In the wake of the 2020 election, big issues around election transparency and election auditing have been front and center in the media. Now America Counts is inviting everyone, regardless of their politics, to participate in auditing vote reporting using our flagship app, Actual Vote.

So it’s very important that we clearly explain that Actual Vote is a trustworthy, non-partisan tool that’s next-to-impossible to abuse for partisan gain. There are seven big reasons for this:

1. Actual Vote users are simply cross-referencing publicly available information sources, an act with little inherent potential for abuse
2. A vote reporting audit with Actual Vote is designed only to turn up evidence about what happened–what then happens with that evidence is determined by the courts, not the app users
3. The Actual Vote process does not require special prerequisites from its users to ensure that it produces valid and actionable results or to prevent it from detracting from election transparency rather than contributing
4. Actual Vote is free, meaning there is no artificial barrier against using it
5. Trying to abuse a vote report auditing tool like Actual Vote is an incredibly difficult and inefficient way to gain partisan advantage
6. It’s very unlikely that Actual Vote is really part of some secret partisan agenda that America Counts is pursuing
7. Widespread use of Actual Vote, including by groups of staunch partisans, will only increase election transparency

We cover each of these reasons in detail in this essay. By the end, you’ll see why you can use Actual Vote and recommend it to others with the confidence that it will produce evidence-based, legitimate vote report auditing results.

But before we get started, let’s recall that America Counts defines “vote reporting” as the following process: 

• Election officials receive poll tapes from voting machines and/or other reporting documents containing vote data,
• They tabulate this data, and
• They officially announce the results to the public

This is distinct from vote counting, which we define as basically everything else involved in measuring the will of the people in an election. Actual Vote enables its users to audit vote reporting in US elections by video-recording poll tapes from voting machines on election night, analyzing the results, and comparing against the officially reported results to check for discrepancies. If one is found, the data can be used as evidence in a legal challenge to correct the official record. Poll tape recordings with Actual Vote recordings are publicly available for anyone to analyze. As you will see, the fact that Actual Vote’s scope is limited to errors in vote reporting, which is a relatively small and simple subspace of the larger election error opportunity space, is a big reason that Actual Vote is much less vulnerable to misuse than other election auditing techniques.

We also note that we define errors in vote reporting as encompassing both unintentional errors and intentional attacks.

We understand what it might look like—America Counts bills itself as a nonpartisan election transparency organization, but you’ve seen other organizations that make similar claims turn out to have hidden partisan agendas. And we’ve announced that Actual Vote is trustworthy, non-partisan, and next-to-impossible to abuse, and made it available to literally anyone who is able to take videos with a mobile device. Meanwhile, you’ve seen how software errors and human errors constitute such a big part of the election security vulnerability space itself, and how partisan-motivated questions about technical procedure can be such big generators of controversy around traditional audits. You’re right to demand a full explanation of why we should believe that Actual Vote, which involves both of these, is indeed trustworthy, non-partisan, and next-to-impossible to abuse. 

So let’s now recall some key properties of traditional audits for purposes of comparison. In particular, while traditional election audits are very important, they require significant money, rigor, and expertise to execute properly. Moreover, if the auditors lack the absolute highest standards of integrity, honesty, and neutrality, a traditional election audit becomes vulnerable to all sorts of errors and deliberate partisan bad intent, and is therefore susceptible to failing at its purpose of reliably double-checking the official election results. In the worst case, the activities might appear like a proper traditional audit on the surface, while in reality they amount to a bad-faith attempt by partisans to manufacture evidence in their party’s favor, conceal evidence in the other party’s favor, sow chaos, and make their opponents appear uncredible. Partisans who endorse such fraudulent activities weaken democracy, detract from the common good, and fail at their responsibility of upholding the US Constitution. Unfortunately, apart from demanding integrity, rigor, and comprehensiveness from election officials and political parties, there’s not a lot that the voting public can directly do to help improve the extent to which traditional audits increase election transparency.

America Counts strongly supports traditional election audits as a critical part of the democratic process when they meet the highest standards of integrity and rigor. But as we shall see, a vote reporting audit with Actual Vote is very different than a traditional audit–and it is these differences that explain why Actual Vote’s potential for abuse is negligible compared to that of traditional audits.

Now let’s carefully discuss the seven reasons that Actual Vote is trustworthy, non-partisan, and next-to-impossible to abuse.

1. Actual Vote users are simply cross-referencing publicly available information sources, an act with little inherent potential for abuse

A vote reporting audit with Actual Vote looks only at two sources of data, both of which are available to the public and perfectly legal to observe: the poll tapes posted on election night in a given area and the official vote total report for that area. Since the official report is generated by a series of steps involving collating and summarizing the data on poll tapes (a process that is known to be subject to error), in principle it should be totally routine and natural to use an app like Actual Vote to cross-reference the two sources. In particular, there’s no place where users, in their capacity as independent auditors:

• are somehow granted access to private ballot data containing individual votes,
• have a way of mishandling real ballots,
• interact directly with voting machines,
• deal in any way with the software used to count or report votes,
• participate in the official certification process,
• have any kind of opportunity to somehow fabricate or delete votes in the official totals,
• or face any other sensitive situation in which their integrity and neutrality might be put to the test

during the course of their activities with Actual Vote.

Instead, a vote reporting audit with Actual Vote amounts to analyzing publicly available data in a “trust but verify” spirit—the same way an investor might analyze a company’s financial statements or a tournament poker player might keep track of the behavior of other players at the table to make sure they’re following the rules. In all cases, one doesn’t necessarily suspect that errors or foul play are afoot (such as overstating revenue or marking the cards). But since the opportunity exists and the temptation is high, it’s a no-brainer to do basic comparisons of publicly-available information to make sure that they match where they should match. And though there may exist ways for participants in the system to gain an unfair advantage, the act of others auditing the publicly available information about the situation places them far enough removed from the system itself that there’s just no realistic way to use the audit itself to try to generate unfair advantage. Indeed, the literal act of auditing with publicly available information inherently has vanishingly low potential for abuse.

Of course, anyone can observe publicly posted poll tapes and cross-reference the data with officially reported results without using an app. The incremental value of Actual Vote is that it provides an easy and organized way to support this cross-referencing activity with hard video evidence that has evidentiary power in a legal challenge to the official results. 

2. A vote reporting audit with Actual Vote is designed only to turn up evidence about what happened–what then happens with that evidence is determined by the courts, not the app users

When Actual Vote users compare data from poll tape recordings to officially reported totals with Actual Vote, they produce evidence about the integrity of the vote reporting. It may turn out, for example, that a group of Actual Vote users discover that everything that’s expected to match in their analysis does indeed match, which would constitute evidence that the portion of the vote reporting that was in-scope for that audit was done correctly. They may report this outcome to America Counts and to the public and take pride in participating in democracy. In this kind of situation, the evidence does not warrant any corrective action.

On the other hand, if the analysis turns up something that doesn’t seem to match where it should match, that constitutes evidence of error in the vote reporting, and indicates that at least one candidate may be improperly benefitting from such errors. Here, corrective action is indeed warranted, but at this point, the “audit” phase is over—the evidence the independent auditors gathered must now be forwarded to an “investigation” phase to be conducted by properly qualified personnel like law enforcement, lawyers, and judges. In the same way, an investor might independently observe suspicious figures on a company’s financial statements or a tournament poker player might independently observe card marking by other players, but their role is merely observers—only a financial accounting firm or tournament referee, respectively, have the authority to actually mount an investigation based on the evidence that the auditing activities turned up. The independent auditors themselves in these types of situations don’t need to be qualified to conduct the investigation to participate in collecting evidence.

And in the case of Actual Vote, as with that of the investor and the tournament poker player, it’s impossible to guarantee that people in the investigation phase will act 100% honorably. But the possibility of a compromised investigation is no reason to not use vote report auditing with Actual Vote to increase election transparency, and certainly not a neutrality flaw in Actual Vote itself. 

The point is that the mere act of doing basic comparisons of publicly-available information in high-stakes situations (like with Actual Vote) is very hard to abuse, in the sense that calling out evidence of error, regardless of what it shows, serves only the universal goal of helping to increase transparency about the official results. And should evidence of discrepancy be uncovered in these situations, and the role of the auditor ends as investigators take over, any subsequent controversy or foul play that enters the picture cannot be due to someone abusing the audit process itself.

3. The Actual Vote process does not require special prerequisites from its users to ensure that it produces valid and actionable results or to prevent it from detracting from election transparency rather than contributing

The vote report auditing process with Actual Vote doesn’t need to assume anything about the integrity, motivation, or expertise of its users. In fact, part of the power of Actual Vote is that unlike in a traditional election audit, you don’t need special qualifications to participate in an Actual Vote audit, meaning there are no arbitrary or systematic restrictions on its potential user base that could cause bias in who is able to participate in using it. Furthermore, you don’t need to be studiously non-partisan or even of the utmost integrity—you just have to have access to a mobile device and be able to take video-recordings of poll tapes. In this sense, Actual Vote is open to just about anyone—for people who are not able to use it, the reason will stem from something other than a bias in Actual Vote’s design (such as a physical disability that prevents them from using a mobile phone). In particular, you wouldn’t expect the number of people who for some reason aren’t able to use Actual Vote to vary by political party.

Some Actual Vote users will only participate in collecting poll tape recordings, which helps to provide data for those who later do the comparison to the official results. To do this successfully, you just need to have a mobile device, be able to safely go out on election night, and be able to make legible video recordings of the poll tapes. Many adults are already able to do this.

Others will continue on to obtain the officially reported results and do the comparison, which requires a bit more effort. In particular, local knowledge, spreadsheet skills, and persistence are helpful for driving towards meaningful results, but they’re not strictly required. (In fact, some Actual Vote analyses, even by the most motivated users, still give a final result of “inconclusive” for any number of reasons.)

But you might be wondering: is there a situation in which Actual Vote can actually harm election transparency in a way directly traceable to naïveté, lack of training, or even willful ignorance on the part of its users? What bad things could happen if Actual Vote users with no bad intent simply mess up while using Actual Vote?

• They might submit illegible or irrelevant recordings on election night, slightly increasing the workload for America Counts’s reviewers who are behind the scenes approving or disapproving recordings. We are able to account for this eventuality in the way we budget time for the review process, so it is at most a minor inconvenience
• They could try to do the comparison but end up with incorrect conclusions due to human error, lack of care, or lack of thoroughness
• They might do what amounts to a superficial and error-filled analysis that misses important information about the integrity of the vote reporting and leads to a result that suggests that the vote reporting was more secure than it actually was. This is an unfortunate property of any independent auditing tool in any situation, but in this situation, the effective outcome is that the status quo stands–which on a practical level is the same as the outcome if there were no audit. As Actual Vote garners greater adoption, increased training, usage, and oversight will help to prevent or correct situations like this
• They might do sloppy work that makes it look like there’s a major discrepancy in the vote reporting, when in fact they’ve simply goofed up a few basic things, and a correct analysis would show that vote reporting under audit appears to be free from error. In this high-stakes case, America Counts will certainly double check their work and find that there is no discrepancy–all the fuss is due only to unintentional error on the part of the auditors
• They could end up making misleading statements about what Actual Vote is and does to other auditors, members of the public, and others, which, on average, will slightly detract from what authority the Actual Vote process has so far garnered until the statements can be corrected

All of these unintentional errors with Actual Vote described above have happened in the past. And this is not surprising–any crowdsourced auditing technology inherently contains the potential for unintentional error from its users, and as such, this is not a shortcoming specific to Actual Vote itself. Meanwhile, as we just showed, the potential downside of such errors is very low. And there is a positive error-correcting feedback loop from the crowdsourcing property of Actual Vote–namely, the more Actual Vote users who work in a given area, the closer we can expect to end up with a portfolio of evidence that gives the correct conclusion about the ground truth about vote reporting in that area, and the more likely that human error on the part of any one user will be caught and fixed.

The inescapable, inherent potential for unintentional error in using Actual Vote therefore does not meaningfully detract from its integrity, and is not a reason to question the validity of independent citizen-run vote reporting auditing. And to help prevent errors like these, America Counts offers free and easy training to all potential Actual Vote users that helps insure against the possibility of such errors.

This means we are able to confidently provide Actual Vote freely to the public without worrying about pitfalls from lack of training, time, funding, or integrity. This is a key way in which Actual Vote differs from traditional audits.

4. Actual Vote is free, meaning there is no artificial barrier against using it

With no cost for using Actual Vote, any person or group regardless of their financial situation can mount a vote report audit with Actual Vote of theoretically unlimited scale. If you have a mobile device, you literally have the power right now to audit vote reporting in your area without paying an additional cent.

This lack of cost barrier, combined with the previous point about no prerequisites, is how America Counts is able to proudly invite everyone, including partisans of any political stripe and integrity level, to use Actual Vote to their heart’s content. There is nothing that stops any group of people from utilizing Actual Vote and working towards increased transparency in vote reporting. In particular, with the ubiquity of smartphones, there is no barrier that systematically prevents a given group from availing itself of the benefits of Actual Vote, while other groups (especially ones with differing interests) experience no such barrier.

5. Trying to abuse a vote report auditing tool like Actual Vote is an incredibly difficult and inefficient way to gain partisan advantage

Now let’s imagine a group of crass partisans with the worst possible motivations bent on doing whatever it takes to get their candidate seated in office, including trying to abuse Actual Vote for maximum partisan gain. Could they find a way to use Actual Vote dishonestly to help their candidate?

They might already be planning to try something like altering or hacking voting machines to produce fraudulent poll tapes that give doctored vote counts. But this would be an attack on vote counting, not a misuse of Actual Vote and the vote report auditing process. So, while the potential for such attacks is indeed a cause for deep concern, it is unrelated to the integrity of Actual Vote.

Instead, to abuse Actual Vote for partisan gain, our crass partisans would basically have to do one of the following two things.

First, they could take actions resulting in Actual Vote seeming to show that the vote reporting is correct, or working to prevent any suspicion-generating Actual Vote analysis from seeing the light of day, when in fact they have successfully attacked the vote reporting in favor of their candidate (by taking actions unrelated to the integrity of Actual Vote itself), and they’re going to get away with it because they’ve neutralized Actual Vote’s vote report auditing capability.

Conversely, they could take actions resulting in an Actual Vote analysis that falsely claims that there was an error in vote reporting, which, if a court could be convinced to “correct” it, will (illicitly) benefit their candidate, and somehow win that court case without their subterfuge being caught. But note that as soon as Actual Vote evidence generates a court case that has the potential to result in the official results being restated, lawyers from the relevant campaigns will surely enlist teams of highly-trained analysts to thoroughly scrutinize the poll tape data, official results, and Actual Vote recordings (all of which are publicly available) to form a legal opinion on the ground truth of what happened. Any claimed error in vote reporting by any Actual Vote users, regardless of its truth value, will have to withstand this fierce legal onslaught, particularly from highly-skeptical defense lawyers. If the legal system is working correctly (which is unrelated to the integrity of Actual Vote), only valid evidence will hold up. It would therefore be next-to-impossible for this kind of attack to succeed.

Nevertheless, let’s speculate about what kinds of attacks in these two categories are theoretically possible for our crass partisans:

• If they planned to attack the vote reporting in a certain area (which is unrelated to the integrity of Actual Vote), and didn’t suspect Actual Vote was going to be used in this area to potentially catch the attack, there’s no incentive to try to abuse Actual Vote itself–instead, they’d just do nothing and hope to not get caught. If, on the other hand, they knew that an Actual Vote user group were doing an audit in that area (opening up the possibility that their attack will get caught), they could try to install an audit committee of their own in that area to produce an Actual Vote “analysis” that deceitfully claims that the evidence shows that the vote reporting in that area was error-free, and have it ready to go should they find out that the other Actual Vote users are preparing to release evidence that there were indeed errors in the vote reporting. Since the competing analyses are based on the same publicly available data, the “analysis” by our crass partisans would be quickly revealed as a fraud
• They could try to plant saboteurs in otherwise-honest Actual Vote user groups in different areas to try to gum up the works of vote report auditing and force Actual Vote to produce incorrect results to try to help their candidate. To succeed at this, they’d have to surreptitiously look over their co-auditors’ shoulders and try to forecast the impact of the analysis on their candidate (which is very different from the act of auditing, in which you pursue the truth no matter where it leads). Then, at such a time as they determined that the analysis was likely to harm their candidate, they’d have to suddenly try to come up with reasons to change the analysis to succeed in sabotaging it. All of this is fantastically difficult to pull off successfully without getting caught
• In a situation where their candidate stands to benefit from errors in vote reporting, they could try to flood the Actual Vote discussion space with all kinds of different purported Actual Vote analyses that are intentionally obtuse and inconsistent in a sort of distributed denial-of-service attack. The intent would be to dilute the signal content of real Actual Vote analyses, slow down the process of vote report auditing, and call into question the very validity of Actual Vote itself. Of course, since the inputs to any Actual Vote analysis are publicly available data, these kinds of fraudulent Actual Vote reports would be trivial to debunk. And since Actual Vote is crowdsourced, any sufficiently honest and motivated user group with valid, actionable evidence should be able to attract the attention of an impacted campaign with their results. So it’s next-to-impossible to benefit from this unless the number of saboteurs is large enough relative to the number of honest people that the actual, ground truth is prevented from percolating to the surface, which is very unlikely. So in this kind of a situation, it’s very likely that the truth wins out
• They might undertake a vast conspiracy in which coordinated teams of fraudsters take correct poll tapes that have already been printed in different voting locations on election night and somehow physically alter all of them to benefit their candidate–both the set of tapes to be posted and the set of tapes to be delivered to the Election Office. Of course, to do this without being detected by any poll worker at any polling place would be next-to-impossible. (If this gambit somehow succeeded against all odds, the doctored poll tapes posted at polling places will match the similarly-doctored poll tapes election officials use for vote reporting, and a valid Actual Vote analysis will report “no evidence of errors in vote reporting” while not being able to detect that there is a conspiracy afoot)
• They might try to digitally alter Actual Vote poll tape video recordings using sophisticated video editing techniques in such a way as to try to make it look like there was a vote reporting error that harmed their candidate when there really wasn’t. They could also use this technique to try to muddy the waters by raising unwarranted and distracting questions about already sound vote reporting that seems to stand to harm their candidate. Of course, all of this is very difficult and time-consuming to do in a remotely convincing way, let alone a way that could stand up to expert analysis during a legal challenge. Moreover, one comparison to any existing undoctored corresponding physical poll tape would alert everyone that there’s a major error somewhere
• They might send an agent posing as an enthusiastic vote reporting analyst to earn Democracy Counts’s trust for the purpose of gaining access to our Actual Vote analyses and surreptitiously sowing errors and inconsistencies in them. The goal would be to try to gain an illicit advantage for their party by wasting our time and money and slowing us down in our quest to find and correct errors in vote reporting. Of course, it would be very difficult to figure out how to effectively target such sabotage to gain a real payoff for their party. Moreover, this is an inherent peril of any independent auditing effort, not a partisan vulnerability unique to Actual Vote

So the fact that Actual Vote is free and open to the public does technically open up the above sorts of theoretical tactics for abuse that are characteristic of any crowdsourced auditing tool. But as you can see, on a practical level, aside from being prohibitively difficult to execute, these tactics are inefficient–the benefit our crass partisans might expect relative to the required amount of work, luck, and negligence/credulity on the part of others makes it effectively next-to-impossible to abuse Actual Vote for partisan gain.

And why would partisan bad actors even be motivated to try to exploit Actual Vote itself as part of their dishonest partisan agenda? Since their stated goal is to do whatever it takes to get their candidate seated in office, they’d naturally prefer to focus on the countless possible attacks on the election machinery itself that will benefit their candidate that are easier to execute, harder to detect (since robust election audits aren’t ubiquitous in the US), and offer a much greater expected illicit return for their candidate per unit of election attacking effort. Indeed–why bend over backwards to try to, say, forge convincing digitally doctored poll tape recordings and then use Actual Vote to convince people they’re real (for a very small expected gain), when it would be way more efficient to attack the vote reporting itself, or exploit just about any other known election security vulnerability? It’s just very unlikely.

Meanwhile, the expected benefits of increased transparency in vote reporting that we stand to gain from widespread use of Actual Vote are considerable. Just imagine teams of independent auditors around the US collecting and analyzing evidence and driving towards conclusions about the integrity of the vote reporting in their area–whatever they may be. In areas in which the analyses show no discrepancy, we have increased confidence in the integrity of the vote reporting. In areas where discrepancies show up, Actual Vote gives a real, practical pathway to forcing the official results to be corrected if warranted, improving the extent to which the election measures the will of the people, and deterring bad actors from future attacks on vote reporting. The potential harm of a partisan rogue muddying the waters with a few bogus Actual Vote analyses, or engaging in any of the other theoretically possible but low-ROI abuses of Actual Vote above, is negligible by comparison. Actual Vote was intentionally designed to have this desirable property.

6. It’s very unlikely that Actual Vote is really part of some secret partisan agenda that America Counts is pursuing

Of course each of us at America Counts has our own personal politics. Our viewpoints differ sharply on some issues. But here are some things we strongly agree on:

• The very idea of democracy itself can’t be inherently partisan
• Working to strengthen US democracy is critical for the common good and much more important than any partisan goal
• Anything as high stakes as a US election should be rigorously audited as a rule–no exceptions
• Independent citizen-run election audits are an important and under-appreciated component of democracy

So we’ve built America Counts as a non-partisan initiative, and its parent organization, Democracy Counts, is also non-partisan. That means we don’t officially endorse any particular party or candidate, and studiously avoid any political bias, or even the appearance of political bias. You can check this in our track record. In the context of elections, we are committed to improving the extent to which elections correctly measure the will of the people and standing by properly-audited election outcomes, regardless of how these outcomes may relate to any of our personal politics. By working with America Counts, then, each of us are implicitly committing to fully supporting whatever proper audits show the ground truth of election results to be, even if our honest and non-partisan efforts end up benefiting a candidate we personally oppose.

If, despite the foregoing, America Counts were in fact secretly bent on abusing Actual Vote to benefit a particular candidate (while continuing to publicly pretend that we’re non-partisan), it’s theoretically possible on election night for us to somehow try to find poll tapes recordings that might support conclusions that hurt that candidate and simply delete or “accidentally lose” those recordings before they get posted on Actual Vote’s public website. The goal would be to try to allow fraudulent vote reporting in areas that illicitly benefits our candidate to go under-audited while allowing vote reporting in other areas to be carefully audited.

Of course, with the volume of recordings coming in on election night, and the difficulty of determining which recordings we’d need to delete, it would be very hard for us to succeed at this. Furthermore, the next morning, only the Actual Vote users who recorded in the affected areas won’t be able to access their recordings, while everyone else will. It would soon become clear that America Counts must have engaged in partisan shenanigans behind the scenes, which would immediately compromise the integrity of Actual Vote, the app that we’ve worked for years to build and promote with the goal of improving election integrity. At this point, all of this effort on our part would instantly be neutralized.

Why would America Counts work hard in words and deeds to convince you we’re a non-partisan organization and release an app like Actual Vote that we’ve shown above to be trustworthy, non-partisan, and next-to-impossible to abuse… only to have our endgame actually be to try to do the next-to-impossible: eke out a minuscule illicit partisan gain through some bizarre, inefficient, easily-foiled-by-our-users attack involving the app itself? It just doesn’t make sense.

Instead, we put Actual Vote in the hands of users everywhere and empower them to independently audit vote reporting as an embodiment of our larger ideal of improving election transparency, whatever that might mean for the outcome of an election. Once the app is out there, there’s not much we can do to stop them from pursuing their own independent vote reporting audits, gathering evidence about the integrity of the vote reporting, and challenging discrepancies even if we did have a secret agenda, short of shutting down the app itself.

7. Widespread use of Actual Vote, including by groups of staunch partisans, will only increase election transparency

We’ve already discussed the positive-feedback loop property of Actual Vote–the more users, the better quality evidence we expect to collect. But what if the only Actual Vote users were staunch partisans with some modicum of integrity and no excess of ill intent? Could their strong partisan motivations when reacting to the novelty and power of Actual Vote impel them to take partisan-motivated but technically-not-illegal actions that end up causing Actual Vote to effectively decrease election transparency?

To answer this, we consider three scenarios involving political Party A and political Party B, and an election in which Candidate A from Party A, and Candidate B from Party B, are running.

For the first scenario, imagine that Candidate A just won State Z in the latest election, carrying critical but narrow victories Counties X and Y as they have been doing for the last twenty years. This year, however, a group of independent Actual Vote users mounts an organized and comprehensive audit of the vote reporting in Counties X and Y, and discovers major discrepancies that constitute evidence that Candidate A might have actually lost to Candidate B had the vote reporting been error-free. Such a scenario is not at all unrealistic to imagine.

A subsequent court case definitively shows that the users’ analysis that vote reporting error did take place in Counties X and Y was completely valid, and that when it’s corrected, Candidate B is the true winner of State Z. (The error here could be unintentional human error, or an attack by a group of bad actors with no explicit connection to Party A–it doesn’t matter for this example). The court case results in Candidate B being declared the winner of State Z, not Candidate A as originally reported. In other words, Actual Vote is working exactly as intended here.

It’s understandable if general members of Party A feel surprise, shock, and anger in a scenario like this.

But you can guess where this is going, right? We’ve constructed the “waaah, no fair” Actual Vote scenario, where certain prominent members of Party A are now prepared to represent some combination of the following sickening ideas:

• The notion that vote reporting in the status quo (in which we win State Z every four years) is subject to errors, and these errors are the only reason we always win State Z, would hurt and embarrass us, so we’re going to feel free to treat it as a vile fabrication by a rubbish app regardless of any other considerations
• “So-called” election integrity and audits are outrageous partisan charades when they hurt us electorally, but we reserve the right to insist that everyone scrupulously follow the rules in situations when doing that would hurt or delegitimize our opponents electorally
• Who can understand all this mumbo-jumbo about poll tapes and precinct-level totals from… what was the app called again? Never mind, doesn’t matter–the simple truth is that we won, then Party B started whining about some BS app, and now everyone has lost their minds. It’s time to move on
• There has to be some implicit equivalence principle that says that that if our fortunes are left worse off by the court case, then Party B by definition has to also have their fortunes decremented by an equivalent amount or else it’s not fair
• Party B are the ones who are really benefiting from election insecurities–look, they’re about to succeed in a brazen partisan power grab and steal the victory in State Z from us based on random videos that random people took on their phones

Part of the goal here is to deceive people (especially angry Party A partisans) into believing that there’s no way the act of using Actual Vote to audit vote reporting could possibly be legitimate if it stands to rob them of their victory (regardless of the means by which it was determined). Of course, no matter how angry Party A members might be about the situation, the fact remains that the Actual Vote analysis that uncovered the error in this situation simply can’t be partisan—because in a democracy, errors in vote reporting (regardless of who they might benefit) should in principle be caught by audits and corrected. And while some angry Party A partisans might be susceptible to this deception at this moment, they must understand that Actual Vote users have only collated publicly available data sources, compared them, and communicated their conclusions–exercising their right to observe poll tapes and work together to improve election transparency. They haven’t been given an opportunity to do delicate and error-prone things like mishandle actual ballots or state official vote report totals.

In a real-life situation in which the official winner of an election changes when errors in vote reporting are corrected, it’s all but inevitable that we’ll observe some version of the “waaah, no fair” scenario. That’s why it’s critical for us to emphasize here that the Actual Vote audit in question cannot in principle be a partisan power grab, no matter what the prominent Party A members from above would like us to believe. Like it or not, residents of a democracy are obligated to accept the results of valid election audits, whatever they may be. In an ideal world, Party A would even be able to join Party B and the nation in celebrating a philosophical victory for election transparency and a more accurate measure of the will of the voters, and that everyone has successfully represented that the same standards of honesty and integrity apply to all parties. Indeed, if a political party is really dedicated to doing their job of serving the people and upholding the Constitution, they’d welcome arbitrarily rigorous and extensive audit activity as a crucial pillar of democracy.

The second scenario is the “what if we get caught but they don’t” scenario. Keep the setup from the first scenario above, but now imagine further that a rogue group of Party B members in a different area succeeds in attacking the vote reporting in that area, resulting in Candidate B winning there when they otherwise would have lost. There’s no audit of the vote reporting in that area (like with Actual Vote, for example), and so the foul play and incorrect result goes undetected. This scheme by the rogue group of Party B members is indeed a partisan attack on the election. But the fact that it will go undetected, and that this part of the official vote reporting will go uncorrected, is not due to partisan bias in Actual Vote. It’s due to the fact that the vote reporting in this area wasn’t sufficiently audited (as it isn’t in most areas in the US). But as we’ve seen, one of the things that makes Actual Vote trustworthy and non-partisan is that it’s open for anyone to use with no artificial cost barrier. In this situation, we warmly invite Party A and Party B alike to use Actual Vote to conduct vote reporting audits in this area—if there is any kind of error or fraud in the vote reporting there, democracy as a whole benefits from it being detected and corrected. America Counts can only provide an election-transparency-increasing tool like Actual Vote. But the fact that we can’t guarantee that Party A and Party B will both avail themselves equally of its power, or that there will be any sort of party-wise symmetry in the impact of the results on each party, is not a partisan flaw in Actual Vote. It’s entirely a property of the extent to which the tool is used (which is bounded only by level of intent, not by cost or training).

The third scenario might be called the “friendly adversarial relationship” scenario, and it describes a best-case side-effect from the subset of partisan motivations that have good reasons behind them. It’s similar to the software development practices of NASA’s on-board shuttle group, which is responsible for writing software for the space shuttle. They are able to produce extraordinarily error-free and robust code in part because they have two teams working on it: coders who write the code and the verifiers who look for flaws in it. These two teams have a friendly adversarial relationship in which each is incentivized to do their specific job to the highest standard because the better the other team does at their job compared to you, the worse your performance rating is. This two-way incentive system is one big reason why the on-board shuttle group is able to write some of the world’s best software.

Now consider that in the US, the Democratic and Republican parties both have an equal duty to uphold the Constitution. And while each have different values and policy goals, they have some roughly equivalent degree of understandable skepticism of the other’s motives and honesty when it comes to conducting free and fair elections. With an independent vote report auditing tool like Actual Vote, they can act on this skepticism by conducting vote reporting audits specifically designed to check that their opponents aren’t benefiting from errors in vote reporting. The fact that their opponents can audit them in the same way will only increase their motivation to audit thoroughly. This kind of friendly, adversarial, and escalating use of Actual Vote by the two major parties would have the welcome side effect of greatly improved overall transparency in vote reporting despite the fact that it nominally stems from partisan-motivated skepticism. This, ultimately, would be a big win for democracy in the US, and therefore not a partisan vulnerability in Actual Vote. In the same way, the competing interests of the coders and verifiers of the on-board shuttle group result in them achieving their broader shared goal of producing outstanding software.

Conclusion

For these seven reasons, you may rest assured that Actual Vote is indeed a trustworthy, non-partisan tool that’s next-to-impossible to abuse for partisan gain, and all may advocate for its widespread adoption without fear of misuse or sabotage.

At the end of the day, the only people who have anything to fear from the widespread use of Actual Vote are those who wish to weaken democracy by allowing errors in vote reporting to go undetected.