Session Description:

If mobile is the Internet, then how we allocate and manage precious resource of spectrum is critical. We’ve got plenty of challenges as we move towards the Internet of Things, augmented reality headsets, smart watches, and disparate connected systems. As other countries push quickly toward high-bandwidth connectivity, what should  America’s priorities should be?

Kirkpatrick: Craig Mundie, who’s the moderator, is another longtime friend of Techonomy who’s been at many of our events and who has inspired us with his own engagement with Washington, and was running Microsoft Research for a long time, has been on the President’s PCAST initiative for a long time, had a lot of input into federal technology policy issues for a long time, ran Microsoft’s entire policy efforts for many years with great effect, and I’m very pleased that he can moderate this session. So Craig, take it away.

Mundie: Thanks, David. Good morning, everyone. Let me begin just by briefly introducing Jessica Rosenworcel, who’s one of the FCC commissioners, and her colleague, also a commissioner, Mike O’Rielly. And we’ve agreed that I’m going to call them Jessica and Mike, so that no one worries about that.

I thought what I’d do to kick this off, because ultimately we want to get around to talking about the policy questions, was just to give a little bit of, I’ll say my own personal view of why does this matter so much. And so I thought, okay, well I’ll just give you a few lists of things that are all happening now or are clearly going to happen in the relatively near future, one I’ll call the drivers of communications requirements, and in a moment, the technological changes that are underway.

So here’s some of the things that are going to happen: smart cities, which is essentially deploying sensing everywhere and using the data that is created there to make things happen; personalized medicine, telemedicine, all these things on a global basis; things like augmented reality, where we’re trying to come up with real time supplementation of the world in which we operate; new models of education—we talked earlier about broadband in the schools, but I mean just getting it to the school isn’t going to be sufficient. You’re going to want to use all these other technologies. Autonomous and connected vehicles that are continuously talking to each other to some extent, even while they operate. A whole new concept of public safety so that the kind of challenges we had around 9/11, where fire departments from one part of the city couldn’t even talk to another part of the city perhaps, or across different public safety organizations. And then, as was just being I think talked about, some of the embedded communications requirements in the Internet of Things that we see happening.

And the list goes on and on. And almost all these things are going to have a dependency on wireless communication, and while we all think today we know what wireless means—it’s Wi-Fi and cellular which predominate, along with Bluetooth, which predominate for most people-there’s really a lot more wireless communication going on around you than you ever really would’ve thought about.

But then, you know, we’ve seen a tremendous evolution. It’s been 100 years really—102, since the FCC was created, which was the result of the sinking of the Titanic, believe it or not. And the radio signals of the ships couldn’t get through and that was really the thing that motivated us creating a model of spectrum management. But all of the things that were dominant as very, very simplistic technological basis of radio communication 100 years ago have evolved very dramatically, and as fancy as it seems to you today, you ain’t seen nothing yet. So here’s a few things, you know, just in antennas, in signal processing, we’re going to have massive multi-antenna arrays, the ability to do beamforming in real time. I know people that are actually building antennas today out of metamaterials, which are synthetic ways of creating antennas that are steerable.

Adaptive arrays, new ways to actually shape waveforms that change the nature of their propagation.

We’re going to start to put software into the networks themselves. We’re doing this in the wired world, in data centers, but eventually that’ll happen in the wireless world.

We’re going to virtualize the networks, the same way we talk about virtualizing computing. We already can see a path to sort of 5G, beyond 4G in the cellular space.

We have free space optical, the ability to beam things down from satellites using lasers, next generation passive optical things, new kinds of satellites. Of course the Internet of Things and machine to machine requirements will ultimately dwarf the amount of talking we do as people. And yet, we’re already driving the network into saturation, but not because of our voice use, but because of our data use. And whole new models that are going to provide more facilitation for these kind of capabilities.

So I thought in the half hour we have, you know, we wanted to talk about spectrum policy: where are we today, how is that happening in transition, and a little bit what we see over the horizon as some of the issues to be faced. So I’m going to begin by asking the two commissioners to talk a little bit about their view of what’s happened in the last almost three years since the PCAST report on spectrum was issued. There had been a lot of efforts prior to that to talk about different ways of thinking about the evolution of spectrum management and use, but for one reason or another, you know, they didn’t happen to all come together. But it does seem that the PCAST report catalyzed some change. The FCC—even though it was framed as a way to get more use out of the government spectrum, it was clear that the FCC was a counterparty to that thinking and has sort of taken up the challenge. And in April, rules were passed that allow a new way to think about using spectrum. So I thought I’d start by letting each of you comment a little bit on what you saw as the good news and bad news of that proceeding and where you think it’s taking us. Jessica, do you want to start?

Rosenworcel: Sure. Well, thank you so much, Craig. That’s a terrific introduction. And for the uninitiated, I’m sure that spectrum policy sounds complicated. But there’s one thing that’s very clear: we are cramming more uses into our airwaves than ever before. And because the laws of physics aren’t going to change, we are going to have to find new ways to be more efficient with how we zone those airwaves.

Historically, the FCC has auctioned off spectrum for exclusive licenses for specific use. The system served us reasonably well over the last several decades, but with all of this new use of the airwaves, we’re going to have to think about them differently, come up with new ways to be more efficient, to cram more uses into this finite space. And I think the PCAST report really gave a jolt to that conversation by encouraging government to rethink its existing uses, by encouraging government to think about how to take more of its spectrum and help repurpose it for commercial use, and then how with that repurposing, we would think about models that were not only about exclusive use, but about shared use, and if we did that effectively we could multiply what we were putting in our skies. And I think the PCAST report gave a tremendous jolt to that conversation and was a terrific initiative from the government.

Mundie: Good. Mike?

O’Rielly: Well, thank you, and thank you for having me here today. I think it’s obvious, you can see there are certainly a number of issues that my colleagues and I work together on, and certainly the two of us are able to. Spectrum policy is so important, and the unlicensed and conversation on Wi-Fi is part of that discussion going forward.

We talked a little bit about—you know, your introduction was so well done in terms of talking about the PCAST report and its impact, and then what does it mean for, or how did it play into our 3.5 gigahertz item. I think that—you know, the PCAST report was done before I got to the Commission, so I’ve used it as a helpful tool, but it has implications in the items that we’ve done since then, and 3.5 is a great example. I think I’ve had some concerns on the 3.5. I support the item in its concepts. I think there were some concerns that I had going forward. I could go through the particulars if it’s helpful. I certainly want to have smaller coordination zones. I don’t want to see, you know, big swaths of the country being excluded, and we can work to reduce those, and did work to reduce those over time, and that’s very helpful.

I also thought that the Commission had thrown a bunch of different ideas, a bunch of experimental pieces, variables, if you will, at the 3.5 gigahertz ban, and so it’s going to be unclear, in my opinion, what of those are going to work and what are not going to work. So it’s throwing so many different things and so many different components at a problem, you’re not sure what was helpful and what didn’t work going forward.

If I had a third part, I would say I think the Commission could’ve spent more time on what’s known as the priority access licenses. It’s how are you going to have technically the GAA, or the unlicensed side interact with folks that are looking for some reliability and some certainty on a license that’s different traditionally than our existing licenses. So I’d like to see some more specifics on that. I think the Commission could’ve done a better job on that. But as a whole, I’m supportive and I think it’s a good product and we’ve got some work to do still.

Mundie: For the uninitiated, the 3.5 gigahertz item that the FCC took up was one that the PCAST report pointed to, and it was one of the primary uses in that band of marine radars. You know, so you look around and you say how many of those boats are sailing around in the middle of the country, and the answer is, well, not many. But under a model where you give out these bands by frequency, you know, without any geographic segmentation, for example, you know, you say, well, if you’re not where a navy radar is, it just is wasted. And so the question raised by the PCAST report was, gee, aren’t there going to be a lot more ways to subdivide things? So when Mike talked about the exclusion zones, that was really, well, how big an area did the navy get to continue to claim? Because when you’re near the coast, they say, “Well, our boats are near the coast, so how much of the country do we not allow to use?” And with I think some urging, they were able to get that to be reduced more and more.

But that’s just one first example that shows, okay, by thinking about coexistence, we don’t any longer have to think about just allocating on frequency. We can allocate on geography. And it turns out, for all the reasons I just gave you in the list of technologies, there’s potentially many, many other ways to think about discrimination, and therefore, coexistence in what historically would’ve been thought to be sort of an exclusive area, we might be able to discriminate on other technological grounds.

That was important, I think broadly, and so I think this 3.5 was a grand first experiment. The other thing that was notable about it, and why this is such an important policy question, if you just look across the bands, roughly 60% of all spectrum is controlled by the United States government. In other words, the government gave it to itself. You know, and then the public gets to put everything else in the other spaces. And that’s true around the world. The US is probably the most congested user of spectrum anywhere on the planet, but there is a process by which these things try to get harmonized through the World Radio Conference every four years.

Rosenworcel: I was going to say one of the things that I think is a big challenge for spectrum policy going forward is figuring out how we rationalize that division of holdings between commercial and government use. Because much of the spectrum that the government controls for government critical missions was allocated decades ago, when our spectrum technologies were not as efficient, and when we stand back and contemplate just how much wireless functionality is now in our commercial economy, we realize we’re going to have to transfer more of that spectrum resource from government to commercial sector use over time. The challenge is we don’t have a really good system for doing that. Historically, it’s been the kind of thing where the government from time to time realizes we need more spectrum in the commercial pipeline, so we knock on the door of federal users like the Department of Defense or the Federal Aviation Administration and we beg, coax, and cajole them to give up a few more scraps for commercial use. And then over time, we wrangle some from them, usually through the political process, and we get the opportunity to repurpose it. That does not strike me as a terrific pipeline for the modern wireless economy, and so we are going to have to figure out how to provide our federal users, who control roughly 60%, with incentives to be a lot more efficient so that when they see us coming for reallocation, they get some gain from reallocation and not just loss.

O’Rielly: And I would agree with that concept, but I would just add to it, it’s going to be two parts, right? You know, it’s always the carrot and the stick, and my colleague highlighted some of the carrots we can provide to federal users to reallocate spectrum. We’re still going to need some part on the stick part of the equation, in my opinion. And I’ve worked in the political process and I wouldn’t say that we were cajoling or asking them or knocking on the door and begging them. We were telling them, “You will give up spectrum.” And so I think going forward we’ll still have a forceful Congress seeking to reallocate spectrum. We’re also going to have to provide some kind of a stick mechanism, and I think that this has been talked about a number of times, putting some type of a price on the spectrum itself, and existing users to have some kind of budgetary consequence for holding onto spectrum. So I think it’s both components, in my opinion.

Rosenworcel: Yes, I think that’s a terrific idea and something that was mentioned in the PCAST report. We don’t have a uniform system for valuing the spectrum that’s used by federal authorities, and the PCAST report recommended a kind of spectrum bucks or synthetic currency to try to account for how much of this tremendously valuable resource is used by the government. And I think the development of that over time will help us create a system of exchange that will make us more efficient with this resource.

O’Rielly: I like the term synthetic currency. It sounds like we can trade that in for some bitcoins.

Mundie: Well, actually, that was what we recommended, but it wasn’t for bitcoins. It was actually for sort of current day budget dollars that actually would allow an agency to start to make the move. One of the things that is always true in any organization, government or business, is that you say, okay, I know I’m supposed to move there, but it costs me money to go from the place I am to the place I want to be. And in an austere budget environment, it’s been difficult to get appropriations to do these things that look out over 10 or 20 years. You know, if you have all these satellites flying around, you say, that’s great, let’s have new satellites that work in a different way. No one wants to essentially just forget those assets that are up there. And so we have a problem of short-term versus long-term incentives against which to lay out these carrot and stick kind of models.

I think there’s also another interesting thing about money as it relates to spectrum. Congress itself has seen this as a revenue generation opportunity for the government, and the big spectrum auctions, whether they’re in the TV domain or in the cellular domain, while they’re in reality small relative to the nation’s total spending, they’re big relative to other ways where the government has historically been able to, on a transactional basis, get money from businesses. And so Congress is in some ways motivated by that process.

But, you know, one of the big discussions is the issue between licensed and unlicensed. You only got to get that tax—bad word, but, you know, that money, the transaction, if you were giving this right to use for some period of time expressly in the past. And unlicensed of course has been a huge driver economically, and while we don’t keep score the same way, the fact that everybody in here is hooked up to the Wi-Fi network here and has a Bluetooth that hooks their phone to their car, I mean all those things were all shoehorned into a tiny little chunk of spectrum that was set aside by the FCC many decades ago—

Rosenworcel: Thirty years ago.

Mundie: Thirty years ago for experimental uses. That was what unlicensed was. It was a place to experiment. And most people’s data traffic today, at home and at work, is going through that grand experiment, and it has had a fundamental effect. And I’d love to hear your thoughts now about the trade-offs between the economic impact of unlicensed, which really evolve very quickly, versus the economic issues associated with the auctions and control, and then how do you see that intersecting the sharing model.

Rosenworcel: So unlicensed spectrum I think was the greatest experiment that the FCC did in spectrum policy in the last 30 years. We had some random scraps of spectrum that were supposed to be developed for industrial, scientific, and medical uses, but nothing really was coming forward there, in part because some of that—

Mundie: We had garage door openers in there too.

Rosenworcel: Yes, we had garage door openers. And in fact, some of that spectrum was facing interference from microwave ovens, which were just on the scene three decades ago and very exciting. In any event, we decided, well, maybe we didn’t have to have exclusive licenses, where we gave them to specific parties for specific uses? What if we just, instead of treating it like a house that you occupied with a lease, we made it like a road and everyone just would obey the traffic rules? This was completely radical thinking three decades ago. Nobody thought it would work. But in time, we had the 802.11 standard and 2.4 gigahertz is where Wi-Fi was born, and there’s not a person in this room that hasn’t benefited from unlicensed spectrum today. Half of us rely on it to go online regularly, and more than that, it contributes about $140 billion dollars in economic activity to the US every year. That number’s only going to grow with the Internet of Things.

But the challenge is every time we set aside some spectrum for unlicensed use, like we did three decades ago, we don’t get near-term revenue like we do when we auction it off and license it. So Congress is going to struggle when they look at these things. There’ll be a bias towards coming up with revenue-generating activity like exclusive use license spectrum, when the truth is we have to think not just about the static value of that spectrum when we auction it off for exclusive use, but about the economic benefit that we all get from that use, and unlicensed spectrum, it’s becoming clear, is an economic engine for the modern economy.

Mundie: Mike, let’s talk about a special case of this licensed versus unlicensed question, which I think is quite fascinating. So now many of the cellular operators have realized that one way to offload the congestion in their networks as data grew was to offload it to the Wi-Fi environment, and then they went out and they bought a lot of the Wi-Fi hotspots, and now they actually have a hybrid that’s sort of half licensed and half unlicensed. But now they’ve developed interesting new technologies like LTE. That was developed originally for their licensed network, and people say, “Hey, I could come up with a version of that and I could stick it down there in the unlicensed band.” So one could say, “Hey, it’s unlicensed. You know, let’s just”—why not? It’s just a continuation of the grand experiment. That’s what brought us from garage door openers to cordless phones in your house to Wi-Fi and Bluetooth—

Rosenworcel: Baby monitors.

Mundie: Baby monitors. They’re all living in there—but now they’re all living in there. And so the question is, you know, now people say, well, now there’s so much in there, are we at risk of having a great new thing that would be superefficient, but it tends to boot all of them out of that space? How do we—you know, we didn’t want to have a lot of rules about it. So how do you see that tradeoff going forward?

O’Rielly: So I think what you’re seeing also is two policymakers that are probably the biggest proponents of unlicensed spectrum use out there—and I’m a huge fan of unlicensed and what they’re able to do. A former employer of mine used to say the greatest thing about unlicensed use is you have no idea what people are going to do with it. It’s the beauty of it. And so I’m a big proponent.

I do think going forward we’re going to continue to have the discussion. There’s still going to be a great need for licensed spectrum. We see the demand in the AWS-3 auction. We’re going to have demand in the broadcast and incentive auction. So there’s still going to be a great need for licensed. Then the real question becomes how do you expand the opportunities for unlicensed spectrum and Wi-Fi? Well, one way my colleague and I have been looking at is how do we expand in the upper five gigahertz band, the 5.9 space, where we have some auto manufacturers and equipment providers who are working on auto safety issues, how can we share that spectrum with unlicensed use. And there have been obviously a number of conversations that have been had and there’s many more to be had, but I do believe there’s a possibility that we can expand into that band and expand the opportunities that you talked about where you see a number of licensed providers offloading traffic in unlicensed bands.

And to answer the second part of your presentation, what do we do with the conversation that we’re having both externally and internally on the LET-U issue, and that is something that I’m very sensitive to. It is something I’ve mentioned before that I want to keep my eye on. I certainly recognize the concerns that people have in the unlicensed space. I’m certainly sensitive to that. But I also want to be careful in trying to foreclose any opportunities. So it’s a balancing act and I want to watch that very closely without putting my—what I’m really concerned about is putting my thumb on the scale of some of the international standards setting bodies. We’ve done a great job of keeping out of those entities and I don’t think it would be a smart move for the Commission to be involved in that space.

Mundie: Okay. So the last thing I want to talk about a little bit, since money always is important, is, you know, one of the things about these new sharing models is that they are built around the idea of dynamic sharing, so that you don’t necessarily say—it’s sort of if you aren’t using it, you could lose it, but not lose it in the sense that you can’t access it at all. You may still have a priority right. But when you’re not using it, we’re going to let other people use it. And because of the software nature of the networks and even the phone systems—and frankly, even the future of cellular is going to be a lot more soft than it was in the past from a control point of view. The ability for these things to move around on a space-available basis looks to be dramatically better in the future. So now we have the idea of being able to have real time auctions, so instead of having a big auction held every so often with billions of dollars at stake one time, why not actually let this become a continuous process? It’s a bit more like buying keywords on your search engine.

Rosenworcel: Right.

Mundie: You know, it’s a continuous big market with lots of money flowing around. And that’s another way to think about, you know, Congress’s desire to say, “Hey, but is there a way for us to at least get some monetization of this in a different way?” And I think that blending these things together and recognizing that there’s so much more dynamic opportunity to use the network itself and the Internet to control this use, that’s a big opportunity and I’d love to hear your thoughts on that, and then we’ll open the floor to questions.

Mike, do you want to go first on that one?

O’Rielly: Oh no, please.

Rosenworcel: Well it’s really like two poles here. One is the old system, whereby we would auction it off at a static point in time, collect revenue, and then functionally hand you a certificate and say you have the right to use this spectrum. The other end is what you’re describing, a spectrum access system that’s software that helps you on a real time basis purchase the right to spectrum to use for a discrete use. And I think that over time we are going to explore more and more things that look like that real time spectrum access system or software system you describe, and the place that we started to do that—not on a real time basis—will be in the 3.5 gigahertz band. And I think that’s really exciting and there’s revolutionary stuff that’s going to happen in that portion of our airwaves and people should pay attention.

Mundie: Yes. I this it’s incumbent on all of us to make people stop and realize that at the intersection of all this intelligence and connectivity is—here’s just another example where public policy can be implemented to the benefit of the society in different ways than we historically did it. You know, I mean like the carriers often will make the argument, “Well, if I don’t have certainty, I wouldn’t make an investment.” But of course right now they’re investing in both the Wi-Fi half and the other half, because at the end of the day they’re economically rational.

So to some extent, as long as you’re willing to keep putting more money into a real time auction—it’s like, hey, if I like my ad word, I’ll just keep paying, bidding it up because I want to keep getting it. You know, we may see the same thing happening with the way people access spectrum.

Let’s open the floor to questions now. Yes, sir?

Audience 1: Yes, I just had a question. Why don’t you just lease or rent out spectrum? Why do you always auction it off? It’s like giving like fee-simple ownership or something. Why don’t you just get a revenue stream out of it?

O’Rielly: I think that’s some of what we were just talking about. Some of it is dictated by the statute itself. We have obligations under the law that we’re required to follow, and that’s for a bigger conversation of the Congress to consider to whether to keep or change. There have been definitely models in terms of different ways to go about this. I think what we’re seeing in the 3.5, and my colleague talked about this very well, is a mechanism where if you’re not using your priority access license it automatically is available for unlicensed use. So you can get what’s known as the PAL, and then if you’re not using it, someone else can, but then you can immediately, when you’re ready to roll out your network or whatever it may be, you’ll be able to reclaim that. So you’re going to bid in an auction that’ll be real time and have smaller terms and conditions, and shorter. So that’s something that we’re kind of trying to experiment with. It’s kind of one of the things I said, we threw like seven different experiments at the 3.5, and hopefully we’ll be able to decipher which ones are working and which ones are not.

Mundie: One way to think about it this is it’s another natural resource. It’s quite scarce. We can’t manufacture more of it. And it’s a bit like most government lands, for example. We don’t give them away or sell them to people. You know, we’ll give you an oil lease or a gas lease or a tree lease, but in the end, we retain the ownership for the public. And I personally think that that’s really what we’re doing here. We have to husband this thing, because we can’t really predict exactly the ways we’re going to want to use it in the future and to some extent the leases were pseudoperpetual in the past—legally not, but there was sort of an expectation—

Rosenworcel: Pseudoperpetual, I like that word.

Mundie: But, you know, now technology allows that to essentially become shorter and shorter and therefore we may get a better utilization.

Another question? Yes, sir, back here.

Audience 2: With some of that real time access, things like that, do you see the hardware manufacturers ultimately sort of pushing back against that? Do you see hardware manufacturers understanding that there’s issues in the complexity of devices and things like that only increasing and if you’re offering some of these real time or near-real time use cases, suddenly the complexity they need to manufacturer into their chipsets becomes much, much higher—higher power demands, longer battery life, all the things that, you know, they’d be pushing back saying, “Look, if everybody’s standardized on one thing, that would make it infinitely easier for development”?

Rosenworcel: Well, this is an ecosystem that is going to take some time to develop, but I’m also confident that the number of use cases for wireless technology is growing exponentially and everyone is going to want to make sure that some of that traffic can get on their devices.

In addition, I want to point out that a lot of our activity right now is in airwaves like the 3.5 gigahertz band that never had a commercial use case before, so it’s greenfield and there’s new opportunity. We don’t have incumbents trying to protect their existing business case like we might in some other spectrum bands.

O’Rielly: Well, can I answer this way—and I completely agree. Two parts: One, Craig made the point that it’s impossible to make new spectrum. But I was in Dallas last week, and if you want to see what the manufacturers and equipment guys are doing, you know, Dallas seems to be a pretty good hotbed place for that and you see that they’re actually testing in higher bands than I thought we were—you know, when people talk about 28 gigahertz, they’re in the 100 gigahertz and they actually—I said what’s the farthest we’ve been able to—you know, they’re testing at certain universities in the 400 gigahertz. And we certainly don’t have any idea what the federal government and the military is able to do. So we’re at 400 gigahertz that we’re trying to make operational, so we’re expanding upwards.

But to your point, I think that the equipment manufacturers are being very dynamic and they’re ready for all of this, so they’re adjusting as well as we do too. So, sure, a static marketplace would be easier to manufactures, but they’re the ones pushing the envelope just as well.

Mundie: I always say it’s kind of funny, there’s a technology that’s important in all these things that people don’t think about. I call it JAMOS. That’s “just a matter of software.”Because almost everything now has got a microprocessor in it and they’re so small and so inexpensive that the only thing that governs whether they can do these things or not is just writing some more code. And that’s why I think in fact in the past it would’ve been expensive for them, but increasingly it’s not that expensive because there’s so much more software and firmware in these things that gives them adaptability.

Another question?

Audience 3: Slightly off topic, but two commissioners at the same time, so: Schools are only open 15% of all the time in the year. Kids are in class 8% of that time. They’ll have an Internet connection 1% of the time. E-Rate Funding is about to expire. We need to wire our schools, but how do we expand Internet access through E-Rate so we can have Wi-Fi cards given out through libraries, so kids can take the Internet home with them so they have 24-hour access? None of us in this room would be successful if we only had access to the Internet 1% of our lives. We’ve got to think bigger than just wiring our schools.

Rosenworcel: I completely agree. I mean when I was in school, homework only required a clear workspace, your siblings leaving you alone, and a number 2 pencil. But today, seven in ten teachers assign homework that requires Internet access. The data from where we work show that one in three households don’t have access, and if you’re a student in one of those households, just getting your homework done is hard. We’ve made tremendous success in updating our E-Rate program. We are now, over the next five years, wiring all of our schools with high speed broadband and Wi-Fi and it’s something we can be really proud of as a nation. But the challenge is just what you described. When those kids go home, if they don’t have connectivity, they’re in trouble. Because it’s no longer a luxury. It’s a necessity to get basic schoolwork done. And I hope over the next year the agency’s going to be able to look at some of the programs it has, including the Lifeline program, and also our efforts to put more Wi-Fi in the airwaves and see what we can do to fix that problem, because I think it’s the new digital divide and it deserves all of our attention.

O’Rielly: I would just answer, we’ve done quite a bit on E-Rate, not everything I agreed with, but that’s understandable. But I would say that one area that I want to look at going forward is, we spend a great deal of money and time on libraries, but the localities determine the hours of the library, so they’re shrinking the amount of time the library’s open. So we can have whiz-bang technology within the building, but then the building’s only open from 10:00 to 4:00. So how do you deal with that issue and how do you make sure the library’s leaving its equipment on going forward and how can we expand the opportunities of the money that we’ve invested as a community into the local library.

Rosenworcel: I think this problem, which I call the homework gap, is a huge one, and it is the new digital divide, and we’re going to have to figure out how to address it, not just for the students who lack access today, but for all of us. Because if they can’t do their homework, they represent the economic future of this country and we’re going to have a problem.