Stats NZ has a new website.

For new releases go to

As we transition to our new site, you'll still find some Stats NZ information here on this archive site.

  • Share this page to Facebook
  • Share this page to Twitter
  • Share this page to Google+

What is ethnicity?

According to the statistical standard for ethnicity (Statistics NZ, 2005a), ethnicity relates to the ethnic group or groups that people identify with or feel they belong to, and is a measure of cultural affiliation. Ethnicity is self-perceived and a person can belong to more than one ethnic group.

Why is ethnicity information important?

The ability to produce ethnicity from administrative sources is a key consideration when determining the feasibility of a census based on administrative data. Ethnicity is the principal measure of cultural identity in New Zealand, and is used across the Official Statistics System. The collection of ethnicity information in the Census of Population and Dwellings is a legislative requirement under the Statistics Act 1975, and the census is an important source of ethnicity data for small areas and small ethnic groups. The census provides the basis for official population estimates of major ethnic groups, and for population projections for Pacific Island, Māori, and non-Māori populations.

Ethnicity data are widely used with other characteristics of the population to inform resource allocation, policy development, and research. The major uses of census ethnicity information and ethnic group population estimates and projections are:

  • to monitor the changing ethnic diversity of New Zealand’s population at the national, regional, and local levels so that services can be appropriately targeted
  • to provide denominators for calculating rates by ethnicity for topics such as fertility, mortality, morbidity, and crime
  • to derive measures for monitoring the well-being of ethnic groups, particularly in the health sector (eg morbidity rates, immunisation rates)
  • to monitor the demographic, social, and economic progress of ethnic groups
  • to assist in the planning of services directed at the special needs of ethnic groups in areas such as education, housing, health, and social welfare
  • to evaluate the impact of government policies on the economic and social well-being of ethnic groups
  • to assist in the allocation of funds from government agencies to ethnic groups.

What are the challenges of measuring ethnicity?

Official New Zealand standards and classifications are available, and these are used by Statistics NZ surveys and the census, and by some administrative sources. Despite this, collection of ethnicity information is challenging for several reasons.

Ethnic group changes with context

A person may give a different response depending on the context. For example, when filling in a self-administered form a person may respond differently to when asked his/her ethnic group by an interviewer. The social or cultural setting may also affect the ethnicity response reported.

Ethnic group changes over time (ethnic mobility)

The ethnic group or groups that someone identifies with may change over time. Longitudinal surveys and administrative databases must allow for ethnic mobility. Ethnic mobility also affects the integration of different datasets, as the same person may have given different answers in different collections. Rather than using both datasets' responses, the decision on what to use must be made case by case.

Multiple ethnicity

People may identify with more than one ethnic group, so provision to collect multiple ethnic groups for each individual is essential. The statistical standard recommends the collection of up to six ethnic group responses per person. The ability to collect three responses is the minimum requirement to meet the standard.

Previous comparisons

Previous comparisons of ethnicity between census and administrative data have been undertaken by Blakely, Atkinson, and Fawcett (2008) and Tan, Blakely, and Atkinson (2010), who compared the 2001 Census with mortality data from 2001–04 and 2004–06. Using a total responses measure of ethnicity (a count of the number of people in each ethnic group, regardless of multiple response) they found generally close agreement between the census and mortality data. However, they found that fewer people had multiple ethnic groups recorded in mortality records than in census data, resulting in greater sole Māori counts on the mortality data.

Statistics NZ (2005b, 2014c) compared birth and death registrations for children who died before their fifth birthday. Both studies found fewer multiple ethnic responses recorded on death registrations compared with birth registrations.

O’Byrne, Bycroft, and Gibb (2014) provides a summary of early findings about the potential use of administrative data sources for census information. However, this first assessment was based on metadata and intended to be indicative only. No analysis of the data itself was undertaken. This paper builds on the finding from O’Byrne et al that ethnicity was ‘likely’ to be satisfied by administrative data and suggested several potential administrative sources of ethnicity information.

Aims and scope

This paper describes preliminary analysis of ethnicity information from linked administrative sources available in Statistics NZ’s Integrated Data Infrastructure (IDI).

The overall aim is to investigate the potential of producing estimates of ethnic populations from linked administrative sources, using Statistics NZ’s Integrated Data Infrastructure (IDI) as a test environment.

Three research questions guide this work:

  • What ethnicity information is available from linked administrative data?
  • What is the quality of ethnicity information in the IDI?
  • What would be required to improve the potential for producing estimates of ethnic population from linked administrative sources?

This paper provides reference information about the statistical concepts and about administrative data sources relevant to ethnicity, and presents findings from analysis comparing information in the census with administrative sources.

We investigate the potential for administrative data to provide ethnicity information comparable to the current census and the official series of ethnic population estimates, with the latest 2013 Census as the base reference.

We include level 1 of the standard classification of ethnicity and do not extend to more detailed ethnicity breakdowns. Level 1 is commonly used in reporting and for public policy. We limited the administrative sources investigated to those available in the IDI in May 2015.

  • Share this page to Facebook
  • Share this page to Twitter
  • Share this page to Google+
  • Share this page to Facebook
  • Share this page to Twitter
  • Share this page to Google+