Mesh blocks: the housing evidence

Armadale: local housing variety, with mixed infill

Armadale, Perth, WA. Deliberately selected photographed examples, with 2025–2026 listing records and ABS 2021 boundaries. They illustrate housing variety; they are not a random sample or every home in either block.

Both photographed homes have three bedrooms. The established house has one bathroom and 708 m² of listed land; the infill home has two bathrooms and 229 m². The land figure for 112C is its listed lot, not the wider parent parcel. These are individual examples, not different whole-block room medians.

Construction source file dated 19 April 2026. The file date does not date each underlying listing. Eight of the nine house-year labels above have a matching construction phrase in available archived descriptions; 37 Sexty Street is explicitly marked as a source gap below. This pair is not included in the broader 2,777-pair age comparison because block A has fewer than five recorded house years.

132 Fifth Road repeats the established brick-house form. 104 Fifth Road shows an older home remaining in the infill block. A mesh block can contain a recurring housing form and still be mixed.

This local source check uses the wider available dated house listings, with the latest retained listing for each address. It is separate from both the April construction-year source and the national 2023–2026 room-count sample.

Medina: a much clearer change in housing age

The neighbouring Pace Road and Nannup Street blocks separate a 1950s cohort from a 2015 cohort. This example was selected because the reported years and address-registration history make the contrast unusually clear.

A · Pace Road

1953

Median reported year, five houses.
Years: 1953, 1953, 1953, 1955 and 1965.

51 current address records. None first registered from 2015.

B · Nannup Street

2015

Median reported year, ten houses.
One 2014, eight 2015 and one 2016.

77 current address records. All first registered from 2015.

Construction file dated 19 April 2026; address source held in May 2026, with national creation-date coverage through 21 April 2026. The 128 address records and 15 reported-year houses are different populations.

All 15 reported construction years have a matching phrase in an archived listing description. This is stronger evidence of different housing cohorts than address-registration dates alone. Registrations can reflect new addresses, subdivisions or administrative changes, so they are not construction dates. These counts cover current matched address records, not every historical address or every completed dwelling.

Broader evidence: closer house ages within a block

Why the sample counts differ

Bedrooms and bathrooms: the source contains dated listings from 1 January 2023 to 10 February 2026, reduced to distinct address observations. Inferred-date records are excluded, housing labels are grouped, and building-level apartment addresses are excluded where a unit address is also represented. A valid address-to-2021-block match is required. Bedroom and bathroom availability differ. These comparisons include houses, apartments and townhouses and require at least one metre of shared boundary.

Construction years: the much wider source contains 2,118,045 address rows, but only 201,027 have a reported year. After address, year-range and conflicting-year checks, 196,492 reported-year address records remain, including 146,468 records labelled houses with no unit identifier. The adjoining comparison includes 21,828 of those houses, in blocks with at least five recorded years and at least ten metres of shared boundary. No predicted construction year is used.

The construction-year source file is dated 19 April 2026. Its original listing-date range was not preserved in the aggregate. It must not be described as the 2023–2026 room-count sample, a current building inventory or council-certified construction history. Reported years may be approximate, wrong or refer to a rebuilt home.

Coverage is concentrated in Western Australia

Western Australia contributes 13,899 of the 21,828 houses (63.7%) in the adjoining age comparison. The source reaches several states, but coverage and attribute availability are uneven. Equal weighting of suburbs does not turn it into a representative national estimate.

Inspect all 2,777 adjoining house-age pairs →

Armadale and Corio: a photographed stock comparison

Single-storey brick houses with tiled roofs, front gardens and three-bedroom, one-bathroom layouts appear in both Perth and Geelong. These two homes were selected to make that physical resemblance visible.

Both homes have three bedrooms and one bathroom. Their listed land sizes are 708 m² in Armadale and 618 m² in Corio. Neither photographed property has a verified construction year, so this comparison makes no claim that they were built in the same year or decade.

Both surrounding block samples also have medians of three bedrooms and one bathroom. That provides local room-count context; it does not establish that all houses in either block have the same appearance, age or condition.

These are the latest retained source-dated listings for distinct house addresses with no unit identifier. Missing room counts are excluded from that room’s median. This local comparison uses wider historical records than the national 2023–2026 room-count analysis.

Physical resemblance is not price equivalence. The Armadale sale was $655,000 on 2 December 2025; the Corio sale was $500,000 on 25 September 2025. They are different properties, dates and markets. Their resemblance can inform research into building form, maintenance and competing stock, but it does not make them interchangeable valuation comparables or explain the difference in sale prices.

The photographs are selected examples, not a random sample or images of every home in the blocks. The separate Medina–Clarkson comparison matches reported construction years and floor areas; it answers a different housing-similarity question.

Medina and Clarkson: similar physical stock across suburbs

The two selected Perth blocks have the same median reported construction year and floor area: 2015 and 112 m². Their centres are 59.9 km apart in a straight line.

Medina block 50181240000: ten houses with years and ten recorded floor areas. Clarkson block 50391659000: eleven houses with years and eight floor areas. Both local maps span 600 metres. Source: April 2026 reported-year and floor-area file, ABS 2021 and OpenStreetMap. These blocks were selected as a clear physical-stock match.

The wider scan found 812 qualifying cross-suburb matches among 860 eligible blocks. Blocks needed at least eight reported-year houses, five floor areas and a reasonably concentrated construction period. Matched blocks are in different suburbs within the same state, at least three kilometres apart, with median years no more than three years apart and median floor areas within 15%.

Similar building age and size can help investigate maintenance, dwelling form and competing supply. It does not establish equal rents, equal prices, similar residents or interchangeable valuation comparables. Location, land, condition, schools and market date still matter.

All ten Medina year labels have direct construction phrases. Nine of the eleven Clarkson labels do; 19 and 23 Perkins Drive are retained and marked as quotation gaps in the table.

Clyde North: observed population differences at the right geography

The 2021 Census area containing Selandra Rise Retirement Village had a median age of 74. The neighbouring Census area had a median age of 30.

Source: ABS 2021 Census General Community Profile, SA1 DataPack, population table G01 and median/average table G02. The Australian Energy Regulator record identifies Selandra Rise Retirement Village at 5 Harmony Chase, Clyde North, VIC 3978.

The orange block also includes ordinary houses on Nunkeri Court. Its figures describe all usual residents of the 2021 area, not just village residents. The blue figures describe all four constituent blocks, not the highlighted neighbour alone.

At mesh-block level the ABS publishes total usual-resident population and dwelling counts. Detailed observed demographics normally use Statistical Areas Level 1, or SA1s. Only when an SA1 is one mesh block can those figures describe that exact mesh block as well. This example demonstrates historical area differences, not similarity between all individuals or today’s residents.

Microburbs also creates estimated mesh-block demographics. Those estimates can add context when clearly labelled, but they were not used as independent proof that people within a block are alike. Income, age or tenure is not evidence of an individual’s safety or behaviour.

What this changes for property research

Keep estimates beside the actual home

Property, Armadale WA	Estimated value, 11 July 2026
197 Seventh Road 3 bedrooms · 1 bathroom · 708 m² listed land	$632,160
112C Fifth Road 3 bedrooms · 2 bathrooms · 229 m² listed land	$667,924

These are Microburbs modelled estimates, not sale prices. Similar estimated prices can describe different housing. They are shown to explain why a valuation needs the property and local context, not to establish a boundary premium or a measured accuracy improvement.

Why the shape of the block matters

Roads, access, land use and property boundaries can make one group of homes share a setting. These selected strips demonstrate actual shapes, not a measured traffic, noise, safety or price effect. The ABS mesh-block design guidance also addresses communities with shared access, housing complexes and physical barriers.

2026 mesh blocks: a new boundary edition, not new Census results

The ABS released the Edition 4 main structure on 22 July 2026. Mesh blocks form the smallest part of its nested main hierarchy: mesh block → SA1 → SA2 → SA3 → SA4 → state or territory → Australia. ABS approximations of suburbs, postal areas and council areas also use mesh blocks, with their own structure and release dates.

That official standard makes detailed data easier to connect and aggregate. Suburb and postcode names remain useful for orientation. A mesh-block code plus its boundary edition preserves a more explicit geographic reference. It is a statistical geography, not a property title or a legal planning boundary.

The official 2021-to-2026 correspondence contains 43 successor rows for the selected Clyde North 2021 block 21327400000. This is a boundary correspondence count, not 43 new developments or a count of people. The imagery dates, boundary editions and observation periods are separate facts.

All new adjoining housing and construction-year comparisons in this note use 2021 boundaries. The Clyde North satellite example separately compares 2021 and 2026 geography. An unchanged code alone does not guarantee an unchanged footprint. The ABS release plan schedules the first 2026 Census results for June 2027.

Microburbs API access

Microburbs holds the 2026 boundary data. The public mesh-block polygon documentation currently describes 2021 boundaries, and the property context documentation describes local block information. Holding the 2026 data does not establish a documented live 2026 endpoint.

Discuss the boundary edition and data access you need →

API documentation and research checked 9 September 2026. No 2026 Census demographic results or undocumented API response are announced here.

What the evidence supports

Mesh blocks provide a common official way to preserve local detail. The observed house-year sample is more similar within blocks on average, while selected maps show substantial diversity inside one suburb and similar physical stock across different suburbs.

The research does not establish universal homogeneity, a causal effect at the boundary, improved valuation accuracy or person-level similarity. Listing coverage is incomplete and uneven. Source dates, property condition and the exact location still belong in a sound property comparison.

Return to the main article →