Development

Panelists discuss the prospect of an inclusive internet, how the technology industry is reaching and interacting with developing countries, and the policy implications of a connected world. HAMMER: OK. Ladies and gentlemen, welcome to today’s Council on Foreign Relations meeting. The topic is “Connecting the World: The Internet’s Next Billion Users.” I am Craig Hammer. I’m a program manager at the World Bank, and I’ll be presiding over our discussion this evening. Our discussion will take two parts this evening. I’ll kick things off with a facilitated discussion until about seven p.m., and then we’ll open up to members for a thirty-minute Q&A, and we’ll end at 7:30 sharp. Let me begin by briefly introducing our distinguished speakers by their titles, since you have their detailed bios. Robert Pepper is head of global connectivity policy and planning at Facebook. Jen Spies is product manager of the Next Billion Users Initiative at Google. And Mike Pisa is a policy fellow at the Center for Global Development here in D.C. Welcome. The implications of adding a billion users to the internet from across the world from a range of cultures and contexts, with all their infinite complexities, is vast. The pendulum swings a full 180 degrees, from the tremendous potential of the social and economic development implications of internet connectivity to the truly terrifying: from the collision between social media and surveillance, to the incitement of genocide through social media platforms. My sense is that we’ll cover a lot of ground in our discussion this evening. But I’d like to begin on a slightly more positive side of the spectrum. Jen, Google’s Next Billion Users Initiative is making products specifically with emerging markets in mind. Talk about the local benefits that can occur with the expansion of large digital platforms in emerging markets. SPIES: Sure. And thanks for having us tonight. I think from Google’s perspective on of the primary benefits of building for the next billion users is that in 2019 if you can make a product that succeeds in emerging markets, it’s really a product that can succeed globally in scale. And I think one of the key factors contributing to that is building a product that works on a mobile device. So I think one of the key trends that we see in this market is what we call the leapfrog effect. So if you think about who are these users getting online today, they’ve often never been on a desktop PC. If you think about my internet user journey, I came online first on a computer, and then around 2010 there was this huge shift to mobile where people actually started spending more time on mobile than they did on a desktop PC. And actually, the internal team structure of these companies, Google and Facebook, was completely pivoted to accommodate this shift in time spent that was seen. If you think about, you know, a twenty-year-old coming online in South Africa today, they’re coming online through perhaps a $20 smartphone, and that’s their experience of the internet. And so understanding how to build products that are mobile first and that are native to that platform is really critical to winning in NBU markets. So I think that that’s a key component that can contribute success both in the U.S. and in these markets. HAMMER: In terms of the local development, though, I mean, speak to what NBU is thinking about in terms of local economic growth and social development. SPIES: Yeah. One of our key tenets is to grow economies in NBU markets, and the internet is a true democratizing force that can help do that. I think if you think about—one of the really big engines of growth we’ve seen is small business empowerment. In many regions in NBU you have shopkeepers coming online. Oftentimes they’re inheriting these businesses from an older generation, and they really understand the power of digital and the power of using their phone and internet platforms to sell goods. And so using platforms like Google and Facebook and WhatsApp to help find new customers, to grow their business, to have secure payments and secure transaction is really something that we’re seeing happening in NBU markets that’s sort of growing the size of the digital economy, and helping local economies, and helping people participate in the digital economy in a way that is accelerating every year. HAMMER: Robert, Facebook is working with governments and others around the world to extend connectivity to the 49 percent of the global population that’s not yet connected. Can you talk about some of the major issues or barriers that you see and what’s happening as you address them? PEPPER: Sure. Thank you. And again, thanks for putting this together. So the 49 percent is a reference to the fact that there’s three-and-a-half—3.8 billion people connected and 3.8 billion people not connected. There’s a little bit more—the ITU just came out just recently and said it’s just crossed that line. That’s a real milestone. So we think of it not as the next billion, but as the next 3.8 billion. And so—and the reason that that’s really important goes to some of the reasons that Jen just talked about. Each year now for the last four—we’re into our fourth year—we do a study with the Economist Intelligence Unit called the Inclusive Internet Index. And it looks at over time 120 countries, fifty-three indicators for each country, looking at availability, affordability, relevance, and readiness. And one of the things that the Economist found in the most recent study, released earlier this year, was that the progress that we’ve seen over the last decade seems to have stalled out for the lowest-income countries. In other words, the upper-income, upper-middle-, and lower-middle-income countries are continuing to improve; the lowest-income countries did not improve year over year. So you look at that, and then you combine that with something we do as part of that project, a study called the Value of Internet Survey, and each year the theme shifts. The theme last year that we looked at was how do people view the internet in terms of livelihood. And this is across the globe, so this is, right, low income—I mean, it’s all one hundred countries. It didn’t make any difference: 75—73 to 78 percent, so call it 75 percent of people in a country said the internet helped them get a job, they used the internet to help them do better at work, they used the internet to learn more as it relates to their livelihood. If that’s the case, the fact that we have this, you know, stalling out at the lowest income, at the bottom of the pyramid, that’s a real problem, right? We don’t know whether it was a one-year data anomaly. We’ll find out when our—we get the latest data in. That’s bad enough, but if it’s a trend that’s really bad. And that’s why, if you’re looking at the economic development and the benefits—whether it’s small business or individuals or healthcare, education—it can’t be about the next billion. It has to be about everybody. And that’s why we’re looking at it in that respect. So in terms of the barriers, going back to the analysis, the barriers to getting another billion, two billion, four billion—by the way, out of the 3.8 (billion), probably a billion and a half are children you don’t want online. So it’s not quite as big a number, but it’s still important. It’s a combination of supply side and demand side. The prerequisite, you need the network. You need the connection. You need a minimum of 3G, preferably 4G connection. It’s all mobile. That’s the minimum that you need. If it doesn’t exist, you have to have it. But that’s only the necessary but not sufficient prerequisite. It’s all about the applications, the demand driving local content, local language, relevant content, including e-gov applications, e-commerce, e-entertainment. Those are all things that you need. And people on the readiness side need to know how to get on safely. So it’s about digital literacy. It’s about knowing how to set your settings. It’s knowing how to prevent not just spam, but fraud, phishing attacks. So all of those things need to be in place for people to be able to get online and then benefit from being online. So some of the things that, you know, we’ve been doing—and we work with government—those are the barriers. We work with governments, using the analysis from the Inclusive Internet Index as a diagnostic, on how to build digital—national digital strategies. But for example, in Uganda, there’s an—multiple operators. Airtel, which is an Indian company which has networks in Africa, has a 3G network. They’re building out a 4G network in Kampala and the cities, but not in the rural areas. In the rural areas, what prevents them from doing that is they don’t have what’s called backhaul, right? They’re using narrowband microwave, which won’t even support a single smartphone, right? They need fiber to the towers. So we coinvested with them and jointly built a 770-kilometer fiberoptic network. It’s an open fiber. We do everything on an open basis. Any of the operators can use dark fibers. They’ve now connected cell sites and they’re converting from 2G directly to 4G. They’re leapfrogging, all right? That’s supply side. That’s a prerequisite. We’ve co-built—we’re building a subsea cable from Brazil to Argentina which is the first new subsea cable to Argentina in eighteen years. The single cable is going to triple the international capacity to Argentina. Again, open fiber; it’s not just for us. So those are, you know—and we can talk more about some of the other infrastructure investments with technologies, but we’re also working on the applications side with zero-rated services, digital literacy, education applications, just like you are. I mean, there—you know, the internet companies all have programs to build local content, local language, relevant content, working with governments to create those kinds of applications. HAMMER: Let’s zoom out a bit and talk about the larger kind of governance context. So, Mike, we’ve had some discussion already of the pros of digital platform penetration to emerging economies. Let’s talk about the greater international cooperation, what it could or should look like. Is some version of some global governance mechanism feasible, from a policy perspective? PISA: Well, so let me start by—I know we’ve heard about the pros, but I think because I focus on governance, and often the discussions about governance kind of drift towards dealing with challenges and mitigating risks, they kind of have a negative—take a negative tint. So before I start talking about the governance issues, I do want to underline, especially from a lower-income-country perspective, the benefits that the things we have I think started to take for granted here—you know, reducing transaction costs, reducing the cost of search, enabling greater access to information—what this means. And I think oftentimes when we’re talking about this through a development lens we tend to forget that some of the benefits that we’re debating here in OECD and rich countries play out very differently in lower-income countries. So one quick example would be the gig economy that shared—that two-sided platforms enable. I mean, here we have concerns about what gig workers—are gig workers losing out in terms of the level of informality attached to their jobs, the level of worker protections? Whereas in poorer countries this is not a debate, right? First of all, those worker protections don’t exist. And they actually can be a step—and most of the economic activity is in the informal economy, so there’s actually a step towards formalization by just participating on one of these platforms. So I think another issue is that these benefits are quite diffuse, so they tend to I think get discounted in debates around the benefits of certain policies. So, actually, I want to start at a national level rather than go into global governance first, if you don’t mind, and that’s because I think you have to understand this from the perspective of national policymakers. And actually, I think where I want to start is the idea of why are governments in the position of assessing or reassessing how they approach data governance? Which inevitably means: Why are they assessing or reassessing their relationship vis-à-vis big tech platforms? And I think it’s two reasons. One is, you know, we’re more aware today of the risks of the misuse of personal data. And, two, there’s a growing recognition of the value of data as an economic input. And so I’ll start with the second, actually. From the perspective of policymakers—and let’s just say policymakers in poorer countries—again, the benefits are diffuse, but they also—there’s also an aspect of very concentrated wealth being generated from personal data. So if you look at—you know, one quick stat is 90 percent of the—let’s just start with market capitalization. Ninety percent of market capitalization of digital platforms in the world accrues to companies in the U.S. and China, and two-thirds of that value accrues to eight companies, and we all know their names. Now, if you’re a policymaker in a poorer country, and you’ve listened to the World Bank and others rightfully tell you that in order to take the next step in the global economy and the digital economy you need to build an enabling framework, and that will lead to wealth creation, you’re going to look—first of all, the benefits, again, diffuse, oftentimes not feeding into GDP statistics for reasons we can touch on, oftentimes not feeding into revenue. And yet, you see the pie growing elsewhere. And to me, I think that leads to—that’s why the data-as-oil metaphor has had so much staying—sticking power. It’s not—we all know that it’s not a very apt metaphor for a number of reasons, right? But there is a feeling among at least the policymakers we’ve spoken to at the Center for Global Development, where I work—and maybe it’s because we talk to folks from ministries of finance and central banks more often than not—but that they’re being shortchanged somehow in this—in this interaction. And I think the worst thing about the data-as-oil metaphor is it leads to this idea that data is inherently valuable or valuable on its own, right? And it’s kind of like the flip of the Hal Varian argument, that the real value that platforms provide is not the data—not the value they derive from the data, but it’s the knowhow and it’s the—you know, the proprietary algorithms and it’s the human capital. And of course, the truth is somewhere in the middle. But once you’ve kind of accepted this frame as data-as-oil, it becomes all about this idea of, oh, well, I have to somehow grasp or hold that data to have a better negotiating position in this discussion. And that leads to forced localization. And I think—I mean, and these localization policies have been around since the internet, but there has been an acceleration of their use and there also has been a change in the rhetoric around them. It’s much more—you see it much more frequently that countries justify the use of localization policy for economic reasons. I think the clearest example is the India—the draft data protection bill, which says we are arguing in favor of localization because it’s going to help India develop its digital infrastructure and its AI industry. And I think—my concern in this space is that right now the policymakers we talk to seem to think that there’s a continuum of policies: on one hand you let big tech platforms run free, and on the other you kind of pursue these forced data-localization policies, and there’s not much in the middle. I do think eventually you’ll have to have some enlightened taxation policies that will fill that middle, and that’s maybe something we can talk about in the future. HAMMER: So, Jen, how are technologies of the internet impacting users in sub-Saharan Africa specifically? SPIES: Yeah. We think about it in terms of a multistep journey. And in sub-Saharan Africa specifically, the first pillar of that is access and affordability. I think if you compare those economies to others in Asia, Southeast Asia, Latin America, data as a percent of income is the highest, there’s the lowest smartphone penetration. And so access to internet, access to smartphones is really the key pillar to participating in the digital economy, to getting online. That’s step one. Step two is what we sort of call ecosystem and platform development. So you can imagine you bought your first smartphone, you bought a prepaid data plan, you’re getting online; if there’s not content that’s locally translated, that’s relevant for you—if you’re going on YouTube to watch celebrities that you care about and you don’t recognize any of the names—that’s not a great experience. And like you said, there’s a lot of job hunting that happens online. I think the behaviors on the internet are fairly similar. Like, people are talking to their friends. They are watching content that they care about. They are searching for jobs or ways to make money. And if that is not comprehensible or relevant or localized to you, then it’s also inaccessible. So a lot of parts of Google are focused on making sure that there are local Indian content creators uploading videos on YouTube and using machine learning algorithms to translate into local dialects and making sure that Google Assistant can translate into vernacular languages that are maybe more intuitive for a first-time internet user to talk to and to access. And I think the third stage, and the pillar that we think about, is platform development and business creation. And so once you have all of these stages of, you know, you’re online, you can afford a device, you can afford internet there is content in your local language, you can sort of participate in the internet as we experience it in the U.S., then there’s this opportunity to create businesses that are localized, that are taking advantage of trends like social commerce, of digital payments, of, you know, people really sort of hacking our products and using them in interesting ways. But you can build an ecosystem on top of that. So I think this trend of leapfrogging to mobile, of really leaning into social commerce and using platforms like WhatsApp to sort of run a business end-to-end on a mobile device, of developing content locally in local languages are all trends we see that are really pronounced in sub-Saharan Africa, but also apply to other NBU markets. HAMMER: Robert, maybe you can speak to some of the public goods that Facebook is doing to help support that leapfrogging process. PEPPER: Yeah. So just picking up a little bit on that. One of the things that we’ve looked at in terms of analysis goes to the devices. Out of the three-point-eight billion people., we believe that there’s about a billion people who actually could receive a 3G or 4G signal but cannot afford the device. In Africa, it’s 240 million people. So I mean, you don’t have to build a new network, right? You have to make the device more affordable, but it has to be a good enough device so that it’s actually either a very high-end feature phone or a low-end smartphone. But that actually works and that’s not hackable, in terms of the device. So you know, as we were discussing earlier, one of the things they are doing is working on that, plus other variables, with the World Bank on its Africa moonshot project, which was just launched. I guess it was at the annual meeting about ten days ago. And if you haven’t—if any people here haven’t seen that, you should take a look at it. It’s a great document. It was actually a working group paper that was released by the U.N. Broadband Commission, put together by the Bank in conjunction with the Bank on the Africa moonshot. And it goes through, you know, what needs to be done. And, you know, it’s—you have to take a deep breath, right? The goal of the moonshot is to have everybody over ten years old in Africa connected by 2030—eleven years, right? They estimate it’s going to cost $100 billion. What’s interesting is about twenty-three, twenty-four billion (dollars) is capex for networks, supply—pure supply side. Eighteen billion (dollars) is the demand side. It’s building the applications, the local content, the e-gov applications. Creating local businesses that do that. There’s a small policy piece and some others. The biggest chunk, over forty-five billion (dollars), is what you don’t even think about, which is opex, maintenance, replacement, right? The really boring stuff that makes it work, right? And so the question is, where does that come from? And that’s going to have to be—and the Bank’s talking about putting a lot of money into it. And the device affordability is a chunk, and that’s one of the things we’re working on. So there are these great opportunities. And in terms of some of the public good aspects, you know, what we’ve seen is that there are literally thousands of businesses that in just sub-Saharan Africa—that exist only online. There is no brick and mortar. So it’s not like there was a shop that then went online. These are businesses that would not exist if they were not online. And they’re selling things not just in their communities but broadly in their—in their region, or their country, and even sometimes to other countries in the region, or internationally outside of their continent. So we know these benefits, but they’re—again, I want to come back on to—there are these, you know, very legitimate concerns and questions. And I think you have to separate—and I thought it was really good the way you laid out, for example, data localization. Data localization in many places is really sort of an industrial policy that, you know, usually starts off as being dressed up as we need it for security. But localizing data actually makes your data less secure. We know that. And the irony, by the way, of the India draft localization bill, if you think about the India BPO industry, their back-office industry, literally it employs millions of people, it’s worth billions of dollars. And it only existed because Indian companies could process data and build call centers for companies based in the U.S. and Europe and have customers in the U.S. and Europe. If there was data localization, right, and it applied in the reverse, that industry would never exist in India, right? They don’t think about it that way. HAMMER: So, Mike, let me ask you one question. I think, just because these issues are very much alive in much of the work where we happen to be, it may not be the case elsewhere around the world, but let’s talk briefly about data privacy, disinformation. As the digital platform market concentration expands, let’s say, to emerging economies, have these issues like data privacy and disinformation manifested in lower-income countries? And how have policymakers in those countries begun to respond? PISA: So I think all countries are grappling with these issues. I mean, I’m not going to go through the list of how they’ve manifested themselves—(laughs)—but I think all countries are grappling with these issues in different ways. I’m actually—since I punted on your question on the global governance I’m going to punt back to that global governance question around these issues, because I think it’s going to point to why governance is going to be so tricky to do at the global level on these questions. So I think you can make an argument around data privacy and data protection. I’m going to stick with data protection, because I think privacy is such a culturally laden term. That there are efficiency gains to be made by having a harmonized approach globally, right, that would allow companies—big tech companies—to apply the same standards globally, rather than having bespoke models for different countries that they work in. And frankly, I mean, you know, countries have been willing—I’m sorry—companies have been willing to change, often at great cost, how they go about their business in order to comply with GDPR because Europe is the world’s second-biggest market. If Benin, or Bhutan, or another country asked—you know, enacts privacy regulations that are quite different and viewed by industry as quite onerous, then they probably won’t make that same distinction. So I think there’s a stronger argument for harmonization in that space. Other rationales for global governance. When you have cross-border spillovers. So I came from U.S. Treasury, where there’s a—you know, a large architecture around protecting against banking crisis spillovers, right? So we have the FSB, we have the Basel Committee, and many other structures. And we don’t have those similar things for digital platforms, even though there may be some spillovers—I think the spillovers are less acute and less immediate. But you can have some arguments around, you know, whether there are spillovers, about how companies managed or regulated in one country affect political stability in another country. And there are more clear spillovers in the areas like cybersecurity. But I think the one argument I want to make is I think the overarching need for global governance in this space is we need to make the experience for individual users, but also for governments of using the open internet better, in the sense that there is now a more clear challenge to the open internet. And actually, I watched Robert’s—his Turing speech in 2015 over the weekend, like we talked about it. He said in 2015, we’re at the crossroads because national governments are reasserting more control over internet policy, and they’re often going towards a more closed model. You said this two years before President Xi in China says, you know, we’re going to make China a cyber superpower, and we are going to treat—we are going to, you know, kind of set China as a model, in his words it was, for countries who want to speed development while preserving independence, which I think is code for closed—have a closed internet system. So I think we’re still at the crossroads. We’ve probably gone a few steps in the wrong direction. But these debates are still happening. And my view is, the reason you want to pursue—you might want to think about pursuing global governance over specific issues areas related to how you regulate digital platforms, is because you want to make governments more comfortable with the open internet model. And right now, there’s a lot of reasons to not be comfortable. Having said all that, I think the challenges of having effective global governance in this space are huge. I mean, just take disinformation. I mean, how countries go about regulating, or monitoring, or treating, or dealing with disinformation is going to depend on how they value privacy, how they value transparency, how they value freedom of speech. As we can tell from debates that have happened here in Washington over the last few weeks, and speeches, that we’re still figuring this out. And why do we have any belief that having, you know, a global discussion on disinformation on digital platforms would be any different than a discussion on freedom of press in the more analog world? So I don’t think we’re likely to actually have productive conversations unless we take very narrow, specific areas where cross-border spillovers are real, and the efficiency gains from cooperation are really clear. HAMMER: Thank you, Mike. PISA: Yeah. HAMMER: So at this time I’d like to invite members to join our conversation with your questions. So let me first begin by reminding everyone this meeting is on the record. When I call you, please wait for the microphone and speak directly into it. Please stand and state your name and affiliation before asking a question. And please do limit yourself to one question and keep it concise. This will allow as many members as possible to ask questions and share insights. So let’s begin here. Q: All right. Thank you. My name’s Dave Harden, and actually my colleague Sara Agarwal is over there. And we run a tech company that has its research and development lab in Ramallah. So imagine kind of all the big challenges that you all have described and think about, and then add the political overview on that. All of our employees have stock options, and there could be a dramatic impact if we were able to drive a big exit. One of the challenges that we have, and this goes to the question of how do you get to the next billion or three-point-eight billion, if they’re only users as opposed to developers, or creators, or value-adders into the internet. And so it’s very easy to pick up rick capital in Palo Alto, or in Cambridge, or in Shanghai. Not so easy to do it when you have operations in Ramallah, and Amman, and you’re struggling to do that. So how do you see kind of the nature of capital over the next—this next ten-year period where you’re trying to increase the number of actors that have access to the internet? Thank you. HAMMER: So I think should looking, I think, closely at Jen and Robert on this one. But I think we can start with you. SPIES: Yeah. I can answer somewhat narrowly, in terms of how the company has thought about expanding in a probably conservative but effective way in these markets. So we’ve actually opened offices globally. And we think about product development offices, which are probably the analog to the startups, and the engineers, and coders, and people building these products. We’ve opened tech hubs in Bangalore and in Singapore, in markets that are just closer than mountain view to our next billion users. So I know there is a balance between rapidly expanding everywhere and trying to be in every market, and have engineers in every market, and trying to build centers where you have sort of hubs and knowledge that builds up over time. I know that’s one way that the companies thought about just sort of expanding slowly and methodically and trying to find a product market fit with various initiatives before expanding too quickly. So it’s perhaps not like on the capital allocation side, but at least internally it’s who we’ve thought about expanding our footprint. PEPPER: No, we think about it very similarly in terms of, you know, where the engineers are, and growing globally. But that doesn’t—I mean, that’s toward Facebook. That’s not really your question. Your question is, you know, where are the—where’s the venture capital going to come from to fund startups in Ramallah, Kinshasa, you know, Cape Town, Bangkok? And that—and that is a challenge. In fact, earlier today I was having a conversation with a former minister from an African country, who is trying—he’s now in the U.S., at one of our major universities, two-year fellowship. And one of the things that he’s working on is trying to figure out exactly that. He was telling me that, you know, there’s capital available in Africa, but the—but the people in Africa who can write the checks are writing checks in Europe, not in Africa. And that really bothers him. The question is, why and what can be done about that. And what he was saying is that, you know, you have to have the right—create an ecosystem in terms of law, local law, and also taxation, bankruptcy law, rule of—you know, rule of law—all of the things that we take for granted—before people are going to write checks, whether it’s at—you know, for a startup, an angel writing a small check, or a VC writing a large check. And that’s not always easy. So this is—this is one of the big questions. The only optimism I see, right, near-term optimism, is if you go back ten years, fifteen years—no, not even fifteen. Ten years, when I was having meetings with startups in China, when I was—actually I worked for CISCO then—they were asking the same question, right? They now have loads of money, capital markets. You went to Europe, you went to Berlin and talked to startups, they were struggling. They couldn’t get capital, right? At that point it was beginning to go to London, right? So I think that they’re—you know, the optimism is that it is happening, but it’s happening too slowly. But that is an issue, because we see huge talent globally. Huge talent—for example, there are startups in—and incubators, for example, in Kenya that are struggling to get capital to get to the next stage, right? It’s not for the lack of talent. It’s not for the lack of effort. It’s not for the lack of market, right? And so it’s—this is something that as a broader global community I think we need to—need to address. HAMMER: Let’s come here, please. Q: Thank you very much. Miriam Sapiro from— PEPPER: Put the mic closer, Miriam. Q: Oh! Can you hear me now? Is that better? Closer? I don’t—I have to eat it, like an ice cream code? How’s that? HAMMER: There you go. (Laughter.) Q: (Laughs.) Mmm, delicious. Anyway, I want to focus for a minute more on the challenges because in some ways part of the conversation hasn’t really changed much in the last five, ten, fifteen years. Pepper’s nodding. And so, whether we’re talking about capital or infrastructure, we’re also talking about governments that either need to get out of the way or, better scenario, is to facilitate the kind of investment that’s needed. And yet, in the last few years, we’ve seen how in the developed world, especially in the U.S., the darker side of the internet has become more obvious. Whether it’s influencing elections, or cybercrime, or extremism, et cetera. And so how—you know, what do each of you, in your different roles, say to the governments, especially middle-income and lower where we really do want to be able to help more people get online, what is the ability that you think we have to try to make a difference in that regard and build, if not a governance structure, which is proving very challenging, at least some ways to acknowledge that we’ve learned some lessons from what we’ve seen in the developed world. And we will be able to try to help the less-developed world learn from them. PISA: So maybe I’ll start on that. I think—so the Pathways for Prosperity Commission from Oxford University Blavatnik School just did a survey of developing country policymakers, and what were their priorities around digital technology and the internet? And their first priority was job creation and the second was digital infrastructure. And much further down were these issues of data protection and data privacy. And then cybersecurity was up near the top, and also revenue collection—which I kind of alluded to before, right? So they want to see some direct benefit into their own coffers, but again the highest priority for poorer countries—at least in the policymakers in that sample—were: How do we make the digital economy work for us? And I think the trust issue—you know, we’ve—my remarks have focused on governance, and I’ve talked about why I’m skeptical of a broad-based global governance framework. I think one step could be the U.S. taking useful steps to regulate the industry from here. And then the other, and I’ll turn this over to you guys, is the large platforms convincing their users and the governments where they operate, the countries where they operate, that they’ve kind of corrected from, learned from some of the missteps in the past. SPIES: And I can give a really specific, but hopefully enlightening, example. I’m a product manager on Google Station, which is Google’s free public wi-fi initiative in emerging markets. And we generally have, like, the most success when we go into a market and at launch we are partnering with local governments and, you know, the minister of telecommunications is on stage with us at launch championing the arrival of free public wi-fi to his country. So we like to look at this as a partnership with local government. And often we find partners in government who are really excited to bring this product to their country and champion it. I think on the—on the technical and regulatory aspect of it, we try to hold our product to the same data standards as GDPR, which means oftentimes when we’re evaluating new markets to go into if the local government has requests around data sharing or data—like, data regulation that is outside of the bounds of that threshold, it prevents us from going into market. And we just don’t have the—I think, the technical capabilities, because we do really value encrypted data and user privacy in a way that’s incompatible with some requests from governments that would like Station to be in market. So I think that’s just a concrete example, where we’re prevented in operating in certain markets where we are holding ourselves to the standard that’s perhaps different than, like, the local government. PEPPER: In terms of the product—very similar. We have something called Express Wi-Fi, which is very similar to the Google Station. And there are issues, right, because there are some countries in which they want it to be open, so they have access to the data. And that’s not what we do. But going to another way to think about your question is, you know, you’re all familiar with Mark’s op-ed back in April, where he called for sensible regulation. It’s not—it’s no—you know, the shift has been—again, from ten years ago to today it was the internet is different, the internet is special, don’t regulate the internet, all right? The internet has now grown up. And we realize that there will be regulation. The question is, what form, what type? And, you know, we think the regulation needs to be not only—you know, how do you—what do you mean by sensible, proportionate, targeted, and not sort of overreaching? And you have to get this balance between—you want to continue the innovation, but you’d need frameworks and guardrails. You also have a continuum of what we mean by regulation. There are some things in which we think it’s completely appropriate, it’s within the remit of government. One of the big issues is election advertising. We think actually just tell us what the law is; we will abide by it. We don’t think we should be making law based upon, or decisions about—we probably have greater transparency than any medium—broadcast, print, certainly more than print. We have more transparency on election advertising than anybody anywhere. You want to take a—you want to find an ad and find out who bought it, where it ran, how many people saw it—in political advertising, we make that available. But that’s not the regulatory side, so what should the framework be? What should the rules be? So there are some things—and we believe, by the way—again, picking up on what Mike said, having harmonized privacy regulation we think is actually very important, because you don’t want—in the U.S. we don’t want 51 different flavors. And globally, you can’t be a global company without having a framework. GDPR’s a good starting point. We abide by it. Could it be improved here and there? Absolutely. But as a place to start for an approach to being able to have a global framework of, in quotes, “governance for privacy,” yeah. And that’s a regulation. But there are other things that we believe are going to be sort of regulated in more traditional self-regulatory ways, but there’s still going to be—so there’ll be industry codes. And the analogies there in the U.S., for example, are the ratings from MPA—it’s now MPA; it’s no longer MPAA—MPA promotion of pictures. So again, that’s a form of self-regulation. So you’re going to see, I think, a range of what the instruments are and the relative participation of governments in that process. But we’re already regulated and the question is what type of regulation makes the most sense to get the balance right between protecting users/consumers, protecting democracy and protecting and fostering continuing—to foster innovation and investment. HAMMER: More questions? Here, please? Q: Thanks. Hi. I’m Sabeen Dhanani with the Center for Digital Development at USAID. One thing we hear a lot about is 5G and how it’s going to completely again change the digital landscape. Given the infrastructure challenges you mentioned earlier and the regulatory challenges, what really—what is the realistic timeline on 5G in some of the emerging markets, and how should we all prepare to deal with the new set of challenges that that might pose? PEPPER: I’m somewhat of a 5G cynic. I think 5G actually is going to be transformational, but not in the way we’re hearing about the hype. Most people talk about—especially at a very publicity-driven level or certainly globally at the political level, it’s being characterized as this is going to be super AK, high-definition video to your smart phone. No. Maybe eventually. That’s not the transformational part of it. To me, 5G’s transformational, first of all, in terms of the types of applications that will be enabled by 5G are, in the first instance, the most transformational ones are going to be on the industrial side. It’s going to be smart factories. It’s going to be augmented transportation. It’s going to be precision agriculture. In previous years people called it the internet of things. That will be enabled by 5G. That is going to have very significant transformational impact in terms of economies globally in the developed world and developing world. That architecture and the types of deployments that are going to happen with that are very different than sort of rebuilding our consumer-oriented mobile networks. They are converting to 5G, but I can tell you that I’ve talked to CEOs of operators in developing countries, one in particular in the Middle East, and it’s a state-owned operator and he was told by the government, the king, this 5G thing, this is really important. I want you to build 5G right away. So he did. He converted his network to 5G. He said, I’m dying. I have no business case. Nobody’ll pay extra. My margins are crushed. I had to make these big investments. He said maybe, I hope someday, I’ll be able to justify that investment. So we hear a lot about 5G and I think it is going to be transformational, but not necessarily in the near term in the ways we’re hearing about. But I think it’s really important. But that’s going to be a very different type of deployment. The last point on that, what’s interesting is Germany just made some 5G spectrum available. Guess who won the auction for it? It was not a mobile operator. It was Siemens. They wanted that spectrum for 5G for smart factories and building smart factory systems. To me that was really interesting and a harbinger, I think, of what we may be seeing. HAMMER: We have a question here. Q: Welby Leaman, from Walmart. I totally buy the importance of social media, search, e-government to create that critical mass of demand for the next billion users to come online. Where do you rank retail in that, among the other sectors? Does retail have a big role to play in pulling people into the digital economy? And since the retail world’s now in sort of a big fight between Walmart and Amazon and others, between sort of a digital native approach and a Walmart approach which would be basically blending increasingly digitally enabled stores with e-commerce, do you see one or the other as more likely to pull people into the digital economy? Because our proposition has been that the blending gives people more on-ramps. So for example digital payment, you can go into a store, digitize your cash in Walmart Mexico and thus join the digital economy. SPIES: I think from everything I’ve seen it is—it can be really tough to go completely digitally native from a retail perspective in these markets. And the challenges often look very different than what a U.S. retailer would think about. So cash on demand, or cash on delivery is still a huge percentage of payments in many of these markets and a lot of major retailers have built the option to pay through credit card, through online payments and then cash on delivery. I think Uber offered this option before offering it in the U.S., Trust and Safety. So there’s a ton of issues with scam and fraud and being able to have a verified checkmark by your badge or by your store and know that you’re not participating in a fraudulent purchase is incredibly important, more so than it is in the U.S. And I think lastly even just logistics. So there’s a ton of focus on being able to map out cities that previously maybe didn’t have a formal address system and it was informal directions that even the national post office wasn’t delivering to, like building an e-commerce company around those logistics is often something that’s going to be done, often it’s proprietary to local startups, if they’re able to crack that in a city. So yeah, I would say that for retail and e-commerce specifically, localization is incredibly important because the challenges are so unique to these markets. Retail is still a much bigger percent of online shopping than e-commerce, and often this blended approach is something that has to happen just by nature of business there. HAMMER: We’ll come here. Q: Mike Jobbins with Search for Common Ground. A lot of the societies that you’re talking about—Congo, Mali, Yemen, where these next billion people live, are as divided as Myanmar was five years ago and have the same level of internet penetration as Myanmar did. So what are the lessons that you take from Myanmar or any of these places as they come online to shift the social norms of accepted and expected behavior on your platforms? PEPPER: So Myanmar’s a good example, but it’s not alone. That I mentioned earlier being prepared and digital literacy. Getting the first 3 billion people, 3.5 billion people online was relatively easy, compared to the next 3.5 billion. You take a look at basic literacy rates, you take a look at who’s connected and who’s not, even in emerging markets, and so the needs for even very basic digital literacy are extremely important. The training—and I mentioned earlier, giving people basic skills that when they go online, how do they set their settings on their phone? How do they—how can they know if there’s a somebody’s—a phishing attack for information? How do they keep their passwords safe, or even know that they’re supposed to? So these actually, I think, are extremely important and—for example, we’re working—we have multiple digital literacy programs that we have in place and others, new ones that we’re building all the time. And in fact, one of the most effective ones which we used to call OTG, On The Ground, which nobody knew what that meant, and I kept saying, OK, it’s a great program but that’s a silly name. They actually now call it Internet 101, which I thought was—makes more sense. The first place they launched it was in Myanmar, and it’s been actually quite effective. And if you’re interested, I can get you information about that. But that to us is extremely important for a safe internet experience going forward, especially in places with low literacy and very limited digital literacy. We’re also working with NGOs, for NGOs to develop their own digital literacy programs in particular countries. SPIES: I’d say for us, often our values are baked into the product itself. With free wi-fi specifically, as we’ve looked into other companies, what you see is there’s a much lower standard of user data privacy and protection. A lot of people have gotten onto wi-fi networks that are not secure and they get hacked in some ways. And so we’ve talked a lot internally about how do you even shift the perception of this product experience in this industry, because what we’re offering is a much more secure encrypted experience that really prioritizes user privacy. HAMMER: (Off mic.) PISA: Sure, I think it’s a great question. I think if you look at the infrastructure that Facebook very belatedly put in place in Myanmar, the thing that’s striking to me is how expensive it must have been to deal with that problem ex-post. They have hundreds of local language content moderators. Obviously the development of new AI systems, those can scale rather cheaply. But I think as digital platforms broadly step into new countries and are used en masse in new countries where they have to kind of get up to speed with local language content moderation, it seems like to me that one question that arises is when is it not going to make sense as a business proposition? Because if you look at—I know the Facebook average revenue per user last year in the United States was $27 per user, and in most—in sub-Saharan Africa was less than two dollars. So at a certain stage—and the next billion users, it’s a long game, right? And it’s a numbers game, but it’s not hugely profitable in the short term, or maybe even the medium term. And then the questions around how you do—how do you mitigate those risks, it seems like it can be a very costly exercise. So I wonder at some point does it become—do companies just kind of wash their hands of certain problematic situations? HAMMER: I think we have one more in the back, please. Q: Hi. Thank you for this session. My name is Ibrahim. I’m from Deloitte. My question is around cybersecurity and future threats with Google’s Sycamore quantum computer. We have things in place that ensure that these nations are secured in the future, especially around cybersecurity, encryption algorithm, and things like that? Also, looking atm, like, the World Bank report, which is set from I believe 2020 to 2030, which is around the time that quantum computer and things like that will start being really widespread. Are there strategies in place to ensure that these nations are also secured from infrastructure and capacity building? SPIES: Yeah, and I wish I could speak more specifically to quantum computing and some of the cybersecurity efforts. I know on the policy side we are working mostly with some of the transnational bodies that are thinking about policies to put in place and what are global regulations that could be a template for countries on this. Some of the points raised earlier, any time you can get a framework that is globally applicable and scalable, that’s easier for a company whether you’re talking about privacy or cybersecurity. So I know we’re participating in those forums, but I don’t have more detail on them. Maybe Robert can speak to the Facebook efforts. PEPPER: We’re not doing quantum at this point. (Laughter.) SPIES: Yeah, just to pass that over to you. PISA: Neither is CGD, so—yeah. (Laughter.) HAMMER: Neither is the World Bank. And so, with that—(laughter)—I am so grateful to each of you for coming this evening. Thank you very much for spending your time and please thank your panelists for the insight they shared. (Applause.)