Our latest podcast is called “Who Runs the Internet?” (You can subscribe to the podcast at iTunes, get the RSS feed, or listen via the media player above. You can also read the transcript; it includes credits for the music you’ll hear in the episode.)
It begins with Stephen Dubner and Steve Levitt talking about whether virtual mayhem — from online ranting to videogame violence — may help reduce mayhem in the real world. There is no solid data on this, Levitt says, but he hypothesizes:
LEVITT: Maybe the biggest effect of all of having these violent video games is that they’re super fun for people to play, especially adolescent boys, maybe even adolescent boys who are prone to real violence. And so if you can make video games fun enough, then kids will stop doing everything else. They’ll stop watching TV, they’ll stop doing homework, and they’ll stop going out and creating mayhem on the street.
This episode then moves on to a bigger question about the Internet itself: who runs it? As Dubner asks: “Who’s in charge of the gazillions of conversations and transactions and character assassinations that happen online every day?”
Internet scholar Clay Shirky, author of Here Comes Everybody: The Power of Organizing Without Organizations, tells us that 60 percent of adults around the world are now connected to the same communications grid. (South Korea, he says, is the “most wired” country.) And this global connectivity is interesting, he says, because it’s not like there is an international body governing what’s online:
SHIRKY: Well, I mean, famously, the regulatory overhead on the Internet is permissive and minimal. In fact, the thing that freaked everyone out about it in the 90s when it was spreading on the wings of the web was that no one was in charge. … There are famous stories of bosses fretting that because all of their employees were suddenly sending international emails that they were suddenly going to be hit by the bill by the people who ran the Internet.
This week’s episode also serves as a prelude to next week’s, which discusses who’s in charge of — well, the whole world.
[MUSIC: Ruby Velle, “Used Me Again” (from Used Me Again)]
Stephen J. DUBNER: We’ve all heard the accusations, again and again and again…
TV CLIP: And again another story we’re following, cyber bullying. This seems to be an all too common occurrence.
TV CLIP: Porn is now available everywhere at the click of a button.
TV CLIP: Pornography is taking the place of good healthy sex education.
TV CLIP: We are exposing people most at risk to a new and toxic drug called virtual entertainment and the worst of it are these violent video games.
TV CLIP: People are so concerned about violent video games, think about your kids acting out violently on real people through social media.
DUBNER: The message is clear: technology makes it easy for people to do bad things, to engage in antisocial behaviors, that they might not otherwise do. But what if we have this question backward? What if – maybe, somehow – what if all that virtual mayhem translates into less actual mayhem? I called up my Freakonomics friend and co-author, Steve Levitt. He’s an economist at the University of Chicago. One of his favorite research topics is crime:
Steve LEVITT: So in theory there are at least three channels through which you could imagine virtual violence spilling over or not into real violence. So the first, and this is the more popular view, is that when you teach kids how to shoot guns in violent video games then they’re more likely to go out and shoot guns in the real world. And certainly it’s easy to understand how that would work. There’s a second view of the world which I think is probably hard to defend which is that if I’m frustrated I can go shoot my fake guns in my video game and I won’t feel the need to go shoot my real gun in the real world. You can also imagine how that would be. And then there’s a third answer, which really is one that economists think about more than regular people, which maybe the biggest effect of all of having these violent video games is that they’re super fun for people to play, especially adolescent boys, maybe even adolescent boys who are prone to real violence and so if you can make video games fun enough, then kids will stop doing everything else, right? They’ll stop watching TV, they’ll stop doing homework, and they’ll stop going out and creating mayhem on the street. But I think evidence that we have, it’s relatively scant, but the evidence we do have is actually that that third one is far more important that either of the first two in influencing real-world violence.
ANNOUNCER: From WNYC: This is FREAKONOMICS RADIO, the podcast that explores the hidden side of everything. Here’s your host, Stephen Dubner.
[MUSIC: Two Dark Birds, “Start All Over Again” (from Songs For the New)]
DUBNER: So, is it possible that virtual mayhem is a substitute for real mayhem? Yeah, sure, it’s possible, as Steve Levitt says, and the data – as preliminary and murky as they are – suggests this may even be true. But this is the kind of question that’s a lot easier to ask than to answer, at least definitively. So let’s take a step back, both in scope and in time, to ask a different set of questions – about the Internet itself:
Clay SHIRKY: You know, it was interesting, I started using the Internet in the early 90’s.
DUBNER: That’s Clay Shirky:
SHIRKY: I teach theory and practice of social media at NYU in the Interactive Telecommunications Program and in the Journalism Department.
DUBNER: Shirky is what might best called an Internet scholar. And as he was saying, back in the early 1990’s…
SHIRKY: I was running a theater company in New York City that staged nonfiction documents. We would take things that weren’t originally meant to be put on stage and make theatrical collages out of them.
DUBNER: Like a for instance, somebody’s journal?
SHIRKY: The first one we did was the transcript of an air traffic control conversation during a plane crash. We did one that was only drawn from materials produced in 1974. We staged the Attorney General’s report of pornography, a whole variety of these things. And my mother, who is a reference librarian, said, Oh, if that’s what you’re doing you should know about this thing we’re learning about in the library. It’s called the Internet; it’s like a great big library. I said, Okay Ma, I’ll check it out, and I….
DUBNER: I love that the Internet scholar learned about the Internet from his mom the librarian.
SHIRKY: Well actually that turned out to be enormously helpful because I never once believed the kind of mirror-shade, cyber-cool thing because it came from my mom and because my first email ever was not someone trying to contact across the ether, but mom writing wanting to know if my shirts were clean.
[MUSIC: Ed Hartman, “Happy Marimba”]
DUBNER: From the outset, Shirky’s interest in the Internet went beyond just using it …
SHIRKY: Having the rich and varied social life I did in those years I spent a lot of Saturday nights at home reading engineering documents and slowly pieced together some sense not just of what you can do with the Internet but really what the Internet does, what the logic of the thing itself is. And in a way having got interested in that just as a side effect of wanting to use it, I ended thinking more about the Internet than I did about using the Internet for other kinds of things.
DUBNER: Now, just think for a minute of how you use the Internet today – you used it to get this radio program, for starters – and then think about how everyone else uses it (and abuses it)… The question I wanted Clay Shirky to answer is simply this: Who’s in charge? Who’s in charge of the gazillions of conversations and transactions and character assassinations, that happen online every day?
SHIRKY: To the first approximation, no one, which is to say by and large these things are unstructured, unsculpted, unedited, unmoderated. That in small conversational groups turns out not to be much of a problem. In fact almost all of the negative social aspects of the network that people point to are a function of group scale. And there’s a huge sociological literature on things like the tragedy of the commons, right? The inability of people to self-moderate in places where there’s large open resources. Well what there’s a commons of on the Internet is other people’s attention. So if you’re dealing with you know, one-on-one as we are, or in a small group, people self-moderate, because in a way there’s no more attention to be gotten than you get just by being a participant. But once there’s a large group, the way you get attention is by acting out. The way you get attention is by attacking people, flaming, trolling, sort of intentionally leading people on and so forth. And so what’s, the huge culture clash over the last 20 years, the time that I’ve been on the Internet has been between the libertarian ethos of the people who have built the systems and the growing need for social controls around very large-scale resources. You see this on Wikipedia especially where almost all of the pernicious effects on Wikipedia come from people editing biographies of living persons as they call it, right, articles that are about people still alive. So Wikipedia has developed a bureaucracy around paying special attention to those kinds of articles. And you can’t have Wikipedia do what it does and have complete freedom to edit because people now know where the problems arise and you start to get this tension between the open-ended culture of the encyclopedia anyone can edit and the essentially the social norms that form around this particular group of articles we have to keep under some control.
DUBNER: Okay, two things, one very quickly, you are affiliated with Wikimedia, yes?
SHIRKY: I am an advisor to Wikimedia, yes.
DUBNER: Okay, so just for the record. Second, then getting to Wikipedia and talking about the friction or the stakes when there are biographies of living people, is, what are the characteristics of an environment on Wikipedia, let’s say, where there is more regulation or intervention needed? Is it necessarily a case where the stakes are “higher?” In other words a biography of a living person, or is it a political arena, a sexual arena, in other words what are the characteristics of the environment that most necessitate policing, regulation, et cetera?
SHIRKY: There are essentially three...You could break it down into three kinds of articles. There’s articles you know in advance will be like that because of their structural characteristics. Biographies of living persons is the most important one. By and large most biographies of living persons don’t get fussed with. But that class of article is important. Then there’s the stuff that are political hot button issues, the article on abortion, the article on evolution, the article on Islam. Those things come and go. But there is a heightened degree of attention because of the sense that those articles could come under attack at any moment. And then there are the things where for no reason anybody can predict in advance they just sort of go pear-shaped. And if you look at the social, if you look at Wikipedia as a social network and you pick any two articles and you look who’s edited both of you look at any two users and you look at which articles they’ve coedited, what you start to find is that almost the most active users are content experts, so the people who are editing the articles about say the solar system are astronomers -- people who know a great deal about that subject. But that the most active editors of all, and there’s a small cadre of them, show up on one day on Pluto because remember when Pluto got kicked out of the planet club, all of a sudden there was a huge argument about Pluto, and the next day they’ll be editing the article about abortion. And those people are people who don’t care so much about the content of each article, they’re people who have committed themselves to the health of Wikipedia as a whole. And you need both things. If it was just people who generically wanted Wikipedia to be good but didn’t know anything in particular about Pluto, the Pluto article would be terrible. But if the people who knew a lot about Pluto didn’t have some defensive cordon they could rely on when the things comes under attack they would eventually get frustrated and leave. So you have -- without these roles being assigned – you seen people going in and slotting themselves into essentially voluntary division of labor in order to keep the system going in ways that are better than if everybody doing the same job as everybody else.
[MUSIC: Espionage, “Girl From Orange County”]
DUBNER: Unassigned roles… voluntary divisions of labor… that doesn’t sound like any kind of hierarchy we’re used to, does it? Coming up on Freakonomics Radio… this creates some obvious problems:
LEVITT: One of the things about the online rant is that they live forever, right? So that someone can say something hurtful and it can stick around.
DUBNER: And: why some people aren’t so worried:
SHIRKY: You hear all of the ethicists saying oh the technology is outrunning ethics, to which the pragmatist answer is that’s exactly what we want.
ANNOUNCER: From WNYC: This is FREAKONOMICS RADIO. Here’s your host, Stephen Dubner.
[MUSIC: The Sound Room, “Just Can’t Help It”]
DUBNER: So Clay Shirky, the Internet scholar, is telling us today about how little oversight goes into the very very very very very vast universe that we call the Internet. So yes, things can get pretty chaotic online – but given its size, wouldn’t you expect it to be much more so? I asked Shirky to give us some numbers on the size of the Internet, whether you measure it in users, dollars, or time spent thereon.
SHIRKY: The interesting question about users is not what constitutes a user, but rather what constitutes the Internet because as the phone networks and the Internet are becoming more and more entangled, the Internet population looks to be something shy of 2 billion, but the number of activated handset accounts is north of 3 billion. And you can make an argument that when you look at kind of the hybrid uses of SMS and the Internet in India, and Kenya, and Egypt and so forth, you can make an argument that the people using SMS on their mobile phones, they don’t have the best devices, they don’t have the best connections but they are certainly connected to the same grid as we are. So depending on how you define…
DUBNER: Okay, so let’s say the biggest grid…
SHIRKY: The biggest, the biggest…the biggest grid if we’re talking about the biggest grid that we want to talk about, which is to say the sum total of two people who could communicate with one another when they’re connected to it, that is about 3 billion people right now.
DUBNER: You mean there are 4 billion who aren’t? And how many of those are under five let’s say?
SHIRKY: Yeah, so roughly a third of the population is under 15, which means for the first time in human history, a majority of the world’s adults are connected to the same communications grid.
DUBNER: Okay, so let’s say we’ll talk just adults. What percent of adults generally as best as you know are connected to what we’ll call the communications grid rather than the Internet.
SHIRKY: About 60 percent.
DUBNER: That’s it, really?
SHIRKY: About 60 percent, yeah.
DUBNER: What the hell are all those people doing with their time?
SHIRKY: What is everybody else doing? Well you can’t be playing Angry Birds.
DUBNER: Sailing or something?
SHIRKY: No, it’s, and it’s now a little bit like the adoption rate in the U.S. because the U.S. relied on commercial spread of wires and antenna, rather than state spread. And because we’re a more rural population even now, there, we’re a more rural population than say, South Korea, right, the famously most wired country. When you cross the 50 percent mark for populations with large rural, you know, large rural populations you start to slow down. So what we’ve got is Northern India, much of inland sub-Saharan Africa, Bangladesh away from the large cities, inland China, and it will be slow going to connect those populations, but it is still happening.
DUBNER: Okay, so 60 percent of adults around the world, do you have any kind of worthwhile dollar figure in terms of overall economic value?
SHIRKY: I don’t.
DUBNER: But we’re talking trillions of dollars.
DUBNER: No matter how you measure. So here’s what I really want to ask. So let’s pretend for a minute that we're talking about a machine that 60 percent of the world’s adults use and it generates or takes in revenue in the trillions and it accounts, and it involves, we’ll make up a number, I don’t know do you have a number of hours per day that that median adult might spend?
SHIRKY: Well you know, I don’t because it’s so variable between countries like Korea, where the number can be, the median number can be six, you know as high as six.
DUBNER: Six hours a day we’re talking?
SHIRKY: But then there’s also the question of how often are you online when your iPhone is in you pocket and doing something on your behalf in the background, are you online?
DUBNER: Let’s be very conservative, we’ll say half an hour.
SHIRKY: Half an hour.
DUBNER: Super conservative.
SHIRKY: That is very conservative.
DUBNER: So now we’re talking 60 percent of adults worldwide, trillions of dollars, half an hour a day, on a machine, or within an institution, if we want to define it as an institution. Now, compare for me then the regulation and policing, and oversight that goes into that machine with a more traditional organization that has 60 percent of the world’s population and trillions of dollars.
SHIRKY: Well, I mean famously, famously the regulatory overhead on the Internet is permissive and minimal. In fact, the thing that freaked everyone out about it in the 90’s when the thing was spreading on the wings of the web was that no one was in charge. And in particular that the thing was designed to be completely oblivious to national borders. Right? And it was a weird accident of broadcast technology, literally signals going through the air, you’d think this would somehow escape the nation-state, but because radio engineering circa 1920 was so crappy, broadcast radio and TV stayed inside national boundaries just because the broadcast towers had to be near the receivers. So the Internet was the first really big group oriented transnational medium. And there are famous stories of bosses fretting that because all of their employees were suddenly sending international emails that they were at some point going to be hit by the bill by the people who ran the Internet. And it took some time to realize that there is no people who run the Internet, and therefore there is no bill, that once you, it’s infrastructure that everybody pays for and everybody gets to use. And the history of the Internet in a way has been the history of building up regulations…And the fight, the regulatory fight around the Internet is precisely around regulation designed to preserve a degree of openness of interconnection and use, versus regulations designed to limit that exactly that thing.
DUBNER: So give us a thumbnail history of regulation of the Internet. Where did the appetite for regulation come from primarily, and how successfully, how successful have those people or institutions been at trying to regulate it as they wish?
SHIRKY: You know, the appetite as almost always comes from multiple places depending on who you’re talking about. So there was a period in the 90’s where people were very concerned about the moral effects of the Internet as they famously were about paperback books and comics. Whenever a new medium comes along people wring their hands about the effects on youth. And it has to be said that the conservatives when they worry about the social effects of a new communications technology are almost always correct. Right? When the Victorians said my God if this telephone thing spreads the way men and women will court one another will be completely blown up compared to our current norms, which, indeed, is exactly what happened. People said rock ‘n’ roll would lead to race mixing, well there you go. So the people worried about the moral effect of the Internet, which is to say it would be impossible to create an environment in which children only learn their parents’ view of the world is in fact completely correct. And there was an attempt at the time with the Children’s Online Protection Act and so forth to bring about a set of omnibus bills that would filter these, you know, filter the sites that were causing the moral panic.
DUBNER: Let me ask you this kind of more, kind of, broadly or maybe more naively, let’s say we could step back 50 years and I would say to you, Clay, I would like to propose a scenario whereby everybody has a computer, it’s on their desk, or it’s in in their pocket, or it’s on their eyeglasses, and in that computer you can use it to output, you can broadcast, or you can take in just about anything. If you want to…Anything that you want to write you can immediately publish to anyone else who owns one. Anything you want to sell, create, think up, draw, steal, invent, et cetera, and anybody can communicate with anybody else, pretty much for free. And I can sell stuff, you can buy stuff including physical goods, organs, you can sell sex, things that are licit and illicit. If I were to describe to you that scenario and I would ask you, Clay, to draw up the regulation, legal and otherwise, that would properly constrain,, what would you do and how would it be different from what we actually have?
SHIRKY: So I have the easiest way of abandoning this question, which is the most disappointing for radio, alas, which is that I’m a Pragmatist, uppercase P, in the tradition of John Dewey. And the Pragmatists regard a lot of the solutions to these problems as things that, you know, as I often say to my students who try and advance to forestall the problems that they might have when developing things, there’s a large class of problems that you don’t solve until you have them. Right? You hear all of the ethicists saying oh the technology is outrunning ethics, to which the pragmatist answer is that’s exactly what we want. If you and I were to sit down right now and say well let’s draw up regulations for time travel and telekinesis we would, it’s a ridiculous question. So although I don’t think, I couldn’t point to the current regulatory environments, decidedly plural and say we have got it exactly right, we can absolutely promise you that anyone looking at this at 50 years removed would have made an absolute hash of this, because it was impossible to imagine what the second order effects of the technology would be. And in fact one of the things that I study when I study social media is they ways in which people got a lot of this right about, you know, a lot of the early theorizing about this network, right about access to data, access to content, and they completely missed or misunderstood the social aggregates, in particular the group communication part inasmuch as it replaces the telegraph, the telephone and the fax machine.
DUBNER: Why do you think that is? I have a theory, which I’m sure is not at all right, but I’m curious to know yours.
SHIRKY: It’s I think for two reasons. One, people have enormous anxieties about social change that they don’t about other kinds of technological change. And I wrote about this a little bit in my first book, “Here Comes Everybody.” I grew up when Popular Mechanics was saying, you know, you’ll fly your flying car to work. And your…
DUBNER: What you don’t?
SHIRKY: Alas, Citibike is about the best I can do at this point. And your housewife, as a wife, will have all plastic furniture that she can hose of. But the idea that gender relations would change, nowhere, nothing in Popular Mechanics ever suggests that anything other than me going to work and my wife staying at home was going to be the norm. And so we would have big arguments about which, is this the Space Age or is this the Nuclear Age? Is this the 1970’s, is this the Space Age or the Nuclear Age? And it turned out it was neither, it was the Transistor Age and the Birth Control Pill Age. Those were the really important technologies. And they were important not because they were big photoready government projects, but what individuals chose to do with them in aggregate lead to social change that no one was in control of. So the idea of trying to guess what a technology will do when regulating in advance is to me I think almost the surest way to guarantee, just by the way the regulation would be structured, that you miss out on the opportunities. You know, when the bioethics group was convened under Bush, this is off the subjects of the Internet, but they made the same mistakes, which is essentially they imagined what we could do with stem cells, sequencing the genome and so forth standing from a position when those things were just barely possible. And that kind of regulatory hand is in general the worst way to figure out what a new technology can do.
[MUSIC: Christopher Norman, “Can’t Let Go” (from EP1)]
DUBNER: So Clay Shirky’s point is well-taken: it’s a fool’s game to try to anticipate how any given new technology will roll out – whether from a regulatory angle or otherwise. As Niels Bohr is said to have once said: “Prediction is very difficult, especially if it's about the future.” Who would have predicted, for instance, that all the nastiness that happens online, and in video games, and so on, might result in less nastiness in the real world? Here’s Steve Levitt again:
LEVITT: Just the empirical reality when you look at the data is that adolescent males spend an inordinate amount of time playing video games. And I don’t know what the actual numbers are, but I think now that those kinds of numbers are starting to surpass TV watching.
DUBNER: Well, and the other fact is that the number of crimes that are committed in the world are disproportionately committed by young men. So it wouldn’t be such a leap to assume that the prominence of the one would potentially decrease the other?
LEVITT: Absolutely. I think anything you can do to keep adolescent and 20-something men busy is likely to reduce crime. Whether it’s working on a job, whether it’s having them in school, whether it’s having them play video games until three in the morning, if the kids are out doing that they’re out not doing regular crimes. So I think the premise is incredibly simple. And it just seems to be the case empirically that what young men love more than anything else is video games. They like that more than school, they like that more than TV, they like that as much as anything we can find. And it’s one of these great examples of unintended consequences. Now, is there a lot of incredibly strong, careful academic work that supports the conjecture I just made? Probably not, but on the other hand I think it just stands to reason that if you find an activity that keeps potential criminals busy for six waking hours a day, then it probably makes sense that they’re going to be doing less crime.
DUBNER: And theoretically, the more portable that activity gets, i.e. a smartphone on the subway, the better off everybody is?
LEVITT: I mean, so you’re taking the idea even further, Dubner, the idea that let’s not just distract them when they’re at home, let’s keep them distracted even when they’re out on the streets trying to play. And I like that idea, the only other possible counter example I can think of is that if you make video games only available at home, these guys never leave their home. And if you make them available everywhere and you can play video games in the subway, then you’re in the subway all the time. And then you take a minute or two off from the video game playing to commit the crime, before going back to it. That could go either way.
DUBNER: Let me ask you one last thing about this. So there are schools of thought about the way people should behave online. And these days with most people being online, everybody is engaged in this at some point. So even if it’s you read a newspaper online, or you read blogs, or you tweet and retweet what others say, there’s a concern about about nasty people, about ad hominem attacks and people being insulting, and rude, and vulgar, and racist and so on, to the point where now some publications are, for instance, shutting down comments sections rather than having to deal with it. So, I understand the cost of that, and I understand why people get upset about it, but I feel, and I’m curious what you feel, I feel that the more people are allowed to rant online, the better off society will be on net because if that virtual ranting can let them get their ya yas out, and their anger out, and their whatever out, then it would stand to reason to me that there’s a smaller likelihood that they’ll actually punch somebody in the face for real at the grocery store if they can punch somebody in the face. And maybe it’s us that they punch on the face virtually, which is one reason that we let people say terrible things about us online. I think it’s kind of great to have that filter and flow. I’m curious what you have to say about that.
LEVITT: Well, I think that’s hard, because number one I think people feel a lot of pain from those kind of online rants. So you’re implicitly saying that the punch in the face is far worse than the online rant, but it’s not completely clear to me that that’s actually true, that the amount of pain…I mean…because one of the things about the online rant is that they live forever, right? So that someone can say something hurtful and it can stick around and haunt you on and on and on. The other things is one of the reasons that people don’t do so much of this sort of ranting and fighting in public is that it’s really socially not accepted very much. And I think there’s a reason it’s not socially accepted because it’s not very costly for the ranter, but potentially very, very costly for the person who gets ranted against. You know, it just reminds me, I don’t know why, of when SuperFreakonomics came out and we have the stuff on climate change, and the online rants against us were untrue, they were absurd, they were ad hominem. It was just a bad, ugly situation where a group of people who were very emotional about what we were saying, working essentially in unison to try to discredit us, and we really didn’t have a good vehicle for fighting back. But what I remember so vividly is I remember, I think it was in Washington D.C., where we were giving a talk, and a heckler got up and began ranting in exactly the same way that the online rants were going against us, and the people in the audience just told him to sit down. And what was so weird was that he sat down. And that he was so bullied, and so easily cowed by a few people saying don’t do that. So he got to say a little bit of his peace and we got to respond to it. But somehow it seemed to be much more productive than if he had just ranted and ranted and ranted and hadn’t let us talk at all, which is really what I think you get online. So I don’t know. I’m not really that much in favor of giving…I mean, sure people can rant, it’s such a hard job for other people to sort out what’s true and what’s false, that rants that can’t be verified for truth, I think are terrible. And we don’t have a good online way of saying whether things are true or not. And that’s where I think the cost is.
[MUSIC: Bronze Radio Return, “M.O.T.R. (Middle Of The Road)” (from Up, On & Over)]
DUBNER: So that’s where we’ll leave it this week -- on virtual nastiness versus real nastiness, how the Internet polices itself. But I’ve got a bigger appetite for this kind of thing. So on next week’s program, we ask a similar but bigger question. If the Internet is, generally speaking, a self-organizing, self-policing organism, could the same be said about the world in general? Is there really someone “in charge” as much as we think? To answer that question, we’ll start – where else? – in a skating rink. We’ll also hear from Bill Bradley, Matt Ridley – he of the House of Lords Ridleys -- Alexi Lalas, and we’ll take to the field with some Ultimate players, Ultimate Frisbee to people who don't know any better, to see what happens when you play a sport with no referees. That’s next time, on Freakonomics Radio.