[throwaway894345]

I don't exactly understand the mechanics, but it seems pretty widely accepted that the polarization of media outlets and the rejection of journalistic ethics are a consequence of the online advertising business model. Before the last decade, a larger share of media business revenue came from sources other than online advertising, and the media was less polarized as a consequence.

As for my speculation about the underlying mechanics, perhaps people who are willing to subscribe to media (rather than consume it for free) are more likely to be interested in the truth (i.e., useful information about the world) rather than having their biases reinforced and their hateful prejudices stroked.

[jjk166]

The online advertising business model simply made it possible for small publishers, which in an earlier era would have been small time tabloids, to reach a wide audience that previously would have required a vast media empire. Back when things had to be printed on paper, it was really hard to get an article in front of millions of people, and throwing lots of shit at the wall to see what sticks was simply impossible. Nowadays, instead of competing with a dozen national papers and a handful of local ones, there were essentially unlimited numbers of competing headlines. In this new land of competition, everyone raced to the bottom. There is virtually no cost in publishing 100 crappy stories to have 1 take off and go viral, indeed that's an optimal strategy.

There's nothing special about advertising revenue. It's incredibly rare that I see an ad on a site that's related in anyway to the article that I'm reading, nonetheless that there is some financial motive to modify the article. I'm no more likely to click on an ad for diet pills on an article saying there was voter fraud then on one saying there wasn't.

If I were required to read every single thing published by an outlet I subscribe to, then I would definitely reserve my money for institutions that published a small number of high quality works. But if I can select which articles I read, it makes sense for an institution to produce as many options as they can so that I'll be more likely to find at least one thing I want to read.

[wutbrodo]

> I don't exactly understand the mechanics, but it seems pretty widely accepted that the polarization of media outlets and the rejection of journalistic ethics are a consequence of the online advertising business model

I don't know that this is true, nor widely accepted. I think it'd be accurate to say that the decline in general _quality_ of journalism is downstream of the dynamics you describe. But there are ways to be click-driven beyond building closed-loop narrative worlds for your readers. The obvious example is sensationalism and disproportionate emphasis on catastrophizing negative news for your readers, and this is entirely possible without being partisan. If anything, committing to a dishonest narrative view of the world cuts off half of your potential pearl-clutching: if the New York Times were breathlessly writing about the dangers of critical race theory the way Fox does, they would get _more_ clicks, not less.

Subscriptions work the opposite way here. Most people don't care about (or aren't capable of?) being informed, and subscriptions to the correct news source serves roughly the same function as choosing a church to hear your Sunday sermon at and make a part of your identity.

As far as I can tell, it fits the evidence much more cleanly to model journalism's narrative/polemical turn as downstream of the broader increase in polarization in society in general. The NYT spins alternative-fact fantasies not because of the economics of ad-based serving, but because of the characteristics of contemporary consumer demand.

[throwaway894345]

> I think it'd be accurate to say that the decline in general _quality_ of journalism is downstream of the dynamics you describe. But there are ways to be click-driven beyond building closed-loop narrative worlds for your readers. The obvious example is sensationalism and disproportionate emphasis on catastrophizing negative news for your readers, and this is entirely possible without being partisan.

No doubt there are other ways, but partisanship seems to be the optimal way at least in today's social media landscape.

> If anything, committing to a dishonest narrative view of the world cuts off half of your potential pearl-clutching: if the New York Times were breathlessly writing about the dangers of critical race theory the way Fox does, they would get _more_ clicks, not less.

That's kind of the point: NYT isn't going to write breathlessly about CRT because that's a conservative boogeyman (to be clear, there are a lot of valid criticisms of CRT or at least the movement born out of CRT, but as it's just entered the conservative lexicon 5 minutes ago they don't understand it well enough to critique it gracefully). NYT have their own topics about which they write breathlessly, consider from whence the 1619 Project came, after all.

The idea that NYT would get more clicks by appealing to both the left and the right assumes that there is a fixed number of clicks, but that's not true. By focusing on right wing partisanship, Fox gets the right to engage far more than they would if the writing were more or less neutral. These additional clicks dwarf the clicks lost because they alienate left-wing and moderate readers. The NYT has roughly the same strategy with respect to left-wing clicks (although I don't think the NYT is exactly the mirror image of Fox). Partisanship grows the market such that either the left-wing partisan or right-wing partisan markets are individually larger than the entire nonpartisan market.

> Subscriptions work the opposite way here. Most people don't care about (or aren't capable of?) being informed, and subscriptions to the correct news source serves roughly the same function as choosing a church to hear your Sunday sermon at and make a part of your identity.

I agree that most people aren't interested in being informed, but I disagree that subscriptions work the opposite way. First of all, there were many decades during which online advertising was nil and during which partisanship was low. IIRC there was some increasing partisanship as a consequence of cable news advertising which is just a less-gripping form of online advertising (if only because online media is truly on-demand and we carry access to it in our pockets 24/7), which is to say this observation supports the thesis that advertising and not subscriptions drive partisanship.

Beyond the actual evidence, it doesn't make sense in theory that subscriptions would drive partisanship. To your point, most people don't care about being informed, but those people generally aren't subscribing in the first place. To the extent that people made their news sources part of their identity (prior to the significant increase in politicization of media), it was to signal their membership in the higher class, not their political identity.

> As far as I can tell, it fits the evidence much more cleanly to model journalism's narrative/polemical turn as downstream of the broader increase in polarization in society in general. The NYT spins alternative-fact fantasies not because of the economics of ad-based serving, but because of the characteristics of contemporary consumer demand.

As previously discussed, if traditional and social media companies' ad-based revenue model were torpedoed tomorrow (say, due to legislation prohibiting ad revenue altogether, or at least online ad revenue) I strongly suspect the market for partisan media would largely evaporate. To say that media companies increasing polarization is just "responding to demand" in the same way that the tobacco industry was just "responding to demand" by making their cigarettes ever more addictive.

[wutbrodo]

> No doubt there are other ways, but partisanship seems to be the optimal way at least in today's social media landscape.

Yes, of course, but the statement I'm responding to was claiming that the economics of ad-based publishing drives agenda-driven "news". I'm saying that this doesn't fit the evidence, as agenda-driven reporting is if anything more consistent with a subscription-based system. The fact that it's agenda-driven instead of broadly sensationalist has roughly nothing to do (at least first-order) with the Internet and everything to do with consumer demand. To put it another way, in a hypothetical parallel universe where ad-funded news was banned, you would still see the turn towards agenda-driven reporting, as it's reflective of much broader trends in journalism and independent of ads-funding.

> The idea that NYT would get more clicks by appealing to both the left and the right assumes that there is a fixed number of clicks, but that's not true. By focusing on right wing partisanship, Fox gets the right to engage far more than they would if the writing were more or less neutral. These additional clicks dwarf the clicks lost because they alienate left-wing and moderate readers. The NYT has roughly the same strategy with respect to left-wing clicks (although I don't think the NYT is exactly the mirror image of Fox). Partisanship grows the market such that either the left-wing partisan or right-wing partisan markets are individually larger than the entire nonpartisan market.

Yes, I agree, but my point is that this is an _exogenous_ factor. Competition from an ads-funded Internet had nothing to do with driving newspapers towards partisanship; an exogenous shift in consumer demand did. I brought this example up not as a description of the world as it is but as a contradiction in your hypothesis. My claim is that consumer demand for partisanship is driving partisan news. Whether it's funded by ads or subscriptions is irrelevant.

> there were many decades during which online advertising was nil and during which partisanship was low

I know it's trite to say that correlation doesn't equal causation, but it's very relevant in this case. The postwar decades of relatively high journalistic standards were a historical anomaly. Yellow journalism existed in the past, well before online advertising, and the strategy that got people to buy newspapers was broad-based sensationalism, not the construction of elaborate filter bubbles.

I get that it's cheap to dismiss correlation without elaborating, but the causal hypothesis (ads -> partisan news) both has data pts contradicting it[1] and a more-plausible alternative mechanism, or at least the sketch of one[2].

> Beyond the actual evidence, it doesn't make sense in theory that subscriptions would drive partisanship. To your point, most people don't care about being informed, but those people generally aren't subscribing in the first place. To the extent that people made their news sources part of their identity (prior to the significant increase in politicization of media), it was to signal their membership in the higher class, not their political identity.

Ha, we differ pretty strongly on this factual question. Though perhaps this is my fault for being a little glib: when I say people aren't interested in being informed, I am very much (perhaps centrally!) including the modern information junkie, the core of the NYT (or Fox, or CNN, or MSNBC, or WaPo) subscriber/viewer base. They would probably claim they like being informed, but my point is that watching infotainment opinion shows and reading whatever the fuck the NYT calls what they do these days is a grotesque mockery of "informing oneself". There are undoubtedly hyper-literate, intelligent news consumers out there that have every subscription and use them all sparingly, but these are a vanishingly small minority; my claim is that anyone who thinks that a single-source news subscription is likely to make you understand the world better instead of understanding it worse is not "informed" by any useful definition.

All of the educated, affluent coastals that make up my entire social group use their NYT subscriptions and John Oliver viewings as nothing more than keeping up with the latest approved narrative[3]. For them, the value of their subscription is the feeling of religious harmony and comfortable, predictable single-mindedness it delivers. Broad-based (non-partisan) sensationalism is much less compatible with subscriptions than it is ads-funded articles.

[1] a broader rise in partisanship that is not plausibly driven entirely by journalism, the fact that the NYT et al has gotten substantially _more_ partisan as they've become increasingly subscription-based, etc.

[2] Ads funding and the decline in journalistic quality were both coincident with the earth-shaking, epochal rise of the Internet. Ads funding doesn't need to be causing this decline in order for them to co-occur, because they both arise from complex changes in the way information is distributed and epistemology is carried out. To use a contrived example, the correlation claim here is like saying "literacy expansion causes urban slums, just witness the correlation across the late 19th-century Western world" and ignoring that industrialization is the shared upstream node in the causal graph.

[3] The pandemic was an incredibly stark example, since for the first time they had significant skin in the game of being informed. No surprises there: everyone did exactly what you'd expect them to do, switching multiple times from "X is a terrible anti-science Trumpist conspiracy theory that people shouldn't even be allowed to think about!!!!" to "X is the science how dare you not believe it" on a _dime_ as soon as their news orgs of choice told them to.