Transcript: Community Broadband Bits Episode 214

This is episode 214 of the Community Broadband Bits Podcast. The CEO of GWI joins Christopher Mitchell to discuss last mile and middle mile networks. Listen to this episode here.

Fletcher Kittredge: You can't have the last mile without the middle mile. It is necessary, but it's certainly not sufficient.

Lisa Gonzalez: This is episode 214 of the Community Broadband Bits Podcast from the Institute for Local Self-Reliance. Welcome. I'm Lisa Gonzalez. Which is more likely to bring better connectivity to a larger section of the community, middle mile or last mile infrastructure investment? Both are necessary and, while there are those who believe middle mile investment inspires private providers to deploy their own last mile infrastructure, that theory has yet to be proven. On the other hand, last mile connectivity depends on solid middle mile infrastructure to connect a local community to the outside world. This week, Chris talks with someone who has expertise with both. Fletcher Kittredge, CEO of GWI, has developed projects that involve middle and last mile connectivity. He and Chris discuss a range of issues including where limited dollars should be focused when both last and middle mile infrastructure needs developing. Learn more about the company at and read more about Maine's Three Ring Binder Project at the tag Three Ring Binder. Now here are Chris and Fletcher Kittredge, CEO of GWI, to talk about middle mile and last mile deployment.

Christopher Mitchell: Welcome to another edition of the Community Broadband Bits Podcast. I'm Chris Mitchell and today I'm speaking with Fletcher Kittredge, the CEO of GWI in Maine. Welcome to the show.

Fletcher Kittredge: Hello. Thank you for having me.

Christopher Mitchell: Fletcher, it's great to have you back again. You've been on I think twice before. Maybe just once. But, you and I have talked many times. I think we'll call this our "middle mile versus last mile show" in which we talk about some of the tensions I think between where to invest to spur better Internet access. To start, I thought we would ask you to just tell us a little bit about the Three Ring Binder approach in Maine and your role within that.

Fletcher Kittredge: The Three Ring Binder is an 1,100 mile very high capacity, dark fiber middle mile network that's built throughout the State of Maine. It goes primarily in rural Maine though it touches urban Maine too. It's in the form of 3 large fiber rings that go around the state and the rural areas. It's dark fiber which means there's no lit service. It makes it possible for carriers to provide service but it requires carriers to do it. It's open access.

Christopher Mitchell: You said it requires carriers to do it. Is that just to distinguish that if I owned a business selling widgets, I myself would not be able to buy dark fiber off the Three Ring Binder?

Fletcher Kittredge: No, not at all. Anyone can buy dark fiber off of the Three Ring Binder but it's just someone needs to put the equipment and light it.

Christopher Mitchell: Right. Okay.

Fletcher Kittredge: It's an anti-monopoly network in a way. It is meant to create to create a platform where anyone could provide Internet access in a rural area. It's not just used by carriers. You have some business that are more sophisticated and have their own network department use it to light up locations between different locations. It's used a lot by hospitals which tend to have their own network and departments. The primary use is carriers who put their equipment on it and use it to provide middle mile services and last mile services in some cases. It is an all-in-one service. There was really nothing in rural Maine. The first thing that was needed to lay down a foundation was the fiber itself. The idea was if you put the fiber in, that would lower the barriers for carriers to provide access to supply what was already there but they had to make an investment themselves in the form of equipment. The hope was -- very flexibly --They would create a public-private partnership where the private sector would -- It would be worth it for them to do it because the cost had been lowered so much but at the same time they would have to put some skin in the game. To give you a sense of the cost, it's per strand, per mile, per month in the neighborhood of $15 to $20 per strand, per mile, per month which is maybe 50% to 1/3 of the weight for commercial fiber. In many of these places, there was no commercial fiber available. Certainly not on an open access basis.

Christopher Mitchell: Right. Because I think like many states, Maine -- You had a situation where, to the extent that there was fiber or high capacity links. It was controlled largely by the incumbent telephone company that had historically served that area which had been Verizon until they sold it off to FairPoint which was an unmitigated disaster according to many of the consumers that I've talked to just in terms of the actual transfer. Things have gotten a little bit better but people are still pretty frustrated. That's just to say that the incumbent then has little incentive to actually make that available to competitors.

Fletcher Kittredge: Right. It was not available to competitors.

Christopher Mitchell: Anyway, that was the middle mile. I don't want to get too far down that rabbit hole. You were involved in creating that project. You helped to make sure it got stimulus funding. It's taken off. It's done very well. The Maine Fiber Company which operates it seems to be doing well. You also have last mile experience. Just as a reminder to folks who maybe haven't been listening to all the back catalog of shows, what's your experience with last mile networks?

Fletcher Kittredge: We started doing last mile networks, about  fiber networks, about 10 years ago. A little over that. Now we -- we're like many companies. We're well in the transition off the copper network to the fiber network. We provide residential and business service, primarily business service, off of last mile networks. Some of those are in partnership with municipalities. Some of them are I think interesting, charismatic projects but then I'm from Maine so something like creating an island off the coast of Maine where there's universal gigabit broadband for everyone, I think, is really neat. A gigabit island it's pretty cool.

Christopher Mitchell: Yeah. Islesboro which we have done a couple of shows on. And I haven't been there, but I've glimpsed it from the shore. I knew that was beautiful.

Fletcher Kittredge: I really hope you get there sometime because it -- If I didn't have family on the mainland, boy, that's where I'll live. It's a neat community and it's a beautiful, beautiful place. There's just something really special about living on an island. The closeness you have with different people. My expectation is now, if there's really cheap gigabit access available to everyone, that it will attract young people who can have jobs where they can live anywhere and have the opportunity to raise a family in an island environment. I was delighted to be able to give my kids the experience of a Maine childhood, but I think an island childhood trumps that.

Christopher Mitchell: Speaking of Trump, whether it's Trump or Clinton, we're going to have a new administration come next year. One of the things that people in the broadband world have been talking about is how to improve access. Sometime over the last year I remember NTIA at a conference saying that they thought the smartest thing a new administration could do to spur better Internet access would be to just build a lot more middle mile access much like we invested in the BTOP Program, the Broadband Technology Opportunities Program, as part of the stimulus investments. I thought it's worth talking about because that raises my hackles a little bit. I've argued that middle mile investment is definitely necessary although, I think, the benefits from it are less than have been expected. My reading of it is basically that, when you build a lot of middle mile, you get a little bit of last mile. I guess, I'm curious if we could just start with a maybe empirical look. What has happened with the robust middle mile you have throughout Maine?

Fletcher Kittredge: You can't have the last mile without the middle mile. It is necessary, but it's certainly not sufficient. To some extent, that's what we've seen in Maine certainly in the first few years. It's starting to tick up now but that thing was completed almost 5 years ago. We have a number of communities that have really good broadband now. Some of these are really rural, remote communities in Maine where you have the best broadband in the state.

Christopher Mitchell: That's mostly served by I think historic co-ops or local companies that are really rooted in the community, right?

Fletcher Kittredge: Yeah. In fact, it's entirely small companies or the independent telephone companies which are not co-ops, but it's all small companies. And it's either the independent telephone companies which really have much better coverage than the big incumbent. It's like the bigger those independent telephone companies are, the worst their broadband is. Then a bunch of new entrants such as Pioneer Broadband up in Aroostook County. In fact, I think all that money however, the last mile has been done with additional government subsidies mostly from the state and then from the ConnectME authority. I'm sorry. The state is the ConnectME authority. Then there's the federal Connect America Fund. It's a drop in the bucket. We're getting to the point where the economics justify doing it with private money, but we're not there yet. It's just that these are rural communities with a lack of density. And the adoption rate is problematic. That's just something it is. There are two dynamics going on that seemed potentially to make things better. One is adoption rate going up. Basically people are less and less and less satisfied with their DSL and wireless base broadband, all the time. DSL really doesn't meet, and a lot of the rural wireless, doesn't meet the definition of broadband anymore. The definition of broadband, the bar is constantly being raised all the time. That's going on. More people are willing to switch to fiber. At the same time, the costs are coming down somewhat in providing fiber as things get more efficient. Those things are going on, but it's a slow trend. I think with nothing else happening, it's going to be a long time til that is complete. And there are many reasons to believe that there's a bunch of communities without some level of government support which will never have sufficient networks. If you think about the power network -- just fundamentally the rural networks -- the power network, the road network, the telephone network, all in rural areas all required government support. The economics are not changed enough by fiber even though it's cheaper than copper. It's still too expensive and quite a few areas to do it. The payback is just not there. There's going to need to be government subsidy for the last mile.

Christopher Mitchell: If we accept that there's going to be government subsidy, I think the question then is how to target it? One of the things that I've said and you can certainly -- I'm not asking you to just go along because this is my show. I love being challenged and being told where I'm wrong --

Fletcher Kittredge: Don't worry.

Christopher Mitchell: When you build robust middle mile, you are lowering the operating cost of a network. If you think that building middle mile is going to change the equation on the last mile, you're missing an important point which is the barrier to investing in the last mile is primarily a capital cost. It's not an operating cost. Operating cost is a barrier. But my argument has been that when you put a lot of money into last mile networks, then you create a business model for middle mile networks. When you just build a lot of middle mile networks, you're not really creating a business model for last mile networks. What do you think about that?

Fletcher Kittredge: I'm not sure I agree. First of all, the difference between capital and operating cost is ultimately just an accounting issue. It's money out the door either way and money that needs to be there. If you're starting up a network, operating costs in a way are capital cost because you're going to have to fund some losses in the early days and you're going to need the capital to do that.

Christopher Mitchell: Right.

Fletcher Kittredge: I really think middle mile is the foundation. You need to do it first. If you build a last mile network that -- New Hampshire FastRoads. There are government funded last mile networks that ended up not working, and to a larger extent, not working because of the high cost of the middle mile. That's one example there and I'm sure there's others. There's one point I really want to make is one of the dumbest things people do and you see this constantly is you build a middle mile network that's just a middle mile network and it's not a last mile network at the same time. That makes no sense. We have a conceptual tool dividing the network into middle and last mile but fiber on the pole is fiber on the pole. If you're running a middle mile network, build it so it can be used for a last mile network for all the places you pass. You probably pass quite a few of them. Many of these rural areas, you're going to run the middle mile right down the highway and that happens to be where all the business and most of the -- Goodly part of the population is. You start with a big win right there. You should never build fiber and not build it to serve every dwelling and building pass. The incremental cost is so low compared to the cost of going back and redoing it. You see people redoing it and paying massively for that all the time. I think, ultimately, Chris, people are trying to come up with a half ass solution that ends up doing nothing. There's no way that you're going to get out of having to subsidize both. It will be easier ultimately to deal with a last mile problem than the middle mile problem. There's more of a conceptual role for the government in it because the middle mile is massively shared infrastructure. It's shared between multiple towns. I can see a day when a community can invest in itself. It's taken us a while to put together the models and tools because we're used to networks getting help from the federal government for power, the telephone, and roads and water, storms, wastewater, and things like that. The change in the world where all of a sudden the federal government doesn't help people anymore. We don't share those expenses amongst ourselves is a new one.

Christopher Mitchell: I think one of the challenges is that when these decisions are made in DC in particular, but also in the states and to a lesser extent in the cities, I think, you have people saying, "Well, let's come up with a solution that doesn't upset anyone." I think that's one of the things that you're definitely pushing back against. I believe that one of the reasons that the stimulus program focused so much on middle mile is because it was less likely to upset incumbent operators. Nobody's ax was really being gored significantly.

Fletcher Kittredge: Yeah. Okay. Yeah. You got regulatory capture going on big time. I'm a Yankee Republican. I believe in investments that pay for themselves. You should get more out than you put in when society invests. I believe that rural broadband is one of those. Ultimately, you build these. You're either going to end up dealing with a depopulated rural area or you're going to end supporting a family in a county, or you can do this investment and create economic growth there that spins up everything because the rural areas depend on the urban areas. If it's farming or natural resources, then the urban areas depend on the rural areas too. This is an investment that society can make that makes economic sense.

Christopher Mitchell: This raises an interesting question. I'm just curious to get your reaction to it which is to say that you can actually do this investment in a very poor way though too which is something that I'm afraid we're doing in Minnesota, which is to say, "We're going to invest in rural fiber but only in areas that have nothing." Which is to say that, if you have a county and you have a population center that has poor DSL, slow DSL, slow cable, we're going to invest in fiber all around that and not allow any government-subsidized competition inside of that area. Our fear is, is that then you actually are going to be pulling apart those main streets. Is that a dynamic that you would foresee?

Fletcher Kittredge: I would think so. The danger of saying bad wireless and bad DSL means you can't have good broadband, big broadband that's necessary for economic growth and survival of the community. Young people don't stay where all they have is DSL. That's just not good enough for, to be part of today's economy and society. If you think it's bad now, just wait for a year or two. There's so much fiber being built in urban areas. It's going to become the economic norm and rural areas are really going to take it at its neck. I have heard from real estate agents, the first question that people ask is, "What's the broadband like here?" They assume that it's got clean water and they assume that it's got power, but broadband is becoming just as important. I'm not quite sure what pulling apart main street means which I think is what you said but I think if you go to a community and say, "Availability of DSL and older wireless broadband means that you are not eligible for government subsidies to build good broadband," is really going to hurt a lot of places.

Christopher Mitchell: What I had in mind when I was saying that is that you had a situation which businesses and people who live in those towns are thinking to themselves, "If I stay here, I can pay often more for much slower, less reliable service or I move 10 miles away into the country."

Fletcher Kittredge: Will businesses and people, over time, make the decision to move out of communities? The answer is absolutely yes. You see that all over the place. We have a real problem admitting that problem, and it's certainly not the only, and probably not the major, but it's a definite component, I think. As a component, the importance of it will rise over time. Just think of what it's going to be like in a few years. Most people, they don't even think a few years out. Communities are not going to be able to participate fully in the economy. It's going to be a real problem. Businesses won't come there. Business will relocate, and the same is true for individuals.

Christopher Mitchell: I think one of the things I mean when it comes to smaller, rural cities, I think Maine's experience, not way up in the north country, but when you're in just the normal areas in between Portland and Augusta or Bangor and what not, they're pretty similar to our Minnesota, Greater Minnesota cities. I think the idea that you would take that population that lives, the small population aggregations that you have and then you disperse them. It's one thing if they move to Portland or to Bangor. It's another thing if they move 15 miles away and you're trying to figure out, "How do we build roads in this low density are? How do we build water systems? How do we take care of all that?" We want people to be aggregating in those areas.

Fletcher Kittredge: Okay. I haven't thought of that before but it's a way of -- Maine is over, from time-to-time over the last 5 decades, suffered from what we call sprawl, which brings a lot of those problems. And yes. This could be something, if you don't provide good broadband in the population centers and do provide it in the rural areas, of course people are going to move. It's going to drive sprawl.

Christopher Mitchell: Let's touch on one final thing before we run out of time. Let's talk about cherry-picking which is when I first started researching what happens when you build robust middle mile, I came across the Alberta SuperNet which was built almost 20 years ago now, it started being built. They connected over 400 population centers in Alberta, which tells you that they're talking about a population center being anywhere of 50 people hanging out or more. They opened it up with such a pricing mechanism that their goal was to spur last mile investment by anyone. When I talk with them 5 or 6 years ago and things might have changed since then, they said that one of the big results that they saw was basically cherry-picking. People would go to the Alberta SuperNet point of presence and they threw up some wireless so they would do some basic investment and they would just try to get a few customers as many as they could and then they would be done. They wouldn't invest further which I think makes universal service even more expensive than it otherwise would be. How do you build the middle mile to make sure that you're not encouraging that cherry-picking?

Fletcher Kittredge: First of all, you've hit on a really important problem. Second of all, the reason it's so hard -- I think, it's going to be so prevalent, is it's hard to avoid and I think it's just to some extent an economic reality. What is the mechanism for getting people, carriers to serve people that aren't economic? We tend to be future-oriented here definitely to a fault. There's some things you can do. First of all technically, if you extend the network to serve those few customers around the middle mile, there ought to be some mechanism, and we're just doing it with private money because the cost is not that much but the benefit is so great. To build it in such a way, that if you build a mile off the middle mile, you plan to serve everyone that's 2 miles out and 3 miles out and 5 miles out. You design your network in such a way that you expect to extend it ultimately to be universal service. That's really important. I think that the best way from a societal standpoint to deal with this is to come up with targeted subsidies that would vary on a much finer grain than what we've had in the past for providing service outside to the people who wouldn't be cherry-pickers or are not economic. What we're looking at and working with theirs is really good. If anyone out there who's listening is planning to or building a fiber system, you ought to look a new -- The software is new. It's called Vetro. It's put out by a company called MVP Solutions. The idea is to use big data to figure out exactly how much it would cost to extend a network to additional locations be it a neighborhood or an individual house or a road or something like that. And that can be updated dynamically as your network grows. I'm a computer scientist, and many of the most successful computer science theories come from nature. If you look at the way that a plant grows, the way the tree grows, where you start out with a spur, a leafing spur, and that grows into a branch and that other branch just grows off of that. You could use software like this to build a network in the same fashion so it grows organically. Your cost are updated. If the network grows and a location gets closer to the network, the cost of extending service goes down, and you could target subsidies in a better fashion where for a given location. 25% subsidy on the cost of construction. And then another needs are 50%, and there are maybe some rural areas that need 100%. As the network grows, different network providers can bid on that. Bottom line is cherry-picking is shaping the world. If you think about it right now, Google and everybody's investing their fiber dollars in the wealthy dynamic southern cities and/or a few wealthy dynamic northern cities, but everywhere else is not getting that. Over time is that the best investments become not available, people go for things that are more marginal, but if this cherry-picking is shaping the network we have today and will on to the future and --

Christopher Mitchell: Right. I think to some extent that was maybe one of the goals of the Connect America Fund although I think perhaps outweighed by their generosity to the incumbents. They did the reverse auctions.

Fletcher Kittredge: There's so much regulatory captures you notice. You just can't do something without taking care of the big boys. It's unfortunate A and B, this is -- What I'm talking about is the technology solution to it. Really harnessing big data and being a more dynamic system. Our governments aren't very good at technology, and they tend to be behind the curve. At some point, hopefully it would get to this point and then you got to watch out for -- Pay a lot of attention to solutions to make sure there isn't regulatory capture involved in that, so you don't end up with a result you don't expect. Conceptually, I think it's a topic that's worth a lot of exploration. You should get a group of us together and we should talk about it.

Christopher Mitchell: Great. We'll keep that in mind for future show. I like that idea. Coming soon to a podcast feed near you.

Fletcher Kittredge: Cherry-picking. It's going to be big.

Christopher Mitchell: Thank you so much for coming on this show to talk about middle mile versus last mile.

Fletcher Kittredge: Enjoyed the discussion.

Lisa Gonzalez: That was Chris and Fletcher Kittredge discussing last mile versus middle mile to expand Internet access. We have transcripts for this and other Community Broadband Bits Podcasts available at Email us with your ideas for the show at Follow Chris on Twitter. His handle is @communitynets. Follow stories on Twitter. The handle is @MuniNetworks. Thank you to the group Roller Genoa for their song "Safe and Warm in Hunter's Arms" licensed through Creative Commons. Thank you listening to episode 214 of the Community Broadband Bits Podcast.