• Summary
• Downloads
• Explanatory Notes
• Related Information
• Past Releases

Understanding migrant statistics in the Census and other data sources


Census

The Census of Population and Housing (Census) collects a range of information about the cultural and linguistic diversity of the Australian population such as country of birth, country of birth of mother/father, year of arrival, main language other than English spoken at home, proficiency in spoken English, Australian citizenship, ancestry, and religious affiliation. It also provides a range of information on education, work (e.g. labour force status, occupation and industry), income, mobility, and characteristics of families and households.

The Census is therefore a leading source of information about migrants and allows for detailed analysis across a broad range of socioeconomic dimensions. However, a number of other data sources are available and therefore this page provides an overview of strengths and key uses of each collection.


Estimated Resident Population (ERP) by Country of Birth

Information on the country of birth of Australian residents is available annually from the Estimated Resident Population (ERP), and it is therefore another key data source for information on migrants. The ERP is the official ABS estimate of the Australian population. Among its many uses are the determination of the number of representatives from each state and territory to sit in the House of Representatives, and the annual allocation of Commonwealth funds for state governments and local government.

ERP by country of birth is mainly based on the number of people counted in each country of birth from the Census of Population and Housing, but also takes a number of other factors into account.

As the Census generally relies on people self-reporting their country of birth, on occasions people fail to give sufficient detail to allow their response to be allocated to a particular country. For example, a person may report their country of birth as "United Kingdom" or "Africa". As part of producing ERP, these responses are redistributed across the countries that make up that region. Persons counted in the Census that do not report a country of birth at all are similarly redistributed.

While the Census makes every effort to count each person in Australia once (and only once), inevitably some people are missed and some are counted multiple times. Generally, more people are missed than counted more than once, and this therefore leads to a net undercount. The net undercount for each country of birth is estimated through the Post Enumeration Survey (PES), a large survey taken soon after the Census.

In addition, a number of people who usually live in Australia are out of the country on Census Night and therefore not in scope of the Census. The number and country of birth of Australian residents temporarily overseas is determined mainly from information collected and compiled by the Department of Home Affairs.
Further adjustments are made for births, deaths and net migration to account for the period between June 30 and Census Night. As a result of these adjustments, ERP is generally higher than the Census count. For example, the 2016 ERP estimate of the number of persons born in China was 526,040, while the number of persons born in China according to the unadjusted Census count was 509,555.

The Estimated Resident Population for each country of birth is updated annually as at June 30 each year, based on births, deaths, and overseas migration. The ERP therefore represents the most accurate and timely statistics available on the population for each country of birth (and for summary statistics such as the proportion of Australians who were born overseas).

However, only an extremely limited number of characteristics of the population by country of birth are available through Estimated Resident Population: age, sex, and (in Census years only) state or territory of residence. For information on other characteristics, such as main language other than English spoken at home, the Census usually represents the best data source.

Estimated Resident Population by Country of Birth is currently published in Migration, Australia (cat. no. 3412.0), with a brief summary table included in Australian Demographic Statistics (cat. no. 3101.0).

Summary of differences between the Census and Estimated Resident Population by Country of Birth

	Census		Estimated Resident Population by Country of Birth
Purpose	Counts the number of people in Australia on Census Night and the dwellings in which they live. Collects information about a range of characteristics of people, including their country of birth, main language spoken at home, ancestry, and religion. Provides a rich snapshot as at Census Night. It is the leading source of information for the characteristics of small population groups and areas, and allows for the analysis of cultural and linguistic diversity at a highly detailed level.		Estimated Resident Population (ERP) is the official ABS estimate of the Australian population. ERP by Country of Birth provides the definitive source for the number of Australian residents born in each country, and their age, sex, and (in Census years only) state or territory of residence.
Frequency	Conducted every five years. Provides a single snapshot of Australia on Census night.		Updated annually. Rebased every five years following the Census. Provides an annual time series of data on the country of birth of Australian residents for all countries since 1996, and for selected countries since 1992. Other data may be available on request. Information by state or territory currently only available in Census years as there is insufficient data on interstate migration by country of birth.
Reference period	Census Night		June 30 of each year
Methodology	Generally a self-completed online or paper form.		Based on Census counts, with adjustments made for vague or missing responses, persons missed in the Census, and residents temporarily overseas.
Scope	Persons usually resident in Australia and present in Australia on Census Night.		Persons usually resident in Australia.
Sources of error	Not subject to sampling error as it aims to collect the information from everyone in Australia on Census night. However the Census is subject to undercount. Subject to vague or missing responses.		Not subject to missing data as vague or missing country of birth responses are redistributed to individual countries, but these methods may introduce a small amount of error. Accounts for persons missed by the Census. However, the adjustment for Census undercount is based on a sample survey (the Post Enumeration Survey) and is therefore subject to sampling error. The scale of this error is very minor relative to the size of the estimate. Any inaccuracies in the components used to update ERP (births, deaths, and net overseas migration) will be present in the estimate and will compound over time. However, the error in these components is small.