Strategic Expression, Popularity Traps, and Welfare in Social Media

TL;DR

An attention-seeking model, distinct from canonical mechanisms of conformity, learning, persuasion, and (mis)information transmission in social networks literature, and the first utilitarian framework defined directly over observable social media platform metrics are introduced, filling a critical gap in the social media literature.

Abstract

Social media platforms systematically reward popularity over authenticity, incentivizing users to strategically tailor their expression for attention. In this paper, we introduce (i) an attention-seeking model, distinct from canonical mechanisms of conformity, learning, persuasion, and (mis)information transmission in social networks literature, and (ii) the first utilitarian framework defined directly over observable social media platform metrics, filling a critical gap in the social media literature. In the model, agents hold fixed heterogeneous authentic opinions and derive (i) utility gains from the popularity of their own posts -- measured by likes received, and (ii) utility gains (losses) from exposure to content that aligns with (diverges from) their authentic opinion. Social media interaction acts as a state-dependent welfare amplifier: light topics generate Pareto improvements, whereas intense topics make everyone worse off in a polarized society (e.g., political debates during elections). Moreover, strategic expression amplifies social media polarization during polarized events while dampening it during unified events (e.g., national celebrations). Consequently, strategic distortions magnify welfare outcomes, expanding aggregate gains in light topics while exacerbating losses in intense, polarized ones. Counterintuitively, strategic agents often face a popularity trap: posting a more popular opinion is individually optimal, yet collective action by similar agents eliminates their authentic opinion from the platform, leaving them worse off than under the authentic-expression benchmark. Preference-based algorithms -- widely used by platforms -- or homophilic exposures discipline popularity-driven behavior, narrowing the popularity trap region and limiting its welfare effects.

Figures (5)

• Figure 1: Authentic Opinions versus Expressed Opinions.
• Figure 2: Authentic Expression by All Agents vs. Autarky. The figure plots the utility gain for each agent type under authentic expression ($c_j = b_j \forall j \in \mathcal{N}$) by all agents ($\Delta U^{\mathrm{auth}}_i$), for a neutral agent (blue line) and opinionated agent (red line). Left Panel ($|b|=0.1$): For low-intensity topics, authentic participation yields positive utility (a Pareto improvement over autarky) for all agents regardless of group size. Right Panel ($|b|=1.0$): For high-intensity topics, minority groups suffer negative utility even under authentic expression, as the cost of exposure to opposing views outweighs popularity and alignment benefits.
• Figure 3: Equilibrium Outcomes vs. Authenticity Benchmark. The figure decomposes welfare effects into Equilibrium levels (dark lines) and Authentic benchmark levels (light lines). Rows 1 & 2: Utility levels under each case for neutral and opinionated agents, respectively. Row 3: The Strategic Effect ($\widehat{\Delta} = U^{eq} - U^{auth}$), isolating the gains/losses under strategic actions. In High-Polarization regimes (left of each plot), neutral agents' posting opinionated content often cause a utility loss for neutrals (Green line $<0$). In Low-Polarization regimes (right of each plot), opinionated agents' posting neutral content benefits neutrals at the expense of opinionated agents (Green $>0$, Orange $<0$). Parameters: $n=100$, varying intensity $|b| \in \{0.1, 0.5, 1.0\}$.
• Figure 4: Welfare (Aggregate Utilities) under Equilibrium and Authenticity. The figure plots aggregate welfare gains from social media interaction at equilibrium ($\Delta \mathcal{W} = \sum \Delta U_i$) relative to autarky and authenticity benchmark outcomes. The blue solid line represents equilibrium welfare, while the red dashed line represents welfare under authentic expression. Notes: Red shaded regions ($\Delta \mathcal{W} < 0$) indicate parameter spaces where social media activity generates aggregate welfare losses, while green shaded regions ($\Delta \mathcal{W} > 0$) indicate aggregate welfare gains. Parameters: $n=100$; $a=0.25$ in Row 1 and 3, $a=0.75$ in Row 3; $\omega^p=2$ in Row 1 and 3, $\omega^p=20$ in Row 2; $\omega^a=1$; $\omega^d=1$.
• Figure 5: Polarization Under RA versus PVM

Theorems & Definitions (6)

• proof
• proof
• proof
• proof
• proof
• proof