This page provides information on the 2023 Census interim coverage and collection response rates, and the source data for major ethnic groups and Māori descent.
The interim rates provide the best indication of census coverage and response prior to the release of the official measures from the 2023 Post-enumeration Survey, which will be available on 9 December 2024.
- Census dataset
- Calculating the interim rates
- Interim coverage rates
- Collection response rates
- Counting people
- Glossary
There are now records for about 5.0 million people in the 2023 Census dataset, with real data about real people used to fill the gaps when people did not take part in the census.
Census dataset
How many people completed forms in the 2023 Census?
Stats NZ received census form responses for approximately 4.4 million people. Of these, approximately 2.5 percent were partial responses, where we received information via the household form but did not receive an individual form for the person.
How many people are in the census dataset now?
There are records for approximately 5.0 million people in the census dataset. This is 0.9 percent (or 44,000 people) less than our current best estimate of the population on 7 March 2023 (census day). By comparison, the official census undercount in 2018 was 2.6 percent.
Of the total number of records, 89 percent come from 2023 Census forms and 11 percent come from other government data.
Calculating the interim rates
Which denominators are used to calculate the interim rates?
We use two denominators:
- the dual system estimation (DSE) benchmark (or estimated usually resident population count as at census night) for calculating the interim coverage and response rates
- the census usually resident population count for the source data (number of people who have been counted from census forms and administrative (admin) data).
Tables 1-6 all use the DSE benchmark as the denominator, and tables 7-17 use the census usually resident population count as the denominator.
Why are the rates interim?
The official measures of coverage and response in the 2023 Census come from the Post-enumeration Survey (PES), with first results expected to be released on 9 December 2024. From now until December, steps need to be performed to produce the 2023 PES and estimated resident population (ERP) results. This process includes:
- linking PES records to census and admin data, using both automated and manual matching techniques
- modelling the matched data to produce measures of census undercount and overcount for the total population and population subgroups
- deriving the base ERP, which uses census and PES data as key inputs.
In the meantime, Stats NZ has calculated interim measures of coverage and response using its best estimate of the population on 7 March 2023 (census day).
Interim coverage rates
What does coverage mean?
Coverage compares the number of people who were counted in the census with the number who should have been counted. It is usually expressed as a percentage of what should have been the complete count (the expected number of New Zealand residents in the country on census night).
The official measures of coverage for the 2023 Census will come from the PES when published on 9 December 2024.
What is the interim national coverage rate?
The interim coverage rate for the 2023 Census is 99.1 percent. This means the interim coverage gap or undercount is 0.9 percent. This is based on census form data and admin data.
Table 1
2023(1) | |
National population | 99.1% |
1. This rate is interim because it does not yet incorporate Note: Uncertainty estimates will be provided with the Source: Stats NZ | |
What are the interim coverage rates for subgroups?
Table 2 shows the interim coverage rates for the main subpopulation groups that have traditionally been harder to count in the census. The interim coverage rates for subgroups range from 97.2 percent for people of Māori descent to 99.4 percent for the Asian ethnic group. Māori descent and Māori and Pacific ethnic groups tend to have lower coverage rates in the census, due in part to their younger age profiles, as younger people are more likely to be missed.
Table 2
2023(1) | |
National population | 99.1% |
Subgroup populations | |
Māori descent | 97.2% |
Māori ethnic group | 98.0% |
Pacific ethnic group | 98.4% |
Asian ethnic group | 99.4% |
15-29-year-olds | 98.5% |
1. These are interim results because they do not yet Note: Uncertainty estimates will be provided with the Source: Stats NZ | |
How do these compare with previous census coverage rates?
The interim national and subgroup coverage rates for the 2023 Census are higher than the official coverage rates for the 2018 Census as measured by the PES. Table 3 shows the comparison against historical data.
Table 3
2013 | 2018 | 2023(1) | |
National population | 97.6% points)(2) | 97.4% (credible interval | 99.1% |
Māori descent | .. (4) | 95.8% | 97.2%(6) |
Māori ethnic group | 93.9% | 95.6% | 98.0% |
Pacific ethnic group | 95.2% | 95.1% | 98.4% |
Asian ethnic group | 97.0% | 96.7% | 99.4% |
15-29-year-olds | 95.2% | 95.3% | 98.5% |
1. These are interim results because they do not yet incorporate information from the 2023 PES. 2. The sampling error (±) indicates the extent to which an estimate from the PES sample might deviate from the true value. 3. Credible interval describes an interval within which the unobserved quantity of interest falls with a particular probability. For instance, a 95 percent credible interval defines a range of values within which the ‘true’ value has a 95 percent probability of belonging. The lower and upper bounds represent the 2.5th and the 97.5th percentiles. 4. Māori descent information was not captured in the PES prior to 2018, and therefore a 2013 Census coverage rate is 5. The 2018 Māori descent coverage rate is calculated using Māori descent (electoral) census data which is used to determine electoral populations. Responses of ‘Don’t know’ have been imputed to ‘Māori descent’ and ‘No Māori descent’. 6. The 2023 Māori descent coverage rate is calculated using Māori descent (output) census data which includes responses of Note: Uncertainty estimates will be provided with the official 2023 PES rates. Source: Stats NZ | |||
Are the interim coverage rates available by territorial authority local board level?
2023 Census dataset interim coverage and composition provides interim coverage rates for territorial authority local board areas in a story map.
For each area, rates are available for the census usually resident population, Māori descent population, Māori ethnic group, Pacific ethnic group, and Asian ethnic group.
Collection response rates
What is a collection response rate?
This is the field response rate at the end of the collection phase divided by the number of people who should have been counted (DSE benchmark).
The official measures of response for the 2023 Census will come from the PES, when published on 9 December 2024.
What counts as a response?
For a response to be counted, we need to have received two or more of the following information fields about an individual on census forms: name, date of birth, meshblock. This information may come from an individual form, dwelling form, or household set-up form. A meshblock is a defined geographic area, varying in size from part of a city block to large areas of rural land.
What is the national collection response rate?
The national collection response rate is 88.3 percent.
Table 4
2023(1)(2) | |
National population | 88.3% |
1. This is an interim result because it does not yet 2. A census field response needs to include two or Source: Stats NZ | |
What are the collection response rates for subgroups?
Table 5 provides the collection response rates for subgroups. There is significant variation in the collection response rates across subgroups, ranging from 76.7 percent for the Māori ethnic group to 91.5 percent for the Asian ethnic group.
Table 5
2023(1)(2) | |
National population | 88.3% |
Subgroup populations | |
Māori descent | 77.0% |
Māori ethnic group | 76.7% |
Pacific ethnic group | 79.9% |
Asian ethnic group | 91.5% |
15-29-year-olds | 84.8% |
1. These are interim results because they do not yet 2. A census field response needs to include two or Source: Stats NZ | |
How do the rates compare with previous censuses?
The interim collection response rates for the 2023 Census are higher than the rates for the 2018 Census, but lower than the rates for the 2013 Census, both at national and subgroup levels.
Table 6
2013 | 2018(1) | 2023(1)(2) | |
National population | 93.2% | 85.8% | 88.3% |
Subgroup populations | |||
Māori descent | ..(3) | 72.8%(4) | 77.0%(5) |
Māori ethnic group | 89.7%(6) | 73.4% | 76.7% |
Pacific ethnic group | 90.8%(6) | 72.0% | 79.9% |
Asian ethnic group | 93.3%(6) | 85.6% | 91.5% |
15-29-year-olds | 90.3% | 79.5% | 84.8% |
1. A census field response needs to include two or more of name, date of birth, or meshblock. 2. These are interim results because they do not yet incorporate information from the 2023 PES. 3. Māori descent information was not captured in the PES prior to 2018, and therefore a 4. The 2018 Māori descent response rate is calculated using Māori descent (electoral) census 5. The 2023 Māori descent collection response rate is calculated using Māori descent (output) 6. Ethnicity was not imputed in 2013 Census data. These collection response rates are estimates Source: Stats NZ | |||
Counting people
How did we count people?
Of the 5.0 million people in the 2023 Census dataset, 89 percent were counted from census forms. The remaining 11 percent were counted using admin data sources (other government data) through a process we call ‘admin enumeration’.
Approximately 2.5 percent of all people in the dataset were counted from partial census form responses. This is where we received information via the dwelling form or household set-up form but did not receive an individual form for the person.
Table 7 shows how different data sources have contributed to the population being counted for different ethnic groups. For example, for the Māori ethnic group:
- 74 percent of the Māori ethnic group in the census dataset filled out an individual form during the 2023 Census
- 4 percent were counted from information provided on the dwelling form or household set-up form
- for the other 22 percent, admin enumeration was used.
Note also, that because some of these people/records have come from partial responses or admin enumeration, they didn’t identify as being of Māori ethnicity on their census form.
Table 7
Proportion of ethnic group | |||
Ethnic group | Individual form | Partial | Admin |
European | 89 | 2 | 9 |
Māori | 74 | 4 | 22 |
Pacific | 76 | 5 | 19 |
Asian | 89 | 3 | 8 |
MELAA | 87 | 3 | 10 |
Other | 86 | 2 | 12 |
Total population | 87 | 2 | 11 |
Note: The numbers are a proportion of the unit data source for each ethnic group. MELAA is Middle Eastern, Latin American, and African. The sum of each row is 100 (±) 1 due to rounding. Source: Stats NZ | |||
Table 8 shows how the different data sources have contributed to the count of the Māori descent population:
- 75 percent of people in the census dataset with Māori descent filled out an individual form during the 2023 Census
- 4 percent were counted from information provided on the dwelling form or household set-up form
- for the other 21 percent, admin enumeration was used.
2023 Census dataset interim coverage and composition provides information on how we counted people by territorial authority local board area, as a whole and then by each of the six ethnic groups.
Table 8
Proportion of Māori descent | |||
Individual form | Partial | Admin | |
Māori descent | 75 | 4 | 21 |
Note: The numbers are a proportion of the unit data source for Māori descent The sum of each row is 100 (±) 1 due to rounding. Source: Stats NZ | |||
Where did ethnic group information come from?
Most ethnic group data was sourced through census responses. Where ethnicity was not completed on the 2023 Census forms, other data sources were used to fill in the gaps, such as historical census data (2018 Census and 2013 Census), admin data, or statistical imputation.
For example, of the people counted in the Māori ethnic group, 73 percent advised us of their ethnicity on their 2023 Census form. For those who did not tell us their ethnicity on the 2023 Census form, 16 percent came from historical census forms, 10 percent from admin data, and 1 percent were statistically imputed.
Table 9
Proportion of ethnic group | ||||
Ethnic group | 2023 Census response | Historical census | Admin data | Statistical imputation |
European | 89 | 8 | 3 | |
Māori | 73 | 16 | 10 | 1 |
Pacific | 75 | 14 | 10 | 1 |
Asian | 89 | 5 | 5 | 1 |
MELAA | 86 | 7 | 6 | 1 |
Other | 84 | 10 | 4 | |
Note: The numbers are a proportion of the item data source for each ethnic group. MELAA is Middle Eastern, Latin American, and African. The sum of each row is 100 (±) 1 due to rounding. Source: Stats NZ | ||||
Where did we source ethnicity admin data?
For ethnicity, the admin data sources are Department of Internal Affairs (births), Ministry of Education (school and tertiary enrolments), Ministry of Health (population cohort demographics), and Ministry of Social Development (benefit recipients).
What is statistical imputation?
The term ‘imputation’ refers to the replacement of missing information with values from a statistical process (in contrast to our methods of sourcing real values from admin or historical census data).
How did we count the person by ethnic group information source?
As an example, for the European ethnic group (table 10):
- 88.6 percent submitted a 2023 Census form and responded to the ethnicity question
- 0.4 percent submitted a 2023 Census form where they didn’t respond to the ethnicity question, so the ethnic information was sourced from historical census data
- less than 0.1 percent submitted a 2023 Census form but didn’t respond to the ethnicity question and the ethnic information wasn’t available from historical census data, so it was sourced from admin data
- very few records required statistical imputation (less than 0.1 percent).
For each ethnic group we have provided the breakdown in tables 10-15.
Table 10
Item source | ||||
Unit source | 2023 Census | Historical | Admin | Statistical |
Individual form response | 88.6 | 0.4 | ||
Partial response | … | 1.0 | 0.3 | 0.5 |
Admin enumeration | … | 6.5 | 2.4 | |
Note: The numbers are a proportion of the unit and item sources for the European ethnic group The sum of the table is 100 (±) 0.1 due to rounding. Symbol: … not applicable Source: Stats NZ | ||||
Table 11
Item source | ||||
Unit source | 2023 Census | Historical | Admin | Statistical |
Individual form response | 73.1 | 0.5 | 0.2 | |
Partial response | … | 2.1 | 1.2 | 1.0 |
Admin enumeration | … | 13.3 | 8.4 | |
Note: The numbers are a proportion of the unit and item sources for the Māori ethnic group The sum of the table is 100 (±) 0.1 due to rounding. Symbol: … not applicable Source: Stats NZ | ||||
Table 12
Item source | ||||
Unit source | 2023 Census | Historical | Admin | Statistical |
Individual form response | 75.2 | 0.4 | 0.3 | |
Partial response | … | 2.6 | 1.6 | 1.1 |
Admin enumeration | … | 10.8 | 7.9 | |
Note: The numbers are a proportion of the unit and item sources for the Pacific ethnic group The sum of the table is 100 (±) 0.1 due to rounding. Symbol: … not applicable Source: Stats NZ | ||||
Table 13
Item source | ||||
Unit source | 2023 Census | Historical | Admin | Statistical |
Individual form response | 89.1 | 0.2 | 0.2 | |
Partial response | … | 1.1 | 0.9 | 0.6 |
Admin enumeration | … | 4.1 | 3.5 | 0.3 |
Note: The numbers are a proportion of the unit and item sources for the Asian ethnic group The sum of the table is 100 (±) 0.1 due to rounding. Symbol: … not applicable Source: Stats NZ | ||||
Table 14
Item source | ||||
Unit source | 2023 Census | Historical | Admin | Statistical |
Individual form response | 85.9 | 0.5 | 0.4 | |
Partial response | … | 1.1 | 1.0 | 0.7 |
Admin enumeration | … | 5.2 | 4.7 | 0.4 |
Note: The numbers are a proportion of the unit and item sources for the MELAA ethnic group MELAA is Middle Eastern, Latin American, and African. The sum of the table is 100 (±) 0.1 due to rounding. Symbol: … not applicable Source: Stats NZ | ||||
Table 15
Item source | ||||
Unit source | 2023 Census | Historical | Admin | Statistical |
Individual form response | 84.5 | 1.0 | 0.4 | |
Partial response | … | 1.2 | 0.6 | 0.5 |
Admin enumeration | … | 8.1 | 3.5 | 0.1 |
Note: The numbers are a proportion of the unit and item sources for the Other ethnic group The sum of the table is 100 (±) 0.1 due to rounding. Source: Stats NZ | ||||
Where did Māori descent information come from?
Most Māori descent data was sourced through census responses. Where Māori descent was not completed on a 2023 Census form, other data sources were used to fill in the gaps. These sources are historical census data (2018 Census and 2013 Census), admin data, parental data, deterministic derivation, or statistical imputation.
For example, of people counted with Māori descent, 74 percent advised us of this on their 2023 Census form. For those who didn’t respond to the Māori descent question on the 2023 Census form:
- 15 percent came from historical census forms
- 9 percent from admin data
- less than 1 percent from parental data
- less than 1 percent from deterministic derivation
- 2 percent were statistical imputation.
Table 16
Proportion of Māori descent | ||||||
2023 Census | Historical | Admin | Parental | Deterministic | Statistical | |
Māori descent | 74 | 15 | 9 | 2 | ||
Note: The numbers are a proportion of the item data source for Māori descent The sum of the table is 100 (±) 1 due to rounding. Source: Stats NZ | ||||||
Where did we source administrative Māori descent data from?
For Māori descent, the admin data sources are the Department of Internal Affairs (births) and the Electoral Commission.
What is parental data?
Where historical census data and admin data are not able to fill missing Māori descent responses, we use Department of Internal Affairs birth records to link children to their parents. Māori descent values for parents can then be used to fill missing information on Māori descent for their children.
What is deterministic derivation?
Deterministic derivation is when a value from a census variable can be used to inform a missing value in another census variable. For Māori descent, a valid iwi value on the 2023 Census form, historical census data, or admin data is used to confirm Māori descent. A valid value in this context excludes ‘Don’t know’ and residual responses (‘Refused to answer’, ‘Response unidentifiable’, and ‘Response outside scope’).
How did we count the person by Māori descent information source?
For example, for Māori descent information (table 17):
- 73.9 percent submitted a 2023 Census form and stated they were of Māori descent
- 0.8 percent submitted a 2023 Census form but did not respond to the Māori descent question, so this information was sourced from historical census data
- 0.3 percent submitted a 2023 Census form but did not respond to the Māori descent question, and this information wasn’t available from historical census data, so this information was sourced from admin data
- less than 0.1 percent submitted a 2023 Census form but did not respond to the Māori descent question, and this information wasn’t available from historical census data or admin data for the individual, but their parent’s/parents’ or grandparent’s/grandparents’ Māori descent data was sourced
- less than 0.1 percent submitted a 2023 Census form but did not respond to the Māori descent question, and this information wasn’t available from historical census data, admin data for the individual, or from parental data, so this information was sourced from deterministic derivation
- less than 0.1 percent required statistical imputation.
Table 17
Item source | ||||||
Unit source | 2023 Census forms | Historical | Admin | Parental | Deterministic | Statistical |
Individual form | 73.9 | 0.8 | 0.3 | |||
Partial response | … | 2.0 | 1.1 | 1.0 | ||
Admin enumeration | … | 12.5 | 7.2 | 0.3 | 0.3 | 0.4 |
Note: The numbers are a proportion of the unit and item sources for the Māori descent census usually resident population count. The sum of the table is 100 (±) 0.1 due to rounding. Symbol: … not applicable Source: Stats NZ | ||||||
Glossary
Census glossary provides further explanation of census terminology.
ISBN 978-1-99-104979-7
