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+

Religious affiliation is the self–identified association of a person with a religion, denomination or sub–denominational religious group.

The following supporting concepts are defined in Glossary and references:

  • religion
  • denomination

Operational issues

Some respondents do not wish to report their religious affiliation. All religious affiliation questions contain an option allowing people to object to answering the question. This needs to be taken into account when interpreting religious affiliation data. In addition, facetious responses can affect the usefulness of the data obtained for this variable.

Explanatory notes

Limitations of the variable

It is important to understand that the religious affiliation variable is not a measure of people’s commitment, attendance or participation in religious organisations. For example, a person who has never formally attended a church service may report a Christian religious affiliation because their parents or grandparents were identified with that group.

Changes since the 1996 review

The classification has remained a hierarchical classification with three levels. However, the lowest level of the classification has been changed from a four–digit code structure to a five–digit code structure. A new level one grouping has been added, called ‘Māori Christian’. Previously, this group was at level two of the classification. The residual categories have been amended in accordance with standard residual policy.

Lack of mutual exclusivity between some groups

The classification suffers from lack of mutual exclusivity in some areas. There are a variety of reasons for this. Firstly, because of the complexity of the concepts being measured. Secondly, because of the ambiguity of some responses. Thirdly, because of the way the data has been collected.


The table below illustrates how three people belonging to the same church may be coded to three different categories in the classification.

What happens when three members of the Korean Agape Presbyterian Church respond to the religious affiliation question three different ways?

If this person’s response is to... Then this person’s religious affiliation is coded to…
mark the ‘Presbyterian’ box Presbyterian not further defined (nfd)
write ‘Korean Agape Presbyterian Church’ in the space provided Presbyterian not elsewhere classified (nec)
mark the ‘Christian’ box and write ‘Korean’ in the space provided Korean Christian (in the ‘Asian Christian’ grouping)

Thus, despite the effort that has been made to classify responses accurately, some groupings in the classification are not strictly mutually exclusive. Another example would be a person who reports ‘Fundamentalist’ – they may also be affiliated with one or more Christian denominations or sub-denominations. But since the respondent has not provided this additional information, categories such as ‘Fundamentalist’ are classified outside the traditional denominational categories.

Other areas of the classification that lack mutual exclusivity are the ethnicity–based categories. For example, people who are affiliated with the Church of Tonga may simply mark the ‘Methodist’ box on the questionnaire module, and therefore would not be included in the ‘Tongan Methodist’ response category.

Sample surveys

Sample surveys that wish to collect information on religious affiliation should adhere to the standard. This means that respondents must always be given the option of objecting to answer (either verbally asked if they object, or shown an ‘object’ category on a ‘showcard’). A showcard of responses may be used, but sample surveys must also collect ‘other’ written or self–identified responses, and code them via the codefile to the lowest level of the classification. This is to ensure that the data from sample surveys will be comparable to Census data and other collections.

  • 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+