I’m Isaac Saul, and this is Tangle: an independent, ad-free, subscriber-supported politics newsletter that summarizes the best arguments from across the political spectrum on the news of the day — then “my take.” Today is a special Friday edition.

First time reading? Sign up here.


Today, I'm writing about the latest stolen election theory, one that is the centerpiece of Dinesh D'Souza's new movie 2,000 Mules. The film is based on the premise that a team of researchers and a massive trove of purchased cell phone data has uncovered that at least 2,000 people were employed across a half dozen swing states to collect and deposit hundreds of thousands of ballots for Joe Biden. Those ballots then tipped the election in his favor.

Let me start with a little bit of throat-clearing.

I consider myself to be both a skeptic and an open-minded person. I am deeply cynical about our government, believe intelligence agencies are covering up the truth about UFOs (yes, I'm still trying to find time for that newsletter), and don't feel any particular loyalty to the major political parties. I generally distrust authority, government agencies, and politicians, but I do believe it's wise to consult expert opinions and advice. I love a good conspiracy, a good cover-up, and a great story, and I’ve learned over the years that identifying motives is often the best way to accurately suss out what is going on.

A stolen presidential election would be an all-timer in every regard, a story so gigantic — a conspiracy of corruption and power so unthinkable — that the idea alone is tantalizing enough I almost want to believe it (aside from, you know, the fact that it'd destroy our democracy).

I'm also a reporter. And a very active and engaged U.S. citizen. The combination of those two things means I've learned a lot about elections, how they work, how we vote, and what goes into making sure our votes are counted accurately.

In 2020, that put me in a unique position. I was a fresh voice on the scene, building trust with both conservatives and liberals, skeptical of the mainstream media (I literally started my own news organization to combat their biases), and had cleared my schedule to cover the before and after portions of the election.

I was ready, and there were signs of what was coming. Despite President Donald Trump defeating Hillary Clinton in 2016, he claimed the election was actually rigged, and there were three to five million illegal votes. Conveniently, this was about the margin Clinton had beaten him by in the popular vote. Trump also claimed none of those three to five million illegal votes were for him — they all had been for Clinton. Heading into the 2020 election against Biden, on the debate stage and on Twitter, Trump repeatedly said the only way he would lose was if the election were rigged. In other words, Trump spent months predicting the election would be stolen before the first actual vote was cast.

All of this produced a rather logical prediction from me: I told readers heading into the 2020 election that there would be a wave of claims of election fraud. I told them that, as in every election, there would be instances of real voter and election fraud. But because so many people were calling the election rigged or fraudulent before the first vote was cast, I cautioned people to move forward with skepticism, to treat viral claims of fraud carefully, and to do their best to verify information before sharing it.

What actually happened surpassed even my wildest expectations.

Election fraud allegations didn't simply rain from the sky. There was not a firehose or a downpour. It wasn't even a Category 5 hurricane. It was Biblical. A Noah's-flood-event of allegations. I had prepared, and I thought I was ready, but I wasn't.

Still, I did my best. Many Tangle readers are here because of my Twitter thread that I kept live throughout the election where I tracked allegations of fraud and tried to investigate them. Six million people saw that thread. I thought it would be 20 or 30 tweets long. It ended up being close to 400. I slept about 15 hours over the course of a week. I did dozens of radio interviews, TV spots, and interviews with various media outlets. About half of the allegations were easy to figure out or debunk in minutes. Some took hours or a couple of days. Some remain unresolved even now.

People alleged Republican poll watchers were barred from polling places or that statistical analysis proved the election was stolen. People said that videos showed election workers filling out people's ballots or shady actors collecting ballots from ballot boxes. People thought they witnessed election workers throwing out ballots; people alleged that dead people voted; people claimed voter turnout exceeded registration; people claimed Trump won every bellwether county but still lost, and many, many more.

I've addressed all of those claims in the past.

Today, a few more obscure allegations remain unresolved, either because the evidence was so flimsy I couldn’t investigate them or because they have since been retracted by the people who alleged them. A few of the claims I addressed in that initial thread, about three or four, ended up being real documentations of voter fraud (read: not election fraud, which is the systemic corruption of an election, but some individuals voting twice or in districts where they did not reside).

If you can take your mind back to 2020, when most of this started, you'll remember the central premise, the granddaddy of them all: Dominion Voting Systems. This is a company that sells electronic voting hardware and software, including voting machines and tabulators used in many swing states. The initial, widespread allegation from Trump and his allies was that these machines were "flipping votes." That is how Biden vaulted to the lead when most of us were sleeping on election night (I was awake). It was how Democrats had stolen the election.

The most prominent purveyor of that theory was the now-disgraced lawyer Sidney Powell. We now know, through contemporaneous text messages from Trump's staff, dozens of court cases, and Powell's own temporary retreat, that it was all nonsense. Unadulterated lies. Per text messages that were leaked from the Jan. 6 investigation, as Trump's own chief of staff and several Republicans said privately, Powell never showed them any evidence of fraud, despite the fact they initially embraced her theory. That is why she was eventually distanced from Trump. And it is why she is, right now, being sued for millions of dollars by Dominion for defamation.

Slowly, though, claims about Dominion fell to the background and were replaced by other theories on how the election was stolen.

One of the most prominent today is actually one that resonates with me. This is the argument that, because major social media outlets and news companies ignored or outright censored the "Hunter Biden laptop story," Joe Biden was able to hold onto his lead. Then, Mark Zuckerberg donated $500 million to election centers across the country, which disproportionately benefited Democratic districts. These two things, many people say, tipped the election in Biden's favor. As far as "stolen election" claims go, it has more merit than most.

There are problems with those claims (like the many Republican donors who also funded elections, or that they sound an awful lot like Hillary Clinton’s claims about 2016), but most importantly is that they aren’t proof of election fraud — which, early on, was the central premise of how the election was purportedly stolen. The sleight of hand matters.

Which brings us to 2,000 Mules.

This film reverts to the original allegation from Trump and his allies, which is that the 2020 election was stolen via a far-reaching election fraud conspiracy. Instead of Dominion Voting Systems flipping votes, though, the story hinges on a network of paid “mules” who were stuffing ballots into drop boxes across the most critical swing states. Rather than arguing the votes were counted inaccurately, the film posits that the votes themselves were illegal.

I noticed that a lot of news organizations "debunked" the movie based solely on the trailer, the kind of fact-checking or journalism that is sure to (or should) further erode public trust. Others, like the Associated Press and Washington Post, did some more thorough fact-checking, though they left some obvious points out or failed to adequately explain a few things.

I'm going a different route. I decided to shell out the $29.99 to buy the movie online (the price was an initial red flag) and write about it. I've divided this piece into four parts: What they got right, what they got wrong, what they failed to prove, and finally some concluding thoughts.

I hope you enjoy it.


What they got right.

As with any political documentary, threads of truth have to exist for your argument to be even slightly compelling or convincing. 2,000 Mules is no different.

The movie was careful to do a few things. First, they noted once or twice that voter and election fraud are bipartisan issues — i.e. that when fraud happens, it is committed by both sides. There were parts of the film which claimed that the Democrats are now the election/voter fraud party (we'll get to that spurious allegation in a little), but generally speaking they presented election integrity as an issue all Americans should care about because both sides have been caught cheating. There is nothing wrong with this assertion.

They also emphasized a core principle of their theory: Trump was deeply loathed, and Democrats would go to great lengths to stop him. I actually think this is true, too. It is not inconceivable to me that Trump caused so much manic hatred from the left that the Democratic party would try to orchestrate the theft of an election to stop him. The simple premise that Trump was that hated, and that Democrats feared him winning that much, that they might commit a crime of this degree is totally believable to me. He is that loathed, and some Democrats were that desperate to win.

Paired with that claim is the idea that Trump's loss, before any ballot was even cast, stolen or stuffed, was in part due to his overwhelmingly negative media coverage. This claim, too, has merit. One study in particular (which the film cites) shows that media coverage of Trump was extraordinarily negative — over 90% of all coverage — in the mainstream press. Liberals would, of course, argue that this was Trump's own doing. Perhaps it was, but it is still an accurate claim.

Related to the film's core allegation, which is founded on cell phone location data, they also talk a lot about how location data is used. The value of this data, and the way it has been used to solve crimes or tap into the interests of cell phone users, is all real. A good example: Plenty of people think their cell phones are listening to them. If you own a smartphone, you've probably had the experience of talking about something out loud with a friend, and then a few hours later getting a targeted advertisement on social media or YouTube for that very thing you were discussing.

In fact, this is not your phone listening, but the apps on your phone leveraging location data. What is actually happening is that your phone is sensing your location near your friend's location, and then it is hitting you with targeted ads based on your friend's interests. It's a cross-pollination, of sorts. And when it happens all day every day, eventually the ad hits "jackpot" by presenting you with something you and your friend ended up discussing. Voila.

In 2,000 Mules, they also discuss how cell phone data was used to track down and prosecute the rioters from January 6. This, too, is true. There have been reams of reporting on how the people who stormed the Capitol were tracked through their own apps. One of the characters in the film claims that law enforcement must have been spying on those rioters beforehand, by virtue of how quickly they were tracked down and arrested. I have no proof that this is actually true but, still, it is true that location data is incredibly valuable, can be used to draw lots of insights, and has been used to solve crimes.

The film also alleges that the rules were changed in many swing states before the election due to the pandemic. This, too, is true. We witnessed widespread adoption of mail-in and absentee voting in 2020 due to Covid-19. In some states, like Pennsylvania, mail-in adoption was quite controversial, with plenty of Republicans loudly and vociferously fighting it (even those who had voted for it months before). All across the country, election workers were processing a volume of mail-in and absentee ballots that they had never seen before, in an election unlike any we’d ever witnessed.

There were three other major plot lines the story touches on that are true: One is that hundreds of millions of private dollars went into funding the elections in 2020. As I wrote in the intro, this is true. I think it is more an indictment of what a lousy job our government does in properly funding our elections, but there is no doubt this happened in 2020.

The second is that we have examples of recent election fraud that forced elections to be overturned. This is also true. The example they reference briefly in the film is from North Carolina, where a Republican candidate leveraged an absentee ballot scam to win a race for a House seat in 2019. The candidate was caught, the election was thrown out, and a new election was held.

Finally, the film talks about the threat of elderly or incapacitated folks being taken advantage of for their ballots in nursing homes or assisted living facilities. This, too, is a real thing that happens. D'Souza attempts to tie this reality to his theory of a network of mules, though he never really gets there. But he is right that there are examples, even from the 2020 election, of nursing home workers forging ballot applications (it should be noted, those ballots were caught during a signature verification procedure, and stopped any fraudulent votes from being submitted).

So, that's some of what the film got right, or the guideposts of their theory that are founded in fact.


What they got wrong.

People love to use the expression that a lie gets around the world twice before the truth even gets its shoes on. I think it's a good expression. I need a new one, though, to explain the phenomenon of how explaining why something is untrue or misleading so often takes more words, effort and time than the misleading allegation you're trying to address (seriously, write in with your suggestions). Hence the length of today’s newsletter. Please, bear with me here.

From the very opening scene in the film, the movie is littered with recycled claims that have already been disproven or are deeply misleading. Given my previous work, many were addressed in my past writing, and I was able to identify them from memory.

The very first clip in the movie is a video of Joe Biden. It purports to show him announcing that he built the biggest "voter fraud" operation in history. The clip went viral during the election in 2020, too, even though it is deceptively edited. What actually happened is that on October 24 of 2020, Biden was answering a question on a podcast about what he would say to people who hadn't voted yet. He was being filmed and recorded for the show. Below, I've transcribed his full answer, and put in bold the only part 2,000 Mules quotes:

Republicans are doing everything they can to make it harder for people to vote — particularly people of color — to vote. So go to IWILLVOTE.com. Secondly, we’re in a situation where we have put together, and you guys did it for our administration — President Obama’s administration before this — we have put together I think the most extensive and inclusive voter fraud organization in the history of American politics. What the president is trying to do is discourage people from voting by implying that their vote won’t be counted, it can’t be counted, we’re going to challenge it and all these things. If enough people vote, it’s going to overwhelm the system. You see what’s happening now, you guys know it as well as I do, you see the long, long lines and early voting. You see the millions of people who have already cast a ballot. And so, don’t be intimidated.

Obviously, Biden is saying he has put together an organization to combat voter fraud. Dinesh D'Souza opens his film with this clip, cutting out all the words around it, and lays it over very scary sounding music. On screen, you then see an actor, at night, in a dark hoodie with a bag over his shoulder, stuffing ballots into a ballot box. From the very first clip — the very first words spoken in the film — there is some misleading reporting going on.

Shortly after, D'Souza pivots into perhaps the most popular claim about election fraud in the 2020 race — and one of the easiest to explain. He shows a series of clips where Trump is ahead, and then recalls that when Americans woke up the next day, Biden had magically taken over the lead while we were all sleeping. D'Souza uses news clips of anchors saying the vote count "stopped" and edits them up nicely with clips from the next day showing Biden's lead extended. The obvious takeaway is that the race was halted by some mysterious forces, and then Biden's winning votes were delivered, and then Biden won.

Of course, for those of us who stayed up on election night, we know that the vote count never actually stopped. And the late surge from Biden was predictable — in fact, I told readers to expect it weeks before the election — because in most swing states mail-in ballots were going to be tabulated last. Since mail-in ballots were predominantly coming from Democratic voters, who were more fearful of Covid-19 and more encouraged by their preferred candidates to vote by mail, it was obvious that the votes counted last were going to disproportionately favor Biden. Especially given that Trump repeatedly told supporters that mail-in voting was unreliable.

We knew all of this, and there is plenty of proof. For instance, you can read this piece from the BBC on November 2nd, 2020, the day before the election, explaining to readers that "the influx of postal ballots this year will make it hard to see who's in the lead early on." Here is what they said (emphasis mine)

Different states have different rules for how - and when - to count postal ballots, meaning there will be large gaps between them in terms of reporting results. Some states, like Florida and Arizona, begin pre-processing ballots weeks before 3 November. Others, like Wisconsin and Pennsylvania, won't touch these votes until election day, meaning they'll likely be slower to count... early tallies may be deceptive. This is due in part to more differences between states in reporting. In some states, ballots cast in person on election day will be counted first. These counts are expected to favour Mr Trump, as his supporters are expected to be more likely to vote on the day.

So, the fact that you went to bed when Trump was ahead and woke up (after the mail-in ballots were counted) to Biden leading should not have been surprising. It played out exactly as reporters and election experts expected it would. It should be noted, too, that the only reason this happened was Republicans in states like Wisconsin and Pennsylvania refused to tally, process and count mail-in votes before Election Day. They wouldn’t pass legislation to do that. This wasn't an issue in Florida and Ohio, which are also run by Republicans but had results quickly on the 3rd, because Florida and Ohio Republicans allowed mail-in votes to be processed before Election Day.

Relatedly, while discussing Biden's late surge, Charlie Kirk claimed at one point that "Donald Trump won almost every single one of the predictive counties that show who is going to be president." This is patently false.

Bellwether counties change every year. When I was in high school, the county I grew up in (Bucks County, Pennsylvania) was a bellwether. Today, it leans Democratic. Kirk makes this claim look real by citing past bellwethers instead of the ones that were used in 2020.

In October, Cook Political, which is widely considered one of the most reliable election outfits in the country, posted a story about the 10 bellwether counties in 2020 that would determine the race. Here are the counties with their eventual winners in parentheses: Kent County, Michigan (Biden). Wood County, Ohio (Trump). Erie County, Pennsylvania (Biden). Sauk County, Wisconsin (Biden). Marshall County, Iowa (Trump). Maricopa County, Arizona (Biden). Pinellas County, Florida (Biden). Peach County, Georgia (Trump). New Hanover County, North Carolina (Biden). Collin County, Texas (Trump).

In other words, Cook Political isolated these 10 counties as the ones that would signal the winner in the race, and Biden won 6 of the 10. He also won the presidency.

In the same moment in the film, one of the people at a roundtable D'Souza assembled says Trump was outperformed by every Republican down the ballot, proof that he was targeted by Democrats and the election was stolen (because this is allegedly an anomaly). This claim, too, is nonsense. Trump ran behind some Senate candidates (as he did in Georgia) and ran ahead of others (as he did in Arizona). This happens in nearly every presidential election with an incumbent and is not unusual at all. Most of these claims include people selectively choosing only certain races to make their point — when a holistic look at results doesn’t back up the thesis.

Sebastian Gorka says in the film that there is no way Biden got more votes than the first Black president. Then, a mere two minutes later, he notes that Trump got 10 million more votes than he received just four years before. Oddly, Gorka doesn't put together that this is simply reflective of a race that drove massive turnout, and not proof of a conspiracy (Biden getting more votes than Obama 12 years later is framed as criminal, while Trump gaining 10 million votes in just four years is framed as proof of his popularity).

Later on in the film, once many of D'Souza's most novel allegations are fleshed out, he brings in election expert Hans Anatol von Spakovsky.

Spakovsky alleges that voter and election fraud are actually quite common, citing recent examples where elections were overturned in North Carolina, Mississippi and Florida. Based on his brief descriptions of the cases, I was able to find the ones he was referencing. They are worth explaining.

The first is the aforementioned North Carolina case, which involved a Republican candidate and political operative who got caught trying to execute an absentee ballot scheme. This case is real, though it should be noted that it undercuts the film’s narrative that election fraud is now predominantly a Democratic issue (it’s also worth noting the film completely ignores a ballot harvesting scheme orchestrated by two Republican operatives recently uncovered in South Philadelphia).

The second reference was in Mississippi. This was a Democratic runoff election for alderman in a town of 5,326 people. The race was overturned because the judge found "irregularities" in the paperwork of 66 of the 84 absentee ballots cast. The race had been decided by 37 votes. The judge apparently uncovered an election fraud scheme being orchestrated by the people who notarized paperwork for ballot applications, then ordered one of them arrested. On top of being a tiny, local alderman's race, the story should be proof that even subtle election fraud committed by people on the inside of elections is very often discovered and prosecuted.

The third example was a Florida race. Far from being an election overturned because of sophisticated fraud, it was actually a tiny city council race decided by one vote. A recount of the race found two previously uncounted votes, which gave the eventual winner a 269 to 268 victory. There was no fraud; it was just a recount executed in a tight race. It's unclear why Spakovsky would cite this example other than to mislead viewers.

It should be noted, too, that in both the Mississippi and North Carolina races, there were numerous witnesses and detailed testimony under oath about the fraudulent schemes. People who literally worked for the candidates in question testified against them, which is something to keep in mind for the allegations that we discuss later.

Spakovsky also cited a real case of illegality in Texas, where a social worker filled out 67 ballot applications for people in her facility who had intellectual and developmental disabilities (again, no illegal votes were cast — she was caught). From news reports I’ve seen, it's unclear who she was trying to help in the election, but the story reminded me that Texas Attorney General Ken Paxton had offered a minimum of $25,000 for election fraud tipsters and got $2.2 million from the state to run an entire election fraud unit. The first tipster he had to pay out was to a Democratic poll worker who caught a registered Republican filling out a second vote in his son's name.

With six staffers, over $2 million, and literally $25,000 in prize money to report voter fraud, Paxton has closed just three cases in 2021, down from 17 in 2020. Mind you, that is not guilty pleas or convictions — just closed cases. 11 million people voted in Texas in 2020, and Paxton has confirmed a few isolated incidents impacting a handful of votes. Save the social worker who illegally filled out ballot applications, they were all run of the mill of voter fraud cases.

There were some other dubious or outlandish claims, but a lot of them were political in nature or claims that have been shot down enough times that I don't feel the need to address them here.


What they didn't prove.

Ok, we've gone over some of the claims in the film that were accurate and some that were misleading or untrue. Now we get to the main course, which is the thrust of the entire film.

D'Souza employs the help of Gregg Phillips, the former head of the Mississippi Department of Human Services, and Catherine Engelbrecht, the head of the vote-monitoring organization True The Vote. D'Souza introduces them by saying True The Vote is a "non-partisan organization," which would be funny if the lie didn't disregard the intelligence of viewers so blatantly. In fact, True The Vote is an offshoot of King Street Patriots, a Tea Party organization founded by Engelbrecht.

Gregg Phillips, who now works at True The Vote and is presented as Engelbrecht’s partner on this project, is a lifelong (and very vocal) conservative. In 2016, he became notorious for being the origin of Trump's claim that three to five million people voted illegally, because he made the allegation before voter history data was available in most jurisdictions (that is to say it was essentially invented out of thin air).

Together, Phillips and Engelbrecht explain to D'Souza (and his wife, who appears by his side throughout the film) that they have purchased several petabytes of cell phone location data. They claim that they have used that data to uncover a massive drop box scheme, where paid "mules" were collecting ballots from nonprofit organizations and then dropping them off in drop boxes in swing states like Georgia, Arizona, Pennsylvania and Wisconsin.

They purport to prove this by mapping the cell phone data location movement of millions of people. Then, they create a criteria: From October 1 to Election Day, They isolate every person who passes or stops at 10 or more drop boxes in a day, and every person who passes or stops at five or more nonprofit organizations tied to political entities. Using this criteria, they isolate over 2,000 people across several Democratic strongholds in swing states (mainly Philadelphia and Atlanta). These "2,000 mules" are then credited with delivering hundreds of thousands of ballots that threw the election to Biden.

As a nice topping, Phillips and Engelbrecht also filed freedom of information requests to obtain surveillance footage from several drop box locations across the country. They present about a dozen clips from this footage purporting to show people illegally stuffing boxes with ballots. In a few instances, they show people wearing various kinds of gloves (including doctor's gloves) in the Georgia Senate runoff, which took place in early January. They explain that clips of people wearing gloves started to show up after December 23, which — it hit them — was right after two women in Yuma, Arizona, were indicted on ballot harvesting charges and caught using fingerprint technology.

The pair tells D'Souza, who is now astonished at the mountain of evidence, that their location data shows these same mules returning to drop boxes repeatedly, sometimes multiple times in a day, and that they have some of the same people in different drop box locations on surveillance footage.

Finally, in the latter half of the film, D'Souza unveils a "whistleblower" in San Luis, Arizona. The woman is backlit, totally anonymous, and says she witnessed what she "assumed" to be mules picking up payments for ballot collections. No other first-hand witnesses from any state are presented.

These allegations are the crux of the new stolen election theory, and the claims that give the film its name.

I'll say up front that I found these allegations both peculiar and challenging. For starters, I had to read a huge amount of information about the way location data works, refresh my memory on how Georgia tracks votes (I was immersed in all the swing states during 2020, and none are exactly the same), and scour news reports from 2020 and the 2021 Georgia runoff to see if the film's central claims lined up. I also spoke to a few election security experts and tech nerds to resolve some claims I didn't quite understand (in all but one case, I was careful to ask questions independent of D'Souza's film, so as not to poison their responses).

Ballot harvesting is not a new kind of election fraud scheme or allegation, but this kind of coordinated attack across several states involving at least two thousand people (at one point, D’Souza speculates the real number is more like 54,000 mules!) was both astounding and hard to wrap my head around.

After probing these allegations, though, I'm quite skeptical.

Let's start with the cell phone data. Throughout the film, D'Souza, Phillips and Engelbrecht repeatedly insist that their data show illegal behavior. But what they show in the movie doesn't prove this — and in some cases even undercuts their claims. In one example, Phillips shows an animation of mules passing by drop boxes and going to political nonprofits. But the animation he uses in the film doesn't match the locations of the actual drop boxes in Atlanta, which were all (by law) required to be on government property.

This may seem silly — maybe the animation was just for the movie? — but if you watch the film you know why it's important. We're supposed to be seeing what D'Souza is seeing in real time, and in the movie Phillips presents this animation to him (and a roundtable of conservatives) as proof. Yet the map appears to be made up, or, at least, significantly altered. This was the first major red flag in the film.

They also allege that the cell phone data is accurate enough to show if one of the mules stopped at a drop box, rather than just walked past it. Best I can tell, this is not true. There are a few kinds of cell phone location data: GPS data (like what’s used in Google maps), bluetooth internet location data, and cell site location data (via pinging towers). D'Souza and his team say they are using location data bought from private apps, which could only feasibly be the GPS data. This can tell you where someone is within five to ten feet. If you're interested, you can read about how this works here.

Despite D'Souza repeatedly claiming otherwise, it is true that the cell phone tower or GPS data could have easily been capturing people walking right past a drop box (and one of the nonprofits, for that matter) without actually pausing there for any extended period of time. Not once in the film do they show “mules” pausing at the various locations more than once in one day. Depending on the kinds of data they obtained, the locations could be showing up via intermediate "pings" and triangulations from cell towers as someone moves through various locations. The upshot is we simply don’t know: We just have to take D’Souza’s word for it, even though it’s totally feasible (if not likely) that the data they collected shows people simply walking past the same locations every day.

Here, it is important to note that drop boxes are intentionally placed in busy places. In Georgia they were on government property, but across the country they’re put on street corners, college campuses, thoroughfares, etc. The reason for this is obvious: They are supposed to both increase voter access and be easily monitored. This also creates problems with the 2,000 Mules theory. Take a personal example: Every day, I walk to my office in the morning. Then, at lunch, I walk to my gym. My gym happens to be next to a Democratic nonprofit. If Brooklyn had a drop box between my apartment and my gym, I would have gotten ‘caught’ in the dragnet.

There are all sorts of other problems with their criteria that they simply don't address. Phillips actually testified before the Pennsylvania state senate in March. During that testimony, he conceded that his group did not get any surveillance footage from Philadelphia despite claiming in the movie they have proven that more than 1,100 mules operated there.

Pennsylvania state Senator Sharif Street explained that he moves around the city with a cellphone, a smart watch, a tablet, and a mobile hotspot to do work remotely. All of which produce location data that can be tracked by private companies and bought by Phillips. Street speculated that he was almost certainly caught in the 2,000 Mules dragnet because he traveled with a staffer who carries two devices and made dozens of trips to drop box rallies and nonprofit offices, and drives past a half dozen drop boxes on the way to work every day. D'Souza and Phillips have no explanation for how they keep people like Street out of the count.

“What this shows you is there were lots of people who were going past these boxes, it doesn’t necessarily show what folks were doing when they got there, right?” Street asked Phillips during the hearing.

“Which is exactly why you need video” on the drop boxes, Phillips said.

To me, this reads a lot like an admission the evidence in the film is not enough — yet D’Souza frames it as being definitive.

As for ballot harvesting — a term for delivering completed ballots for people other than yourself to drop boxes or polling places — it is legal in many places. In the states 2,000 Mules focused on, it is mostly illegal, with some exceptions for family, household members, or someone in your care.

The very first "mule" D'Souza, Phillips, and Engelbrecht focus on is someone who was caught in their dataset and appeared on a Fulton County, Georgia, drop box surveillance video. Here is a screenshot of the "mule" they purport to be delivering illegally harvested ballots:

Screenshot from 2,000 Mules

As it turns out, the Georgia Secretary of State's office had already investigated the man when Phillips presented the evidence to state officials before the movie's release. “We investigated, and the five ballots that he turned in were all for himself and his family members,” Republican Secretary of State Brad Raffensperger said during a debate last week. His office also said they looked into several other videos in the film and had found no illegal behavior. Nowhere in the film do D'Souza or Philips make any mention of the fact that law enforcement has looked into their claims.

Of course, this doesn't mean every person in every clip that D'Souza showed was doing something legal either. Just that several of the main characters in his film had already been investigated and exonerated, and that D'Souza and Phillips don't disclose this to viewers.

D'Souza, Phillips and Engelbrecht also allege over and over again that they have videos of the same people appearing at different drop boxes in the same day and placing ballots in them. This would, actually, be a major red flag. Given that in these states you can only drop off ballots for family members, people who are disabled, people in your care, etc., one would think you'd only ever make a single trip to a single drop box in a day.

Yet, not once in the film do they show the same person at multiple drop boxes in a day. Not once. Over and over again, they say they have people delivering to multiple drop boxes on surveillance footage, but despite showing a dozen surveillance footage clips they don't show one "mule" who goes to multiple drop boxes as their location data purportedly proves they do. It is one of the most glaring omissions in the whole movie, because it is so central to their overall allegation.

To D'Souza's credit, he responded to this criticism. He tweeted this photo, with the caption: "I keep hearing there is no video of the same guy at more than one drop box. The difficulty is that many drop boxes had no surveillance at all. Whole states refused to comply with the rules and install video cameras. But okay, here you go!"

A response D'Souza put out on Twitter

There are a few things wrong with this reply. First, this photo is of the same person at the same drop box on two different days. D'Souza (and his team) alleged repeatedly that these “mules” were making several stops at several drop boxes in the same day. That is the entire premise of their geo-tracking cell phone data, if it is to be believed. Second, he also says they have millions of hours of surveillance footage in the movie (to bolster their claims), but in this response frames it as if they don’t have enough footage to back up their claims. It is a bizarre about-face.

Third, this person — to anyone with family or friends who work in the care world — appears to be a nurse or assisted living employee of some kind (he is in scrubs with an identification lanyard in both images). In Georgia, it is legal for these caretakers to drop off ballots for people who live in their facilities. As we mentioned at the top, there have absolutely been cases of these caretakers illegally applying for ballots or filling them out for the incapacitated.

But it is just as likely (in fact, much more likely!) that this person is doing a good deed for the people in his care rather than breaking the law. He is delivering the ballots in broad daylight, in his work scrubs, at the same drop box, on two different days around the same time. Further, there is no evidence presented that he is one of the "mules" caught in the cell phone dragnet. I don't find anything shady about this at all.

On top of that, it also proves that despite their petabytes of data and millions of hours of video footage, D'Souza has only produced this one person as proof of someone visiting a drop box twice (and it's someone at the same drop box on different days).

The introduction to this entire theory of geolocation was also built on a flimsy premise. Phillips starts by explaining to D'Souza how he had used a similar geolocation tracking method to prove only a few people were in the vicinity of a shooting in Atlanta, Georgia, and that data he gave to the FBI helped revive a cold case murder which was eventually solved. As Phillip Bump explained, though, almost nothing he said about this case was true:

But looking at the case more closely, you see how the impression you’re meant to have is wildly misleading. The shooting led to the death of Secoriea Turner on July 4, 2020. It was far from a “cold case” — police arrested a suspect about two weeks later after he turned himself in. A second suspect was arrested in early August 2021 — not by federal law enforcement but by state officials. There is no indication that geolocation data played a role in either arrest, much less data provided by Phillips’s team.

D'Souza also does not explain the ballot harvesting scheme they say they've uncovered. Even supposing that they had one person in multiple locations, or even one person delivering dozens of ballots (which they don't ever show in the film — typically it looks like five or six), they have no explanation for how the scheme works. This is critical. Statistical significance isn’t enough to find a theory of causation, you have to accompany it with a functional explanatory mechanism.

For starters, why drop boxes? In states like Georgia, these people could have returned multiple ballots through their own mailboxes or the Post Office. Why would they use ballot drop boxes on government property, which any schemer or election official would know are being filmed? Second, absentee ballots in Georgia were only given to voters who requested them. Returned ballots have a barcode, registration information, and voter signature. Ballots delivered without that information weren't counted in Georgia. Ballots without matching signatures or information would be flagged. So whose ballots were these?

If they were being stolen and illegally filled out, we'd see a wave of invalidated double ballots. Georgia held probably the most investigated election in the country, yet that didn't happen.

So, what did happen?

This, or any of the details of how the delivered ballots were invalid, is never explained in the movie. Georgia election reporter Mark Niesse addresses this well.

There are a few other claims worth addressing to wrap this up.

First, the film makes a salacious claim about people delivering ballots with gloves on in the Georgia runoff. They allege that these gloves started appearing after two women were indicted in Arizona on a ballot harvesting scheme and were caught with fingerprint technology. Yet, I could find nothing in the indictment or news reports about it that says anything about fingerprints being used to catch the women.

As for the gloves, the answer is so obvious I'm struggling to understand why it needs to be explained: It was January in Georgia, the coldest month of the year, and we were in the middle of a pandemic. A few people wearing medical gloves seems... totally normal to me. You can go see pictures, videos and stories about people wearing gloves and masks to drop off ballots all over the country. I saw a lot of that in person in 2020. And the people who were wearing other gloves were also wearing hoodies, beanies, and long pants. Because it was cold.

D'Souza shows many of these people taking selfies with the ballots or pictures of the ballot box. This, too, is alleged proof that they are documenting their crime to get the money they are owed. This, to put it nicely, is what we call a motivated conclusion. I took a selfie of myself with an "I voted" sticker after I voted. Who is to say these people weren't taking a photo to simply let the people whose ballots they had know their votes had been deposited?

Even more bizarre is that they wouldn't have to take any pictures. If this were really a sophisticated scheme, the people running it could simply check public records to see if someone had voted. I can look up whether you voted right now. It's not hard. In other words: They could check whether a ballot was delivered and tallied by looking at a state's public records, and then pay the people out accordingly. That'd be a lot more reliable than a photograph of a random ballot box, and it wouldn't require a person incriminating themselves on surveillance footage.

Of course, most suspiciously of all, there is only a single witness to this entire event.

Again: D'Souza is alleging a baseline, guaranteed “2,000 Mules” involved in this scheme, and floats that as many as 54,000 mules actually took part in it across the country. Yet they produce one whistleblower, who remains anonymous, and says she witnessed what she assumed was people being paid for their roles as mules. Then, with all the cell phone data they allege to have, they couldn't contact one of the participating 2,000 (or 54,000) mules and convince them to go on record? Or even to share their story anonymously?

This alone is beyond belief. As D'Souza himself pointed out in other documented cases of election fraud, the examples always have witnesses. In North Carolina, people on the candidate's own staff testified against him. Yet we are supposed to believe that none of the more than 2,000 people who participated in this scheme have gotten caught, blabbed to a friend, been prosecuted by law enforcement, or could be compelled to share their story with the press?


Some concluding thoughts.

I have a philosophy about taking in arguments like this: You should judge the content of the argument, not the person behind it.

This, to me, is critical. It's why when I approached some people for advice or reactions to certain allegations in the film, I was careful not to mention I was asking about D'Souza's film, because he is considered a bit of a huckster (even by some on the right). It's also why, up until now, I haven't said much about the people who made this film or are featured in it. I want you to judge the film by the merits of its claims, and the issues I raise about those claims.

But it would be a disservice to you, the reader, to not at least give some space to describe the people making these allegations. They are, after all, accusing thousands of people of federal crimes, claiming the 2020 election was stolen by Democrats, and calling for people to be prosecuted. And as I said at the top, the best explanations include understanding motives.

I already mentioned Gregg Phillips’ history. You can go read about him in articles from 2017, when he first claimed three to five million people voted illegally without ever proving it. It should be noted that Phillips actually backed off those claims after they were elevated by Trump, conceding he didn't have the goods. It should also be noted that he has been repeatedly accused of government misconduct and abusing his positions for financial gain when he was a government official in Mississippi. In 2017, he owed the U.S. government $100,000 in unpaid taxes, according to The Guardian, and in 2017 the Associated Press revealed that he was registered to vote in three different states. He replied by suggesting this was proof not of his own criminal activity, but of how broken the system is.

In short: Phillips is a lifelong conservative activist with a very clear record of impropriety who, along with Engelbrecht, D'Souza presents in the film as "non-partisan."

D'Souza is more well-known, and it seems relevant that he was one of the people pardoned by President Trump in 2018 for illegal campaign contributions (he pleaded guilty). D'Souza has a long track record of being a conservative provocateur, and this is not his first foray into filmmaking or documentary making. In 2017 he published an entire book claiming the Democratic party's platform was similar to that of the Third Reich (all I'll say is he did a very poor job arguing this). This month, Trump premiered 2,000 Mules the film at Mar-A-Lago with many of his closest political allies in attendance.

Clearly, D'Souza is monetizing his claims of fraud. He charged $30 for this movie and says it has already grossed over $10 million. I know a lot of people believe there are strong motives for Democrats or mainstream media to cover up stolen election claims, but I'd like to point out here there are even stronger motives to make these bombastic claims and put them behind a paywalled film. $10 million worth of motivation. I can say with certainty that if I announced today I had changed my mind about the 2020 election and now believed it was stolen, it would be an incredibly profitable career move.

Finally, there is Catherine Engelbrecht. She is one of the leaders of True The Vote which, again, D'Souza presents as a non-partisan organization in the film. On the contrary, True The Vote has a decades-long history of challenging elections where Democrats won, a very open and public track record of supporting Tea Party Republicans, and plenty of controversies to its name. They have been accused of voter intimidation and making unfounded allegations of voter fraud. As far as I can tell, in the last 12 years they have documented some instances of actual voter fraud, but also repeatedly made outlandish, headline-grabbing claims that were never proven.

In 2020, Trump supporter and North Carolina money manager Fred Eshelman sued True The Vote for $2.5 million after making a donation to their Validate The Vote campaign that he said they were not providing updates on. The case was eventually thrown out, but is a recent example of how the group's work has upset even its staunchest supporters.

At the end of the film, viewers are encouraged to get out and vote in 2022 and 2024. They are also told to visit Engelbrecht's website, which I did. Immediately, you are prompted to donate money and join as a supporter. Again: None of this should be a prerequisite to disbelieving their arguments, which is why I tried to address the claims independently first. I’ll ask you to subscribe to Tangle or donate to support our work at the end of this piece.

But it isn't irrelevant, either.

True The Vote itself, on the face of it, appears to be after a goal I support: Better tracking of voting rolls and voter data. The organization has absolutely brought forward some legitimate challenges to fraud or election results. But it is simultaneously clear that they are anything but non-partisan; its principal characters all have their own checkered history with the truth, and each has clear financial incentives for leveling the claims they make in the movie.

In the end, the most damning part about 2,000 Mules is not what the film's creators tell their viewers, but what they leave out. In September, True The Vote presented the Georgia Bureau of Investigations (GBI) with their claims. GBI responded by saying they lacked sufficient evidence to even merit an investigation.

“What has not been provided is any other kind of evidence that ties these cellphones to ballot harvesting,” GBI Director Vic Reynolds says in a Sept. 30 letter. “As it exists, the data, while curious, does not rise to the level of probable cause that a crime has been committed.”

More recently, True The Vote has refused to cooperate with Georgia's election officials who have requested the names of people allegedly caught harvesting ballots. As they note in the film, the cell phone data is attached to individuals, so if the 2,000 Mules creators felt they really had them pegged, it shouldn't be hard to prosecute them. They could just hand the identifying phone numbers of the mules to GBI investigators. Yet they haven't. Why not?

It's gotten so bad that the State Election Board actually issued subpoenas to the organization last month seeking documents, recordings and names of individuals involved. True The Vote still hasn't turned them over. Why would a group that claims to want justice not hand over what it says is irrefutable evidence to the one group that could prosecute the case?

All of this — the lack of cooperation, the obvious financial incentives, the history of the people behind the films, and the many easily refutable and misleading claims in the movie itself — makes me quite certain the premise of 2,000 Mules is nothing more than the latest "stolen election" fad.

And in this case, a rather unconvincing one.


Want to share this?

You can share on Facebook by clicking here.

You can share on Twitter by clicking here.

You can share via email by clicking here.


Want to support our work?

You're currently on our free list. We were going to send today's email to subscribers only, but we decided to unlock it for everyone. If you want to become a member and support our independent journalist and get future Friday editions, subscribe here.

You can drop something in our tip jar to say thank you. Or, you can just forward this email to friends.

Feel free to write in and let us know what you think.

Thanks for reading, and we'll see you Monday.