This article is a review of the study
Bursztyn, Leonardo, Benjamin Handel, Rafael Jiménez-Durán, and Christopher Roth. 2025. "When Product Markets Become Collective Traps: The Case of Social Media." American Economic Review 115 (12): 4105–36.
Do you think TikTok/Instagram has more pros or cons? A recent study by L. Bursztyn, B. Handel, R. Jimenez-Duran, and C. Roth revealed that the majority of active users perceive these platforms as detrimental. Despite this, they continue to scroll through their feeds and return to the apps repeatedly. What are the reasons for this phenomenon?
The key to understanding lies in the negative externalities for those who refrain from using social media. But before we proceed, it is important to understand the general mechanism. The authors have demonstrated that social media consumption is associated with multiple reciprocal effects.
First of all, there is a positive network effect: for a given platform, a larger number of users may increase the benefits of joining, by expanding the network of individuals available for interaction. Similarly, going to a concert or having dinner with friends is often more enjoyable than doing the same things alone.
On the other hand, along with network effects, users face negative outcomes that reduce the overall utility of social media. This issue has been the subject of other research on this topic. For instance, studies by Allcott et al. (2020) and Braghieri, Levy, and Makarin (2022) demonstrate that the expansion of social media and their frequent use can adversely affect individual well-being and mental health of students, which, in turn, can have a detrimental effect on their academic performance. One potential explanation is the tendency of students to engage in unfavorable social comparisons. Another common adverse effect may be procrastination.
Finally, there are also negative externalities for nonusers of social media. If a platform has already become part of daily interaction among a person’s peers, abstaining from it may cause a substantial loss in individual utility. The authors indicate that, within the context of social media, externalities can be driven by mechanisms such as social exclusion or a fear of missing out (FOMO).
Thus, positive and negative effects act simultaneously and in opposite directions. Therefore, the overall utility of social media is subjective. A survey conducted by the authors indicates that a considerable proportion of respondents expressed a negative perception of these platforms, with 33% of TikTok users and 57% of Instagram users indicating a preference for a world without these social media platforms.
Naturally, this raises the question of why they continue to use them. Most pointed to the fear of missing out on something important, with 40% of TikTok users and 76% of Instagram users expressing this concern. This is the primary expression of the aforementioned negative externalities. Another significant reason cited by participants was the entertainment value (31% and 21%, respectively): “It’s a very good source of entertainment and it’s always something to do when bored”.
In previous studies on this topic, negative externalities were disregarded, and it was assumed that zero social media consumption automatically resulted in zero utility. The authors of the present article sought to test this assumption and thus established an experiment to provide a more precise evaluation of the utility of social media, incorporating external effects into the analysis.
These results could help reconcile the seemingly contradictory findings in the social media literature of a large consumer surplus coexisting with negative effects on well-being. The underlying reason for this phenomenon is that abstaining from social media results in a significantly greater degree of disutility in the form of externalities compared to active usage. In other words, people continue to scroll TikTok or Instagram not because they benefit from it, but rather out of a desire to remain connected to their networks and avoid feelings of exclusion.
The authors designate the resulting state as a "social media trap" which is similar to a Prisoner’s Dilemma. It would be preferable if no one used TikTok or Instagram. However, given the widespread active usage and content sharing on these platforms, joining them is less harmful than abstaining. It is important to note that such a trap can arise from social forces even with fully rational expectations and without behavioral frictions, such as lack of self-control and naivety.
The experiment was conducted among a sample of over a thousand students from various colleges in the United States. The authors' decision to focus on college students was motivated by several reasons. First, they are among the most active users of social media. Second, literature (Braghieri et al., 2022) points to a link between social media usage and the rising incidence of depression among college students. Third, while other college students may not represent the entire network of participants' friends, they constitute a significant subset of students' social networks. TikTok and Instagram were selected as the subjects of the study because of their potential impact on individual well-being, which is of primary concern.
The experiment sought to determine the extent to which students would be willing to deactivate their social media accounts for four weeks under different circumstances. To ensure that participants actually completed the task and deactivated their accounts, researchers reviewed profiles and requested participants to upload screenshots of their app usage.
The monetary reward was determined through a series of surveys divided into three steps.
Step 1 employed conventional tools (the Becker-DeGroot-Marschak mechanism) to evaluate the willingness to accept, that is, the amount for which a student would agree to give up social media, provided that others continue to use it as usual. This step corresponds to standard methods of assessing individual consumer surplus (i.e., utility), which fail to account for negative externalities. Results of the Step 1: on average, students agreed to give up TikTok for $55 and Instagram for $47.
In Step 2, an assessment was conducted to measure the compensation that the student was willing to accept if the other students at his college also deactivated their accounts. To enhance the credibility of deactivation study, participants were reminded that similar "deactivation" experiments had been conducted previously (Allcott et al., 2020 and Mosquera et al., 2020), and the anticipated level of agreement among students was 90%. Results of the Step 2: on average, students agreed to give up TikTok for $39 and Instagram for $37.
The difference between results of the first and second steps determines the size of the positive network effect. Students agreed to lower compensation if they knew that others would also give up social media. The average difference was 29% for TikTok and 21% for Instagram, in comparison to the Step 1 estimates.
In Step 3, students were provided with a scenario in which one randomly selected participant, whose anonymity was guaranteed, had to make a decisive choice between: (i) maintaining access to social media for everyone or (ii) deactivating the accounts of all participating students, including himself. The results showed that, on average, participants were willing to forgo rewards by choosing to deactivate everyone's accounts for free. On average, respondents were willing to forgo a payment worth $24 for TikTok and $6 for Instagram. These values represent estimates of negative consumer surplus (utility) that account for externalities.
The results of the Step 3 show that on average active users of social media platforms derive negative utility from the platform’s existence. Students were willing to pay $24 (TikTok) and $6 (Instagram) to disconnect not only themselves but also all students at their college from the selected platform. This can be interpreted as a scenario in which social media would not exist. In other words, a significant portion of users experience negative utility from the platform, and it would be more beneficial for them if it did not exist.
Based on these findings, it is possible to estimate the negative externalities to nonusers. This effect is equal to the difference between the results of the third and first steps of the survey, yielding an average value of (-)$79 for TikTok and (-)$53 for Instagram.
When these figures are compared with the negative utility values for social media users (-$24 and -$6, respectively), it becomes clear that negative externalities are greater in absolute terms. Therefore, the option with less adverse impact in this case is consuming TikTok and Instagram. This outcome is what constitutes the social media trap.
The reviewed article challenges the traditional argument that the mere existence of a product implies positive welfare for its consumers. In other words, consumption does not necessarily mean benefit – sometimes it is just a less detrimental option.
A similar trap is probably the throwing lavish weddings, which can impose a substantial financial strain. Due to prevailing social norms and social comparison, choosing not to organize such an event still appears to be a less attractive option.
In conclusion, we would like to know your opinion: are TikTok and Instagram more beneficial or detrimental to you? And how do you feel about weddings? Share your opinion by taking a short survey at the link below (30 sec.).
Bursztyn, Leonardo, Benjamin Handel, Rafael Jiménez-Durán, and Christopher Roth. 2025. "When Product Markets Become Collective Traps: The Case of Social Media." American Economic Review 115 (12): 4105–36.
Allcott, Hunt, Luca Braghieri, Sarah Eichmeyer, and Matthew Gentzkow. 2020. “The Welfare Effects of Social Media.” American Economic Review 110 (3): 629–76.
Braghieri, Luca, Ro’ee Levy, and Alexey Makarin. 2022. “Social Media and Mental Health.” American Economic Review 112 (11): 3660–93.
Mosquera, Roberto, Mofioluwasademi Odunowo, Trent McNamara, Xiongfei Guo, and Ragan Petrie. 2020. “The Economic Effects of Facebook.” Experimental Economics 23 (2): 575–602.
