BERL Data Newsletter - May 2019

2018 Census update

On 29 April 2019 Statistics New Zealand finally announced the first release of 2018 Census data will be on 23 September 2019. This first release will include usual resident population counts for New Zealand, along with population counts by region, local authority, and the new statistical areas. 

In addition to announcing the first release of 2018 Census data, Statistics New Zealand also ran a small number of Census technical workshops in early May. These workshops focussed on the methodology Statistics New Zealand used to fill in the gaps in the 2018 Census, using Statistics New Zealand’s Integrated Data Infrastructure (IDI).

The main takeaways from the workshop:

1. In determining who did not complete a Census form, Statistics New Zealand used birth records, death records, arrival and departure records, along with IRD tax registration data to determine who was alive and living somewhere in New Zealand as at March 2018. This data was used to match records with the people who had fully or partially completed the 2018 Census.
2. This process revealed that 526,000 people (or 11 percent of census population) did not complete any part of the 2018 Census.
3. Out of these 526,000 people, around 165,000 could be placed into a household, while the remaining 361,000 people could only be placed within a meshblock.
4. Statistics New Zealand used data from 2013 Census, for the following variables in the 2018 Census for the 526,000 people:
   1. Country of birth
   2. Maori Descent
   3. Number of children born (for those aged 45+)
   4. Languages spoken
   5. Ever-smoked
   6. Regular smoker
   7. Years since arrived in New Zealand
   8. Ethnic group
   9. Religious affiliation
   10. Highest secondary school qualification (Updating from MOE records if finished high school after 2013)
   11. Highest post-school qualification (Updating from MOE records if finished a qualification after 2013)  
5. Apart from the iwi affiliation variable which will not be reported in the 2018 Census, other variables such as disability, unpaid activities and quality of homes will be of a lower data quality. This is because these census questions have no other data sources that Statistics New Zealand could use to fill in any gaps to boost the data quality.

So what does this mean for us?

Overall the Statistics New Zealand announcements on 29 April mean that by the end of 2019, we will have some 2018 Census data available for use in our research and project work.  

In using the Census data that will be released on 23 September, we as researchers, will have to consider the following issues carefully:  

• The data quality of each variable and population group we use from the 2018 Census. We will need to ask ourselves what is the level of imputation or administrative people included in these variables and population groups.  
• Combine 2018 Census data with data from other sources (including other Statistics New Zealand data) to double check that we are happy with what the 2018 Census data is telling us.
• Iwi affiliation data will not be available in 2018 Census data releases, we will need to either make use of the IDI or use a more complex methodology in undertaking projects that focus on a particular iwi.  
• Data for new or changed questions in the 2018 Census, such as quality of homes, will most likely be of a lower quality. This is because these questions cannot be imputed from other sources (such as 2013 Census).
• Sense checking of all data will be crucial when working with 2018 Census data, and ensure we understand and state the limitations of the data we are working with.

Urban influence classification - draft paper

This draft paper released by Statistics New Zealand sets out their proposed methodology for an urban influence classification. This classification has been created to classify rural areas and smaller urban areas based on their proximity and linkages to larger urban areas, or level of remoteness. This urban influence classification will enable Statistics New Zealand to effectively determine a set of functional urban areas.

Statistics New Zealand used an OECD definition (OECD (2012) Redefining urban: a new way to measure metropolitan areas), which defines a functional urban area as: ‘a functional urban area consists of a city and its commuting zone. Functional urban areas therefore consist of a densely inhabited city and a less densely populated commuting zone whose labour market is highly integrated with the City’. 

In order to meet this definition Statistics New Zealand decided that a functional urban area will have more than 10,000 residents and will consist of:

• Core urban areas
• Secondary urban areas where at least 40 percent of the working population commute to the core urban area
• Satellite urban areas where at least 40 percent of the working population commute to the core urban area
• Rural settlements and rural hinterland where at least 50 percent of the working population commute to the core urban area.

In addition, Statistics New Zealand decided that the functional urban areas should be split into three sub-groups, which are:

• Metropolitan areas with an estimated resident population of 100,000+
• Large urban areas with an estimated resident population of 30,000 to 100,000
• Medium urban areas with an estimated resident population of 10,000 to 30,000.

The following are some examples using their new classification:

• The Wellington metropolitan functional urban area will consist of the Wellington City as the urban core, with Porirua, Lower Hutt, Upper Hutt, Paraparaumu, and Waikanae as secondary urban cores, and Paekakariki and Featherston as satellite urban areas.
• The Hastings large regional functional urban area will consist of Hastings as the urban core, with Havelock North as a secondary urban core, and Clive as a satellite urban area.
• The Queenstown medium regional functional urban area will consist of Queenstown as the urban core, and Arthurs Point, Lake Hayes, and Arrowtown as satellite urban areas.

The full paper on the methodology and description of the first draft areas can be requested from Statistics New Zealand.