So here’s something a little different for you all. I wanted to write something about the current hullabaloo in the U.S. over alleged fraud in the election and the way that many have responded to the allegations. It’s not a particularly Anglican or even Christian post (although having finished it I think it does shine a little light onto some of the reason I write some of the pieces here on davidould.net
Sit back, get your favourite drink, and I’ll try to take you on an interesting journey. I want this to be the most exciting thing you’ve ever received from an (ex) accountant.
How, as a trained auditor, would I check or assess the electoral process?
I need to point out I’m not a specialist in election oversight, just a trained auditor in systems/processes more generally. So I’m applying from one area into another similar area. There could be some reading along who are specialists in this field and I welcome their input and correction.
Let’s treat the election process as a flow of information. We start somewhere (an electoral roll) and end up at the end with a final position (an accurate count of votes). Systems analysis assesses both the plan of the system itself and then its implementation. Systems work best when there are clear controls upon them; checks and processes that ensure information flows through accurately. Examples would be requiring a signature or having a third party oversee an activity.
As an auditor I’d want to check the following:
- Does the system make sense in principle (i.e. designed to move information accurately, all controls in place)?
- Were the controls actually used correctly?
- Did information actually flow correctly through the system? This then breaks into two main questions:
- Did any information at the start of the system carry through all the way to the end as it should? (i.e. is there a clear trace all the way through from electoral vote to final vote being counted accurately?) – this is a check for undercounting of votes
- Was every final result a genuine one? (i.e. can I trace it all the way back through the system from a final vote to someone on the electoral roll?) – this is a check for overcounting of votes.
- Finally, does the final result actually make sense? Does it fall within the limits of what I might expect to see and, if not, is there an explanation for that difference in expectation?
So those are the basic principles.
Applying the Principles
So let’s begin to apply those principles to elections:
1. I don’t see any suggestion that the systems being set up in various states are themselves flawed. This is a separate argument to one about whether (for example) late ballots ought to be received. As an auditor I don’t care about that, my question is whether whatever rule has been decided upon has been properly implemented. So, if state X decides it will accept ballots for the election as long as they are postmarked NO LATER THAN 6 DAYS AFTER THE ELECTION DAY (to pick a silly example) – are they actually applying that rule correctly?
It follows that one of the major claims being made is that these rules are not being properly applied. In some places, the claim is, late ballots are being accepted CONTRARY to the rules that were set out. Further, it is claimed that officials were then changing the appearance of those ballots to appear legitimate or even simply ignoring the rules. This would then be an example of…
2. Are the controls being properly applied?
By my reckoning this is where the majority of the complaints are coming. It is now beyond question that scrutineers were kept from observing the processing of some mail-in ballots, so much so that a court had to order that they were let back in to see (and even then there is documented intransigence from some officials).
So take the silly six-day example above. It’s really important that when late postal ballots come in the chronological cut-off (whatever it may be) is strictly applied. But if a scrutineer is unable to stand within a reasonable distance AND be given the opportunity to reasonably ask to see a sample of ballots of their own choice and timing then there is no check upon the control.
The control (someone checking the date of the ballot) is legitimised by the presence of the scrutineer. Remove the scrutineer and there is now no confidence that the control is being applied correctly. No wonder there was uproar! Especially when it looks like the “other side” are getting really good results from the mail-in ballots that they won’t let my side scrutinise. Any fair-minded person would surely concede this.
How Dare You Impugn My Integrity?
However, this raises another question. Does the very asking of the question not itself cast an unnecessary accusation about the integrity of the process and the officials? In the first few days I saw a number of people arguing to the effect of “stop asking all these unnecessary questions and let them just get on with the job”. Is that a helpful approach?
Not at all. Auditors will operate under the principle of “healthy scepticism”. i.e. we assume that the process COULD be broken and then put ourselves in a position to be assured that it is not. The auditor is not there so much as to sniff out the secret Swiss bank accounts as to affirm and validate the company’s entirely reasonable claim that there are no Swiss bank accounts.
How can I claim that I have no secret side-hustle in Panama that I’m syphoning money off into? Well, I invited the auditors in to make it clear.
In the same way, the healthy scepticism of the scrutineer provides the assurance that is required by both scrutineer AND officer that the process is working properly. The scrutineer does not so much challenge the probity of the officer but, actually, is the best way of demonstrating it conclusively.
But that means the converse is also true. The deliberate exclusion of the scrutineer is therefore the worst thing to do for making the process look dodgy. To my mind, BOTH sides of the election ought to have been outraged by the exclusion of scrutineers, especially the Democrats, because it utterly undermined the apparent probity of what was already widely acknowledged was going to be the most controversial and disputed part of the whole process.
This is also why the complaint in Detroit that there were no effective controls in the middle of the night on boxes going into the counting room was also very important. We have seen footage of someone unload some boxes onto a trolley and then wheel the trolley all the way into the count room PAST a security guard who didn’t even check credentials, let alone the contents of the box.
The answer, we were told, is that it was a camera man wheeling in his gear. Case closed, leave it alone.
Except that doesn’t help at all. Because remember the healthy scepticism. Assume the process is broken and look for clear evidence to establish that it’s not. In this case assume that something dodgy (presumably boxes of false ballots) are being wheeled in under cover of darkness. We’ve been given what appears to be a reasonable explanation but no actual evidence. The important thing here is that under the necessary healthy scepticism the burden of proof shifts. In order for the system to have integrity it must be able to be held up to scrutiny and pass the test.
The issue here is that it was held up to scrutiny – somebody filmed the box being wheeled all the way in. At the moment that it clearly crosses into the secure zone without being checked we have a serious issue irrespective of what we ultimately discover is in the box.
I can’t get a frozen pizza out of Costco without it being checked at the door against my receipt but this looks like a mass delivery from Dominos.
So now we have a clearly documented breakdown of the controls. It could be a box of doughnuts, an inflatable castle for break time, a camera or a set of fake ballots. There is no way of knowing. The control was set up specifically to assure us that the area was secure and it has apparently been shown to have failed.
Again, my contention would be that both sides of the political aisle ought to be concerned at this.
But it was just one incident…
Nor, as an auditor, would I be satisfied at this point with an explanation of the actual box that was filmed. That’s not the point. More than likely it was camera equipment and the two people who testified (I’m just speculating at this point) that they saw it being taken out of the box assure me of that (and all the more if those testifying are Republicans, not Democrats – for what I hope are obvious reasons). But that’s not the point now.
Remember, I’m concerned about the system, not just the incident.
I now have a much bigger question in my mind. I no longer care about the box that I’ve seen so much as ALL THE OTHER BOXES THAT I MIGHT NOT HAVE SEEN! And that really worries me because I know one thing for sure – THE CONTROL WASN’T BEING APPLIED!
Again, I would make an appeal – can we not agree across the aisle that this is a serious issue? It contributes to undermining confidence in the integrity of the process.
So what WILL satisfy you?
At this point I want to take a step back and talk about the sort of evidence that we do or don’t rely upon. Auditors of financial systems will do less and less of the flow-through testing outlined in 3 above if they’re satisfied with the system itself and especially it’s controls. This is, I would think, important in elections because it’s pretty hard to test one piece of paper all the way through.
We prize a secret ballot. So we actually make it very very hard to know who voted for whom and we deliberately try and block that chain of evidence in order to protect the principle of secrecy. Of course, if you remove this direct and obvious trail (as you need to to make the ballot properly secret) then you compensate by making the controls around the flow even better. Want high confidence in a secret ballot? Make the controls really really good. Want everyone to think the controls are really really good? Let them be thoroughly scrutinised as they’re being used.
Also, what sort of evidence do we prefer? Well, we have a clear hierarchy.
The least reliable is internally-generated evidence. Let’s say I’m testing my child’s maths class. How do I know whether they’ve learned enough?
I could just ask the child. Have you got this right? They say to me “yes”. Well all good. On we go.
Except of course, it rather defeats the purpose and begs the question. It’s obviously more complex in our election system. We have internal tally sheets, we have computer programs and reports and so on. These are all helpful but they’re still all internally-generated.
So I go up a tier. The next in the hierarchy is third-party evidence. Did my child get it right? They present for me a certificate from the exam board. This is far more reliable for me (although the reliability will still have to be assessed – is it the HSC board? Is it “Smith’s Maths Assessment Corporation”?)
It’s the difference between accepting a company’s own cashbook as their record of how much money is in the bank and looking at the bank statement.
A good example of third-party evidence in the electoral flow is a voter’s signature on a ballot or attendance record. It could be the post office stamp on the envelope. It comes from outside the system and the auditor can therefore place more reliance upon it to assess the system itself.
But it’s not the best form of evidence.
The best evidence is auditor-generated evidence. It’s where the auditor themselves do the work to gather the evidence. Think of the auditor doing the stock take themselves (or at least doing part of it to satisfy themselves that the count is correct). It’s the auditor doing their own actuarial calculations to value a superannuation or insurance fund’s liabilities.
And it’s the scrutineer who watches over the application of all the controls in the electoral process. Their presence is the very best evidence. And the higher the stakes, the more critical it is that everything be seen to be functioning as it should.
The best scrutiny? Get a Democrat and a Republican together as a single scrutineering unit. One of the basic controls against fraud itself is to have more than one person doing a job. Even better when the incentive to collude in deception is zero. So the best scrutineering happens not when we have independent scrutineers but actually when they are from competing political parties. That way the scrutineers scrutinise each other’s scrutiny!
The worst scrutiny? Exempting one of the parties.
Putting it all together
At this stage I hope a few things are evident.
- External scrutiny is absolutely vital for the demonstrable probity of an election process. It has long been our practice in elections all over the world to not only allow but to encourage scrutiny. Proper scrutiny provides the “auditor-generated” evidence that assures everyone things are working as they should.
- Controls exist for a purpose – they also demonstrate that things are working as they should and that the final result can be trusted. When a control appears to break down we have some serious questions that have to be asked BEYOND THE INDIVIDUAL DOCUMENTED INSTANCES OF BREAKDOWN – this is really important for some of the questions being asked about the U.S. election. It simply won’t do to say “well we answered that particular instance” when the event has raised much wider concerns.
- Healthy scepticism is vital. Every time someone questions the process and gets the right answer it actually increases our assurance and confidence in that process. Unless we ask “are these officials acting in the right way?” we will be unable to get an answer and then affirm “it really does look like they did” and the goal of all this scrutiny is to say with as much confidence as possible “it really does look like they did”.
Some Further Observations
Some of the complaints raised over the last week or so have arisen out of a lack of understanding of the systems.
One example will suffice: video circulated of someone writing out ballots inside the counting room. Looks dodgy? But what was going on was a classic example of re-writing a ballot that a counting machine couldn’t read properly. Someone else was at the table checking all the work and a security guard passes by. Now technically a scrutineer hovering nearby would add to our confidence but the process being filmed was actually legitimate. I personally circulated this video asking “what’s going on here?” (Or something similar) and I got a pretty prompt answer.
Of course, just because a process is legitimate, doesn’t mean it was followed legitimately. But that’s why we have the controls and the scrutiny of those controls.
We must also concede that there has been a fair degree of people rejecting complaints or defences simply because of their source. It came from a Trump supporter? Well then it must be nonsense! It’s a Trump lawyer? Well then it’s just a delaying scam.
The flip side is also true. “Pennsylvania and Illinois Democrats have a history of cheating in elections”. Well they’re obviously doing it again.
As a logical argument it just doesn’t hold. Of course, it’s compounded by the fact that, yes, Trump and his supporters can blow louder than an elephant with hayfever and, yes, unfortunately PA and IL Democrats do appear to have a bit of a history with electoral irregularity. But that in itself doesn’t prove anything. What it really does is further increase the auditor’s healthy scepticism of whether they can trust the system to be working well or whether the complaint has legs. But it doesn’t dismiss it.
To dismiss a complaint out of hand because of its source OR to be certain that the election must be rigged because of where it is being held – well that’s just unreasonable.
Where to from here?
Our electoral processes thrive on being transparent. That transparency comes from well-planned and executed systems that have the proper controls that are themselves open to full scrutiny. Where any of those parts don’t work we have a serious breakdown in confidence.
At the time of writing there are some outstanding questions about the application of controls in the U.S. elections and their proper scrutiny that deserve a full and proper answer alongside a number of (let’s put it generously) more misguided complaints.
Who should be wanting those answers to be fully and clearly given through the right system (the courts)? Well the Republicans certainly want to have answers. But two other parties ought to be getting on board too: the Democrats and the Election Officers. And the rest of us too.
Ultimately when we allow these challenges to be properly handled and followed through to their conclusion it underwrites the public confidence in the process, rather than undermining it.
I couldn’t believe some of the things that I was seeing over this past week. Scrutineers pushed away from being able to watch a count? No controls over material being introduced into secure spaces? But what astounded me even more was the rush to seek to brush aside these complaints as all being simply politically motivated or even just deliberately vexatious.
In the same way, the haste with which some would accuse officials of corruption has far exceeded what would be merited by a simple healthy scepticism.
Asking and Answering the Right Questions in the Right Way
The bottom line for me appears to be this:
Asking the question is not the same as a full accusation. At the same time the blanket rejection of a legitimate question sure doesn’t look good.
When I was auditing I saw this all the time. We would regularly find small things that had gone wrong. So we’d ask a question of clarification. What happened here? Was it evidence of a more systemic problem or an explainable one-off? Depending upon those first answers we’d then ask ourselves some more: What would be the consequence anyway of this control failing? Ultimately what difference would it make overall to the opinion we offered at the end of the whole process?
The response to those first questions would often depend upon two related factors. Did the person in the process who was being questioned really understand what we were doing and why? Did they ultimately comprehend that the question was designed to increase confidence in their probity, not attack it? And did they understand the overall intent and purpose of the audit as a whole – to allow the scrutineer to stand back at the end of the day and say “yes, what is being reported by this system does really look like it’s true and fair”? Did they understand that the questions were a necessary, even vital part of the whole thing?
Underlying that is a relational trust of the auditor. So there’s a tension there – how long can you keep auditing the same company? Not too long that you no longer look independent, but long enough to build up experience and relational trust.
A Perfect Storm of Ignorance and Bad Faith
And these two last factors are why, to my mind, it’s all gone horribly wrong in the U.S. Those complaining and especially responding often had no real idea of what the actual arguments and motivations were. And there was also plain bad faith.
Some of the complaints were ill-founded. Some of the rejection of the complaints showed no awareness of how proper checks and balances in the system are meant to work and how they can fail.
And fuelling all of this ignorance was what is now widely acknowledged as the most antagonistically polarised political environment we’ve ever experienced.
It is, perhaps, one of those moments where we all need to step back and ask how we’ve fallen into one of those traps. As we think about the questions that have been asked we need to ask ourself some too.
- Have I dismissed a complaint out of hand because of who it came from?
- Have I assumed the veracity of a complaint because of who it was about?
- Have I actually not understood the potential implications of a single apparent breach upon confidence in a whole system?
- Have I claimed that one single breach must indicate the whole system is broken?
I’m sure there’s more but I trust you get the point.
So What Now?
I do wish Trump wouldn’t claim so loudly that the election must have been stolen from him. It makes him look like he’s not really handling the evidence and claims properly.
I do wish Biden wouldn’t have claimed that the election result was now beyond doubt. It makes him look like he’s hurriedly dismissing legitimate questions that still need to be answered.
Both approaches are actually unhelpful. What is needed right now, founded in a proper understanding of how checks and balances function to uphold a perception integrity in the electoral system, is the patience to see the challenges through. If both sides value that perception of integrity then both sides ought to be defending the whole process.
That’s it from me. I’m sure not everything above is perfect but I trust it might cause some of us to think again about the whole thing while we wait for God’s instituted authority (Romans 13:1) to be beyond doubt and the good grace to accept that final decision or ruling, whatever it might be and because the whole thing had demonstrably proper and full oversight.