The following are the outputs of the real-time captioning taken during the Tenth Annual Meeting of the Internet Governance Forum (IGF) in João Pessoa, Brazil, from 10 to 13 November 2015. Although it is largely accurate, in some cases it may be incomplete or inaccurate due to inaudible passages or transcription errors. It is posted as an aid to understanding the proceedings at the event, but should not be treated as an authoritative record.
***
>> CHRISTIAN DAWSON: Thank you so much for joining us here today. I appreciate it. We're going to have a fun workshop. It is going to be participatory. I hope you have come to help us think through some of the problems that we're going to be addressing here today. I think that there are big and important issues that we will be able to along with my colleagues lead the discussion around. From the topic that we are dealing with today is called local infrastructure deploying internet infrastructure globally. My name is Christian Dawson and I'm chairman of a trade association called the I2 Coalition. So we are guys the rack and stack servers in data centers and build the internet infrastructure. In that capacity, we're very interested in being able to make sure the internet infrastructure flourishes globally. On my way over here this morning, I was sitting next to somebody on the bus and she was from the Philippines. I was telling her a little bit about the pal we were doing today and she said to me, "You know, I would really love if you come over and talk to us in the Philippines. There's a tremendous amount of work we need to do in order to improve our infrastructure for business." I'm hoping that I can come to your session that you can connect with people and facilitate conversations about how we can improve the fundamentals of the internet core infrastructure. That breaks my heart and that's exactly why I want to be here leading the conversation about how we can improve the internet infrastructure. I'm talking not just about some of the really cool technologies that we're seeing out here like if you go out to that tent, you see the amazing satellite technology that they've got now to actually make satellite internet fast, which blows my mind. But we're talking about the fundamental undepending of how it works in the community right down to the core data center infrastructure. We're talking about the AXP. We're talking about data center infrastructure. We're talking about the ISPs and the interconnection points that connect the community and to facilitate the hosting of local content within a community. So we're gonna drill down into the kinds of things that need to happen in order for an entire internet infrastructure Eco system to take route into the community and we're going to talk about the importance of why a community should want and make even discuss why the community may not want to focus their attention on that sort of thing. So I am excited to be able to lead that conversation. I'm very excited about having a great group of panelists here to help us guide the conversation with the workshop. To my right here, I have Robert Guerra who is going to be the community engagement co‑chair and also the repetiteur for the day. Next to him, I have Michael Kenday who is with the (inaudible) society. On our remote participation line, we have two additional panelists. As we move forward and we start getting into questions, I'm hoping that people when they first speak and go into who they are and why they are ‑‑ why their role is relevant to this particular conversation before jumping into their conversation. But for now, we'll leave it at names and affiliations. What I'd like to do first is I would like to ask Michael to help guide us in getting our roots in the scope of the issue by talking us through a survey that he recently did with ISOC. And so if I can very quickly turn the mic over to you, I will pull up the slides that I supported you over and that will be a great way to help us start up the conversation we will have today.
>> MICHAEL KENDAY: Good afternoon. I am the chief economists with the Internet Society. I'm going to talk about ‑‑ okay. Fantastic. So I will talk about a study that we did recently that addresses some of the issues raised here on local content hosting and we did this study in Rwanda and I think it holds in many, many studies here today. Increasingly, people are starting to think about local content and making the internet more relevant for people because in many, many countries including Rwanda because of the mobile and mobile broadband, availability far outpaces adoption. In some countries including Brazil, where surveys have been done and asking people why are you not on the internet, increasingly, the majority say it is just because they don't find it relevant. They don't think it's interesting and they don't see the need. It is much, much less availability and cost that. Depends on countries and many countries, cost will still be an issue. But the point is in general finding local content that's relevant to people, making it accessible helps create interest, brings people online and helps people who are already online to make people more familiar with it. The project was three related points. First in most and it's not all developing countries there is already local content being created. It might be the newspapers, magazines, some e‑Commerce sites, blogs, everything else, but there is local content available. But it almost always tends to be hosted abroad even if there's data centers in the country and IXP, often the data centers are quite empty because everything is being hosted abroad. From Africa, it is being done from Europe and in here, it is Miami or somewhere in the U.S. Performance of the website is fairly poor mainly because of latency as we'll see and that keeps usage down. We needed a lot of data. So we did a ‑‑ we had a partnership with a ministry and the ministry of ICT in Rwanda that brings the content together and put together the resulting study up there in case you want to take a look at the full study. I will go quickly through the theory and what we found. The content developer looking around, finds prices for hosting and almost always finds it is much cheaper to hosts massive data center in London or somewhere where prices are very, very cheap. So they'll send their content overseas. They put in the data center and every time someone wants to see it, the IXP has to carry the content back over quite expensive international connections. It imposes a large cost on the IXP and economic terms we call that a negative extrality. It is imposing a huge cost on someone else to save money. Those high transit costs that you need to incur every time you bring the content back, um, don't just impose a cost, but because it is expensive, it is usually under provisions, the international transit the IXPs don't buy as much as they need. It also adds a lot of latency aside from just the distance the content has to travel. So it adds a lot of latency which impacts users when they're trying to download a website. Now, if it is hosted locally, again starting from the lower right, the cost of hosting is a little higher or maybe a lot higher depending on the country. You put in the local data center where it is made available through the IXP typically to any IXP. The costs are very, very low for transmitting the traffic within the country meaning there's enough demand and enough capacity available and the latency is very low. It travels much, much faster and in order to not just say this and have it taken at faith, we gathered a lot of numbers from the industry so that we can show this and we showed with one IXP, well, one content developer had a website they had in the U.S. and we showed they were saving about $111 a year on hosting it in the U.S. They also shared the amount of traffic that website was generating and it was incursion a cost of $13,500 on the ISP. So that's a fairly significant cost for $111 saving and that's a great illustration of a negative externality. It was also imposing large latency costs. So we measured latency and if you were trying to get a local website, latency was very, very low, less than 10 milliseconds if you understand the numbers. If had to get it into the U.S. typically up to 400 millisecond or higher and latency was high meaning throughput was very, very low very slowly. That has a significant impact on usage. We all know if a page is loading slow, you just move on and find something that will be faster. And so kind of by nice lung while we were there, AKAMAY put one of the content delivery servers. They put a server in the country and were kind enough to share some of the results with us. What was the impact on throughput and usage and they were quite significant. So the upper right hand where there's all that green, that is a speed that the throughput that users were getting. 90% were in that light green and that's about 500 kilobytes per second and then the darker colors above are higher speed. So almost everyone was getting very slow throughput. The day that they turned on the server, you get to the left‑hand side. The green part has shrunk. Only 50% are getting 500 kilobytes per second. Everyone else is getting faster speeds exact same access, exact same content, but it is hosted locally. And some people were getting or 5% were getting 20 mega bits or higher which is equivalent of what you get on 4G networks here. So that's understandability. So that's an amazing change in speed just by putting the server in the country. They didn't tell anybody. By the way, it is going to start going faster and then they showed that usage doubled. They said usage doubled within three months. Purely based on the higher speed. And that's the message that we're trying to get across now to the developers. If you're making content for people to see it whether it's because of interest to them or you're making money off of advertising or something else, bring the local and even if it costs a little bit more will double, triple your speed hopefully and that's a powerful message and that's the way we can get around this externality that the developers, of course, want their pages to be seen. Otherwise, why bother.
So in the course of the project, we came across, you know, a couple recommendations. There's more in the report. So the first thing we noticed is that there's a real lack of awareness of hosts abroad. Particularly in many cases, there was someone in the middle. Someone develops a small website or pays someone to develop their website. They put it abroad and we found cases that the person who owned the website had no idea where the website was sitting because they developed and just ships them off to London or even some in Texas. So there's just a general lack of awareness of what the impact is and that it could be faster and it could be better locally. There was some perception and some fact that the local hosting just didn't match what was abroad. I think some of it was based on older assumptions and one of the interesting things is put everyone in the room and the developer said that the data centers are no good and the data center said we've been doing this, this and this. So some of it is just awareness and getting people over that hump. The data centers clearly need to work on capabilities to make sure they can match what is offered abroad. In some countries, it wasn't an issue in Rwanda and it wasn't an issue for local content providers. But for some international content, there can be some issues around liability if you start hosting content locally versus abroad. So really this is where we came to on the first report and I'm happy to say that the government and the developers seem to have bought into this and now they're doing a second phase where they're going to start helping to subsidize, but not require, but help subsidize bringing websites back into Rwanda and we're starting to gather data to have a before and after picture. I'm hoping next leave to present those results to show what happened to the local Eco system when everyone started bringing their content back home. Thank you.
>> CHRISTIAN DAWSON: Michael, thank you very much for your study. I think the insight you have here about Rwanda would play out in some of the other regions elsewhere. My background is a Cloud guy. I can tell you when you were talking about cost savings, I completely understood that component, but it did all come down to latency. When you improve that customer experience, you improve the experience of the person who is hosting the content. The entire Eco system happy and that has a multiplier effect. The simple fact we all understand that nothing can travel faster than the speed of light including content over a fiber line. And so since that travels in the speed of light, keep it shorter by making it local. People tend to understand that.
Luckily, we have David Vivar who can give us a local perspective on a local story and developing infrastructure in this area, building a data center in order to do to ‑‑ to be able to host content to local users the. I was wondering if you can take the mic and start telling us about the story.
>> DAVID VIVAR: Thank you for the introduction. We had quite a personal experience that actually mirrors a lot of what Michael has shared already with us. A lot of our user base actually started hosting from our data center in Orlando, Florida. And due to a demand, high demand of those users to have the content locally hosted to improve on performance and the issues which are very troublesome as their operations go over time. So that was the driving factor and continues to be the driving factor and host operations and infrastructure here locally. We started with things to begin with catering to users and over time and that base grew. We sought out to develop plans to develop better infrastructure that was more resilient by certain standards that were known by certain organizations like other places and obviously learning from the bigger thing like Google, Microsoft, et cetera. And they have best practices and had been working and had been building it here locally JOAO PESSOA. It is instituting standards and we hope to continue to pride the services with user base that have had or have amended operations to improve the actual content early to the end users.
>> CHRISTIAN DAWSON: Fantastic. Thank you very much. So the first question that I want to do that I want to pose to our panelists down here and then jump to our remote panelists as well in order to allow you guys better framing on this particular subject and also if you haven't spoken yet, introduce yourself in the context. At which point after we have done that, I will give you a chance to talk and I'm hope think for future questions we all treat this as a workshop in anything that I ask or that we discuss we can have an open forum and anybody can speak and be a part of this conversation. What I'd like to do for this very first question is ask: How can we make ‑‑ you showed us the importance of making local hosting a viable option Amidst other global options, but how can we achieve that goal? How can we make that transition? Michael, would you give us your first take on not just the need, but some actionable house ‑‑
>> MICHAEL KENDAY: How we can make it affordable?
>> CHRISTIAN DAWSON: Basically not just convince people, but there are a lot of challenges. You have to ‑‑ you want to do more local hosting. The barriers to entry in different communities are going to be different. In some areas, they're going to be raw infrastructure issues, unreliable power, unreliable sources of funding. And I guess there is some extent to which it really is ‑‑ your answer is going to be different for any community you enter into. But you still have core compelling arguments that you're going to carry into each community and say this is why this is important. This is why you should invest your time, your money, your resources into developing those sorts of things and what are those core arguments.
>> MICHAEL KENDAY: Okay. That's a great question and that's why we're trying to demonstrate that it is good for the content. It travels to the speed of light. The speed of light to Europe is 50 milliseconds and it is the fact that it is so congested. That's how you get up to half a second speeds because in peek times, it is very slow. When you think about trying to compete with a huge data center in London, there's a few things that, you know, first pop out. The cost of power that you mentioned and back up powers are hard to deal w I think an argument can be made for the government to, you know, help subsidize the cost or find alternatives for the power. And then there's the scale issue that has to build up and I think that one thing is that the government should try and use the local data centers rather than building their own to help build up the scale and help build up the capabilities. There's a key training component. A lot of the people we spoke to said that they just felt that the type of services that they were offering people locally didn't have the capabilities and to give you an example, we talked to a company that did live streaming of conferences in Rwanda and they looked around to find capability. They were using a company in Colorado to process it and send it back to Rwanda. So light speed with all of that. That was a skills issue. They clearly recognized is they to try and move it local. And then just this awareness that we've been trying to work on that is a big issue. And then finally, one issue that we were talking about. Typically in a lot of countries starting out, there's no big independent neutral data center. It's usually the ISPs that are hosting the data center and we said to them, you know, you're imposing an externality on yourself. You're raising the price on the data center meaning you are willing to pay content that doesn't go in your data center. Why not lower the cost of the data center and save a lot of money on bringing the content back from Europe because your prices were too high. So I think there's a lot of parts that have to be addressed to deal, you know, the perception of the skills, the actual capabilities and then just awareness of the issue. We're trying to do our part and then panels like this and other things help.
>> CHRISTIAN DAWSON: I will simplify the question which is basically: Do you have any initial insights on how to overcome the challenges in developing countries?
>> A few of the different possibilities that I would offer to think about how we spur, the infrastructure development ‑‑ first thing we need to do is connecting whatever you prefer. It is a difficult and B, requires a multi‑faceted approach because the underlying challenges are complex and change a lot by country and region. Remember people to the full diversity of the internet. We have seen increase efforts in the name of connectivity and of access to connect people only to some small parts of it. We find that deeply concerning. We believe that access of the web spurs and economic social developments and spurs opportunity and education and spurs for people to create on the web. And this is touched on a bit, but this idea that the provision of local and relevant content can be an incredible motivator and I will come back to this in a second. So we need to have that creation of local content. But I think there is also a virtuous cycle that can happen here. By incentivizing and creating new pathways of opportunity to create local content, that helps to drive demand for local infrastructure as well. I think this is part ever the larger Eco system in order to increase connectivity and meaningful levels of connectivity. I think when we talk about the traditional barriers to access and things like infrastructure, affordable, capabilities and incentives, we have a couple things that we're doing at Mozilla to try different things and pieces of this. One is to try and tackle some of the affordability and changing some of the economics as well as try and change the incentives for objects. One is a partnership with your phone in Bangladesh. They have 30 megabytes they can use on whatever they want. They can do that 30 times a day. And then another partnership is with the Cliff phone. Almost 20 middle eastern African markets. The user gets unlimited talk, text and 500 megabytes of data for 3 to 6 months depending on the market. And so it is to give people a chunk of the internet, a taste of the internet and demand the rest. That's what we're trying to test out. We're doing things like that that are very product focused, but I think we're also doing a lot in terms of skills. I think that, you know, we have seen a certain amount level of content and one of the reasons Facebook is so powerful in the internet access is because they have been relevant. I think when we look at things like Bangladesh, we did a small scale study. We want to do more around. This Bangladesh, we took the ‑‑ first we start off with research and we asked people you have connected to the internet and they said no. I don't have my Facebook account yet. No. Really. Have you been to other websites. And they say no. I don't have my Facebook account yet. It is the need for digital skills and literacy. What we did on another study in Bangladesh is we gave the control book the normal on boarding string when you buy a mobile phone and then we gave the test group that normal operating experience, plus a two‑hour monitoring with the web. How is this all working and what we saw and then we gave them two gigabytes of data for free, which is a lot in Bangladesh in exchange for coming back a month later and telling us what they did with it. And so we saw that 20% higher data usage by the group that had the digital skills training, we saw them exceed that free cap we gave them and it helps to paint a picture for further investment. We saw as more diverse array. It is an investment and scape ants and I think it is a really untapped potential in terms of thinking through this sort of Eco system approach to build connectivity to build infrastructure, to bring people online. And then finally I'm going to give all my secrets away. I think that we have an opportunity in looking a little lower in terms of the power of Smartphones. We just released a study with GSMA and the power of Smartphones to spur power of development. Smart phones have so much more potential for creation and for enabling people to make things that are local and relevant and, you know, we need to be designing for that. In terms of principles and user central design, we need to be making a shift from primarily push applications and sort of just sending to these communities to enable that creation and participation online. And I think that what we found in our research is that there's a huge lady end appetite among users to create original content. And that these markets have not yet been sensitized other than publishing to their friends and families. They have a real opportunity to design thing. I think that one way to approach this is really pushing more on the capacity to create, innovate and that brings us greater connectivity. Thank you.
>> CHRISTIAN DAWSON: I think that is a fantastic acknowledgment that even I sort of laid out at the very beginning this sort of traditional view of what the internet's infrastructure looks like from a data center centric environment that the evolution of the internet is making it so this is changing. It is broader. It is richer and so much has to do with mobile communications. And the things that you're talking about just illustrate to me the complexity of the situation that we've got, but also the importance of getting it right and getting it right on tomorrow's terms instead of today's terms. So I really appreciate that insight. David, you talked to us a little bit about the development of your own infrastructure, but ‑‑ we have now spent a week nearly in your glorious city and I am really enjoy its and your beaches are absolutely beautiful. This is obviously not a developing or region, but you certain probably have some challenges that you need to overcome as you have been developing your infrastructure resources and I would love to see if you also have some insights and recommendations for developing reasons.
>> DAVID VIVAR: Yes. One of the major ones that stands out has been connectivity and it is part of local service providers and ISPs that have capacities to commit to before JOAO PESSOA became the site. We created a site for the data center. And after sharing things of evolution that went back from our starting point a couple years ago, about 2 and a half years ago is when we took quite a bit of notice in development that had been around and that (inaudible) as far as local infrastructure with ISPs and connectivity and they're playing with the network like level 3 to (inaudible) and having that study and evaluate it in the capacities to decide what to control and what is going to be a better site for us. Infrastructure connectivity ‑‑ they have things provided by the local utility was driving back there. Since those are the two things we really count on, everything is going to be booked on site. The infrastructure, the resiliency can be built on site, but the two main factors that (inaudible) are having a right power infrastructure from the local utility and having the connections to the rest of the country. There's a lot for that matter.
>> CHRISTIAN DAWSON: After we go through this first round, that's something I would like to get into a little bit more with the crowd because that's so vital. The core resources and how you address core resource issues in developing countries I think is something that we need to explore further down the road in the workshop.
>> DAVID VIVAR: Yes. There's a lot of studies that have been out there available through a simple search on Google. The organizations are out there that have a hand on exposure and experience to the subjects indicated in the field and in the industry. (inaudible) they actually had global resources. My personal favorite was developed from the local telephone days and it stays best practices from day one that had been getting evolved. We have people in our organization that really provide the ins and outs and have the experience and hands on and can share that knowledge with anybody who wants to learn it.
>> CHRISTIAN DAWSON: That's great. Now I'm really excited to go on a participation channel who is also somebody who can speak to hands on experience in developing internet infrastructure. Fiona, would you be able to walk us through your story and talk about your challenges and maybe some insights and how to best approach this problem? I show Fiona as talking.
>> Audio.
>> CHRISTIAN DAWSON: Fiona, I'm sorry. We're not hearing you well. Once we figure out how to address the audio and I believe that it is an audio level issue. Okay.
>> Can you hear?
>> CHRISTIAN DAWSON: Well, is this ‑‑ oh, wow. I did. But ‑‑ so, is this something we can work on and move on for now?
>> (speaking).
>> CHRISTIAN DAWSON: Hello, Fiona. Go ahead and chat with us at will. I'm terribly sorry. We're still having volume issues here. So I'm actually going to go.
While we address that, I will ask Robert to help us there. What we will do ‑‑ I will jump into a couple of other questions that build off of what we're talking about now. As soon as we get the ability to Lupe, Fiona and Baraka, I want to get their thoughts immediately. But jumps into specific questions, the first one I have jumps into a study where you talked about how bringing in a CDM made all the difference in that study. Now, does a CDM, does it undercut the need for local content storage? Because a CDM then takes a content and replicates it everywhere. If you as a provider that is hosting your content in London or in the United States somewhere else can then replicate it into your local community, do you no longer need to host it locally? That's a question for everybody.
>> For the little local content providers, that never really came up. They can't even afford to mirror. They really have to have it in one place or another. So the CDM was really bringing the international content. So, you know, I think AKAMAI has updates. Microsoft has updates. They carry parts of other websites, like Facebook. The business model would support taking a small website and then distributing it and they don't really need distribution everywhere. For some of them, 80%, 90% of their traffic was going to Rwanda anyway. So they didn't need to think about distribution.
>> CHRISTIAN DAWSON: How many people in this room hosts some kind of content somewhere? Even if it's on YouTube or somewhere? Content you created or own in some way. Anybody here use a content distribution network in order to do it? Accelerate the content in that way. Some? There you go.
>> Any good news, Robert?
>> Hello? Hello?
>> There we go. Hello, Fiona on the video conference. I'm sorry we're having problems. This is Mike from Google and we talk a lot about CDMs in the past. I work for Google and I'm involved in CDM stuff as well. As Mike was saying, small content hosts generally don't use CDMs. They keep their content in one place. It is more generally scaling your content and making it available all around the world. If you do have some sort of global platform such as YouTube or blog, you can take advantage of Google CDM or if you use another company, you can use a CDM to help get your content to many places. I think local content for very local audience is better is to have the content where the users are rather than trying to replicate it all over the world.
>> CHRISTIAN DAWSON: One of the reasons I am excited to talk to you, Fiona is she as an exchange point. I think often they're seen as the fundamental building block of Internet in a local economy. Do you that and branch out to connectivity throughout a local community and that's something that there's been a lot of conversations around at IGF. I think that moving on to a data center centric conversation in a more holistic view of the internet infrastructure is what we're hoping to bring as sort of new conversation into this area, but a lot of the same challenges that you face setting up in IXP is the same. You may be talking about setting up in the same location. Before I move on to the next question, let's see if we can bring Fiona and Brock. Anything yet?
>> Apologies. If any of you are on the Webex, please make sure to mute your sound. Fiona, if you have headphones, please use them. That may have been causing the audio feedback. I ask you in the back if you can connect, Fiona, again with audio or just video so we can just try to get her, please. Thank you. In the mean time, we're having issues with audio and video. So what we're doing is defaulting back to slower internet days and we're asking Fiona to answer the question via text and we'll read out the answer. Thank you.
>> CHRISTIAN DAWSON: So I have received a few notes from Barak Atiendo. He would like to ‑‑ he would like to share with us that there are four points that he'd like to go over with us today in addressing how local infrastructure is local development in this region. And that first point is that one will contribute to its capacity building and the years have been trained locally instead of relying on foreign experts. Number two is focusing on contribution towards local economies by creating employment opportunities, which is infrastructure deployment. Three, contributions socially by insuring that the youth who are the largest consumers are engaged to develop solutions and services. And number 4, the development of informed societies due to on board of citizens. I think that is very insightful looks as to why we're talking about here is extremely relevant and why we should focus in developing regions. So thank you, Barack, for those notes. As we are continuing to have these issues, we would ‑‑ we would love for either one of you to be able to type in any additional comments and we would be happy to read them aloud. Remember this is the workshop format and the things we're discussing here we would like to talk about inclusively. The next thing that I would like to focus is what David brought up. Having effective internet infrastructure, you need have effective resources that go into it. Generally we're talking about power, power that does not go down or power Fit does go down, you have effective means of making sure that your data center does not go down. That may be having effective batteries and generators on hand. If you have generators, making sure those generators can access easily and continually the fuel that it takes to keep going through any planned or unplanned outages. And those are sometimes big barriers to entry for certain communities. And there's no real easy answer to this. But the question that I have is: Are there ways in which internet infrastructure providers look to developing in local economies can influence that? Developing relationships with power companies trying to make sure they build out the core infrastructure they need so you can build on top of it.
>> Building that relationship and having communication with those sources for them to build a capacity for the needs and the demands that will arise as you see the path to your end goal. Having that right relationship with the right source is a big help. That in addition to having to build the resilient infrastructure that you want to build and building a relationship with local vendors and they need it have the original infrastructure in place. Also a big plus, other things to bear in mind, in certain countries, there's a high import tax like here in Brazil. Keeping that in mind and turning to economize one of the major quests to build the data center is by trying to acquire strategically anything that's manufactured locally, you can bypass those high taxes. So those are all the driving points to build something effectively and cost effectively to achieve the end goal.
>> CHRISTIAN DAWSON: That's great. I do have a message from Fiona. Actually, I think that she's continuing is to write. So what I'll do is I will relay one story quickly and let her complete her thoughts. I will say something that dub tails from that and then I will go ahead and read her statement afterwards. So you're talking about going into an economy and telling them that what you're doing is important. That the raw power infrastructure should listen to you, that the people who want to fund you want to listen to you and that certainly has its challenges too. I want to tell you guys about a story that happened to me just a couple of months ago where I was having a conversation with a high ‑‑ a high powered United States government official. Worked for a state in the United States and I was telling him what it is we did with the Internet infrastructure and I mentioned the word data center and he said to me okay. I'm going to stop you right there. Governments, especially local governments don't care about data centers. And I said why. Well, they don't create new jobs. They take up all the power in the region and they're filled with things that are bought from foreign places. So we don't care about them. I think that there was a fundamental lack of acknowledgement of the incredible things that come out of all the layers of what happens in a data center. There may be very few employees directly in that data center, but the amount of business that it empowers, the amount of income that it empowers from all the things that happen as a result of what happened in the data center was not acknowledged in that comment, but it's a challenge that we have wanting to develop that ‑‑ those kind of resources to explain to them that the benefit and the impact of what happens in a data center goes well beyond what happens within its walls.
>> DAVID VIVAR: We can attest to that. It is major impact on a local community with a local region for that matter of the state or the country. Not everyone has the opportunity or the right resources to build infrastructure and by enabling the local region to host locally, you enable other people that somewhere along the stream benefits from it whether it's local or indirect or simply the ability to provide the service and low latency which drives up the operation and the demand for those end resources that operate ‑‑ that have those operations and promote local high level usage. You have many spots that are hard to see from a depot or at a first glance just by the word data center. But there is also many other areas that are impacted by the operation of that local data center.
>> CHRISTIAN DAWSON: Jim came up to me and said we can try Fiona one more time. So let's give it a shot. Fiona, can you go ahead and talk. I'll type it here as well. She has given me a statement. I will go ahead and read it. It's not working? Okay. That's fine. So from Fiona. Local infrastructure is building the local economy in Kenya, the development of the IDP became a core infrastructure in the internet Eco system to a point that its existence had impact on communication and procurement government policy. Its growth has significantly influenced the push for localization of content, Google and AKAMAI. KIXP has not only received support from the government, ministry of ICT and regulator. This relationship is attributed to the government to facilitate infrastructure development that is supported the IXP and the Internet Eco system. Okay. (no audio) (audio dropped)
>> How can you manage it fully. It is all margins, right? You kind of have a very good relationship. It is their expectations and in terms of substations and powers that are certainly and did helps us to (inaudible).
>> CHRISTIAN DAWSON: Great. Thank you. There is another question or comment, please.
>> I will (inaudible) we also involved in a project in the U.K. and we've been awarded TV wide face is and they're regulated and television switched from analog to digital and freed up compassion ‑‑ they will access this possibility and it can be used to sort of access CD and for remote communities that don't currently have to have broadband access. They connect it to this TV wide space and have a large (inaudible) barometers and it's greater and flexible for not having that line of site. I wanted to flag that and see if you have the possibility of usage.
>> Great. Thank you for that and also as we shift more to digital, what are some opportunities that come up. Are is there any other ‑‑ another question that was great to hear some feedback from, I think the other thing too is the issue of taxation has come up. So I might suggest that when you wrap up, there's recommendations and maybe what is done elsewhere to try do that would be particularly helpful. In terms of global connect, Mike spoke with (inaudible) yesterday. He was becoming about it. There's a variety of different initiatives. The IGF comes out of the outcome of the two and it has always been any unanswered question with regards to funding. It's been really difficult to get governments and others to try to support initiatives. But governments have come forward in what I would call like‑minded efforts to try to address some key issues and try to build coalitions to push things forward. There are coalitions that have worked around free of expression and human rights and they recognize other key aspects to make sure everyone is connected online and there's also connectivity. So being able to put together a global network and I think something they were telling me is that yes it is in U.S. But there are order countries on board to try to support that. So probably what we will see is a time of funding and larger coordination with international bodied to make that happen. What I say for folks here is how they have ideas that falls under that to get it in sooner or later to make sure it's a success. Before I give the floor back to Christian, I want to hear from all of you with regards to last mile issues. A couple things we talked about here is if anyone wants to make a comment about it or maybe (inaudible) that sort of success that again Brazil, Latin America region or elsewhere. Are there any folks that want to make a comment on any issue? All righty. Christian, over to you.
>> CHRISTIAN DAWSON: And more broadly, before we move up to cross the panel, if there are any additional comments that people want to make in general reactions to any specific thing that we discussed hire today, we'll do that before I move to wrap up. Okay. Well, I'm going to start to wrap up. I'm lucky enough to have wrap‑up statements from both Barack and Fiona. I'm sorry you didn't see them both today. Barack shared with us, as well to protected infrastructure from vandalism. Academia is educating the populous and creating customer awareness. That's the statement he wanted to leaf us with. Fiona said government policy is ‑‑ in Kenya, we've been able to work on infrastructure. Since we have not made the headway we would like to achieve, there are a number of private partnerships in Kenya that have facilitated infrastructure development. Why don't we move to last spots for the rest of the panel. Maybe we'll start with you, Michael.
>> MICHAEL KENDAY: Sure. Recently the U.N. broadband commission, the EU came out with a report in September that said broadband internet growth is slowing down. New user rates are slowing down. It is even more alarming when you realize that it is not because of access. It is far more people that could have availability through the mobile broadband. And I think we have an illustration here of the challenges. It is not just this linear ministry of the ITT and the regulator, the companies and the spectrum and everything in place so they can offer service. That's already been done in a lot of countries. It can be expanded through wide spaces and other things as we heard, but really now we're starting to talk about hinges that get more and more challenging and need more and more government and private sector and civil society involvement, local content generating more local content, hosting more of it locally, the taxation issues are hard for the ministry of finance to give up. Getting the skills. We heard some interesting examples of teaching people and the benefit that has, the power issues. So there's really a lot of things and it is not just one ministry and one set of company that can solve it anymore. It will really take the whole government and the whole private sector in civil society to really pitch in and solve this and hopefully not let the slow down become the new normal, but try and speed up the new development so we can get the 60% and beyond.
>> I think a comment and a question. I had a report showing that taxation can often make up 20 to 40% of the total cost of owning a global phone. I think stressing the importance that government embraces policies friendly to consumers is virtually continuing investment. One statistic that I find interesting with the mobile side of this is there are some 250 million Android devices out there that are not connected to the internet. These are data capable, but not data enabled phones. And so we have seen the price of devices falling and data falling. We have seen coverage, you know, 3G increasing at a 4G level and exploding. There are things that need to be done on infrastructure affordability. But I think as we look to sort of potentially get people online, we increasingly come back took this question of skills and of content and of literacy and designing for that and promoting that. And again, I think that is encouraging more demand that happens to put on the pressure on government and on industry to build out that infrastructure and have the appropriate things to do so.
>> Coming from the power sector, having greater incentives and greater help by federal government in terms of exceptions ‑‑ excuse me. Exceptions and incentives to build locally would be a greater impact. The more infrastructure that can be built locally, the more you can promote businesses that affect the local community directly. It makes them very troublesome, very expensive to come and build here is the majority of the infrastructure that is imported was twice the amounted it would normally cost in a developed country. It doesn't make sense and (inaudible) personnel with regard to that knowledge and having that ‑‑ having that comprehension would help a lot. If the data can be hosted locally, that means the services need to be taxed locally which is actual revenue the government is using on. They need someone who will host their data and it is hosted abroad and that server is paid abroad without the local sales tax. Then that's what the government is missing on is they can capture that and getting that understanding. I think they can see the bigger picture and they incentivize the growth and promote it better because overall, everybody has something to do (inaudible) local.
>> CHRISTIAN DAWSON: Robert, would you like to share any final thoughts?
>> ROBERT GUERRA: There were things at the local level and which are key and crippling in some ways. There is the data center that picks up all the power and nothing else and what does it create and things that kind of came up was that it's a bit more nuance and what's key is having data, but also utilization at the local level is that there also needs to be a coordination with regards to digital skills in terms of improvement engagement and this training will also drive up a local access. So there's this other component that also needs to be there as well and that those are probably partner organizations, NGOs and other stakeholders to be aware of the internet. The governance is taking place and so I knowledge let's not forget about that and that's the way to perhaps involve those other stakeholders as well. I will turn it over to you, Christian.
>> CHRISTIAN DAWSON: Thank you very much and I would like to close this out with this thought. The internet is awesome. We are here at IGD probably all of us because ‑‑ IGF because we see this as this incredible life changing, world changing tool. Many of us may be here because we see problems with it we want to fix, but ultimately, we know it is incredibly powerful and incredibly meaningful and it is something we need to cultivate. What we need to know about the internet is that even about we use words like the world wide web or Cloud that make it seem like it is just in the air, the Internet and its infrastructure is made up of real hardware in real places and then real people doing real things in real locations. And we don't think of the work or the mobility of work that goes into the internet and making it run and making all these different components of it all around the world tick, but the internet's infrastructure really is the heart of the Internet that we're all here to protect and grow. And that heart isn't always in the local economy. Think about how you can bring that home. Think about ways in which you can develop the internet's infrastructure where you live and make sure the benefits that Michael illustrated that come from developing local infrastructure can be a part of it. The speed, the cost benefit. Is there are good reasons to engage in the kind of things we were talking about. We hope that we will be able to share some ideas and thoughts that help you carry things on past IGF that will allow you to see where you can take the internet infrastructure in your local community. So thank you so much for your time today. I really appreciate it.
[APPLAUSE]
***