Thanks to Jeff Hoel for this transcript of Episode 21 of the Community Broadband Bits Podcast. Benoit Felten of Diffraction Analysis joins Chris to talk about his case-study of Stokab, the municipal network in Stockholm, Sweden. Listen to this episode here.
Lisa Gonzalez: Hello, and welcome to the Community Broadband Bits Podcast. This Lisa Gonzalez, from the Institute for Local Self-Reliance andmuninetworks.org .
In our 21st episode, Christopher Mitchell and his guest take us to Stockholm, Sweden. Chris interviews Benoit Felten, cofounder of Diffraction Analysis. The two discuss Stokab, the municipal fiber network created over sixteen years ago by the city of Stockholm. Benoit has authored a detailed case study of the network, and he shares the story with us. Benoit and Chris discuss the ripple effect of the network, and lessons learned from this world-class municipal infrastructure. Now, to Chris and Benoit.
Christopher Mitchell: I'm here talking with Benoit Felten, with Diffraction Analysis -- cofounder of Diffraction Analysis -- who has written a number of important research papers, some of which we've discussed before on other shows. Benoit, it's good to see you again.
Benoit Felten: Hello. It's good to see you too.
Chris: I think the last time we were on a show together, we were talking about bandwidth caps, and there was a paper that put you on the map for a number of people. The lunacy of them, I guess.
Benoit: I think the issue is still around. But, yeah. [laughs]
Chris: Right. Right. Yeah. Unfortunately, we don't change things just by making a strong, factual case against them. But I wanted to talk today about a network in Sweden that has had a lot of attention over the years. And you've published an excellent case study of it, that is available to everyone. And so we'll talk about that in a minute. But first, can you tell us a little bit about what Diffraction Analysis does?
Benoit: Diffraction Analysis is a research and consultancy company that focuses, not exclusively but let's say strongly, on issues around the renewal of broadband infrastructure. It's our belief that the infrastructure that's powered our broadband needs for the last decade -- 'cause that's really how old broadband is, when you think about it -- that infrastructure is no longer suited to the needs of most users in most places. Renewing it, or changing it, raises a lot of issues about who should do that, with what money, and with what -- in what timeframe, I guess. And so, our job is to try and understand what works, what doesn't work, who's doing what.
One of the really important things -- that I probably don't stress enough, and I should stress more -- is that a lot of these issues are not about the technology. The technology is well-understood. There are a number of arbitrations that -- or, choices -- that can be made about the technology. But very few of these successful projects are successful because they picked the right technology. They're successful because they had the right strategic outlook, and because they implemented properly. So, increasingly, a lot of our work is about not just how do you build a next-generation infrastructure, but how do you market it, what are the optimal models to sell it. And that's the key to success.
So, we advise customers in many different countries. If I go east to west, I would start with New Zealand and India, then many European countries -- Eastern Europe, Western Europe -- U.K., and then the U.S., basically. So, we have a fair number of customers across all these countries. And, interestingly enough, even when you look at emerging markets, the issues are kind of the same everywhere. So, it's an interesting market. Interesting times. And, of course, the U.S. is a particular focus because politics have kind of -- have a grasp on what's happening in the U.S., much more than in many other countries. Or, at least, in other countries, you have the feeling that policy-making can be a counterforce to corporate -- well, I'm not going to call it corporate greed. That would be may be slightly excessive. But at least a corporate takeover of the market.
Chris: Right. I think you have a sense that regulations aren't the root of all evil in other countries. Whereas, that seems to be a dominant sense in American politics.
Benoit: Yeah. I would even go one step further and say that, looking from the U.S. market from the outside, you get the feeling that policy is actually actively helping capture of, you know, customers for a few established corporate players.
Chris: You said it's not about the technology. And I think that really bears emphasizing. And, I'm remembering, Lance Armstrong wrote a book, and his slogan was, "it's not about the bike." And, of course, now we've found out that that's very much true. Right? [laughs] Um ...
Benoit: I'm not sure where you're going with the parallel here, but ... [laughs]
Chris: Well, it's just that I think that, you know, a lot of communities get hung up on technology. And the reality is that there's so many other things that are important. Mostly, I'd say, probably, relationships, and business models, and things like that. But you -- you've become an expert on what's happening internationally. It'll be very valuable for our listeners to get a sense of what IS happening outside the U.S., and what's possible. And so, I don't think, for the purposes of this conversation, we have to necessarily bring it back to something that -- lessons that are practical for the U.S. But if we drill into -- is it Sto-KAHB? How do you say the city?
Benoit: Well, I'm not sure my pronunciation is very Swedish anyway. But, you know, they understand me when I say it like that. So, that's enough for me at this stage.
Chris: Excellent. So, you're out of Paris, France, right?
Chris: And you traveled to Stokab to meet a lot of these people, and to do an in-depth case study, which we'll link to on the website. So, I'm curious if you can give us a little bit of background on what they did.
Benoit: Right. So, first of all, let me start by explaining the genesis of that. I was having a discussion with some of the policy people at Google. And they have been aware of Stokab as an interesting model. As far as they were aware, and as far as I was aware, no one has really ever done an in-depth case study into how it came to be, and how it performed over the 15 years it's been around. And so, basically, Google said to me, look, you know, we'll pay for you to do it. We don't want to interfere. We don't want any involvement in your conclusions, or anything. We're just aware that, you know, it takes time and money to do that kind of analysis, and so, we want it to be out there. So, just to be completely transparent, there was a form of sponsorship of that paper by Google, which is completely overt. There's nothing -- there's nothing fishy about it. But I think it's important to stress that up-front.
What's interesting with Stokab is, first of all, that it is probably the largest and most successful municipal broadband project in Europe. And the second thing that's really interesting is, it started -- it was -- the vote to start Stokab occurred in early 1994. So, we effectively have 16 years of history of that project. Which is more than we have on any other municipal broadband initiative around the world, as far as I'm aware. And so, that alone makes it an interesting case study -- or topic for study. Because we have that history. How we analyze that history, and the lessons that might be relevant for other projects is a different matter. But at the very least, understanding how it came to be, and how it was successful, is a really interesting topic.
The important aspect here is that the City of Stockholm owns a number of businesses, in various areas. Some in energy or utilities. Some in public housing, and things like that.
Chris: Um hum.
Benoit: And so, Stokab didn't come out of nothing. It came as one additional business that the city felt was required to effectively serve its citizens. One of the things that stuck me -- one of the first interviews I made is with one of the -- the leader of the opposition party when Stokab was voted in as a project. And, as we stress in the paper, one of the reasons why this happened was that there was political consensus about the necessity to make it happen.
Chris: Um hum.
Benoit: And what he said to me was, no city would imagine having its roads run by private entities. And, for us, this broadband infrastructure is just roads.
Chris: Um hum.
Benoit: It's nothing more, and it's nothing less. And, as such, it's necessary. And we had to make it happen.
And I like that analogy, because it shows that they have a way of looking at infrastructure that, even by European standards, isn't as widespread as you'd think. I know, by U.S. standards, it certainly isn't. But even by European standards, that's fairly forward-thinking. To say the least.
Chris: Right. And so, they actually -- they run this network exactly like the roads. Right? I mean, this is a network that Stockholm, you know, has built. And -- but offers no services on.
Benoit: Absolutely. And one of the -- so, it's interesting to understand how the political consensus occurred. Obviously, you would expect -- using traditional terminology is complicated, because what's "liberal" in Europe is what's "conservative" in the U.S. But anyway. So I'm going to try to avoid that terminology. But, basically, "pro-market," let's say -- "pro-free-market" -- parties, you would assume, would not be in favor of that kind of venture. And what they said was, well, we understand it as a necessary infrastructure. We don't want it to disrupt the market. And in order for that to happen, that infrastructure has to be sold as infrastructure. So, the only thing Stokab sells is dark fiber point-to-point connections. And that has effectively allowed for a thriving, competitive service market to be built on top of that infrastructure.
Now, I'm always weary of analogy. And the road comparison didn't come from me; it came from them. So I don't really know how you would compare it to roads in that sense. But the important thing is, they're not competing with the private market. Serving everyone on an absolutely equal basis. And, in fact, transparency in their offerings and business dealings was also a prerequisite from, you know, that original political decision.
So, that's really the world view that they had when they launched it.
Chris: And so, when they launched this, they didn't just put dark fiber in major corridors, right? They ended up taking it very deep. And so, in the U.S., and many places, we typically think of dark fiber as going from point A to point B. But point A and point B are not houses. They're typically, you know, maybe a school or a large institution. But Stokab, they actually run the fiber much deeper into neighborhoods. Is that right?
Benoit: Yeah. Although -- Absolutely. But you have to remember that there's 16 years of history here. So, ...
Benoit: ... it started -- it started by connecting, you know, hospitals, universities, public institutions, large businesses, as the contracts were signed.
Chris: Um hum.
Benoit: And, gradually, until about 2003-2004, that kind of mesh densified. But they were still not actively addressing -- ah, houses. In fact, there's a bit of a specific to the Swedish market which has an impact on exactly how their strategy unfolded here. And that is that, inside of a building, no one can own any infrastructure, apart from the building owner. Or owners. So, the model that is prevalent in my country, France, or in most of the U.S., where a telecom operator would deploy its copper network all the way into your home, but still own that copper wire -- that's legally not possible in Sweden.
So, what happened in Sweden was that the real estate owners, when they started realizing the potential of fiber services, turned to Stokab and said, where do we have to pull fiber to connect to your network? So, it happened from the other side.
Chris: Um hum.
Benoit: It came from the homes. And, typically, a real estate provider will either, himself, build the in-building fiber and pull it back to the nearest node -- which is usually a few hundred meters away from the building -- or they will contract out to an intermediary who will manage that network, but not own it. And that intermediary, having a long-term contract, can actually afford to roll out the network for that limited distance, and complete the loop, effectively.
But it's important to understand that Stokab doesn't get into the buildings.
Chris: Right. I think it's such a key point there. That regulation that stops the existing telco from being able to monopolize a multiple-dwelling unit simply because they were the first one there seems to have been a major deciding factor, as I recall your paper, because these real estate companies were such an essential part of making sure much of Stockholm had access to this network.
Benoit: So, my answer to that would be, yes and no. Right?
Benoit: Real estate players clearly had a very important role. And I'm going to come back on that for a minute. But the fact that the incumbent operator didn't own the copper that went into the building didn't stop them from being a monopoly provider for most of their history.
Chris: Um hum.
Benoit: Because you still have the cost of deploying that network. Even if it has to be borne by the real estate provider, one way or another, it's still a very important cost. The deciding factor was that the public housing companies and some of the private social housing companies realized that this was a way to not only enhance the services that their tenants would get but also dramatically improve the way they would manage their buildings. I recently had a discussion with one of these Swedish public housing companies, and they were telling me that their response time, when there are heating issues -- And, typically, you know, heating is not a trivial matter, by any stretch of the imagination.
Benoit: And they were telling me, the difference between being able to stop a heating leak within an hour, versus have that go on for 24 hours -- that difference financially is huge. It's -- you know, it's immediately thousands of Euro-equivalents.
Chris: So they built smart buildings, essentially, with this broadband infrastructure that they were building to -- They used it not only to benefit their tenants but also their operations.
Benoit: Absolutely. And then, there was a kind of a viral effect, because real estate providers that were not serving their buildings with fiber, or had not done that, started to realize that they weren't filling their buildings as easily as they used to. Whereas the other guys were. And some of these largest real estate providers actually had buildings with fiber and buildings without fiber. And they were -- you know, looked at them comparatively and said, oh my God, that's the difference. That's why, you know, we have a 20-percent non-occupancy rate on these buildings and a 1% non-occupancy rate on those buildings. And, again, if you translate that in revenue, the difference is so huge that, you know, connecting those homes with fiber becomes a trivial decision financially.
Chris: Um hum.
Benoit: I think part of the issue that we see in many other countries -- and that's not just the U.S. Part of the issue is -- and this is where you're right, in terms of the role of the real estate players. Part of the issue is, until now, most building owners in most countries -- at least in the developed world -- they have never to spend a dime on getting communications services inside the building. Because the price of that monopoly provision from the incumbents or the cable operators was, well, I'll pay for it, but I'm the only one there. I think the Swedish real estate companies -- the company is in Stockholm, but this also happened outside of Stockholm -- they were much more aligned with thinking that, you know, are we willing to look at a business case for this where WE spend money? I am pretty certain that any building owner in New York City today who made that calculation would realize very quickly that if they paid for fiber access to their building, the positive outcome of that would be huge.
Chris: Um hum.
Benoit: It's just that their mindframe is, no, no, these guys have got to pay for it, not me.
Benoit: Do you see what I mean? So, I think it's more a matter of how they view the world than the actual economic analysis behind it.
Chris: What are the key takeaways? I'm curious, as you dug into this -- Obviously, communities can't just go back 20 years to duplicate, but they have to move forward. And so, what are the lessons that you think are the most valuable, for any community around the world?
Benoit: So there's two aspects of that. One of them is, you know, what can we learn, in absolute terms, from what Stokab did? And the second one is, how do you replicate that?
Chris: Um hum.
Benoit: There are two important aspects in the first one. The first important aspect is, it worked. Right? It's been very successful. We keep hearing, over and over again, that, you know, it's a waste of taxpayers' money. It can't work. Well, it does. OK? So, next time someone says to, it can't work, you can say, well, why did it work there, if it can't work?
Chris: And, to be clear, you've -- in your paper explains, this is a network that did not cost the taxpayers anything, ...
Chris: ... and, in fact, has been operating in the black for a long time.
Benoit: Well -- and that's the second point I was going to make. This was not financed with taxpayers' money -- at all. There was one year where Stokab lost a significant amount of money because they overstretched in the network deployment, and the commercial take-out from that wasn't sufficient to basically compensate the excess deployment that they had done. And so, on that year, part of their debt was refinanced by the holding company that manages all of these public assets that the city has. But because many other of these companies were in profit that year, effectively, there was no external money filling that. Now, they were -- there was one single year where they were not profitable, in their 14 years of existence. Now, they're churning cash for the city very, very significantly. They're a great asset to the city. And even though they are reinvesting a lot of that money in network expansion. So, by the end of this year, they will have 400,000 homes eligible for service in Stockholm. Which is, basically, every multi-tenant dwelling connected. And, originally, they weren't going to do single homes. And I think even that they're starting to look at now, because they think it can work just as well.
That's a really important thing, too. This was not financed with taxpayers' money. Basically, the backing of the city obviously allowed them to borrow money at, you know, good rates. And that is what sustained them.
Now, they did this in 1994. In 1994, the competition they were facing were brand-new cable operators that were not operating broadband at all. And that didn't serve customers with broadband until the middle of the first decade of the 21st century. And an incumbent that was basically operating on copper for 99 percent of its needs. So, they were well ahead of the game. And that was a key element to their success. The big difference that you would find today, in a city the same size as Stockholm, is, for the last ten years, private businesses have been cherrypicking buildings and institutions that are obviously profitable to connect with fiber. And so, that cherrypicking effectively degrades a global business model, if you were to have a long-term strategy of connecting the whole city.
Chris: Right. Universal service.
Benoit: Right. You could argue that as a consequence of that, that their experience is not replicable. I think what is replicable is -- two things. But it's replicable because the market has changed as well. So, sure, the cherrypicking happens. But, back then, the idea of a city partnering with a private venture to basically build a balanced business model, where both the areas that a private company would not find profitable and the areas that they would go for anyway would be covered by the same business model, and one would balance the other with some amount of public money, either subsidies or investment, to compensate the non-profitable areas -- that's now become not just envisageable but actually implemented in many places. So, that's the first element that I think is really important.
The second element is, their cost of deployment in 1994 was probably three to four times what it would be today.
Chris: Um hum.
Benoit: Because the technology has evolved a lot in the meantime. And, you know, whatever technology they were using back then, to trench, to dig, to blow the fiber, etc., these were really early days of fibering the access, to say the least. So none of these technologies was stable. There was a lot of risk and uncertainty about it. All of that's disappeared now.
So, if I had to basically stress one message, it would be that. I'm still not convinced that you could replicate Stokab on the scale of a city like Stockholm. I think those large cities -- there -- too much cherrypicking has happened in the last ten years to make a project like that easily imaginable, basically.
Chris: Um hum.
Benoit: It's maybe not impossible, but I think it would be very complex. However, in Tier 2 cities, very little of that fiber infrastructure exists today, if any. And we're still talking cities of, you know, several hundred thousand of residents. So, in those places, I think the Stokab model can really represent a good starting point. Maybe not as an exact replication. But there are definitely elements in there that can point a community willing to bring state-of-the-art infrastructure to its citizens and businesses towards a model that would work.
Chris: Right. And I think some of that can be -- we're about to release a case study on Santa Monica, where they effectively traded off time for money, in that they built a lot of their network when the streets were already open. And so, if you have a city that has this opportunity, they can even keep costs low by just marrying the burying of this fiber and conduit with other projects that may be opening the streets already. And so, you don't need to have the kind of success that Stokab had in order to pay it off. You keep the costs a lot lower, and so you can bring benefits while generating fewer revenues and still operating in the black. And so, I think, you know, the goal is never just copy these, but to learn from them. And I think you've really helped us with that.
Benoit: I fully agree. And just on that point, when I was having these discussionswith the politicians who took that early decision, one of the things they told me, which I thought was really interesting was, at the time, they had no idea how much this network would power. I mean, you know, they weren't technology visionaries or anything.
Chris: Um hum.
Benoit: One of the key reasons that everybody agreed to do that was, because of the market liberalization, they could imagine dozens of service providers asking to dig in the streets, to lay bits of network here and there. And they envisaged an absolute urban chaos as a consequence. And so, one of their most important drivers was actually to limit the amount of roadworks by, in effect, centralizing and, you know, forcing a collaborative model onto all of the market players who wanted to access. And the benefits to the market players have been huge. One of the really interesting aspects of Stockholm is, they have four 4G networks in operation today. They have a level of competition in LTE that is unrivaled in the world. And the only reason that was possible is because fiber access to cell sites is much cheaper than it is anywhere else. The urban angle leads you to a very healthy competitive market, much healthier than it is most places in the world.
Chris: I understand that you are doing some more research into municipal networks around the world. Is that right?
Benoit: Absolutely. In fact, in -- probably in a week from now, we're going to release a report that analyzes the structures and the financing mechanisms of many of these in Europe. The reason we chose to focus on Europe is that the European Commission has set up some specific financing mechanisms and constraints for a municipal network. So, obviously, there's a lot of them. But they operate under certain rules that are interesting to kind of dissect and discuss. So, that's coming out -- yeah, within a week from now.
Chris: Great. We're going to have you back on, to learn more about the lessons from that.
People -- I recommend Googling, or searching, the Internet for "Diffraction Analysis." It's diffractionanalysis.com . And if you search for that and "Stokab" -- S-t-o-k-a-b -- you'll find this case study. So thank you for coming on the show.
Benoit: Thank you very much, Chris. Au revoir.
Lisa: That was Chris talking with Benoit Felten from Diffraction Analysis. Be sure to visit diffractionanalysis.com , where you can download the complete case study and learn even more about Stokab. If you have any questions or comments, we encourage you to e-mail us at email@example.com . Our handle on Twitter is @communitynets . This show was released on November 13th, 2012. Thanks again to Fit and the Conniptions for the music, licensed using Creative Commons. The song is called, "Got My Modem Working."