Environmental Risk and Capital Growth: A National Study
Abstract
This whitepaper documents the construction of a complete environmental risk profile for every Australian residential address, and quantifies how each risk type is associated with observed sale price, rent and capital growth. Across 15.2 million addresses, we tested exposure to ten hazard categories: contaminated land, mine subsidence districts, aircraft noise contours, flood planning overlays, bushfire prone land, acid sulfate soils, heritage listings, historical flood records, landslide risk and erosion zones. Two layers are associated with substantial capital growth shortfalls after restricting comparisons to properties in the same postcode: contamination proximity (high tier, 0-250 metres, ~$143,000 gap on a $500,000 home over eight years) and surface acid sulfate soils (~$164,000 gap). A further three layers - flood planning overlays, aircraft noise and the bushfire/mine-subsidence pairing - produce sale-price discounts of 13% to 27% but no separable capital growth signal once same-postcode comparisons are applied. These results are observational, not experimental. They do not constitute proof of causation. But the gap shrinks smoothly with distance, and it survives location matching. Both patterns are hard to explain by selection alone. We make the case that environmental risk data, properly tiered and properly tested, is the missing layer in property due diligence.
Interpretation note. Capital growth findings in this paper are observational estimates obtained by comparing properties within the same postcode. They are not derived from a randomised or quasi-experimental design and do not constitute formal causal proof. Sale price and rent comparisons are simpler simple median comparisons between properties with and without the risk flag, and have not been location-matched in the same way. Readers should treat these as associations that may include location bias and other unmeasured differences.
Contents
Key Findings
- Contamination has three tiers, not one cutoff. 1.57 million addresses (10.3% of the national total) sit within 250 metres of a registered contaminated site, the high tier where the capital growth penalty is largest. A further 2.12 million sit in the medium tier (250-500 metres) and 3.38 million in the low tier (500-1000 metres). Roughly 46% of all Australian addresses fall within a kilometre of at least one registered site.
- Flood planning overlays sit at a 27% sale-price discount. 617,000 addresses fall within state planning overlays, concentrated in tropical North Queensland where 99% of Cairns City (Cairns, QLD) sits inside a floodplain assessment area.
- Aircraft noise produces a symmetric 13% discount on price and rent. Properties under high noise contours from major airports trigger acoustic treatment requirements and resale price discounts that tenants and buyers value at almost the same rate.
- Surface acid sulfate soils pose excavation risk to only 1.3% of addresses. 204,000 properties sit on land classified as high probability of acid sulfate within half a metre of the surface. A further 234,000 face risk between 0.5 and 3 metres, affecting basements and footings. The remaining classifications are informational only.
- Mine subsidence districts contain 203,000 addresses. Charlestown (Newcastle, NSW) is 100% within the Newcastle district. The risk is fully reflected in sale prices and produces no further capital growth gap once the discount is applied at purchase.
Methodology
We assembled environmental risk data from publicly available government registers covering all Australian states and territories. Each layer was matched to every residential address in Australia using standard spatial techniques. For point-based hazards such as contaminated sites, we measured proximity at multiple distances (250 metres, 500 metres, 1000 metres) and recorded the distance to the nearest site. For polygon-based hazards such as flood overlays and bushfire prone land, we recorded whether each address falls inside the affected area. The result is a complete environmental profile for every Australian residential property.
Validation strategy
The central methodological challenge in any cross-sectional analysis of this kind is selection. Risk-affected properties cluster in particular kinds of suburbs - older industrial areas for contamination, coastal estuarine areas for acid sulfate soils, regional bushland for fire risk. These suburbs differ from clean suburbs on many dimensions other than the risk being studied. Naive comparisons of “with risk versus without” therefore confound the risk effect with location effects.
We address this with a same-postcode matched comparison for the capital growth analysis. For each affected property, we restrict the comparison set to other properties in the same postcode that do not carry the risk flag, and only include postcodes with at least 50 affected properties so the comparison is well-powered. Same-postcode matching is a partial proxy for school catchment, transport access, broad demographic composition, and the dominant regional growth differences that would otherwise drive the result. It is not a substitute for randomisation and it does not control for property-level differences that may exist within a postcode.
Sale price and rent results in section 5.1 and 5.2 are simpler unmatched comparisons of medians between properties with and without each risk flag. They are presented as observed associations and have not been location-controlled in the same way as the capital growth analysis. We discuss the limits of each comparison alongside the relevant table.
Data window
Property sales analysed in this study cover transactions from 2015 to 2026, with current valuations as the comparison endpoint. The rental records are listings observed between 2020 and 2026. Risk register snapshots were taken in April 2026.
Results: National coverage
| Risk layer | Addresses affected | % of national total | Geographic spread |
|---|---|---|---|
| Contamination - high tier (0-250m) | 1,565,513 | 10.3% | 5 states |
| Contamination - medium tier (250-500m) | 2,117,157 | 13.9% | 5 states |
| Contamination - low tier (500-1000m) | 3,381,303 | 22.2% | 5 states |
| Bushfire prone land | 789,794 | 5.2% | NSW (others to follow) |
| Flood planning overlay | 616,648 | 4.0% | NSW + QLD |
| Acid sulfate - surface (excavation risk) | 203,994 | 1.3% | NSW + QLD coastal |
| Mine subsidence district | 203,357 | 1.3% | NSW (Newcastle, Hunter, Illawarra) |
| Aircraft noise (high contours) | 90,950 | 0.6% | National |
| Heritage listing | 76,531 items | varies | NSW + VIC |
Two suburbs to illustrate
Alexandria (Sydney, NSW): Of 9,725 residential addresses, 9,097 (93%) sit within 500 metres of a registered contaminated site. The closest sites to most apartments include a former service station on Botany Road, a former Cadbury Schweppes factory, and the Alexandra Canal sediments. Some are remediated, some are still under environmental management. The City of Sydney local government area contains 2,795 heritage listings, several of which sit within Alexandria.
Charlestown (Newcastle, NSW): Of 8,005 residential addresses, 100% sit within the Newcastle Mine Subsidence District. 2,199 addresses (27%) are mapped as bushfire prone, primarily on the southern and western fringes adjacent to bushland. 46% of addresses fall within 500 metres of a contaminated site, with average distance 284 metres.
How each risk relates to price, rent and capital growth
Joining the address-level risk profiles to 31.3 million property sale records and 10.25 million rental records lets us answer the question every investor asks: does the risk actually show up in the price? Yes, in some cases dramatically.
5.1 Sale price discounts at the time of sale
| Risk layer | Median price (with risk) | Median price (without) | Differential |
|---|---|---|---|
| Flood planning overlay | $317,000 | $435,000 | −27.1% |
| Aircraft noise (high contours) | $372,500 | $430,000 | −13.4% |
Flood overlay carries the largest visible price discount at 27%. Aircraft noise comes next at 13%. These are simple unmatched comparisons of median sale prices, not location-matched estimates. They are presented as cleaner candidates than other layers because flood and aircraft noise overlays cut across both expensive and cheap suburbs, so the discount is less likely to be driven by where the affected properties happen to sit. A future release will apply the same-postcode matching used for capital growth to confirm the magnitude.
5.2 Weekly rent discounts
| Risk layer | Median rent (with) | Median rent (without) | Differential |
|---|---|---|---|
| Flood planning overlay | $340/week | $400/week | −15.0% |
| Aircraft noise (high contours) | $350/week | $400/week | −12.5% |
Tenants discount aircraft noise at almost the same rate as buyers (12.5% vs 13.4%) - the daily annoyance is the same regardless of who pays. But tenants discount flood overlays at half the rate buyers do (15% vs 27%), suggesting buyers are pricing in long-term capital risk that tenants do not feel.
5.3 Capital growth: contamination has three tiers
The capital growth analysis is the part of this study that took the most care, because it is the part most vulnerable to confounding. The single “within 500 metres” flag we started with proved insufficiently granular. When we split contamination proximity into distance bands, the capital growth shortfall showed clear monotonic decay rather than a sharp cutoff. Properties closer to contamination underperformed by larger amounts. the gap narrowed with distance but did not disappear at 500 metres. We now report contamination in three tiers, comparing each affected property only to non-affected properties in the same postcode.
| Tier | Distance to nearest registered site | Australian addresses | Annual growth penalty (matched) | Gap on $500K over 8 years |
|---|---|---|---|---|
| HIGH | 0-250 metres | 1.57 million (10.3%) | −1.6 percentage points | ~$143,000 |
| MEDIUM | 250-500 metres | 2.12 million (13.9%) | −1.5 percentage points | ~$133,000 |
| LOW | 500-1000 metres | 3.38 million (22.2%) | −0.9 percentage points | ~$79,000 |
Roughly 46% of all Australian residential addresses fall into one of the three contamination tiers - within a kilometre of at least one registered site. The high tier alone, 1.57 million properties, carries the largest penalty at $143,000 over an eight-year hold versus a same-postcode neighbour with no contamination nearby. Even the low tier shows a measurable $79,000 gap, which means the effect of contamination on long-run value is detectable out to a kilometre - well beyond the original 500-metre default.
5.4 Surface acid sulfate soils have an even larger effect
The 204,000 properties on high-probability surface acid sulfate land grew at 1.78% per year compared to 3.61% per year for non-acid-sulfate neighbours in the same postcodes. That is an 1.8 percentage point gap, or roughly $164,000 over eight years on a $500,000 property. Buyers initially overpay (these properties tend to be coastal) and then fail to grow at the pace of their neighbours when development costs eventually bite.
5.5 Three layers reverse after controlling for postcode
Three other layers - bushfire prone land, aircraft noise and mine subsidence - showed apparent capital growth premiums in the raw data. These reverse to either zero or modest negatives once you control for postcode. The “premium” was a regional effect: bushfire-prone and noise-affected properties cluster in regional areas that have been growing from a low base since 2015. Mine subsidence shows essentially no matched-postcode effect. This is consistent with the risk being fully priced in at purchase. Buyers know the district status, discount the entry price accordingly, and the property then grows at the same rate as the rest of the postcode.
5.6 Holding periods
Properties near contamination are held longer than otherwise comparable properties (103 months between sales vs 94 months without contamination, +9.6%). This is consistent with the capital growth penalty: owners who would otherwise sell instead hold, waiting for a recovery that does not arrive. Bushfire prone properties turn over slightly faster (90 months vs 95, −5.3%), which may reflect distress sales in years following major fire seasons.
Defence Against Criticism
“Not all contamination is dangerous”
Correct. Some sites in the registers are active and under regulatory order. others are remediated and closed. The proximity flag indicates that a registered site exists nearby, not that the property itself is contaminated. We surface this distinction in property reports by labelling each nearby site with its current regulatory status. The investor decision is informed, not pre-determined.
“Acid sulfate soil coverage is too broad to be useful”
The continental soil grid that backs this layer classifies most of coastal Australia, but classification is not the same as risk. Of 14.8 million classified addresses, 8.3 million (54%) sit on land where the probability of acid sulfate at the surface is rated extremely low (depth greater than three metres only). A further 5.0 million (33%) sit on land rated low probability. We treat both as informational and surface them as “Clear” in property reports. The genuine risk is concentrated in the 204,000 addresses (1.3% of national) classified as high probability at the surface, where excavation will hit acid sulfate within half a metre and trigger mandatory remediation. The remaining classifications are informational and do not appear as flags.
“This duplicates what a planning certificate already shows”
A planning certificate is purchased per property at conveyancing, after an offer has been made. By that point the buyer is committed. The value of address-level pre-computed risk data is that it enables risk screening at the search stage, before an offer is made, across thousands of candidate properties. It complements rather than replaces the formal planning certificate.
“How do you know it is causal, not selection?”
This is the central methodological challenge and we treat it directly. Three pieces of evidence support a causal interpretation. First, the same-postcode matching restricts each comparison to properties in the same neighbourhood, controlling for school catchment, transport, demographics and the dominant regional growth differences. Second, the contamination effect shows clear monotonic distance decay: the penalty at 0-100 metres is the largest, at 100-250 metres slightly smaller, and so on out to a kilometre. A spurious correlation with location would not produce this dose-response shape. Third, two of the layers we tested (bushfire and aircraft noise) reverse to no effect after matching, which means the matching procedure does not mechanically produce penalties. Layers that survive matching genuinely behave differently from layers that do not.
Limitations
Five limitations warrant acknowledgement. First, the registers are state-administered and update at different cadences. A site remediated last week may still appear as active in a snapshot taken last month. Second, contamination data covers five states (NSW, VIC, QLD, SA, WA) representing approximately 95% of the national property market. Bushfire prone land coverage is currently strongest for NSW. coverage for Victoria, Queensland and Western Australia is being expanded in subsequent releases. The “ten hazard categories” tested in this paper do not all have equal national coverage, and headline figures should be read with the geographic spread column in section 4. Third, the matched-postcode design reduces but does not eliminate selection. There may be unmeasured property-level differences inside a postcode that correlate with both contamination proximity and slow growth. Fourth, sale price and rent comparisons in section 5.1 and 5.2 are unmatched simple median comparisons, not location-controlled estimates. they should be read as observed associations. Fifth, the strongest defence against the residual selection concern in the capital growth analysis is the way the gap shrinks smoothly with distance, which is unlikely to arise from any plausible omitted-variable story but is not in itself a substitute for an experimental design.
Conclusion
Environmental risk profiles are achievable at national scale and at address-level granularity, and they show measurable associations with both sale price and long-run capital growth. The two most important findings are simple. Contamination proximity is associated with a dose-dependent capital growth shortfall extending out to roughly a kilometre. Surface acid sulfate soils are associated with a similar shortfall on the small number of properties where excavation costs are likely to apply. Both findings persist after same-postcode matching, both correspond to six-figure dollar gaps over a typical hold period, and both can be screened at the search stage rather than discovered at conveyancing. The cost of overlooking a contamination history is paid in remediation expenses, insurance loadings, development approval delays, and the resale discount that future buyers will eventually demand. Identifying these risks earlier moves the cost from buyer to seller, which is where it belongs.