The Platforms & Publishers Relationship, 2018

Tow Center
Jul 13, 2018 · 41 min read
Campbell Brown and Mark Thompson at Columbia Journalism School on June 14, 2018

On June 14, 2018, the Tow Center for Digital Journalism at Columbia Journalism School launched its report entitled “Friend and Foe: The Platform Press at the Heart of Journalism,” as part of the Center’s ongoing, multi-year study into the relationship between large-scale technology companies and journalism.

Following the report launch, Tow Center Director Emily Bell moderated a panel discussion at the New York launch of Reuters Institute’s Digital News Report 2018, which is transcribed below. To get updates on similar events at Columbia Journalism School, you can sign up for Tow Center’s newsletter.

Note: The panel discussion (video) has been edited for clarity and concision.

Rasmus Kleis Nielsen, Erica Anderson, Lydia Polgreen, Campbell Brown, and Mark Thompson at Columbia Journalism School on June 14, 2018

Emily Bell:

To introduce a panel that needs no introduction: To the far right, we have Rasmus Kleis Nielsen, director of research at the Reuters Institute for the Study of Journalism, whom you’ve already met. Then we have Erica Anderson, who is global head of everything and US partnerships lead at Google. Next to her is Lydia Polgreen, who is global head of everything at HuffPost. Then, Campbell Brown, who is head of news partnerships at Facebook — so basically head of everything at Facebook as well — and Mark Thompson, chief executive officer at The New York Times.

Welcome, panel. Let’s just start with some of Rasmus’s provocations, specifically that really broad question of “what have we learned?” I sometimes wonder, what is it that platforms think they’re trying to achieve when interfacing with news organizations? And for news organizations — we’re constantly talking about sustainability — is that still what we’re trying to achieve? Mark, let’s start with you.

Mark Thompson:

Thank you. Look, economic sustainability and profitability are a necessary but not sufficient concern when trying to build effective digital revenue streams for high-quality journalism. That task is far from over. It’s been going badly for The New York Times. I don’t think we’re out of the woods on that. That’s important. What’s so interesting about the way it’s going is that it has taken us into the realm of civics: the role of journalism in society, the roles and responsibilities, or lack thereof, of major platforms in supporting plurality and debate. Also, what kind of public do we want? Basically, no one says they want only trusting audiences. We want critical audiences, audiences who are testing ideas, often skeptical about institutions. In a way, I think we need to engage with this theme and not endlessly with the idea of economics.

EB:

Alright, yes, we ought to pivot to civics. Campbell, let me ask, what does Facebook think it’s doing?

Campbell Brown:

Well, there’s going to be change. Very frankly, there was a huge change this year in terms of how we do ranking. Referencing the slide that talked about whether Facebook is moving away from news — with less news on the platform — I don’t think that way at all. We had a very broad definition of news before, and it included clickbait, sensationalism, and a lot of things people don’t want to see on the platform. So, what we’ve done is condense the definition of news and turn down the volume pretty massively on clickbait, sensationalism, and stories that had misleading headlines. We made a ton of ranking changes to take down that type of less-quality news and try to turn up the volume on how we define quality news with a whole number of different levers — from surveys to the community about who they trust and who they don’t, to signals that we developed with outside academics and publishers to try and define quality so that we can give that a bigger boost. Overall, it’s a change in the ratio of news on platforms.

EB:

I’m going to put a pin in that, because I know Mark has some strong views about that. In an earlier meeting, he had some very specific things to say, but we’ll get down the line before we get that aggressive.

Lydia Polgreen:

That’ll be fun. (Laughter.)

EB:

Lydia, you’re one of these digitally native-born creatures at the Huffington Post, but you have also shifted your strategies since you started. How is it connected? How does that play into what we’re seeing?

LP:

Lydia Polgreen and Campbell Brown at Columbia Journalism School on June 14, 2018

I think there are lots of fascinating directions to go with that question. There was a lot in this report about the economics, and Mark is right that we need to focus on the civics, but these things are actually deeply connected. One of the things I read in the report was that 26 percent of people in the United States and 18 percent of people in the UK would consider paying for news in the future. That’s encouraging; it’s also a tiny slice of the population, and as we think about business models — as we think about distribution — I am really concerned that there is a strong focus on the subscription model as being the panacea and savior of the news business. If we are living in a world where 74 percent of Americans are not willing to consider paying for news in the future, who’s going to inform those people? How are they going to become informed citizens? How are we going to have a solid and stable society if we’re moving towards a world in which our main hope for the future of journalism is people paying for subscriptions?

Now, we’re very fortunate in that we have great news organizations like The New York Times that make the vast majority of their journalism available for anyone to read. Just a tiny percentage of the Times’s audience actually becomes subscribers, and it’s a fantastic thing that they are able to execute a highly successful pay model on that. But I do think a lot about the audience that’s never going to pay for the news and how we serve them. And that’s the category that I think is seeing the greatest impact from the platforms. We’re also seeing a tremendous amount of consolidation there, so there are fewer voices. You are less able to sustain your voice as an independent news organization, and it’s interesting because there is a very sharp partisan divide. You have ongoing concerns on the far right that are supported by patrons, and then on the left you’ve got a handful of organizations supported by patrons and donations. But in the center left, and less so on the center right, there is a tremendous battle going on over how organizations like HuffPost, BuzzFeed, and Vox — who I think are doing good, vital journalism — are going to survive in this climate providing free consumer journalism that is really important to our society.

EB:

Erica, you get a say. Google has recently said it is moving money to market, $300 million dollars I think it is, to the Digital News Initiative. Why are you doing that, and why now?

Erica Anderson:

It’s great to be here. Thank you for hosting this. I would preface this by saying that in March we looked around at the company — and we’re doing a lot, we are doing a lot. Having been in the news industry for about 15 years, we have a digital news initiative focused on training, we’re doing a lot of product working groups to develop products for the industry, but we really need to align and have unified communication internally, and also externally. So, Google News Initiative, which we announced in March, is first and foremost about helping journalism thrive in the digital age — and saying that out loud and being accountable to that. It’s about three things. Number one, it’s about elevating and strengthening news. Number two, economics. It’s about building new financial models in addition to the ads model, if you want to subscribe to Google News for free. And number three, it’s about empowering technological innovation.

So, your question, why now? Because there are a lot a people inside the company who care about this. There are a lot of things happening and we think there needs to be more. So, the $300 million — right now we are figuring out what the structure and form look like. We know we’re going to work on product; we’ve seen a lot of product working groups. For Subscribe with Google, I think we worked with 18 working groups, and over 80 news organizations, because ultimately we know that we can’t build products without close collaboration with the industry. We’re figuring out right now what the fund looks like. So we’re working on that, but ultimately having quality content on our platform matters, so people come to Google looking for information they can trust.

EB:

Alright, Rasmus, from your 30,000-foot view, as we hear a lot of talk about elevating quality journalism and helping journalism thrive in the digital age, how do you think about this as an active researcher?

Rasmus Kleis Nielsen:

Rasmus Kleis Nielsen at Columbia Journalism School on June 14, 2018

I think in terms of what content platforms service and how much, there are really only two kinds of criticisms of platform companies: One is that they either do too much or do too little to judge content. That’s one, versus the other: It’s that they either serve too much or too little of other people’s content. So, it’s either that Facebook has too much news on it or too little news on it. It either does too little or too much to rank news. It is similar with Google Search — either there’s too little done to rank content or too much done to rank content, and similarly there’s too much or too little of other people’s content in search responses. I think the most important thing for me as a researcher is to provide evidence for this discussion so that other people can make up their minds from their own values and principles about what they think the right response is to this. It’s not my role as a researcher to tell people how to think. What I will say is that I think the discussion around whether trust is a useful proxy for trustworthy — I’m perhaps more optimistic and positive about that development than some people are. I think the results suggest that people can simultaneously love a content provider while, as Mark suggested, be critical about the veracity and accuracy of the information provided. We have known this for a long time in Europe. Tabloid newspapers are widely read, but not actually necessarily trusted even by their own readers. I don’t think there is anything wrong with reading this kind of stuff. I do it all the time myself. I love entertaining stuff! I love sports and gossip. The question is how you engineer something that allows people to pursue their interests, not only accurately and factually so we can properly cite that information, but make sure that the systems that service content at scale do not avoid judgment on veracity or accuracy and trustworthiness when it comes to truly important things — election results and the like. I think there is a lot of work to be done there. I think the work around trust we’ve seen from the Trust Project and the work around the trust survey is wholly encouraging.

EB:

Mark, I’d like to come back to you, because we just heard these platforms talking about categorization, elevating quality news, etc. You’ve had some fairly sharp things to say, particularly in relation to Facebook’s categorization.

MT:

You know, I’m afraid so. I contrast it with the kind of constructive dialogue we’ve had with Google — we have plenty of issues with Google and there is plenty more we want to talk about — but there’s a lack of substantive progress with Facebook. People know the topic, and the approach that’s been taken to advertising on the Facebook platform for news providers, and the attempt to blur and treat news providers as if they were party advocates. Several journalism industry bodies, representing 20,000 publications, are now involved with this around the world. Campbell wrote a blog post yesterday [saying] that many publications are happy. I’d love to meet one. We have 20,000 unhappy publishers.

I thought I’d give you an example of what is happening. I have a slide. I’ll show you a couple of New York Times ads on the Facebook platform that have been rejected by the Facebook algorithm and placed in a public archive, which is for ads with miscreant content. This is an ad for the New York Times Cooking section, and the dangerous political thing you are looking at there is a pistachio and rosewater pastry. I would say the cake may contain nuts — frankly, Facebook may contain nuts — but it does not, to my understanding, contain political content. When you think about the excellence of algorithms and their ability to help in the matter of editorial decisions — now to be fair, we pointed it out to Facebook, and they took it out of the political archive — but the extent to which the machine can mistake that for politics gives you some idea about the subtlety of what’s going on here. You can go to the next slide, which is more typical. Another rejected ad. Here is an ad featuring a New York Times news story: “Trump pulls out of North Korean summit with Kim Jong-un.” This has been rejected and is now in the archive of ads for political content. We have a fundamental objection to the idea that reporting of politics — and I accept that Trump is a politician and what he does is politics — can be blurred with politics itself is. There is a distinction between capturing and reporting at arm’s length on an activity and being part of the activity. We have a long history of striving to be impartial. And it’s different, fundamentally, as a category than the actual activity itself. Moreover, it is completely clear that the enemies of high-quality journalism want to blur the lines. They want to blur the limits between arm’s-length reporting of politics and politics itself. They say there’s no difference. And I’m afraid, unintentionally, Facebook is supporting the enemies of high-quality journalism.

EB:

Campbell?

CB:

I think there’s a fundamental misunderstanding of what we’re doing on the platform. So, I think I should start by explaining. We, given what happened in the last election, made a decision that we would move toward more transparency in all advertising through Facebook, that we would share an archive saying who was paying for the ads. They would be labeled and say “paid for by” so that information would be publicly searchable. Anybody could go see who was advertising about what. Because of what happened in the last election, because the focus was around politics, around fake news, our starting to address transparency focused on all advertising that contains political content or issues. Now, obviously we recognize there is a distinction between news content about politics and political advocacy advertising. These are very different things. I was journalist for 25 years. I think I know the difference. So, what we did was create a distinction, a differentiation within the archive, where there is an archive that says “news content” and there is a separate archive that says “political advocacy content.” I don’t think there’s any blurring there. The information is very clear and very transparent for anybody who is interested in seeing who is advertising on our platform. This is the first time that I have heard journalists argue against transparency. (Interruption by Mark Thompson.) No, no, no, Mark, I am not finished.

MT:

We are not arguing against transparency.

CB:

You are exactly arguing against transparency. We are asking you to be transparent about all content that you boost on the platform. We are going to be doing this with all of our advertising. We believe this is really important for people to understand, and the idea that somehow we are conflating — with two separate archives that are clearly labeled — advocacy or politics and journalism is completely false. People are smarter than that. What conflates advocacy and journalism is when journalists do advocacy, and that’s happening a lot.

EB:

But that’s not necessarily what we saw there. That’s not a case of advocacy by The New York Times

CB:

I’m not suggesting it was. It was a piece of news content. We’re talking about Trump. It was a boosted piece of news content about politics. We think that should be transparent. We want everything to be transparent.

EB:

So, isn’t one of the problems, in general, a blurring of distinction by all parts of the equation here? I think most people are really shocked to see that something they consider as an article is treated as an ad, because they probably don’t know that The New York Times pays for [them] to see that. We could say we’ve been seeing that blurring with advertisements even within different news organizations. Mark, do you want to add on?

MT:

If I may say so, that’s a different distinction. You can debate that or not debate that, but this is not about the boundary between it — this is about the underlying theme. That is a New York Times article. Yes, we want to put it in front of you because we hope you read more Times articles, but it is a Times news article. I really want to pick away at this. Yesterday, Campbell, in her blog, says, “We recognize the need that news content was different than political advertising.” This isn’t true. It just isn’t true. What happened, in reality, was Facebook came to us and they said, “From now on we’re going to have to label all your content as a political ad paid for by The New York Times.” Everything from The Times was going to be a political ad, and so this idea that there’s never been a distinction, it’s completely blurred in Facebook’s head.

And quite honestly, Campbell, why? You’ve now got 20,000 publishers around the world — and this is going to pile up. You think this is going to help Facebook? This is about the election — it’s about a catalogue of horrendous mistakes made by Facebook. To be fair, Mark Zuckerberg has admitted the scale of what went wrong. Trying to put it right this way is just going to dig you deeper into the hole.

EB:

Campbell?

CB:

Yeah. What I hear is, you’re all for transparency from Facebook, except when it applies to you.

MT:

So every single advert from The New York Times is being labeled as an ad for The New York Times?

CB:

That’s what it’s labeled as. It’s labeled as a piece of content paid for by The New York Times in an archive that says “news content.” And I just wonder whether this is not about The New York Times not wanting to be transparent about the money they spend, how much money they spend, how they do their marketing. I mean…

MT:

To be clear, what we don’t want is Facebook to set itself up as a judge around what political content is.

EB:

Let me stop you right there. Lydia?

LP:

OK. This is not picking on you, Campbell, but I sat in a room where Mark Zuckerberg used phrases like “broadly trusted” and “common ground” in a way that, frankly to me, felt downright Orwellian. I don’t want trust to be a popularity contest decided by users on Facebook. The idea that this organization could somehow decide who’s trusted and who’s not. I mean, we’re living in an environment right now where one segment of the United States population essentially doesn’t trust anyone but Fox News, so if we’re going to take the entire population of Facebook and use that to decide what’s broadly —

CB:

Can I correct? This is again just a misunderstanding of how this works. Our broadly trusted survey of people on the platform is one of thousands of signals that go into determining what shows up in a person’s News Feed. It’s not like, “Oh, well, the Huffington Post is broadly trusted — they’re going to be in your News Feed and The New York Times isn’t broadly trusted, therefore they’re not.” This is a survey that contributes to many, many different signals that go into making a determination of what appears in someone’s News Feed. So, for example, you mentioned earlier the Trust Project indicators. That’s another signal we use to make a determination about what is and isn’t quality. We survey people about what they find informative, what they don’t find informative. That’s another signal. We look at how much time people spend reading a news article. That’s another signal. Literally thousands of signals determine what appears in your News Feed. So, the idea of Mark [Zuckerberg] saying, “You know what, we’re going to start surveying the community and try to add additional signals like, do people find a news organization broadly trusted?” will be one of the thousands we consider. This is a good thing. I think we’re moving closer and closer toward how we do this. The idea that this is a determination about whether a news organization rises or falls in News Feed is just not how News Feed works. It’s not how our algorithm works, and just listening to you both and the misunderstanding that exists here makes me think that we’re just not doing a very good job of explaining ourselves on these issues. We need to be much more aggressive and out there about trying to clear up what are real misunderstandings.

EB:

So, on this basis of who gets to decide or who constructs what comes next in journalism, it seems clear to me that there’s a part of Facebook, I’m sure, that would quite like all of these publishers to get off the platform, maybe?

CB:

No! Not true at all! Look, it’s part of our mission that we want to have an informed community. That means having high-quality news on the platform.

EB:

Do you know what high-quality news is?

CB:

That’s hard. That’s why you want so many signals helping to make that determination.

LP:

Again, I think there’s been a strong focus, understandably, from the top people at Facebook on two kinds of news: local news and investigative reporting. And those two are held up as the paragon of the most important, virtuous news. Those are great forms of journalism. There are lots of other kinds of journalism out there that need support as well, and that people want to read and engage with. And again, I’m not saying you’re only going to force local news and investigative journalism down people’s throat, but we all live rich and complicated lives, and the idea that these two forms of journalism are the ones that sort of pass the smell test and everything else can just, you know —

EB:

Well, there’s also the question of Sinclair Media.

LP:

That’s a great point. Local television news scores at the absolute top of your findings, both in terms of “what’s a news source that you used in the last week?” (proud to say that Huffington Post was number one in digital news sources in the United States) and, in trust — local television news counts as number one. So, to me that is an on-fire question for us: What does the future of local television news look like? I spent many weeks crossing the country last year and frequently stopped and did interviews on the local newscast. Campbell, did you ever work in local news?

CB:

I did. I worked at three local stations.

LP:

So you know how deeply trusted that connection is between the local TV anchor, the local TV reporters, and that community. The type of service journalism they provide. The idea of that being co-opted and perverted by political agendas, having them sandwiched into the report about the parade, the weather, and the traffic, is to me probably the most terrifying thing in the news right now.

EB:

Erica.

EA:

Erica Anderson at Columbia Journalism School on June 14, 2018

One of the announcements we made in March was about authoritativeness. One of the things that was happening was during breaking news situations, with top stories, for example, relevance was what was bringing up content or showcasing content immediately after every breaking news event. And we know that it takes news organizations, I don’t know, five, 10, 15 minutes to vet and verify a story, and it takes conspiracy theorists a bit less time. So, we’ve begun to look at authoritativeness as a signal of quality.

We think that can be very nuanced if you add on top of that what the Trust Project has been doing. So, proximity matters. Who runs the news organization? Do they have an ethics policy? There’s this whole world of Trust Project signals that we think are really important, and we’ve supported them for multiple years to figure this out, outside of Google. Then there’s also this move to understand authority, and when authority is important over relevance. That’s something that will be continually developed. But I’ll say this: Right now part of the signal of authority is what publishers self-select to be a part of Google News. I do think ultimately this is about transparency, and on the platform side the various platforms do need to keep having much more open conversations about how decisions are made and how we view these things, because that will ultimately build trust in the ethics, certainly inside of Google.

EB:

One thing that both of your organizations can do is open up and add much more data to your APIs. If you try to study Google AMP, it’s really difficult. You have to do a lot of manual collection. It feels as though the opportunity to introduce that transparency is there.

EA:

Well, this is what we’ll talk about with the Google News Initiative. We hope to involve a lot of researchers. Literally this is something we should have an open conversation about with you and other media researchers. What are the things that need to be opened up? This is the moment to have those conversations.

EB:

Mark, I want to come back to you because of something you said at the event on Tuesday where you talked about, again, why constantly focusing on trust is not a particularly useful metric in constructing an open market for news. Now, one of the questions I’m going to put to the entire panel is: How is it even possible to think about an open market given how people behave and given the evidence of what they concentrate on? We complain a lot about journalism standards. What are the solutions to this?

MT:

I think if you start with a citizen-centric view of the world, one of the most important things is that you’ve got an economic environment which is sufficiently strong. It makes the practivity of journalism, at a local and national level, possible. But also, there’s the idea that there are multiple, different points of editorial decision making and accountability. So, a citizen, in a sense, can over time choose between those points. I think everything should be formulated, for all the reasons that Rasmus said earlier on, in line with giving citizens the genuine choice. And I’ll say that each news organization itself should not just be profoundly transparent, but completely transparent. All of of our marketing data on Facebook is about to be made available, but more generally our process should be transparent. At The New York Times, there is an editor, you can find out where he lives, you can go see him. You can sue the organization. We’ve just had a four-part documentary all about this for Showtime, the ad for which, by the way, was blocked by the Facebook algorithm.

There needs to be worry about the algorithms, particularly. And Mark Zuckerberg’s remarks, which I know come from the heart and are intended to help, make my blood run cold when he starts talking about how he thinks about community, and about what we trust in. I think he’s got a terrifyingly naive perspective on news. The issue being not just with Facebook, but with Google. In the end, making a bit available to researchers, being there to talk about algorithms, isn’t enough given how important it’s becoming and given the risk, in the end, that these machines and machines led by proxies become more determinant of what people look at and read than human editors. It is vital not just to understand how they work, but that they themselves can be exposed and interrogated, and ultimately are available to politicians and others to view. It’s part of our society; it’s far too central to democracy and politics to be left to proprietary ideas.

EB:

But what about the idea that actually we’re just reconstructing a different type of plurality — Rasmus, I’d like to hear your views on this — but with slightly different gatekeepers? Maybe in the AT&T and the Facebook era, everything is actually moving away from one set of, forgive me, Mark, white men to, forgive me, Rathmus, another set of white men.

MT:

I think if there were going to be fewer than four or five local gatekeepers, we would worry. With 20,000 publishers, we still have a lot of structural plurality. But because of the way things are going, with the number of censors in decision making declining as a lot of legacy media is going or has gone bust, and because of the consolidation and the aggregation of the big players, I think plurality is the biggest single risk we face.

EB:

Right. Rasmus, you see this as a professor of political communication. Is Mark right that plurality is as big a threat as he says? Is this new world order desirable? Is it inevitable?

RN:

First, I think a precedent to establish in this conversation is that while there are many reasons to be concerned, and many concerns that we should seriously discuss as a society, I do not know a single person under 40 who would rather have the media environment that I grew up with in the 1990s than the one we have today. And if they want it, I am happy to take them to the provinces, to the outer reaches of Denmark, and show them what that’s like to live in a small village with access to one regional newspaper and very little else.

We live in a world in which essentially two organizations, Google and Facebook, serve news to more people than any organization has done in the history of mankind, and the way in which they shoulder that responsibility is absolutely decisive. When we look at how they capture that responsibility, the most rigorous, independent, evidence-based research I know of suggests — and this is not only the Oxford Institute but many other people, too — that discovery through social media and search increases the diversity of people’s news consumption. It does not reduce it. The filter bubble question is a really important and urgent one to look at — and one we should keep exploring — but at this stage the evidence suggests that discovery through social media and search drives diverse discovery, and increases political participation and civic engagement. And, importantly, it does so particularly and powerfully for those least interested in news and for young people. In a society in which so many other forces are increasing in potency, I think it is actually critically important to appreciate in this moment the way in which platforms operate.

Now, are there other considerations? Absolutely, yes. Around including their role in the market, in terms of advertising, and including their role in accounting for data. I think what is important as we have this discussion are essentially two things: One is that we try to have a basis of some sort of evidence that is not based on a moral panic advancing the latest utterance about what the world might be like. Let’s at least try to understand what’s actually going on. It’s too big and too important to address in a different fashion. That’s the first. The second: I think we need to accept that there are no solutions in this space. We live in irreducibly diverse societies, where we have different values, and we will not agree on how to do this. Just on the panel we see different principles and points of views, and that’s why I think it’s so important that we not only bring evidence to this discussion, but ultimately appreciate and move away from the way in which this discussion has played out for too long — which are unilateral announcements from individual companies into a forum like this, where I really appreciate that it’s uncomfortable to have sort of a spiky conversation. But it’s also really important, and I think it’s a sign of maturation from companies as well as communities, that we now can disagree in public to help each other understand it. Then each of us can ask based on our own values, “What do you think this should look like?” I mean, I already have my own research, but what do you think? This is the collective conversation we need to have.

Lydia Polgreen, Campbell Brown, Mark Thompson, and Emily Bell at Columbia Journalism School on June 14, 2018

EB:

I am going to ask the audience what they think in perhaps another two minutes, so get your questions ready. But on that point, you’re right. Facebook is an enormous structure of power; in economic terms, in civic terms, etc. But not as big as Google, which has its tentacles absolutely everywhere, and the fact that Google is getting really favorable ratings from news organizations is great for that particular relationship, but is it healthy when we have reports from organizations supported by Google? Is putting reporters in newsrooms in areas where Google is also selling software solutions and putting self-driving cars on the road OK? I just worry, as a complete outsider, that the relationship is just too cozy or is in danger of becoming too cozy, and that’s really significant. I’ve been looking at Lydia. Is there a danger here? You can say no.

LP:

I think there is a tremendous danger. That said, if Google wants to write HuffPost a check… (laughter). No, I think there is a tremendous danger in co-option. Look, there is a lot of scar tissue among news organizations who’ve collaborated with the platforms. You know, you spin up Facebook Live, you spin up this thing, or that thing. First they’re supporting it, and then they stop supporting it, which is fine. It’s their business. It’s up to them whether to do it. But I think any news organization that is depending on an experimental handout from a platform company is doing it wrong, and I think Campbell would agree with me on that, right? Many news organizations declined to take part in the newest Facebook Watch for precisely that reason. They just looked at the terms and said, “This doesn’t work for us,” and walked away. I think there is a tremendous problem in the concentration of power, but to me the much bigger problem is the concentration of economic power of these platforms. Do we really want a world in which 95 percent of advertising revenue is sucked up by two companies? The answer is probably no.

EA:

These are really good questions. They’re important questions you’re raising about independence, and we should have that conversation. Report for America, let’s talk about that, and let’s talk about the Google Fellowships that we’ve been doing for years. We have fellows around the world and they are selected by the organizations that are going to have the fellow. They pick their own projects; we have absolutely nothing to do with the selection. In fact, Neiman’s a good example. Neiman has had a Google fellow for years ever since it started, and they hold us accountable. The fact that you’re asking these questions is important, because it means we should be talking about this more, and we should share these principles about church and state when we give money away. Frankly, we really care about independent reporting and we think we should be held to account, too. It’s our expectation that when we do give money away, these organizations are independent to cover things like they should and they will.

CB:

Just two quick points I want to make. There is a fundamental tension that I think is always going to exist when media companies and tech companies try to partner with each other. Media companies need stability in order to build a business in the long term; they need to know what’s coming. Technology companies are exactly the opposite; they want to change, they want to innovate constantly. And I think we haven’t been great in the past about just being completely upfront with partners who want to experiment with us about how thrashy and rocky a ride this is going to be. So, we’re trying now, for example with Facebook Watch where we’re funding new shows, to say, “Look, this might not work. So, if you come onboard, know that. This may not work, but if you’re game to experiment, we’ll fund it.” Because we all want to find new formats, new ways to tell stories, and a lot of these organizations do not have the resources laying around to be able to experiment with these new ideas. We can help them try new things, and it may not work, none of it may work, but I do think when we’re trying to reach a new generation of people which are not consuming news the same way, you have to have some sort of format where we can test and try to experiment. And, just a broader point: I get that this is affecting Mark’s business, our decisions around political ads. I understand that. It’s something we’re deeply concerned about, Mark. I hear you. But, when we make this decision, I think you have to take a step back. I hear Mark’s concern, but this is about the integrity of our elections. We have a mid-term coming up. You know what happened on our platform during the last election. We have very big ways of trying to address many of these problems beyond the ones we’re already doing, which is trying to tackle the fake accounts, but also by just opening up everything and being as transparent as possible around the advertising that takes place on the platform — and that’s what we’re doing. We’re trying to throw it all up. That way reporters can search a publicly available database of everything that’s out there. And the reason it’s focused on political content and issues is because that’s what happened … that’s what was at the heart of the last election. It wasn’t about: You target trying to sell laundry detergent. It was about politics.

EB:

Just a very quick question. Would this just not be much better done by human beings? I mean, is it that hard?

CB:

You mean the the review process?

EB:

Yeah.

CB:

Oh, sure, sure. Capturing a cooking thing is a bug, and those things are going to happen and we’ll have to fix them. But I don’t see what’s wrong with saying, “The New York Times paid for a boost to this story about Donald Trump.” There it is, you can see it. There’s nothing wrong with that, and we label it in an archive that says “news content.” So, that’s what [Mark’s] opposed to. That we label it in an archive that say “news content,” not an archive that says “political content.” That’s a different place. You have to go search somewhere else for that. I have a really hard time understanding why the transparency of having your ads, your boosted content, appear there is such a problem. I view my job as to advocate for news organizations within our really big company with a lot of competing interests. But on this issue, I just think you’re wrong. I’ve been pushing like crazy for Facebook to be more transparent, and I’m going to keep pushing and it’s going to impact a lot of people, and I do think it’s the right thing to do.

EB:

Thank goodness I can turn it over to the audience now. If you can just say who you are and, as somebody said to me the other day, make sure there’s actually a question mark at the end of whatever it is you’re saying.

Audience member:

My name is Damaso Reyes, and I’m director of partnerships with the News Literacy Project, which is partnered with many of the organizations represented on this stage. Obviously, the media and technology landscape is very different from when most of the people in this room grew up, but the way that we’re teaching our young people about how to deal, and interact, and understand the media environment or the information landscape isn’t. I have a bias, but how important is media literacy, news literacy, and information literacy in solving the demand-cycle problem? Making sure people are aware that, “Hey, that was a great piece, but that’s an opinion piece, not a news piece.” Or when I see five people arguing on CNN, or MSNBC, or Fox, there’s very little news content there; that’s actually opinion. How, as tech companies and media companies, can we address the supply side?

EB:

Great question.

EA:

To Google, it’s very important. We just announced a $3-million grant to the Poynter Institute working with Stanford. I think your project is amazing, too. What Stanford is doing is teaching people critical reasoning skills. So, sure, you see an article telling you to ask yourself these 10 questions — we think that’s good. We also need to get in there earlier and, say, educate them about how to be like a fact checker. That’s some of the research they’ve done. They’ve given students of different ages websites to look at and asked if they’re trustworthy or not. A lot of those students stay on the page, they read, they look at the citations and do most of what fact checkers do — which is, they do lateral reading. They start pulling up other stuff, they see who’s the source, who’s paying for this, etc. We just have to invest in using more hard resources to do exactly what you’re doing. It’s hugely important to Google and you’ll see more coming from that.

CB:

I know Google’s been doing a lot on this, and this is one of the things that isn’t always as sexy as the fake news stuff and everything else, so it doesn’t get as much attention. But we are partnering with [the New Literacy Project] on developing a curriculum to give to schools around news and digital literacy. I’m so glad you’re doing it — Facebook is sponsoring it; you’re doing all the work. It’s about building this muscle that I don’t think any of us needed to have. We just didn’t have that many sources of news. We didn’t have to be as discerning as our children do. I have two kids. They’re going to grow up literally never having to open a textbook in their lives. We’re looking to give them the skills to be discerning about different kinds of news. Knowing the kinds of questions to ask is something they need to learn, and we don’t teach it in schools. So, developing this curriculum, which is what the News Literacy Project is doing, is pretty incredible.

MT:

If I can just say, one of the issues we’ve got is a lot of signals to audiences which are available in today’s newspaper. If you turn the literal, physical page of a newspaper to the opinion page — you know it’s a platform for debate rather than a news story. Or, a cable news TV show is obviously a discussion framing a political subject. We’ve lost those signals. They’re not there anymore. I think those signals aren’t apparent, particularly on a social feed or when you’re looking at the screen of Google AMP stories. I think we need to work together to put those signals back. I think the best people to put those back, by the way, are the originating publishers. We know what the intent — the intentionality — is. I think a lot of the supposed problems around distinguishing news from politics, which is around opinion, comes when people can’t see the labels we put on ourselves, so the reader knows exactly the status of what they’re looking at. The Times has a history of being a platform for ideas and debate going back a hundred and fifty years. Our traditional readers are used to it. We need to get our new readers used to the idea that these are ideas that stimulate news.

LP:

I do think that we’ve long overestimated people’s discernment even in the print era. I remember when I was a young reporter at The New York Times and they explained to me that the way to know the difference between a straight news article and a news analysis was that one was justified bracket right. (Laughter.) Oh, that will make it clear. But I never forgot it. I think we should remember this is actually a real issue.

EB:

Question over there.

Audience member:

Hi. Mike Millin, soon to be a fellow at the Social Science Research Council and my question is for Lydia. Earlier you cited the statistic that only 25 percent of people would be willing to pay for a subscription, and you seemed to suggest that is too low. What would be the appropriate number? And since we’re moving towards expecting that journalism subscriptions are a staple expenditure, is that putting a burden on a particular segment of society?

LP:

It’s a fascinating question. I think that the portion of Americans who have directly paid for news has always been pretty small. I don’t know the historical data, but I’d guess more people are paying directly for news now — though I actually don’t know. The reality is that a lot of people pay indirectly for news. Let’s say you have a cable bundle. Part of that bundle, a tiny sliver, is going to support the newsroom of CNN or MSNBC that, I assume, is not captured [in subscription numbers]. It is inspired in many ways by the fantastic work The New York Times has done in this area. I used to work there. I’m a huge fan of it. One of the things that we often talked about was the curious reader, and how important it is that The Times reader is not just an affluent person. It’s any person who’s curious. I think that’s a great way to think about who the ideal consumer of the product of The New York Times is. Now, let me ask another question. What about people who don’t think that news is important? What about young people? It’s not the most inspiring slogan in the world, but work for the incurious! They need to be informed.

I’m kind of kidding, but not really. I do think a lot about this question as we talk about the “un-newsed” and “under-newsed,” similar to the unbanked and the underbanked. Those news consumers who are never going to pay for news, people who aren’t particularly news literate. That’s a huge part of the population, not just in the United States but around the world. I spent the last 18 months of my life thinking about how to serve quality news to those audiences that, frankly, are probably never going to pay for news.

EB:

Rasmus.

RN:

Thanks for raising your question. I think it is an incredibly important one. Let me first say I believe strongly in the value of journalism, but — here comes the but — a lot of news publishers, when they talk about paying for news, what they effectively mean is, “We think you should pay for all the things we already do.” Hmm? It’s about prizing the product based on the cost to produce it. “We are not making money, therefore you should pay.” This is not sound business advice, right? I think this is where we need to start. There is a question: How does news create value for people? For actual people out there? Not people who come to us here, but the ones outside, 20 blocks, 25 blocks away. How does news create value? What are the problems it solves, and how can you convince people that it solves their problems? And when we see the numbers, I think it behooves us to think of social inequality. When you look, for example, at the Bureau of Labor statistics data on expenditures, the amount of money Americans spend on media has declined over time as the percentage of their income. Also, amongst poor people. When we’ve done studies in the UK, people in our studies say they have not consumed any professional news journalism in the last month, which by the way is about seven percent in this country. That’s a high number. When we ask about their media habits, they use Netflix, they have a smartphone, a data subscription plan. Subjects spend quite a lot of money on media, but they don’t spend any time with news and they don’t spend any money on news. Why? Well, because they do not feel the news creates any value for them. It does not make their lives better. In fact, it makes their lives worse. They find that it’s depressing, that it’s disempowering. They find it’s increased their alienation. It’s the kind of thing they would not let their children watch. This is a real, serious question for journalism. What is the problem that you solve? Who do you solve it for? And that has to be the starting point for every conversation. What is the value you create?

EB:

Isn’t journalism, though, like good government? It should work when people aren’t paying attention to it and that’s one of the problems with it? It has to do something to be transactional instead of just functioning?

RN:

That would be nice, but that’s not the world we live in. If it was, we would actually have good governance everywhere.

MT:

There’s a real risk of publishers focusing on the end of the story [and getting people to subscribe]. But that’s not the central question. The central question is: Can you offer genuine, fervent, valuable content to a user, in a way at the moment, that can build their engagement over time? If we can build the engagement, we can figure out the rest. I’m not saying subscriptions are the answer. To me, it’s a fatal mistake to believe it’s somebody else’s job to pay for it when you really just need to change it.

EB:

Jeff Jarvis, let’s hear your question.

Audience member:

I’m Jeff Jarvis. Here’s a question for the non-platform folks: Lydia, and Mark, and Rasmus. I’m about to start a project trying to aggregate the signals of quality, and lack thereof. External signals, besides things that the platforms already do. I think we in journalism need to help them make better decisions about promoting forms of quality news and advertising. So, my question is, what are the signals of quality in journalism that you would wish both platform technology companies and advertisers would try to push toward?

MT:

We’re comfortable with the idea of fighting to try to win our audiences on Facebook and the internet as a whole, to stand on our reputation to convince users one by one by the experience of our journalism. This stuff is difficult and dangerous, in my view, to somehow imagine it could be codified. One simple suggestion is there are a number of well-established industry bodies who could gather groups of publishers together. You know, well-established, properly vetted individual organizations with certain standards. But I’m reluctant to surrender a list of automatic signals, which say together that they demonstrate trust. It’s a complicated thing. There are times of majority error when mainstream media gets it wrong, and some untrusted outlier gets the truth.

CB:

Can I just respond? I’m one person and I have a small team, and so in our conversations with people I find it’s better to talk to publishers than to their lobbyists, because we tend to get the straight story and things don’t get lost in translation. If you would prefer I talk to your lobbyist, Mark, going forward I can do so, but just from my perspective in trying to solve problems and solve them quickly, getting a whole bunch of publishers together in a room is a lot more effective.

MT:

You’ve spent the whole session talking about your understanding on this topic, not just with us, but with other publishers and our entire industry. I think, just since I have the mic, the idea that Facebook is transparent is ludicrous. It’s ludicrous! Maybe Facebook is going to become a transparent organization. It’s worth remembering Cambridge Analytica, and that Facebook’s first reaction when the Cambridge Analytica story surfaced was to take legal action to attempt to get it suppressed. And the kind of Mickey Mouse-world in which Facebook is in favor of transparency and publishers are not is nonsensical! We’re completely comfortable in being transparent about our marketing on Facebook, but the idea that Facebook currently operates in a transparent way is ludicrous.

CB:

I didn’t suggest that, Mark. I think we have a lot to do in terms of trying to be far more transparent. I know you’re wound up, but we’re moving towards transparency in our advertising.

LP:

I’m very much in Mark’s camp on this question. I would rather wait for government to regulate large, unaccountable companies rather than to give the tools to those companies to regulate us. I wouldn’t want to give any ammunition to them, and say, “Oh, here, we’ve agreed you’re the sheriff and these are the rules by which we’re all going to play.” I think that would be suicidal on our part. Our independence, our identities, require us to think of ourselves as having our credibility judged not by our achievements, by Pulitzers, by big audiences, and those things; we need to grind out trust one user at a time, story by story, video by video, and that’s really going to be, from my perspective, the only validation that matters.

EB:

Right, over here.

Audience member:

I’m Gabriel Sams. I’m in news and journalism partnerships at reddit. I’m curious to know how you consider discourse on platforms, or on your own, as a complement to the content that you’re distributing or producing?

EB:

That’s a great question.

MT:

To me the fundamental object is to try and get as positive a dialectic as you can between the incoming stream of professionally reported and curated content, and public discussion of what that might mean and debate about what that might mean, and quality control on our own platform. We have very little control of what happens off our platform. We have had a theory going back to 1895 about intelligent and civil debates, and we’re pretty ruthless. If we can’t moderate to achieve that then we won’t run that content. You know, we’ve never had a problem with advertising from news organizations or with advocacy advertising. We’ve never had a problem distinguishing that from our news. We’ve never had a single complaint about confusion between news and advocacy, because we’re very careful about how we label advocacy. The same thing with content moderation. We want to have a standard.

LP:

One of the biggest changes we’ve made at HuffPost since I came on is that we actually shut down our open contributors network. For me, that was really important, because while it was an extraordinary innovation when it was invented in 2005 back when HuffPost first launched and it was signature part of the brand, fast-forward to 2018 and our problem is cacophony. Our problem is confusion about what is news and what is just someone spouting off. So, we’ve come from these kind of radical internet roots of being a platform for everybody to putting a much greater emphasis on trying to make it clear to our audience that these are our products and these are the things that we put our brand behind. Now that we’ve reached that equilibrium point where I think we’ve made that known, we do want to get back into that engagement game, having a deeper conversation. One of the things that is very admirable about The New York Times is its approach to getting great content from its readers and its commenters. They’re really interesting people, and I think that’s something you build over time with a very deeply engaged audience. I would look to a model like that for how we would want to start engaging with our readers now that we’ve got our house in order.

Audience member:

I’m Ben Jackson. I run a consulting firm and before that I was at Vice and The New York Times. I’ve built some algorithms in my day, and one of the things that stuck out to me was hearing about the thousands of signals on Bitcoin’s newsfeed. The first thing that came to mind was: I have no idea how this was made. That goes back to this fundamental question that you all can’t open your algorithms in public without allowing them the gains, so my question for Mark and Lydia, if you could put yourself in the shoes of the platforms, short of open sourcing their algorithms what’s the first step that you would take to start building trust?

EB:

Great question. Imagine, Mark, you run Facebook. What would you do first?

MT:

I think this is a difficult problem. I mean, to be honest, as a publisher this is not my problem to solve —

CB:

You had a lot of opinions on how to solve other things…

MT:

As it happens, some of the suggested solutions sound to me worse than the disease, actually…

CB:

So give us better ones. That’s the question.

MT:

I worry about having a kind of statutory or judicial algorithm review board as has been suggested. I think, as it were, handing control over all of this to politicians or their proxies runs its own manifest risks. And not literally, but one imagines that President Putin and President Xi Jinping already have such boards, as it were, and they control the way algorithms work. But, honestly, I want to say the first thing the platforms may have to consider is a much simpler approach to news, i.e. fewer signals, and the algorithms not continuously changing and optimizing — but longer periods where they’re stable and understood. The idea of enough stasis in the algorithm, that over a number of years everyone — the public, the politicians, the publishers, the platforms themselves — can see how it works, and then the adjustments are public. The idea is that this is essentially public but voluntary. Something else we should get is structured competition between platforms. I hope that involves civic objectives, as well as financial objectives. Taking less money from political advocacy units wouldn’t be such a bad plan as well. What went wrong in the last election was not the election process, but the role Facebook played in propagating false news in the election. Looking hard at quality control in terms of advocacy and advertising, that’s what we do at The Times. I know we’re doing a lot smaller quantity of advertising, but the principle is that if you don’t quality-control advertising, you’re going to end up with a problem, I’m afraid to say.

CB:

I have to say one thing. The best way to quality-control advertising is to shine a light on it: transparency.

LP:

I’m of the belief that transparency alone is not enough. One has to be proactive, because transparency puts the burden on the consumer, rather than the person who’s making the profit from it. If we’re going to say transparency, well, let’s put this drug out there and see how many people drop dead. We’ll put the results in a newspaper and that way we’ll all know what not to take.

I actually look at this from a slightly different perspective. I think that news organizations probably should be far less dependent on these platforms in general, but when I look at a platform I admire and I think is actually very positive in today’s news ecosystem, I would say Apple is a really good one. It’s interesting because Apple is going to make value judgments about what is news and what is not. They are going to have an editorial staff that’s going to say, “Yes, Fox News, The New York Times, Huffington Post, these are news organizations, and we as a company decided our value is that we want to create a product that serves people quality news. We’re going to take it upon ourselves to decide what’s quality news. We’re not going to worry about people squawking and saying, ‘You’re biased left, you’re biased right.’ We’re just going to bite the bullet and say, this is the product.” And guess what? People love that product. As a news consumer, I think it’s a fantastic product. Now, I’m not saying I want Facebook to say, “This thing indicates this level of quality, and this thing indicates that level of quality,” but I’m perfectly comfortable with a platform saying: “This is our platform, and we’ve decided that xyz players are the ones that are allowed to play on it — and they’re the ones we’re going to promote.” I think that, to me, is the type of platform that is best.

RN:

This is one of the areas of principal disagreement, where I think it’s good that Lydia has shared her opinion. This is sort of a whitelisting approach, where you make a positive judgment about one thing and say the rest is going to go. I would add to that another approach is the “do no harm approach,” where you try to prevent truly atrocious crap on your platform but otherwise try to establish ways to rank content. I am perfectly fine with Snapchat making decisions about which publishers are working with Snapchat Discover. I will be absolutely terrified of Facebook and Google saying, “We will only work with these two publishers and no one else.” I think that would be terrifying.

LP:

That’s fair.

EB:

That’s a great way to finish. I want to congratulate Rasmus on his research, and I want to thank the panel.

Tow Center

Written by

Center for Digital Journalism at Columbia Graduate School of Journalism

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade