Adam J. Berinsky’s research while affiliated with Massachusetts Institute of Technology and other places

An enormous body of literature argues that recommendation algorithms drive political polarization by creating “filter bubbles” and “rabbit holes.” Using four experiments with nearly 9,000 participants, we show that manipulating algorithmic recommendations to create these conditions has limited effects on opinions. Our experiments employ a custom-built video platform with a naturalistic, YouTube-like interface presenting real YouTube videos and recommendations. We experimentally manipulate YouTube’s actual recommendation algorithm to simulate filter bubbles and rabbit holes by presenting ideologically balanced and slanted choices. Our design allows us to intervene in a feedback loop that has confounded the study of algorithmic polarization—the complex interplay between supply of recommendations and user demand for content—to examine downstream effects on policy attitudes. We use over 130,000 experimentally manipulated recommendations and 31,000 platform interactions to estimate how recommendation algorithms alter users’ media consumption decisions and, indirectly, their political attitudes. Our results cast doubt on widely circulating theories of algorithmic polarization by showing that even heavy-handed (although short-term) perturbations of real-world recommendations have limited causal effects on policy attitudes. Given our inability to detect consistent evidence for algorithmic effects, we argue the burden of proof for claims about algorithm-induced polarization has shifted. Our methodology, which captures and modifies the output of real-world recommendation algorithms, offers a path forward for future investigations of black-box artificial intelligence systems. Our findings reveal practical limits to effect sizes that are feasibly detectable in academic experiments.

The dynamics of random transitive delegations on a graph are of particular interest when viewed through the lens of an emerging voting paradigm: liquid democracy. This paradigm allows voters to choose between directly voting and transitively delegating their votes to other voters so that those selected cast a vote weighted by the number of delegations that they received. In the epistemic setting, where voters decide on a binary issue for which there is a ground truth, previous work showed that a few voters may amass such a large amount of influence that liquid democracy is less likely to identify the ground truth than direct voting. We quantify the amount of permissible concentration of power and examine more realistic delegation models, showing that they behave well by ensuring that (with high probability) there is a permissible limit on the maximum number of delegations received. Our theoretical results demonstrate that the delegation process is similar to well-known processes on random graphs that are sufficiently bounded for our purposes. Along the way, we prove new bounds on the size of the largest component in an infinite Pólya urn process, which may be of independent interest. In addition, we empirically validate the theoretical results, running six experiments (for a total of N = 168 participants, 62 delegation graphs, and over 11,000 votes collected). We find that empirical delegation behaviors meet the conditions for our positive theoretical guarantees. Overall, our work alleviates concerns raised about liquid democracy and bolsters the case for the applicability of this emerging paradigm. This paper was accepted by Martin Bichler, market design, platform, and demand analytics. Funding: This work was supported by the Michael Hammer Fellowship, the Office of Naval Research [2016 Vannevar Bush Faculty Fellowship, 2020 ONR Vannevar Bush Faculty Fellowship, and Grant N00014-20-1-2488], the Office of Secretary of Defense [Grants ARO MURI W911NF-19-0217 and ARO W911NF-17-1-0592], Simons [Investigator Award 622132], the Open Philanthropy Foundation, and the National Science Foundation [Grants IIS-178108, CCF-1733556, CCF-1918421, CCF-2007080, IIS-1703846, and IIS-2024287]. J. Y. Halpern was supported by a grant from the Cooperative AI Foundation [ARO Grant W911NF-22-1-0061]. Supplemental Material: The online appendix and data files are available at https://doi.org/10.1287/mnsc.2023.02470 .

We investigate the impact of exposure to biased media content on political attitudes, focusing on a specific government-controlled broadcaster in Poland. In an experiment (N = 1,943), we randomly assigned participants to either a treatment condition, where they read excerpts from a representative sample of articles published by this media outlet on divisive political topics, or to a control condition. We do not find evidence that exposure to these excerpts influenced political attitudes (such as distrust in the opposition, perceptions of outsider threat, and meta-perceptions of these attitudes in society). Additionally, there was no evidence that mentioning the source of the information affected the persuasiveness of the news articles, nor that party preferences or political interest moderates any of these effects. These findings raise questions about the causal role of biased media in political polarization during periods of high political discord.

Misinformation is a major focus of intervention efforts. Psychological inoculation—an intervention intended to help people identify manipulation techniques—is being adopted at scale around the globe. Yet the efficacy of this approach for increasing belief accuracy remains unclear, as prior work uses synthetic materials that do not contain claims of truth. To address this issue, we conducted five studies with 7,286 online participants using a set of news headlines based on real-world true/false content in which we systematically varied the presence or absence of emotional manipulation. Although an emotional manipulation inoculation did help participants identify emotional manipulation, there was no improvement in participants’ ability to tell truth from falsehood. However, when the inoculation was paired with an intervention that draws people’s attention to accuracy, the combined intervention did successfully improve truth discernment (by increasing belief in true content). These results provide evidence for synergy between popular misinformation interventions.

Recent advancements in generative artificial intelligence (AI) have raised widespread concern about the use of this technology to spread audio and visual misinformation. In response, there has been a major push among policymakers and technology companies to label AI-generated media appearing online. It remains unclear, however, what labels are most effective for this purpose. Using two pre-registered survey experiments (total N = 7,579 Americans), we evaluate the consequences of different labeling strategies for viewers' beliefs and behavior. Overall, we find that all the labels we tested significantly decreased participants' belief in the presented claims. When it comes to engagement intentions, however, labels that merely informed participants that content was AI-generated tended to have limited impact, whereas labels emphasizing the content's potential to mislead more strongly influenced self-reported behavior, especially in the first study. Together, these results shed light on the relative advantages and disadvantages of different approaches to labeling AI-generated media.

Debates on how tech companies ought to oversee the circulation of content on their platforms are increasingly pressing. In the U.S., questions surrounding what, if any, action should be taken by social media companies to moderate harmfully misleading content on topics such as vaccine safety and election integrity are now being hashed out from corporate boardrooms to federal courtrooms. But where does the American public stand on these issues? Here we discuss the findings of a recent nationally representative poll of Americans’ views on content moderation of harmfully misleading content.

Researchers have tested a variety of interventions to combat misinformation on social media (e.g., accuracy nudges, digital literacy tips, inoculation, debunking). These interventions work via different psychological mechanisms, but all share the goals of increasing recipients’ ability to distinguish between true and false information and/or increasing the veracity of news shared on social media. The current megastudy with 33,233 US-based participants tests nine prominent misinformation interventions in an identical setting using true, false, and misleading health and political news headlines. We find that a wide variety of interventions can improve discernment between true versus false or misleading information during accuracy and sharing judgments. Reducing misinformation belief and sharing is a goal that is accomplishable through multiple strategies targeting different psychological mechanisms.

Content moderation is a critical aspect of platform governance on social media and of particular relevance to addressing the belief in and spread of misinformation. However, current content moderation practices have been criticized as unjust. This raises an important question – who do Americans want deciding whether online content is harmfully misleading? We conducted a nationally representative conjoint survey experiment (N=3,000) in which U.S. participants evaluated the legitimacy of hypothetical content moderation juries tasked with evaluating whether online content was harmfully misleading. These moderation juries varied on whether they were described as consisting of experts (e.g., domain experts), laypeople (e.g., social media users), or non-juries (e.g., computer algorithm). We also randomized features of jury composition (size, necessary qualifications) and whether juries engaged in discussion during content evaluation. Overall, participants evaluated expert juries as more legitimate than layperson juries or a computer algorithm. However, modifying layperson jury features helped increase legitimacy perceptions – nationally representative or politically balanced composition enhanced legitimacy, as did increased size, individual juror knowledge qualifications, and enabling juror discussion., Maximally legitimate layperson juries were comparably legitimate to expert panels. Republicans perceived experts as less legitimate compared to Democrats, but still more legitimate than baseline layperson juries. Conversely, larger lay juries with news knowledge qualifications who engaged in discussion were perceived as more legitimate across the political spectrum. Our findings shed light on the foundations of procedural legitimacy in content moderation and have implications for the design of online moderation systems.

The spread of misinformation through media and social networks threatens many aspects of society, including public health and the state of democracies. One approach to mitigating the effect of misinformation focuses on individual-level interventions, equipping policymakers and the public with essential tools to curb the spread and influence of falsehoods. Here we introduce a toolbox of individual-level interventions for reducing harm from online misinformation. Comprising an up-to-date account of interventions featured in 81 scientific papers from across the globe, the toolbox provides both a conceptual overview of nine main types of interventions, including their target, scope and examples, and a summary of the empirical evidence supporting the interventions, including the methods and experimental paradigms used to test them. The nine types of interventions covered are accuracy prompts, debunking and rebuttals, friction, inoculation, lateral reading and verification strategies, media-literacy tips, social norms, source-credibility labels, and warning and fact-checking labels.