In the central scenario, a 22.2 percent population decline by 2076 coincides with 3.5 percent more households. The same run can still produce a shortage of roughly 54,000 habitable homes after the city has lost 266,000 people.
Evidence status: Modelled. Sensitivity: high. Depends on GSX:POP-009:A01, GSX:POP-009:A02. This is a hypothetical region, not a forecast for a named Gauteng municipality. Housing demand follows households, so a falling headcount is not itself a falling housing demand. The shortage also depends on the stock deteriorating and on repairs being misallocated.
C01
Related questions
What happens to house prices when population declines?
Can housing scarcity persist while population falls?
How should cities and infrastructure adjust when a country shrinks?
2
Assumptions
Assumptions
These are the registered model assumptions. A result stays conditional on the ones a claim names.
A01Hypothetical Gauteng-scale city-regionThe model follows a hypothetical region from 2026 to 2076. It starts with 1.2 million people, about 420,000 households, and 430,000 habitable homes. The zones are stylised. They are not labelled real Gauteng neighbourhoods.1.2 million people in 2026
A02Central shrinkage pathIn the central scenario, population falls by 0.5 percent a year and average household size falls from 2.86 to 2.15.0.5 percent population decline per year
1
Model
Model
The registered model statement. It is a scenario model, and the manuscript remains the full specification.
M01Shrinking city-region accountsThe model separates population, households, school-age children, and network length across four zones. Housing demand follows households. School demand follows children. Infrastructure cost follows the network, not only the litres sold.GSX:POP-009:M01
2
Results
Results
Modelled scenario results. They are not forecasts.
R0154000 habitable homes short in 2076By 2076 the region has 934,000 people and 434,000 households. Population is 22.2 percent smaller and the number of households is 3.5 percent larger. Misallocated construction and repairs can leave a shortage of roughly 54,000 habitable homes.Central scenario, deteriorating stock and misallocated repairs
R02366.2 billion constant rand, fifty-year present value under managed shrinkageManaged shrinkage keeps school-seat use near 82 percent rather than 47 percent and reduces the fifty-year present value of the modelled system bill from R384.8 billion to R366.2 billion. Under delayed adjustment, satellite and outer-edge coverage fall to 41 and 72 percent by 2076.Managed shrinkage compared with delayed adjustment
Sensitivity
High · 1
Moderate · 1
Low · 0
Limitations
This is a hypothetical region, not a forecast for a named Gauteng municipality. Housing demand follows households, so a falling headcount is not itself a falling housing demand. The shortage also depends on the stock deteriorating and on repairs being misallocated.
The rand saving is not the whole argument. Delayed adjustment leaves the satellite town and outer edge at 41 and 72 percent maintenance coverage by 2076 while the centre stays above cost. Uniform policy can transfer money from viable networks to networks that no longer have enough users.
2
Claims
Claims
Each claim names the assumptions it uses and the limitation that travels with it.
C01GSX:POP-009:C01In the central scenario, a 22.2 percent population decline by 2076 coincides with 3.5 percent more households. The same run can still produce a shortage of roughly 54,000 habitable homes after the city has lost 266,000 people.Modelled · high sensitivity
C02GSX:POP-009:C02Managed shrinkage in the model keeps school-seat use near 82 percent rather than 47 percent and lowers the fifty-year present value of the modelled system bill from R384.8 billion to R366.2 billion.Modelled · moderate sensitivity
Sources and reproducibility
The PDF and complete Markdown source are both published.
Full manuscript: 5 397 words across 29 indexed headings. Source checksum: e2327ae42da0…. Code recorded: yes. Data recorded: yes. Environment recorded: no. Published files: 3. the published statement does not rely on a recorded random seed A public code repository is not published.
