Report of the Auditor-General

The Auditor-General has completed a review of vote counting errors in the 2023 General Election.

The Electoral Commission sought an independent review after data entry errors were found in the 2023 General Election official results. The errors were corrected, and the results republished on 9 November 2023. 

The errors had no effect on the outcome in any electorate or for the allocation of seats in Parliament.

The Office of the Auditor-General’s report was presented to Parliament on 7 May 2024.

General Election 2023: Independent review of counting errors (Office of the Auditor-General)

Electoral Commission response 

The Electoral Commission welcomes the report and will fully implement the Auditor-General’s recommendations to strengthen the way votes are recorded, counted and checked.

Work underway on 7 May 2024 includes:

  • a comprehensive internal review of the delivery of the 2023 General Election and post-election period
  • adopting a new assurance framework 
  • audits of post-election and enrolment processes to identify any further gaps or improvements 
  • reviewing operating manuals and training to improve the implementation of quality assurance controls during delivery of elections 
  • prioritising improvements to post-election processes to improve the integrity and timeliness of the official count
  • approving short-term changes to strengthen the official count process for any by-election required before a full review of post-election processes is completed.

The Board of the Electoral Commission will closely monitor and report on the delivery of these and other improvements to ensure New Zealanders can continue to have trust in the delivery of elections. 

Updates on the implementation of the Auditor-General’s recommendations will be posted on this page.

Possible dual votes

During the Auditor-General’s review, it was found some possible dual votes were included in the official results.

A dual vote is where a voter appears to have voted more than once. An investigation is undertaken to determine whether it is a clerical error, for example, an incorrect marking of the roll, or a suspected case of dual voting.

There were cases where possible dual votes (known as apparent dual votes) that hadn’t been fully investigated were included in the official count. This does not mean they were dual votes, but that they could have been.

The total number of apparent dual votes that could have impacted candidate counts is 321 votes across 40 electorates. In each of those electorates, including where there were recounts, the number of apparent dual votes left in the count is less than the winning candidate margin, so could not have affected the results.

Electorate breakdown of apparent dual votes [PDF, 74KB]

Back to top