What happens to media as the digital age enhances their ability to engage consumers? As a part of our “Lost in the Super Market: Navigating the Digital Age” collection, American Compass convened contributing writers Matthew Crawford (UVA’s Institute for Advanced Studies in Culture) and Peter Suderman (Reason) to discuss the ramifications of the “Attention Economy”. Wells King (American Compass) moderates.
Transcript
Wells King: Matt, let’s start with you. Your essay expresses relatively more concern about the ways in which the internet and the digital economy commodify attention. Could you just explain a little bit more? What is the attention economy and what sort of challenge does it pose?
Matt Crawford: Yeah. I guess the attention economy isn’t simply a function of the internet or anything digital. In fact, this is a problem I first started thinking about really in a different setting. I was once in the supermarket checkout line, and you know how there’s the little machine where you put your card in to pay for your groceries and then you confirm the amount and then you enter your pin? Anyway, there’s a few steps like that. Well, during those intervals when I’m staring at the screen waiting for the next step, it’s showing me advertisements because some genius figured out that someone in that position is a captive audience. That’s when I started just to notice this phenomenon everywhere.
When you go through airport security, the bottoms of the bins are often papered with advertisements, which can make it pretty easy to miss your keys or a thumb drive against the visual clutter. Then I pull down the tray from the seat in front of me on the airplane and the top of it is papered with an ad. This seemed to peak maybe, I don’t know, 10 years ago, but you just have this sense that every available moment was getting monetized, any moment that might otherwise offer a bit of private headspace. I think there’s a bigger backdrop for this that has nothing to do with the internet, but it’s simply a logic of escalation.
Peter Suderman: Yeah, I think that’s right in many ways and it’s an interesting point. You can think of even stuff like sports stadiums, NASCAR. The NASCAR cars are covered in their sponsor logos. I’m not a big NASCAR watcher, but my recollection is that even the drivers’ uniforms are all just patched up with this sort of thing. This seems to me to be not that big a problem. That’s who’s paying for it and in some ways that offsets the cost of quite expensive vehicles, which are very expensive to maintain and to run, and it makes it possible to have the sport and to have it be available and be a mass popular product.
In the digital world, there’s actually been a backlash to this. You said that the peak was maybe a couple of years ago. I think we’ve actually seen some of that, some of this sort of attempt to advertise at least in every spare minute of your time come back because now services are offering the opportunity for people to say, “Hey, I don’t like that,” and say, “Well, okay, you can pay us instead.” Instead of charging you effectively your time, they are charging you money. Spotify offers an ad-supported version for free or they will offer a paid version where you never hear an ad. The same goes for Hulu and several of the other big streaming services.
I think this just seems to me like a natural evolution of the market that isn’t really that big a problem, because wherever there is time and wherever there is attention, people are going to try and figure out how to monetize it to some extent, because I think one of the things that we ended up agreeing on, maybe not all that surprisingly, in our essays was that time is valuable and it’s going to be valuable to you and to every individual and it’s going to be valuable to others. Trying to figure out how to spend that time and how to make use of it is the challenge for every individual, but it’s also the challenge for every organization that is trying to make money off of their product.
Matt Crawford: Yeah. Just to pick up your point and refer it back to the airport example, in the airport there’s this chattering of CNN that can be hard to avoid. You can avert your gaze from it, but the fields of view that haven’t been claimed for commerce seem to be getting narrower and fewer. In an airport, unlike say a NASCAR stadium, it is a public accommodation in the sense that you have to pass through it in order to just conduct the business of life.
Now, interestingly, I once had the great pleasure of being in the business class lounge in Charles de Gaulle Airport in Paris. When you walk into that space and the airtight doors whoosh shut behind you, the difference is amazing. The only sound I heard was the occasional tinkling of a spoon against china. There were no TVs or no ads on the walls. I think you can understand this as that there’s something like an attentional commons, and as the commons get appropriated for private gain, now, again, I’m talking about a basically public facility like an airport or a train station, if you want your attention back, you going to have to pay for it. You’re essentially leaving the commons for a private club, which is what the business class lounge is.
When you think of it in these terms and you think of attention as a resource, it seems to me we need a political economy of this resource and we need a concept like the attentional commons, because absent protection by some kind of, I don’t know, limiting principle, you do have an incessant cacophony that just impinges on you.
Peter Suderman: I take your point to some extent. There’s advertising everywhere, even in commonly used public spaces. At the same time, I think these things are easier to ignore than maybe you’re suggesting here. I distinctly remember being in Atlanta Hartsfield Airport, which is not an advertising-free zone in any sense, and missing a flight literally because I was so engrossed in an article in the Atlantic.
I was reading good text, good pros, and I wasn’t paying any attention to CNN on the airport or to the basic fact that Coca-Cola basically sponsored every single object in the building because it’s Atlanta, because I was able to direct my attention elsewhere because something else was engaging it. That brings me back to another one of the points that I made in my piece, which is that I think that anybody who is trying to do business, trying to make a living, even just trying to write is trying to engage people’s attention. Right? We write in ways that are designed to lure people in, to engage people, to get them to read the next sentence, the next page. In some ways that is just what advertisers are doing, that is just what social media is doing, is they want you to stay with the product. When I write, I want readers to stay with the thing that I am creating. When Facebook creates a new feature, they want people to stay with that feature that they are creating because that’s the business they’re in just as the business I’m in is the business of writing.
Wells King: Now, Matt, you highlight in your piece the ways in which tech firms don’t merely try to capture attention but try to actually capture data and then manipulate behavior from that. Is it fair to say that the digital economy then, in the way that it captures attention, is somehow novel, and the way that these firms are using it is novel and something of a unique challenge?
Matt Crawford: I think so, and I think we can get to that point, but a good segue would be to respond to Peter’s last point. As a writer, you’re producing content that you want to be engaging and appealing, so there is a kind of continuity there, I guess, with digital content.
Now, when we talk about the platform firms such as YouTube and Twitter, the entire premise of there being unregulated, under Section 230 immunities, is that they are not producers of content and that they don’t exercise editorial control. What we’re talking about when we talk about wanting to regulate those things is quite different than talking about someone writing a novel.
What is the problem then? There’s an entire discipline called persuasive design that’s quite prominent that you can take courses on it at Stanford University. The point of persuasive design is to maximize time spent on a platform. As it turns out, the way you do that is through the techniques of behavioral conditioning. It turns out that random reinforcement with little rewards offered at different intervals can induce a kind of compulsive behavior. Now, within the ambit of classical economic thinking, there’s no way to really account for the distinction of compulsive behavior other than just another instance of preference satisfying behavior.
Let me just give one example. Oh, no. Let me stop there because I imagine Peter might want to respond to that.
Peter Suderman: I think just one small thing is that Section 230 is often misinterpreted and it’s really just a law that says that neither companies nor individuals acting online assume legal liability for content created by other people, and it really doesn’t have anything to do with a publisher or platform distinction or with a lack of or political bias or anything like that. It’s just whether you’re a big company or a small company or an individual, content that is created by someone else you don’t have any legal liability for. It’s actually, it’s sort of simpler and narrower in some ways than a lot of people seem to think. But in terms of your description of behavioral targeting and design is interesting. I think obviously the tech folks study it and are interested in it because they want people to engage with their products. I would just say two things in response to that.
One is there’s a really fine, and maybe I’m not even sure how much of a line at all, between the kind of behavioral design that you’re talking about that seems a little bit scary because it’s designed to hook people and just good design that is designed to make things useful and make people like using products and make them good to use because a product that is functional and effective and good to use as one that is enjoyable to use and also one you’re going to keep coming back to, and so it’s going to lure you in, in some sense, but it’s also, still, I think, not all that different from analog media. It may be a little more advanced, and the people who are designing these things may think about them a little bit differently, but you can go back to the era of movie cereals and remember that what they learned was that if you designed those cereals in a certain way, if you wrote them in a certain way each week, you could bring people back with a cliffhanger at the end.
What is a cliffhanger if not an attempt to induce a behavior in the viewer who is also the customer who is someone who they want to come back and pay and pay their money and sit in front of that screen again and again and again. I just don’t see this as being all that novel. I think even some of the people inside the industry, inside the tech world sometimes think of it as novel, they’re wrong in some ways. This is just people designing products that are designed to be used, designed to be useful, and designed to be engaging for customers, viewers, listeners, whoever it is who’s actually using the product.
Matt Crawford: Okay. Well, I guess a couple of things to say. I want to come back to your point about a kind of historical continuity because I think that’s really interesting and whether there is any kind of break and how to locate that, but as to the idea that what we’re talking about is products that are engaging, and so we’re happy to pay our attention to them willingly, I think if we’re careful… What is the product here, and who is the customer? When you’re talking about the platform firms, you’re not the customer because you’re not paying for this, so what is the product and who is the customer?
Well, the product is predictions about your behavior and fine grain characterizations of what kind of proclivities you have. The product is that based on that accumulation of behavioral data that you generate when you’re online. The customer is any firm that can use these predictions to try to steer and nudge your behavior in some way. To say that the content is appealing, that is the condition for this market working, but since I’m not the customer, the appealingness of the product isn’t the dimension on which this market’s working.
Now, as you say, there’ve been, I mean, there’s been advertising for, what, more than a hundred years of… at least a couple hundred years, I would say, and so, yeah, there’s been a refinement of techniques. Again, I guess there has to be some kind of content that is appealing that is distinct from, well, is it the meta-level in which the capturing of your attention is the basis for then having this third thing, this third party come in, which is the advertiser, and so you’re making the viewer of the content available to the advertiser. All of this is, I think, quite distinct from… In your piece, you draw parallels with novel writing and alarms over novel writing as corrosive. Do you want to just rehearse that for us and-
Peter Suderman: Yeah, I mean, there were concerns even back almost 200 years ago about young women spending way too much time reading novels. This was sometimes a part of Jane Austen’s fiction and she was writing about, but just of that era, there was the sense that young women had become too engrossed in novels, that they were just sucking people’s time away from them and taking them away from their real world. I think there is a real parallel there to the kinds of complaints that we hear now about young people spending too much time online, whether it’s with video games or with TikTok or whatever the social media platform or of choice is at the moment is, that there’s always, I think, adult concern about how adolescents are spending their time.
Usually, because adolescents are spending their time in some way that seems wasteful, that seems a little bit socially maladapt, that just seems like it’s not productive, like it’s not helping anything, and like you’re… and so it just gives rise to complaints that kids are riding their brains doing whatever, playing pinball, reading science fiction novels or comic books or Jane Austen novels. I think in a lot of ways, the concern about internet use, especially amongst adolescents, is driven by some of that same sort of generational separation between adults who didn’t grow up with this stuff and kids who just see it as part of their everyday. You can see that in some of the political rhetoric, particularly from lawmakers like Josh Hawley.
Matt Crawford: Yeah. Okay. Maybe fuddy-duddies have always been saying, “The world is going to hell in a handbasket.”
Peter Suderman: Or at least that teenagers are doing things that we don’t like and are destructive to their minds and corrosive to society. I just don’t see it as particularly new.
Matt Crawford: Yeah. Well, I’ll tell you this. If you try to get a 10-year-old or a 13-year-old interested in literature, it’s a bit of an uphill battle… I don’t know if you have kids… whereas if you leave the kid in a room with an iPad and walk away and come back a couple hours later, I guarantee you, when you come back, the kid is still going to be on that iPad. These two facts are not independent of each other. I mean, maybe it’s only after enduring crushing, crushing boredom that you finally pick up that Jane Austen or Moby Dick or whatever, but the availability of easier enthrallments that don’t really make much demand on you can tend to short circuit that educational process.
They would have said, “Well, that’s a meaningful distinction, and therefore, one of them we should be worried about.” It turns out that kids who read comic books when they’re kids end up being better readers as adults, and it’s actually pretty productive, and we should want kids to be reading comic books rather than reading nothing if we think that reading is something that we should value. I’m just not completely convinced that the concern about iPhones and iPads and the internet is really all that different.
But I actually want to go back to another point that you made just a little while ago about who the real customer here is because you said, “Well, on Facebook, you’re not paying for anything. You’re not the real customer.” I mean, I think that’s an interesting point in a lot of ways, but what you’re just describing is a two-sided market. We’ve had those in media for a long time. In particular, we’ve had them in the world of newspapers and magazines, and so even back when most newspapers and most magazines weren’t online or were only trivially online in the 1990s, and you might have to pay 3 or $4 at the newsstand stand for an issue of time magazine or the Sunday New York Times, the price you were paying for the print edition was just a tiny, tiny amount of the revenue that those organizations were picking up at, in some cases didn’t even literally cover the cost of printing and paper. They made all of their money off of selling the eyeballs of their readers to advertisers, to a first approximation.
There’s this a slight exaggeration, but they effectively made all of their money off of selling there, and then they would sell also, they would sell their lists if they had mail subscriptions, and so they would divide up those lists by the information that they could glean about their readers, and they would sell that off to advertisers. That’s just what Facebook is doing. That’s just what some of these other digital media companies are doing. They’re doing what TIME magazine and Newsweek and USA Today have done for decades. I just don’t see a huge difference. I just think two-sided markets have been around for a long time and to be concerned that two-sided markets are a novel thing that came about when Mark Zuckerberg invented Facebook is I think not the right way to view that kind of market or that kind of media. And these are all in some ways forms of media. The New York Times and Facebook are obviously different in crucial ways, but one is social media, the other one is prints news media, which is now mostly online.
Wells King: Well isn’t it the case though that one of the real key distinctions here is that these are platforms? And so not only do they operate as kind of natural monopolies, but also by virtue of some of their computing power, the way they’re able to analyze these data, that they also exercise a pretty novel degree of corporate power. Is that a meaningful distinction, Peter?
Peter Suderman: So I think the word platform is a completely reasonable word to use colloquially to describe Facebook or Twitter or any of their peer organizations. It has no legal meaning. And so this is what I was saying about 230 is that there is no platform publisher distinction built into section 230. And so for legal purposes, as the law stands today, there’s been a lot of I think confusion about this. What section 230 simply says is that Facebook doesn’t have legal liability for content created by other people. It also says that I don’t have legal liability. So if I am in a Facebook thread and somebody says something that might be legally actionable, I am not legally responsible. Or if I started the thread for example. And so the platform publisher distinction is one that I think you can talk about in the sense that Facebook is not primarily in the business of hiring editors and writers in the way that the New York Times is, but it’s not a legal distinction and doesn’t in any way change their legal status or what they’re responsible for under section 230.
Matt Crawford: Yeah. While Peter’s emphasized the continuity between an ad based media company and something like Facebook, one thing that I think is novel is that it’s a two way street. So in the New York Times… I guess you can write in a letter to the editor. And now of course they have comments sections. The engagement algorithms of social media are based on the realization that when people are contributing their own content they become more invested in it. And one effect of this has been these kind of creation of these sort of communities, right, that have gotten it kind of polarized into these different tribes. Because one of the powerful things Facebook discovered is that you can induce a kind of emotional contagion is the phrase. So if you curate their news feed in such a way as to show them things likely to make them angry, they’re going to spend more time on the platform, they’re going to share more of links with others, and contribute more content of their own.
So there’s a lot of people have been complaining about this obviously that our political divisions are in some significant measure seem to be artifacts of these engagement algorithms. So the fact that it has this sort of fast-pace networked collective and omnipresent character does seem to have opened up new avenues for having really an outsized influence on our public life. And again, it’s also these algorithms that achieve a kind of operant conditioning, to use the behaviorist term, because it’s those little likes and retweets and such that give you a little bit of social affirmation, a little shot of dopamine. And that is the mechanism by which these things become compulsive. And people of course do report. They self-report that they spend too much time on these things, which seems to indicate there’s some kind of gap between their sort of revealed preferences in staying on these things and their more considered preferences of how they’d rather be spending their time.
Peter Suderman: Yeah. So I think again that’s interesting to think about, but I would still say there’s continuity, right? The emotional contagion you described as if people are posting things that are likely to make other people angry. Well one of the most sort of well worn terms of phrase in the newspaper world is that if it bleeds, it leads, right? If you can put something that is going to outrage people on your cover, you will get attention for that and you will get reader engagement. And newspapers have known that and been doing it for decades if not centuries at this point.
I also think it’s interesting, you started by saying that they create communities and it seemed like you were going someplace, it was clear you were going someplace negative with this. You were like, you’re concerned about the creation of communities. And to me, I think that community creation is actually one of the features and one of the things that social media and internet connection has given us that we didn’t have. And that is, if not completely novel, at least a little bit different. And I have affinity communities over just sort of shared interests, but also some very, very close friends who I have been able to keep up with online that I just don’t think I would have been able to stay close with in an analog world, including a group of college friends who yes, we were physically close together in close proximity for a couple of years and many of those close-knit college groups grow apart. What we did was we started emailing every day. And this was back in the early days of Gmail. We had a listserve. We were all moderately tech-savvy and many of those guys went into IT careers. But we became really good friends, not just by living in dorms together, but by playing early online video games together. And that was what brought us together and that was our shared experience.
And this is something that I think critics of tech and of social media often overlook is that social media is social and can in fact create tight bonds and communities and especially in the last year or so when a lot of people have felt constrained and not able to have in-person, real world, real life relationships or contact with other people. I think we were much better off having social media and having digital connection which has allowed people to maintain perhaps imperfect, but still much better and much more robust and full communities and connections with other people than they would have been able to otherwise.
Matt Crawford: Well I’ll second that wholeheartedly. Yeah. Since the conversation sort of revolved around the concept of this, is it something to worry about here. That’s what I’ve been focusing on. But let me affirm my own version of what you just said. The technical forums that have sprouted up… So I’m a mechanic and engineer and I build stuff. The progress in knowledge that… The rate at the state of the art and various engineering fields has progressed in the last 20 years is mind-boggling because of these communities, of these technical forums. So what I was referring to with the sort of hate tribes of politics, that certainly does not characterize internet culture altogether, but it has been a highly consequential one. I don’t know what to do about it. I mean I don’t have a prescription for that. But clearly I think social media has played some significant role in pushing our politics into a kind of existential crisis, kind of French Revolutiony kind of vibe that has something to do with this sort of emotional contagion and kind of a scapegoating mechanism. So it’s a mixed bag.
Peter Suderman: So that seems to me more like a problem with politics than it does with social media. And I think people should probably spend less time engaging with politics, whether it’s online or whether it’s offline. And I say this as somebody who writes about politics for a living. And I think people have invested too much of their personal identity in political identities. But I don’t think that’s necessarily specifically a social media problem. I think that’s a problem to some extent with politics itself and with politics as a substitution for other forms of identity.
Matt Crawford: Yeah. And again, I took this, the point of having this conversation, not to be a sort of wholesale condemnation or defense of the internet, but rather more specifically focusing on whether the platform firms ought to be in some way kind of, I don’t know, made to put their effects on our common life somehow more accommodated within their business model, because if it turns out that stoking sort of politicized hatred is good for engagement, good for time on device, and therefore the profit motive lines up with that, I mean that’s consequential. So we might want to in a principled way say we shouldn’t interfere with that, but I think we need to sort of look concretely at the effects of that.
Peter Suderman: No, I’ll just say you you’ve agreed with me on a couple of things here so far. So I will agree with you a little bit here, which is just to say, I don’t love Facebook in particular of all the social media companies. I don’t remember exactly what it was I did to my account, but I certainly haven’t accessed it in several years at this point and have felt like I’m better off for it. At the same time, that suggests to me that we have the capacity to choose to log off and that the compulsive effects that I think that critics are ascribing to Facebook are just not all that strong. You even see that in the daily user numbers where Facebook’s daily users, even in the midst of a pandemic where people just wholesale moved their social lives online, they actually lost daily users in the last two quarters of 2020 and that’s because people do have the choice.
And this to me seems like a really sort of critical thing to just focus on is that individuals have the choice to engage with these companies and their products or not. And they’re not, Facebook in particular, Twitter in particular, they’re not commons, they’re private spaces, and you can either be on Twitter or not be on Twitter. You can be on Facebook or not be on Facebook. And I think we need to recognize in the political discourse here is the individuals have these choices and can make them. And in fact, if you look at Facebook’s numbers, in North America at least, people are choosing to log on.
Matt Crawford: Yeah, I think it’s this idea of that it’s simply a individual choice in an open market isn’t quite right. There is a quasi-compulsive character to these things. If your job’s, hunting, you’re not going to not be on Facebook, not be on LinkedIn. If you’re a journalist, you’re not going to not be on Twitter. We call them platforms, because there’s a network effect such that everyone has to be in the same place. So to say that you can simply opt-out of these things. If you’re lucky, you can. But there are a lot of circumstances in which people simply feel they have to be here, which does make them, in a sense, public utilities, or at least places of public accommodation. And there’s a whole obviously tradition of law on public accommodation that it needs to sort of comport with and that we affirm as a society as being important and worthwhile.
Wells King: Can we trust users to self-regulate? And should policymakers see the choice that users may have as adequate to opt-out of the attention economy or at least to manage the degree to which their attention is commodified?
Matt Crawford: So, yes, I guess the point to notice is that the burden of self-regulation has increased in the absence of state regulations. So when porn is available 24 hours a day in a device that I carry around in my pocket, this has to be consequential, maybe only in the sense of crowding out other things that I might be doing and yes, that it’s still a matter of individual choice. But on the societal scale, you have to wonder about a kind of… The costs of self-regulation are not zero. They don’t show up anywhere in economistic thinking, but they do show up in the kind of anxiety and failure perhaps to cultivate other types of pleasures that demand application and unpleasant practice, like learning the guitar or something like that. So this is consequential.
Wells King: Peter, Matt earlier noted that users themselves say that they use these products compulsively and even more than they would care to use them. To what extent should we depend on their stated preference or their revealed preference?
Peter Suderman: Yeah. I don’t think it’s unimportant that people self-report feeling like they spend too much time on social media. At the same time, people also say they don’t work out enough, people say that they eat too much ice cream, people say they don’t spend enough time with their kids. And people also report feelings about their time use in ways that comport with kind of what media and political elites are telling them is good time use and bad time use. And so this is why I actually think it’s important for people who have any sort of position in the discourse, or whatever we’re going to call it, to emphasize the role of individual choice. In the same way that we emphasize the role of individual choice with exercise, with diets, and with building relationships with your family members, is that you do have choices. You can opt-out of Facebook if you think that Facebook is bad for you.
You can choose to spend less time on Twitter. You can choose go running every day. Sometimes it’s hard. And for some people it’s harder than others. And we should recognize that, it’s not going to be the same experience for everyone. But if people don’t like what they’re doing, they can choose to do differently. And I think that it is incumbent upon those of us who write and speak about these things to not try and craft one-size-fits-all top-down solutions that are going to have at minimum a bunch of unintended consequences and other effects that I think people aren’t prepared for and I think would be quite negative and may not work at all. Instead, we can say, “Look, you can take control of your own life in your own time.” Many people have and many people do. And that is the struggle for all of us every day is to manage our own lives in our own time. And we have to take that struggle seriously.
Wells King: Well, Matt, Peter, thank you so much. This has been a fantastic discussion, with great defined areas of agreement and disagreement. And again, you can check out the essays on the attention economy as a part of our Lost in the Supermarket: Navigating the Digital Age collection. Thanks to you both.
Matt Crawford: Thanks to you both.
Peter Suderman: Thank you.
