Layers of regulation: CPUC maintains grip on telecoms infrastructure


Kicking down barriers to competition, progressively.

There’s always a danger of reading too much into a single, seemingly routine decision by the California Public Utilities Commission, but I’ll risk it. Earlier this month, the CPUC granted a certificate of public convenience and necessity to Schat Communications LLC, which is a sister company of, an Internet service provider in eastern California. In doing so, the commission determined that Schat is a “telephone corporation” under Californian law and can be regulated as such.

Schat isn’t offering traditional voice or other consumer services that the commission is allowed to regulate. Its retail offerings are all Internet protocol (IP) based – i.e. VoIP and plain vanilla Internet access – which are specifically exempted from commission oversight by state law. In technical jargon, those belong to the network layer – Layer 3 – or higher of the standard Open Systems Interconnection (OSI) model, which provides clear and globally recognised definitions of the roles and the relationship between the technologies that make digital communications possible. The CPUC did not specifically reference the OSI model, but the language it used in the Schat decision usefully parallels those definitions.

The commission endorsed Schat’s assertion that physical infrastructure – Layer 1 in the OSI model – is something completely different from what state law considers IP services, which live on Layer 3 or higher (for brevity’s sake, I’m ignoring Layer 2, arguably a grey area)…

Schat Communications’ proposed middle-mile network will consist of conduits, ducts, poles, wires, cables and other property over which Schat Communications will offer transmission services to other certificated carriers (i.e., telephone corporations) as well as to business customers. Carriers purchasing Schat Communications’ transmission will then be able to offer telephone services to their customers. As such, Schat Communications’ middle mile network will be comprised of “telephone lines” and by such offering, Schat Communications will be a “telephone corporation” pursuant to the [Public Utilities] Code.

The commission is clearly saying that Layer 1 infrastructure is still within its regulatory domain, whether or not IP-based Layer 3 services are riding on it. If it follows this line of thinking to the logical conclusion, physical plant – all the way down to last mile copper – will remain regulated, even if telecoms companies completely convert to IP technology.

More broadly, the decision bridges a 100-year philosophical divide. The modern CPUC’s roots lie in California’s early 20th century Progressive Era, which brought regulation of natural monopolies and dismemberment of purposeful ones. Cyberspace may be limitless, but the physical dimensions of our world haven’t changed. The cost and difficulty of building public utility infrastructure – the defining hurdle of a natural monopoly – remain. By establishing a clear framework, based on objective technical standards, for keeping static physical infrastructure in a traditional regulatory regime, the commission is creating neutral ground where vigorous competition and the Internet’s libertarian ethos can flourish in the 21st century.

  • My first thought is that the FCC should really define what goes on at each layer. For instance, even in your discussion above it is not at all clear that the PSTN layer 3 switching is comparable to the IP layer 3.

    Then they need to understand the relationships and tradeoffs between layers as determined by not only density and value of traffic but also resulting from “monopolistic” bottlenecks that ripple across the InformationStack. Such things as bundled services, on-net vs off-net pricing, etc… all play a role in determining level of competition and open-ness.

    Importantly, the “vertical” boundary points–geographic, market segment, or application–defined by previous “arbitrary regulations” are blurred, constantly shifting or completely gone in a high-definition, two-way, mobile, collaborative world.

    It’s important to recognize that it’s not just IP vs PSTN, but the fact that most media wasn’t even digitized 20 years ago. So it starts with a basic understanding of what “digital” implies in terms of pricing, networks, consumption, etc… If the regulators can’t resolve the fact that a voice minute costs $0.0000008 to transport anywhere in the world (think of a standard def video session as using ~ 20x as much, which is still pretty low) with current last mile pricing, then we’ve got serious problems.

    Bottom-line is that the FCC needs an objective framework that models past, present and future consistently across rapidly depreciating supply (capex and opex) and constantly shifting and nearly infinite demand. See macro approach here: The natural conclusion, btw, is that a vertically integrated service provider model cannot be maintained; something a few in the industry are beginning to recognize.

    • If IP transition is successful, PSTN regulation becomes a moot point. So long as Layer 1 resources are available at something akin to market rates and/or makable with a reasonable amount of capital, Layer 3+ can be competitive enough to be left alone. Or at least treated no differently that any other consumer service.

      If everyone has a chance to play, it’s OK if some (a few) win and some (most) lose. That’s what capitalism is about. The only requirement is that there are sufficient winners at any one time to ensure that any given player, or wilful combinations of players, cannot control the market.

      • Ok, so where are the interconnect points in layer 1? At the curb? At the node? At the end of the middle mile? Or just 12 nationwide as Verizon is proposing and doing with VoIP providers to cut out middle mile and WAN transport providers? The number of questions in layer 1 alone are vast. Then there are layer 2 modulation and technology and operating issues: what standards, where are the electronics stored, what type of sharing of poles, COs, towers and customer premises (think MDUs and rooftops) exists?

        As for layer 3 (addressing and switching), well that is big. Number portability or resale? Who owns the number? Who owns IP address? What about email addresses? What about twitter handles? I think the regulator (and all stakeholders for that matter) need to, at a minimum, set guidelines at each layer and horizontal “zone” which can then be used in arbitration and adjudication over questions/issues/challenges regarding competition.

        As we get higher in the infostack I agree that regulatory and policy issues become less onerous as we’re seeing play out in the enterprise and smartphone application markets.

        What it comes down to is trying to understand or assess how many players at each layer are required to keep competitive forces vibrant. Furthermore, how do regulators (federal and state working together) monitor pricing so that it reflects marginal cost at each boundary point (layers and horizontals).

        Marginal cost (and hence pricing) is set ex ante in the market by numerous demand/supply assumption iterations and should, on balance follow a trajectory reflecting a combination of moore’s and metcalfe’s laws on a cost per bit basis. This issue and outcome is rarely if ever discussed, let alone understood by 99% of the industry stakeholders. Ask the FCC to produce a study on the economics of Google Fiber or 802.11n or ac.

  • Pingback: CA: Layers of regulation: CPUC maintains grip o...()

  • Pingback: Broadband infrastructure can – and sometimes will – be regulated in California | Eastern Sierra Connect Regional Broadband Consortium()