STV time is up
Time to remove a complicated voting system
For those suffering from insomnia at 2.30 in the morning, can I recommend a taxpayer funded website stv.govt.nz
On second thoughts it may not just make your eyes glaze over. It could also make you bloody angry about how some local government political activists, aided and abetted by their bureaucrats, have conned their voting public to adopt the ridiculous voting system known as Single Transferable Vote or STV.
I raise the subject again after the results of a recent by-election at the Dunedin City Council were finalised this week.
There were fourteen candidates to replace the late Jules Radich, a sitting councilor and former mayor who died suddenly earlier this year.
The most popular candidate was another former mayor Aaron Hawkins. Of all those who voted, and there were only 31,629 of them or 33.31 percent of those eligible, Hawkins received 7740 first preferences. Jo Galer was number 1 on 5527 voting papers and in second place.
To me that says Hawkins was the preferred choice of more voters than anyone else and should have been elected.
But no. Using the STV system Jo Galer has ended up winning the by-election and will take her seat at the DCC table next week.
There’s a real irony here. When Hawkins became mayor in 2019 he was second in the popular vote behind conservative Lee Vandervis. Hawkins was about 400 votes behind Vandervis when the number 1 rankings were collated. Vandervis was the city’s most popular choice to be mayor seven years ago but was shafted from the top job by the vagaries of STV.
And just how exactly does STV work?
I wish I could explain it in plain and straight forward language But even the stv.govt.nz website says this: “Votes are counted using specially developed computer software. Because parts of votes are transferred, the counting is too complex to be done by hand.”
That’s just extraordinary.
There are numerous local authorities in this country using this voting system which is essentially opaque and because of its complicated nature I believe it lacks real integrity.
I’ve read various academic and encyclopedic explanations in support of STV.
The basic premise is that STV is a much fairer system because one party or one person cannot dominate an election and that a successful candidate will have at least some modicum of support from people whose first preference was for someone else to have the job.
There is a degree of logic in that but democratic cycles in New Zealand are short and simplicity is always the best answer to any problem. If we have an easy to understand system whereby he or she with the most votes is the winner then that winner can easily be voted in or out in three years depending on their performance in the role.
Wellington City Council has used STV for most of this century. Arguably the most politically aware city in the country voted by referendum in 2002 to adopt STV and reinforced that with another public vote six years later. The system gave them Tory Whanau.
Its popularity in the capital is probably based on the politically active left knowing that even if their support was not enough for a win in a straight out race, having enough of their candidates on a list that need to be ranked by voters will help those candidates get elected when the computer gets to work.
The big issue with STV is its complication. Any election outcome determined by a computer programme that includes fractions of votes can surely not be regarded as a system with integrity.
One academic paper I read suggested even some candidates cannot understand the system. Isn’t that a significant issue?
At last year’s Local Body Elections 15 out of 78 councils used STV. They’d adopted that system either through a public poll, as was the case in Wellington, or because the councilors themselves voted to do it, as happened at the Otago Regional Council in 2023. The then left leaning ORC decided that they could consolidate that block using STV for the 2025 and 2028 elections.
The move completely backfired as the ORC political composition shifted significantly to the right in 2025 – mainly because a ward in the more populous hinterland was installed to replace one in Dunedin – and by 2028 the ORC is likely to be out of existence anyway in local government re-organsiation.
Those changes will require a new Local Government Act to replace the one that the Helen Clark regime put in place in 2002.
Here are two suggestions for that new legislation:
1. Make the duties of local government prescriptive. Tell them what they have to do and nothing more. Remove that woolly clause about the current purpose of local government providing “for local authorities to play a broad role in promoting the social, economic, environmental, and cultural well-being of their communities, taking a sustainable development approach.”
2. Insist that all local authorities adopt an easy to understand “most votes win” system, otherwise known as First Past the Post.
If it’s good enough for all our local electorate races in the general election to be decided this way then it’s good enough for local government too.
I sort of feel sorry for Aaron Hawkins this year. But then he shouldn’t really have become mayor seven years ago.


I just had a look through the STV link and had no trouble understanding the fairness of STV at all. This is how it works ONLY when a single candidate will be elected. Lets start with some facts:
31,629 votes were cast.
7740 first preference votes went to Hawkins. A majority (quota) is 15815 (rounded up) so Hawkins first preference votes is well short - in fact 23,889 first preference votes were NOT for Hawkins Under FPP he would have won. Hardly fair because 23,889 voters did not vote for Hawkins - he just got the most votes in a crowded field.
So now we get to rank based on 2nd preference. How many second-preference votes did Hawkins get? Down the list we go adding in the second preference. If we still dont have someone with exceeding quota (15815 or more) we go to 3rd preference and so on until one candidate exceeds quota. Then we have a winner.
Clearly another candidate garnered more 2nd preference and lower voters than Hawkins.
While the 'first-round' leader was ultimately outnumbered, the result was fair.
Note that where a voter does not rank a candidate by preference, they presumably count as zero. For example if a voter indicates only a first preference (1) and counting goes to second preference (or more) then each candidate gets a zero vote from this voter.
This may seem complicated but it is fair and not hard to audit if it works this way. Note again that this only applies where a single candidate is to be elected.
Although I understand your objection to a system that didn’t give the win to Aaron Hawkins this time, my sentiment is that it worked to save the general populace from him. Of course, if people don’t bother to vote, they can’t moan about who they end up with.