As the lead agency for government-held data, we make sure we protect your privacy when we're working with your data to produce insights and statistics about New Zealand.
Here's what we do to protect your information. Integrated data is a rich source of information about New Zealand people, households, and businesses. By bringing together information from different sources, we can produce insights and statistics more efficiently. Approved researchers can also examine data in ways that haven't been possible before.
We understand the value of this information to Aotearoa and our role as its data steward. We ensure this data is used safely in ways that benefit all New Zealanders.
Since 2012, we've used a 'five safes framework' to protect privacy when we make de-identified data available to approved researchers. We only provide access to integrated data if all our 'five safes' conditions are met: safe people, safe projects, safe settings, safe data, and safe output - see The five safes framework.
How we link data
We integrate datasets by linking information using identifiable data, including first and last name, date of birth, age, sex, and country of birth. At this stage of the process, the data is not accessible for analysis or research.
Data integration starts with a central set of data formed by combining government tax, birth registrations, and visa application records.
This set of data forms a central point for other government data we hold, and Stats NZ surveys, to be linked to. Once we link the data, we de-identify it before it is made available for statistical and research purposes.
Stats NZ staff who look after linked data are the only people who work with identifiable information. Different teams within Stats NZ then analyse and produce statistics and other information from de-identified data.
How the linked data is used
Approved researchers can access de-identified linked data through the Integrated Data Infrastructure (IDI). The IDI is refreshed regularly with updated information from various government agencies, non-government organisations, and Stats NZ surveys. See Integrated Data Infrastructure.
We conduct research using the IDI to look into the feasibility of a future census based mainly on administrative data. This includes using IDI data to produce an experimental administrative residential population that can be analysed and compared with Stats NZ's official measures of the population.
We carry out all data linking in accordance with:
- Statistics Act 1975
- Privacy Act 1993
- Stats NZ's Information privacy, security, and confidentiality policy.
How we're using linked data for the 2018 Census
We fast-tracked the development of new and innovative methods for using administrative data to help compensate for lower than expected participation in the 2018 Census.
We used data from the 2013 Census, and other government data, to fill in information missing from the 2018 Census dataset. To do this, we linked administrative data to form a list that is a good approximation of the New Zealand population. We then linked this list to the 2018 Census forms data to work out who was missing, and therefore, determine which administrative data could be added into the 2018 Census dataset accurately.
This approach has built on the research that has been underway for seven years, to explore how a future population census could use mainly administrative data.
2018 Census: How we combined administrative data and census forms data to create the census dataset has more information about how we added administrative data to the 2018 Census dataset.
How we look after your information
We take our role to look after all the information held within the organisation seriously.
We retain identifiable census and administrative data securely, and access to this data is highly restricted. This data includes names, addresses, and other identifiable information such as date of birth. Access to all data is strictly controlled, and identifiable information is only accessible to a small number of staff.
We make sure all the information is as secure as possible - it is protected during transit, during the linking process, and once the process is complete, so that no identifiable personal information is shared with anyone who should not see it.
Everyone employed by Stats NZ, and all the approved researchers who use the information we hold once it has been de-identified, sign a statutory declaration of secrecy before they start working for us, or use the information. This declaration is a lifetime obligation.
We never share identifiable information with other government agencies without consent and our published statistics and research do not identify individuals, households, or businesses. We remove all names, addresses, and other identifiable information before we make data available to other government agencies and researchers for statistical and public interest research.
In addition, confidentiality methods (such as data rounding) are applied to all statistical and research outputs to prevent re-identification of personal information.
Our legislation
The Statistics Act 1975 empowers Stats NZ to collect a variety of information about people, businesses, and other organisations using both surveys and through information already collected by another organisation, such as a government department. The Act governs the use of the data we collect about individuals, households, and organisations. Our data collection is for statistical purposes, and for research that's necessary to support decision-making by New Zealanders, businesses, and government.
More information
2019 Privacy impact assessment for creating the 2018 Census dataset by combining survey and administrative data (full privacy impact assessment)
