America Needs a Public Interest Approach To Solving Big Tech Harms To News
America Needs a Public Interest Approach To Solving Big Tech Harms To News
America Needs a Public Interest Approach To Solving Big Tech Harms To News

    Get Involved Today

    France and Australia have moved forward with plans to force Google (and, in Australia, also Facebook) to pay existing media companies for linking to news content. In theory, the laws adopted in France and Australia are designed to compensate these news organizations for the benefit Google and Facebook derive from linking to news. But both France and Australia have made it clear that these companies cannot simply choose to stop linking to news sites. This goes beyond the idea that Google and Facebook are unjustly enriched by linking to news content, and beyond the idea that news organizations should have some copyright-like right to control who links to them. In an effort to ensure the continued production of news — an important public interest goal — France and Australia have created a structure that will both preserve the existing market power of dominant platforms and the market position of the largest news publishers by requiring negotiations between the two. Worse, by framing this as a negotiation to prevent unjust enrichment, the Australian and French regimes threaten to bring news blackouts to the internet similar to the blackouts of television programming on cable and satellite services when similar “retransmission consent” negotiations fail.

    Finally, while the current proposals limit themselves to Google and Facebook on one side and newspapers and other news providers on the other, these proposals establish the principle that linking transfers value from the linked source to the linker — an unjust enrichment where the linker should pay the linked to website. This constitutes a radical change that threatens the fundamental nature of the internet as it exists today. If this principle becomes established in law, we can expect it to expand, undermining the interactivity that makes the internet such an important medium of speech and civic engagement.

    Unjust Enrichment v. Public interest Obligations

    First, Public Knowledge (and many others) agree that democratic society needs robust journalism to survive, and that the current economics of the internet do not sufficiently support the journalism we need. In particular, we need a combination of strong, independent local journalism and competing sources of national and international reporting that meet rigorous journalistic standards. We need to find ways that support professional journalism, while fostering those elements of the current internet environment that facilitate citizen reporting and citizen engagement.

    Nor is the question here whether or not Google or Facebook have market power — particularly in the advertising market. Issues about whether these companies use their power in the market to disadvantage publishers or favor their own vertically integrated products are very real and need addressing. That’s why I wrote a whole book about how to combat their power, and why Public Knowledge supports the ongoing antitrust lawsuits in this area. But the proposals from France and Australia (as well as similar proposals in the United States) will actually aggravate this market power by making the largest news companies beneficiaries of these unfair practices. This is why Public Knowledge, and other public interest organizations, have pushed to address these problems with targeted solutions and creating a regulatory agency that would directly attack the mechanisms of market power wielded by digital platforms.

    We can distinguish between the approach advocated by Public Knowledge and the approach taken by France and Australia as the “public interest obligations” approach v. the “unjust enrichment approach.” The public interest obligations approach does not turn on whether or not dominant platforms engage in bad behavior or receive an unfair benefit. Rather, the public interest obligation theory recognizes that we have a classic market failure. The current market does not produce a valuable public good (here, news). We therefore place obligations on those most capable within the current market (dominant platforms) to correct this market failure by supporting local news production.

    The unjust enrichment approach rejects this public interest formulation. Instead, the theory of unjust enrichment posits that the current market structure would produce the public good (news), but the market does not produce news because some parties (dominant platforms) unfairly benefit from the product (news) without adequately compensating the product creators. News producers argue that linking to their content gives tremendous value to dominant platforms and that if dominant platforms had to negotiate with news producers for a fair market price, then we would eliminate the crisis in journalism because “the market” would justly reward news producers.

    The unjust enrichment theory currently has many supporters internationally and in the United States. We have already seen efforts to exempt news organizations from antitrust laws so they can collectively bargain with dominant platforms. As demonstrated in France and Australia, the unjust enrichment theory doesn’t work because dominant platforms don’t receive significant commercial value from linking to news (as compared to their other sources of value). And it cannot work without further creating a new substantive right on the part of news publishers to prevent being linked to, except on their terms. As a result, both France and Australia have gone well beyond requiring market-based collective negotiations to a regulatory system that employs market rhetoric.

    What Happened in France?

    In 2019, the European Union adopted a new Copyright Directive on digital copyright. Among other things, Article 15 of the Digital Copyright Directive requires digital platforms as defined — basically Google and Facebook — to pay news publishers for any incidental use of their content other than a bare link. (You can see the EU press release explaining Article 15 (previously Article 11) here. A less complimentary view here.) In the EU, member states must implement EU level directives through national law. France promptly passed its own version of Article 15 and ordered Google to negotiate with French news publishers for payments to link.

    Predictably, Google said no. Instead, since Article 15 (and therefore the French version of Article 15) clearly allowed Google to publish a naked link, Google announced that is exactly what they would do. This would put French news publications at a significant disadvantage in search results, since people gauge whether a search result is responsive to their request by reading the extract or thumbnail that appears with the link. Bare links generally don’t generate a lot of traffic.

    At this point, the French Competition Authority (FCA) intervened. It told Google that refusal to use extracts or thumbnail sketches of links from French news companies constituted an “abuse of market power.” The FCA ordered Google to engage in “good faith negotiations” to determine how much they would pay news providers. Period. Not using the “copyrighted material” from newspapers was therefore no longer a legal option for Google.

    Last week, Google announced a new negotiated agreement with French news publishers. Google announced a framework under which Google will negotiate individual licensing agreements with French news publishers. However, rather than feature these in Google’s organic search, they will post the licensed news publishers in a new product called Google News Showcase. Google is rather vague on the details of the framework, stating that criteria for compensation include “the publisher’s contribution to political and general information, the daily volume of publications, and its monthly internet traffic.” In other words, the largest and most established news publications will get the largest license fees.

    In addition to favoring larger publishers over smaller publishers, Google News Showcase illustrates how this negotiation approach fails to get at what genuinely drives Google’s advertising dominance. Google News Showcase addresses some publisher concerns, but It still keeps users on Google’s website and gives Google access to the same information as before. Publishers and users would do much better in a system where Google quickly facilitates users going to the publisher’s own site, instead of finding ways to keep the user on Google, as Google News Showcase does.

    It remains to be seen if the FCA finds this framework acceptable, or if it will require further “negotiations” to reach what it considers a “fair” result.

    The Australian Approach

    With the advantage of this history, Australia decided to craft a policy that would absolutely guarantee that Google and Facebook “licensed” Australian news content. The Australian Parliament is now considering a Mandatory Code of Conduct (text here, official memorandum of explanation here) which would compel “designated platforms” (those found by the Australian Competition and Consumer Commission (ACCC) to have market power, i.e., Google Search and Facebook) to negotiate with any registered news organization that earns over $150,000 (AUS) a year and complies with the Australian media and news code of ethics.

    To ensure this works out that designated platforms pay a proper amount, the statute requires:

    1. Designated platforms cannot refuse to carry a registered news organization that demands payment, or in any way disfavor the news organization in its search or recommendation algorithms.
    2. Designated platforms cannot refuse to carry Australian news if they carry news from any other country. “News” is defined incredibly broadly so as to make it effectively impossible for designated platforms to provide any content that discusses Australian current affairs from outside Australia without triggering the requirement to carry all Australian registered news organizations.
    3. Designated platforms must honor each individual request, but news organizations have the option to band together under a single representative to “balance” their negotiating power.
    4. If the parties don’t reach agreement, then it goes to a mandatory arbitration panel supervised by the Australian Communications and Media Authority. The parties then submit their “good faith best estimate” of appropriate payment by the designated platform which reflects: (a) the value to the designated platform (monetary or otherwise) of the covered news content; (b) the value to the news organization of having its content on the platform; (c) the cost of creating the covered news content; and, (d) whether a particular remuneration amount places an “undue burden” on the commercial interests of the platform.

    There are some other things in there as well. Notably, the designated platform needs to turn over clickthrough data to the registered news organization (in a manner consistent with Australia’s privacy law), and must notify a news organization 14 days before changing its algorithm in a way which could impact them. These provisions address some significant concerns by publishers that platforms insert themselves into the relationship with their customers and that algorithm changes in the largest platforms can have a serious impact on their businesses. While not endorsing or opposing these specific solutions, I will note in passing that PK has highlighted these issues in the past as legitimate concerns about the market power (intentional or not) of dominant platforms.

    The French and Australian Approaches Have Very Bad Policy Consequences

    Once again, PK supports the basic idea that news is a public good and that giant digital platforms should pay to support production of news. PK has some very specific proposals regarding how to do this, consistent with the public interest. You can see details of our “Superfund” proposal (to clean up toxic misinformation and address news deserts) and what public interest obligations we would impose on Google, Facebook, and other large platforms here, here, and here.

    The approaches taken in France and Australia illustrate some of the very bad results that flow from taking this “unjust enrichment” theory rather than adopting a “public interest obligations” theory. I’m not talking about just the usual problem PK has with copyright maximalism and creating a new “intellectual property right” for the benefit of the large content providers.

    The Australian and French approaches reinforce and reward media consolidation. If you are a media corporation, this is a feature, not a bug. But after 20 years in the media reform movement, I can’t believe the answer to the problem of ensuring production of local news from diverse sources — especially traditionally marginalized voices — is to have a government-supervised wealth transfer between the biggest digital platforms and the biggest media conglomerates. That a handful of smaller, independent organizations might also score a payday doesn’t go to the heart of the problem — ensuring that the public good of news production necessary for democracy actually happens.

    This will end up reinforcing the existing structure (and — I expect — will actually exacerbate concentration and further lock out independents). History warns that regimes which rely on negotiations among the players end up driving consolidation because the parties need leverage. We saw this repeatedly in traditional media consolidation. Cable operators argue they need to consolidate to negotiate retransmission fees, which prompts broadcasters to consolidate so they can gain greater leverage against cable.

    Arguably, the French and Australian models try to eliminate the need to consolidate by allowing for collective bargaining by publishers. The French model expressly requires a negotiated framework between the news industry as a whole and the relevant platforms. The Australian model is structured for individual negotiations, but also allows a single negotiator to represent an unlimited number of news organizations. But this has the same impact as the consolidation of the past. The large players will control these negotiations and smaller stakeholders, local newspapers and citizens, won’t have a say.

    Third, the various formulas to determine the value of the news content will all invariably slant toward the biggest news organizations. That’s not simply because those are the companies in the room with the biggest muscle on the formula (and the ones Google and Facebook have the biggest political incentive to pay off and make happy). It’s because the notions of “value” — such as overall ratings/readership in the offline world — are all slated toward the biggest existing providers. After all, the largest news organizations must give the greatest value to the platforms, since they are the most popular. If you are buying the theory that people are using Google or Facebook because of the presence of links to news stories, it naturally follows that the most popular news outlets, controlled by the largest companies, add the most value and should get the biggest payments. Q.E.D.

    For all these reasons, despite the facially neutral structure, the Australian and French approaches essentially end up with the biggest media companies and the biggest platforms duking it out, with the little guys scrambling for crumbs and trying not to get crushed in the process.

    The “Unjust Enrichment” approach fundamentally undermines the architecture, history, economics and basic functioning of the internet. Today, no copyright or “adjacent” right applies to linking. In theory, we could create such a right under the “unjust enrichment” theory. But there is no intellectually principled way to limit this expansion of copyright to links and snippets to news organizations on one side vs. dominant platforms on the other. A copyright is a copyright is a copyright; copyrights are rights held “against the world,” not just “dominant” players. From a legal standpoint, such a right bars anyone — big or small — from linking without a license. Just ask any individual sued for hundreds of thousands or even millions in copyright damages. Nowhere in copyright law does a right spring into existence only when we think the entity holding it does worthwhile or socially valuable things. (Notably, retransmission consent for broadcasting is not a copyright, but a benefit conferred under the Communications Act and regulated by the Federal Communications Commission. This contrasts with the mechanical copyright for original programming created under the Copyright Act of 1976.)

    I generally tend to regard slippery slope arguments as somewhat suspect. But it is fair to ask here what is the limiting principle that prevents this from spreading to all linking or snippets by anyone, not just linking by dominant platforms (or their users) to news. If the answer is “because news is a public good,” then we should adopt a public interest theory. It is hardly worth undermining a central feature of the internet to pay for the public good of news when a straightforward obligation, such as Public Knowledge’s Internet Superfund proposal, supports the public good of news without undermining the history, architecture, and economics of the internet.

    Do we want to introduce news blackouts to the internet? If this is a question of negotiating a license, then Google and Facebook can simply decide the price demanded is too high and walk away. That’s how a negotiation works. We see this constantly in the negotiations between cable operators and broadcasters. In the three decades since Congress required retransmission negotiations, the number of blackouts has continued to rise — and at an accelerating rate. If we seriously believe that news is a public good critical to democracy and self-governance, we cannot have significant publishers pull links from the largest digital platforms as a negotiating tactic. We recently saw one broadcast group threaten to pull their signal and deprive millions of Americans of the opportunity to see the Super Bowl live. Imagine the largest news publishers collectively threatening to block linking from Google or Facebook in the lead up to a national election as a negotiating tactic?

    The alternative, as France and Australia have demonstrated, is to have government come in and set the price. Either indirectly as in France (where the French competition authority will simply force Google to keep negotiating until Google agrees to an arrangement it considers fair), or directly as in Australia (where the law dictates mandatory arbitration based on the factors listed in the statute). But once we have government setting the price, how do we pretend this is a “free market” negotiation that is producing fair market value? It basically becomes a government supervised wealth transfer from dominant platforms to the largest news publishers. It would make far more sense to have governments decide the best public policy for the future of news rather than outsource that determination to monopolists and cartels.

    This incredibly inefficient and unfair process fails to get at the root causes of the problem, or limit the market power of dominant platforms. These payments don’t actually address the many structural problems in modern media that have existed for over two decades — as I discussed extensively in this blog post here. Instead, they perpetuate them by funding the existing structure. The accelerating loss of local news and the decline of journalism were well underway in the Aughts, as was the loss of ad revenue to online advertising (when it was supposedly Craigslist that was killing ad revenue). While Google and Facebook potentially engage in conduct that captures advertising information from news publishers and pushes them to use affiliated services and formats such as AMP, nothing about the “unjust enrichment” theory actually addresses these unfair practices.

    No tweaks or changes will produce a better outcome for any remedy based on “market negotiations” between the news industry and dominant platforms. You cannot get good policy if every single element of the policy is based on magical thinking combined with a complete disregard for existing facts. Instead, we see Google respond by highlighting its own Google News Showcase product. The negotiation becomes how to buy off politically powerful publishing interests rather than anything resembling fair market value to remedy an unjust enrichment. News publishers have no incentive to negotiate for policies that will address the real problems with their business model for producing quality journalism, nor do dominant digital platforms. Rather than pursue a fictitious unjust enrichment strategy, governments should adopt policies designed expressly to promote the production of quality, independent journalism and policies that would limit the market power of dominant platforms. 


    This is one of those situations where there is just so much bad that I haven’t even covered the more nuanced concerns. Whatever happens in Australia or France, we must not allow this “unjust enrichment” theory to become the basis for policy in the United States. We have a growing movement to use the full power of antitrust enforcement combined with smart regulation to target the harms of dominant platforms. We also have a choice of public interest approaches that will actually foster the production of more local and independent news — including providing opportunities for traditionally marginalized communities to provide their own reporting. It would be profoundly unfortunate if proper concern for the future of journalism and well-founded concerns about the exercise of market power combined to reinforce every existing market failure in both the news business and the digital advertising business.