arXiv subjects

Zaruhi Hakobyan

Publications and source records attributed to Zaruhi Hakobyan.

3 recordsLinked to original sources

arXiv2026

Whom Should a Platform Amplify? Truth, Engagement, and Networked Polarization

Social-media platforms allocate attention by deciding whose content receives reach. We study this problem in a network coordination model in which users learn from observed signals but also care about coordinating with others. Amplifying a source, therefore, spreads information while potentially making that source a more influential coordination reference. For any finite directed network, we show that the effect of an intervention on truth-tracking accuracy decomposes into an information gain and a change in network-amplified bias risk. The latter depends on how sources' influence profiles interact with the network's entire bias composition, so the most informative source need not be the best source to amplify. Broadcasting can improve the use of common information, while amplification of directionally biased sources can reduce accuracy. Integrated communities can outperform segregated ones; users may reject informative cross-cutting links, and engagement-oriented platforms may favor connections that reduce collective accuracy.

econ.TH↗

arXiv2026

Buying the Right to Monitor:Editorial Design in AI-Assisted Peer Review

Generative AI acts as a disruptive technological shock to evaluative organizations. In academic peer review, it enters both sides of the market: authors use AI to polish submissions, and reviewers use it to generate plausible reports without exerting evaluative effort. We develop a three-sided equilibrium model to analyze this dual adoption and derive a counterintuitive managerial implication for journal policy. We show that when AI capability crosses a critical threshold, reviewer effort collapses discontinuously. This transition creates a welfare misalignment: authors benefit from a weakened ``rat race,'' while editors suffer from degraded signal informativeness. Characterizing the editor's optimal constrained response, we identify a strict policy reversal. Before the AI transition, editors should tighten acceptance standards to curb rent-dissipating author polishing. After the transition, conventional intuition fails: editors must loosen acceptance standards while investing in AI detection, because further tightening only amplifies dissipative polishing without improving sorting. We prove analytically that this sign reversal is a structural consequence of the reviewer effort collapse under log-concave quality distributions. Ultimately, addressing AI in evaluative systems requires treating monitoring and loosened selectivity as complementary design instruments.

econ.GN↗

arXiv2026

Strategic Expression, Popularity Traps, and Welfare in Social Media

Social media platforms systematically reward popularity over authenticity, incentivizing users to strategically tailor their expression for attention. In this paper, we introduce (i) popularity as a strategic expression mechanism, distinct from the canonical mechanisms of conformity, learning, persuasion, and (mis)information transmission in social networks, and (ii) a utilitarian framework for measuring user welfare that maps directly to observable platform metrics, filling a critical gap in the social media literature. In the model, agents hold fixed heterogeneous authentic opinions and derive utility gains from the popularity of their own posts -- measured by likes received, and 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. Homophilic algorithms that match users with preferred content -- widely used by platforms -- discipline popularity-driven behavior, narrowing the popularity trap region and limiting its welfare effects.

econ.GN↗