Barring any paradigm-shifting event in the 2020 presidential race (of which there are still months of possibility), I believe the election is already so far tilted that it could be called over and that Donald Trump will win. I’ll explain this below—and why this is sour conclusion has implications well beyond this year.
There have been two very good pieces of political journalism that I’ve read recently that have brought me to this conclusion. First and foremost is an article from The Atlantic titled “The Billion-Dollar Disinformation Campaign to Reelect the President“, which has the secondary tab title of “The 2020 Election Will Be A War Of Disinformation”. The second is from Politico and is titled “An Unsettling New Theory: There Is No Swing Voter“.
Please read both.
I’ll be doing some summarization here but both are deeper and better written than anything I could muster. My goal is to ty these two pieces together, in addition to some other observations, and get to my thesis here that the 2020 election is likely over due to circumstances allowed for in micro-targeted “advertising” and the budget and large-scale strategy that the Trump campaign will deploy utilizing it.
The ‘No Swing Voter’ Theory
We’ll start with electoral politics and voter turnout. The theory laid out in the Politico article is that actual voter turnout matters more to winning elections than moving blocs of swing voters one way or the other. The election is decided by WHO VOTES rather than who voters VOTE FOR. Now, this is based on one polling analyst (which is a stretch to call it truth), but has some profundity. And it might help explain why one Trump campaign goal, voter suppression, is a target.
As the article says,
“Bitecofer’s theory, when you boil it down, is that modern American elections are rarely shaped by voters changing their minds, but rather by shifts in who decides to vote in the first place.”
Turnout here is the name of the game. That there are no swing voters I think is untrue and the headline here is misleading. People will move from party to party, but the article’s thesis (or the political scientist herself) is that it won’t matter. Getting people to vote will move numbers.
Have we seen this play out accordingly? Well, comparing the 2016 election to the 2012, Michigan, Wisconsin, and Ohio all had less total votes casted in the more recent election. Florida had nearly a million more votes, but added more than 1.3 residents in that time.
It’s an interesting theory and one I would have likely just kept in the back of my head as the 2020 election nears. But then I read the Atlantic article and it doubled on top of it.
Because as that article, written by McKay Coppins, describes in incredible detail, the Trump campaign is going to be waging an all-out disinformation campaign for this election designed not only at swinging voters, but also at suppression. And it’s not happening with the hope of working, it’s happening with a trove of micro-targeted data and analytics at the very people who have the best chance at being suppressed. Oh, and it’s already happening.
Let’s dive in.
Political Advertising? No. Micro-Targeting!
The Trump campaign plans to spend $1 billion dollars on its re-election campaign.
Read that again.
The legal limit on campaign spend in 1984 for a presidential campaign was $20.2 million. The Reagan re-election campaign said it had every intention on spending that. We’ve come a long way.
This raises a bunch of interesting points like how an incoming candidate will compete with that, the sheer size of a presidential campaign budget, the enormous avalanche and bombardment of targeted ads that swing states will likely see, and the profit windfall for those that get that money (see: Facebook).
I’ll spend a good deal of this piece talking about those swing states. Consider that Clinton could have won the 2016 election by flipping the three states with the smallest margin of victory , and most of these margins were under 2%.
She didn’t win these, of course, and there’ve been a cadre of reasons for it. But I think the real reason came down to targeted ad spent on individuals in those states by Trump and his digital team. With such a small number needed, the budget can be directed intensely on that population and maneuvers that may turn a person into a swing voter, or a non-voter entirely, can be magnified and mega-personalized.
Indeed, what the Atlantic article focuses on is the campaign strategy from the Trump side; and since “disinformation” is in the title, you probably know where this is going. Because they weren’t (and won’t be in 2020) targeted with ideas of what they’re passionate about, they were targeted with ads meant to confuse, obfuscate, and anger.
And we’ve heard a lot about that. But what we should be hearing more on is where those triggering ads WENT, WHO received them, and WHAT DATA informed the campaign that they were the right people to target. Those three caps-locked points above, taken into a totality and strategy, is called micro-targeting and boy is it getting big.
Here’s the way McKay Coppins, the author of the Atlantic article puts it:
“An ad that calls for defunding Planned Parenthood might get a mixed response from a large national audience, but serve it directly via Facebook to 800 Roman Catholic women in Dubuque, Iowa, and its reception will be much more positive.”
Trump’s political advertising operation was run as ‘Project Alamo’ in 2016 and the same people are back for more in 2020. I don’t know if the name has stuck but you can sense how strongly Trump believes in this from his comments during the most recent State Of The Union. He praised the “beautiful Alamo” which threw everyone who has ever seen the actual Alamo off (proof—and others went a totally different direction). The disappointingly small Alamo isn’t really all that beautiful; but a large operation that got him elected by targeting the “persuadable” mass is certainly worth praising in Trump’s book.
The War On Truth
What is disinformation anyway? That’s a definition a bit above my paygrade but I can say that it’s an attack on truth. Meaning both that it can be a variety (stretching truth, omitting truth, or just straight lying or deceiving). It’s not necessarily propaganda but it certainly can be. And sometimes it can be done without attribution, so that a person isn’t exactly sure where something came from. Picture a hyperlink that someone (even myself) might use that is linked to words that ostensibly the link would source, and yet when you click the link it’s something different or a slant of what the original author wrote. That’s an example of disinformation.
The purpose of it though is meant to obscure some kind of truth through deceptiopn. But as I’m struggling to explain even, this is something more vague and more complicated than we’re used to. Disinformation has been around for centuries, as long as we’ve been using perception to tell stories of what we believe is true; but the internet, weighed down by trillions of pieces of content is the first place where you can be showered (or bubbled in) completely by disinformation.
This is a point made in Coppin’s article where he describes the goal as “jamming the signals, sowing confusion”. I align this to something like DDOS attack on one’s own mental sensibility—your brain so overloaded by noise its attacked to the point of submission. And you start to change your mind. It’s been used now by several high profile politicians and leaders around the world.
Disinformation campaigns, like micro-targeting, are not new. In fact, for the former it was actually Obama’s first campaign in 2008 that began this on a Presidential election scale—though his data set and capabilities were inarguably more nascent. It’s the crossroads of these—where voters can be targeted to a micro and specific degree and then fed disinformation of all disorienting types—that’s new and concerning. It’s now been used in elections across the globe and will continue to be. Trump’s 2020 campaign like his 2016, is simply the biggest stage of them all and the the one with potentially the most profound effects.
And that, the stage and stakes being the biggest, means it’s worth spending the $1 billion on for an election budget and those who want a Trump win (large-pocketed conservatives with now-untaxed offshore bank accounts, foreign nationals sowing discontent, the Peter Thiels of the world (or CEOs of companies he’s invested in, etc… etc…) want a completely exercised and maximized campaign of disinformation to go down.
Where exactly will it go down? Mostly on Facebook and other visited internet sites. But that’s not the only place.
All Politics Is Local, So ‘Buy Local’ Indeed
One of the scariest parts of Coppins’ foreboding story comes nearer to the end—after paragraphs of the Trump campaign’s plans. It has to do with the purchasing and/or creation of local “news” sources popping up in key places; usually representing a titled point of view.
He describes the birth of registers like the Arizona Monitor which have come and gone with little proof other than cached sites and a trail of endorsements. Following the Breitbart model, they take a conservative point of view and often feature bombastic titles, ostensibly aimed at the part of the population truly concerned by a liberal bias in the media. As Coppins found, several of these online-only publications come from a company called Locality Labs.
If micro-targeting doesn’t work (for maybe a group of folks that aren’t tech-abled enough to actually see Facebook ads), they may just be drawn in by the new newspaper in town which probably lets them know that every single candidate on the left will take their guns away on the first day of office. That’s just old-fashioned targeting, really, and with online content being cheaper than ever, it’s easy to pull off.
This is concerning, surely, but I think it may be peanuts in a way (or a small-scale way of supporting a much larger, more concerning operation). And that bigger threat is the data-focused threats repeating from 2016. I’ll go into these below so you can see what elections are not up against. And when you think about how these tactics can tip the scale, remember to consider just how sensitive the scale is. Because, as the Washington Post explained, the 2016 election was “effectively decided by 107,000 people in these three states. Trump won the popular vote there by that combined amount.”
I know, I know, you’ve heard all about Cambridge Analytics. Maybe you watched The Great Hack. Maybe you binged a whole slew of articles about the company in 2016 after the election where they worked with the Trump campaign or in 2018 when they shut down and transformed into Emerdata.
And the outrage was deserved. The firm had access to 87 million Americans Facebook data, enough to make “profiles” about them and extrapolate out to their friends and connections. They said they had up to 5,000 pieces of data on every voter in the U.S. And they sought to weaponize it. Because the data was just used anonymously for aggregation purposes, it was divided into groups with those that they felt they could evolve put into a bucket called “persuadables” and then they went after those with tactics. They presented the data to their clients (Trump, Brexit, etc…) and then “won” by targeting those in that bucket.
Here’s one fact about that operation that’s stuck with me since the 2016 election. We know now that in addition to data pieces, Cambridge had access to private messages sent on Facebook. And that if you took part of the “quiz” that allowed them to collect this, it also allowed them to scrape data on 1,500 of your friends, uninvited I may add.
So if you’re someone who is friends of friends with someone who happens to be a little bit racist, they would know that. And indeed, one could imagine a scenario where they would find “persuadables” around those who may hold such prejudices and then feed you ads that stoke that little bit you may have inside you (even if you knew or didn’t). So you may see ads like these. Or if someone you knew put up something about Hilary’s political record you may see ads like these; this essentially leveraging the power of a Facebook-built network. And if you didn’t want to vote for Trump, these may have had you think twice about voting for Hilary (by the way, Cambridge folks take credit for the “crooked Hillary” meme from that campaign) and convince you not to vote at all (back to The No Swing Voter theory above).
Think about the level of marketing this allows for and the number of “swing” people that could be persuaded by fear to make action (vote one way for instance). And then think about the narrow margins that Trump won some states by….
Wisconsin hadn’t gone Republican in a presidential election since 1984. Trump didn’t even win the primary there—Cruz did, handedly actually. In the 2016 election, Trump got 1.405 million votes to Clinton’s 1.382. Less than a percentage point difference. It’s quite a turnaround in the difference points, the winning party, AND the voter turnout from 2012. In ’12, Obama carried the state with 1.620 million votes and Romney ended with more votes than Trump with 1.407. Total votes in 2012 topped 3 million which 2016 did not. 3rd parties carried much more in the ’16 election; Gary Johnson received over 3% of the vote himself.
Wisconsin changed its voting laws (like many states, specifically swing states, hint hint) between the elections which could account for the lower turnout but it’s not hard to see the voter suppression efforts in effect here. The difference in Democrat votes was almost 250,000—the size of Madison. And the Gary Johnson effect doesn’t quite explain it, he was the Libertarian, theoretically taking votes from the extreme right (though I can imagine some Sanders votes going to him).
Going back to Cambridge, I included this here to show both the horizontal (total number of data points) and vertical penetration (specificity and ability to micro-target) of what a firm can do with data. Once they have that plan, it’s just a matter of finding the place to do it. And luckily, they didn’t have to look far. There was one platform, Facebook, that reached nearly everyone in America in 2016 and though it’s main platform has taken a bit of a reputation hit, still reaches just about everyone in 2020 with the additions of Instagram and Whatsapp. Moreover, with Twitter and others banning political ads, it just gives a campaign more budget to go all-in on Facebook.
The Role Of Facebook
I think more than ever that Facebook will have a soured legacy—something not far from the tobacco industry but not so directly misunderstood. Put simply, it’s a powerhouse that acts recklessly, apologies minimally, and has a huge stake in how its perceived. It’s also massively wealthy and tied to some very important people.
Here’s Coppins on Zuckerberg’s captain-ing of how to move Facebook on after being implicated in the Cambridge stuff from 2016. (Keep in mind here that Zuckerberg alleges no wrong doing in the massive trove of data and access Cambridge got in 2016).
“After the Cambridge Analytica scandal broke, Facebook was excoriated for its mishandling of user data and complicity in the viral spread of fake news. Mark Zuckerberg promised to do better, and rolled out a flurry of reforms. But then, last fall, he handed a major victory to lying politicians: Candidates, he said, would be allowed to continue running false ads on Facebook.“
The bolding of the last line is mine but it needs bolding. Yes, Mark Zuckerberg did say that false ads could run. Not just political ads, ads that were paid for by now deep-pocketed political campaigns (POCs with unrevealed intentions even) and that those ads could be demonstrably false.
How could he allow this? Easy. The journalists will save us!
Coppins on Zuck again:
“In a speech at Georgetown University, the CEO argued that his company shouldn’t be responsible for arbitrating political speech, and that because political ads already receive so much scrutiny, candidates who choose to lie will be held accountable by journalists and watchdogs.”
Ahh, yes. Journalist and watchdogs! We love them. We love them to patrol the truth running around our company. And we trust them, with their small salaries to save the sanity, the trust, the very integrity of a platform run by a company with a market capitalization of $610 billion dollars ($100 billion more than the GDP of Argentina) and a median employee salary of $240,000.
But why should they be responsibly for someone lying on their platform. The press can help! The press will show who is being false and who isn’t. Except that our press is also being bought by those with a stake in politics (see above) AND the very way people GET to journalism is through social networks like Facebook where the LINK that would get them to the article is being micro-targeted to them based on their already-existing conditions.
Moreover, the press is the very institution that Trump’s warred with the most, and with struggling revenue numbers and both the consolidation of local news AND new “publications” popping up to spread disinformation (see above) the American press may well be at its weakest point in centuries. As it tries to be this arbiter of truth that Zuckerberg says it can and should be, it’s under threat from other public institutions.
And then you remember what I said above that the very mechanisms that one could and would use to reach the masses (a social network for instance) are the very ones bombarded with alternate press and disinformation. And, oh yeah, they’re making an insane amount of money off of this because they advertise TO you instead of printing journalism for you.
The short answers is that Facebook does not want to admit that its advertising is compromised in any way at all. This is the bread & butter of Facebook—where advertising in just last quarter brought in $16.6 billion in revenue.
All of that money comes BECAUSE the platform is able to micro-target. This is the whole point. And just as Cambridge used this to persuade people to vote or not in 2016, Facebook uses this to help businesses sell their product. They need to toe a line where they cannot say that micro-targeting changes elections entirely because they risk exposing the power (and potentially evil-ness) of their platform, but they have to say advertising is effective because it’s how they make money.
So they just say they’re not responsible and move on. Yikes.
Meanwhile, there’s a billion dollars from Trump (and untold billions more thanks to Citizens United) that is ready to promote disinformation on the most powerful of platforms exactly as that platform says it doesn’t care.
So Why Is The Election Over?
The question then is what this all means for November 2020. Because the Democrats can utilize these same tools, right? Absolutely, and they will. They already are in the primary—the cat is out of the bag and this the future of political campaigning.
But Trump will have the distinct opportunity of being the incumbent. That means a few things: (1) a headstart — so while the Democrats need to micro-target in primary states (and none have been remotely a swing state yet), Trump can focus on either flipping voters to his side in Michigan, Ohio, Pennsylvania, or Florida or just simply suppressing those who may not be exciting about the battling field of Democrats. And then (2) the incumbent has an advantage in politics simply by being known and not a mystery. This is part of the incumbent advantage phenomenon and it’s likely why the last four of five Presidents have served two terms.
I find these arguments to be, well, less arguable than what others have pointed out (things like Trump’s propensity for stretching the truth or playing on the fear of voters rather than presenting his actual record on events). But I won’t go into that further. I think the points above, and the focus of the entire Coppins article on ‘disinformation’ combined with the two points above show why this election may be over before we really even get started (or a Democratic candidate does).
Of course, as I qualified, there’s months for disaster to happen for the Trump administration that changes, swiftly, the minds of millions. So we’ll see on that.
For now, we’re looking at an Electoral College that rests on the back of what is likely less than a third of a thousandth of our country’s population, so a “war” of the kind described here represents a complete and existential threat.