The Heritage Premium Regime Shift

1. Background

Heritage protection in Australia operates through a patchwork of state-level registers and local planning overlays. In New South Wales, local environmental plans schedule heritage items and conservation areas. In Victoria, the Heritage Overlay embedded in the planning scheme restricts alterations to listed places. Queensland maintains a state register through the Department of Environment, Science and Innovation. Western Australia publishes the inHerit register via the Department of Planning, Lands and Heritage.

The investor question is simple. Does buying a heritage-listed house help or hurt capital growth? The research literature has answered both ways, with older studies finding a modest positive effect and more recent work finding the premium shrinking or turning negative. None of the published work covers the post-pandemic cycle in any detail.

2. Data

Our primary dataset is our Microburbs paired dataset of completed property holds. Each record represents one completed buy and sell of the same property by the same owner, with buy and sell prices, buy and sell dates, and the physical location of the property. We rebuilt this table through 13 April 2026 to capture the most recent transactions. After standard cleaning the file contains roughly 10 million completed holds across Australia, of which 1.43 million ended between January 2020 and April 2026 in the three states where we have statutory heritage polygons loaded into our geographic infrastructure: New South Wales, Victoria and Queensland.

Heritage status was assigned by locating each property on the map and testing whether its point falls inside any of five statutory polygon layers: the New South Wales environmental plan heritage layer, the Victorian Heritage Overlay in the planning scheme, the Victorian Heritage Inventory, the Victorian Heritage Register, and the Queensland Heritage Place Register. Around 4 per cent of the qualifying holds in these three states fall inside at least one of these layers.

Rent and yield figures are sourced from our standard suburb median tables at the most recent snapshot, 12 April 2026.

3. Scope and limitations

The analysis is restricted to New South Wales, Victoria and Queensland because these are the only three states where we have statutory heritage polygons loaded for spatial matching. Western Australian and South Australian heritage registers exist but have not been loaded. The Queensland result is based only on the state register: local council heritage overlays, which contain the majority of statutory heritage in Queensland, are not loaded. The Queensland numbers should be read as indicative.

The heritage flag is static. A property that was added to a heritage overlay in 2024 is treated as heritage for its 2015 sale. This is a small but systematic bias in favour of apparent pre-2022 heritage outperformance.

The within-suburb pair matching excludes fully heritage-protected suburbs entirely. There is no non-heritage control group in Paddington (Sydney), Balmain (Sydney) or Carlton North (Melbourne), so these suburbs drop out of the pair-matched results by construction. They appear only in the state-level aggregates and in the time series.

The 2026 figure covers only January through 13 April 2026 and is composition sensitive. All references to 2026 in this paper should be read as partial year.

4. Findings

4.1 The regime shift

The central finding is a decisive change in direction beginning in 2022. Heritage-protected properties were growing faster than non-heritage properties in all seven years from 2015 to 2021. The heritage advantage ranged from around 11 per cent relative outperformance in 2021 to around 33 per cent relative outperformance in 2015 and 2016. In 2022 the direction reversed. By 2024 the heritage growth rate was running at roughly 36 per cent below the non-heritage rate. By 2025 the gap widened to about 41 per cent slower and the partial-year 2026 figure shows heritage growth at roughly 45 per cent slower than non-heritage growth.

Figures compare the mean annualised capital growth of heritage and non-heritage holds completed in each year, across New South Wales, Victoria and Queensland. Relative outperformance is computed as the heritage rate minus the non-heritage rate, expressed as a share of the non-heritage rate.

4.2 State-level averages

For the 2020 to 2026 window as a whole, heritage properties grew roughly 11 per cent slower than non-heritage in New South Wales, around 21 per cent slower in Victoria, and around 20 per cent slower in Queensland. The Queensland figure reflects only 743 heritage holds drawn from the state register and is indicative rather than decisive. The New South Wales and Victorian figures rest on 34,271 and 24,568 heritage holds respectively.

4.3 Within-suburb price premium (size-matched)

The price premium is our cleanest finding because the comparison is harder and the confounders are easier to control. We form matched cells inside each suburb, where a cell is a group of sales of the same property type with the same number of bedrooms and the same number of bathrooms. Within each cell, we compute the average heritage price and the average non-heritage price, and we express the difference as a percentage of the non-heritage average. We then take the median of those per-cell percentage differences across all qualifying cells. Matching on suburb controls for location. Matching on bedroom and bathroom counts controls for two of the largest drivers of price variation within a suburb. Land area, building area, parking, condition and period quality are not controlled, so the residual premium should be read as reasonable evidence of a heritage effect rather than a clean causal estimate.

Across 728 matched New South Wales house cells covering 9,089 heritage sales and 34,377 non-heritage sales from 2020 to 2026, the median heritage premium was 3.3 per cent, or approximately $56,000 on a typical $2.05 million non-heritage base. Across 699 matched Victorian house cells covering 10,403 heritage sales and 51,881 non-heritage sales, the median heritage premium was 5.3 per cent, or approximately $116,000 on a typical $3.57 million non-heritage base. Units show smaller premiums in both states, with a slight negative in New South Wales units.

The mean premium is larger than the median in every cut of the data. Mean premiums for the same matched house comparisons are 20.7 per cent in New South Wales and 28.9 per cent in Victoria. The divergence between mean and median reflects the skew of the heritage sale distribution: a small number of heritage houses command extreme premiums (unique architectural pieces, large blocks with protected landscaping) that pull the mean up. The median is the honest description of what the typical buyer experiences.

Time series of the matched price premium

The stable pre-2023 premium was roughly 7 to 10 per cent in New South Wales and 8 to 12 per cent in Victoria. From 2023 onwards the premium has become unstable and lower on balance. In New South Wales the 2023 median was 3.3 per cent, the 2024 median was around minus 4 per cent, and the 2025 median bounced up to around 20 per cent on a much smaller cell count. In Victoria the 2023 median was around minus 7 per cent, the 2024 median was around 6 per cent, and the 2025 median was around minus 3 per cent. The pre-2023 band of 7 to 12 per cent has not held in any post-2022 year. Year-on-year composition effects are larger than in the pre-2023 window because cell counts fall as heritage sale volume drops.

Matched cells require at least two heritage and two non-heritage sales in the same suburb, property type (house), bedroom count and bathroom count. Median per cell, then averaged across cells. 2026 is partial and omitted because the cell count is below ten.

The practical reading is that the price premium and the capital growth rate moved together. From 2015 to 2021 both indicators favoured heritage. From 2023 onwards both indicators have moved against heritage. The price finding is more conclusive than the growth finding because the matched comparison controls for location and size directly, reducing the inner-ring location confound that dominates the state-level growth story.

Earlier drafts of this paper cited unmatched price differentials of 10 per cent for New South Wales houses and 26 per cent for Victorian houses. Those figures were arithmetic averages of within-suburb means and included size differences between the heritage and non-heritage comparison groups. The bedroom and bathroom matched figures presented here are the preferred estimates.

4.4 Rent and yield

At the suburb level, areas with high heritage concentrations command higher weekly rents in absolute dollars but systematically lower gross rental yields. Suburbs with no statutory heritage exposure show an average gross yield of about 4.3 per cent. Suburbs where more than half of the house stock is heritage-affected show an average gross yield of about 2.9 per cent. After adjusting for the overall price level of the suburb, a small residual yield gap of roughly one-third of a per cent remains between high-heritage and low-heritage suburbs in the same price bracket. This residual is descriptive. It may reflect the older average age of heritage stock and the higher maintenance burden that comes with restoration-compatible upkeep, as well as pure heritage effects.

4.5 Illustrative suburbs

To make the pattern concrete we highlight two illustrative cases from the high-exposure end of the distribution. Both are approximately fully heritage-protected and drop out of the within-suburb pair matching for that reason. Neither example proves the finding on its own. The validity of the result rests on the 1.43 million holds behind the aggregate numbers, not on these two suburbs.

Paddington, Sydney is the archetypal New South Wales terrace heritage suburb. Every one of the 1,671 house holds we observed in Paddington since 2015 falls inside at least one heritage layer. The current median house price sits around 3.73 million dollars with a gross rental yield of about 2.3 per cent. Paddington was one of the strongest-performing Sydney suburbs through the 2015 to 2021 window and has lagged the wider Sydney market through the 2022 to 2026 cycle.

Carlton North, Melbourne is the Victorian equivalent. All 951 house holds we observed in Carlton North since 2015 are inside the Victorian Heritage Overlay. The current median house price is around 1.51 million dollars with a gross rental yield of about 2.5 per cent. The same pattern is visible: strong pre-2022 growth, conspicuous underperformance since.

5. Discussion

We do not claim that heritage protection causes slower growth. Three alternative explanations are plausible and are not distinguishable in this dataset alone.

First, heritage clusters are concentrated in inner-ring locations, which have lagged outer-ring locations in the post-pandemic cycle. What we label as a heritage effect may simply be an inner-ring effect.

Second, heritage stock is older on average. Older stock requires more maintenance and offers less scope for renovation-driven capital gain. What we label as a heritage effect may simply be an age-of-stock effect.

Third, the 2022 inflection in direction coincides with the onset of the interest rate rising cycle. Buyers in heritage-dense markets tend to be more leveraged than buyers in outer-suburb markets, and the interest rate sensitivity of capital gain is typically higher in leveraged markets. What we label as a heritage effect may partly be a rates-transmission effect.

These three candidate explanations are not mutually exclusive. Statistical controls for state, sell year, distance to central business district, bedrooms, land size and year of construction would be required to separate them. Until such work is done the result in this paper should be read as an observed pattern with an honest framing, not as a causal statement about heritage protection.

6. Implications for buyers

Investors should stop assuming that heritage protection adds to capital growth. That assumption was correct for the decade to 2021. It has been wrong for every year since 2022 and the gap is widening. Buyers who are considering a heritage purchase for growth reasons may need to wait for the outer-suburb cycle to cool, or accept that their growth is likely to come from elsewhere in the cycle.

Investors buying heritage for lifestyle, character or long-horizon capital preservation reasons have a different calculus. The matched price premium is positive on average and is larger than zero in most of the 2015 to 2022 window, but since 2023 it has been unstable and lower on balance. The typical matched heritage premium now sits in a single-digit range that bounces around zero. Buyers should no longer assume they are building on an expanding price floor.

Buyers agents running growth mandates should weight heritage status against their buyer’s time horizon. For clients buying for growth over the next three to five years, the current regime suggests avoiding heritage-dense suburbs. For clients buying for ten years or more, the regime may reverse again if macro conditions swing back.

7. Conclusion

Heritage-protected residential property in New South Wales and Victoria delivered roughly 20 to 33 per cent faster capital growth than non-heritage property through the 2015 to 2021 window. From 2022 onwards the direction has reversed and heritage has lagged by 18 to 45 per cent. The size-matched price premium has also compressed: from a stable 7 to 12 per cent through 2015 to 2022, it has become unstable and lower on balance since 2023, with several years showing near-zero or slightly negative median premiums in both states. Capital growth and price premium moved together.

The regime shift is visible in every year from 2022 onwards and in each of New South Wales, Victoria and Queensland independently. What it does not establish is causality. The inner-ring location correlation, the older stock correlation and the interest rate cycle all offer plausible alternative explanations. Buyers should treat heritage status as one signal among many, not as a standalone growth lever.

Microburbs Research. Sydney, Australia. April 2026. This paper is a descriptive analysis of observed capital growth patterns. It is not investment advice. Past performance is not a reliable indicator of future performance. All figures are derived from completed property transactions recorded in our property-holds database through 13 April 2026, joined to statutory heritage polygons published by New South Wales, Victoria and Queensland government planning authorities.