My Interview With Mathew Ingram Regarding Section 230
[Note: I did this interview with Mathew Ingram in late February–before the recent flood of new anti-Section 230 activity in DC that I still need to blog. I had always planned to share it here, but the issue got backburnered as the pandemic killed 120k+ Americans and shut down our country. I’m sharing it now because so many people in DC are now prioritizing the destruction of Section 230 over addressing the many other devastating problems in our country.]
Ingram: …Eric, could you start by giving us a brief overview of your thoughts on Section 230, and whether or not you think it needs to be amended — and if so how, or if not, why not?
Goldman: The main arguments in favor of Section 230 haven’t changed since 1996:
First, Congress enacted Section 230 because it could not imagine what the Internet was capable of becoming. That remains true today. Though we now have Internet incumbents and a well-developed Internet architecture, we still can’t reliably anticipate what the Internet might look like in 25 years or even the next 10. I’m confident the Internet will continue to evolve in unexpected and socially beneficial ways. As it did in 1996, Section 230 provides the legal infrastructure for that next generation of innovation.
Second, Section 230 reduced the barriers to entry in the user-generated content (UGC) industry. In 1996, that ensured a wide and diverse set of new marketplace entrants. Even though we now have Internet incumbents who would be delighted to see their competition thwarted, Section 230 continues to keep the door open for new entrants. The next Google- or Facebook-killer will emerge only with Section 230’s assistance.
Third, Section 230 eliminated the “moderator’s dilemma”–the concern that Internet companies won’t try to moderate content if they face liability for any mistakes in their moderation efforts. If Section 230 is limited, then the moderator’s dilemma comes back into effect, and Internet companies will make choices we wish they didn’t make–such as simply eliminating certain categories of UGC because it’s impossible to perfectly manage those categories.
Because the principal rationales for Section 230 remain as true today as they were in 1996, I continue to support the law as it is. However, I do think Congress could improve its efficacy by coupling it with a federal anti-SLAPP law in the cases where Section 230-preempted lawsuits are brought for anti-social or pro-censorial purposes.
Ingram: Thanks, Eric. In my conversation with James Grimmelmann from Cornell on Galley, he argued that Section 230 was outdated and needed to be amended, because among other things he said that it was “too strong as applied to truly bad-faith actors who rile up their user populations to post incredibly harmful stuff that is technically coming from a third party so that the platform itself is immune.” How would you respond to that?
Goldman: It’s true that we have seen some rogue actors invoke Section 230 to advance their anti-social agendas. Nevertheless, I don’t agree with Prof. Grimmelmann’s conclusions for at least three reasons.
First, Section 230 does not apply to federal criminal prosecutions. Rogue actors can be, and are, taken down using those existing crimes.
Second, other law enforcement agencies have been pretty clever about stretching their legal authority to target rogue actors. For example, Section 230 has not stopped the FTC from shutting down “revenge porn” operators.
Third, rogue websites rarely last very long. They flame out quickly due to a combination of bad public relations, legal risk, and monetization challenges. Indeed, vendors to rogue websites often intervene to terminate services–and those vendors often rely on Section 230 to make their termination decisions.
Ingram: Thanks, Eric. When I spoke with Olivier Sylvain of Fordham Law earlier this week, he repeated arguments that he made in an essay for the Knight First Amendment Institute in which he discussed how the implementation of Section 230 has missed a key point — namely, how platforms like Facebook use data to target users and shape their experience, in addition to just publishing their content. He argues that the information we see “might be qualitatively different or, worse, disproportionately harmful” but under the current law, courts never get to evaluate whether this disparate treatment is unlawful or otherwise harmful. How would you respond to that kind of argument?
Goldman: All publishers seek to target their users and shape their experience as part of the ordinary publication process. Obviously different media permit different ways of targeting users and shaping experiences. Still, I don’t get a lot of valuable insights from distinguishing between how online “platforms” remix content and other types of publication processes.
Having said that, I’ve noted the scenarios where Facebook’s algorithms may cause ads to be delivered with discriminatory impacts even though the advertisers’ selected display criteria did not signal any potential discrimination. In that case, Facebook’s algorithms may be creating discriminatory effects where none would have occurred otherwise. I think it’s a tough legal question figuring out how Section 230 applies to that circumstance.
Ingram: Thanks, Eric. When I spoke with him earlier today, former Homeland Security secretary and NSA general counsel Stewart Baker said that Section 230 needs to be amended because the protection from legal liability is essentially a subsidy, and these large companies don’t need or deserve that subsidy any more. He also argued that the large platforms suppress conservative speech, and therefore don’t deserve protection because they are infringing on people’s rights to speak. Any thoughts on either of those points?
Goldman: It’s incredibly myopic to want to reform Section 230 because Google and Facebook benefit from it. The Internet ecosystem has a vibrant and extraordinarily long tail of services that benefit from Section 230’s “subsidy.” Changes to Section 230 would disrupt or eliminate the services in that long tail, with the unintended consequences of further entrenching the Internet giants like Google and Facebook. I explore the “subsidy” point and the competitive consequences of Section 230 in this short essay.
Implications that Internet companies uniquely target “conservative speech” are really not credible. All content moderation decisions privilege some content over other content. People who didn’t get their desired moderation outcomes will routinely gripe that the decisions were “biased” against them. Due to the Internet services’ massive scale, they all produce mountains of anecdotal evidence of “bias” against every class or viewpoint of speakers. However, this doesn’t prove that the Internet services somehow were biased against those speakers; instead, the evidence simply proves that Internet services make a large number of content moderation decisions.
(I also remind readers of the troubling overlap between “conservative” speech and repugnant speech that absolutely should be moderated).
Finally, Internet companies aren’t “infringing on people’s right to speak.” They are applying their editorial standards to decide what content to publish. Users do not have a legally enshrined “right” to speak on Internet services. Instead, obligating Internet services to carry speech they don’t want to carry would infringe the Internet services’ constitutionally protected rights to publish (or not publish) content as they see fit.
Ingram: Thanks, Eric. Danielle Citron and Benjamin Wittes have argued that the protection of Section 230 should be narrowed, to the point where it would only be available to those who make “reasonable efforts” to moderate the content on their platforms (although how to define what is reasonable is not spelled out). Do you see any merit in that suggestion?
Goldman: No. The proposal, by design, would reduce Section 230’s substantive scope. Some might cheer that; I do not. More importantly, the proposal would functionally eliminate the substantial procedural benefits that Section 230 provides. I explain that angle here. As a result, the proposal grossly misapprehends the mechanics that make Section 230 effective as an immunity. I consider the “reasonableness precondition” proposal to be equivalent to repealing Section 230, just packaged in less obvious terms.
Ingram: Thanks, Eric. One last question: Why do you think Section 230 has come under fire from so many different quarters in the past year or two? There seems to be a sustained attack from government and those with partisan political agendas, but also from some in the legal community as well. Is it just because Facebook seems so dominant and is implicated in so many negative things and no one can think of what to do about it?
Goldman: We’ve seen a broad-based techlash in the last half-decade, and Section 230 has been ensnared in that techlash downdraft. What has prompted the techlash? There are many causes, but a few specific incidents stand out: Russia’s interference with the 2016 presidential election, including voter manipulation via social media; Facebook’s Cambridge Analytica data leakage; and the Nancy Pelosi “cheapfake” video (where her speech was slowed down to give the impression she was drunk). The Nancy Pelosi video perhaps has been most damaging of all because Congressmembers were shocked and furious that Facebook wouldn’t protect someone who looks just like them from such a blatantly false attack.
The massive size and power of Google and Facebook also contributes to the techlash. We generally distrust big and profitable companies; and anti-elite populism has tainted public perception of Google and Facebook in some circles.
Separately, regulators view Google and Facebook as threats to their own political power, and regulators routinely push back on such threats. In particular, Section 230 provides a convenient bargaining chip for Congressmembers trying to advance their regulatory agenda. If Facebook and Google value Section 230, Congressmembers increase their negotiation leverage by putting it in jeopardy.
Similarly, corporate rivals and enemies of Google and Facebook have targeted Section 230 as part of their own power gambits. For example, technology companies like IBM and Oracle, and entertainment companies like 21st Century Fox and Disney, have spoken out against Section 230. The two things these companies have in common: (1) they have essentially no first-hand exposure to UGC, so they face no downsides if Section 230 disappears, and (2) they view Google or Facebook as threats to their business for various reasons (e.g., Oracle’s total war against against Google spilling out from Oracle v. Google; and the entertainment companies’ long-term strategy to undermine legal protections for UGC in every way they can). When other technology companies speak out against Section 230, regulators then misleadingly claim that the “tech community is divided” about Section 230, which masks the unanimous opposition to Section 230 reform among the companies that actually will be affected.