Skip Navigation
Fellows

A Lie Just for You in 2020

Microtargeting is a powerful tool that leaves too much room for misbehavior and disinformation.

September 21, 2020
facebook
Shutterstock

Face­book recently announced that it would not allow new paid polit­ical ads in the week before Elec­tion Day. This is too little too late given the deluge of microtar­geted ads that have already inund­ated Amer­ican voters in the past year.

While broad­cast ads are still the norm, more candid­ates and polit­ical play­ers are opting for the narrow­cast so that they can reach you directly in your social media feed. During the second half of the 20th century, the main way that major party candid­ates for the pres­id­ency got their message out to nation­ally dispersed elect­or­ates was through broad­cast ads on radio and TV. But nearly as soon as the inter­net allowed, polit­ical ads have been moving online. And as plat­forms have become more soph­ist­ic­ated treas­ure troves of personal inform­a­tion, polit­ical ads are increas­ingly microtar­geted to subsets of the popu­la­tion. An expec­ted $6 billion is likely to be spent on in polit­ical campaign this cycle, and 20 percent, or $1.2 billion, will to go to digital ads.

So what’s so wrong with microtar­get­ing online polit­ical ads? The trouble is that when they contain misin­form­a­tion or disin­form­a­tion, they are the mental equi­val­ent of a drive-by shoot­ing: They are quick. They inflict damage, and then suddenly, they are gone.  

As I explained in my book Polit­ical Brands, “market­ing a [commer­cial] brand can be microtar­geted to a partic­u­lar demo­graphic, or income group, or subcul­ture.  For example, Dove soap, which targets women with images of empower­ment that are ostens­ibly against look­ism, and Axe soap, which targets men with sexist appeals, are both sold by the same company (Unilever), just to differ­ent target demos.”

The same thing can be done with polit­ical propa­ganda such that an untruth is repeated at a targeted subset of the elect­or­ate, until the idea is accep­ted as the truth. At the end of the day, Unilever wants to you to buy soap, and politi­cians want you to buy ideo­logy.

In one sense, target­ing key voters with tailored messages is not new. In the pre-inter­net age, polit­ical campaigns microtar­geted partic­u­lar voters using snail mail with messages that did not appear in broad­cast campaigns. So for example, if the Demo­cratic Party had a sense that a partic­u­lar Repub­lican house­hold was envir­on­mental because they subscribed to Back­packer magazine, that house­hold could be targeted by direct mail with messages that touted a candid­ate’s stance on the envir­on­ment. Or if the Repub­lican Party had a sense that partic­u­lar Demo­cratic house­hold was strongly pro-law enforce­ment because they had given to the Police Bene­vol­ent Asso­ci­ation, it could target tailored law-and-order messages to that home.

One factor that mitig­ated how much of this target­ing by mail could occur is that post­age costs add up fast. For instance, in 1984 both the Demo­crats and the Repub­lic­ans had racked up remark­able post­age fees during the elec­tion. But now, each digital ad can be sent to a voter at near zero marginal cost. And the effect­ive­ness of target­ing by street address can be under­cut by how mobile Amer­ic­ans are. By contrast, even when a woman moves from Topeka to Tall­a­hassee, her Face­book page (and other social media) is port­able. While mail to her old house may not reach them, ads on her Face­book pages will follow her around like barnacles on a ship.

How are voters targeted? There is a cottage industry of big-data wizzes who will sell polit­ical campaigns the abil­ity to microtar­get voters. Each company has its own secret sauce, but the basic compon­ents are similar: start with the voter file that is publicly avail­able from the state. This gives the name, address and party affil­i­ation (or lack thereof) of the voter. Add to that other publicly avail­able data, which in many states that will include marriages and divorcesgun owner­ship, or crim­inal records. In Flor­ida you find boat owner­ship if you have enough inform­a­tion.

Then added to those voter files is commer­cially avail­able inform­a­tion, which can include your web brows­ing history if you have not taken any privacy precau­tions online. And then there is the really creepy aspect that some polit­ical advert­isers will try to guess your person­al­ity type. This is what Cambridge Analyt­ica allegedly did for the Trump campaign in 2016.  This was after Cambridge Analyt­ica targeted black youth voters in in 2010 in Trin­idad with a campaign that discour­aged them from voting with a slogan of “Do so. Don’t vote.” The point was to try to increase voter apathy. That company is now defunct, but its owners have just moved on to start other compan­ies.

What the voter file allows the microtar­geter to do is to send targeted messages to divorced, gun-owning Flor­ida Repub­lic­ans with a habit on click­ing on Fox News stor­ies online. This is likely to very differ­ent from the message sent to single Demo­cratic voters who click on Pro Publica head­lines daily in Ohio.  And plat­forms like Face­book enable this type of gran­u­lar target­ing of voters. As MapLight explained, “Not only does Face­book allow advert­isers to spread polit­ical misin­form­a­tion, but it also allows them to target plat­form users on the basis of personal inform­a­tion such as age, gender, educa­tion, income, multi­cul­tural affin­ity, ZIP code, or interests. While microtar­get­ing can help NGOs, non-profits, and polit­ical chal­lengers to reach their audi­ences in a cost-effect­ive way, it can also enable foreign inter­fer­ence and voter suppres­sion.” 

Politi­cians are stereo­typed as talk­ing out of two sides of their mouths. But with microtar­get­ing messages on complex issues, politi­cians can send out a hundred differ­ent messages on the same topic, just microtar­geted to differ­ent constitu­en­cies. Think of health care — a politi­cian could put out dozens of vari­ations on messages on health care, just to differ­ent audi­ences. Microtar­get­ing may contrib­ute to polar­iz­a­tion as the algorithms in most social media plat­forms are designed to tell the viewer what she already wants to hear — confirm­ing bias after bias.

The less scru­pu­lous the politi­cian, the more danger­ous microtar­get­ing is. If the politi­cian is will­ing to lie to voters, then microtar­get­ing is gift that keeps on giving because now the politi­cian can lie person­ally to each voter in a tailored and manip­u­lat­ive way. If the politi­cian has no scruples, then why not tell iden­ti­fi­able pro-life voters that he is anti-abor­tion and tell iden­ti­fi­able pro-choice voters, that he is for repro­duct­ive rights? Tell the gun owners that he is pro-guns and tell the gun sense voters that he’s for restric­tions? This isn’t a hypo­thet­ical, as Face­book has contin­ued with its puzz­ling stance of letting politi­cians lie in paid advert­ise­ments on its plat­form.

In May, Media Matters excor­i­ated Face­book for allow­ing paid polit­ical ads that touted misin­form­a­tion about voting fraud. And a civil rights audit of Face­book criti­cized the plat­form for being “far too reluct­ant to adopt strong rules to limit misin­form­a­tion and voter suppres­sion.” And in July, ProP­ub­lica found polit­ical ads on Face­book making “unsub­stan­ti­ated claims that House Speaker Nancy Pelosi is trying to fix the elec­tion using mail-in voting.”

Finally, another differ­ence between direct mail target­ing and digital target­ing is that at least with mail, there is a phys­ical arti­fact that a voter can show the media if there was a partic­u­larly mislead­ing or obnox­ious message from a polit­ical campaign. But with most digital ads, until very recently they were ephem­eral.

In 2018, Google launched its Ad Library of polit­ical ads, and in 2019, Face­book finally launched a library of polit­ical ads. These are steps in the right direc­tion, although real­ist­ic­ally, most users won’t research the polit­ical ads that just shocked or titil­lated them online.

In 2020, more voters should flag troub­ling or pecu­liar digital polit­ical ads they see in their social media feeds to the media — espe­cially if the ads seem to discour­age voting.

The views expressed are the author’s own and not neces­sar­ily those of the Bren­nan Center.