This piece was co-authored by Nason Maani, and his version is cross-posted here
“You are what you eat” is clearly an oversimplification, but there is a truth to it. We are heavily influenced by what we consume in a physical sense. The same is true of our consumption of information, of ideas, particularly when our attention is in short supply, and competed over in what has been termed the “attention economy”.
It feels there is more information around us than there ever has been, and as a result, one might think, we consume a varied information diet. Perhaps in the morning we scroll through one or two news websites, then perhaps our notifications on a social media platform like Linkedin, X, Facebook, or BlueSky. Maybe we receive a couple of newsletters from Substack, and some “forwarded many times” messages from groups on WhatsApp. Then there are some alerts on our phone from Google or Apple News. After 30-45 minutes, over our morning coffee, it can feel as if we have sampled quite a range of news, work updates, life updates, sports, science, views and ideas from people in the wider world. But what is shaping that information? What biases what we see and read?
The definition of bias in research terms is a systematic error that distorts measurement away from truth. Just like there are innumerable forces that bias accurate measurement, there are also innumerable forces that bias what we think. We do not think it is possible to remove all sources of bias that shape the information we consume. But if we could understand what is biasing the information we consume, we can come to conclusions about the world around us that are clearer and close to truth.
By way of analogy, let us imagine we are working as a security guard for a warehouse at night. We sit at a desk with a bank of CCTV monitors and check them to keep track of what is going on. If some of them are actually a video on a loop without realising, others are too low resolution to pick up on individual people’s movements, and some are pointed at useless areas like the ceiling, well, all these are biasing against our ability to detect intruders. Even the presence of the bank of screens itself is biasing. We might have a false sense of security because we can see 20 screens in front of us that we can diligently check between. However, because we are focused on that task, we do not realise what we cannot see, or what is wrong with what we can see. Sometimes there is really no substitute for walking around the warehouse ourselves, for seeing things up close with our own eyes, and not through the set of lenses that have been set in front of us (remembering of course that even our eyes can trick us sometimes).
So, what are some of the biases in what we read this morning over our coffee? We organize our thoughts on these biases in three areas for convenience: Who is writing it, who is putting it in front of us, and why we are seeing it.
First, who is writing what. Let us consider the news and commentary side. Who is writing what we are reading? At the most basic level, who is writing what is shaped by the make-up of society itself, and as a result is not a representative sub-section of the population but reflects the social life of ideas. For example, in the UK, the Sutton Trust in their 2025 report found that while only 7% of the population is privately educated, they hold dramatically outsized sway in terms of news and media. 50% of newspaper columnists are privately educated, as are 47% of political commentators, 38% of BBC executives and 45% of podcasters. This can influence what stories feature, how, and the extent to which newsmakers are likely to directly experience the consequences of such coverage.
Further, most news media organizations are competing for our attention, and rely on advertising or subscription to grow and continue operating. To do that, they need to comply with, and optimise their content, for what will appeal to new consumers, what will keep people coming back to the site, and what will align with how ideas are discovered on social media. That is the case even when what is most eye-grabbing or engaging is not what is most important, most factual or most balanced. At core, news outlets are rewarded on the basis of subscription and advertising revenue, not on the basis of how many truths they expose. Of course there are exceptions, but that does not change an underlying bias towards certain types of stories and away from others. Slow-burn, positive stories are far less profitable or engaging than high-profile, negative ones.
So there is inherent bias in news and commentary, but what about bias through who is writing and sharing in our social networks? While it might look like a large number of people are constantly contributing to our social media feeds, in reality most people are not present on most platforms, and of those who are, a small minority of people are providing a majority of the content and are followed by most people. Who is in each of these networks themselves represent skewed samples of the wider population in terms of age, geography, interests, income, education and language. All these dynamics affect what we see and when we see it, and conversely, what we don’t see, who we do not hear from.
This is even more complicated when we think about the technology that is serving us up that information. Often it is social media platforms, each with their own biasing features. For example, who turns up on our Substack feed is shaped by who is accepting paid subscriptions and how many. Many platforms are built on keeping us in the platform in order to view ads, and to do that, the platforms actively disincentivize any content that contains an external link (because that could lead us away from them). So, for example, articles linked to news outlets or academic papers on Linkedin will appear less frequently than posts that do not offer such hyperlinked “ways out” of Linkedin. Conversely, people willing and able to pay for the “premium” service on such platforms will feature more often on our timelines. This dependency on seeking and maintaining attention shapes the algorithms that serve you up content. In some cases, this can be highly damaging.
Another way the attention economy, the commercial bias towards attracting our engagement, affects what we can see is through algorithms that base what we see on what we have interacted with in the past. This makes financial sense in that a given platform is trying to retain our interest by showing us things we have shown an interest in before. So, if we view carpentry videos, we are offered more carpentry videos, and ads for carpentry tools. However, the same is true of opinions or ideas we have engaged with in the past. This is clearly a key dynamic driving polarization—the digital content we consume dictates that we are exposed to more of the same, creating what are known as “filter bubbles”, in which what we see is filtered based on our past activity, not what is happening in the present. Let us take a relatively current example of a protest occurring in a city exposed to an oppressive police force. People who viewed police shooting videos in the past will be served up more police shooting videos. People who previously viewed videos of protesters looting shops and commentary to that effect will be served up more looting videos. As a result, perceptions of reality will diverge based on past behaviours rather than reality, as people remain trapped in their evolving, and in some cases gradually more extreme, filter bubbles. It is worth remembering that we are all in such filter bubbles, and as with advertising more broadly, it is something we wrongly assume powerfully affects others, but never ourselves. In reality, it is very hard to be immune to our digital environments, just as it is to be immune to our physical environment.
So let us go back to why you are seeing what you are seeing. We are seeing it in part because of who we are (which brings its own biases), but also because of who is writing it, on what platform it is presented, and what the incentives of that platform are. In economic terms, what we are navigating is more than bias, but an attention market in which incentives reward certain voices, certain emotions, and certain kinds of content, while filtering out others in invisible ways. In such a system, the most visible ideas are not necessarily the most accurate or the most morally praiseworthy, but the ones best adapted to capture and hold attention. This matters for ideas. Attention shapes ideas, ideas lead to action, and action shapes the conditions in which we live, and therefore, our health and wellbeing.
None of this means retreating from information or disengaging from the world. But it does suggest the need for a different posture in which we recognise that what reaches us has already passed through multiple economic filters. It has its biases. We have previously written about idea factories, how ideas come to be in the first place as products of processes and institutions. Well, the attention economy affects how those ideas, those products are advertised and presented to us, the marketing of ideas. If attention is a determinant of what we come to know, then it is also a determinant of the societies we build. Learning to see the economics of attention seems therefore a critical part of understanding the life of ideas, the groups that discuss them, and what shapes the world of the possible.
__ __ __
The Ideas about Ideas project
The Ideas about Ideas project, co-written with Dr Nason Maani, and cross-posted on the Money, Power, Health substack is a series of reflections of how thought lives in the world. It follows ideas as they emerge in institutions and everyday life, as they move across borders, as they harden into orthodoxy or fade into silence. It argues that ideas are not possessions but shared conditions of living, and that to see them clearly is to recognize both their power and the obligations they impose. Fundamentally, this project is built on the notion that without ideas we have nothing, and that the business of creating a healthier, better world rests on ideas.
__ __ __
Also this week
On March 31, WashU Public Health will host Georges C. Benjamin, executive director of the American Public Health Association; Lori Tremmel Freeman, CEO of the National Association of County and City Health Officials; Scott Harris, immediate past president of the Association of State and Territorial Health Officials; and Laura Magaña, president and CEO of the Association of Schools & Programs of Public Health, for a conversation about how public health can find its way forward in challenging times. Learn more and register for the event here.