Thursday, January 26, 2017

Trump's Digital Campaign and The Effect of Social Media on Our Political Lives.

Below the break is a letter I sent to Democratic Chairman Jaxon Raven's on President Trump's digital campaign. It represents my concern that "Trump's campaign represents the most egregious confluence to date of money, big data, high level computation and neuropsychology". Chairman Ravens responded positively with concern to my letter noting that WA state Democratic party will be attempting to address these and other concerns very soon.  My own analysis is that the use of social media is that single most important driver of political activism in the United States.

My name is Ryan Ferris. I am the elected PCO for Bellingham (Whatcom County) Precinct 208. I am  member of the Whatcom County Democrat Campaign Committee.  I blog at  I have spent some time these last two months attempting to unpack the Republican national victory and DJT's Presidential victory. Below are thoughts and research on Donald Trump's digital campaign and questions on the implications of such for our party.

Thanks for your time,

Ryan M. Ferris
Democrat PCO Whatcom County 208

At least one part of this Presidential election deserves serious technical and political analysis: the effect of the Trump campaign's social media + data analysis + neuro-psychological operations upon the electorate. Obviously, both technology and psychology  have always had strong influences upon politics. But the combination of smartphones, data analysis, social media is clearly something much more than just a few steps beyond Thomas Paine and the printing press or Walter Cronkite and the television. Donald Trump, with the help of hedgefund manager Robert Mercer (1,2,3,4),  Reince Priebus, Cambridge Analytica, Alexander Nix, Steve Bannon, Jared Kushner, Brad Parscale (and apparently dozens of anonymous top tier data analysts and psychologists headquartered in San Antonio and London),  assembled a first class multimedia propaganda machine during this Presidential campaign. The story of amassing the "Project Alamo" database with 4K - 5K data points on 230 million U.S. adults is an important read. That story is told in these links here: (1,2,3,4).

I saw Trump's Facebook strategies play out every night in my feed.  I didn't realize how in-depth they were at the time. In the end, Trump's Facebook and big data campaigns were much more than fake news.  While Hilary Clinton's campaign was spending millions on TV advertising, securing Republican endorsements and receiving the largesse of  (anti-Trump) article after article in the NY Times and Washington Post, Trump's team was reaching deeply into the psyche of America with millions of dollars of  targeted Facebook advertising, raising hundreds of millions through small donations, and syncing his rally appearances and tweets with real time data analysis of potential supporters. His campaign wasn't created or run by impulsive madness. It only seemed so at the time to those of us who were Democrats. Trump's campaign was a tailored, purposed and aggressive reach into the subconscious of Americans, supported and backed by top tier data scientists and psychologists.  Again,that story is told in these links here: (1,2,3,4). Cambridge Analytica reputedly worked on 44 Republican campaigns in 2014 and 50 in 2016.  To get a brief but poignant summary of how Cambridge Analytica used big data and personality testing in the Presidential campaign, see CEO Alexander Nix presentation here:

An important part of this story is the effort to use online, targeted advertising to "depress the vote" of identity groups that could have voted for Clinton.  There is nothing new about voter discouragement. But the fabric of Facebook's architecture allows for outreach into identity groups and profiles with repeated depth and focus. The granularity of micro-targeting has to be looked at to be believed. I've looked at micro-targeted data for elections that I can't share.  The best public example of micro targeting research I could find online were ESRI (ArcGIS) Tapestry (1,2,3,4).  What I can say is that the firms that sell micro-targeted data specify individual profiles similar to the ESRI Tapestry 67 household segmentations. Two more (accessible) background articles and sites on data brokers are here: (1,2)  Another critical part of Facebook's reach is the explosion in global network data.  Mobile network data for 2016 was 6.2 exabytes (EB) per month. Cisco predicts 14.9 EB per month by 2018.  Over 4 billion smart phones will be in use by 2018, arguably providing the largest most connected network in human history. There are currently 1.66 billion mobile active users  (MAU) on Facebook.

I also wonder if anyone is thinking  or imagining how the next gen (VR) of social communication is going to effect politics. With VR (Virtual Reality) coupled with smartphones, simulated social engagement could have exceptionally powerful effects upon the electorate. Samsung is coupling VR (from Oculus Rift) with its latest cellphones. Some of the headsets are $99. In any case, Facebook bought Oculus Rift for $2B in 2014. Mark Zuckerberg (for example) is a big fan of VR and Facebook is developing VR as the next big social media experience. I think it is possible this tech will be here in a big way by election 2018.

Trump is a kind of penultimate opportunist and salesperson.  He reminds me very much of the floor traders I serviced for a number of years in San Francisco's investment banks: aggressive, focused, in the moment, conditionally reflexed for ruthlessness in sales opportunities. Trump takes care of his friends as he needs them, one of whom is Senator Doug Ericksen. Trump thanked the Senator by name at his Everett rally this year. Can we expect the "Project Alamo" database and Cambridge Analytica's "psychographics" to be used against us in LD 42 election in 2018?. Did some form of these tools already help Vincent Buys and Luanne Van Werven amass their huge vote totals in this last election?  How would we deal with that type of data power and reach in Whatcom County for off year elections?  

Here are my questions:
  • Does the Democratic Party have a committee that is assessing the effects of social media + data analysis + complex neuropsychologcial metrics and manipulation upon the electorate?
  • Are we responding to these new tools? 
  • Does the coordinated campaign in Seattle have matching weapons sets?
  • Is that something we should be developing at a local level?

There's a narrative that can be discerned among current news stories that describes how Donald Trump came to big data; a confluence of genius and money that was not entirely happenstance, but serendipitous all the same. However, there is also another more important thread: Trump's campaign represents the most egregious confluence to date of money, big data, high level computation and neuropsychology.  We can no longer stop the combination of these technologies being used to define electoral results.  These technologies are being used at local levels to commandeer red states or turn blue states red. 

Most of us know what the outlines of the Republican plan has looked since 2009:
(1) Work in the Rural areas first.
(2) Appeal to the disenfranchisement of  America's dis-empowered rural classes with various forms of media "programming" back by ALEC funded legislation and candidates.
(3) Win the Statehouses
(4) Win the Governorship
(5) Control the state, the budget the expenditures on law enforcement, the voting boards, the voting systems, voting requirements.
(6) Presto! Red state for life!!

I call this process "red locking" or more succinctly "Walkerization" in honor of what has happened to the previous populist and progressive Wisconsin. This process has happened not just across the South but in states like Wisconsin and Arizona as well.  I maintain that WA state is deep inside the process of being turned into a "red" state. It would be the first of the "Western Wall" (WA, OR, CA, HI) to fall.  It would be disappointing if there were an entirely new set of campaign tools in use by Republicans that Democrats do not yet have; especially if such tools depended less on issue engagement and more on personality testing coupled with social media outreach and if those tools were responsible for Democratic losses in 2014 and 2016.

Again, thanks again for your time,

Ryan M. Ferris

No comments: