Causally estimating the effect of YouTube’s recommender system using counterfactual bots
Open accessThe DOI matches its Crossref record; 8 of the 8 references it cites were found in a registry or an index.
Abstract
In recent years, critics of online platforms have raised concerns about the ability of recommendation algorithms to amplify problematic content, with potentially radicalizing consequences. However, attempts to evaluate the effect of recommenders have suffered from a lack of appropriate counterfactuals—what a user would have viewed in the absence of algorithmic recommendations—and hence cannot disentangle the effects of the algorithm from a user’s intentions. Here we propose a method that we call “counterfactual bots” to causally estimate the role of algorithmic recommendations on the consumption of highly partisan content on YouTube. By comparing bots that replicate real users’ consumption patterns with “counterfactual” bots that follow rule-based trajectories, we show that, on average, relying exclusively on the YouTube recommender results in less partisan consumption, where the effect is most pronounced for heavy partisan consumers. Following a similar method, we also show that if partisan consumers switch to moderate content, YouTube’s sidebar recommender “forgets” their partisan preference within roughly 30 videos regardless of their prior history, while homepage recommendations shift more gradually toward moderate content. Overall, our findings indicate that, at least since the algorithm changes that YouTube implemented in 2019, individual consumption patterns mostly reflect individual preferences, where algorithmic recommendations play, if anything, a moderating role.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Deep Neural Networks for YouTube Recommendations
Covington, P., Adams, J., & Sargin, E. (2016) · Proceedings of the 10th ACM Conference on Recommender Systems · doi:10.1145/2959100.2959190 · Cited by 3,539
Auditing radicalization pathways on YouTube
Ribeiro, M. H., Ottoni, R., West, R., Almeida, V. A. F., & Meira, W. (2020) · Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency · doi:10.1145/3351095.3372879 · Cited by 580
Self-citationHow do social media feed algorithms affect attitudes and behavior in an election campaign?
Guess, A. M., Malhotra, N., Pan, J., Barberá, P., Allcott, H., Brown, T., Crespo-Tenorio, A., Dimmery, D., Freelon, D., Gentzkow, M., et al. (2023) · Science · doi:10.1126/science.abp9364 · Cited by 363
Asymmetric ideological segregation in exposure to political news on Facebook
González-Bailón, S., Lazer, D., Barberá, P., Zhang, M., Allcott, H., Brown, T., Crespo-Tenorio, A., Freelon, D., Gentzkow, M., Guess, A. M., et al. (2023) · Science · doi:10.1126/science.ade7138 · Cited by 315
Right-Wing YouTube: A Supply and Demand Perspective
Munger, K., & Phillips, J. (2022) · The International Journal of Press/Politics · doi:10.1177/1940161220964767 · Cited by 265
Examining the consumption of radical content on YouTube
Hosseinmardi, H., Ghasemian, A., Clauset, A., Mobius, M., Rothschild, D. M., & Watts, D. J. (2021) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2101967118 · Cited by 171
Self-citationUsers choose to engage with more partisan news than they are exposed to on Google Search
Robertson, R. E., Green, J., Ruck, D. J., Ognyanova, K., Wilson, C., & Lazer, D. (2023) · Nature · doi:10.1038/s41586-023-06078-5 · Cited by 126
Haroon, M., Wojcieszak, M., Chhabra, A., Liu, X., Mohapatra, P., & Shafiq, Z. (2023) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2213020120 · Cited by 64
Cited by
Misunderstanding the harms of online misinformation
Budak, C., Nyhan, B., Rothschild, D. M., Thorson, E., & Watts, D. J. (2024) · Nature · doi:10.1038/s41586-024-07417-w · Cited by 156
Self-citationMilli, S., Carroll, M., Wang, Y., Pandey, S., & Dragan, A. D. (2023) · arXiv.org · doi:10.48550/arxiv.2305.16941 · Cited by 200
Ye, J., Luceri, L., & Ferrara, E. (2025) · Proceedings of the 2025 ACM Conference on Fairness, Accountability, and Transparency · doi:10.1145/3715275.3732159 · Cited by 74
“While some studies, particularly on YouTube, allow bots to interact with algorithms (e.g., following recommendations to study radicalization [18, 19]), we refrain from inducing interactions in the current study for several reasons.”
Introducing v0.5 of the AI Safety Benchmark from MLCommons
Vidgen, B., Agrawal, A., Ahmed, A. M., Akinwande, V., Al-Nuaimi, N., Alfaraj, N., Alhajjar, E., Aroyo, L., Bavalatti, T., Bartolo, M., et al. (2024) · arXiv · doi:10.48550/arxiv.2404.12241 · Cited by 5
Metadata
- Authors
- Hosseinmardi, H., Ghasemian, A., Rivera-Lanas, M., Horta Ribeiro, M., West, R., & Watts, D. J.
- Venue
- Proceedings of the National Academy of Sciences
- Published
- 2024-02-13
- Volume
- 121
- Issue
- 8
- Pages
- e2313377121-e2313377121
- ISSN
- 0027-8424 (ISSN-L)
- 1091-6490
- OpenAlex
- W4391773993
- Cited by
- 49
- References
- 8
- FWCI
- 35.97
- Type
- journal article
- Language
- en
- Open access
- hybrid
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Causally estimating the effect of YouTube’s recommender system using counterfactual bots. https://app.cochat.ai/papers/10.1073/pnas.2313377121
- The paper
- Hosseinmardi, H., Ghasemian, A., Rivera-Lanas, M., Horta Ribeiro, M., West, R., & Watts, D. J. (2024). Causally estimating the effect of YouTube’s recommender system using counterfactual bots. Proceedings of the National Academy of Sciences. https://doi.org/10.1073/pnas.2313377121
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.