Home / Blogs

Let American Telcos and Cablecos Merge - But Declare Infrastructure a Utility

While there is much discussion in the United States about the mergers of Comcast and Time Warner Cable, and of AT&T and DirectTV, issues such as this are generally discussed from a very narrow perspective and, we maintain, from the wrong underlying telecoms regime operating in that country—one that has stifled competition in the telecoms for nearly two decades.

The same wrong parameters apply to the endless debates on net neutrality an issue that is, by the way, largely of significance to the US market alone. Without wholesale competition net neutrality is a far more serious issue: with proper competition the issue will mostly sort itself out.

The irony is that the way these issues have been addressed in other developed economies is based on lessons learned from America several decades ago: ‘Let the market sort this out through competition’.

Over the last few decades competition in the American telecoms market has weakened, while the opposite has happened in most other developed economies.

We have now reached a situation globally where it has become clear that over the next few decades the old copper telco and HFC cableco networks will need to be replaced with deep fibre—and eventually all the way to premises with FttP infrastructure. When that point is reached it will not make economic sense for America (or any other country for that matter) to have a collection of competing FttP networks. At that stage it will be clear that these networks are another utility, just like water, gas and electricity.

The construction of such networks is very expensive, and in the larger countries will take 10 to 20 years to complete. Planning and design are necessary before the end results become visible. So while in many cases FttP may still be a decade or more away, now is the time to make the strategic decisions about these investments. Since this means that there will be a looming monopoly in infrastructure, it also means that government policy decisions will need to be made now to ensure that the market’s proper functioning will be safeguarded during the process.

As has been shown by the net neutrality and recent merger proposals, the network operators involved can no longer maintain their current business models. ARPU for many operators is falling, which is having a detrimental effect on infrastructure investments which need to increase. So, as we are seeing in other sectors of the economy, costs need to be taken out of many of the existing old-world business models. The same economic transformation is needed in the telecoms industry, and here also network operators will need to change their old business models. Instead, they are trying to avoid doing so through merger and acquisition activity. This process will indeed take costs out of their models, but it will lead to monopolies being formed.

That being the market reality, it is necessary to look at the effect this will have on other elements of the market, such as competition, innovation and customer services. It is well established that under monopolistic conditions all of these three elements will suffer, so we need to find ways to counteract the negatives of what constitutestoday’s market reality.

The solution is not too hard to find, however. It is to declare the infrastructure a utility. Yet it is in the implementation that the real problem lies: given that these mergers will lead to a monopolistic market situation, appropriate government policies and regulations will need to be put in place to counteract those negatives.

The positive in all of this is that when looking at the new digital economy developments it is evident that real innovation, competition and customer services are already largely taking place on top of utility infrastructure. So the obvious solution is to ensure that this level of competition can take place on top of that infrastructure.

Why is it so difficult to facilitate these changes, since most people would consider that it makes sense?

The problem is that over the last two decades the various American governments have taken their eye off the ball and let the key network operators write the rules and regulations that best suit them. The government and the FCC have been made captive of the highly successful telcos’ lobby (tellingly, these telcos now call themselves ISPs simply to confuse the issue and avoid regulation). Telcos have become so powerful that they believe they are too big to be subjected to government policies that could turn them into utilities.

This is exactly the battle that is now taking place in relation to all of the key telecoms decisions that need to be taken in Washington.

As all the issues have been turned into a political and regulatory game by the operators, the reaction to this will have to be similar.

If the operators want to offer special services to content providers—contrary to the principles of net neutrality—and if they want to consolidate further, they should be able to do so as long as they separate access to their infrastructure from the services they offer. With that infrastructure rapidly becoming a monopoly (especially at regional levels) access to it should be regulated. With open, equal and transparent access and interconnect rules, whatever the incumbent players and others want to do on top of that network could largely be left unregulated.

And from here the message will be the same as the one that America has been sending out to the world for over a century: may the best ones win.

By Paul Budde, Managing Director of Paul Budde Communication

Paul is also a contributor of the Paul Budde Communication blog located here.

Visit Page

Filed Under


One System, It works? Rick  –  Jun 4, 2014 7:21 AM

And the end result, of course is a single large communications monopoly. Maybe they will call it the “Bell” system because it rings a bell in the back of some 30 year old memories…?

Monopoly Paul Budde  –  Jun 4, 2014 7:29 AM

As long if it is an open, wholesale-only monopoly I don’t have a problems with that Rick. At the same I fully realize that we have a long way to go. Current developments are opening an opportunity for change. However, if we don’t act we will get a fully vertically integrated monopoly and that would be far worse.

Re: Monopoly Rick  –  Jun 4, 2014 9:20 AM

I agree with you whole heartedly, Paul. In the interest of full disclosure: I'm an old Bell Labs alumnus. I do appreciate the benefits that competition has brought to the telecom world over the last 30 years, but I mourn the loss of the great center of learning and research that was the old Bell Labs. And similarly, I can only hope that a return to regulated monopoly would revive the public service ethic we had in the old Bell system.

monopoly Paul Budde  –  Jun 4, 2014 10:36 AM

It is just so sad that something like this is hardly discusable in America as this is seen as government interfering in the marketplace, ignoring basically the private monopolies of the incumbents.It also looks like the court system is stepping into the political vacuum and - dominated by rather conservative forces-  pushing the national good and the national interest further and further away.

The future is multifaceted Michael Elling  –  Jun 6, 2014 2:52 PM

Another way of looking at this is that we’re in the final phase of a process begun 30 years ago, when the US started down the path to this wonderful new mobile, digital, high-definition future with the vertical or logical separation of the telco monopoly.  We have not seen sustained (and successful) complete structural separation here, nor for that matter anywhere else in the world where the lower, middle and upper layers have been completely disintermediated in the last mile or edge.

To get to a ubiquitous and low-cost 4/8K VoD, seamless mobile BB, 2-way HD video collaboration, and IoT future we will need to see precisely that happen.  No govt or central authority can assure a satisfactory implementation of and outcome to capacity (particularly upstream), latency, QoS, security, privacy, redundancy, etc… cost and performance issues in the last mile.  Only a competitive marketplace can.

But we’ve seen that mandated interconnect policies have developed low-cost, highly scalable and rapidly achieved competitive market solutions over the past 30 years.  There is no reason that a modified Title II (interconnect applied multilaterally and simultaneously in layers 1-2 in COs, head-ends, pedestals, poles/towers, and access to buildings/towns) along with prioritization and fast lanes couldn’t provide the optimal outcome.  A win-win compromise for the two opposing sides.

In the process the cloud would move rapidly to the edge. Current under-utilized assets would be valued as “beachfront” property when it comes to the above performance and cost issues with respect to clearing supply and demand both north-south between infrastructure and application/content layers and east-west between agents.  Moreover, we’d open up an infinitely large revenue opportunity for centralized procurement of edge access.  Think of managed VPNs for large enterprises and institutions across numerous boundaries and networks at the edge.  The IP stack cannot do this with it’s flat-rate, bill and keep approach.  Time for compromise and implementing in the last mile what has proven to work in competitive core (WAN) voice markets, data, and mobile.

Multifaced Paul Budde  –  Jun 6, 2014 10:48 PM

Good points Michael and I certainly agree with you. I call these the ‘value added elements of infrastructure’ and there are both great developments here and stunning innovation.
I should perhaps have been a bit more precise. I was talking about the fixed ‘pipe’ I can’t see two or more fiber networks being rolled out to each house or each street. That’s where I see the monopoly element. In Australia, New Zealand, Singapore, UK and parts of France you see various flavors of structural separation being implemented. In Scandinavia that is done in a more voluntary ‘social’ way. Most of Europe does have proper wholesale arrangements in place that also leads to a similar outcome: sharing of that basic infrastructure. The US is the odd one out here.

Sharing of facilities Michael Elling  –  Jun 8, 2014 3:58 PM

Paul, there is no layer 1 last mile monopoly given what we know today.

This was an artificial fiction created by AT&T;and other information monopolies in the first half of the 1900s in collusion with an ignorant (or paid for) regulator.

The 3 key factors that will drive multiple, generative, low-cost, last mile solutions will be:
1) hetnets, frequency reuse, and customer mobility
2) centralized procurement will drive need for redundancy, backup and asynchronous layer 1 solutions splitting up and downstream
3) the last couple hundred feet will be part of the real estate costs.  We’ve already “REIT"ed the core and part of the edge (towers).  The next step is the CO’s, poles and drops between home/premise and curb/pedestal. 

You are thinking in fixed, edge subscription terms.  Not mobile, centrally procured terms, where the end-user will expect multi-screen flexibility ubiquitously and securely.  Nor are you taking into account network effect and how rural demand (or small market/small country demand) is naturally influenced and driven by urban or large market demand (and vice versa). 

Think of the infinite demand developing out of the above 4 trends I point to which are here today and will scale rapidly with interconnect near the edge in layers 1-2 and balanced settlements.  These are all solutions that dramatically compress time and space and digitize everything we know in the real world.  The past 30 years will look like child’s play in comparison and result in multiple fiber networks being built out to the edge.

Hetnets, etc Paul Budde  –  Jun 9, 2014 8:30 AM

Very interesting scenarios Michael and these technologies/scenarios are indeed emerging. However, I am not sure if this will develop along competitive lines. You are correct I was referring to the fixed lines (as I mentioned in the article). For the foreseeable future I still see fixed an mobile as separate developments. In whatever way you turn this, high use of rich media over mobile networks remains problematic (spectrum, end user costs). Fiber will go deeper and deeper into the neighborhoods and streets (but I can’t see multiple providers doing that). At a certain stage you could indeed see that mobile could be used for the last few hundred feet – basically a mobile station in each street –linked back to fiber. However, I wouldn’t be surprised that if indeed mobile access from the street is going to be developed – something that will take considerable time – that before that happens, the fixed operators will have signed up many customers directly to their fiber network (at the same time some of the incumbents already operate both infrastructures and will simply merge this). Within the current regulatory framework I can’t see new players entering this infrastructure play, so we are stuck with the forever merging smaller number of (vertically integrated) infrastructure players that currently dominate this market (both fixed and mobile) and therefore I remain concerned about monopolistic behavior that would hamper the sort of developments that you describe in any competitive way. The current players will most certainly use those technologies but that will all remain within their own market domination. Anybody interested in pursuing your scenarios on any large scale will hit the brick wall of the incumbents and under the current regulations they can dictate access/interoperability and they will decide if others can use them and how they can use it.

Once again I have no issue with the scenarios as you mention them, my worry is that I can’t see this being deployed in any sort of open network context, as the regulations stand,  these developments will be dominated by the incumbents.


Centralized procurement Michael Elling  –  Jun 9, 2014 11:16 AM


Completely missed is the point about centralized procurement.  This comes via “balanced settlements” or “prioritization and fast lanes”.  The IP stack lacks price signals and incentives to clear (marginal) supply and demand north-south and east-west in the stack.  Spam, DNS, IPv6 upgrade, etc… are all resolved by balanced settlements serving as price signals.

(I am working on a model where net neutrality (which was a convenient fiction created by competitive apologists unwilling to admit to the IP stack’s policy warts and commercial origins out of the competitive PSTN core in response to the remonopolization of the US market) is given up in exchange for interconnection out towards the edge in layers 1 and 2.  I believe the latter applied evenly to COs, headends, pedestals and poles is sufficient to move the competitive cloud to the edge.  The full weight of core supply/demand will clear the highly marginal and variable demand at the edge.  I would welcome your input on this compromise approach.  I do not believe full structural separation is possible, wise or even necessary.  Nor do I believe the govt has ever been good at setting price levels.  But it has been good at mandating physical and logical interconnect: equal access, dial-1, # portability, RF constraints, etc…  The markets develop the efficient price and business models around these policies.)

Not only do balanced settlements lead to value conveyance driving network effect, they lead to coordinated and virtuous infrastructure investment and most importantly edge procurement or subsidization from the core. Large institutions and enterprises pay for secure, committed information rates (CIRs) for telework, telemedicine, teleeducation, tele-sensing, etc…

The latter is infinite in terms of revenue opportunity as the entire economy becomes digital.  Networks become the most efficient way of promulgating transactions; B2B, B2C, C2C.  Sessions driven by or married to transactions.  The commercial foundations for balanced settlements are already developing in the world of “big data” where the markets are matching preferences and anticipating intent to real-time ad serving.  Aren’t these very similar to the same issues surrounding mediation/billing and congestion control?

We had a taste of centralized procurement with 800 and VPN in the US, but the latter were swamped by the privacy invading advertising model as no settlement systems existed to pay for webs 1.0, 2.0, and to some degree 3.0 (mobile apps).

Speaking of which, I think you have to speed up your assumptions on the mobile consumption front.  Right now we can look at the mobile model two ways (and this is from the perspective of the US market, which is the most advanced large 4G/LTE market).  First is we pay for 1 dollar of service and receive only 30 cents of real value or performance due to the 70% user generated/driven offload.  Compounding this is the fact that 5 year wireless TCO is now 13% capex and 87% opex.  About half of the latter is due to backhaul and site rental.  Clearly neither are sustainable business models.  What if the cable cos start charging for offload (transit) under the new FCC rules?

But at the same time, people’s consumption of video over mobile is skyrocketing.  This is principally due to the cloud and video on demand.  People are consuming long-form video in small snippets; smoothing out demand and consumption across the entire day.  As well, they are increasingly watching live-streaming on the go or at work.  The result is the mobile device is becoming the first and sometimes even primary screen for a large portion of the population.  This will be an even greater trend in developing markets; contrary to people’s expectation of “expensive mobile”.  And then of course 2-way video over mobile will develop, which no one has a solution for in terms of upstream capacity and latency.

We see that wireless is already highly dependent on fixed and therefore less independent of each other than most, and yourself, perceive.  Putting the above issues together we will see fiber quickly driven to the edge by demand and competition (fixed and wireless). And the process will not stop as evolution of the individual’s and overall economy’s propensity to consume sessions will be endless.

These are the lessons learned from the competitive digitization of the 1980s-90s.

Rethinking our conversation Paul Budde  –  Jun 11, 2014 1:55 AM

Michael it looks like we will either not agree or we are talking past each other. I have gone back over our previous emails to rethink the issues.

I am not quite sure if I understand what you mean by ‘centralized procurement”, but I twill try to respond by addressing the points backwards from our conversation.

I understand that you argue that a commercial model based on ‘balanced settlements’ of charges for transactions or services will fix everything - providing capital for building out last-mile infrastructure (as per my previous comments, ignoring here the distinction between fiber and wireless for the moment) and ‘back-pressure’ to limit congestion and over-use. And that ‘demand’ (not sure for what?) is “infinite”.

However the ‘balanced settlements’ system of payments between international carriers up to the 80s was shown to be a fragile house of cards that progressively collapsed when operators were bypassed by alternative lower-cost operators and infrastructure owners who didn’t follow the same gentlemen’s accounting rules. That horse has bolted. There is no reason to suspect a similar accounting regime, foisted onto the last-mile access network operators, would fare any differently - the moment one competitor network offered an alternative charging model that resulted in lower charges for a particular form of traffic, or a fundamentally incompatible technology or service structure everyone would be forced to follow suit or lose market share, and the system would collapses again - it is inherently unstable commercially.

In your June 06 post you say “There is no reason that a modified Title II (interconnect applied multilaterally and simultaneously in layers 1-2 in COs, head-ends, pedestals, poles/towers, and access to buildings/towns) along with prioritization and fast lanes couldn’t provide the optimal outcome”. I don’t believe this is correct - interconnect can be mandated to be provided at any location, but if it isn’t commercially viable for a competitor to build out to those locations, then interconnect won’t happen in practice. This is why interconnect rarely occurs at pedestals and poles and other locations deep in the access network - the economies of scale for the newcomer of building fresh new infrastructure deep into the network to service smaller and smaller groups of potential customers doesn’t work as well as interconnection further back, at the CO, or the switching center, or at the extreme example a couple of nationally distributed interconnect points. It relies on a regulator setting a sensible regulated price for the interconnect facility - something that rarely works satisfactorily, and usually takes many years to approach. It also relies on the underlying network ‘playing nice’ and providing equivalent performance for the service to the interconnect point - involving more regulation, and more lawyers.

The argument about increasing demand for new services, more video, two-way video etc is well made - but increasing demand doesn’t necessarily result in more infrastructure being built to service that demand, unless the organisation laying out the capital to build the increased capacity - or a parallel competitive network - can get a adequate return on their investment. This is not a given - if the public have increased demand, but aren’t willing or able to pay any more, then the result is frustrated customers, not new or upgraded infrastructure. When you say “the individual’s and overall economy’s propensity to consume sessions will be endless” and “infinite in terms of revenue opportunity as the entire economy becomes digital”, I cannot see where you take account of the fact that the individual or the economy’s wallet-capacity is not endless or infinite, and the increasing volume of traffic will need to be carried for little increased revenue - or in fact decreasing revenue, as the competitive model relies on each operator attempting to provide more value for the same money, or the same value for less money, extracted from the customer to win and keep the customer.

A competitive infrastructure won’t be built unless the new operator believes it can capture enough customers from the existing operator to pay for the new infrastructure - customers that are accustomed to paying less each time they switch providers - and this is less and less likely the more regional/rural the area. Every last-mile network - fixed or wireless - needs a threshold scale of customers connected to be commercially viable, and in the worst case a duplicated competitive last mile will see both (or each) network become sub-scale until enough competitor networks collapse that the remaining network(s) achieve sufficient scale again - a highly inefficient waste of capital, and undesirable from a customer experience. In any case it rarely comes to that - competitor proposals can do their own modelling, and if there isn’t at least a reasonably surety of achieving profitable scale and good return on capital outlay, the competitor network will never be built - leaving a very real layer 1 last mile monopoly in that area.

So IMHO my issue still stands - how, in areas with such a layer 1 monopoly, do we as a society ensure the customers in that area can still access the same innovation, capacity growth and advanced services as might be offered in other areas where competition is viable and flourishing, and where service providers need to compete and innovate to retain customers?

The reality of the current commercial operating model is that increasing customer demand is not, in itself, sufficient for the operator to invest funds to satisfy the demand. Only where there is a real threat of the customer switching away to a different provider to have the demand satisfied will the commercial operator invest to satisfy the demand itself.


Paul,Centralized procurement (like the 800 model) is Michael Elling  –  Jun 19, 2014 3:35 PM


Centralized procurement (like the 800 model) is a managed service VPN that crosses boundaries (fixed and mobile) that provides a consistent bandwidth (CIR) and QoS. Few people have a true appreciation for this model because of the lack of vertical (or logical) separation models outside the US in the 1980s-90s. 

Examples would be 2-way HD telework, tele-education, tele-health, where the end-user’s access is subsidized by the central organization.  Costs would be embedded in costs of doing business, or tuition, or premiums, or subscription fees, etc…

Example: AT&T;goes to Johnson & Johnson and offer a 100,000 endpoint HD video collaboration solution that guarantees 20-mbs synchronous for $20/end point might be an enormous tradeoff between real-world costs (real-estate, commuting, productivity) and network costs. This requires AT&T;to get out of its layer 1 silo and move traffic freely across other providers (fixed and mobile).

In addition to being affordable, it has to be high capacity and high quality (50-70 inch screens to replicate “being there”) and high reliability (4-5 9s; no jitter), and likely redundant/backup facilities, to ensure broad-based uptake and adoption.  Think of AT&T;having an agreement with a 2nd wired or wireless facility for $1-5 per month as a backup.

These can be corporate or institutional intra or extranets.  They can also be on-demand.

Multiple VPNs into a home or premise can increase revenue per endpoint at higher margin (lower marketing, churn and support costs) than the current ad-hoc, subscription-based revenue model only.  Also, it will drive faster and more consistent upgrades of the edge.  What happens when we go to 4K video collaboration; wait it’s already here!  Or maybe holography or VR?

There is effectively no end to the tradeoffs between the current physical/analog world and the new virtual/digital, networked world.

The regime of bilateral settlements was byzantine and fraught with inefficient subsidies.  Market driven pricing that clears supply and demand north-south and east-west are the only way to go.  Bill and Keep provides no price signals or incentives.

Read my blog at ivpcapital dot com / blog.  It has a broad range of discussions of historical pricing and cost structures as well as discussion of new settlement models.

I am presently working on an analysis of the cost/performance (and hence pricing) requirements for capacity (particularly upstream), latency, QoS, security, and redundancy for 4K VoD, 2-way HD collaboration, seamless mobile BB, and the internet of things where each user is a node for “n” processors and sensors.


Thank you Paul Budde  –  Jun 20, 2014 10:19 AM

Thank you Michael for the extra valuable information.
Much appreciated.


Centralized procurement Michael Elling  –  Jun 20, 2014 2:31 PM


You asked the question.  I responded.  Your response smacks of sarcasm, instead of discussing the merits of centralized procurement.  That’s unfortunate.

But then 99% of the market didn’t forsee the impacts of divestiture in 1983, nor the opportunity for data in 1990, nor the impact of 10 cent digital wireless in 1996, or the potential demand elasticity with open (data) access in the cellphone (aka layer 2/wifi offload in the smartphone).


Michael we have to accept that we Paul Budde  –  Jun 20, 2014 11:05 PM

Michael we have to accept that we think differently about this. There is nothing wrong or sarcastic about that.

I wonder how the second-tier access line provider will feel about expending the capital to upgrade their network to provide 5-9s high speed symmetric 4K & VR capable links with guaranteed CIR and QoS for $1-5/month wholesale revenue. But perhaps you are right and that this will happen.

Thanks for having this discussion with me.


Future revenue models Michael Elling  –  Jun 22, 2014 6:18 PM


I didn’t say $1-5/month.

I said $20 per month for a MSP intra/extranet.

An extra $1-5 might be backup to another provider; ensuring, along with mobility drivers, that there will be multiple last mile layer 1 competitors.

Multiply this across “n” possible “applications” per user and hence premise.

This is consistent with the demand elasticity we saw with competitive voice in the 1980s-90s and data (aka internet) applications in the 1990s-2000s.  Over the past 10 years edge-users bought smartphones and broadband to access apps.  I call this “core-driven” demand.  Going forward we need to see centralized buyers have the opportunity to pay for the access, while the access becomes more competitive farther out to the edge.

Where we differ is our perspective on final demand, which today is limited by high cost/bit and lack of market driven balanced settlements (not reciprocal comp or 2-sided takings; which are both tarnished terms and not what I believe will serve as needed price signals.).


Comment Title:

  Notify me of follow-up comments

We encourage you to post comments and engage in discussions that advance this post through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can report it using the link at the end of each comment. Views expressed in the comments do not represent those of CircleID. For more information on our comment policy, see Codes of Conduct.

CircleID Newsletter The Weekly Wrap

More and more professionals are choosing to publish critical posts on CircleID from all corners of the Internet industry. If you find it hard to keep up daily, consider subscribing to our weekly digest. We will provide you a convenient summary report once a week sent directly to your inbox. It's a quick and easy read.

I make a point of reading CircleID. There is no getting around the utility of knowing what thoughtful people are thinking and saying about our industry.

Co-designer of the TCP/IP Protocols & the Architecture of the Internet



Brand Protection

Sponsored byCSC


Sponsored byVerisign


Sponsored byDNIB.com

Threat Intelligence

Sponsored byWhoisXML API

New TLDs

Sponsored byRadix

IPv4 Markets

Sponsored byIPv4.Global

Domain Names

Sponsored byVerisign