Social media is an integral and pervasive aspect of modern living. There are 4.88 billion social media users worldwide, equating to 60.6 percent of the global population (Kemp, 2023). On average, people actively engage in 6.7 different social media platforms monthly for 2 h and 26 min daily (Kemp, 2023; Kepios, 2024). Facebook, YouTube, Instagram, and TikTok are the world’s most widely used social media platforms, each claiming at least a billion or more active monthly users (Kemp, 2023; Kepios, 2023a, b, c). There were only slightly more than 2.5 billion active social media users in 2015, just under 30% of the world’s population. With the adoption rate of social media doubling in just eight years—there were only 2.5 billion active social media users in 2015—the number of people using social media will likely see continued growth.

Computer-mediated virtual social interactions increasingly resemble in-person interactions (Bailenson, 2018; Nass & Reeves, 1997). While this technological transformation offers unprecedented opportunities for connectivity and communication, it has also raised critical public concerns and prompted substantial research inquiries (e.g., Kross et al., 2021; Piazza & Bering, 2009). The consequences of prolonged and intensive social media usage have become a focal point of attention, with growing awareness of potential detrimental implications for mental health, privacy, and societal dynamics (e.g., Howard & Parks, 2012; Kross et al., 2021; Smith et al., 2012). As we grapple with the multifaceted impact of social media usage, it is evident that an evaluation of how our psychology reacts to virtual social interactions is imperative.

This paper outlines a conceptual framework, proposing how social media appeals to our evolved social needs and how social media features may interact with our psychological mechanisms to prompt “social media ills.” These “social media ills” encompass the less-than-desirable outcomes commonly associated with social media usage, such as problematic use and relationship dissatisfaction (Bányai et al., 2017; Sbarra et al., 2019). We will first provide a brief introduction to social media by discussing the working definition of social media and some of its common features. Then, we will discuss some commonly observed consequences of social media usage. Subsequently, we will cover some psychological mechanisms humans have evolved to facilitate survival and reproductive fitness over our socially complex evolutionary history. Specifically, we will focus on how these evolved psychological mechanisms are sensitive to specific social information, which will set the stage for explaining our amenability to social media and its vast reach of social content. Applying the evolutionary mismatch framework (Li et al., 2018), we proposed that negative consequences of social media use are likely the result of the departure of modern functioning on social media from the normative operation of our evolved psychology over our evolutionary history. We address how evolutionary novel features of social media can disrupt our evolved psychological processes geared toward monitoring social information and produce negative outcomes such as lowered self-esteem and social media addiction. Finally, we will also propose suggestions on how policies can be crafted to minimize social media ills while maximizing the benefits of social media. Directions using the evolutionary mismatch framework for future investigations on social media are also discussed.

A Brief Introduction to Social Media

Definition and Features of Social Media

Given the rapidly changing usage and functionality of social media, defining what constitutes “Social Media” is a challenge (for a systematic review, see Aichner et al., 2021). For the purpose of this review, we base the definition of social media on the typical working definitions in the existing literature. That is, social media broadly refers to highly interactive internet-based computer applications that serve as platforms for individuals to create, discuss, and modify user-generated content and for communities to emerge through these activities.

Understanding social media requires understanding what makes these media “social.” Social media arose from the post-millennium paradigm shift in the usage of the World Wide Web (Kaplan & Haenlein, 2010). Contents are no longer created by specific individuals or organizations but generated and modified continuously via collaborations by participating sets of users (also known as Web 2.0). Social media platforms (a) allow users to reveal their identity to others online to varying extents, (b) signal to others their social availability, (c) act as the medium by which conversation between users takes place, and (d) facilitate the exchange, distribution, and the consumption of content. Consequently, social media also (e) fosters relationships between two or more people, which in turn (f) forms groups and communities. Finally, through interactive features such as “likes,” reviews, and comments, users can (g) infer the reputation of others, themselves, and the quality of the content distributed.

Social Media Features and Impact

Undoubtedly, social media has dramatically facilitated communication and social interaction. Social media platforms provide on-demand opportunities for connecting and engaging with others, irrespective of the hour or geographical distance. These readily available modes of communication are particularly crucial for fostering social interaction (see Kietzmann et al., 2011, for a review). Its social features, combined with the massive volume of users, also allow for more information creation and information to achieve greater reach (Lipsman et al., 2012). In this way, not only do individual users benefit from timely access to relevant information and opportunities (Utz, 2016), but businesses, institutions, and enterprises can also enjoy the wide reach of audience within social media, which is integral to communication arsenals in modern economies (Kim, 2018).

Social media is unique in its affordance of remote socializing and the ability for users to articulate and make their social networks visible (Boyd & Ellison, 2007). Through their social network, users can traverse their connections and those created by others in their network, allowing them to connect with like-minded individuals not found in their immediate network and form communities. Support networks can be built and maintained by sharing personal experiences and self-disclosure. Online communications provide support networks that allow for hope, camaraderie, and assistance to individuals facing challenges and enduring difficulties (i.e., mental illnesses; Bucci et al., 2019; Naslund et al., 2014, 2016). The option users have in controlling the extent to which they reveal their identity online also affords anonymity, which can be especially beneficial for individuals with stigmatizing physical and mental health conditions seeking to establish social connections (Berger et al., 2005; Highton-Williamson et al., 2015). Results from a meta-analysis across 38 empirical studies suggest that positive and honest self-disclosure is moderately and positively associated with psychological well-being (Chu et al., 2023). Collectively, engaging in social media can bring about improved psychological well-being and life satisfaction (Kim & Lee, 2011; Valenzuela et al., 2009) as well as reduced stress via perceived social support (Marciano et al., 2022; Nabi et al., 2013).

However, rapid technological advancement toward unparalleled connectivity without a thorough assessment of its foreseeable impact begets negative consequences (Haidt, 2022). Empirical research has also revealed robust, adverse effects of social media use. Being privy to other people’s seemingly exciting social activities can induce a sense of loneliness and the fear of missing out (FOMO)—the fear that everyone is having a great time without an individual (Valkenburg et al., 2022). Social media platforms are often flooded with content showcasing the highlight reels of people’s lives. Social media users meticulously choose and present the most positive and attractive aspects of their lives, such as flattering photographs, expensive goods, and personal successes (Blease, 2015; Gonzales & Hancock, 2011; Siibak, 2009). These curated contents induce body image dissatisfaction and low self-esteem (Blomfield Neira & Barber, 2014; Harriger et al., 2023). Over time, the regular use of social media can result in negative moods and a reduction in subjective well-being and life satisfaction (Sagioglou & Greitemeyer, 2014; Valkenburg et al., 2022).

In more severe cases, social media use increases the likelihood of social media addiction and depression (Bányai et al., 2017; Morrison & Gore, 2010). More frequent visits to social media platforms and more platforms accessed were associated with greater symptoms of depression and anxiety (Lin et al., 2016; Vannucci et al., 2017). Individuals who used between 7 and 11 different social media platforms were three times more likely to display more severe depressive and anxiety symptoms than those who used two or fewer social media platforms (Primack et al., 2017). Recent studies also indicate that heavy use of social media is consistently associated with adverse mental health outcomes at non-trivial levels, especially for girls (Twenge et al., 2022). The adverse impact of social media use on mental health among adolescent girls was greater than binge drinking, sexual assault, obesity, and drug use. Evidently, persistent engagement in social media use exacerbates these adverse effects (Scott et al., 2019). Yet, paradoxically, despite its negative influence on psychological health, people find it hard to disengage from social media.

Explaining Social Media Amenability and Its Consequences

Despite the evidence on how social media can be harmful, the understanding in this area is far from definitive, particularly for why people remain amenable to social media despite the negative impacts and emotions they bring. Some theoretical models have been proposed to explain the appeal of social media. According to the cognitive-behavioral model of pathological internet use, individuals may consume social media excessively because it serves as a coping strategy for real-world negative emotions and troubles (Davis, 2001). This is especially so for individuals with underlying psychopathology (e.g., social anxiety), who may possess maladaptive cognitions about themselves (e.g., negative self-appraisal) and the world (e.g., social media is their only friend). Within this framework, the excessive consumption of social media is reinforced by the positive experience they may have on social media. The socio-cognitive model of internet addiction is an alternative model that attempts to explain social media’s underlying appeal to people (LaRose et al., 2003). According to this model, when people rely on social media to relieve themselves of dysphoric moods, they develop a habitual reliance on using online media to counter their negative psychological state. Over time, people become less conscious of their consumption patterns, which diminishes their ability to regulate their usage. Consequently, this encourages even more consumption, creating a downward spiral and, eventually, problematic use.

While these theories elucidate how excessive consumption of social media may develop due to the opportunities afforded by social media features, the underlying assumptions behind these models limit the parsimony of their explanations. Specifically, these models assume that (a) only people with underlying psychopathology are amendable to social media and (b) being on social media provides a rewarding experience such that people are encouraged to continue its usage. Both may not be the case, as evidenced by past research and other anecdotal evidence. In other words, as valid as these theories are in the provision of critical insights into how social media features exploit vulnerable psychological processes, they do not sufficiently address why human beings—with and without underlying psychopathology—would be so amenable to social media even if it makes them feel bad.

Some theorists have also argued that it is not social media per se that is harmful, and hence, it should not be responsible for mental health problems and other undesirable outcomes resulting from social media use. Instead, these less-than-desirable outcomes stem from people’s psychological vulnerability (George & Odgers, 2015; Kreski et al., 2021; Odgers, 2018; Odgers & Jensen, 2020). These include those who experienced difficulties in their offline lives, such as coming from low-income families, having a history of victimization, and having behavioral problems. Evidence supporting such propositions often finds a relationship between offline social difficulties and online risk or that youth of specific demographics are more likely to experience harm. For instance, it was found that adolescents at risk of being bullied are also at risk of cyberbullying (Kowalski et al., 2014), and adolescents who have positive social relations offline are also more likely to have positive online social interactions (Valkenburg & Peter, 2007). It was also found that adolescents from lower-income families are more likely to experience spillovers of negative experiences on social media and offline conflicts (Odgers, 2018). While it is clear that some vulnerable youths are far more likely to experience harm than others, it does not sufficiently explain why people without psychological vulnerabilities are also susceptible to the negative effects brought about by social media. These lines of research also stop short of integrating insights from research examining the impact of social media features to explain why social media use entails different outcomes for different groups of people. Focusing on the characteristics of vulnerable groups overlooks how social media can generally be harmful.

Evolved Mechanisms for Social Living

The core aim of this paper is to advance an evolutionary mismatch framework for examining people’s amenability to social media and “social media ills,” which we define as the behavioral problems and psychological harms stemming from people’s amenability to social media and social media use. We propose that people’s amenability to social media and social media ills are manifestations of evolutionary mismatches between our evolved psychology and the evolutionarily novel features of social media. In short, we argue that the roots of social media amenability and social media ills lie in how evolution has shaped several psychological processes that generate adaptive behavior to aid social living in our ancestral past. As an evolutionarily novel environment, social media has introduced several input cues that our evolutionarily endowed mechanisms are not designed to process. As such, these inputs continuously draw people in to use social media and consequently elicit social media ills. Accordingly, central to our proposal is the argument that people, regardless of underlying psychopathologies or psychological vulnerabilities, are all attracted to social media and susceptible to social media ills, even when their experience on social media may not be positive. By employing an evolutionary-informed analysis, we hope to provide an integrative approach that complements existing theories addressing proximal causes of social media problems. By focusing on how evolution shapes psychological mechanisms that are ill-designed for interactions on social media, we seek to provide an ultimate explanation of people’s social media amenability and social media ills (see Tinbergen, 1963, for a distinction between ultimate and proximal explanations).

Before expanding on the series of arguments for how people’s behaviors concerning social media use are manifestations of evolutionary mismatches, we must first identify the evolved psychological mechanism that emerged through natural selection (Li et al., 2018). This necessitates a shift in attention to the environment of evolutionary adaptiveness (EEA) and the problems ancestral humans faced that impacted their survival and reproduction (Lewis et al., 2017). EEA refers to the set of recurrent selective pressures that shape a given adaptation (Lewis et al., 2017; Tooby & Cosmides, 1990). It is important to note that EEA does not refer to any specific environment across time per se (e.g., the African savanna during the Pleistocene). Instead, it describes recurring adaptive problems that persist over an organism’s evolution history, shaping a particular adaptation. As such, for every adaptation an organism possesses, a distinct EEA favors the adaptation. Suppose understanding social media requires an understanding of the sociality of social media; then, understanding how humans are susceptible to the influence of social media would require an examination of (a) which evolved psychological mechanism social media exploits and (b) how these evolved psychological mechanisms emerge to facilitate social living. Hence, in this section, we will discuss these mechanisms—which are now susceptible to the influence of social media—that humans have evolved to facilitate living in social groups in the first place. It is important to note that each adaptive problem favors a specific adaptation. Therefore, our rich evolutionary history of social living has likely led to the development of many psychological mechanisms. Our discussion here is necessarily non-exhaustive. Our current discussion focused on evolved psychological mechanisms in which operations on social media platforms and their interaction with social media features are mismatched to those of our ancestral environment.

The living conditions in ancestral environments were likely dangerous—people were frequently exposed to threats ranging from the weather, predatory animals, and other rival groups. Recurrent obstacles and impactful threats to survival and reproduction exerted strong selection pressures (Lewis et al., 2017). Consequently, important psychological adaptations evolved to address frequently encountered and highly impactful adaptive problems. Group living allowed humans to cope with an environment with predators and an inconsistent supply of shelter, water, and food (Boinski & Garber, 2000; Foley, 1998; Wilson, 1975). A greater number of organisms living in close proximity mitigate the risk of predation as collective vigilance and flight distance reduce the probability of any particular individual dying to a predator (Alexander, 1974; Hamilton, 1971; Hill & Lee, 1998; Olson et al., 2015; Powell, 1974; P. W. Sherman, 1977). Collective actions such as foraging and hunting are more efficient with group living, as foragers cooperate by covering different parts of the habitat and exchanging information about food availability (Kurland & Beckerman, 1985). Reciprocal food sharing is also possible by group living. Excess food availability has diminishing returns for sated individuals (Winterhalder, 1996a); distributing excess food resources not only facilitates cooperative defense of the common food source (Winterhalder, 1996b), but it also allows those with excess food resources to generate reciprocity debt (Kaplan et al., 1985). Division of labor, group defenses, and communal parenting—afforded by group living—provided a buffer against external threats (Kenrick et al., 2003). Evidently, collective effort and cohesion are imperative for human survival and reproductive fitness (Aronson, 1999; Baumeister & Leary, 1995; Brewer & Caporael, 2006; Darwin, 1871), and as such, humans evolved as group-living animals who are obligatorily interdependent on one another.

Paleoanthropological evidence further limned the sociality of the human evolutionary past. Primates, in particular, tend to have an unusually large brain size, which is primarily driven by the size of the neocortex (Jerison, 1973). According to the social brain hypothesis (Dunbar, 2003; Hill & Dunbar, 2003; see Dunbar, 2009 for a review), the difference in neocortex size can be explained by the differences in the complexity of social skills between primates and non-primates (Byrne & Whiten, 1988). The extent of human sociality was exemplified by the finding that the average network size (124.9 to 153.5 individuals) of a modern human closely resembles estimates (150) that the human neocortex allows (Hill & Dunbar, 2003). Social relationships are cognitively costly due to the many-fold demands of maintaining cognitive and behavioral coordination with many social others (Dunbar, 2009). The greater the group size, the greater the demand for neurological infrastructures to support complex social cognition (Dunbar, 1992). Hence, the correlation between group size and neocortex volume is an emergent property of an organism’s ability to juggle complex relationships (Dunbar, 2009).

Given the intense selective pressure that favored group living, humans’ need to belong and the desire for social acceptance evolved as mechanisms to encourage affiliative actions and boost the chances of being accepted into social groups (Buss, 1990; Leary, 1999). By this account, the human brain not only evolved in size to manage the complexities of living in social groups but also possesses several well-designed psychological mechanisms to promote behaviors that would enhance social inclusion and avoid exclusionary status.

Social Monitoring

Humans may have evolved a social monitoring system that regulates social belongingness to enact adaptive behaviors for maintaining inclusionary status within a group. This social monitoring system allows individuals to notice social cues, signal potential rejection, prompt behaviors, and engage in interpersonal techniques that lead to greater belonging and inclusion (Pickett & Gardner, 2005). Much like how biological mechanisms prompt hunger to signal potential caloric deficit to increase the sensitivity of information about food availability, the social monitoring system is activated by situational (e.g., immediate threat of social exclusion; Gardner et al., 2000) or chronic (e.g., individual differences, see Pickett et al., 2004) need for belonging, which increases sensitivity to social cues. These social cues may signal positive or negative inclusionary status and often involve the behaviors of others toward an individual (e.g., detecting a friend being eager to end a conversation) or how others treat someone else (e.g., seeing that someone else is invited to a party by others). Importantly, these social cues may include a range of subtle verbal and non-verbal social cues that individuals can perceive and process (see Fichten et al., 1992). To illustrate, determining whether a conversation partner is bored often entails considering cues beyond what the person says, such as whether they avert their gaze or exhibit signs like tapping their feet. When a perceiver is attentive and observant, they can interpret these signals and recognize the need for a change, such as altering the topic, to prevent potential rejection by the partner. When the social monitoring system picks up on potential rejection, it activates a heightened sensitivity to social cues and social information to attune individuals to opportunities for social inclusion. In an experimental study, participants experienced either acceptance or rejection during an internet conversation (Gardner et al., 2000). Participants who received either a dyadic or group rejection better recalled social (vs. non-social) activities from reading a diary than participants who received social acceptance. Likewise, lonely participants remember social events better and are more attentive toward social cues in the face and voice (Gardner et al., 2005).

Sociometer

From an evolutionary perspective, social relationships characterized by stability, mutual concern, and continuation into the future prompted better chances of survival and reproductive fitness than relationships that did not (Baumeister & Leary, 1995). The central tenet of the sociometer theory (Leary & Baumeister, 2000; Leary et al., 1998) is that self-esteem is an evaluative system that assesses the quality of our actual and potential relationships. This assessment is carried out in terms of how much other people value social interaction with us (Bale & Archer, 2013). Thus, on top of avoiding being socially excluded, people are also concerned with how other people regard their relationships with them to be valuable, meaningful, and close. According to the sociometer theory, self-esteem functions as a psychological gauge that monitors the quality of an individual’s interpersonal relationships (Leary & Baumeister, 2000). High self-esteem reflects the perception that one is a valued partner in social relationships, whereas low self-esteem reflects the perception that one’s social inclusionary status is low. When one experiences relational devaluation, their inclusionary status may dip below satisfactory levels (i.e., belongingness is threatened).

The sociometer activates emotional distress to signal and motivate the individual to engage in behaviors that can restore relationships (Leary & Baumeister, 2000). Meta-analysis has shown that social rejection tends to lower mood and self-esteem, motivating behaviors to regain control or a sense of belonging (Gerber & Wheeler, 2009). Across several studies, self-esteem is positively related to how people feel included or excluded by others (Leary et al., 1995). These behaviors may appear aimed at restoring self-esteem on the surface but are more fundamentally aimed at repairing their social acceptability. Leary and colleagues (1998) manipulated state self-esteem in four experiments by providing them with real and imagined feedback about themselves. While socially rejecting feedback consistently lowers state self-esteem, accepting feedback demonstrates diminishing returns in positive self-esteem. The ogive function characterizing the relationship between social information and self-esteem supports the sociometer theory. People are sensitive to social exclusion and loss of inclusionary status. Once people’s self-esteem becomes negative, they are spurred into restorative behaviors (Gerber & Wheeler, 2009; Leary, 2005; Leary et al., 1995, 1998). However, social acceptance is assumed when self-esteem is relatively high, and further positive feedback yields little improvement (Leary, 2005; Leary et al., 1998). Hence, the sociometer plays a critical role in detecting social information that conveys the quality of social relationships and acceptance in social groups.

From the view of sociometer theory, self-esteem is an output of a psychological mechanism that guides decision-making in social situations, such that appropriate responses implicating one’s group membership can be generated. Past studies have shown that people with higher self-esteem are more likely to draw other’s attention to their desirable qualities (Baumeister et al., 1989). In contrast, those with low self-esteem are more focused on avoiding the expression of bad qualities. As having an attractive romantic or sexual partner is highly desirable, people who perceive themselves as attractive are likelier to express higher self-esteem (Bale & Archer, 2013). Accordingly, attractive individuals are more confident in establishing and maintaining romantic relationships. Likewise, individuals with higher self-esteem are always eager to engage in social activities, regardless of whether social acceptance is guaranteed, as they anticipate positive social outcomes (Anthony et al., 2007). Conversely, those with low self-esteem are more selective in their willingness to join group activities due to their potential negative self-presentation (Anthony et al., 2007). They prefer to participate only when social acceptance is guaranteed, where the quality of group relations is likely high.

The sociometer and social monitoring system operate in tandem in that a feedback loop exists between the sociometer and the social monitoring system—the former is a motivational gauge that accesses the quality of inclusionary status. The latter is a perceptual system that monitors the environment for social cues and guides an individual toward opportunities for social connection (see Pickett & Gardner, 2005). When the sociometer detects a deficit in inclusionary status, the social monitoring system will be activated to monitor social cues, prompting behaviors, and social interactions to restore inclusionary status. Suppose subsequent social interactions successfully restored inclusionary status. In that case, a person’s self-esteem—a gauge of their relational value—returns to a state of equilibrium or increase. Otherwise, further attempts will be made to restore the inclusionary status. When expecting or experiencing exclusion, individuals tend to seek out opportunities for inclusion. Adaptive behaviors, such as increased effort into group tasks (Williams & Sommer, 1997), conforming to group norms (Williams et al., 2000), as well as compensatory cooperation (Ouwerkerk et al., 2005), are some of the well-documented responses to social exclusion. Likewise, behaviors promoting social inclusions, such as prosocial behaviors, reliably reduce feelings of loneliness (Lanser & Eisenberger, 2023), downregulating the social monitoring system.

Social Comparison

People have an innate tendency to compare themselves to others (Festinger, 1954)—a tendency so pervasive that some scholars regard it as an inevitable consequence of social interaction as long as relative information between self and others is available (Wood, 1989, 1996). Social comparisons fulfill an adaptive function of reducing uncertainty and facilitating adjustment to threats others pose. Adequately sizing up one’s potential or actual competitor is crucial to survival and group functioning (Gilbert et al., 1995a, b). Animals, including humans, assess comparative differences between themselves and others to avoid losing situations with a gross disparity in strength or ability (Buunk & Brenninkmeyer, 2000). In such encounters, judgments of relative probabilities of making a successful challenge occur. Individuals with superior strength and abilities are in a favorable position to challenge. In contrast, individuals with inferior strength and abilities are placed in an unfavorable position where displaying submissive behaviors and avoiding attacks or challenges is more beneficial (Gilbert et al., 1995a). In other words, comparing oneself to others allows organisms to compute their resource-holding potential—the potential for individuals to win resource contests or challenges given their ability and motivation relative to competitors (Parker, 1974; Rudin & Briffa, 2012)—thus facilitating survival and reproductive fitness by optimally directing competitive effort. In humans, social comparisons essentially assist individuals in obtaining accurate knowledge of their opinions and abilities (Festinger, 1954). Social comparisons vary by direction of comparison, namely whether one compares with superior (i.e., upwards comparison) or inferior others (i.e., downward comparisons). On one hand, upward comparison can motivate people to aspire to be more like their comparison target. For instance, when people are motivated to develop mastery over a task, high-performing others become a source of information (Butler, 1992; Darnon et al., 2006). On the other hand, upward comparison can also be threatening (Muller & Fayant, 2010). In an experimental study, upward social comparison is induced by simultaneously providing participants with performance feedback on two different tasks (Tesser & Campbell, 1980). When participants were led to believe their performance was poorer than the confederates’, they avoided choosing to work on that task. This finding suggests that people avoid contests with superior others when making upward comparisons. On occasions when people involuntarily submit to superior others, they inhibit aggressive behaviors and experience low-arousal negative affects such as depression (Allan & Gilbert, 1995; Buunk & Brenninkmeyer, 2000; P. Gilbert & Allan, 1998; Price et al., 1994). Upward social comparisons are also associated with dissatisfaction and decreased self-esteem (Emmons & Diener, 1985; Taylor & Lobel, 1989; Wheeler & Miyake, 1992).

In contrast, downward social comparison involves people comparing themselves to individuals who are worse off than they are. Such comparisons allow people to adopt a different standard of evaluation, which changes their perception of the situation to be more acceptable—or, in some cases, favorable (e.g., VanderZee et al., 1995). This subsequently facilitates the improvement of one’s mood and well-being (Pyszczynski et al., 1985; Wills, 1981). Chronic pain patients who employed downward comparisons experienced lower levels of depression (Jensen & Karoly, 1992). In essence, in social comparison, people derive various outcomes, including an evaluation of themselves (Festinger, 1954), regulation of emotions and well-being (Taylor & Brown, 1988), and aspirations to improve their skills or abilities (Wood, 1989).

Evolutionary Mismatch and Its Manifestations in Social Media

The basic tenet of evolutionary psychology considers cognitive and affective processes and behaviors generated as outputs of evolved psychological mechanisms. These mechanisms emerged from natural selection to address adaptive problems in survival and reproduction in ancestral environments (Buss, 1995; Lewis et al., 2017; Tooby & Cosmides, 1990, 1992). Some psychological adaptations are species-typical; they are inherited and developed reliably because they consistently and effectively address challenges to survival and reproduction over alternate designs over human’s evolutionary history (Buss, 1995). Each mechanism receives inputs from the environment as stimuli and processes these inputs with algorithmic decision rules to generate outputs in the form of cognition, emotion, attitudes, and behaviors (Buss, 1995; Lewis et al., 2017).

Psychological mechanisms evolved with their respective EEA, which meant that our evolved psychology is well-adapted to the conditions of the evolutionary past. However, the modern environment we live in dramatically differs from that of our ancestors. An evolutionary mismatch occurs as the environment changes more rapidly than the time required for psychological mechanisms to adapt (Li et al., 2018). This adaptive lag affects psychological mechanisms by causing them to detect either missing or substantially different input, generating outputs mismatched to its evolutionary function (Li et al., 2018, 2020). A classic example of the evolutionary mismatch is our preference for food high in caloric value (i.e., sweet food). Such preference is well-adapted for a calorie-scarce ecology (Eaton et al., 1996). However, in modern environments where manufactured sugar is abundant in food, this evolved preference now leads people to overconsume sugar, resulting in lifestyle diseases such as diabetes (Gluckman & Hanson, 2006). An evolutionary mismatch can occur when natural or human-induced environmental changes are forced upon an organism (Li et al., 2018). For instance, certain night-active insects rely on moonlight to provide navigational cues, and the widespread use of artificial lighting has disrupted their previously adaptive behavioral pattern (e.g., Málnás et al., 2011). This type of mismatch is known as a forced mismatch. Environmental change can also cause certain environmental stimuli to become favored over the stimuli the mechanism is designed to process (Li et al., 2018). For example, Giant Jewel Beetles rely on the size, color, and texture of potential mates as cues to their suitability. However, they seem to prefer to mate with beer bottles, primarily because of their relatively large size compared to female beetles (Gwynne & Rentz, 1983). This type of mismatch is known as a hijack mismatch.

Psychological mechanisms detailed in the previous section—social monitoring system, sociometer, and social comparison—evolved throughout human evolutionary history as they provided adaptive value in ensuring one is included as a valuable member of social groups. However, the digital revolution humans have been experiencing for the past decades constitutes merely a microscopic period in human evolutionary history, which spans about 100,000 years back to the late Pleistocene (Rogers & Jorde, 1995). The digital space that implicates modern living across multiple domains of life vastly diverged from humanity’s past hunter-gatherer lifestyle, creating the potential for evolutionary mismatches in our everyday lives (Giphart & van Vugt, 2018; van Vugt et al., 2024). For instance, exposure to information is likely limited to local ecology due to geographical limits. However, with modern connectivity, modern humans are constantly exposed to remote issues, such as natural and economic disasters elsewhere. This input can overload our evolutionarily endowed cognitive system, generating undue stress and decreased work productivity (Hughes et al., 2024). Likewise, social media is an evolutionarily novel environment comprising features one would not have encountered in the ancestral past. Social media now allows people to interact with other people they would otherwise never meet, exchanging social information with evolutionarily novel qualities and supernormal social cues that psychological mechanisms prioritize (e.g., likes and followers). In the following sections, we will discuss how various social media characteristics prompt psychological mechanisms governing social functioning in ways that produce maladaptive consequences (Li et al., 2018; van Vugt et al., 2024). Specifically, we highlight how social monitoring, sociometer, and social comparison processes take in evolutionarily novel cues on social media—namely, exponential social information, weak and latent ties, vague social cues, curated social information, and anonymity—as input to produce a suite of behavioral and psychological outcomes such as excessive social media usage and dissatisfaction (see Table 1).

Table 1 Evolutionary mismatches on social media: evolved psychological mechanisms hijacked/forced by evolutionarily novel cues on social media to result in social media ills
Full size table

Exponential Social Information

Social media platforms enable users to forge new social relationships, nurture existing ones, and engage with a diverse array of people with unparalleled ease. Registered members can actively search for other users through a search engine, peruse their profiles, and establish connections. Over the majority of human evolutionary history, people resided in small tribes and villages where any event concerning one person would typically be no more than three degrees of separation away from anyone else (Christakis & Fowler, 2009). Consequently, social information held substantial relevance and importance due to the intimate size of the village community (Yong & Li, 2018). As such, we likely evolved to be highly sensitive to all kinds of social information. An average adult Facebook user boasts a network of 338 “friends” (Blease, 2015), far more than a person from ancient times would have ever encountered throughout their entire lifetime (100–230 individuals; Dunbar, 1992; Hill & Dunbar, 2003).

Social media satisfies people’s need for belongingness (Nadkarni & Hofmann, 2012). However, social media also exploits our evolved psychology to treat all social information as important, regardless of the personal relevancy of the information. In this manner, the social monitoring system and sociometer are forced to monitor the extensive volume of social data on social media for social cues and inclusionary status. Self-esteem is negatively related to time spent on social media and problematic social media use (Gori et al., 2023). Furthermore, self-esteem mediates the relationship between anxious attachment styles, time spent on social media, and problematic social media use. These sets of findings illustrate the operation of the sociometer on social media; those who are insecure about their social relationships spend more time on social media, sometimes to the point of problem use. Another study revealed that students dedicated more time to observing content on Facebook than actively creating and posting content (Pempek et al., 2009), indicating that social media platforms feed on people’s desire to monitor social information.

Furthermore, current evidence suggests that exponential social information on social media can compete and usurp attentional resources that would otherwise be allocated to in-person social interaction (see Sbarra et al., 2019 for an in-depth review). Among spouses, husbands’ use of social media negatively affects marital quality (Dew & Tulane, 2015). Despite promoting well-being when online relationships are satisfying, social media are linked to a more negative offline social relationship, reducing well-being (Hu et al., 2017). Social information exhibited on social media is so appealing that being away from social media can become a source of distress. “Fear of missing out” (FOMO; Przybylski et al., 2013) is defined as a persistent concern about potentially missing out on enjoyable experiences that others might be having when one is not present. FOMO represents the cognitive bias related to one’s perceived social resources and is a key driver of the severity of problematic use (Dempsey et al., 2019; Elhai et al., 2020; Franchina et al., 2018). Arguably, this anxiety reflects the functioning of the sociometer, which becomes ensnared by the continuous stream of social information. Consequently, individuals are compelled to stay connected to consistently monitor social cues from their online “friends” to assess their inclusionary status. Neurological evidence appears to support this view. FOMO has been linked to activating the right middle temporal gyrus, a brain region associated with processing social inclusion signals (Liu et al., 2021).

Weak and Latent Ties

The ease with which people can establish connections with others adds to the extensive “friend” networks found on social media, which often include individuals who are not close friends or people they have never met in person (Duggan et al., 2015; Osman, 2023). People can leverage their networks’ connections to expand their networks further and develop meaningful friendships. However, within these vast online social networks, the relationships between social media users and most of their connections are typically either latent—individuals have only fleeting awareness of each other—or weak, individuals only occasionally reach out to each other for information or other forms of minor support (Ellison et al., 2011).

Yet, our evolved social monitoring system and sociometer have not adapted to the novel affordance of establishing weak and latent social ties. Social interaction in humans’ small ancestral habitats is likely with those with whom they have strong ties. This is because close bonds and strong ties are the foundation of cooperation among social animals (Mitani, 2009; Seyfarth & Cheney, 2012; Silk, 2009). Across multiple social species, natural selection has favored individuals motivated to form strong, enduring friendships that aided each other’s survival and reproductive successes (Seyfarth & Cheney, 2012). In the previous section, we discussed the many advantages of group living due to cooperation. These advantages require strong ties between group members, particularly when cooperation is (a) not risk-free or (b) yields unequal benefits for participants. Strong ties reduce defection risk and increase reciprocation odds (Bissonnette et al., 2015). This is not to say that weak ties are always disadvantaged within the EEA. Weak ties aid the diffusion of information to and from the peripheries of one’s social network, facilitating information exchange (Granovetter, 1983). Weak ties may also be significant when the division of labor creates within-group segments and when coordination is required between segments (Granovetter, 1983). However, because the adaptive problems that strong ties can solve are highly impactful on one’s fitness and more likely to be frequently encountered, adaptations that emerge from natural selection are more likely to be attuned to strong ties as a core design feature (see Lewis et al., 2017, p. 358).

By allowing content and communication to be distributed beyond geographical limitations, social media has also facilitated the establishment of social relations that would not have existed otherwise. Consequently, our primitive mind cannot differentiate ties forged on social media from strong ties forged from deep and meaningful personal interactions. Human’s inability to distinguish between entities that do not exist within the EEA has been well documented. Parasocial relationships—unilateral relationships people have with media characters or celebrities—are made possible (Rubin & Step, 2000). These relationships engendered feelings of intimacy and closeness with media characters mirroring a real social relationship (Dibble et al., 2016). People who watch certain kinds of TV shows appear to engage in parasocial friendships and reported being more satisfied with their friendships as if they had more friends and socialization (Kanazawa, 2002). This mimicking effect of parasocial relationships prevailed on social media. In an experimental study (Rasmussen, 2018), participants were randomly assigned to view content from popular (> 2 million subscribers) or moderately popular (100,000 to 250,000 subscribers) beauty vlogs—videos in which content creator channels document daily life, thoughts, and activities—or cartoon clips of similar video length on YouTube. Results suggest a positive association between content creator’s popularity with feelings of knowing the content creator and feeling as though viewers are content creators are their friends.

These studies demonstrate that our evolved psychology frequently places a similar emphasis on our extensive yet weak or unilateral online relationships as it would on ancestral humans’ strong but limited social networks. This means that the social monitoring system and sociometer now monitor social information from weak or unilateral ties that may not significantly contribute to one’s inclusionary status. Monitoring not just large but also weak social relationships amplifies the level of effort individuals dedicate to monitoring their social surroundings, leading social media users to expend an undue amount of time and energy in their pursuit of monitoring inclusionary status and social cues.

Vague Social Cues

Social media platforms frequently integrate “like” buttons, which provide people with a convenient method for offering social validation. Additionally, the visible count of a user’s followers serves as another indicator of being validated by others. Instagram’s prestigious “blue tick” further amplifies this notion, typically granted to individuals with a considerable following and influence. One of the most important implications of our desire for belongingness is that the more we like others, the more driven we are to cultivate close relationships with them. In doing so, we use approval and liking to build, maintain, and assess the quality of our relationships with them. If we generate content or express opinions others approve of, they will approve of us, too. Content creators use likes and following to determine which content garners the highest reach and approval. Conversely, content creators and users chase likes and followers using trendy, appealing content. Several research studies show that interacting with social media content with varying levels of likes can influence us in ways. For example, people tend to see social media news with more likes as more credible (Luo et al., 2022). After users post content that receives more likes and comments, they experience positive emotions such as excitement, motivating more frequent posts (Stsiampkouskaya et al., 2021). Conversely, users change the type of content they post if they feel sad after not receiving the expected level of engagement. People also remember content that they post better if they receive more likes (Zell & Moeller, 2018).

One reason social media cues are so powerful is that they capitalize on our evolved need for social validation—they give social media users opportunities to gauge social approval. People are fixated on receiving likes and gaining followers, as failing to respond to social disapproval compromises our reputation. Having a good reputation is an essential social commodity people strive for (Baumeister, 1982; Van Vugt & Hardy, 2010); reputational information allows people to know who they are dealing with, which can determine the extent of cooperation they may receive from others (Dunbar, 2004; Emler, 1990). Accordingly, people are concerned about their reputations, as having a bad reputation may result in stigmatization and social exclusion (Cavazza et al., 2014; Kurzban & Leary, 2001), disqualifying them from participating in beneficial or harm-mitigating activities.

Nevertheless, “Likes” or an additional following on social media may not indicate social approval per se. People may give “Likes” for several reasons and interpret “Likes” in various ways rather than as actual appreciation of the content (Jungselius, 2019). For instance, receiving a significant number of “Likes” on Instagram from someone outside one’s own network is sometimes interpreted as a strategy to gain “Likes” back in return. Moreover, it was shown that Facebook content receives more “Likes” and comments due to the dynamic recommendation of content based on the Facebook algorithm (Wang et al., 2013). The more a user posts and receives comments, likes, and follows, the more likely the algorithm makes them visible to other users. This recommendation system makes “Likes” and follows more of a reflection of artificial algorithmic decision-making. Given the marketing incentives to be prioritized by the algorithm, some businesses and content creators are also known to purchase underground services that artificially inflate “Likes” and following counts (Ikram et al., 2017), further diluting the quality of “Likes” and followers as a metric of social approval. In addition to being indicative of someone’s reputation and others’ approval of our content, these social media metrics have also acquired an evolutionarily novel quality of being a performance medium in a competition of being favored by social media algorithms.

Yet, the sociometers interpret ambiguous metrics such as Likes and followers as inclusionary cues. In an experiment where the number of likes for a personal photograph was manipulated, those who received more likes for their photograph reported heightened self-esteem, whereas not having enough likes resulted in poorer mood and lower levels of self-esteem (Burrow & Rainone, 2017). The social monitoring system is likewise sensitive to social media cues as it would be to social cues in the real world. Participants who received fewer likes from a fixed number of other social media users reported feelings of higher levels of psychological distress and negative affect (Poon & Jiang, 2020). This effect is even stronger for those who are already bullied and socially excluded offline (Lee et al., 2020). Neuroimaging studies further corroborate with psychological studies. If likes cue social approval, and social approval sated our need for belonging, then having likes on social media should manifest biological markers of rewards. Indeed, fMRI imaging shows that brain activity following the experience of receiving like activates brain regions associated with rewards (L. E. Sherman et al., 2018).

The desire for approval on social media creates an economy of likes and followers. With the vast number of users on social media platforms, having enough likes and followers is a moving yardstick. As such, the hijacked sociometer may likely signal to people that their inclusionary is under threat, which can explain why people are fixated on getting more likes and followers (Nesi & Prinstein, 2019). Consequently, some people develop an unhealthy fixation on sharing content that may compromise their self-image, including images of themselves in revealing attire or participating in sensational or controversial actions to garner increased attention (Chua & Chang, 2016). Posting sexual content puts girls at risk of predation, sexual harassment, and bullying (Haidt, 2024). Although preliminary, evidence also suggests social media metrics like follower and following count can reduce well-being. A higher follower and a lower following count are associated with more addictive social media use and a higher likelihood of cyberbullying victimization, which in turn reduces happiness (Longobardi et al., 2020).

Curated Social Information

Social media allows users to identify themselves to the extent they desire. This feature not only allows users to control how much information about themselves they want to reveal, but it also allows users to curate and present a particular version of themselves to the audience they prefer (Caldeira et al., 2021; Johnson & Ranzini, 2018; Márquez et al., 2023; Zheng et al., 2020). In doing so, they fulfill self-presentation motives by showcasing the aspects of their identity they wish to emphasize while concealing those they prefer not to highlight. This digital self-curation offers a level of control over personal image and narrative that allows individuals to easily manipulate or magnify their actual levels of attractiveness and success to portray a “perfect” life and enhance their reputation (Blease, 2015; Gonzales & Hancock, 2011; Siibak, 2009; Vogel & Rose, 2016). Yet, our psychological mechanisms may not have evolved to discount the artificially generated success or physical attractiveness cues and continue to respond to these cues as though they are people’s actual levels of achievements and physical attractiveness.

Human learning is biased toward displays of skills, success, and prestige because these cues probabilistically contain instructive information that aids survival odds and reproductive success with little learning cost in energy, time, and opportunities (Henrich & McElreath, 2007; Kendal et al., 2009). However, on social media, information about people’s relative standing regarding wealth and social status is less likely to reflect one’s actual fitness. Due to high social interconnectivity on social media, people are frequently exposed to the presence of ultra-high-net-worth individuals (Michaelidou et al., 2021) and their pretenders, who often flaunt their wealth on social media (Chen, 2022). These individuals are, at best, rare in ancestral environments, where resource hoarding yields little fitness benefit (Winterhalder, 1996a). Because our evolved psychological mechanisms behind social comparison are strongly biased in interpreting these cues as people’s actual wealth and status, people are likely to experience envy and dissatisfaction with their lives (Yong & Li, 2018). When people are drawn to the perceived affluence of others, the perceived material gap between self and others leads to life dissatisfaction (Yang & Oliver, 2010).

Likewise, physical attractiveness standards are usually inflated on social media, sometimes to extreme degrees. What is considered beautiful is likely highly heterogeneous and dependent on the specific advantages conferred upon those who possess the traits across a multitude of human ecology (Bovet, 2018). Yet, women and adolescent girls are frequently exposed to unrealistic beauty standards on social media that distort the social comparison process (Scully et al., 2023). Furthermore, photos expressing one’s beauty may also be digitally manipulated and non-representative relevant mate competition. In an experiment exposing original or manipulated photos to adolescent girls, girls with higher social comparison tendencies reported poorer body image (Kleemans et al., 2018). Results also indicate that manipulated photos were rated more positively than the original, indicating that social comparisons are made based on distorted information. The negative effect of social comparison on social media can evidently extend to negative body image concerns. When people engage in social comparison with online friends and celebrities on social media, they may develop body image dissatisfaction and, in serious cases, body dysmorphic disorder symptomologies (Ho et al., 2016; Modica, 2020; Ryding & Kuss, 2020; Scully et al., 2023).

Although social comparison promotes body image dissatisfaction in both males and females, some research has suggested that the presentation of idealized beauty impacts body image slightly more negatively for adolescent girls than boys (e.g., Feingold & Mazzella, 1998; Haidt, 2024; Hargreaves & Tiggemann, 2004). This asymmetry likely reflects the nature of intrasexual competition among females, where female reproductive values are strongly indicated by physical appearance and attractiveness (Buss, 1988; Buss & Schmitt, 1993). With physical appearance as a primary definition of mate value for females, it is unsurprising that girls pay more attention to other’s expressions of beauty on social media and subsequently be more negatively affected by social comparison online.

Anonymity

Social media enables anonymity. While some users prefer to present a curated version of themselves, others prefer to minimize their self-presentation on social media. People may not only share minimal personal information on their profiles without revealing their complete identity, but they may also decide to create social media accounts using pseudonyms or screen names instead of their real names. They may also create temporary or disposable accounts, which facilitates concealing one’s identity on these platforms (Leavitt, 2015). While this can offer opportunities for free expression and open dialogue, especially for sensitive issues and stigmatized groups (Rains, 2014), anonymity on social media platforms also carries the potential for harm.

Without a mediating communication medium for the majority of human evolutionary history, communication in close proximity likely dominates the majority of human interaction. This is evident by the complex layering of evolved communication systems that facilitate face-to-face interaction in various domains of life (Thibault, 2008, p. 292). This premise also implies that social interaction in human evolution past is likely conducted with identities largely known to the participating parties, given the small group size and the lower degree of separation within a social network (Christakis & Fowler, 2009; Hill & Dunbar, 2003). Considering how social interaction occurs in ancestral conditions is important because not all social interactions took place with good faith over our evolutionary history. When opportunity arises, personal interest sometimes precedes collective interest (Axelrod & Hamilton, 1981; Ferriere et al., 2002). Within-group relations are occasionally aggressive or even violent (Buss & Shackelford, 1997). Given sufficient incentive, dominating individuals may enact aggression against ingroup members to gain certain benefits. The act of bullying, for instance, offers benefits such as access to social and material resources, deterrence from being aggressed against, and some conferences of prestige (Volk et al., 2012, 2022).

Punishment strategies via moral condemnation (DeScioli & Kurzban, 2013) and ostracism (Kurzban & Leary, 2001) emerged to deal with adaptive problems associated with within-group conflicts. Under the possibility of punishment, when individuals encounter opportunities to enact antisocial or aggressive behaviors, they face a cost-and-benefit trade-off of physical retaliation, condemnation, and ostracism against the direct benefit they would gain from their transgression. However, human’s ability to punish cheaters and aggressors hinges on their ability to identify perpetrators. In order to condemn or eject undesirable group members, people must be able to determine (a) what wrong has been done and (b) who did it. Anonymity on social media decouples the transgression from the aggressor’s identity. With anonymity, aggressors can expect no reputational damage. Given the remote nature of the transgression in cyberbullying, there will also be no immediate risk of bodily harm from retaliation. By concealing the aggressors’ identity, future retaliation is also made unlikely. Remote transgression no longer requires the computation of relative differences in physical strengths and prowess between the aggressor and the victim. As such, anonymity allows aggressors to bypass social comparative processes regulating competitive efforts, which includes aggression against co-specifics. By conferring users the ability to conceal their identity, the restricted means for aggressors to be meaningfully punished likely encourage abuse. Indeed, studies show that anonymity promotes online antisocial and aggressive behaviors. Anonymity emboldens people to cyberbullying on social media (Barlett, 2015), and cyberbullies also typically expect no consequences for their actions (Pettalia et al., 2013). Other studies have also found that online forums that allow for anonymous comments see double the rate of incivility and ad hominem attacks in discussing sensitive topics compared to forums that do not permit anonymity (Santana, 2014). Laboratory experiment further corroborates these findings. Precisely, when social interactions took place in experimentally manipulated conditions of anonymity, lack of visibility of social partners, and lack of eye contact—typical characteristics of anonymous social media interactions—it results in the highest level of incivility between participants during chat sessions (Lapidot-Lefler & Barak, 2012). Online trolling—the act of antagonizing other online users for personal entertainment—is also facilitated by anonymity (Nitschinsk et al., 2022).

Moving Forward: The Digitalized Future of the Social Animal

So far, we have examined how evolutionarily novel features of social media force and hijack people’s evolved psychological mechanisms regulating different social processes. We suggest that people’s amenability to social media and many social media ills—harm and undesirable outcomes stemming from social media use—are caused or exacerbated by the mismatch between features of social media and our evolved psychological mechanisms regulating different social processes. Although our evolved psychological processes were adaptive in the environment of our ancestors, they can bring about negative consequences in our modern and novel social media usage. While an evolutionary psychology perspective points to the fact that humans are inherently social animals and are “addicted” to social information, it does not imply that social media ills are inevitable. By considering how social media features feed into our evolved psychological mechanisms regulating social processes, we suggest ways social media companies and policymakers can effectively redirect people’s tendencies to lessen these problems.

Controlling Exponential Social Information by Emphasizing Strong Ties

The ease of establishing social connections on social media and the social network reach it affords contribute to the exposure to infinite social information. The forced mismatch between this affordance of social media and (a) human’s evolved neocortex size to maintain a limited network size (i.e., 150 individuals Dunbar, 1992; Tooby & Cosmides, 1996) and (b) evolved inclination to tend to social information associated with strong ties results in the monitoring of an evolutionarily novel large social network size most of whom are people who are not close friends or people they have met (Duggan et al., 2015; Osman, 2023). In this manner, the social monitoring system and sociometer are forced to monitor the extensive volume of social data on social media for social cues and inclusionary status, regardless of the personal relevancy of the information and if it significantly contributes to one’s inclusionary status, leading social media users to expend an undue amount of time and energy in their pursuit of monitoring inclusionary status and social cues.

As it became increasingly evident that social media poses a health threat to users akin to that of other harmful substances (Twenge et al., 2022), the call for regulating social media use to curb its effects—especially that of social media addiction—has been on the rise in recent years (Office of the Surgeon General, 2023). Proposed regulations for managing excessive social media use involve limiting the consumption of social media (Faulhaber et al., 2023; Root & Ashford, 2024). In an internal memo, Andrew Bosworth, a longtime Facebook executive, likened social media use to sugar—moderate social media consumption was harmless, whereas large doses of social media were harmful (‘Lord of the Rings, 2020 and Stuffed Oreos’, 2020). Accordingly, to prevent the detrimental effects of social media use, one should exercise self-control to limit consumption. The Social Media Addiction Reduction Technology Act (SMART Act, 2019) was also introduced in the US Congress, which calls for the ban of infinite scrolling—the endless stream of content on a user’s feed—and autoplay, a feature allowing content to be displayed immediately following the end of the previously displayed content, to impose caps on the amount of time people spend on social media (SMART Act, 2019).

Implicit in measures championing limiting the usage of social media is the assumption that the usage of social media itself causes social media ills. As Haidt (2021) articulated, “Social media platforms are not like sugar.” Social media has rewired the way people gather information in this information age (Haidt, 2024); limiting and banning people from using social media is not feasible as it impacts everyone, not just individuals who overindulge. Moreover, imposing social media consumption discounts the benefits of social media; for instance, parasocial interactions on social media can increase trust between brands and consumers (see Labrecque, 2014). Hence, more deliberate and careful solutions are required to minimize the harmful effects of social media use while maximizing the opportunities it brings.

Advocates have rightfully identified that social media exploits human psychology and brain physiology. However, it is not the usage of social media per se that causes social media ills; it is the exposure to a large amount of social information, most of which belongs to weak latent ties and strangers, that accompanies social media usage. From an evolutionary psychology perspective, humans have evolved to process social information. Specifically, we have evolved to pay attention to limited social information, with constraints imposed by biological capacity (Dunbar, 1992, 2009; Hill & Dunbar, 2003),

As such, rather than limiting social media consumption, adjusting the exposure to smaller network sizes on social media—one that is more evolutionarily familiar than current evolutionarily novel large social networks (Lim et al., 2021)—may be a more effective solution. Also, since social information from individuals we share strong social ties with provides more important information about social inclusion and belongingness, algorithms can be adjusted to prioritize social information belonging to close others rather than others with a higher degree of separation in one’s social network. This proposal can also allow the occasional social content from social media “influencers” and businesses to reach a desired reach while maximizing the match between social information and those that evolved psychological mechanisms are designed to process. Defining an information boundary allows social information from close others to fulfill its adaptive function of satisfying our innate need for belongingness while, at the same time, allowing people to benefit from the rapid diffusion of information from weak ties (Granovetter, 1983) that have proven helpful in various aspect of life, such as for job-seeking (Brown & Konrad, 2001).

Interrupting the Hijack: Promote Conscious and Educated Use of Social Media

As we have discussed, social media contains features such as buttons, a visible count of followers, and the “blue tick.” These evolutionarily novel features hijack our psychological mechanisms regulating social inclusion assessment as they make it convenient for people to quantify reputation, social acceptance, and inclusion; consequently, people rely heavily on it and are fixated on pursuing “likes” and followers (Stsiampkouskaya et al., 2021). Additionally, social media rewired individuals to “perform”—engaging in activities aimed at impressing others to create and develop a personal brand (Haidt, 2021, 2024). People’s psychological mechanisms regulating social comparison are also hijacked to attend to curated and manipulated physical attractiveness and success cues. In essence, people prefer to use these cues to gauge inclusionary status and identify comparison targets.

Much like TV friends and pornography (Kanazawa, 2002; Money & Ehrhardt, 1972), humans did not encounter these stimuli in the ancestral past, but these stimuli share similar attributes to actual social cues and comparison targets; now, these stimuli replace actual social cues as input to psychological mechanisms regulating social inclusion and social comparison. People will remain amenable to these evolutionarily novel social cues because humans have not evolved to learn how to differentiate these novel social cues from actual social cues. Perhaps, the solution to managing the hijacked psychological mechanisms assessing for social inclusion and regulating social comparison is to consciously override people’s automatic responses to these stimuli through social media literacy education.

Media literary—media knowledge and criticism—has successfully reduced media realism, desirable media portrayals of social actors, and harmful media effects (Jeong et al., 2012; Vahedi et al., 2018; Xie et al., 2019). Extrapolating from studies on media literacy, social media literacy programs can be developed to increase awareness of the perceptions of social cues online (e.g., Schreurs & Vandenbosch, 2022). Specifically, there is a need to highlight that there is no one universal interpretation of “likes” (Jungselius, 2019), and hence, they may not accurately reflect the opinions other people hold (Prinstein, 2023).

People are also generally aware of the tools others use to manipulate their photos in curation, yet psychological mechanisms lead them to trust what they see (Prinstein, 2023). Social media literacy programs can also include critical thinking about people’s curation efforts, which can help override the automatic processes and make salient that what they encounter online may not reflect real life (Prinstein, 2023). Successes from fake news literacy programs that emphasize the authentication process, such as looking for credibility cues, lend confidence to such methods to facilitate discernment (Chan, 2022). Given the pervasiveness of digital media in modern life, it is no longer a remote consideration for integrating media literacy, including social media literacy, into the basic educational system of developed societies.

Restoring Accountability

Social media platforms allow for the flexibility of how much social information one is comfortable providing when setting up one’s accounts; people may create social media accounts using pseudonyms or screen names instead of real names and conceal their identity on these platforms. Anonymity is evolutionarily novel—the introduction of anonymity rids accountability, significantly impacting cost–benefit computations implicated in various mental processes. Specifically, the cost–benefit analysis tilts asymmetrically, such that the cost of engaging in behaviors that cause harm becomes almost non-existent, thus enabling toxic behaviors like cyberbullying. As discussed, robust research findings suggest anonymity promotes antisocial behaviors (Barlett, 2015; Lapidot-Lefler & Barak, 2012; Nitschinsk et al., 2022; Santana, 2014).

To curb antisocial behaviors on social media, it is necessary to introduce accountability that reflects familiar evolutionary states. This can be achieved by introducing some extent of user verification while allowing users to use pseudonyms and screen names. Social media platforms should consider a verification process—through a third party or nonprofit—to ensure that the user is identifiable and has met the necessary social and legal requirements to partake in certain activities on social media (Haidt, 2021, 2024). Reintroducing some form of accountability—even perceived accountability—highlights the potential cost of carrying out antisocial behaviors. Accountability facilitates punishment as people are more willing to judge transgression as wrong when circumstances are unambiguous about the perpetrators and their actions. In a series of judgment experiments (DeScioli et al., 2011a, b), participants were tasked with judging if an actor was wrong in a scenario that resulted in the death of others. They were also asked about the years in prison the actor deserved. The experiment varies the cause of death in the scenario as direct (vs. indirect) and the transparency of evidence of the actor’s wrongdoing. Across all four studies, people deemed the actor’s action wrong when the circumstance was transparent and thought the actor deserved a greater punishment, even when the cause of death was indirect. The willingness of others to punish is likely taken into account in the cost–benefit computation toward committing a transgression, as studies have also shown that people are far less likely to display antisocial behavior when the possibility of punishment is made salient. Across two studies, participants made less direct attempts to steal money from another party when they were aware of the possibility of receiving a financial sanction (DeScioli et al., 2011a, b). It must be, however, essential to realize that punishment must be tangible and have direct consequences for the perpetrator. Simply banning and removing individuals from the platform does not constitute meaningful punishment. At best, it deprives users of the benefits of the platform with no restorative justice; at worst, users can recreate another account and continue the cycle of toxic behaviors (e.g., Kou, 2021).

Anonymity can be beneficial—it allows for free expression and open dialogue for sensitive issues and people from stigmatized groups (Rains, 2014); as such, policing blanket policy to remove anonymity removes this opportunity for people. The suggestion of introducing user verification while still allowing people to be anonymous online provides the possibility of keeping this benefit of social media while minimizing the harmful behaviors that people may be encouraged to participate in, given the perception of complete anonymity.

Reversing the Mismatch with Social Policies and Changing Social Norms

Throughout this review, we have characterized social media as an evolutionarily novel environment filled with input cues that our evolved psychological mechanisms are not designed to process, thus generating maladaptive behaviors associated with problem use. By this view, perhaps, the most theoretically parsimonious intervention we can suggest is to limit and supervise the exposure of evolutionary novel inputs (i.e., social media contents and features) in mismatched environments (i.e., social media) to vulnerable minds. Doing so involves introducing social policies and changing social norms regarding social media use to alter the perception that being constantly connected to the online world is necessary. Several experts have offered helpful recommendations along these lines. For instance, Haidt (2024) provided recommendations beyond simply restricting screen time and access to potentially harmful online content; he also advocated for restoring age-appropriate developmental tasks away from social media (Table 2). Schools and parents were recommended to design phone-free time and spaces to allow children to focus on adult-supervised enrichment activities without the interference of smartphones and social media. By introducing programs and activities that promote self-efficacy and social skills, we may be able to restore formative environments that more closely reflect those of our evolutionary history, which facilitates the development and acquisition of socially, ecologically, and culturally relevant skills (see Periss & Bjorklund, 2011). The emphasis on moving beyond restricting smartphone and social media use is important because merely imposing blanket restrictions (e.g., age restriction and screentime restriction) fails to consider individual differences in youth’s maturity and competence that determine the extent of vulnerability (American Psychological Association, 2024; Todres, 2023). Blanket restrictions also remove the potential benefits of social media use and create an unwarranted impression that social media use does not lead to harm after certain conditions are met (e.g., after a certain age).

Table 2 Summary of proposed interventions and recommendations by Haidt (2024)
Full size table

Nevertheless, we recognize that our recommendations in the present review require significant political, social, and cultural paradigm shifts. Furthermore, we also acknowledge several systematic societal problems, such as social inequality, which may contribute to, exacerbate, and complicate issues associated with social media ills (Odgers, 2018; Odgers & Jensen, 2020). These problems require concurrent addressing to maximize the mitigation and prevention of harm. The mission ahead to reduce social media ills for researchers, practitioners, policymakers, companies, educators, and individuals may be long and arduous. However, we have faced similar crossroads in the past, where there have been positive changes following the introduction of social policies and a change of social norms regarding smoking (e.g., Jha et al., 2006) and seat belts use (e.g., Shults et al., 2004). These past successes offer an optimistic outlook on the digitalized future of the social animal.

Final Note

Social media is unquestionably an integral part of people’s lives in this modern world, shaping lives both positively and negatively. It is naïve to conclude that social media is good or bad. More importantly, the critical question pertains to the precise mechanisms through which social media generates adverse consequences (e.g., addiction, FOMO, low self-esteem) in its use. Despite the evidence on how social media can be harmful, the understanding in this area is far from definitive. Although existing research and theories have revealed who is most vulnerable to social media harm (George & Odgers, 2015; Kreski et al., 2021; Odgers, 2018; Odgers & Jensen, 2020) and outlined how problem use develops and manifests (Davis, 2001; LaRose et al., 2003), it does not explain why social media are so appealing to people, that problem use has become such a pervasive concern. This article aims to provide an evolutionary framework for understanding why people are amenable to social media and how their amenability can result in “social media ills”—the adverse outcomes commonly associated with social media. Specifically, this article outlines the case for potential evolutionary mismatches between social media and psychological mechanisms governing social functioning. We argued that sociometer, social monitoring system, and social comparison processes can be “hijacked” and “forced” by the novel features of social media, compelling people to consume it, sometimes to excessive extents, and resulting in negative outcomes such as lowered self-esteem and social media addiction.

The application of evolutionary mismatch theory in the context of social media has yielded valuable insights. For instance, while social media usage is often linked to declines in self-esteem, recent research findings indicate that this may not hold true, particularly when individuals have social media networks of evolutionarily novel sizes, such as 500 friends (Lim et al., 2021). This arises from the fact that humans have evolved with a neocortex size optimized for maintaining a social network of approximately 150 individuals (Dunbar, 1998). Consequently, the abundant social cues available on social media may not be effectively processed by the psychological mechanisms governing social comparisons. This supports the notion that evolutionarily novel inputs of social media influence the functioning of our evolved psychological mechanisms. Applying the principle of the evolutionary mismatch hypothesis, we also have to discuss the potential direction policymakers, social media companies, other institutions, and individuals can take to tackle social media ills.

We would like to stress that we do not intend for our arguments to be presented as a competing framework with existing research or theories. Instead, we intended our framework to complement by addressing the ultimate cause of social media ills. In our view, not sufficiently addressing the root of social media ills would be akin to treating the symptoms without the cause. With this in mind, we contend that adopting an evolutionary perspective is helpful in elucidating why people are so drawn to social media, sometimes leading to unfavorable consequences. Additionally, this approach facilitates the identification of factors—beyond psychological vulnerabilities—that may predict problematic social media use or result in some people being more susceptible (i.e., easily influenced by) to social media features. For instance, intelligence has been recognized as a factor underlying responses to evolutionarily novel stimuli (e.g., Kanazawa & Hellberg, 2010; Kanazawa et al., 2022). Individuals who were less intelligent faced greater difficulty in distinguishing television characters from real friends (Kanazawa, 2002, 2006). While this has not been comprehensively examined, existing evidence suggests that intelligence can influence how people interact with social media.

Aside from advancing theoretical knowledge, a greater understanding of evolved psychological mechanisms and the role of evolutionary mismatch—particularly, mismatches inducing negative consequences or costs to individuals and society—may be able to lead to more informed problem-solving and public policy (Griskevicius et al., 2012; Li et al., 2018, 2020). Ignoring evolved mechanisms and how they respond to mismatches can yield ineffective interventions. For instance, people are slow to adopt pro-environmental behaviors because the threat of climate change does not present itself in ways that are amendable to human’s evolved psychological biases (Van Vugt et al., 2014). Humans tend to disregard threats with impalpable consequences. This is because our primal minds are optimized for problem-solving in a natural, ecological context with tangible links between behaviors and the environment (e.g., Cosmides, 1989). A slow-moving environmental threat like climate change hardly persuades human intuition. For information about climate change to motivate immediate action, such information is best presented in simple, concrete, and affective terms (Heath & Heath, 2008).

As social media and technology advance, novel social cues will likely get more realistic. With the emergence of generative artificial intelligence (Stokel-Walker & Van Noorden, 2023), we are not far from a future filled with artificial stimuli that evolved psychological mechanisms maladaptively preferred or forced to process. Evolutionary mismatch can potentially introduce negative consequences for psychological health (Li et al., 2020). Understanding the mismatch is essential for scientific research and developing effective ways to address modern problems and their undesirable outcomes in the contemporary world.