The NTIA and Section 230

Download this Public Interest Comment (PDF)

The NTIA proposal includes several provisions that would narrow the scope of Internet intermediaries’ liability when they remove or restrict access to content provided by others. It would also require the intermediaries to disclose their content moderation policies in a form that is understandable by consumers and small businesses. Those two sentences of course do not capture all of the legal subtleties involved, and this comment takes no position on the legal issues raised by the petition. However, I believe that in deciding whether to propose a regulation in response to the NTIA petition, the FCC should be fully aware of the analysis required to identify the likely economic effects of the NTIA proposal and other alternatives the FCC may consider.

The duties of the FCC’s Office of Economics and Analytics include preparing “a rigorous, economically-grounded cost-benefit analysis for every rulemaking deemed to have an annual effect on the economy of $100 million or more.”[1] Relevant economic effects could be costs, benefits, transfers, or other positive or negative economic effects. A rulemaking based on the NTIA petition would likely require a full benefit-cost analysis.

The rules requested by the NTIA could create significant economic impacts by altering Internet intermediaries’ content moderation practices and/or altering investment in new and improved services or innovative new companies. Given the large value consumers receive from Internet intermediaries and the size of investments in this industry, even a small regulation-induced change in the companies’ economic incentives would likely generate an annual economic impact exceeding $100 million.

Consumers clearly derive enormous benefits from Internet intermediaries. For example, a 2019 National Bureau of Economic Research (NBER) study estimated that use of Facebook created $213 billion in consumer surplus between 2003 and 2017.[2] Another NBER study estimated that one month of Facebook use creates a total of $31 billion of consumer surplus in the US.[3] Laboratory experiments found that students place significant value on other Internet intermediaries as well.[4] Indeed, since there are 172 million[5] US users of Facebook alone, a regulatory change that altered the average value of the service by just 59 cents per user would have more than $100 million in economic impact. Similarly, a National Economic Research Associates study estimated that adding five seconds of advertising per web search would increase web browsers’ ad revenues by about $400 million annually;[6] thus, if a regulatory change led to a two-second increase in advertising per search, the effect would exceed the $100 million threshold.

The rule NTIA requests could have economic impacts beyond its direct effect on consumer surplus generated by incumbent firms offering their current suites of services. A 2019 study by the Copia Institute presents some comparisons which suggest that Section 230 liability protections (or similar policies) help companies attract more venture capital investment and improve their chances of survival.[7] The study compares the experience of companies in the US versus the European Union; US digital music companies versus US social media and cloud computing companies; and intermediaries in several other countries where liability protections identifiably changed. This study does not control for other factors that might affect the results, so its conclusions are only suggestive, but the pattern suggests that more extensive data analysis could be informative.[8]

A 2015 study by Oxera took a different approach, combining literature reviews with interviews of 20 experts to assess how liability protections for intermediaries affect intermediary start-ups. It found that stronger liability protections are associated with higher success rates and greater profitability for start-ups.[9]

Whether a regulation-induced change in venture capital funding for Internet intermediaries, or their success rate or profitability, should count as a benefit or a cost depends on whether the current level of startup activity is above or below the economically optimal level. That is a key question a full benefit-cost analysis should help answer. My point here is a much more limited one: the NTIA’s proposal could very well affect investment flows by more than $100 million annually.

Thus, in one way or another, the NTIA proposal is likely to have economic effects that exceed the $100 million annual threshold and hence require a full benefit-cost analysis.