavior.
TIME LAG
TODAY
The neighborhood’s public school is perceived as deteriorating.
THEN
Market-rate households with children send them to private school.
LATER
They move, after an average of around four years.
FINALLY
The neighborhood’s social structure changes.
The most structuring effect is not immediate. An indicator measured at delivery, or even at twelve months, would not have captured it.
RETAIL PROGRAMMING
Two retail-programming approaches were compared. They do not differ in their initial attractiveness, but in the population that can actually use them over time.
PRESTIGE PROGRAMMING
Premium brands, high-end dining, exclusive cultural offer. Initially attracts market-rate households.
WHAT HAPPENS AT 5 YEARS
Retail exodus: the offer becomes inaccessible to some residents and loses its local customer base.
MIXED PROGRAMMING
Accessible neighborhood shops, dining at varied price points, popular cultural offerings.
WHAT HAPPENS AT 120 MONTHS
Stable retail trajectory that supports residential social mix rather than displacing it.
THE LOOPS
The mechanisms identified in this case do not operate as a linear chain but as a loop : a perception changes residential behavior, which changes the population, which changes the local offer, which in turn changes perception.
These loops are those observed in this simulation. They do not constitute a universal causal model of mixed neighborhoods.
COMPARISON
Qualitative comparison derived from the simulated trajectories. No configuration is cost-free : sizing to expected need protects the development budget and weakens the trajectory ; prestige programming improves initial attractiveness and destabilizes retail at five years ; investing in the commons stabilizes the trajectory and commits the developer over the long term.
THE MOST ROBUST
30 / 20 / 50 social mix facilities at a higher standard enhanced primary school accessible mixed retail
THE MOST FRAGILE
Programmed distribution public facilities sized beyond expected need
THE EVALUATION HORIZON
WHAT IS MEASURED ON DAY 1
Compliance of the delivered program with the 30 / 20 / 50 distribution.
WHAT IS MEASURED AFTERWARDS
Who stays, who leaves, who arrives — and at what pace.
WHAT STABILIZES OR DOES NOT
The school, public facilities, the neighborhood’s retail dynamics.
WHAT THIS IMPLIES
The decision must be evaluated over the horizon on which its consequences appear, here projected over 120 months.
DECISION
PROJECTION, THEN OBSERVATION
The developer selected the strategy combining social-mix programming according to the initial political objectives (30 / 20 / 50), public facilities sized to the standard of the metropolitan area’s affluent neighborhoods — enhanced primary school with a structuring educational project, high-quality sports facilities, generous green spaces, local safety arrangement —, mixed retail programming, and governance of social mix through a monitoring system for trajectory indicators at 12, 36 and 60 months.
At 48 months from rollout, on delivery of the first phases, the programmed social mix is effectively in place. Perceived quality of the public school reaches the standard of the metropolitan area’s affluent neighborhoods. The exodus of market-rate households is zero at this stage. The monitoring arrangement is operational, with a review every 36 months. The methodology has been adopted by the developer for two other mixed-neighborhood projects in preparation.
This result is an observation at 48 months. It does not demonstrate the trajectory at 120 months, which remains a projection. The gap between the two is precisely what the monitoring arrangement is designed to measure.
TAKEAWAY
The problem was not building a mixed neighborhood. Programming was enough for that. The problem was giving households that could leave a reason to stay — and that reason had nothing to do with the housing distribution.
It depended on public facilities sized above expected need, a primary school built around a structuring educational project, and a retail offer that all residents could use. These parameters do not appear in the political objectives of a development operation. Yet they determine its trajectory.
POSSIBLE FUTURES
A
PROGRAM SOCIAL MIX
30 / 20 / 50 distribution and public facilities sized for expected need
B
ATTRACT THROUGH PRESTIGE
Facilities and retail programming oriented toward attracting market-rate households
C
INVEST IN THE COMMONS
30 / 20 / 50 social mix, facilities at a higher standard, enhanced school, mixed retail
METHOD
Trajectories were projected over 120 months, modeling population, schooling, retail-activity and perceived-safety dynamics.
This case adds a capability to the library : simulating not only how a project is received, but how a population and its environment transform one another. Residents arrive, stay, leave, choose schools, consume, create activity, change the perception of the neighborhood — and this new composition in turn produces new behaviors.
This case comes from a simulation conducted for a French developer. The organization is not named, no personally identifiable data entered the system, and the detailed results remain the property of the commissioning organization. The comparisons between configurations published here are qualitative ; the quantitative values cited are those documented by the commissioning organization.
The simulation focuses exclusively on the residential trade-offs of a synthetic population faced with different urban-programming configurations. It is neither a real-estate market study, nor a housing-policy evaluation, nor a forecast of land values.
Your next decision
Describe your need. We can point you to the right level of support.
State your decision. See the future it produces.
