Skip to content


New Zealand’s electoral system is a dynamic and integral part of its democratic framework. This article provides an in-depth exploration of the various aspects of New Zealand’s elections, including the electoral system, voting processes, and data analysis resources available from within the scgElectionsNZ package.

NZ Electoral System

New Zealand employs a non-compulsory voting system to elect 120 members to a unicameral House of Representatives using a Mixed-Member Proportional (MMP) voting system. This system comprises:

  • 72 Electorate MPs elected from single-member electorates using the First-Past-The-Post (FPTP) voting method.
    • 7 of these seats are Māori electorates for individuals of Māori ethnicity or ancestry who choose to place themselves on a separate electoral roll. Both Māori and General electorates overlap.
  • 48 List MPs elected from nationwide, closed (i.e., the order of the candidates is determined by individual parties), party lists. There is no distinction between Māori and General electoral roll voters for these seats.

Voters thus cast two votes: a Candidate Vote for their preferred local candidate and a Party Vote for their favoured party.

Explore the summary dataset for an analysis of the party and candidate votes at the national-level. The number and type of seats can also be viewed using this dataset.

Party Vote

The allocation of the 120 parliamentary seats is based on the proportion of Party Votes each party receives. After subtracting the number of electorate seats won, the remaining seats are allocated to parties as a top-up.

The order in which seats are allocated is determined by the Sainte-Laguë formula:

\[ quotient = \frac{V}{2s + 1} \]

where:

  • \(V\) is the total number of votes for a party, and
  • \(s\) is the number of seats already allocated to that party.

For a comprehensive understanding of this allocation process, refer to the Seat Allocation Explained article. Electorate-level Party Vote data can be found in the party_votes dataset.

Candidate Vote

FPTP voting results for electorates are available in the candidate_votes dataset Note that this dataset currently includes only successful parties, with others grouped under ‘Other’. Use the add_data function to identify the type of each electorate (i.e., Māori or General).

The majority dataset lists each electorate winner, the majority, and the margin of victory, calculated as follows:

Majority

\[ M = V_{1} - V_{2} \]

where:

  • \(M\) represents the Majority,
  • \(V_{1}\) is the number of votes received by the first candidate, and
  • \(V_{2}\) is the number of votes received by the second candidate.

Margin of Victory

\[ MoV = \frac{M}{T_{votes}} \]

where:

  • \(MoV\) represents the Margin of Victory,
  • \(M\) is the Majority (as calculated above), and
  • \(T_{votes}\) is the total valid votes cast in electorate for all candidates.

For insights into the use of these datasets, see the Party and Candidate Votes article and the Majority article.

Overhang Seats

In circumstances where a party has won more electorates than it would be entitled to from the proportion of the Party Vote, those candidates retain their electorates and the total number of MPs elected to parliament increases. These are referred to as overhang seats. Learn how to identify overhang seats in each election from the Summary article.

Eligibility

For parties to be eligible to have seats allocated to them, they either need to receive five per cent of the Party Vote or win at least one electorate. The Successful variable in the summary dataset indicates party eligibility in election results.

Before MMP

In 1993, New Zealanders voted to change their voting system from FPTP to MMP. The preMMP_results dataset covers data from 1890 to 1996, allowing comparisons with the post-MMP Party Vote. See the Summary article for an example of how to use these data.

Turnout

The New Zealand Electoral Commission calculates voter turnout as follows:

\[ {T} = \left( \frac{{V_{cast}}}{{E_{total}}} \right) \times 100 \]

where:

  • \(T\) represents the Turnout percentage,
  • \(V_{cast}\) is the Total Votes Cast (includes all valid, informal, and disallowed votes), and
  • \(E_{total}\) is the Total Electors on the Master Roll (including both General and Māori electoral rolls).

Explore the electorate-level turnout data in the turnout dataset, national-level data in the turnout_historic dataset, and how to use both in the Turnout article.

Note that in the latter dataset, the Party Vote turnout is used for the national figure. The figures for 1978 have also been amended to disregard the hundreds of thousands of duplicated names on the electoral roll. More on this can be found in the footnotes on the electoral commissions website here.

Vote Types

New Zealand’s electoral system encompasses a range of vote types, categorised by method and validity.

Methods of Voting

  • Ordinary Votes: Cast in-person on election day.
  • Special Votes - Within New Zealand: For voters outside their electorate, late registrants, those needing anonymity, and those in hospitals or prisons.
  • Special Votes - Overseas: For New Zealanders abroad, ensuring that the 1 in 5 Kiwis overseas can still vote.

Validity of Votes

  • Valid Votes: Properly cast and meeting all criteria.
  • Informal Votes: Where voter intent is unclear, due to improper filling, illegibility, or multiple markings.
  • Disallowed Votes: Not counted due to late receipt, unauthenticated special votes (not witnessed or signed), being given the wrong ballot paper, or ineligibility of the voter.

The vote_type dataset offers an in-depth breakdown of these types. For analysis examples, see the Vote Type article.

Split Voting

Split voting occurs when voters choose a candidate from one party for their Candidate Vote and a different party for their Party Vote. This phenomenon reflects the flexibility of the MMP system, allowing voters to support a local candidate irrespective of their preferred party for proportional representation.

Explore the national-level flow of votes in the split_total dataset and the electorate-level data in the split_electorate dataset. For practical examples, refer to the Split Voting article.