The Wrongful Deregistration of Flux
On March 24th 2022, Flux was wrongfully and involuntarily deregistered by the AEC . This is in spite of the fact that we were (and remain) an eligible political party. Unfortunately, this was not unexpected.
Flux’s deregistration was wrong and unjust. This is a direct result of recent changes to legislation that targets minor parties. One particular change forced the AEC to repeat the membership test for all parties, but at a larger scale. The membership test is flawed , though. Even if we could find a way to keep Flux alive, we should ask ourselves: is this a battle worth fighting?
Even if we succeed in overturning this decision (which is possible), hurdles like this will keep coming, and keep coming, and keep coming. If we got big enough to be a threat, then new hurdles would be invented – for example: even more legislative changes around funding, disclosure, deposits, and refunds (and we are already at a significant financial disadvantage). Parliament has changed the rules on us more than once, and they’ll do it again. It’s not a fight that we can win – at least, not in an enduring way; not every time.
The tagline on our website reads: “It’s time to fix Democracy.” Turns out, we can’t. Our rulers could, but why would they do that?
Flux was founded as a movement based on a set of principles and ideas. Registering a political party was a vehicle to bring Issue-Based Direct Democracy & Digital Democracy to a wide audience during elections – the times people were most interested in politics. The intention was to build a system to create better legislative outcomes in parliaments. Over 8,000 members have joined, we’ve contested 5 elections at both federal and state levels, and released an app which connected people to their federal parliament – to review legislation and vote on bills. This, and so much more, was accomplished by volunteers who gave their time & energy in service of better ideas.
So. This is it, then.
To all of you: thank you for being a part of Flux. To our volunteers: thank you for your time, your energy, and your passion. To our candidates: thank you for putting yourself out there on our behalf. And, of course, thank you to our voters.
This email covers a few things:
- A summary of the wrongful deregistration
- What this means for Flux
- Party matters, esp. regarding membership and personal information
- What to do if you want to do something
- Links to supporting documents
You can expect one, or maybe two more emails about wrapping things up. The next will probably be in July or August.
Till, then, thanks again to you all for giving this a go. It wasn’t a fair shake, but it was a shake at least.
– Daithí & Max
# Summary of Wrongful Deregistration
The short version is that the system is rigged and the outcomes of AEC membership tests are often predetermined. How are such tests predetermined? Methodological bias – parties with more than 1650 are at an unfair disadvantage, and in some real-world cases an eligible party has a 0% chance of success.
In order to assess Flux’s membership list for eligibility, the AEC’s audit process insists on using a small subset of party members (1650), as opposed to reviewing a full list of all of our pre-validated members (4680).
Using these smaller sample sets introduces a major error into the process which is never corrected. The flawed outcome informs the decision of the AEC. For the AEC to consider that a party meets the legislative requirements, it must pass this test.
We have provided the AEC with our full membership list which vastly exceeds the legislative requirement, but the AEC rejects the list as it does not “conform with the AEC’s membership testing parameters” .
The AEC ignores the fact that their process incorrectly fails for parties with membership that exceeds the AEC’s maximum list size of 1650.
The AEC has ruled that Flux is no longer a registered political party.
That’s the concise overview. The full context is more complex.
The membership test involves: the party submits a membership list, and then after some administration, members on that list are randomly sampled, contacted, and asked to confirm or deny their membership. If someone says “no”, this counts as a membership denial. If too many people deny membership then the party fails the test. On first glance, this sounds reasonable – and it could be done fairly. But it’s not done fairly.
This test artificially restricts the number of members a party can submit to 1650 – just 10% more than the minimum required (1500). It gets worse: not only are some members excluded without replacement, but at most a denial rate of 15% is tolerated (regardless of how many members the party actually has). Practically speaking, this heavily biases high quality membership lists and younger parties. Statistically speaking, this invalidates the test (since the reduction in sample size is never accounted for).
One consequence of this is that, if a party gains members, it can get harder to prove sufficient membership. That’s a serious issue, but the AEC ignores it.
Despite this fairly obvious methodological flaw, there is no explanation or evidence that the AEC will consider. In all cases, they claim the test is fine because it was developed by the ABS. The commission will literally ignore statistical arguments – we provided 3 different decisive criticisms of the method in our response  to the AEC, and they were all ignored. The AEC remains steadfast in its irrationality. They don’t care about facts.
If you want to judge the AEC’s nonsense for yourself, you can read the AEC Notice of Deregistration  which includes results for our second membership test. (Note that we didn’t even request another membership test – what’s the point if the method is broken?)
If you are wondering: with Flux’s current membership (4680 verified members), the accuracy of the AEC’s method for Flux’s first test was 28% – that means it has a false negative rate of 72%. The chance of Flux surviving the AEC’s process, given two tests with the same list, was optimistically around 48%; worse than a coin flip. The AEC considers this “rational, fair and practical in all the circumstances” . These numbers come from a statistical analysis of the AEC’s method  that proves that the AEC’s method is broken (the paper has numerous additional criticisms).
The accuracy of Flux’s second membership test was 0.002% ± 0.0012%. Statistically, we expect that it results in a false negative 99.9980% ± 0.0012% of the time. “[Rational], fair and practical in all the circumstances” indeed.
This is not the first time that we’ve been mistreated. If we remain registered, it would not be the last.
# What this Means for Flux
Flux is going to wind up. This was the plan prior to the wrongful deregistration, too. If we happened to pass the second membership test that the AEC performed, we would have voluntarily deregistered. If Flux was going to end anyway, we wanted it on our own terms.
The truth is that this game (elections, democracy) is rigged against us in more ways that we know. The other obvious ways are financial: federal elections are profitable for major parties (combined they receive more than $50M each election); and they can run candidates at zero cost – whereas, for us, it’s $2000 per candidate. This isn’t to say that we would have won any elections if those issues had been fixed. We weren’t good enough, and if we ever came close then new rules would be created to stop us.
If the system is rigged, what would we be saying if we continue to play by their rules? We’d be saying that it’s still worth it and that the rules are fair enough to compete by. We’d be the victims and we’d be implicitly giving them our ‘okay’. It’s not a good system to participate in and it isn’t a good way to live.
# Party Matters
Over the next few months or so, we’ll shut down systems and wind things up. The membership database will be deleted along with backups. The website will remain up in an archived state. Any other infrastructure, like the backend software, will be published on our github and licensed under an open source license (probably MIT).
State parties (NSW and WA) make things a little more complex, we’re still figuring out exactly what to do there.
# Things you could do – and whether you should
This decision by the AEC is wrong and unjust, and there are things you can do before it’s “too late”. Should you, though?
What does it mean to spend your time and energy fighting government injustice? Doing that is hard, tiring, frustrating, and possibly (probably) futile. What will change? If the AEC refuses to change, what then? Why would the AEC even want to change?
We can’t win – so why fight this battle? It’s just a tax on your life. It’s an extra cost that you have to endure because you’re trying to fix (or at least band-aid) a broken system. That’s what participating in our political system looks like – participation outside tribal politics, that is.
If you want to change things in the current paradigm: sell out, get friends in high places, play the game, acquire power, and so on. Then, when you finally achieve enough power to actually change things, ask yourself Why aren’t I? That’s what always happens – people say they want to change things, get elected, and then they don’t. That’s why Flux existed. That’s also why Flux won’t succeed.
In regard to fighting the AEC on this matter, here’s the advice: don’t bother.
Max… had trouble taking his own advice here, and has applied to the Commission to review the deregistration  – it’s 34 pages total. Will it succeed? We’ve read a lot of AEC responses to allegations that the membership testing method is bad, and they have always failed. So probably not.
# What are the options for doing something?
# Apply for a review under s141(2) of the Electoral Act
Even now, the process is not complete. From the AEC’s statement of reasons:
33. Under s 141(2) of the Electoral Act, a person (including an organisation) affected by the decision who is dissatisfied with the decision may make a written application to the Electoral Commission for internal review of this decision within 28 days after the day on which the decision first comes to the notice of that person. There is no fee payable for requesting an internal review.
34. Requests for review of this decision should be addressed to Mr Tom Rogers, Australian Electoral Commissioner, and emailed to firstname.lastname@example.org or posted to Locked Bag 4007, Canberra City ACT 2601.
35. In accordance with ss 141(2) and 141(3) of the Electoral Act, an application for review must:
• be in writing;
• specify the name of the applicant;
• specify an address of the applicant; and
• set out the reasons for making the application.
36. If you wish to apply for additional time beyond the 28 days to make an application for review of the delegate’s decision, please also include the reasons for the application for additional time.
If this is the first that you know of the AEC’s decision to deregister Flux, then you have 28 days to make an application to the Commission (you’ll need to explain that you only learned of the deregistration on this date unless you submit on/before April 21st). Everyone in Australia is affected because if our Electoral Commission is irrational then the country’s democracy suffers, so anyone can make a submission.
Note that these reviews – handled by the Commissioner himself – involved 3 people. One of those is the statistician at the head of the ABS. Would he act as the Commissioner’s lacky and sign off on their standard excuse? Or does he perhaps have enough integrity to listen to a statistical argument; such as  and ?
If you apply under s141(2), then you might also want to apply for more time to submit an application for review, so that you might e.g., lodge FOI requests with the AEC (though only do this if you actually are lodging FOI requests).
# Lodge FOI Requests
 has a list of FOI topics that might help. I expect the AEC will put up a fight, e.g., by claiming that a request is not clear enough. Additionally, FOI requests may come with some charge, but that’s case-by-case. I don’t know how much that charge would be.
# If you know a statistician
If you know a statistician, you could ask them to review  and . If those seem reasonable, then a short application under s141(2) might help (though it would rest on that person’s claim to be authoritative).
To be honest, the AEC would probably disregard this unless it comes from a professional statistician. They might disregard Max’s submission for this reason, regardless of the content.
# Supporting Links / Citations
: AEC Notice of Intention to Deregister (2022-01-13) https://xertrov.github.io/aec-membership-test-simulator/docs/BDLPN0-aec-flux-jan-13_unencrypted.pdf
: Flux’s Response to the AEC (2022-02-13) https://xertrov.github.io/aec-membership-test-simulator/docs/Response-to-AEC-rego-20220213.pdf
: “AEC Party Membership Test Methodology is Rigged! A Statistical Analysis of AEC Methodology and Graphs (of PMFs)” https://xertrov.github.io/aec-membership-test-simulator/
: AEC Statement of Reasons WRT Flux Deregistration (2022-03-24) https://web.archive.org/web/20220326093243/https://www.aec.gov.au/Parties_and_Representatives/Party_Registration/Deregistered_parties/files/statement-of-reasons-voteflux-org-upgrade-democracy-s137-deregistration.pdf
: AEC Review of Decision to Deregister SUPA (2021-12-15) – The quote is from point 8 on page 3 https://web.archive.org/web/20220320030110/https://www.aec.gov.au/Parties_and_Representatives/Party_Registration/Registration_Decisions/2021/notice-of-decision-with-reasons-SUPA.pdf
: Max’s submission to the Commission to review : https://xertrov.github.io/aec-membership-test-simulator/request-review-flux-deregistration
: Ideas for FOIs that might assist in uncovering AEC wrongdoing: https://github.com/XertroV/aec-membership-test-simulator/blob/b076029bd6e34f474dfdbffcad5e6bf0485d04a0/README.md?plain=1#L94
Comments powered by Talkyard.