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

15:40-17:00, Friday, November 13, 2026

I-206, Boxue Building

Dr. Hakobyan is a microeconomist whose research focuses on Information Economics, Networks, and Dynamic Games. She earned her PhD through the Department of Theoretical Economics at the Higher School of Economics，Moscow and the Luxembourg School of Finance at the University of Luxembourg. Dr. Hakobyan currently serves as a Research Scientist at the University of Luxembourg. She is also an Affiliated Researcher at CESifo in Munich, Germany. Her research explores important questions in economic theory, particularly how information, strategic interactions, and networks shape economic decisions and outcomes.

Recommendation systems determine more than what users see: amplifying a source can also make that source a more important reference point for others. This creates a challenge for accuracy-oriented regulation because the exposure patterns that best promote truth tracking need not coincide with those preferred by users or favored by engagement incentives. We study this tension in a network coordination model in which users observe public and private information, have heterogeneous directional biases, and value coordination with others. Feed composition therefore affects two margins simultaneously: it changes information exposure and the coordination weight placed on recommended sources. We characterize source selection under an accuracy mandate and show that neither informativeness nor a source's own bias is sufficient to determine whom to amplify. Optimal ranking trades off information gains against the network-wide distortion generated by additional coordination influence. We then identify two limits to accuracy-oriented design. First, network design can reduce directional distortion but cannot generally eliminate it completely. Second, accuracy-improving changes need not be privately implementable when users control their exposure. This conflict persists even under complete information: every unilateral cross-cutting reallocation of coordination weight away from an echo chamber raises truth-tracking accuracy while lowering the deviating user's utility. An engagement-oriented platform can reinforce this tension by favoring more validating exposure patterns at the expense of accuracy. The results distinguish information delivery, coordination influence, and user adoption as separate margins of platform design.

For more information of the seminar, scan the following QR code(s) to join Tencent QQ group (904 544 292) or WeChat group named "IAER Seminar", please.

QQ Group

WeChat Group 

(QR code is valid until November 16, 2026)