This was a big week for administrative regulation of the internet in the United States. Specifically, the regulatory framework known as "net neutrality" was voted to be rolled back by the FCC. We've previously discussed the changing disposition of the FCC, and conservative administrations and their likely impact on TripleLift's business, both of which are a good primer for this Lift Letter.
What is net neutrality?
Net neutrality is a regulatory framework in which all internet service providers have to treat all traffic the same. This means they can't slow certain types of traffic or providers of content down, they can't charge different rates for different types of traffic, and they can't have additional preferential treatment (e.g. fast-lanes, zero-rating - which is the practice of not counting certain traffic against a bandwidth cap, etc). The concept can apply, theoretically, to any country and their regulation of the internet - but the recent regulatory change only applies to the FCC in the US. The EU has a statute that is pro net-neutrality but leaves significant power for member states to craft enough workarounds that it is effectively neutralized, so it is on a country-by-country basis (the Netherlands and Slovenia are particularly pro).
The FCC itself was created by the Communications Act of 1934. This was significantly revised in 1996 by the Telecommunications Act, parts of which we discussed previously in Lift Letter on media ownership. Title II of the Communications Act, as amended, defines a "common carrier" and provides a set of restrictions on what these entities can and cannot legally do. A common carrier in common law (i.e. jurisprudence) is an entity that provides transportation for people, goods, etc on regular routes, at set rates, to the general public - and it is generally required to provides those services without any discrimination on rates etc, as long as there is availability. In the Communications Act, digital "common carriers," per 47 USC 203, are likewise not able to provide preferences or disadvantages based for otherwise similar communication services. Telecommunications services had historically been subject to the regulations of common carriers, and were explicitly so regulated in the Telecommunications Act.
In 2015, the FCC issued the Open Internet Order, voted in favor 3-2 by the then democratic majority (with two members that are now in the republican majority, Pai and O'Rielly, dissenting), that categorized broadband internet as common carriage, regulated by Title II of the Communications Act. The order specifically outlines that legal content cannot be blocked, throttled or prioritized in any way. There are certain exemptions for "reasonable network management", but this is given a relatively firm boundary to avoid abuse.
For some historical context, much of this conversation dates back to 2005, and the Supreme Court case of National Cable & Telecommunications Ass'n v. Brand X Internet Services. The court used a common legal doctrine known as Chevron deference to rely on the administrative agency (the FCC, in this case) and its determination that, at the time, broadband services were to be considered "information services" rather than telecommunications services (the court notably didn't itself decide that broadband services were information services, but instead decided that it was reasonable to rely on the FCC's determination). In 2010, the FCC released their Open Internet Order of 2010, in which they tried to effectively create net neutrality by limiting throttling, discrimination etc of broadband services, but this order did not change the determination of broadband as being an information service (it instead relied on the Telecommunications Act, section 706). Verizon successfully sued that this rule was illegal, all the way through the Supreme Court, in the case of Verizon v. FCC. The court ruled that the restrictions in the 2010 order were only valid when applied to common carriers, not information services - which, at the time, broadband services were. Thus the 2015 order cast broadband as a telecommunications service, thus being able to be regulated like a common carrier, and thus gave it the FCC the grounds to impose the net neutrality type restrictions that it had sought.
What just happened?
The FCC is run by five people, three appointed by the president and two of the minority party, with one chairman among the five. When Trump assumed office, the composition swung to three republicans and two democrats. While Ajit Pai became the chairman in Trump's administration, he was already on the FCC - nominated by Obama (but in one of the republican seats, per Mitch McConnell's original nomination). Ajit Pai is a staunch opponent of net neutrality and led the effort leading to its repeal, but Brendan Carr was the new Republican member that in effect added the deciding vote of three Republican members. Pai et al released a draft order in November that would restore broadband to being an "information service" (Title I, as opposed to Title II), which was officially adopted by a 3-2 vote along party lines this past week (and defended by Pai, awkwardly, in this video). As broadband is once again an information service, the restrictions against throttling and traffic discrimination per both the Open Internet Orders of 2010 and 2015 are no longer applicable.
The new FCC order is decidedly free market. The idea is that there would be a number of competitive offerings that would spur innovation and investment by the ISPs, which in turn would benefit the American people. As a result, the order added requirements that ISPs be transparent about commercial terms, network management practices, and performance. This would allow consumers to be informed about the plans and potential restrictions, and to choose what they want. The order also vests significant authority with the FTC to police privacy practices of the ISPs.
The discussion about whether broadband services are telecommunication services or information services has been both highly technical and highly emotional during the FCC rule-making process. The latest FCC order makes a fairly nuanced and technical argument that broadband is an information service, and their argument derives substantially from three parts: 1) the language of the relevant statutes actually supports this position, 2) the FCC has long held that broadband is an information service, and only at the behest of Obama in 2015 did it reverse course, while the internet did not fundamentally change during that time, and 3) its actual functionality more closely resembles an information service than a telecommunications service, based on the technical definitions of each (e.g. network-level caching, DNS lookups, etc v. time division multiplex access) - it does significantly more than simply move packets, and, at best, it makes use of telecommunications services, but provides significantly more as part of its offering. That said, popular culture is often not debating the issue at the technical level, but is instead thinking about the downstream impact that it may have on the internet, and the sort of relationship that end users ultimately want with their ISPs, and thus it has become emotional as well as technical.
One should note that that this matter has been an open question for a long time. There are already legal appeals challenging this decision and - as just happened - a different president in 2020 could result in an FCC that reverses this outcome. This far from a done deal - but, right now, it is the rule (that goes into effect next year, assuming all goes to plan). Legislation could also be drafted that, at any time in the future, simply makes a final call on this question and removes this decision from the hands of the FCC (the FCC is simply an executive agency operating under the authority granted to it by legislation to make these decisions). There was also some commotion around the potentially fraudulent anti-net neutrality comments that were submitted to the FCC during the public comments period, but that is largely beyond the substance of the issue because this ruling would have been the same without any such comments.
Why get rid of net neutrality?
Casting ISPs as Title II common carriers would effectively cause them to become public utilities. The notion of a public utility is that of an entity that has to invest such significant sums to provide its service, and that duplication of that investment is so ineffective, that its prices need to be regulated to avoid monopolistic practices while still ensuring profit. The reality is that these entities often stagnate, fail to operate effectively, fail to invest, and do not always serve the best interests of their customers. It is noted that 2015 - the year of Title II classification - was the first year since the 2009 recession that nationwide broadband investment declined, and the decline has continued year over year through the present. Meanwhile, DSL experienced a marked increase in investment when it was recategorized from a Title II to Title I service in 2005. Further, utility-style regulation may eventually be appropriate for ISPs, but given the dynamic nature of internet growth, assuming a certain static model of how they should work, and depressing incentives for further incentives, may actually undermine the overall growth of the internet.
Why do we need net neutrality?
Taken to its extreme, a world without net neutrality would be one in which internet gatekeepers (telcos, cable companies, etc) could decide which sites load quickly and work well, which do not, and which simply don't work.
There are concerns around free speech, and all that it entails. Speech here means anything on a website - so Comcast could block sites critical of Comcast, for example. But more generally, corporate speech could also be blocked. We previously discussed network level ad blocking, which is disallowed under net neutrality principles, but without net neutrality, ad providers could be required to pay for access. The same could be applied to any type of corporate or individual speech. More broadly, net neutrality provides equal access to information for all internet users - but this access may be undermined in the future. This might mean that wealthier people could simply access a greater percent of the internet, and thus have more access to more knowledge, than people would couldn't afford the "all-access" plans.
The challenges are perhaps most immediately obvious in a few tangible examples. In the case of Verizon, which owns Yahoo! and AOL, it could exclude its own services from data caps, and implicitly slow their competitors (by using low data caps that degrade performance after that cap is hit, but free access to its own services). Similarly, AT&T could make TimeWarner content available for free, and prioritized, while punishing rival services (directly addressed in this Lift Letter on their merger). This creates a world where content distributors and content owners no longer have a separation of concerns, and may incentivize vertically integrated content and distribution behemoths (it should be noted that any such prioritization would need to be clearly disclosed even without net neutrality). As discussed in prior Lift Letters - this outcome has mixed effects. Apple, Facebook, and Google have substantially monopolized internet eyeball distribution while, in effect, freeloading on the content creation and distribution created by other industries. Allowing the broadband providers to have more power in this relationship may (or may not) level the playing field, but it can certainly upend the existing dynamic.
T-Mobile and Netflix created a zero-rating deal, in which Netflix doesn't count against T-mobile's data cap. This is a clear violation of net neutrality, yet enforcement was not prioritized. Consumers may have liked their free Netflix, but it means that Hulu, Youtube, and any other potential competitors (e.g. Disney's upcoming service) were implicitly disadvantaged. They would likely need to pay some amount to T-mobile to get an equivalent deal. This would increase the cost of doing business for these content companies, which would indirectly get passed down to the customer. This would also mean that any potential newcomer would need to pay material fees to reach parity that they likely wouldn't be able to afford, precluding competition beyond incumbents.
Not long ago, Facebook released a platform called Free Basics. This was initially rolled out to provide free internet to rural India. The catch, of course, was that Free Basics effectively only included Facebook and a few other websites that Facebook chose to include. This was perhaps the most egregious net neutrality violation conceivable, and it was rejected by Indian regulators under a similar framework as net neutrality. Similar offerings could theoretically be made available without net neutrality that exclude vast portions of the internet altogether. That said, whether free-ish offerings for very limited internet is better than, effectively, no internet, is a matter of opinion.
How will this change ad tech?
This is a very difficult question to answer, and nobody really knows. Assuming net neutrality is fully rolled back, there are some potential outcomes that include:
1) Network-level ad blocking becomes legal - this would mean that companies like TripleLift would need to pay certain ISPs (most likely mobile ones) to bypass their content filters for ads - perhaps subsidizing the fees for the consumers. This is not a great outcome, but it could be one where more byte-efficient advertising proves more cost-efficient. It may potentially be the most likely.
2) Payment for speed - ad tech companies could pay to have their ads prioritized relative to others, especially in the mobile context, to prioritize delivery. This could potentially increase performance, and provide an opportunity to analyze the potential ROI of such an arrangement.
3) A certain set of apps like Facebook, Instagram, etc get zero-rated on various distribution channels, incentivizing users to remain on their favorite apps, potentially increasing the value of products like Instant Articles - significantly hurting the "independent" ecosystem, of which TripleLift is a part (unlikely...)
4) No change - the repeal acts the way Pai predicts and, by and large, things get better
5) No change - legal challenges delay the effective date of the rule until it's overturned by a democratic president's nominee in 2020.
6) No change - players like TripleLift largely fly under the radar of major distribution channels
But really, it's very challenging to say for sure. The main focus is assuredly with major content distribution channels and high-bandwidth activities like video. Video ads are an increasing part of the ad mix, but still pale in comparison to Netflix, Youtube and Facebook video consumption, which is likely going to be the initial focus. Ultimately, it adds unknown unknowns to our business, which we'd simply prefer to live without.