Whole Cit­i­zens

11:28AMFebruary 18 2018Daniel Tompkins

Archive

An­a­lytic Ac­tivism

In the pre­vious ar­ticle, I took a look at David Karpf's An­a­lytic Ac­tivism— ex­am­ining how po­lit­ical cam­paigns have re­acted to a hy­brid media en­vi­ron­ment.1 In that ar­ticle, I also imag­ined a public cit­izen, a par­rhe­si­astes, to par­tic­i­pate in the es­tab­lish­ment of a dig­ital public do­main for vi­brant po­lit­ical speech and dis­cus­sion be­yond so­cial me­dia's "echo cham­bers"— dis­parate in­for­ma­tion bub­bles fil­tered of ide­o­log­i­cally op­po­si­tional con­tent.2

Karpf's col­lec­tive analysis of how media and an­a­lytics func­tion within the modern po­lit­ical arena is in­cred­ibly in­sightful and en­gaging— but how did we get here? Ad­di­tion­ally, how can past po­lit­ical cam­paigns func­tion as models for the strategic im­ple­men­ta­tion of emerging tech­nolo­gies in fu­ture races and ac­tivist move­ments?

Pro­to­type Pol­i­tics

Daniel Kreiss, in Pro­to­type Pol­i­tics: Tech­nology-In­ten­sive Cam­paigning and the Data of Democ­racy (TICDD), does well to ad­dress these ques­tions— analysing the his­toric adop­tion of the In­ternet and the evo­lu­tion of its in­volve­ment in po­lit­ical races.3

Kreiss looks specif­i­cally at the transfer of knowl­edge from one cam­paign to the next, and at how in­di­vidual cam­paigns can func­tion as ex­plicit pro­to­typ­ical models for fu­ture elec­tions— es­pe­cially as their re­la­tion­ship with Karpf's hy­brid media en­vi­ron­ment con­tinues to mu­tate.

This in­volves a crit­ical ex­am­i­na­tion of the po­lit­ical party (as well as third-party po­lit­ical data con­sul­tan­cies) as a vessel for car­rying in­for­ma­tion from one cam­paign to the next:

During and after elec­tions, par­tic­ular cam­paigns are trans­formed through meaning-making processes into pro­to­types for some ac­tors, a model for fu­ture cam­paign prac­tice, and a set of claims about the world that are ac­tion­able for prac­tioners.4

The bridging of this in­for­ma­tion— how an­a­lytic pro­gram­ming, or­ga­ni­za­tional knowl­edge and con­stituent data moves be­tween races and other po­lit­ical move­ments— is a key theme throughout TICDD.

Lastly, Kreiss breaks down the pur­suit of tech­nical teams in iden­ti­fying the whole cit­izen. This is the con­cept of how dis­parate pieces of in­for­ma­tion are cap­tured from the spec­trum of hy­brid media to pro­duce a more pre­dictable, or wholistic, model of an in­di­vidual voter:

Prac­ti­tioners seek to be able to rep­re­sent whole cit­i­zens through data as a way of re­lating to and lever­aging their media use, psy­cho­log­ical dis­po­si­tions, and so­cial re­la­tion­ships for elec­toral ends.5

Com­plex Net­works

In César Hi­dal­go's class: Net­works, Com­plexity and Their Ap­pli­ca­tion, we've been taking a look at the math­e­mat­ical models de­scribing net­work topog­raphy and be­havior.6 Our read­ings cover a breadth of pro­fes­sions and topics, from viruses and epi­demi­ology to dig­ital and so­cial net­works; how­ever, what is quite in­cred­ible about net­work sci­ence is that it cre­ates very ac­cu­rate models that are ap­plic­able across a broader range of dis­ci­plines.

On that point, I think it's in­ter­esting to un­der­stand the in­ternal or­ga­ni­za­tional net­works of cam­paigns in ad­di­tion to their re­la­tion­ship with voter data and the con­stituents them­selves— specif­i­cally, how they or­ga­nize and be­have as net­works.

There are a few in­ter­esting prop­er­ties in re­gards to the "scale-free" net­work model— a net­work in which the de­gree (con­nec­tivity) of nodes fol­lows a power law dis­tri­b­u­tion.7

  1. One prin­ciple of a scale-free net­work is that new links that are added will nat­u­rally show "pref­er­ence" for more highly-con­nected nodes, or hubs if you will. Put an­other way, the ex­isting nodes with more links are more likely to af­ford con­nec­tivity to newly added nodes. This is part of the theory of pref­er­en­tial at­tach­ment.

This makes a lot of sense if you imagine Google or Face­book as highly-con­nected hubs of the World­WideWeb. If anyone is putting a web­site onto the In­ternet— per­haps con­necting for the first time; then, they are much more likely to con­nect to these al­ready highly-con­nected and ac­tive hubs.

Sim­i­larly, web­sites that don't re­ceive a lot of at­ten­tion are un­likely to form con­nec­tions with new nodes.

  1. An­other quality of the scale-free net­work is that (as­suming some con­stant of growth) the age of the node has been found to cor­re­late with its number of con­nec­tions. That would in­di­cate that the sooner a node has been es­tab­lished within a given net­work, the greater chances it af­fords for fur­ther con­nec­tions.

Ap­plying Net­work Logic

It's re­mark­able how rel­e­vant these two simple prin­ci­ples are when ap­plied to po­lit­ical cam­paigns— par­tic­u­lary, given that some of this re­search in net­work sci­ence had been es­tab­lished decades prior.

Looking at the broader so­cial and dig­ital net­works of in­for­ma­tion and people that are cre­ated over the du­ra­tion of a po­lit­ical cam­paign, there ap­pears a likely can­di­date for topo­graphic and be­hav­ioral analysis.

In fact, I would argue that these net­works are the essence of the po­lit­ical race— which is es­pe­cially re­mark­able con­sid­ering the rapid rise and fall of this in­fra­struc­ture from elec­tion to elec­tion.

In this sense, you might say that the cam­paign it­self is re­ally a ma­chine for lever­aging this net­worked space— of in­for­ma­tion, people, and skills. An in­ter­esting as­pect of con­cep­tu­al­izing the cam­paign in this way is in the emer­gence of a set of uni­versal prin­ci­ples that govern its be­havior like in net­work sci­ence.

In a way, this it how Kreiss writes Po­lit­ical Pro­to­types— looking not only at the win­ners, but finding cor­re­la­tions in the tac­tical methods, hi­er­ar­chies, and ef­fi­cacy of cam­paigns (along with other be­hav­ioral qual­i­fi­ca­tions).

The models Kreiss fo­cuses on— Howard Dean's pres­i­den­tial race, Mc­Cain, Oba­ma's 2008 and 2012 cam­paigns, and Hillary's run in the 2008 pri­mary— all ex­hibit unique and evolving in­ter­faces with media, con­stituents and their data.

Re­call the second pre­dic­tive prin­ciple of scale-free net­works. We would ex­pect that the sooner a cam­paign is able to begin de­vel­oping their mar­keting strategy— building a brand, a nar­ra­tive, and a data­base of people, the better its chances of reaching a larger au­di­ence (forming more con­nec­tions).

Ad­di­tion­ally, as I noted ear­lier, the De­mo­c­ratic and Re­pub­lican Com­mit­tees play sub­stan­tial roles as data vaults for their re­spec­tive can­di­date. Con­sid­ering the DNC or the RNC as a highly-con­nected hub— and re­calling now both prin­ci­ples of the scale-free net­work, it would seem that these user pro­files should play a tremen­dous part in the out­come of a race.

The problem, which Kreiss re­peat­edly em­pha­sizes, is that the data often ne­ces­si­tates a time-in­ten­sive trans­la­tion as it's handed off from one or­ga­ni­za­tion to the next— each with its own unique work­flow and de­vel­oper stack.

The Strength of Weak Ties

In an in­ter­esting net­work sci­ence ar­ticle, The Strength of Weak Ties (1973), the au­thor— Mark Gra­novetter — wrote:

Some have in­di­cated that early in­no­va­tors are mar­ginal, that they 'un­der­con­form to norms to such a de­gree that they are per­ceived as highly de­viant'. Others find that those named more fre­quently adopt an in­no­va­tion sub­stan­tially ear­lier.8

This is given against the "per­ceived risks of adop­tion of a given in­no­va­tion"— a struggle cen­tral to the po­lit­ical race.9 Kreiss, as well as Karpf, makes it clear that the po­lit­ical pro­to­type can be used to more strate­gi­cally draw con­clu­sions from pre­vious cam­paigns.

In this way, it can serve as a basis for the evo­lu­tion of dif­ferent dig­ital so­cial media plat­forms and the force of their in­volve­ment in the po­lit­ical race. De­spite this an­a­lytic in­sight, its ex­tremely dif­fi­cult to ag­ilely adapt to this changing en­vi­ron­ment.

If, for ex­ample, you look at Google's un­suc­cessful chal­lenge to Face­bookGoogle Plus, it can be a bit of a gamble with the at­ten­tion and time of the tech­nical team versus the po­ten­tial suc­cess of the plat­form.

Of course, this is a risk that has won races. Oba­ma's nick­name, the first "YouTube" Pres­i­dent— or the "Face­book" Pres­i­dent— came about be­cause of his rel­a­tive mas­tery in adapting to these media, adopting them to pop­u­larize his cam­paign (and him­self as a can­di­date).

Going a step fur­ther, these net­work the­o­ries could even be used to an­a­lyze the vi­a­bility of these po­ten­tial plat­forms for ef­fec­tively dis­sem­i­nating in­for­ma­tion on the progress of the cam­paign, the can­di­date, and their views on policy.

Mar­keting groups em­ploy ac­tors, mu­si­cians, and ath­letes to en­dorse a product be­cause they are es­sen­tially highly-con­nected nodes that are now con­nected to that product— and will en­courage fur­ther con­nec­tions through this pro­mo­tion.

In the same way, pres­i­den­tial can­di­dates must push to be­come these highly-con­nected nodes. Celebrity politi­cians, like Reagan or Trump, al­ready have a pow­erful ad­van­tage through their pres­ence and en­tan­gle­ment in this net­work and its di­verse media.

Others will ded­i­cate years of work to building these con­nec­tions— em­bed­ding them­selves across dif­ferent so­cial cir­cles, and con­structing phys­ical and dig­ital mailing lists.

Po­lit­ical "Sci­ence"

Of the more sci­en­tific— per­haps ex­tended to the al­go­rith­mi­cally de­rived— models of net­works and voter be­havior, the po­lit­ical writer Sasha Is­senberg quotes the con­ser­v­a­tive opinion of Prince­ton's then-pres­i­dent, Woodrow Wilson:

I do not like the term 'po­lit­ical sci­ence'. Human re­la­tion­ships are not, in any proper sense, the sub­ject matter of sci­ence. They are the stuff of in­sight, and sym­pathy, and spir­i­tual com­pre­hen­sion.10

Throughout the ma­jority of ar­gu­ments and analysis on po­lit­ical cam­paigning, it seems, the dis­cus­sion of ethics and morality have been ab­sent. I think Wilson brings up an in­ter­esting point. Even in the ver­nac­ular lan­guange of cam­paigns— "whole cit­izen", or "per­suad­able voters"— there is an ob­jec­ti­fi­ca­tion of the con­stituent.

Value is af­forded to cit­i­zens only if they're able to pro­vide one of two things: a do­na­tion, or a vote. In fact, I found it dis­turbing to re­alize that it's common prac­tice for many cam­paigns to sup­press the vote of people who are un­de­cided, but un­likely to vote in the can­di­date's favor— they'd rather they didn't vote at all.

It seems as though, in the rush of tech­no­log­ical in­no­va­tion and al­go­rithmic de­ci­sion-making, cam­paigns have di­min­ished the "whole cit­izen" to a set of sta­tis­tics.

Mer­i­toc­racy

On the other side of the spec­trum, working in con­tem­po­rary trading and eco­nomics, Ray Dalio has imag­ined that qual­i­fying people into some­thing like a mer­i­toc­racy would create enor­mous ef­fi­ciency in the work­place— en­cour­aging ex­pert man­age­ment and pro­duc­tion.

Along with Alexander Nix of Cam­bridge An­a­lytica— which has de­vel­oped a sim­ilar psy­cho­graphic mod­eling tech­nique for iden­ti­fying per­suad­able voters (in the Brexit Leave cam­paign, and ad­mit­tedly in the 2016 pres­i­den­tial elec­tion, here in the US)— this lack of eth­ical con­sid­er­a­tion ap­pears in­sen­si­tive to di­ver­sity and per­sonal unique­ness.

Ad­di­tion­ally, this same model has been used as a pow­erful de­mo­graphic sur­veyor for more ef­fec­tive pro­pa­ganda. These same tech­niques have been de­vel­oped for "pyscho-so­cial" war­fare: We're seeing it now in the US, in the up­roar over the Russian "trolls".

I think the way in which this in­for­ma­tion is so pub­licly trans­parent— pub­lished ar­ti­cles on the SCL web­site— makes it seem as though this is non-lethal. How­ever, as in­for­ma­tion about the role of for­eign in­ter­fer­ence reaches fur­ther au­di­ences, it seems clear that some major re­form is needed.11

The Guardian put out an ar­ticle last summer with a great quote given by Alex Younger, head of MI6, in De­cember of 2016. I think coming from his role of au­thority, it should be taken quite se­ri­ously:

“The con­nec­tivity that is the heart of glob­al­i­sa­tion can be ex­ploited by states with hos­tile in­tent to fur­ther their aims.[…] The risks at stake are pro­found and rep­re­sent a fun­da­mental threat to our sov­er­eignty.”12

Footnotes

  1. Karpf, David. "An­a­lytic Au­di­ences." An­a­lytic Ac­tivism, 2017, 93-122. doi:10.1093/​acprof:oso/​9780190266127.003.0004.

  2. Karpf de­tails these echo cham­bers in his book (cited in note 1).

  3. Kreiss, Daniel. Pro­to­type pol­i­tics: tech­nology-in­ten­sive cam­paigning and the data of democ­racy. New York, NY: Ox­ford Uni­ver­sity Press, 2016.

  4. Ibid. 3

  5. Ibid. 3

  6. MAS.581 Net­works, Com­plexity, and Their Ap­pli­ca­tion, Cam­bridge, MA. Class, MIT Spring 2018. Pro­fessor Cesar Hi­dalgo.

  7. Power law dis­tri­b­u­tion is a dis­tri­b­u­tion graph (com­pare it to more reg­u­larly ref­er­enced "bell curve") that ex­hibits higher dis­tri­b­u­tion to­ward the y-axis as­ymp­tote (ex­tremity). One ex­ample is the dis­tri­b­u­tion of wealth in the United States. A 2015 ar­ticle pub­lished by For­tune Mag­a­zine cited a study es­ti­mating that "the wealth­iest 160,000 fam­i­lies have as much as the poorest 145 mil­lion fam­i­lies.

  8. Gra­novetter, Mark S. "The Strength of Weak Ties." Amer­ican Journal of So­ci­ology 78, no. 6 (1973): 1360-380. doi:10.1086/​225469.

  9. Ibid. 8

  10. Is­senberg, Sasha. The vic­tory lab: the se­cret sci­ence of win­ning cam­paigns. New York: B \ D \ W \ Y, Broadway Books, 2016.

  11. SCL, Strategic Com­mu­ni­ca­tions Lab­o­ra­to­ries. A rel­a­tive of Cam­bridge An­a­lytica serving "gov­ern­ments and mil­i­tary or­ga­ni­za­tions world­wide" (meta-de­scrip­tion, taken from https://​web.archive.org/​web/​20170613170147/​https://​www.sclgroup.cc/​home ); how­ever, many of the client-target re­la­tion­ships have been wealthy first-world to volatile third-world.

  12. Cad­wal­ladr, Ca­role. "The Great British Brexit Rob­bery: How Our Democ­racy was Hi­jacked." The Ob­server, May 7, 2017. Ac­cessed Feb­ruary 26, 2018. https://​www.the­guardian.com/​tech­nology/​2017/​may/​07/​the-great-british-brexit-rob­bery-hi­jacked-democ­racy .