By Bronwyn Howell
This week’s six-hour international outage of Facebook and its Messenger, Instagram, and WhatsApp services provided a costly wake-up call for both the Facebook conglomerate and its more than 3.5 billion users. While Facebook employees somewhat embarrassingly found themselves physically shut out of their premises, commercial users relying on Facebook’s platforms to communicate with and sell to their customers found themselves unable to trade. Likewise, billions of individuals were unable to communicate using apps they rely on. The outage was particularly harmful in low-income economies where, by dint of Facebook Zero sponsorship arrangements with local mobile network operators, Facebook has become a de facto internet service provider (ISP) enabling connectivity to the World Wide Web.
Predictably, the outage has been accompanied by outrage prompting calls for government and regulatory intervention. Rep. Alexandria Ocasio-Cortez tweeted that the outage justified breaking up Facebook. Similar sentiments were echoed as Facebook whistleblower Frances Haugen testified on Capitol Hill. Russia’s Foreign Ministry Spokeswoman Maria Zakharova cited the outage as evidence that Russia should assume sovereignty over the internet in its domain. And while European Union Commissioner for Competition Margrethe Vestager stopped short of calling for regulation, she noted the incident illustrated a need for more competition.
The question remains: What could regulation have done to avert this week’s outage?
First, it appears the outage arose as a consequence of human error. The root cause has been attributed to a Domain Name System problem — namely with the “address book” that tells servers worldwide where to direct internet traffic. Changes to the configuration of routers directing traffic between Facebook’s data centers were flawed. It took time to resolve because network engineers, prevented from resolving the issue remotely, had to physically travel to the centers to make the necessary changes, which had to be coordinated globally.
No amount of regulation can foresee or prevent human error of this type. The issue can beset any firm of any size in any jurisdiction. It is a costly and embarrassing situation for Facebook, but ultimately this is an issue that can be addressed only by the firm itself. The fall in share price, loss of advertising revenues, and reputational damage leading to customers switching to alternative platforms already provide sufficient incentives for Facebook to take internal measures to reduce the likelihood of the issue recurring. A further dose of regulatory discipline is unlikely to achieve anything further.
Second, it warrants questioning whether the extent of the costs borne by end users were to a large extent avoidable had prudent “insurance” been sought prior to the outage. As Vestager correctly identified, competition provides the best means of ensuring Facebook maintains exemplary performance in users’ interests. None of the individual Facebook services caught up in the outage has a monopoly (at least in developed economy jurisdictions). This is best illustrated by the fact that Facebook itself used one of its rivals — Twitter — to communicate with its users during the outage.
While Facebook may be large because its consumers prefer its services, they have options, and there are no barriers to “multi-homing” (belonging to multiple platforms in the same category). While Facebook languished, rival sites flourished: Messenger rivals Signal and Telegram saw spikes in both traffic and user sign-ups. While users may have faced a temporary inconvenience in having to sign up (and coordinate with other parties to whom they wished to communicate), those who had already anticipated and planned to move to an alternative in advance would have been barely troubled.
The likely largest losers in the outage were those using Facebook services alone to distribute digital content and sell products using Facebook’s Marketplace. Multi-homing is more expensive for these firms, but it nonetheless provides insurance in the event of things not going to plan. As we have seen in the pandemic world, having alternatives is valuable when the unexpected occurs. Perhaps rather than regulating Facebook, more could be gained by encouraging greater consumer awareness of multi-homing’s benefits, thereby enabling competitive forces to do the heavy lifting.
However, a caveat regarding the context in which the services operate warrants consideration. No country can regulate Facebook outside of its own jurisdiction, but regulations in one will inevitably have consequences in others given the firm’s global presence. Facebook may be used differently in low-income economies than in high-income ones. While in high-income economies, users are accustomed to having access to large amounts of data traffic enabling multi-homing, in low-income economies Facebook Zero has powers similar to those of an ISP, so perhaps it should be accountable to a different set of obligations reflecting these differences. That is, the activity, not the firm, should be the subject of regulation. Of course, this should apply to all regulatory intervention. Anticompetitive actions rather than artifacts of existence should be the yardsticks.