What CALD means, and how to measure it with Census data

What CALD means

CALD stands for culturally and linguistically diverse, and in Australian reporting it describes people born overseas, people who use a language other than English, or both. The AIHW notes there is no single way to define a CALD background, and depending on the report it can also cover people with limited English or Aboriginal and Torres Strait Islander people.

The ABS's Standards for Statistics on Cultural and Language Diversity, endorsed in 1999 and republished unchanged in February 2022, recommend that the older label NESB (non-English-speaking background) be discontinued for data and that no other single measure take its place. They set out standard variables instead, so a CALD figure is in practice one or more of those variables reported under a common name.

The ABS variables

The standards' minimum core set has four variables, and the Census collects each of them:

• Country of birth.
• Main language other than English spoken at home, which the 2021 Census asked as the language a person uses at home.
• Proficiency in spoken English, answered by people who use another language at home and based on their own assessment.
• Indigenous status.

The full standard set adds ancestry, each parent's country of birth, first language spoken, languages spoken at home, main language spoken at home, religious affiliation and year of arrival. Agencies also set their own minimums. Under the Queensland Multicultural Policy, Queensland Government agencies must collect country of birth, preferred language and whether an interpreter is needed, and only the first of these is an ABS standard variable.

Why the choice of variable matters

Applying different organisations' definitions to the 2021 Census put between 6.3% and 43% of Australian children aged 0 to 17 in a CALD group, according to Abdul Rahim and colleagues (2024). The gap between variables also differs from place to place, as the two councils below show.

Joondalup, a council in Perth's northern suburbs, is above Australia on birthplace and below it on language because about two-thirds of its overseas-born residents were born in five English-speaking countries (England 14.5% of all residents, South Africa 3.7%, New Zealand 2.6%, Scotland 2.1% and Ireland 1.7%).

Fairfield, a council in Sydney's south-west, is well above Australia on all three variables below, though the variables still differ there. 22.6% of residents use Vietnamese at home and 17.3% were born in Vietnam, so Vietnamese speakers outnumber Vietnam-born residents by about 9,900 people.

Choosing a measure

The variable to use follows from the decision the figure supports, and a few common cases map like this:

• Interpreters and translated material: English proficiency, with the individual language figures to show which languages are needed.
• Outreach in one language: that language's share of people using it at home, which counts people born in Australia as well as overseas.
• Settlement services: year of arrival, such as the share of overseas-born residents who arrived from 2016 to the 2021 Census.
• Second-generation communities: the share with both parents born overseas.
• Cultural and faith-based services: ancestry and religious affiliation, which people report whatever their birthplace or language.

Where a funding body, contract or report specifies its own CALD definition, using the same one keeps figures comparable. The AIHW's aged care reporting, for example, counts people born in countries other than the main English-speaking ones.

Measuring CALD with Censible

Censible doesn't publish a single CALD figure. It has the variables a CALD figure is built from: the share of people who use a language other than English at home and figures for 36 individual languages, English proficiency, the share born overseas and 49 individual countries of birth, both parents born overseas, year of arrival, 30 ancestries and 12 religious affiliation figures.

Each is available for suburbs, councils, postcodes and statistical areas, and for catchments drawn around a site as a circle, a drive or walk time, or a shape, compared with Australia, a capital city or another area. Censible doesn't group countries into the main English-speaking ones, so a measure like the AIHW's has to be approximated from the individual countries it has (England, Scotland, Wales, Ireland, New Zealand, the United States, Canada and South Africa), which leave out Northern Ireland.

Limitations

• Census answers are self-reported, and English proficiency is each person's own assessment of how well they speak it.
• The 2021 Census asked about the language a person uses at home, while 2011 and 2016 asked about the language spoken, so part of any change across those years may come from the new wording.
• Censible's percentages leave out people who didn't answer a question, while ABS QuickStats includes them, so Censible's shares are slightly higher: 39.0% of people who answered had both parents born overseas in 2021, against 36.7% of all people in QuickStats.
• Censible shows the share who use each language and the share with low English proficiency separately, so its figures don't show which languages the people with low English proficiency use.
• Religious affiliation is an optional Census question.
• Figures describe an area, and the ABS standards caution against assuming an individual's needs from the characteristics of a group they belong to.
• The latest figures are from the 2021 Census, carried onto 2026 boundaries, until the first 2026 Census results arrive in June 2027.