There’s a study that circulates every few months pointing out that most people who share articles on social media haven’t actually read them. They see a headline, maybe a preview snippet, and hit share. Then it gets reshared by people who also didn’t read it, creating chains of engagement around content nobody’s consuming.

The study is always shared by people who find this appalling. Often by people who didn’t read the study. The irony is delicious.

But here’s the thing: this behavior makes perfect sense once you understand what social media sharing actually is. It’s not about distributing information you’ve carefully evaluated. It’s social signaling. And once you see it that way, the whole pattern becomes obvious.

Sharing Is Performing

When you share an article on social media, you’re not really recommending reading material. You’re performing a version of yourself for your audience. The headline and source tell your followers something about your values, interests, and what team you’re on.

Sharing a New York Times article about climate change signals you care about the environment and trust mainstream media. Sharing a contrarian Substack piece signals you’re an independent thinker who questions the consensus. Sharing anything from Fox News or CNN immediately sorts you into a political category.

The content of the article barely matters for this purpose. The headline and source do all the signaling work. Actually reading the piece would just slow you down and might complicate your feelings about whether it supports the point you’re trying to make.

This isn’t unique to news. People share workout tips they’ll never follow, recipes they’ll never cook, life advice they’ll never take. The act of sharing creates the appearance of engagement with ideas without requiring actual engagement. It’s performance all the way down.

Headlines Are Optimized for Sharing, Not Accuracy

Publishers know this, obviously. That’s why headlines have become increasingly detached from article content. The headline’s job isn’t to summarize the story anymore. It’s to generate shares from people who won’t read past it.

This creates a weird dynamic where the headline can be technically accurate but give a completely misleading impression of what the article actually says. The article might be nuanced, but the headline is black and white. The content might be speculative, but the headline is definitive.

Nobody cares because nobody’s reading the article. The headline gets shared, people argue in the comments about what they assume the article says, and the cycle continues. The publisher gets traffic from the minority who click through, engagement from everyone arguing, and ad revenue from both.

It’s a system perfectly designed to maximize engagement while minimizing actual information transfer. Working as intended.

The Emotional Response Loop

People share articles that make them feel something: validated, outraged, superior, vindicated. The faster you can trigger that emotional response, the more shares you get. Actually reading an article takes time and might complicate your emotional response with nuance or context.

Better to just read the headline, feel the feeling, and share. Your followers who agree will like it. Your followers who disagree will argue with you. Both groups will boost the post through engagement. Everyone wins except anyone who cares about accuracy or understanding.

Social media platforms optimize for this because engagement is their metric. They don’t care if you read the article. They care if you interact with the post. Comments, shares, reactions—these signal to the algorithm that the content is valuable. Whether anyone actually consumed the content is irrelevant.

Identity Over Information

Here’s the uncomfortable truth: most people aren’t on social media to learn things or exchange information. They’re there to reinforce their identity and connect with their tribe. Sharing articles is part of that identity performance, not information sharing.

This is why you see the same articles shared repeatedly within ideological bubbles. They’re tribal markers, not news. “This article confirms our worldview” gets shared endlessly within a community even though everyone’s already seen it and nobody’s reading it again. The sharing itself is the point.

Fighting this by yelling “you didn’t even read it!” misses the mechanism. They weren’t trying to read it. You’re complaining that people are using a hammer wrong when they’re trying to turn a screw. Sure, it doesn’t work well, but they’re solving a different problem than you think.

What Publishers Learned

Smart publishers figured this out years ago. They optimized for shareability over readability. Click-through rates matter, but social shares matter more for reach. An article that gets 10,000 shares but only 2,000 reads is more valuable than one that gets 500 shares and 2,000 reads.

This incentivizes exactly the wrong things from a journalism perspective. Write provocative headlines disconnected from nuanced content. Front-load emotional triggers. Don’t worry too much about accuracy past the headline. The economics reward viral performance over informational value.

Some publications resist this. They write accurate headlines and nuanced content and accept lower engagement numbers. God bless them, but they’re fighting economic reality. The market rewards the behavior we all claim to hate.

Can This Be Fixed?

Probably not through individual behavior change. Everyone who shares articles without reading them knows they’re doing it and why. The people doing it don’t think it’s a problem because they understand they’re performing, not informing. The people who think it’s a problem aren’t the ones doing it.

Platform changes could help. What if shares required you to demonstrate you’d spent time with the content? What if misleading headlines were flagged? What if algorithmic reach prioritized completion rates over shares?

These solutions are technically possible but commercially terrible. Platforms make money from engagement, not information quality. Reducing shares would tank their metrics. They won’t voluntarily kneecap their business model to improve media literacy.

The Real Problem

The deeper issue is that we’re using social media for something it’s not designed to do. We expect it to be a venue for informed public discourse when it’s actually a performance space optimized for emotional engagement. Those are fundamentally different things.

Maybe we need to accept that social media is terrible for distributing journalism and stop trying to make it work. Save the actual reading for newsletters, websites, and platforms designed for consuming content rather than performing identity.

That would require journalists and publications to give up on social media as a primary distribution channel. Given how much traffic it drives (clicks from people who didn’t read the share, but clicks nonetheless), that’s not happening.

So we’re stuck. People will keep sharing articles they haven’t read. Publishers will keep optimizing for shares over reads. The informed will keep complaining while participating in the same system. And the platforms will keep printing money from the whole mess.

At least we’re all clear on what’s actually happening here. That’s something, right?