- Date published:
7:41 am, February 25th, 2020 - 62 comments
Categories: Abuse of power, auckland supercity, elections, electoral commission, electoral systems, local government, Politics - Tags: auckland, online voting, polling booth, postal voting, voting
Stuff has an article up about last years local body elections in Auckland. As usual when it comes to politics as reported by mainstream media, the interesting parts are at the end.
The target had been to achieve a turnout of 40 per cent, but the result slipped 3.2 per cent lower than in 2016 when 38.5 per cent voted.
This was despite trialling some new techniques.
The council ran a range of promotional programmes in the run-up to voting, including marketing campaigns, taking vote collection boxes to public markets and transport hubs, and running more than 50 “one stop shops” where people could enroll and vote.
The analysis report links the success of the one-stop shops to the boost in special voting, with more than 6000 special votes cast – half as much again as in 2016.
“We are attributing the jump to the fact that many people told us they had not received their voting papers through the mail,” Marguerite Delbet, the general manager of Democracy Services, told Stuff in October.
“Also we ran more than 50 ‘One Stop Shops’ where special votes were cast – we estimate we processed approximately 1700 votes at those.”
They also ran an online survey about online voting and got a predictable sampling error.
Auckland Council randomly surveyed 1,871 residents online about their involvement in the 2019 election, among which 29 percent of non-voters said they had started filling out their papers, but ran out of time.
When asked to choose between the current postal voting system, and online voting in future, 66 per cent said they would prefer to vote online, with 26 per cent sticking with postal.
Among the non-voters, 28 per cent said they would have been more likely to vote had polling booths been used, like they are in general elections.
I can understand that. It took me an hour to read the booklet and to fill out the Auckland voting. It then took me close to an hour to find out where there was a remaining postbox and to bike to it and back to send off my votes.
Voting costs me both through my rates, but more importantly through my time.
As someone who has spent the past 30 years programming for a living both on networked systems and embedded ones, I’m deeply in not favour of having online voting systems. In fact I can’t think of a programmer who is, if they stop to think past the convenience.
I don’t think that there is any way to make online voting systems robust without spending the kind of money that the banks do for their online banking systems.
I know that the local councils and even the electoral commission simply don’t have the hundreds of millions of dollars required to ensure robustness and ongoing development. The stuff article talks about Auckland City spending $1.6 million on trying to increase turnout. Kind of drop in a bucket.
They’d have to spend massively more than that each year just to maintain the code base against potential threats. Online security isn’t cheap and it is really expensive when you’re dealing with online systems that only get used every 3 years.
Plus the risks involved in subverted or aborted election are potentially enormous because of the kinds of resources that are at issue. Just the costs of determining that there has even been a problem within the voting system are immense. How do you audit a process when you can’t see what got missed? Ask the Stats department…
I’d be perfectly happy to prove this whenever a online voting system is put in place. I’ll do my absolute best to crash or subvert it, both as an individual, and in cooperation with other skilled people. It is a task worth doing just to look for vulnerabilities. Hell, it may even prove the concept.
I’d point out that any such online voting project should have an open source policy. It is the most effective way to get the online communities to probe it for vulnerabilities and to satisfy us that it is even possible. None of this crap closed source ‘commercial secret’ stuff please. As a concept that just leaves you in the position of knowing that you’ve got obvious bugs, having a monumental faith in a limited number of programmers/testers and praying a lot. The kind of thing that entrepreneurs do – which is why they go bankrupt so regularly.
But there is an alternative. A simple paper based widespread poll booth system does the same thing at a fraction of the cost.
Works for the general election. Gets a higher turnout. Has a pretty simple distributed system with lots of whistle blowers. It has and inherent audit process.
It also takes me about a quarter of the time to perform a vote in a general election than it does for the local body elections.
Plus simplifying some of those frigging election votes would help a lot. Start by demanding a upfront substantial deposit from candidates.