Facebook’s news feed algorithms are designed to keep users engaged on the platform for as long as possible. This is achieved by showing users content that is most likely to provoke a reaction or spark a long thread of comments. The algorithms select which posts appear at the top of your news feed based on complex formulas that take into account factors like how recently the post was shared, who shared it, and how much engagement it’s already received. This system skews towards eye-catching, emotional or controversial content while burying posts from close connections if they don’t get enough likes and comments.
How does the Facebook news feed work?
When you log into Facebook, you’re presented with a constantly updating list of stories, photos, videos, ads, and more – this is your news feed. It’s tailored specifically for you based on your connections and interests. Facebook’s algorithms select posts to show you from the thousands of potential stories in your network. They aim to sort through every post from your friends, family, groups, and Pages you follow to curate a feed that keeps you interested. The goal is to maximize your time spent on Facebook and interactions with ads.
So how does Facebook decide what you see in your news feed? The algorithm looks at thousands of factors, but some of the main ones are:
- How recently the post was shared
- The type of post (photo, video, status update, etc)
- Who shared the post (close friend or acquaintance)
- How much engagement the post has received
- How relevant the post is to your interests
- If you have interacted with the person posting in the past
- Paid promotions
Facebook’s news feed formula is constantly evolving and changing based on new data and tests. Their overall goal is to keep you engaged for longer to view more ads by showing you posts you’re most likely to interact with. This leads to selective, personalized feeds for every user.
How do Facebook’s algorithms manipulate users?
While Facebook’s news feed algorithms are designed to merely improve relevancy, many have accused them of having more insidious effects. There are a few key ways Facebook is thought to manipulate users:
Promoting controversial & emotionally-charged content
Facebook’s algorithms favor posts with high engagement and comments. This incentivizes people and pages to create controversial or emotional content to grab attention and go viral. The danger is this can promote the spread of misinformation, especially if it provokes strong negative reactions.
Filter bubbles & echo chambers
Since you’re more likely to see posts that you already agree with and interact with, this can reinforce your existing beliefs and biases. It limits your exposure to opposing views and creates an “echo chamber” effect.
Influencing moods & self-esteem
Facebook studies have shown that when people see positive posts from their social circles, it can lift their mood. But frequently seeing friends enjoying life can conversely make some users feel isolated or depressed. There are similar effects around social comparison and self-esteem.
Addictive nature
By purposely showing you content Facebook knows you’ll engage with, some believe its algorithms promote addictive usage, especially among vulnerable demographics like teens. The fear of missing out and dopamine hits from notifications encourage constant checking.
Weakening real world relationships
Since Facebook determines who and what you see most based on algorithms rather than your true closest connections, it can distort your real relationships. People interact less with their strongest ties due to seeing their content less frequently.
Undermining democracy & journalism
There are growing concerns about the impact personalized news feeds have on shared public discourse and access to factual news. This “splintering” of the social web makes it harder to build societal consensus.
Examples of manipulation
Here are some real-world examples of how Facebook’s algorithms can manipulate users:
News feed experiments
In 2012, Facebook conducted secret psychological experiments by manipulating which posts different groups of users saw in their news feeds. One altered feeds to show more positive or negative content to affect the user’s own posting emotions. This sparked outrage once revealed.
Facebook “dark posts”
“Dark posts” are targeted Facebook ads only visible to the chosen demographic. This has enabled shady political ads and propaganda sent selectively to certain users to influence voting and public opinion.
Blue feed, Red feed
The Wall Street Journal created different news feeds for self-declared liberals and conservatives. Each group saw vastly different information and news stories fueling their biases.
News feed addiction
Ex-Facebook executive Chamath Palihapitiya has stated: “The short-term, dopamine-driven feedback loops that we have created are destroying how society works.” Many ex-employees feel they helped build an addictive, psychologically manipulative system.
FOMO & social comparison
Seeing friends constantly enjoying life can make some users feel like they’re missing out or inadequate. But in reality, people tend to post only their highlights. Facebook fuels these social comparison tendencies.
Are the algorithms really biased?
Facebook denies actively skewing feeds to manipulate in an overtly political or unethical way. They stress neutrality and just recommending relevant content. However, most algorithms have inherent biases based on their purpose and input data. In Facebook’s case, the core bias is toward engagement and profits. This does produce unintended societal consequences in favor of emotions and misinformation.
The view from Facebook
Facebook argues its algorithms aim to improve enjoyment by showing users what they care about most among thousands of posts. They say the content comes directly from the user’s network, not mandated by Facebook. They try to balance relevance with diversity of views displayed.
Neutrality is impossible
Critics counter that neutral algorithms do not exist. Even simple choices like optimizing for engagement produce unavoidable biases that reshape reality for users. Allowing provocative content to spread virally has consequences even if not directly intended.
All engagement isn’t equal
Maximizing overall engagement as the key signal for relevance means not all types of engagement are valued equally. Thoughtful discussion gets lost in favor of likes, comments and shares. Facebook is biased toward quantity over quality.
Ultimately, Facebook’s algorithms are designed to benefit its business model even if unintentionally harming society. Their engagement-driven results cannot be entirely neutral.
Can users avoid algorithmic manipulation?
Given how core algorithms are to Facebook’s services, it’s impossible to avoid their influence completely as a user. But there are some steps you can take to minimize algorithmic manipulation:
- Turn off notifications and check Facebook less frequently
- Unfollow pages and groups that post lots of negative or one-sided content
- Consciously connect more with people who have different views from you
- See alternate viewpoints directly on news sites rather than just in your feeds
- Install tools or browser extensions that demote algorithmic recommendations
- Diversify your news sources and social channels
- Limit Facebook usage and take breaks
While you have some control, the only way to fundamentally change how users are manipulated is by pressuring Facebook to alter their engagement-based algorithms. But its business model depends on showing you whatever keeps you locked in.
Conclusion
Facebook’s algorithmic news feed curation has revolutionized how people discover, consume and interact with content online. But optimizing for engagement has societal downsides in how it can polarize users, spread misinformation, and addictively manipulate moods. While Facebook emphasizes neutrality, critics argue that algorithms inherently reflect and promote certain biases. Users can take steps to mitigate algorithmic manipulation risks, but ultimately Facebook would require significant reforms to prevent unintentional harms resulting from its engagement-driven algorithms.