Fast, affordable Internet access for all.
Transcript: Open Access and Incumbent Challenges Podcast Episode 128
Thanks to Jeff Hoel for providing the transcript for Episode 128 of the Community Broadband Bits podcast with Eric Lampland of Lookout Point Communications. Listen to this episode here. 00:05:
Eric Lampland: But the monopoly is created by the physical infrastructure ...
00:08:
Chris Mitchell: Exactly.
00:08:
Eric: ... of those particular things, and so, yes, that does need to be replaced.
00:15:
Lisa Gonzalez: Hello. You are listening to the Community Broadband Bits Podcast, from the Institute for Local Self-Reliance. My name is Lisa Gonzalez. In episode 80 of the podcast, we introduced you to Eric Lampland, Founder and Principal Consultant at Lookout Point Communications. In that episode, Chris and Eric discussed indirect cost savings from municipal networks. This week, Eric stopped by our Minneapolis office for a discussion about "open access," today and in the future. We find that many of the communities we study that deploy or consider municipal broadband networks see an open access arrangement as the preferred business model. Traditionally, this would mean that the municipality would provide the infrastructure and providers would offer commercial or residential services to customers over the infrastructure. We also find that a large percentage of those communities find it difficult to implement this model. In our conversation, we dig into some common challenges associated with municipal open access networks. We also talk about some possible cures, and we look at how the very definition of the term "open access" is changing. Every week, we bring you the Community Broadband Bits Podcast advertisement-free. Please consider contributing in any amount to help us in carrying on this service. Just visit ILSR.org and click on the orange "donate" button.
01:38:
Chris Mitchell: Welcome to another edition of the Community Broadband Bits Podcast. I'm Chris Mitchell. Today, I'm in the office with Lisa Gonzalez.
01:47:
Lisa: Hey there.
01:47:
Chris: And we've got Eric Lampland back, the Founder and Principal Consultant for Lookout Point Communications.
01:53:
Eric: It's good to be here with you, Chris.
01:55:
Chris: Thank you for coming back in, Eric. We had a great conversation about two weeks ago, and I fouled up the recording. So we're going to attempt to recreate that. And I really appreciate your patience, and coming back in to, once again, discuss "open access" with me.
02:10:
Eric: We've discussed open access, I think, for about ten years, so doing this a couple of times is not an issue.
02:16:
Chris: Yes. We've been talking about it a long time. It's been a passion of yours, which is why I wanted to have you in here. And you've worked with a number of communities that have, I think, followed a similar trend that we've seen in the industry, which is that most communities fundamentally would prefer to build the infrastructure but not engage in the service competition with a Comcast or an AT&T or someone like that. Even through most cities would prefer not to engage in that level of competition, we find that most cities ultimately DO end up engaging in that level of competition. I was hoping that you would help us to sort that out a little bit. Why is it that cities that would prefer open access ultimately often find they can't make it work?
02:55:
Eric: I think the simple reason is that it's very difficult to get legacy providers to operate their services across other people's infrastructures. And so there hasn't been an option not to produce your own retail services.
03:13:
Chris: So, you and I are in St. Paul. And if St. Paul built a fiber optic network -- in this mythical future where we have a mayor that cares about the future of our community --
03:25: [laughter]
03:26: Chris: not that I feel personally about this -- what you're identifying as the problem is that the -- Comcast and Centurylink would refuse to use a fiber optic network that was even superior to what they had in the ground. They're the legacy providers that you're identifying.
03:43:
Eric: Right. Well, the intention, the old business model for service providers is a simple one, and that is that they own the customer by owning the connection to the home or to the business. And when you take that business model, and you take away the connection, that doesn't reflect well on the way they've set their business models.
04:02: Chris: So, when you're talking about the legacy providers, you're talking about Comcast and Centurylink. And they just don't have an interest in using an even superior fiber optic network a city may build.
04:14: Eric: They have an interest in using a superior fiber optic network, if they built it.
04:20: Chris: Right.
04:20: Eric: But they don't have an interest in using somebody else's.
04:24:
Chris: Right. And I think one of the reasons for that is, these are companies that are fundamentally built on a business model that was not one that was existing in a competitive environment. And that, frankly, it would be difficult for them to suddenly change tens of thousands of employees to be operating in a competitive environment.
04:42:
Eric: I don't want to speak directly to why -- to what the change would be for legacy providers, because I don't think they know it. I think it's rather simply that they're comfortable doing what they do.
04:57: Chris: Um hum.
04:57:
Eric: And they're not comfortable entering into a different monetization approach which they do not understand.
05:06:
Chris: The result of that is -- regardless of what their reasons are, and I think you're absolutely correct, that they probably don't even know -- I mean, I'm always worried about giving too much agency to massive bureaucratic organizations.
05:17: Eric: Right.
05:17: Chris: They're -- they don't have cohesive set of interests. They have competing interests that are fighting within their own bureaucracy. So, let's leave that to the side. And let's look at what that means for a city, then. So, a city that wants to build this network, because they know that the legacy operators won't operate on it -- I think that changes the approach that a city can use. You can't just go out and borrow $100 million -- or $300 million if you're a big city -- to build a network if you are not going to have significant buy-in from the legacy operators and you want to use an open access system. Is that right?
05:51: Eric: Well, that is correct. There are actually multiple models in which a municipality can pay for a fiber infrastructure today
05:59: Chris: OK.
05:59: Eric: One of those models -- and the traditional approach -- is to pay for them out of retail services. And within that model, and as a subset of that model, open access has been thought to solve certain other kinds of political problems.
06:17: Chris: OK, so we have -- now, this is open access as a solution to a political problem.
06:23: Eric: Correct.
06:23: Chris: What is that political problem?
06:24: Eric: The political problem is that many governments do not feel that they are in a position to compete with the private sector. And that they would like to enable the private sector. It's interesting that in this particular section of the economy, that particular point of view is maintained, whereas in other sectors of the economy, we've passed that view. There was a long time ago when sewer systems were private, when electrical systems were completely, you know, individually competitive.
07:00: Chris: Um hum.
07:00:
Eric: And we can look back on, you know, even roadways that do that type of thing. But the truth is that we as a society have moved forward to a point where cities are very comfortable in providing roads and sewers and water and so forth and so on. But in this sector, they have not yet made that transition.
07:21:
Chris: Right. And I think it remains to be seen whether we'll live in a future where, 10-15 years down the road, we'll have a monopoly regime, saying -- recognizing that fiber optic networks are a natural monopoly, and that there is no real viable way to have competition when you have private ownership or -- this or that. I, frankly, don't actually think we're going to get there, for broadband and Internet-type services. But I think it's an interesting conversation. But one thing I want to note is that even we've started on how open access is very difficult, we have seen cities that have figured out ways of moving forward with it. We've seen Danville. We've seen Palm Coast. FastRoads. There's all those public utility districts in Washington state. Many of them have built in an incremental-type fashion. The Washington state ones have not found a way, in all cases, of paying down their debt. And many of those have wholesale power sources of revenue. So they have extra money floating around. And they're used to cross-subsidizing for, you know, sewer. So they may not have an issue with building some of this infrastructure. But in places that absolutely want the network to pay for itself -- where you want only the revenues of the network to pay for the costs of having built the network -- it seems like the most viable model has been this slow, incremental expansion.
08:45:
Eric: Well, I'm not so sure I agree with that. But -- And I think that there's something that's important that's implied in your comment. And that is, what is the definition of "open access"? In the cases of Danville and various different places that you mentioned, the approach to open access has to either do with the selling of components of the physical infrastructure, such as fiber optics -- dark fiber, we call it -- or the allowance of people to use exclusively some part of the infrastructure. That's a very limited view of open access, in my opinion.
09:31: Chris: So, let me just push this out a little bit. You and I were recently in Salt Lake City.
09:35: Eric: Yes.
09:35:
Chris: And this is something that I've heard regularly with criticism with regard to UTOPIA, which is an open access network. But, for the most part, you can only buy vanilla from different people. Right? You don't have 31 flavors. You really have, you know, a whole bunch of different companies that are selling one flavor. And I think that's what you're getting at here. Is, you have a different vision for how open access could work -- and, you think, should work.
09:59:
Eric: There are multiple aspects of this. I think the large piece, which I won't address right away, is the perspective of open access as we see it in places around the country now is really an orientation to providers. It is how a provider can use that access to obtain customers. But once obtained, those customers belong to that provider. So it's really a -- it's a bit of the same version of what we have today. The other side of that is a customer-oriented open access, in which the customer has the ability to elect the choice of provider, and may, indeed, have multiple providers at the same location. That is not -- we haven't seen that type of open access yet develop.
10:54:
Chris: But when you say services and providers, what I'm confused about is, you know, in UTOPIA, you have a choice of many different providers.
11:04: Eric: Right.
11:04:
Chris: Now, they're all selling pretty much the same thing. Now, you question is basically, you know, is, how can we set this up so that those service providers actually have differentiated offerings?
11:15:
Eric: No, what I'm really saying is that, you know, in normal marketplaces that we have -- if you're buying jeans, for example, right? -- you have a lot of people who sell jeans. But they have found ways to differentiate their products -- this jean from that jean -- it might be the design on the back pocket or it might be the fit, however it's so determined. There's no reason in the world that video, for example -- video services, entertainment services -- can't find similar differentiations. And when they find those differentiations, then open access within the video service arena would be something that would probably we welcomed by many people. Today, unfortunately, we still are stuck with the fact that -- in the case that we're making an example of here -- video services -- that video services are thought to have some hundreds of channels, most of which you don't watch, and has a fairly large price, and as long as you have two competitors that are doing exactly the same thing, then you're right, there's no differentiation of service and there's no benefit to the customer or the provider. So, the notion on services is really kind of twofold when you get into open access. The first is, do we have competitive understandings -- or, do we have differentiated services between providers of a given aspect of service -- such as video, such as voice, such as various kinds of healthcare services and so forth? The second piece of that comes into play in that most people are making conversations about what has traditionally been thought of as communications services. That thing that we call "triple-play" -- voice, video, and data. But in fiber optic networks, we're talking about orders of magnitude in capacity that are greater in fiber than they are in the copper infrastructures of today. And so the types of services that will begin to occur in an open access network are much greater than just voice, video, and data. Certainly, some providers, like Comcast, have actually included things like home security, right? Well, home security is an easy extra service, right?
13:37: Chris: Um hum.
13:37:
Eric: But healthcare is an easy extra service, right? And if you wanted to do something with your fitbit, you know, wrist, and somebody had a service that amalgamated that with your trainer at the local gym, that would be a different service. Right?
13:56:
Chris: Right. However, what I think -- when I hear you say that, my thought is, well, what you're describing is things that are data.
14:04: Eric: Hmm.
14:04:
Chris: And so, you know, at home right now, I have Comcast. And I could have a home security system riding over my Comcast connection. And I can do some home-healthcare-type stuff over my Comcast data connection. So how would you vision of open access differ from my idea, which is that I have a data connection and I can do a hundred different things over that?
14:26:
Eric: Well, I think that's kind of interesting that you think of the data connection, OK, as something where you can do hundreds of things over it, but you think of a fiber optic network as where you can only do three things over it. The fact of the matter is, everything we do over networks today is a data connection, whether it's voice, video, or these things that we think of as Internet. Right? If, for example, I had a service that I offered out of my local gym, that was your or my -- preferably me -- physical training activity, and that the service that was offered actually included tracking my physical activity during the course of the day, that brought it back to my trainer at the gym, and so forth and so on. Could I run that as a web-initiated-type service? Well, I could. OK? But when I think, in fact, that I've got a local function here, going through the expense of a web service and a local level may not be the best way to do that service. There may be many alternatives about doing that kind of delivery of service. So, you know, when we think in terms of a web application, we think in terms of a large-scale aggregation of clients; when we think in terms of a local application, we think of a smaller group of clients, with a more focused need. The range of services is much greater than we currently generally think about. 16:00: Lisa: So, where do you feel that we are in terms of these differentiated services now? I mean, how close are we to getting that type of thing now?
16:09:
Eric: It's emerging rapidly. And it will -- you know, one of the things that you have to be careful about -- that I have to be careful about, and I've been in this business for a long, long time -- and a long, long time ago, we didn't foresee many of the web services that we have available today. And today we would only enjoy the wonder of these emerging services when they actually occur.
16:36: Lisa: Um hum.
16:36: Chris: You didn't foresee the Yo app ...
16:39: [laughter]
16:40:
Eric: I didn't foresee a LOT. But healthcare services, for example, that involve children at schools, your own doctor, the specialist that might service your family or your school, the parents that might get involved, is a very localized healthcare service function that is not a web-type-based orientation. So there are many things that could come forward. Farmers in the field who measure their water in their underground wells to determine what kind of irrigation -- is a local service. It's not something that you do at a national level. Right? Or an international level. So, what the services will specifically be -- and we certainly have good examples of services that are beyond voice, video, and data, OK? -- it's all data. It doesn't make any difference anymore.
17:36:
Chris: Well, let me ask you this: so, right now, one of the drawbacks of what I've been describing, which is thinking of the fact that -- I get a Comcast connection, and I can use that to watch Netflix. Well, it turns out that the way the network's architected, Comcast makes decisions that impact Netflix's ability to deliver me content. And, similarly, if you and I, who -- we live a mile and a half apart from each other -- wanted to have some sort of direct connection to each other, and we're both on Comcast, well, those packets might take this incredible route all the way down to St. Louis, or something like that.
18:11: Eric: Right.
18:11: Chris: So, you know, how does that sort of architecture play into the fact that we may want to have a network that's built where we keep more traffic locally, and that sort of thing?
18:24: Eric: Any decent network architecture attempts to localize as much common traffic as possible. And that's been true in this business for a long, long time. I think I want to pick on your example a little differently, if I can. OK?
18:42: Chris: You never do what I ask you to ...
18:43: Eric: I know.
18:45: [laughter]
18:48: Eric: It's just the way I am. If you're talking about you and I both having a choice of Comcast, and they're limiting a certain amount of traffic of one kind or another, or the performance is not as good, this is exactly what we want to see in open access. We want to see another company, other than Comcast, be available to the user, to -- who doesn't do that type of limitation. And that is product differentiation -- kind of the standard form we expect in competition. If those two are available on an open access network, Comcast and our hypothetical other company, then we leave it to the customer to make a choice of the product he chooses to buy.
19:38: Chris: Um hum.
19:38: Eric: And thereby, the kinds of behaviors that we see today in legacy providers, it will be their own incentive not to do so -- not to create bad behavior. OK?
19:51: Chris: Um hum.
19:51: Eric: Because they'll lose customers if they do.
19:53:
Chris: Right. if they have the choice. Now, I think one of the next questions, though, is, if we have this open access network, but it's been built basically copying a Comcast-style architecture, because someone like me says, you know what, we really need gigabit Internet access, and let's build this network so that we have gigabit Internet access, right? We're going to offer one service: Internet. Right? Which is how, I think, a lot of people think about it.
20:21: Eric: Right.
20:21:
Chris: How do design choices that we make in building that network, regardless of whether it's open access, if we just see it as being a one-service network -- I understand the irony of describing the Internet as a one-service network, but I think you know what I mean: the idea of just having a big pipe of transferring packets -- how will we build a network to make sure that we can accommodate your vision of open access?
20:45:
Eric: I think there are two things I want to say. The first is, the Internet today is an example of a perfectly open access network. There is customers who choose which sites to go to. They are customers who choose which sites not to go to. And it's totally within the customer's control. Now, when you think of that model and you try and put it into this other form that we've had -- the legacy providers right now -- they are starkly different. But what we're trying to accomplish is, effectively, the type of service selection that we know and we actually have today. Only we want to apply it against all services. And we want new services to be added when they come around, you know, just as quickly as can possibly be done. Now, some of those services that we were talking about earlier are local. They have local quality. So, they might have something to do with your school and your students, your local university and the students there. Which may, in fact, have both local and national relationships. But the way you architect networks -- OK? Which was really the core of our question. And I avoided it. I apologize. But -- the core of your -- the question is, how do we architect it?
22:17: Chris: Right.
22:17:
Eric: Well, we don't architect it the way we used to architect it, any more than we would think in terms of the speed as being only one factor of an architecture. You know, it's sort of like saying, on my 1972 Volkswagen beetle I put on high-speed performance tires so it can go faster. It doesn't work that way. You know? You need the entirety of the architecture. And one of the things that has come forward recently is an architecture choice called "software-defined networks." Network function virtualization. Now ...
22:55: Chris: SDN. NFV. 22:56: Eric: Very good. You know. And the way this particular kind of new architecture is constructed, it allows us to add and withdraw services at will, rapidly, quickly, and even uniquely, to a certain set of customers. It also allows for automatic provisioning of services, such that a customer can go to a web page -- perhaps the service provider's web page -- and select multiple different services off of that that are instantaneously provisioned to that person's home. There are many aspects that are coming forward in this particular new architecture. But, you know, we don't design modern-day jets that have buggy whips attached to them.
23:47:
Chris: Right. Well, let me just throw something out that I think is important. Which is the idea of what a service provider is. A service provider could be the church down the street. Right?
23:57: Eric: Yes.
23:57:
Chris: And so, what would happen is, potentially, a person who, let's say, has broken a leg, and attends this church could sign up and receive the church sermons, and through the magic of this SDN, you would have a local connection that would be provisioned at a very high capacity to deliver this. The payment would be transacted one way or another way. That would be easily, you know, basically be abstracted away from the user. Which is a different way of saying, you wouldn't have to worry about it. You know, all these sorts of things would happen instantaneously, and easily. I mean, that's fundamentally the goal, right?
24:33:
Eric: It's fundamentally the goal. But in this goal, we're also seeing all manner of other changes. So, I think one of the ways of looking at a service provider is that a service provider's function is really to operate the physical network, and to interface with those kinds of services that are requested on it.
24:53:
Chris: Wait. I would have said the network -- you mean the network operator. Like a monopoly network operator operates the physical stuff.
24:59:
Eric: OK.
24:59:
Chris: And a service provider is any one of -- is what I would have expected.
25:03:
Eric: Well, today -- today, when you're talking about a service provider, you talk about Comcast, you talk about any number of those legacy providers. And the interesting piece is that we -- when we have -- when we use that language, we are implying both their physical network as well as the services over which -- over what -- that they are in fact selling us.
25:28: Chris: Right.
25:28: Eric: So there are really kind of two functions going on.
25:32:
Chris: Right. Well, I mean, in fact, if Comcast, you've actually got them separated now. Comcast does physical stuff. Xfinity delivers services. 25:40: Eric: Um hum. Right. And Comcast has other services, like NBC television, that is available on many, many systems besides Comcast's network. So we're seeing changes happen at a number of different levels. We're seeing monetization change, between -- For example, myself. I happen to be in that group of people who have "cut the cord," as they say. OK? And I pay certain services -- Netflix, Amazon Prime, the various Hulus, and so forth and so on -- a subscription-base, that doesn't go to what used to be the legacy provider collection for video entertainment.
26:25 Chris: Um hum.
26:25: Eric: So, a service provider, meaning a person who operates this physical network, in this case -- I'm using that term -- has an interest and role in the way the monetization structures go forward, and in the actual operation of the network itself. So -- And those two are very closely related. One of the things that might well be the role of an open access service provider is to provide an interface for users to be able to select services that are openly offered across that network. We see a little bit of that today in programming guides on smart TVs and Tivos, on various different video entertainment boxes, where they've all got certain kinds of apps. But, you know, for those of you who have tried it, like I have, you know, you can find that it's even confusing between how they display the apps on a smart TV, on different smart TVs in the same house, on different devices -- BlueRay players, Tivo players, ...
25:50: Chris: Right.
25:50:
Eric: ... and so on, down the line. Wouldn't it be nice to have a service provider that provided an interface that was easy to negotiate and navigate, and allowed you to select these various different -- we'll call them services, OK? And perhaps interact on them multiply on the same display device, right? Could well be one of the reasons you'd want that service provider's services.
28:09:
Chris: Right. But fundamentally, I think, we can sort of head toward the end of the show with the thought that we can't do any of those things until we get rid of the monopoly on the end user, basically, which, currently, cable companies tend to have.
28:23: Eric: But the monopoly is created by the physical infrastructure ...
28:27: Chris: Exactly.
28:27:
Eric: ... of those particular things, and so, yes, that does need to be replaced. Now -- but -- and so, earlier, you mentioned that you don't see this coming together as a public piece. I do. OK? I don't see how we can avoid it. Because the cost of deploying physical infrastructures -- multiple physical infrastructures -- to businesses and to homes is a foolish expense. A singular connection today -- well -- provides it all.
29:03:
Chris: I'm totally with you on that. But I think anyone who's closely observed our healthcare debates over the last 20 or 30 years may argue, we can do foolish things for a very long time.
29:14:
Eric: We can. We can. I'm just very glad that the sewer system that connects my house also connects your house and my neighbors' houses, and so on and so forth.
29:26: Chris: Right.
29:26: Eric: And they really don't care what any of us put down those ...
29:29: Chris: Right.
29:29 Eric: ... particular pipes.
29:30: Chris: I do still toss urine out my window, though, ...
29:33: Eric: Oh, you do. Oh, good.
29:35: Chris: So, thank you for coming in.
29:37: [laughter] 2
9:39: Lisa: For more on "open access," follow the tag at muninetworks.org . Learn more about Eric's firm at lookoutpt.com . Send us your ideas for the show. E-mail us at podcast@muninetworks.org . Follow us on Twitter. Our handle is @communitynets . Thank you again to Dickey F for the music for the show. His song, "Florida Mama," is licensed using Creative Commons. And thank you for listening. 30:06: