So, what I want to talk to you guys about is the maybe some of the less positive [inaudible]. Some of the less positive aspects of [inaudible]. And, and where this comes from is a research I've been doing in Iraq and Afghanistan. And one of the things that goes on in these places, is, as the wars have been going on in Iraq and Afghanistan, The ICT penetration has gone way up, right? So, this is kind of an estimate of the proportion of the population of Iraq. It's, you know, a country of about, about, twenty some 22, 23 million people. This is an estimate of the proportion of the population covered by the network of this company, Zain, which was the monopoly provider in the parts of the country that experienced most of the war from 2003 through 2008. And so, you know, as you can see Zain, Zain is kind of coverage is growing dramatically throughout the war. I can show you some pictures of this. What these are is this is just a map of Iraq. That kind of do you guys have a pointer? Yeah, I have this one. Okay, yes. I see, like, so in this case, it's cell phone coverage. So, each dot here represents a, a, a tower, that typically has kind of kind of three antenna on it. So, you could sub those two slides, too, if you want. No, it's okay. Just a point will be great. Okay. Cool. So, so just to highlight to those guys who aren't familiar with the rocks geography. You know, you've got you got Baghdad right here, you've got Bosford down here. And basically, what, what these are showing you is in the year of 2004, the black dots are the towers that were active at the start of that year. The red dots are the towers that were built during that year and the gray dots are all the towers that will actually be built and down here, this is just the number of towers added per month. And so, what you can see, as, as the war is going on the, cellular network is being built out. And they build out in kind of a sensible way, Right? You start with kind of putting in cell phone coverage around the main population centers, and then you start to build on the main transport routes down to the south, and out to the out to the, out to the west, to Syria. And then, you know, you start to kind of build the network out in 2005. Fill it in. You know, You continue to build things in 2006. And you notice at like the height of the war, from mid 2006 to mid 2007, when the violence got really, really bad. basically, construction activity slows down a great deal and then picks right back up as violence starts to come down in the main parts of the country. And, you know, they continue to, to build out, flesh out the network in 2008 and 2009. Some of this is improving coverage. Have you guys talked about how cell phone networks work yet? No, we didn't. Okay, yeah. So, some of this is, like, cell splitting and building out the network so it can cover more calls at higher quality as, as demand and usage goes up. And some of this is covering new areas you know, where they want to they want to provide coverage along these highways so that people have continuous use and, And they can bill more and what not. What's interesting about this is if we normally think about like, cell phones as kind of an unalloyed good, but if you think about the problem of producing violence, right? So, organizing an ambush, building bombs getting people together to kill your enemies for all of these things, efficient communications are really important right? So, the most obvious one is, if I want to get a bunch of guys with guns together at a particular point and time to kill you I need to be able to call you, you, and you, three of you guys and say, alright, we know that he's going to be traveling down this road at this time so you position yourself here, put yourself here in a blocking position. We'll be here and here. You stop him with some, some fire. We'll get him in a cross fire. He's done and we've accomplished our mission. Doing that is much easier if you can, you know, call each other or text each other, in the same way that it's easier to get together with a bunch of friends for a movie, Right? So, kind of coordination is easier. You can also think that it's easier to put out people who are kind of spotters or will tell you when your targets coming. You can also put out things like this. So, this is a weapons cache from Iraq. So these are some old land mines. These are rocket propelled grenades. This is a different kind of bazooka-type device. And this is a 155 millimeter artillery shell. You know, hundreds of thousands of these things were scattered around Iraq after the end of the war because we didn't secure any of the armories during the drive into Baghdad. And what the insurgents have done here is they've attached a cell phone to it as a detonator. And so, you know, you could use cell phones like this in a few ways. You could use them as a timer so you actually set the alarm on the cell phone. And then solder to the circuit to the speaker, a wire that you would then connect to a blasting cap inside the 155 millimeter artillery shell to set it off. So that didn't required cell phone coverage, but you could also use them in other ways. You could use them so that they would be command detonated, Alright, so you would place a call to the phone. So, you see a coalition vehicle driving by. When it gets to the right spot, You call the phone. The ringtone sends an electronic signal, blows it up. So, that was very useful for the insurgents. They quickly though, the, the, the coalition forces quickly figured out that this was going on so they started jamming. So then, these guys started setting, setting these with things where you would prime the circuit and then call it and then it would be set to explode when the call was terminated, Right? And so, when the jammer got to the point where it shut off, The cell phone signal to the phone, Then the bomb went, boom. The problem with this, of course, was you had to, like, figure out the range of the jammer, right? As this vehicle is driving down the road, at what point does it stop the signal? And then you had to kind of set the triggering device sufficiently far ahead of the sufficiently far offset from the explosive itself. That, as the vehicle is driving down the road, Gets to this point, The range is out here. Cuts off the signal to the phone here, That then sends the signal back to the bomb here to trigger. So, you then got into these very kind of morbid arms races where the guy, the coalition guys would then put poles on the fronts of their vehicles of varying lengths to put the jammer out at different lengths from the front of the vehicle to mess up this process, Right, or they would start doing things like flying drones over the highway ahead of a convoy that would systematically kind of shut down the cell phone network and the IDs would go off as the drone went down and then they could drive down it. So, there's the point is kind of the introduction of this communications technology which normally we think of as being really good, cuz it helps people do things like get fair prices for their produce or either talk to their friends, or coordinates business. It also has this malign side to it. And so, you know, you might think that introducing cellphones and kind of, and better communications technology has this bad effect. You know, you also saw during this time lots of, lots of discussions of how the Internet was a useful tool for radicalization and how people could take advantage of websites and the ability to post videos of them, you know, engaging in what they considered to be heroic acts and kind of victorious attacks against the crusader enemy. There was lots of discussion about how that was making it easier to motivate the insurgency, Alright? So, from all those perspectives, you might think this was a bad thing. But, you know, on the other hand you might think that, look, for the insurgents in Iraq and to a lesser extent in Afghanistan, if the people you were fighting knew your geotemporal coordinates so they knew where you were going to be when, they had basically an unlimited ability to put very fast moving pieces of metal into your body, Right? They had vehicles flying overhead that could drop bombs on any point in time and space at anytime of day or night, Right? They had kind of massive numbers of helicopters and highly trained individuals who could go into pretty much anyplace anytime they wanted to and shoot you in the head. And so, being, kind of your critical vulnerability was your ability to operate without being known to the people you were fighting, Right? So, this is not like rural African insurgency. This is something very different. And so, in that sense, cell phone communications create a problem for you, and they do it in two ways. One is, if your guys are talking on their phones, Right, one thing that the, that the counter insurgents in Iraq and Afghanistan were very good at is picking up signals intelligence, alright? So, if you have people out there who are doing what any person with a cell phone does is you want to talk to your friends and make it easier to coordinate your activities be that kind of good or that creates a huge vulnerability. But the other is for the people who aren't in the, aren't kind of fighting the, the coalition in the case of Iraq, you make it a lot safer for them to react to things this. So, this is a tip card that you could read if your TAs had their resolution properly set. [LAUGH] but basically, so this is one from 2010, in Baghdad, but basically throughout the war, coalition forces were distributing things like this. And, you know, since it's kind of hard to read, what this one basically says is, you know, have you seen someone doing something bad and you want to do something about it, but you don't want to pick up a gun, well, just call us and tell us who they are and we'll deal with them, Alright? And so, in if, if, if this is going on, then the introduction of, of kind of cellular communications can have a really big effect on the insurgency in a couple of ways. One is, it's a lot safer to text this in than to call it, cuz no one can overhear you texting. Moreover, landline penetration in Iraq was really low. And it's a lot safer to pick up the phone and call in some information than it is to actually meet with someone from the, the coalition or the Iraqi police. And in fact, kind of the way informants had to be handled to prevent them from being identified as informants is the coalition forces or the Iraqi forces would go to like an apartment block, pull out all the men. Choose five, ten percent to kind of take them one by one through an interrogation. Choose five to ten percent to beat up, Including their informant. And then send them back, and that way, it wasn't clear who the informant was. And this is something that the British did in Malaya. Again, it's a very common thing. That was what you had to do for your informant to survive, because otherwise he would be identified as an informant. If he can call in that information, you no longer need to go through that process. And so, it's a lot easier to get information. And so, all of that leads to the prediction that introducing this technology might need, lead to lower levels of violence, Right? Now, since in lots of places in the world where there's kind of high levels of political violence, the ICT infrastructure isn't great. And since it's really easy to put the stuff in, you might want to know which way this thing, this thing runs. And, you know, so we did some analysis. So, what this is, is this is just another plot to show you data, cuz you're engineers. How many of you guys are engineers, by the way? Everyone? Almost? Pretty, is it everyone? You know a couple of non-engineers? Okay. I'm with the non-engineers. So, so what this, is this is just the monthly attacks against coalition and Iraqi forces from February 2004 through February 2009. And each of these plots is a different district of Iraq. So, District of Iraq is about 200,000 people in most of the country, About 900,000 people for places like Al-Muqtadir, which is an area in central Baghdad. And in blue, I've just plotted a number of insurgent attacks per month. And you can see there's loads of variation across the country, both in kind of the timing of the peaks. Right here, in Hit it happens at a very different time than here in Mahmudiya. So, Hit is an area in Anbar Governor, mostly Sunni, Mahmudiya is in a mixed area in Baghdad. And then in red is the number of new towers built in that district in that month, Right? So, you can see, like, in Karch, so this is central Baghdad, mostly the green zone, you know. They built a ton of towers there, early on cuz it's like the wealthiest, most densely populated part of the country, so they need to provide a lot of service. But even when the violence is really bad in the area, they're still adding more coverage. And so you know, what we do with this is we take this and we do some statistical analysis where we basically study the relationship between changes in the number of towers in an area and changes in the rate of insurgent attacks, Right, To try and get at. Is this technology is this context like a force for good or a force for, for evil? And it turns out in this context that it is a force for good. So, so basically, increasing the density of coverage within these large districts kind of increasing the number of towers by one standard deviation, decreases the attacks by about ten%.. So, it's 1.1 fewer attacks in the next week. So now, just to, just to kind of benchmark this. A tower cost between fifty thousand dollars and two hundred thousand dollars to set up, Depending whether you're setting it up on an apartment building or building it on a green field site. We need to build the tower. So you think about the cost of one attack. Well, if no one gets killed it's pretty low. But if one person gets hurt, That's, you know, millions of dollars in legacy health care costs for the VA. If a vehicle gets destroyed, that's hundreds of thousands of dollars to replace. If someone gets killed, right, You have kind of however you want to value life. Most insurance companies do it at about four million dollars for kind of a working age person with a family. And so, so, so there's kind of a big, big decrease and it's a little bit bigger in the places that were most violent, city and mixed areas. The other thing though that we did which was kind of neat, is we looked at specific local coverage areas. And we compared what happens when you turn coverage on, in an area, where you introduce new coverage versus one there wa, where there was already coverage, and what you're doing is just making the network denser. And so, these areas are about four kilometers in radius in urban areas, about twelve kilometers in rural areas. And when you turn tower on, towers on, you get about one less attack per fifteen day period in these little catchbin areas, for the ones that introduce new coverage, relative to the ones that don't. So, you're, you're kind of driving down violence, so that's nice. So that's the story in Iraq. So, anyone want to guess what this is? You guys are engineers, man, come on. [inaudible] Huh? [inaudible] Locations of ambushes? Yeah, exactly. This is the decibel level on the GSM network of the largest telco in Afghanistan, around Kabul, Right. So, this is Kabul. And each of these little dots is a tower, and you can actually kind of, you can see, the resolution's not great here, but you can actually see the three antenna going out from the tower. Each covers about 120 degrees. And the cutoff here is the negative, I think it's, I think we did this one at the negative 110 DB level, which is about the, the lowest you could possibly get a call through on the Afghan network. And so this is Kabul, and this is the road up to Bagram Airbase, Alright? And so, so what's interesting about Afghanistan is it's a topographically complex area. So, you know, the reason, for example, that these towers don't provide any, this tower here doesn't provide any coverage here. There's a big mountain in the way, right? And so, this thing is set up in the hills to provide coverage for some population centers. You know, this is kind of an overall picture of what, what the coverage in Afghanistan looks like. So, the blue areas are covered areas from this largest telco. And the little dots here are estimates of the population for each one kilometer cell actually, each 30 arc second cell, from land scan which, which is this project that does population estimates. And what's interesting about this is, is, just to highlight two things, is, this is really a sparsely populated country. These guys cover between 60 and 80% of the population of the country depending on which estimate you use. And like [LAUGH] it is not a lot of blue on the map. The other thing though that's interesting about this is that if you zoom in on some areas so this is Kandahar. This is kind of the heartland of the Taliban. Kandahar Airfield is here. This is the, kins of second largest logistics base in the country. This is Lashkar Gah, kind of the center of opium production in the world. You see that the cell phone guys, they're covering the main roads reasonably well. And this is the, this is the main road around Afghanistan. But there are all these populated areas along river valleys where they're not providing any coverage. And, so you might, you know, you might wonder why that, why that is. And, you know, just to, to give you some background on this, part of it is, there were like 200 people with phone service in Afghanistan when the US and NATO invaded in, in November 2001. Roughly 65% of the population now has a cell phone, Alright? So, the country went from, kind of a sixteenth century level of communication with the outside world to better than most places in Africa in, in, in, ten years. It's kind, kind of remarkable. So some of this is just takes awhile to build this stuff out. You know, and this is kind of what it looks like. This is just the number of towers active in the 40 most violent districts of Afghanistan. And in the big cities like Herat or Kunduz or Mazar-i-Sharif, You know, they, they're building more towers. That's not surprising. What is surprising though, I'll show you this before I show you the last one. Or, I don't know if this is surprising. Is something happening in Iraq that never happened in, in Afghanistan, that never happened in Iraq. So, in Iraq, the insurgents would threaten the telecommunications providers if they didn't keep their towers up. Alright, so there are lots of reports from the guys who maintain the networks, saying we're, you know, we have this problem which is whenever a, a tower goes down, say, because we let the generator run out of fuel, we get a call from someone threatening our lives, because they really like having the towers up. Alright. The opposite is true in Afghanistan. So, each of these dots represents a tower. And the ones in black, These are all the towers that this firm has shut down, because someone from the Taliban has come to them and ,, you need turn this tower off. The ones that are in yellow, that show up here as impacted, You know, so these guys down here around Jalalabad. These are ones where someone from the insurgence has come to them and said this tower will be off between six:00 p.m. And six:00 a.m. or we'll blow it up. Alright. And so, they're only providing service for part of the day. Now, why would the insurgents do that? People kind of in Afghanistan clearly love their cell phones cuz they've all been buying them as fast as they can. Well, if you think that people are using this to call in information on you, Then you might not want them to be able to talk at night when you're moving around. And so, what that, what that means though, is that for us to do the kind of estimation that we did in Iraq, in Afghanistan, is not possible. Because whether or not you have coverage is a function of a bunch of stuff having to do with the market incentives of the cell phone company, but also whether the insurgents are active in an area and want people to be able to use their cell phones. Alright, so you can't actually do the estimation the way you did in the other case. That said we did do some, some, we looked at some interesting stuff so what this is showing you is it's just kind of the correlation, bivariate monthly correlation between revenue at a tower and the number of attacks. And each dot here is a district of Afghanistan. So, like, Kunduz here, right? In Kunduz, when the number of attacks in a given month goes up by one on average, revenues in thousands of dollars from the towers in Kunduz go down by about $500, Alright? So, but, you know, it's kind of all over the map, so there's no clear relationship there. And, you know, so that's really all I wanted to talk about is just, just to highlight the fact that, that communications technologies that you guys are studying here are mostly in kind of well-ordered, well-mannered, nicely developed western countries viewed as an unalloyed good, Right? And there's a huge literature that says, even in developing countries, they're great, Right? There are these papers that show that when you introduce cell phone coverage along the coast of India, all of a sudden the market prices for fish across different markets along the coast, converge to one price. Because what happens is, the fisherman out doing the fishing, as soon as they haul in their catch, they get on their phone, they figure out where they can get the best price, and they go there, Right? And so, the introduction of cell phones there basically disintermediates the market. It takes out the profits that all the middle men in the market were making and gives it to the fishermen, Alright So, people look at that and they're like, great, markets are more efficient. The same thing happens in grain markets in Niger. And so we generally look at this technology as not only good but in these contexts where there's political contestation, it's not at all clear that that's the case. And this is why in places like Kenya Egypt, Thailand, when there's kind of a protest and violent activity or people are acting politically in ways that are threatening to the regime, one of the first things that happens is there are efforts made to restrict communication or restrict the ability of people to talk to each other and coordinate. So, it's kind of a more, in some sense if you are interested in places that are not yet kind of decisively organized in the ways that the United States or most Western European countries are, All of the technologies you are talking about have lots more, lots of kind of more interesting complicated implications.
