Algorithm assessment report 2018

A cross-government review of how government uses algorithms to improve the lives of New Zealanders was undertaken in 2018. The review aimed to ensure New Zealanders are informed, and have confidence in how the government uses algorithms.

Algorithm assessment report - full report [HTML]
Algorithm assessment report - full report [PDF 10.1 MB]
Algorithm assessment report - one-page summary [PDF 151 KB]

About the review

In 2018, the New Zealand government undertook a review of how government agencies use algorithms to improve the lives of New Zealanders. Led by the Government Chief Data Steward and Government Chief Digital Officer, the review initially focused on operational algorithms that result in, or inform, decisions that directly impact people or groups.

Government Chief Data Steward
Government Chief Digital Officer

Fourteen government agencies self-assessed the algorithms they use to deliver their functions, focussing on areas most directly impacting decisions related to people.

The 'Algorithm assessment report' summarises these self-assessments. The report aims to ensure New Zealanders are informed about the use of government algorithms and the checks and balances that exist to manage their use.

The report drew on the ‘Principles for safe and effective use of data and analytics’ developed by the Privacy Commissioner and the Government Chief Data Steward and made recommendations to improve transparency and accountability in government algorithm use.

Principles for safe and effective use of data and analytics

Response to the algorithm assessment report

Responding to the recommendations from the report, in 2020 the Government Chief Data Steward has:

released the ‘Algorithm charter for Aotearoa New Zealand’ in which signing agencies make a public commitment to the transparent and accountable use of algorithms.
ensured that government agencies can seek external advice from privacy, ethics, and data experts by continuing to convene the independent Data Ethics Advisory Group.
continued work to ensure that data ethics are embedded at all levels of the government analytics workforce, through improvements in training and professional development for data analysts and decision makers.

Responding to the algorithm assessment [PDF 201KB]

Responding to the algorithm assessment [HTML]

Data Ethics Advisory Group [HTML]

Key documents

These documents provide more information on the process used to produce the Algorithm assessment report.

Commissioning the work – Report: Review of Government algorithms
Agency submissions: Algorithm assessment agency submissions: June – July 2018

Report: Review of Government algorithms [PDF 749 KB]
Algorithm assessment agency submissions: June – July 2018 [PDF 1.68 MB]

Some documents contain redacted information under the following sections of the Official Information Act 1982:

9(2)(a) to protect the privacy of natural persons.
9(2)(f)(iv) to maintain the confidentiality of advice tendered by Ministers of the Crown and officials.

In preparing this information release, Stats NZ considered the public interest considerations in section 9(1) of the Official Information Act.

Official Information Act 1982

More information

The Algorithm Charter
Blog: Algorithm review underway to increase transparency and accountability (18 July 2018)
Media release: Government to undertake urgent algorithm stocktake 923 May 2018)
Media release: Government will move quickly on AI action plan (2 May 2018)

Summary of findings

Algorithms have an essential role in supporting the services that government provides to people in New Zealand and to help deliver new, innovative, and well targeted policies to achieve government aims.

In one year:

3.7 million people pay Pay As You Earn (PAYE) tax or file a personal tax return.
277,410 people receive working age benefits and 753,319 receive New Zealand Superannuation.
More than 700,000 people are referred for specialist surgery.
There are around 2 million accident compensation claims.
More than 1.5 million children, young people, and students enrol in the education system.
158,900 referrals about children at risk are made, and further action was needed in 33,000 cases.
41 million mail items and 496,000 import sea containers are screened, and 13.9 million travellers are assessed at the border.
More than 1.1 million events are responded to by the police, and more than 100,000 cases are prosecuted.

The report presents an assessment of information reported by 14 government agencies about the computer algorithms they use to deliver their functions. It focuses on algorithms that are used in decision-making processes that affect people in direct and significant ways.

Key findings and recommendations

Humans, rather than computers, review and decide on almost all significant decisions made by government agencies. As agencies continue to develop new algorithms, it’s important to preserve appropriate human oversight and ensure that the views of key stakeholders, notably the people who will receive or participate in services, are given appropriate consideration.
While agencies are applying a range of safeguards and assurance processes in relation to their algorithms use, there are opportunities for increased collaboration and sharing of good practice across government.
There is also scope to ensure that all of the information that is published explains, in clear and simple terms, how algorithms are informing decisions that affect people in significant ways.
The government commitment to Treaty-based partnership should also be reflected, embedding a te ao Māori perspective into the development and use of algorithms.
In addition to connecting expertise across government, agencies could also benefit from a fresh perspective by looking beyond government for privacy, ethics, and data expertise. This could be achieved by bringing together a group of independent experts that agencies could consult for advice and guidance.

Contact us

If you’d like more information, have a question, or want to provide feedback, email datalead@stats.govt.nz.

Content last reviewed 15 June 2023.