Salisbury — LGA Conditions Intelligence
Internal Migration Reversing, Overseas-Offset Growth (provisional)Salisbury recorded negative net internal migration in every one of the four years to 2024-25, and the losses have been widening, not narrowing: -1,468, -1,762, -1,891, -2,031. That makes it the deepest and most consistently worsening internal-migration decline of any LGA built in this series to date — a genuinely different shape from Launceston’s roughly flat negative run. Yet Salisbury’s population is still growing (151,019 in 2025, up from 147,521 in 2021), because net overseas migration (+7,938 over four years) comfortably exceeds the internal loss (-7,152) — the same overseas-offset mechanism already seen on Darwin, at outer-metropolitan Adelaide scale. Salisbury sits directly adjacent to Playford, this series’ first South Australian card and a strongly positive absorber over the same period — two neighbouring northern-Adelaide LGAs pulling in opposite directions.
Salisbury’s four growth phases decelerate in strict sequence — Boom +13.2%, Stall +8.9%, Recovery +5.1%, Current +2.4% — each phase slower than the one before it, with no inversion or rebound anywhere. No other LGA built in this series shows this clean a monotonic slowdown; every other card’s trajectory has at least one phase that breaks the pattern (an inversion, a rebound, or a reversal). ERP has still grown every phase since 2001 (113,849 → 151,019, +32.7% overall), just at a steadily diminishing rate.
Combined population change in 2024-25 (-2,031 internal + 603 natural + 1,810 overseas = 382 persons, ≈159 implied households at 2.4 persons/household) is far exceeded by the 707 dwelling approvals recorded the same year — a substantial supply surplus relative to modest population growth, though Salisbury’s population is still growing, unlike Launceston where a similar approvals-ahead-of-demand pattern coincided with outright decline. The population-components reconciliation checks exact: -7,152 (internal) + 2,712 (natural increase) + 7,938 (overseas) = 3,498, matching the ERP change (147,521 → 151,019, 2021 → 2025) to the last person.
The Census unoccupied dwelling rate counts dwellings recorded as unoccupied on Census night — it is not a market vacancy rate; at 5.0% it is among the lowest recorded in this series. Salisbury’s 4.4% high-income share and SEIFA IRSD/IRSAD decile of 2/2 are both new series lows (previous lows were Launceston’s 7.0% and 3/3) — the most disadvantaged demographic profile built in this series so far, directly adjoining Playford’s younger, less-disadvantaged profile.
- Bedrock figure is population-weighted, not a simple average of suburbs. Readings are on APN Bedrock™ methodology v1.1. Salisbury’s population-eligible localities (141,815 residents across 25 localities) produce a population-weighted mean of 33.4 on node
24110. A simple, unweighted average gives 33.6 instead — a negligible 0.2-point gap, the smallest found in this series. At 0.67σ below the national population-weighted mean, Salisbury sits in the lower half of this series; Fraser Coast, at 1.15σ below, is the lowest. This card uses the population-weighted figure, consistent with the platform’s existing LGA-rollup convention. - Classification badge is indicative. “Internal Migration Reversing, Overseas-Offset Growth” reuses the same badge language as Darwin because the underlying mechanism is genuinely the same (negative internal migration more than offset by overseas migration, net population growth) — not a forced relabel. The platform’s formal 4-way classification query has not been ratified.
- No crime/safety data included in this version. APN Sentinel™ crime-pillar data (node 24121/24122) exists at suburb grain nationally but carries open, jurisdiction-specific data-quality caveats not yet resolved platform-wide — deliberately excluded from this card rather than presented with unresolved caveats attached.
- Demographic profile is 2021 Census vintage. Given Salisbury’s continued (modest) population growth since 2021, the 2021 figures likely understate current population and household counts slightly.
- Card format approved 19 August 2026; the Bedrock population-weighting convention used above was adopted platform-wide in October 2026.
