Journalism is undergoing a crisis in the internet age. Institutional trust is low, readers are siloed in echo chambers, and misinformation spreads like wildfire. Personality-driven commentators dominate the media. Facts are less important than authenticity, as audiences place trust in vibes over references. A public goods problem is at play; original journalism produces value to society, but influencers front-run them to capture the profits. These trends are especially pronounced in the younger generation who increasingly receive information through social media rather than news publications. But this is old news.
Above: Journalism
Even older news is The Market For Lemons (Akerlof 1970), which explains market failure in used cars. “Lemons” is a slang term for crappy cars which plague the resale market. Since buyers cannot distinguish lemons from peaches (good used cars), the market price for used cars averages them together. Sellers with peaches are unable to price them appropriately high and opt to exit the market. This adverse selection increases the proportion of lemons among used cars, which in turn further drives down their average price. These effects compound in a vicious cycle to unravel the market.
I suspect something similar is occurring in journalism. Let’s look at the criteria:
Journalists are incentivized to pass off low-quality information as high-quality
Audiences are unable to accurately assess the quality of information
Journalists are unable to credibly signal quality to audiences
Points (1) and (2) seem true to me, though point (3) is iffier. Let’s dive in.
Journalists are incentivized to pass off low-quality information as high-quality
Low-quality news is cheaper to produce. It’s hard to do your research, check sources, and think through tough issues. It’s much easier to skip that all and just state your claim with charisma and confidence. Moreover, sensational news is more profitable as it baits more clicks. Of course, this all fails if readers don’t trust that your news is factual. It behooves you to convince readers that your journalism is high-quality.
While this effect has always been at play, the internet age magnified it. In the past, spreading news required infrastructure. You need access to printing presses which are owned by publications who want to protect their reputation. You need to pay the cost of printing and distribution, so you only want to push your best work. Today, it is practically free to create and distribute information, so self-publication (like this blog) exploded. The cost barrier is lowered, so lower-quality stuff makes it over the threshold.
Audiences are unable to accurately assess the quality of information
Distrust in institutions is rising; politics, science, and economics seem corrupt or rigged against the everyman, and journalism has not escaped this scrutiny. In our modern social media landscape, authenticity is the prized attribute of successful influencers. Notably, truthfulness seems to take a backseat; the speaker need only believe in their claims, not necessarily have validated them. There’s a whole ecosystem of news regurgitators (mostly white men with strong jawlines) who just consume “real” journalism and offer their reactions. If they make any flagrantly erroneous claims, they tend to sweep it under the rug and move onto the next shiny thing. Their audiences don’t seem to mind.
As evidence, conspiracy theories are bigger than ever. Vaccines causing autism, 5G mind control, alien encounter cover-ups, flat earth, etc. I encounter normal-seeming people who believe their wifi is causing cancer and that placing crystals around it will protect them. Clearly these audiences are unable to assess the veracity of the information they’re receiving.
Journalists are unable to credibly signal quality to audiences
Historically, journalists signaled quality through their publication. Established institutions like the Associated Press or the Financial Times would use their channel as a stamp of veracity. In contrast, lower-quality publications like Weekly World News or local TV stations were infamous for their outlandish claims.
This seems to have degraded. Old publications seem to be leaning into clickbait headlines in order to compete with new upstarts. The unlimited space of the internet (in contrast to the limited space on a newspaper) allows institutions to sell their credibility through editorials. I’ve seen plenty of crypto charlatans brag about their Forbes features, glossing over the fact that it was a paid editorial. This is similar to how ivy league universities create lower-quality international campuses and charge exorbitant rates to allow Saudi princes and Chinese oligarchs’ children to purchase prestigious degrees.
With that said, the fact that people are paying for Forbes placements suggests that there still is some quality signaling at play. Credible journalists do seem to be able to charge for their high-quality services. A subscription to the Financial Times costs $75/mo, so buyers must believe it’s better than free slop. Credibility-signaling services like Ground News have cropped up. This is analogous to used car resellers like Cars.com, who offer their stamp of approval that a used car is not a lemon.
Unravelling
If journalism is undergoing a market for lemons style unravelling, what would we expect to see?
The proportion of low-quality news should be growing. Credible journalists should be unable to profit off their skills and leave, while cheap fraudsters should proliferate.
Consumption of high-quality news should decline as audiences are unable to distinguish it from low-quality slop.
Publications should be attempting and failing to offer stronger credibility signals
Unfortunately, I am a low-quality news source and am not willing to put in the effort to check if these phenomena are actually occurring. That is left as an exercise to the reader.
This was one of my all time favorites in grad school! I had to do a replication study for my Micro II class. You also get to use the Displaced Workers dataset (I think it's a part of the CPS) which is an insanely fascinating data set. Tons of information in there but relatively low N so tough to get accurate results.