Is Social Media a Trap?
Discussion of "When Product Markets Become Collective Traps: The Case of Social Media"
The source recording for this episode has no speaker labels, so this transcript runs as continuous dialogue without attributing individual lines.
I went to Duke and everyone at Duke cares about basketball. So imagine you're the one person that doesn't really enjoy sports. The fact that the basketball team is there and is very good is a huge negative for you and you play for everyone to no longer go to the basketball game. So I think the problem is that there any social activity is going to draw in people who'd rather be doing some other social activity or sometimes a non-social activity at all. Yeah, so think of a kind of the battle of the sex is framing where there's a population that likes social media and a population that doesn't like social media, but we all have to coordinate.
Right. And, sure, you want to fight on this about the social equilibrium, but don't frame this as an economic story. Don't frame this as a we're losing efficiency. Frame it as what it is, which is I prefer a culture that looks like this. And that applies, of course, to so many parts of culture, right? And this is what the culture wars are about. This is Justified Posteriors brought to you by the Digital Business Institute at Boston University's Questrum School of Business. I'll connect everyone to another episode of the Justified Pastereors podcast where we read economics papers and get persuaded by them. So you don't have to.
I'm Seth Benzell coming to you from the Hotel California like a collective trap of Southern California. And I'm Andrey Fradkin representing the no FOMO movement that I'd like to propagate throughout society. That's a great place to start, Andrey. Do you come and experience FOMO? Is that an is that an emotion that you have? So FOMO or the fear of missing out for the neophytes listening to this podcast is a common feeling. You'll fight means it doesn't know it's low means. You make too many assumptions about our audience, Seth. But anyway, do I feel FOMO? I do feel FOMO. I feel like not as much as maybe I used to, but I definitely see people doing stuff.
And I'm like, I'd rather be doing that than whatever it is I'm currently doing. I definitely experienced that emotion. How about you? Great question. I would say FOMO, Qua FOMO, in the sense of I don't necessarily want to go to this party, but I am jealous, anxious about not, but experiencing some sort of anxiety over the fact that I am not at the party, but the fact that I might not even want to be at the party. So that's kind of common to me. And I would say I'm kind of sympathetic to the view that if you show a photo of a bunch of people in my high school class, a bunch of people in my freshman dorm, smiling and bonding with each other while, I'm in my solo dorm by myself, I think FOMO is a real emotion that I've experienced.
And I would say that there's even a possibility that social media accelerates that and makes it more vivid. Now, this is a bit off topic on the college, FOMO. But when you say party, do you mean conference? When I say party, I don't mean conference. What I have in mind is your typical stank drinking party. And you can see someone preferring to live in a world where there's less social pressure to deal with that. They're the cultural critique of social media, which is that it pushes us to do showy superficial things, because that's where the pressure is. Maybe we're trapped performing for each other in a glass menagerie.
Well, as John Paul Sartre once said, hell is other people. So, this paper definitely takes that idea seriously. What's the title of the paper? The title of the paper is when product markets become collective traps, the case of social media. This is a paper by Leonardo Bertson, Benjamin Handel, Rafael Jimenez, and Christopher Ross for the non-economists listening to this. This is quite an accomplished author team. They're definitely known for their careful experimental work and innovative work in general. And this is a paper that has had a lot of, let's say, media attention and just overall and so it's got a newsy hypothesis. Yes.
I had the opportunity to meet with Rafael Jimenez and Biconi just last week. Could not meet a more gracious open to criticism of his work guy. So, Rafael, if you're listening to this, any criticisms of this paper come from a place of critiquing only your audacity and boldness and actually putting these ideas out there. Yes. So that's a great intro. So what are the main claims of the paper? Maybe we should state those before we get to the evidence and presumably we'll state our priors after we state the podcast, the sort of the narrow claim of the paper and then maybe we'll expand out from there.
The narrow claim of the paper is that users of TikTok and Instagram, US college campuses are in a collective trap in the sense that they all prefer to use the platform given that everyone in their community is using the platform, but they would prefer to live in a world where no one used the platform at all. And I would say the paper both talks about this in a sort of average person sense most of the time, but you could also see a version of this argument about the median person instead of the average person. What's your hypothesis? Do you do you buy that at all?
How do you evaluate that going in? I think it's a pretty reasonable hypothesis. Lots of aspects of society seem to suffer from this. The rat race, the red queen effect, and so on, where people feel compelled to invest in something long past its actual utility to them. It doesn't seem crazy to me that this could be an example of that. You can think of other examples such as trying to be fashionable or even trying to be well read. Oh, yeah, but that's a rat race. OK, well, maybe we have to start and take a step back and there are positive sub rat races and then there are negative sub rat races, right?
So good. So, capitalism itself, right, is a rat race. So companies are very actively competing with each other. And maybe if they all sat together in a room, they might decide that competing so intensely is not a great idea. What do you want to do? That's illegal for perhaps the force of competition or harness for a greater purpose, which is promoting the productivity of the economy and increasing innovation, hopefully. So we have a theory of why competition might be very intense in that setting and social media. Maybe it's we're all competing to take the coolest photo to Instagram or to pretend we look as hot as we possibly could, even though in reality, we don't look as hot, whatever it is.
And you never do that. We were never lie to you listeners. The only thing that we'll lie to you about is how wonderful our voice naturally sounds, because we're going to post edit this to make it sound better. Yeah, a lot more in the raw version. OK, so what do you think, Seth? I think that even though the paper does not focus on this distinction, I think there's a massive difference between the hypothesis that the median or a large percent group of people are made worse off and the hypothesis that the average person is made worse off. If we're really trying to take a step back and think about all of the positives and all of the benefits from social media, it's clear that there are some people who have extremely compelling, powerful social media profiles that teach people that distribute information, that entertain, that educate and can make giant profits for themselves.
A full analysis of these platforms would also take into account the profits of platforms themselves, the benefits of people getting matched to products with advertisements or any of those individuals, you might question every element there, to what extent is that a huge social positive? I think you add them all together, taking into account this long tail of people who really benefit for the platform. It's very implausible to me that the average person is made worse off by these platforms now. And that's sort of like some sort of caldorhic sufficiency sense. I think that platform or caldorhic sufficient. If you want to tell me that 20% of people are sadder that they're 10, 20% of people are made 5% more miserable because of Facebook or maybe even 40% of people are made 10% more miserable because of Facebook.
That's a hypothesis that I'm open to. It's not the kind of argument that makes me want to slow down modernity and, return capital V. But I do think there are certainly losers in the new equilibrium. There are people who would have been better off without the platform. Yeah, and I guess, we'll get to the methodology, but it's also important to think about not just at the binary of turning on or off the platform, but changing the rules of the platform, right? So oftentimes what's bundled in this evaluation of these platforms is maybe they're addicting if people are able to have caps or caps imposed on the amount of usage that could kind of preserve, it could have kind of yield a much more unambiguous benefit of a platform versus one in which a non-negligible share of people are completely addicted to it.
Right. So this is a little bit of a false dilemma. So you can imagine them asking the follow-up question, how much would you be willing to pay Mark Zuckerberg to make the beautiful utopian version of Facebook that isn't addictive, makes everyone love each other. And maybe you could solicit a big value from that, but we're getting a little bit ahead of ourselves because we haven't talked about that paper's methodology yet. Talk about the methodology. Just get started. I think I think the thought experiment that the authors are posing is one where you're kind of stuck asking people what they think, right? You can't wave your magic wand and actually make it so that everyone stops using TikTok, right?
So we're asking people hypothetical questions. Kind of is maybe the right word here as credible as possible, credible as possible. So this is kind of this finesse that the authors try to do. Let's start with the simplest case because this is a case that's been done previously in literature as well as, we want to know for a person, how much do they value a particular online service? So the simplest thing you could do is you can ask them, well, how much would we need to pay you so that you don't use the service for some amount of time? That's a simple thing to ask.
And oftentimes you get very reasonable answers there, but a worry with that is, of course, that you might not, believe that this is a real choice you're making, who's going to enforce the fact that you're not going to use this particular platform. So a subset of experiments in the literature, what they've done is they've actually enforced this. So if you say you're going to accept, 20 bucks not to use Facebook for a month, the authors like literally shut you out of your own Facebook account. So you can't do that. That's kind of what are the main approaches in the literature about this individual decision to use a particular product or service or platform.
That is, that's kind of, I'd say, not maybe not a solved problem, but there's definitely a way that people do it that is more or less accepted. What the authors of this paper point out very, very correctly, and also Seth has a paper on this topic. So I can't disregard his contributions, but that, of course, there isn't just one value from how much Facebook is valuable to you. Depends on how many other people you interact with also use Facebook. And so what the authors do, kind of the second step of what the authors do here, in addition to what I just mentioned, where they ask people how much they'd be willing to accept to not use Tiktok and Instagrams.
They ask, they say, well, we're these really wealthy researchers. And there's a chance that we could recruit everyone on your campus to turn off this platform. And suppose we did this. How much would we have to pay you to not use the platform? OK, so that's what the authors do. Now, this is essentially getting at the difference in valuation for this person from the platform with and without their classmates on a college campus. And that difference, presumably, is capturing the network effect. This is the famous network effect that's so important in social media platforms. Get that from the other students on the college campus.
Yes. And that is very important. Of course, the social network is not just about the college campus. It's a global social network, right? There are people like your family and friends who don't go to the same college as you who you might be communicating with on set platform as well. Or entertaining third parties. Yes, like ourselves. We love it when you guys connect to us. Do second parties at best? Yes. And if you do really want us to make a Tiktok account, please submit. Please tweet hashtag it. Justify past your year's Tiktok when 2025. Yes, OK, clearly we are trapped in this collective social media trap.
I want to add a couple of details to what you just said. The first is the sample size we're talking about isn't enormous here. It's about a thousand people and a little over a thousand people total. And we're going to split into three samples because we're asked some about Instagram. We're asked some about Tiktok and there's a subset of the Instagram. Those interviewed about Instagram who do not even use Instagram. So it's not a giant population. And so we're getting split up. If they are all college students, the survey population is overwhelmingly female, 68% female. And so take that into account a little bit. We do think there is some evidence that women are particularly susceptible to the native social effects of social media.
We saw that from the famous Instagram leaks, which vaguely hinting at a connection between teen girls and suicidality. So yeah, so maybe you think that this is a population that's really proud. This is they did this survey right as negatives of social media are in the news. You might imagine that's in the air as they're taking this survey. Are there any other considerations you want to bring in Andrey before we evaluate the numbers again? Well, I haven't mentioned, of course, that there's a kind of a third innovation of course, addition to just asking about the network effects, which is they give a person the following hypothetical.
They want to kind of put them in a position of making a decision, whether having Instagram or TikTok on a campus is overall good or bad. So what they do is they say that with some probability, we're going to pick a person and that person could be you and we are going to ask them, should we, should we turn off TikTok for the entire campus or not? And why this is important is it allows the researchers to then measure kind of how much would a person that not having anyone on campus in addition to themselves, have access to this platform. Right. So that's getting, that's the key at getting at the trap.
Yeah, it's framing it as kind of like the Kickstarter problem is how I would think about it, right? It says, would you be willing to pay $50 if enough of your friends signed up for this $50 plan, right? And even though ingenious, if someone believed it at soliciting the correct information, what it has to imagine that this is not a very few people taking the survey or taking that question literally. Yes, although the authors say that, when they asked whether the participants believe what they just told them, 40% of them said, yes, well, I believe the claim, the thing is the claim is for a case that's impossible.
No, no, they said that, 40% of the people believe that this would actually have. Oh, wow. Incredible. That's some flaws. There makes me trust this nevertheless. Yeah. So I think we're going to shut down. I, let me, let me do a quick look up what specifically that 40% is. While you look at it, I'll also add that they also ask questions just of the form. Do you see social media as bad for your campus? And people say yes to that, right? So there's even if you don't believe the numbers, there are large groups of people reporting that they do not like social media.
Yes. So before we discuss the methodology a bit more, I mean, maybe now is a good time to introduce the actual numbers. Okay, cool. So what do they find looking at TikTok first? They find that the average person in their sample would be willing to pay $55 to keep using TikTok, then they say that if they were the only people on campus using TikTok, they would value it at $39. And then finally, they believe that they would be willing to pay $24 on average for no one on campus to use TikTok, including themselves. So dramatic numbers, people in a universe where everyone else uses TikTok, they value it at $55 in a universe where nobody else on if everyone on campus were to stop using TikTok, they would be happy $24.
So indicative of perhaps a $70 negative social externality of TikTok use on TikTok, they would give social externality of TikTok use on other people on Instagram. Those numbers are a little bit smaller, but in the same direction, people valued Instagram at $47 a month on average in a case where only they on campus were able to use Instagram, they would value it at $37 a month. And finally, the amount that they would be willing to pay on average for no one on campus to use Instagram is $6 a month. So directionally consistent with these theories of a trap, really, it's just the last number that tells you what's a trap, the other numbers don't directly speak to that question.
Yes. And it's important to note that they also ask people who are not current users of the platform. OK, so maybe let's talk about these numbers. I think one important detail I'd like to add before we get too deep into it. I think I might have admitted this from the previous discussion, which is when you solicit these valuations, there's kind of two ways you can do it. One way you can do it is you can say, give me a number, any number. And another, so like basically free response, how much do you value Facebook go? I don't know, $55. Another way you can do it is as a series of yes, no questions.
Would you give up TikTok for $30? Would you give up TikTok for $45? They take the second approach. So they randomly select some of these yes, no valuations to ask about. But importantly, that means that there is a maximum valuation sort of built into this. TikTok can only be so good because they're not asking, do you value TikTok at $1,000, $10,000? Similarly, in the negative direction, there is a negative cap for the final question. So as you think about these average numbers, I would encourage the readers to take into account that if you think there is a long tail of people who benefit a lot, we value the platform a greater than $200 that is not being picked up in this.
Yeah. And certainly the authors make some alternative assumptions about those gap numbers. And right. And I do think an interesting aspect of it is for those who value it at $200, maybe they value it so much because of a network effect that's very different than the people who value it at $5. So like think about being, a power user of Twitter who follows, a selected, sample of some of the best researchers in the world versus someone who is just getting, the four you feed of a bunch of random famous people, a completely different story, right? So kind of the various aspects of this may be heterogeneous across users.
I understand what it is. Yes. I think if you were to ask me how much do I believe these numbers on their face, I think that positive valuations, I think those numbers are. If we're thinking about those as averages, I would think those are underestimates. I think if you look at conditional, I'm using TikTok, how many hours a day people use TikTok. Over $55 is very plausible for those power users. The intermediate case, how much do you value the platform when no one on your campus is using the platform? One thing I find kind of pretty implausible here is that there is a bigger drop off from losing access to your on-campus friends in TikTok than on Instagram.
The reason I think that is because when I think about the kinds of connections that people value on Instagram, sure, it's international celebrities. But it's also people in my family who I want to keep up with, pictures of their don. Whereas for TikTok, you have to imagine that most of the content being consumed is from professional creators. But I think you're going to disagree with me on this, Andrey. Am I completely misunderstanding these platforms? I think the network effects on two platforms are different. On TikTok, it's not really about messaging between your friends on the same campus. It's more like, here's a really funny video that everyone is talking about, and I want to see it too.
And on Instagram, it is really about the sharing aspect of it. You really want that like from your friends, this photo you posted, or you want your friends to see the story that you posted, that you went to some cool place and vice versa. So there's just, and that kind of leads us to think about what the collective traps might be, right? So the collective trap on TikTok is like, oh God, everyone's watching these dumb videos, but I have to be in the loop on them because I want to, appear like a normal person. And on Instagram, it's like, you're kind of keeping track of your, of your buddy's idealized lives, you know?
So yeah, I mean, it's interesting. I wouldn't say the strongest prior about these valuations. I think both platforms are certainly very useful even without classmates using it. Okay, cool. Now, how about this, the negative direction? So the social cost of using that platform? If we think that what's driving this is kind of the FOMO, I'm missing the party. That's the kind of FOMO I empathize with. That would make me think that Instagram would be the bigger trap and TikTok would be the smaller trap. And yet we have a much bigger trap estimate for TikTok at $24 versus $6 of loss. How do you think that?
I think a lot fewer people use Instagram. This is constantly evolving, of course, and is winning. But I think at the time the study, I'd say TikTok was way more important. TikTok was in the ascendant. Yes. Yeah. So just do naive measures of, our spend and stuff. It's just a lot more. So if you think there's a product market trap, then it will be bigger for that. Okay. So that you just, the way you interpret these numbers is that just the magnitudes for TikTok are larger because people are using it more. And the network effects are kind of heterogeneous. We can't exaggerate, speculate about what that should be worse.
Yeah. But I guess I want to kind of shift the discussion here. A little bit about like, how will the people understand these decisions to begin with? Right. This is a very hypothetical question. Right. Hard, hypothetical. Let's say all your friends stopped using TikTok. What would they be doing instead? Okay. So one thing they could be doing instead is watching YouTube. Now the authors say, they point to a previous study that says, actually, when you shut off, access to a platform, most people, or most of that time allocated is shifted to non-digital activity. Sure. Short term, you pull a real live term equal.
Yeah. I don't know. That's a fresh last destiny. So I guess my sense is that you really have to have in mind, what are people going to do instead? Now, then there's a version where what people are doing instead is they're going to, go hang out with each other in person a lot more often or they can play intramural sports or hanging out at the bar or whatever, or they could be using another digital platform that may or may not be, close compliment to TikTok and Instagram. Depending on what you think will happen, that should really affect your valuation. Right. Right.
Inside this $24 number is a lot and especially the 39th. So I think it's easier for people to think about everybody stopping than everybody else stops. I use it, which is such a bizarre case, like literally you're imagining becoming a social exile. Well, I think, I think it's kind of the vice versa. Just, the authors of this paper, they say, maybe you're going to be like an information gatekeeper. You're going to be so cool because you're the only one who gets to view the TikTok videos first and share them with your friends, you know. And then I guess it's a cool result that is a smaller number, right?
Yeah. But you see how like that origin, the idea of there being sort of like decreasing returns to having the platform as more people have the platform, right? That's something that should kind of push us, I guess, is that consistent with the trap? I think it would push us towards that intermediate. Yeah. The hypothesis would be that this intermediate case would be larger. It's a bizarre, it's a network effect and it would be on people not using it. I guess that's very consistent with a collective trap. Yeah, it is. Yeah. I think I think the story in the paper is very coherent, right?
That what the story that they're trying to tell. We don't believe this necessarily. We have to attack the methodology, the hypothetical questions or the sample or things like that. Right. And so maybe let's talk a little bit more about the sample. I already mentioned it was female dominant. We think it's college student. Do you think maybe college students would be particularly susceptible to these negative vibes, perhaps? Yeah. Yeah. I think there would be. But at the same time, I think everyone, we're social creatures. So I think different versions of this apply to different people. I always think about, does the outside option have its own collective trap, right?
So, let's say we'd all be doing methods. So imagine you're the one person that doesn't really enjoy sports. The fact that the basketball team is there and is very good is a huge negative for you. Right. Exactly. For everyone to no longer go to. So I think the problem is that there any social activity is going to draw in people who'd rather be doing some other social activity or sometimes on social activity at all. So are we just learning something about social behavior more broadly, just that there's a lot of people that if there was no party, they'd be happy, but they have to go to the party, right?
Yeah, I like this. Right. So think of a kind of the battle of the sex is framing where there's a population that likes social media and a population that doesn't like social media, but we all have to coordinate, right? And it's not a collective trap. It's just that, well, you're the group that had to go to the musical instead of going to the boxing batch. And you know, sure, you want to fight on this about the social equilibrium, but don't frame this as an economic story. Don't frame this as a we're losing efficiency, frame it as what it is, which is I prefer a culture that looks like this.
And this is what the culture wars are about. Should people get married early or not? Should we make one type of career prestigious or another? All these things are similar types of debates. And some of this sometimes the answer may just be that they're incommensurable, right? Imagine if you actually thought that life with TikTok was incommensurable with life without TikTok, which I mean, maybe, right? Then you're just kind of wildly flailing with these numbers. Yeah. And I guess the other thing I should really, like I come back to is the issue of addiction, and how well people understand their own addictions, two things, this might point to a fact that like, if you ask, people what they want to do, they want to exercise every day, they want to, they want to meditate and sleep well, and so on.
But then when they make their choices every day, they don't always, make good choices there that might be addicted to social media or video games or browsing the internet or whatever. If you ask them like, oh, if you just got rid of this thing that you're addicted to, would you want that? And you say, yeah, I want that. But actually, you just get addicted to something else. Yeah. When we evaluate this, should we take the person who just wants one more TikTok video and their amazing, intense desire for that? Is that the true self that we should be taking into account in our economic analysis, or rather the cold headed, sitting in a lab trying to think like a professional survey taker, trying to have a theory of mind of this survey taker as I take a survey version of myself.
That said, I think that it's important to put the study into context, right? So we have this broader discussion about laws, for example, laws controlling how much teenagers are allowed to use social media. Yeah, in Australia, we just had. Yeah, go ahead. Yeah. I was just going to say Australia has passed the law. Apparently, within the next year, we're going to have a complete ban on under 16 use of social media. Yeah, and individual schools are certainly banning phones or banning social media usage. And there's this movement led by Jonathan Haidt that kind of blames a lot of society's ills on social media.
So it's not just this paper that is kind of pointing to something negative about social media. But I let's talk about the political economy here for a second. Sure. I think the groups like, what are the groups that are ending up being restricted? It's the groups that don't have vote, right? If you really thought that this was a trap, and the only thing we need to fix the trap is get a majority of voters to ban it. Why can't we get a majority of college? Why can't we get a majority of the population to the man, right? And so I guess that's my question. Why don't we keep on if this is popular, why don't we ban it than the people who felt instead of people who felt that?
Well, I think the answer might be your standard answer is that, yeah, a bunch of people might weakly prefer to not have social media around. But the people who benefit from social media are really, really benefiting from it. And you know, that includes a lot of constituents, of course, right? That includes people who we think might be benefiting it for the wrong reasons. People who are just spreading bad ideas or harmful content, it might include people who are spreading good ideas that they think are valuable for the world. It might include also various businesses, right? So advertising, I guess this is a question about how we feel about advertising.
But if you think that, giving people information about products is something society should be doing, then maybe that's a good thing. Right. Yeah. So we return to this idea that there are definitely benefits, but we're definitely think that it's plausible that a big percentage of the population prefers a world without it. And I think that where that leaves me is just kind of anticipating this populist versus elite politics about it, right, is that there'll be the school arm keep this away from me and my family faction. And then there'll be the tech elites telling us, no, it's all good. Just keep by logging it.
I am maybe more sympathetic to the latter view than the ominous way that I just presented it. But I think that's going to be the politics for this for the foreseeable future. I think populist benefit extraordinary amounts from social media, right? Oh, that's a fair play. That they might talk a big game about how they don't like the specific people running specific social media companies, but they want those social media. Social media is great for populist. Right. But you can imagine the populist deciding on the one platform that's okay, right? Then all the other ones. Yes. TikTok is a different case. TikTok is about national security.
I really think it has very little to do with addiction. All right. So given that the leading proposal is to sell off TikTok to an American company, I just don't. I think we're a little mixing. Mixing metaphors there. Sell it to Americans if they'll make it more addictive. Yes. All right. So I think that brings us to our posterior Andrey. To what extent did this paper move your view about whether I guess narrowly TikTok and Instagram or traps on college campuses and we're broadly whether in general they're a net negative for society. I'm surprised by how many people said they would prefer a world without it.
I think that to me was a bigger number than I expected. And so that didn't move my beliefs. But at the same time, I don't love how hypothetical this choices. And I'm also worried just about the respondents kind of catering to what they believe the desires of the experiment. They're clearly trying to point in this direction. Right. Yeah. But I think it's a very neat study. And I think we shouldn't throw out the data just because there might be some issues with how to interpret it. Agree with you that the number and the value people are willing to put on these sorts of banning social media outcomes.
I did find surprising and does move my view about maybe how the for rearing people from benefits most from Instagram to is hard the most from Instagram. Like maybe in my brain traveling up the percentiles that are being made worse off. Maybe going into this, I thought, yeah, there's a percentage but it's five, 10 percentage of the society that is worse off from these. And maybe I'm going to start updating that to like the 20th percentile, 20th percentile. But at the end of the day, I think that this analysis is both underpowered and poorly designed when it comes to soliciting that long tail evaluations. And therefore I remain pretty much convinced that they're getting on average and inefficiency from social media usage in this like caldery existence.
All right. Well, thanks for listening to another episode of Justify Past serious. Let us know what you think and please review, like and comment as much as possible. Yeah, unless you if you don't, you're going to have fun though, from all the cool kids liking and subscribing. Thanks very enough.