Who Wants to #MAGA?

These are (allegedly) some of the ads that Russia's Troll-Farm was pushing on Facebook to try to influence the election. These ads were supposed to make "persuadable voters" vote for Donald Trump--apparently it worked.

Question One: Who Would Be 'Persuaded' By This?

What kinds of people would look at this and think "Yeah--that's the America I want."? Well, we have some clues. For example, a test was done where subjects were shown one of two pictures:

They were then asked about their feelings on housing assistance programs. The image of the black man greatly impacted responses among Trump supporters. Not only were they less supportive of housing assistance programs--but they also expressed higher levels of anger that people were getting assistance.

That's interesting. What else do we know about strong Trump supporters? Well, we know how they answered this Morning Consult poll question:

A poll of the various types of Trump Supporters identified five distinct categories--butthe two largest groups (American Preservationists and Staunch Conservatives), Muslim Bans and Racial Identity was pretty important:

In other words: the ads speak to people who are ideologically against Muslims and speak to a sense of a "white America."

Question 2: How Did Russia Know This?

If you were going to spend hundreds of thousands of dollars trying to influence an election, how would you know who to show these ads to? Does Facebook have a "Advertise to American Preservationists" option?

No--it does not. The Omnivore has advertised on Facebook to try to figure out how they see the world and that wasn't an option.

No, the secret is probably in Data Mining. Organizations like Trump's Cambridge Analytics were able to sift through Social Media and identify persuadable voters based on a variety of contextual clues in their profiles. This data was, apparently, ground-breaking and frighteningly accurate.

Could the Russians have gotten their hands on something like this and used it to target ads?

We'll probably never know for certain.