Stories
Slash Boxes
Comments

SoylentNews is people

posted by Fnord666 on Tuesday December 17 2019, @12:34PM   Printer-friendly
from the browse-with-circularly-polarized-glasses dept.

How Facebook's Political Ad System Is Designed to Polarize

Amid the tense debate over online political advertising, it may seem strange to worry that Facebook gives campaigns too little control over whom their ads target. Yet that's the implication of a study released this week by a team of researchers at Northeastern University, the University of Southern California, and the progressive nonprofit Upturn. By moonlighting as political advertisers, they found that Facebook's algorithms make it harder and more expensive for a campaign to get its message in front of users who don't already agree with them—even if they're trying to.

[...] The paper, still in draft form, is a follow-up to research the group did earlier this year, which found that Facebook's algorithms can dramatically skew the delivery of ads along racial and gender lines even when the advertiser doesn't intend it. That's because while Facebook allows advertisers to design their audience—that's ad targeting—the platform's algorithms then influence who within the audience actually sees the ad, and at what price. That's ad delivery. Because Facebook wants users to see ads that are "relevant" to them, the algorithm essentially pushes a given ad toward users it thinks are most likely already interested in its message. This, the researchers found, can reinforce stereotypes. For example, of the users who saw ads for jobs in the lumber business, 90 percent were male, even though the intended audience was evenly split between men and women. (Facebook is also facing litigation for allegedly allowing advertisers to intentionally discriminate.)

For the new study, the team decided to explore whether the algorithm also skews political ad delivery along partisan lines. Because the company doesn't share that information, they had to run a number of experiments, essentially going undercover to figure out where targeting ends and Facebook's algorithms begin.

[...] What seemed to most bother the political strategists I spoke with was not so much the existence of that machinery as its invisibility. In one of the cleverest twists of the experiment, the researchers created a neutral voter registration ad that secretly served code to make Facebook think it directed to one of the campaign's sites. In other words, to users, the ad was completely neutral, but Facebook had been tricked into thinking it was partisan. Lo and behold, the skew was still there—and it could only have come from Facebook's end. And, significantly, it would indicate that the algorithm was determining the ad's relevance not by the content, but purely by who it thought was behind it.

"This ultimately comes down to a lack of honesty and transparency on the part of Facebook—and that is toxic for our democracy," said Betsy Hoover, a former campaign strategist and the cofounder of the progressive tech incubator Higher Ground Labs, in an email. If the platform is pre-judging which voters should hear from which candidates, regardless of the message, it could be locking campaigns into filter bubbles they aren't even aware of.


Original Submission

 
This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 1) by obfiscator on Tuesday December 17 2019, @01:41PM (12 children)

    by obfiscator (6554) on Tuesday December 17 2019, @01:41PM (#933251)

    I'm not sure I agree with the harm that this is causing. Could it be causing harm? Yes. Does the existence of biasing mechanisms guarantee it is causing harm, or that Facebook is at fault? No.

    From TFA: "For example, job ads targeted to both men and women might still be seen by significantly more men."

    If I choose to advertise in a given magazine, I might target an ad to both men and women. That doesn't guarantee that it will be see equally by both, depending on the exact distribution of readership for the period where my ads are running. And while this is obvious in some cases (e.g., I'm not going to target an ad to a general male audience if I'm advertising in Cosmo), it's not so in others (what's the distribution of male/female readership of Time or Newsweek?). The primary difference with online ads is that we have much easier ways to actually measure this, so the bias can be seen more clearly.

    It's clever how the researchers tested this and what mechanisms they found, but I don't think it's enough to claim the sky is falling yet.

  • (Score: 4, Insightful) by meustrus on Tuesday December 17 2019, @01:58PM (11 children)

    by meustrus (4961) on Tuesday December 17 2019, @01:58PM (#933255)

    The primary difference with online ads is that we have much easier ways to actually measure this, so the bias can be seen more clearly.

    Are you arguing that Facebook's algorithm isn't responsible for making these decisions? If it was just a cross section of Facebook users, one would expect everything to bias at the same level of "people who use Facebook". Facebook is not 90% male. Facebook decided what subset of their users would be interested in being lumberjacks.

    Advertisers used to have the choice of advertising for lumberjacks in Cosmo if they wanted to. Facebook seems to be taking that choice away.

    --
    If there isn't at least one reference or primary source, it's not +1 Informative. Maybe the underused +1 Interesting?
    • (Score: 0) by Anonymous Coward on Tuesday December 17 2019, @02:24PM (1 child)

      by Anonymous Coward on Tuesday December 17 2019, @02:24PM (#933265)

      Because Facebook wants users to see ads that are "relevant" to them,

      Because what are the odds that the Foxnews bunch would actually want to listen to what the Democrats want to say? Same goes for the the other side.

      one of the cleverest twists of the experiment, the researchers created a neutral voter registration ad that secretly served code to make Facebook think it directed to one of the campaign's sites. In other words, to users, the ad was completely neutral, but Facebook had been tricked into thinking it was partisan. Lo and behold, the skew was still there—and it could only have come from Facebook's end. And, significantly, it would indicate that the algorithm was determining the ad's relevance not by the content, but purely by who it thought was behind it.

      And guess who complains about "neutral content" just because they think the Russians are behind it?

      • (Score: 5, Interesting) by meustrus on Tuesday December 17 2019, @03:22PM

        by meustrus (4961) on Tuesday December 17 2019, @03:22PM (#933290)

        And guess who complains about "neutral content" just because they think the Russians are behind it?

        It used to be all red-blooded Americans. I'm not sure why some people in this country have adopted the orange one's love of certain pseudo-communist dictators, but it's frankly un-American.

        Not that I personally care whether the trolls are actually Russian agents. I much prefer the XKCD approach to content moderation [xkcd.com].

        --
        If there isn't at least one reference or primary source, it's not +1 Informative. Maybe the underused +1 Interesting?
    • (Score: 2, Funny) by Anonymous Coward on Tuesday December 17 2019, @02:54PM (1 child)

      by Anonymous Coward on Tuesday December 17 2019, @02:54PM (#933277)

      Facebook decided what subset of their users would be interested in being lumberjacks.

      Well, it seemed obvious, it isn't like they made an ad for lumberjills...

      • (Score: 1, Insightful) by Anonymous Coward on Tuesday December 17 2019, @03:17PM

        by Anonymous Coward on Tuesday December 17 2019, @03:17PM (#933287)

        This is why we need a "Groan" mod.

    • (Score: 1) by obfiscator on Tuesday December 17 2019, @04:12PM (6 children)

      by obfiscator (6554) on Tuesday December 17 2019, @04:12PM (#933312)

      I agree with you: Facebook's ads are not being delivered to a representative cross-section of Facebook's users.

      Let's consider Clear Channel Outdoor, though. They set up billboards on highways, and they also advertise on bus stops.

      I would argue (without any data to back me up on this), that the demographics of people who see those two ad locations are different.

      If you go to Clear Channel Outdoor and say, "I have this ad for lumberjacks", and they say, "Okay, we're going to put that on our billboards because lumberjacks all have cars and don't take the bus", is that different to what Facebook is doing? Your lumberjack ad is not being shown to a representative sample of Clear Channel Outdoor "users."

      Or does this only become a problem with the same company provides both the ads and the service, i.e. if a hypothetical Clear Channel company also controlled all means of transportation?

      • (Score: 3, Interesting) by meustrus on Tuesday December 17 2019, @04:34PM (2 children)

        by meustrus (4961) on Tuesday December 17 2019, @04:34PM (#933320)

        Is that how Clear Channel Outdoor works? They would make secret decisions about where to place ads based on their opinions about how effective those ads would be?

        That doesn't smell right. It costs them more to make that distinction, and there's likely no benefit because of how hard it is to measure the effectiveness of public-space ads.

        --
        If there isn't at least one reference or primary source, it's not +1 Informative. Maybe the underused +1 Interesting?
        • (Score: 1) by obfiscator on Wednesday December 18 2019, @07:55AM (1 child)

          by obfiscator (6554) on Wednesday December 18 2019, @07:55AM (#933652)

          Is Facebook making a secret decision, i.e., is the consequence intentional? Or is a by-product of algorithms that no one really understands (most black-box machine learning algorithms)?

          My original post was in response to the statement made in the title of this news post: "How Facebook's Political Ad System Is Designed to Polarize". After reading the articles, I came away with the impression that it wasn't designed to do that; it was an unintended consequence. And I tried to illustrate that with the Clear Channel example: sometimes, active decisions made can have unintended consequences, but that doesn't mean it's malicious.

          I am not fan of Facebook, and would not lose a minute of sleep if they disappeared completely. But, at best, I feel the articles demonstrate that we don't know if the system was designed that way or not. The articles demonstrate undesirable behavior, but do not demonstrate if it's intentional, except for maybe the latest research they did. However, the new paper is still on arxiv and has not yet been peer-reviewed, so I am not giving it the same weight as their other work. I am trying to figure out if I should increase my dislike of Facebook for this. I'm happy to do so, but I want to make sure it's warranted, hence why I am focused on the intent.

          Re-reading my original post, I may be drifting a bit. But that's fine, conversations evolve. If you're okay with it, I'm okay with it!

          • (Score: 3, Interesting) by meustrus on Wednesday December 18 2019, @04:04PM

            by meustrus (4961) on Wednesday December 18 2019, @04:04PM (#933755)

            I'm OK with it.

            It's reasonable to assume Facebook wasn't trying to make it hard for advertisers to reach their intended customers. They were trying to deliver ads to people who would like them. But there are some problems with that.

            Advertisers are often trying to reach new markets. An algorithm designed to serve the most palatable ads is obviously going to be skewed towards products someone is already familiar with, and is clearly counter to the goal of Facebook's advertising customers.

            This is really just an extension of how social media keeps its users in filter bubbles. We as a society have been talking about this for many years. If it wasn't intentional to begin with, it sure is by now.

            ...as for the headline, I always assume headlines have some amount of hyperbole, because they almost always do. "Designed to polarize" is just an edgier, politically-specific version of "Designed to not surprise".

            --
            If there isn't at least one reference or primary source, it's not +1 Informative. Maybe the underused +1 Interesting?
      • (Score: 2) by dry on Wednesday December 18 2019, @04:30PM (2 children)

        by dry (223) on Wednesday December 18 2019, @04:30PM (#933768) Journal

        As a driver, I see more ads on bus stops then I do billboards. This seems to be by design, the bus stop ads are pointed at traffic, traffic that often stops for lights and such close to the bus stop, which has its ad pointed at the traffic rather then the people waiting for the bus.

        • (Score: 1) by obfiscator on Thursday December 19 2019, @10:52AM (1 child)

          by obfiscator (6554) on Thursday December 19 2019, @10:52AM (#934146)

          That makes sense. It must depend on the bus stop, too. The bus stops I see around here have two ads: one facing in where people wait, and the other facing out. Some are clearly not visible from traffic based on the layout of the road. Not sure if that's intentional, or just a result of, "Let's use the same shelter everywhere". I imagine the advertising company is aware that some are not visible to drivers. I would be curious to know how (if?) they pass that information on to their customers, e.g., we give you 100 locations, 5% of which are not optimally laid-out for viewing from the road.

          • (Score: 2) by dry on Friday December 20 2019, @02:14AM

            by dry (223) on Friday December 20 2019, @02:14AM (#934477) Journal

            I do note that most ads I see at bus stops are general ads, things like internet, that can be aimed at both drivers and bus users, how they're sold, I don't know. Could also be regulations involved, I know billboards are generally illegal on public land here and bus stops are usually on public sidewalks. The bus shelters are all the same as well here.