Ethnicity 2013 Census (information about this variable and its quality)

##Where the data comes from

Question 11 on the individual form.

##How this data is classified

Ethnic group (total responses) level 1

1 European

2 Māori

3 Pacific peoples

4 Asian

5 MELAA (Middle Eastern / Latin American / African)

6 Other ethnicity

9 Not elsewhere included

More detailed information is available at lower levels of the ethnicity classification.

Census ethnic group – grouped total responses 2011 (new for 2013, contains a revised level 2)

1 European

10 European nfd

11 New Zealand European

12 Other European

2 Māori

20 Māori

3 Pacific peoples

30 Pacific peoples

4 Asian

40 Asian

5 Middle Eastern / Latin American / African

50 Middle Eastern / Latin American / African

6 Other ethnicity

61 New Zealander

69 Other ethnicity nec

9 Not elsewhere included

94 Don't know

95 Refused to answer

97 Response unidentifiable

98 Response outside scope

99 Not stated

Data that has classification categories aggregated is also available. See below 'Further information about this data'.

For further information about this classification, refer to the:

• 2013 Census data dictionary
• Statistical standard for ethnicity
• Classification Code Finder.

For background information on classifications and standards, refer to the Classifications and related statistical standards page.

##Subject population

The subject population for this variable is the census night population, as this question applies to all people in New Zealand on census night. However, data on ethnicity is usually output for the census usually resident population.

The subject population is the people, families, households, or dwellings to whom the variable applies.

##Non-response and data that could not be classified

Non-response

'Non-response' is when an individual gives no response at all to a census question that was relevant to them. The non-response rate is the percentage of the subject population that was coded to ‘Not stated’. The response rates here are for the census night population.

• Non-response rate for 2013: 5.3 percent of the census night population, of which 4.7 percent were substitutes.
• Non-response rate for 2006: 4.0 percent of the census night population, of which 3.2 percent were substitutes.
• Non-response rate for 2001: 3.8 percent of the census night population, of which 2.8 percent were substitutes.

Not elsewhere included

Non-response and responses that could not be classified or did not provide the type of information asked for are usually grouped together and called 'Not elsewhere included'.

• 5.5 percent of the subject population was coded to 'Not elsewhere included' in 2013, compared with 4.3 percent in 2006 and 4.2 percent in 2001.

For more information on non-response and substitute records, refer to the 2013 Census data user guide.

##How this data is used

Data from this variable is used:

• to compile a wide range of demographic estimates and projections, and to derive measures for monitoring the well-being of ethnic groups, particularly in the health sector (for example, morbidity rates)
• to determine ethnic group well-being based on analysis against other census variables such as income, employment, education, and housing
• as an important source of demographic information for small areas and ethnic groups with small populations within New Zealand
• by a range of organisations and groups to determine whether there is a need for services, or whether these need to be delivered in another manner: for example, targeted health services, or services in another language
• in the allocation of health funding, development of tailored programmes, and monitoring of results across different ethnic groups
• by local governments(particularly those in regions experiencing demographic, economic, and social change) to effectively target consultation with the community, and in the planning and delivery of services
• by Māori and other ethnic groups in planning and engaging with government agencies
• to produce time series data on how ethnic groups have changed over time, and to show the changing patterns of immigration
• by central government for the compilation of population statistics and information on the key drivers of demand for health services, and for decision-making on funding provisions to district health boards
• by politicians to help them better understand the demographic make-up of their electorate
• to assist in meeting policy commitments related to the Treaty of Waitangi
• in conjunction with data from other sources, such as schools and hospitals, to monitor trends in areas such as educational achievement and disease.

##Data quality processes

All census data was checked thoroughly during processing and evaluation, to ensure that it met quality standards and is suitable for use. These quality checks included edits.

All data must meet minimum quality standards to make it suitable for use.

Quality level

A quality level is assigned to all census variables: foremost, defining, or supplementary.

Ethnicity is a foremost variable. Foremost variables are core census variables that have the highest priority in terms of quality, time, and resources across all phases of a census.

Mode of collection impacts – online form compared with paper form

The online forms had built-in editing functionality that directed respondents to the appropriate questions and ensured that their responses were valid. As a result of this, data from online forms may be of higher overall quality than data from paper forms. The significance of this will depend on the particular type of analysis being done. There will always be a mode effect but this cannot be measured. Statistics NZ design and test to minimise the effects of mode for all questions.

There were differences between how the forms were completed online and on paper for this variable:

• On the online form, the ethnicity question had to be answered in order for the respondent to submit the form. Non-response to this question was possible when forms were completed on paper.
• On the online form, it was only possible to give text responses if 'Other' was marked. When forms were completed on paper, it was possible to give a text response but not mark the 'Other' tick box.
• Responses in the text field on the online form were limited in length.

##Quality assessment of data and data quality issues for this variable

Overall quality assessment

High: fit for use – with minor data quality issues only. 2013 Census variable quality rating scale gives more detail.

Issues to note

• Non-response rate for 2013: 5.3 percent of the census night population, of which 4.7 percent were substitutes.
• Substitute forms and census undercount are both indicators of non-participation in census. The high number of substitutes in 2013 may indicate a higher undercount for some ethnic groups. Estimates of net census undercount (gross undercount less overcount) from the 2013 Post-enumeration Survey are scheduled for release in March 2014, and will include data on four broad ethnic groups: European, Māori, Asian, and Pacific peoples. The Post-enumeration Survey is an independent sample survey run soon after census day, to check the accuracy of census coverage by measuring undercount and overcount. Historically, the net census undercount rate for New Zealand has been about 2 percent, with Māori, Asian, and Pacific peoples having a higher undercount than the national rate, and the European population a lower undercount.

For more information on non-response and substitute records, refer to the 2013 Census data user guide.

##Comparing this data with previous census data

This data is highly comparable with data from the 2006 and 2001 Censuses but the following issues should be noted.

Changes in the data over this time period can generally be interpreted as real changes. There may be a small component of change over time that is due to minor changes in the collection, definition, or classification of the data. The exception is in the New Zealander response, which has fluctuated over time.

Comparisons of data from 2013 and 2006 with 2001 Census data

Although the 2006 question was consistent with those used in 1991 and 2001, the output was not consistent, due to the revised classification used for the 2006 Census. 'New Zealander' responses were no longer coded as 'NZ European' but were classified under 'Other ethnicity' at level one of the classification. 'New Zealander' was output at the lowest level of the classification.

'Middle Eastern, Latin American, and African (MELAA) ethnicities were included in a separate category at the highest level of the ethnic classification. Previously these responses had been coded under 'Other'.

Note that for the last three censuses, six ethnic responses were captured and coded, rather than three responses in 1996 and previous censuses. This variable is output for both six and three responses, so that data can still be compared back to 1996. While the number of respondents specifying multiple responses is increasing over time, the number giving more than three is still low (about a quarter of a percent) so the impact on data quality is minimal.

The 2013 census is the second census in which the Middle Eastern / Latin American / African (MELAA) grouping is used at level 1.

The New Zealander response in 2006 and 2013

Public discussion about the term 'New Zealander' occurred during the months leading up to the 2006 Census. A media campaign encouraged people to write in a New Zealander response in the census. This campaign impacted 2006 ethnicity data with 'New Zealander' responses rising from approximately 80,000 (2.4 percent) in 2001 to just under 430,000 people (11.1 percent) in 2006, making it the third largest ethnic group in the 2006 Census. Analysis showed that there was no corresponding rise in the New Zealander response in other data sources. See Profile of New Zealander responses for more information.

The European grouping decreased from 80.0 percent of the population in 2001 to 67.6 percent in 2006 and is broadly proportional to the increase seen in the 'Other ethnicity' category ('New Zealander'). There is therefore a discontinuity in the time series of ethnic group by size in the 2001/2006 period.

In 2013 there was very little public discussion about the 'New Zealander' term, and the number of 'New Zealander' responses dropped back to just under 66,000 (1.6 percent of the census usually resident population). The European grouping increased again to around 74 percent. While this percentage is not as high as in 2001, this is probably due to the rapid rise in the Asian ethnic grouping (11.8 percent of the census usually resident population in 2013 compared to 9.4 percent in 2006 and 6.8 percent in 2001).

Since 'New Zealander' made up more than 97 percent of the level 1 'Other' grouping, it is reasonable to combine 'Other' and 'European' categories when comparing across censuses.

##Comparing this data with data from other sources

Census is the only information source that provides comprehensive information for small areas and small populations. However alternative sources of information about this subject are available:

• National and subnational ethnic population projections
• New Zealand General Social Survey (NZGSS)
• Household Labour Force Survey (HLFS)
• Household Economic Survey (HES)

Data from these alternative sources may show differences from census data for several reasons. These could be due to differences in scope, coverage, non-response rates, data being collected at different periods of time, alternative sources being sample surveys and as such subject to sampling errors, or differences in question wordings and method of delivery (self-administered versus interviewer-administered). Data users are advised to familiarise themselves with the strengths and limitations of individual data sources before comparing with census data.

Census data is used as the baseline for population projections so the accuracy of the projections tends to deteriorate with time elapsed after the census date. Projections are limited to selected geographic areas, broad ethnic groups, and highly dependent on assumptions

##Further information about this data

All percentages in census publications have been calculated using 'Total people stated' as the denominator.

When using this data, be aware that:

• Ethnicity is a multiple response variable. This means an individual can be counted in more than one ethnic group. The 4,011,399 people (census usually resident population) who gave a valid response to the ethnicity question provided 4,538,985 responses.

• The number of substitute forms required has been increasing in recent censuses: in 2006 it was 3.2 percent of census night population, compared to 4.7 percent in 2013. As New Zealand has become increasingly ethnically diverse, there has been an increase in complex responses to the ethnicity question. Coding standards have been developed to deal with multiple-worded responses. Some multiple-worded responses refer to one ethnic group, such as 'Fijian Indian', 'Turkish Cypriot'. These are coded to 'Indian', and 'Turkish' respectively. There are responses that may be hyphenated or linked in some way, or written without linkage, that need to be classified and coded as two responses. For example, Polish-Hungarian would be counted as Polish and Hungarian in census data.

• Ethnicity data is output in three different ways:

  • 'Total responses' – this type of output shows more detailed categories, such as English, Irish, Samoan, Tongan. A person whose ethnicities were, for example, English, Irish, and Māori would be counted in each of these three categories. The total responses will therefore be greater than the subject population, as the individuals may be counted more than once, ie the total will be greater than 100 percent.
  • 'Grouped total responses' – if two or more of an individual's ethnic groups fall into the same broad ethnic group category, then they are counted only once in that category. Grouped total response output uses the six level 1 categories: European; Māori; Pacific peoples; Asian; Middle Eastern / Latin American / African, and Other ethnicity. A person who reported their ethnicities as, for example, English, Irish, and Māori will be counted once in the European category and once in the Māori category. As with total responses, the number of grouped total responses will be greater than the subject population, as individuals can provide more than one response.
  • 'Single and combination responses' – this type of output has single categories for people who reported only one ethnicity, and combination categories for people who reported more than one ethnicity. People are counted once in the single or combination category that applies to them.

• There has been an increase in the proportion of people stating multiple ethnicities. In 2001, 9.0 percent of respondents to the ethnicity question stated that they had more than one ethnicity. This percentage had risen to 10.4 percent in 2006 and 11.2 percent in 2013.

Contact our Information Centre for further information about using this variable.