|
 |
July 30, 2003 PERVERSE INCENTIVES....Instapundit weighs in with a typically sober assessment of the Pentagon's decision to kill its "terrorism futures market":
Since no one in the pro-war camp seems to understand the actual issue here, let me take a stab at it. The problem isn't that the Pentagon's idea wouldn't work — that's an empirical question that can only be answered with actual research — the problem is the perverse incentives it supplies. Suppose, for example, that a terrorist killed Dick Cheney and a couple of days later there was a photo of some guy in the New York Times grinning because the government had just sent him a check for $10,000 thanks to his shrewd investment. Does it still sound like a good idea? Do we really want to create a government sponsored market that creates a whole class of people who are actively rooting for terrorist events to occur? Remember, there are some things that are bad ideas even if it turns out that they work well for their ostensible purpose. It's the law of unintended consequences, something that Paul Wolfowitz apparently understands better than the blogosphere. (And given his recent experience with it, I suppose that comes as no surprise....) UPDATE: Brad DeLong has some expert commentary, although it's more about whether the plan would actually work, not about the issue I raised. Posted by Kevin Drum at July 30, 2003 09:40 AM | TrackBackComments
From the NYT article: Senator Pat Roberts, the Kansas Republican who serves on both the Intelligence and Armed Services Committee, called the market plan "absurd" and added, "It seems to me they are way off base and somebody should bear that responsibility, and I think we know who that is." It's all Tenet's fault! The CIA failed to prevent the Pentagon from going public with this idea! Posted by: Haggai at July 30, 2003 09:48 AM | PERMALINKThe whole idea is silly, to be honest. The idea that somehoww a market is absolutly the best at parsing information is silly. Ok, the market is the best in a perfect market, where all information is known to all parties. That is not the case. This is very silly. You might as well read blog posts to predict a terrorist attack. (And you might have more success, to be honest) Posted by: Karmakin at July 30, 2003 09:51 AM | PERMALINKWouldn't the system also be open to fraud? If there were large monies to be made if an assisnation took place--wouldn't this incent some nut to begin thinking about how to get it done? Just like Wall Street analysts, we'd have terror analysts talking up there angles and some of them would be trying to make existing problems worse in hopes of a pay off down the line. Posted by: Ken at July 30, 2003 09:53 AM | PERMALINKInsty's being the high-minded arbiter between left and right again, I see... Posted by: jesse at July 30, 2003 09:53 AM | PERMALINKLet me see a show of hands of people who thought that this would really create "a whole class of people actively rooting for terrorist events to occur." Not to mention- who cares who is rooting for what? I have been rooting for the Steelers to win the Super Bowl for the last twenty years- you see where that has ended every January. Posted by: John Cole at July 30, 2003 09:54 AM | PERMALINKThe notion of a futures market on terrorism doesn't work as a predictive tool though, so the whole matter at this point is moot anyway. BTW, insurers have been in the business of assessing risk for centuries, and they've found that while you can profitably cover a broad range of people who may be at risk, you cannot predict exactly which person will suffer a given loss. (Those who try may end up in the slammer as was the case with Fred MacMurray in Double Indemnity.) Posted by: David W. at July 30, 2003 09:55 AM | PERMALINKThere's also a flip side to the problem you just mentioned, Kevin. If (and correct me if I am mistaken) this terrorism futures market was intended to be a way to predict terrorist acts based on where speculators put their money, the signal that sends will encourage the government to act to prevent such acts, thereby reducing the chance I'll earn money. So a bunch of speculators bet that Dick Cheney gets assassinated. The government looks at the trend, gives Cheney extra security, steps up intelligence, etc. Turns out that there was plot and it's stopped. Now, that's a good thing. For me, though, it's economically bad. I lose out. In fact, unless I have some power to counter the government's moves, I have to rely on government failure in order to win. Since the government has access to resources that I don't have if I want to ensure my outcome, this makes for a way too risky speculation. Posted by: Brian S. at July 30, 2003 09:56 AM | PERMALINKNot to mention- who cares who is rooting for what? I have been rooting for the Steelers to win the Super Bowl for the last twenty years- you see where that has ended every January. -- Well, when you have money invested in the team and the ability to affect events to ensure the quality of your investment, your "rooting" is a lot different than Jim Schlub who still swears Kordell was a good QB. The more accurate comparison would be Plaxico Burress laying down $500,000 on the Browns when the two teams are playing. Posted by: jesse at July 30, 2003 10:01 AM | PERMALINKBrian S writes: "So a bunch of speculators bet that Dick Cheney gets assassinated. The government looks at the trend, gives Cheney extra security, steps up intelligence, etc. Turns out that there was plot and it's stopped. Now, that's a good thing." More likely, the plotters saw the trend too, and avoided the high-priced contract entirely. The government, based on the trend, steps up security around Cheney, but the actual plot is hitting something else, which is unprotected. Another problem is that most of the trades would be based on White House announcements. If the market were running right now, there would have been a spike in contracts dealing with 9/11 style attacks, due to the recent warning, not due to traders having any special insight. I shouldn't have to point out the flaws in a predictive device which is mostly driven by White House announcements. Posted by: Jon H at July 30, 2003 10:03 AM | PERMALINK>>Does it matter that this sends exactly the wrong signal to the Pentagon about the consequences of efforts to find original ways to fight the terror war? Yes. so basically we are supposed to - nay, we HAVE to - embrace every scheme cooked up in the Pentagon by the neocons - no matter how illogical, unconstitutional, or just wrong that it is, because to reject it would send "the wrong signal" to those poor sensitive military types from a presumably ungrateful country... This ranks up there with the self-serving Iraqi invasion justifications from late last summer when the meme was that because Bush had invested time in it that we HAD to invade because otherwise it would make him look bad. Just how old is Glenn Reynolds anyway - is he out of middle school yet? Posted by: Andy at July 30, 2003 10:04 AM | PERMALINKMy favorite Insty quote was: "Whether or not the Pentagon's idea is a good one depends on details I don't know about. But the lame criticism makes clear that the critics are -- as usual -- clueless on the subject." In other words, I don't know anything about this market but I do know those who oppose it are wrong. This is a law professor? Posted by: achilles at July 30, 2003 10:04 AM | PERMALINKSuddenly I see it from yet another angle...why would anyone invest in a market where the popularity, and therefore (according to theory) the likelihood, of my contract being fulfilled will create data US Government will then use to undermine the future, and thus my contract? If ten thousand people bet that Bremer will be assassinated, won't the government then take action to prevent this eventuality? Or is the theory so voodoo-religious in nature that it takes all of this into account and believes, as it seems to, that the results produced by the market will be accurate indicators anyway? The government is hoping we'll all bet on people like Osama Bin Laden getting killed or captured...or are they? How much money are they prepared to shell out? Posted by: theperegrine at July 30, 2003 10:06 AM | PERMALINKThe only way I could see this working would be a one-shot opportunity. You find out where people are putting their money that isn't predicated on publically available information and investigate those people. Unless you can find out where investors are getting their information, or let people invest anonymously, the market can't work. And the latter choice is a far worse idea than the former, but the former will destroy the usefulness of the market after one occurence. Posted by: jesse at July 30, 2003 10:09 AM | PERMALINKPeregrine: Great minds think alike, eh? See my post above. Jon H.: Good point. I hadn't thought of that - the market could be used as a way for terrorists to refine their own plans. Posted by: Brian S. at July 30, 2003 10:14 AM | PERMALINKjesse- You've struck on another element of the project that bothered me: The market would be so dominated with patriots betting on 'good' things to happen for 'us' and savvy investors trying to 'beat the system' that the results couldn't possibly be useful except to reveal information about the people who are betting. If everything I bet on comes true, will I end up in the oubliette below John Ashcroft's toilet? It's not like I have family to miss me if I mysteriously vanished... Posted by: theperegrine at July 30, 2003 10:17 AM | PERMALINKThis is a completely overblown issue. A huge market already exists that terrorists can manipulate and make money on attacks. It's called the US stock market. So that's a nonsensical complaint. I do fail to see the overriding benefit of this new proposition though. Posted by: spc67 at July 30, 2003 10:18 AM | PERMALINKYou also have the problem that those behind terrorist attacks have an incentive to game the system. Markets only work as information gathering devices if all the participants are playing to win. Posted by: J. Michael Neal at July 30, 2003 10:19 AM | PERMALINKThe whole concept is one more example of PNAC ideas which occur in the abstract. Remaking the middle east is a good idea, if you ignore the fact that real people in the middle east might disagree with your methods. A market in terrorism futures is a good idea, if you ignore the fact that real people might take the opportunity to game said market for their own ends. There's an entire school of thought in economics called behavioral economics that recognizes that rational decision making is, often, not the natural state of humans. Neocons do not get this concept. Posted by: Flory at July 30, 2003 10:20 AM | PERMALINK"I'll take Civilization Will Be Destroyed by 2005 for $200, Alex." Posted by: The Moron at July 30, 2003 10:21 AM | PERMALINKFirst off, if I'm betting $10 grand there's gonna be a terrorist event at the Super Bowl, doesn't it behoove me to buy a clunker and blow it up in the parking lot? In full view of the TV trucks? Secondly, in response to Instapundit - what the neocons have never
gotten is that appearances matter. If you appear to be rewarding bad
behavior, that's what you'll get. not the best argument, kevin. no offense, but if that's the way markets worked, then why wouldn't people short a ton of microsoft stock, then shoot bill gates? for that matter, why didn't everyone involved with 9/11 short the entire market just prior to flying the planes in? the answer is that you have easily investigatible transactions in a regulated market. no one would be able to profit anonymously, because their accounts would be opened through regulated brokers. there are arguments against this (i pointed out on my blog that a futures market doesn't work right when the government stacks the odds - e.g. the higher the market sets the probability of assassination, the more resources the government would devote to protecting the involved leader). however, the possibility that you're giving financial incentives to anonymous terrorists isn't one of them. Posted by: Scott at July 30, 2003 10:28 AM | PERMALINKfor that matter, why didn't everyone involved with 9/11 short the entire market just prior to flying the planes in? Um, bad example. There were reports of a few investors massively shorting insurance companies and airlines in the days before the attacks. Last I heard they were looking into it for connections to al-Qaida. Anyone remember hearing news on how these investigations turned out? I don't recall any follow-up... Posted by: Kevin Brennan at July 30, 2003 10:36 AM | PERMALINKThe best comment on this scheme I have seen, and I apologize for not remembering who it was that made it, was that such a marketplace as DARPA proposed should be set up privately within the intelligence agencies, allowing low-level analysts to participate. Arguably, these people actually have access to the raw information that could result in a useful "market summary" and also they are the ones that get censored by layers of bureaucracy in the process of creating intelligence analyses for the policy makers. Posted by: cafl at July 30, 2003 10:43 AM | PERMALINKI agree that there is a set of perverse incentives. But that doesn't mean that the idea should be shot down with stupid sound bites. Welfare sets up perverse incentives, but we try to mitigate the incentives in different ways. Off the top of my head we could say you don't get the money if you are involved in the plot. It is just like if killing your father and being denied money from his estate or will. Maybe that kind of thing wouldn't be enough to save the problem. But notice that the idea was shot down BEFORE we even got to the proper question in the debate; that is the question of mitigating the perverse incentive. Posted by: Sebastian Holsclaw at July 30, 2003 10:43 AM | PERMALINKKevin Brennan writes: "Anyone remember hearing news on how these investigations turned out? " I think it was something like "Nothing conclusive", or something to that effect. Which probably means "It was Saudis, and we weren't allowed to look too closely." Posted by: Jon H at July 30, 2003 10:44 AM | PERMALINKBut you guys are misreading Kevin. he's not talking about market manipulation, he's talking about the simple fact that in such a market, some people will get payoffs for bad things happening, and that's just too perverse for the government to encourage. Anyhow, as Ben Graham told us 70 years ago, in the short run, the market is a voting machine, in the long run, it's a weighing machine. A futures market is all about the short run, so its predictive value in fighting terror is pretty minimal. Posted by: howard at July 30, 2003 10:45 AM | PERMALINKNot just insurance and airlines. Companies with offices in the WTC were also reported to be heavily shorted on Sept. 10 at the Chicago exchange. Posted by: BruceR at July 30, 2003 10:49 AM | PERMALINKnot the best argument, kevin. no offense, but if that's the way markets worked, then why wouldn't people short a ton of microsoft stock, then shoot bill gates? for that matter, why didn't everyone involved with 9/11 short the entire market just prior to flying the planes in?-- Well, there were reports of that happening, actually. And the point of terrorism is that you're already commited to the act. You're saying that people never work to manipulate the markets in their favor by doing illegal things, but Enron alone sinks that boat. Posted by: jesse at July 30, 2003 10:51 AM | PERMALINKAnyhow, as Ben Graham told us 70 years ago, in the short run, the market is a voting machine, in the long run, it's a weighing machine. A futures market is all about the short run, so its predictive value in fighting terror is pretty minimal. Intelligently linking Ben Graham and 21st century terrorism. Oustanding! Howard, you're good. Sebastian writes: " the idea was shot down BEFORE we even got to the proper question in the debate; that is the question of mitigating the perverse incentive. " That's not the proper question. The proper question is, why would terrorists attack something they know we've increased security on? They would know, because it would be public information, through this market. Further, why wouldn't terrorists look at the market and use it to find suitably under-protected targets, or just to rule out any target specified in the market? Perverse incentives are a secondary issue; the main problem is that the market wouldn't work. Posted by: Jon H at July 30, 2003 10:53 AM | PERMALINKspc67, how is this a completely overblown issue just because another market exists? If you could manipulate two markets with your actions, wouldn't you split your money across both to increase your benefit/risk ratio? Of course, the real concern here is that the Pentagon's plan essentially goes out of its way to encourage such manipulation (read: encourages terrorism). And it would if it worked even remotely like the Pentagon anticipated. None of this gets at why I don't think this is an overblown issue, even if it's already been killed. The issue I have with this is that it was even promulgated. The stock market is wonderful for some things (like long term return on overall investment, long term meaning 10+ years, overall meaning across a broad, broad range of investments). It's absolutely terrible as a short-term (within the year) predictive tool for relatively discrete events. That the neocons don't "get" this fits right in with many of the results of their policies we're seeing today in the national economy, national security, and in foreign affairs. Like Rush, they are convinced that there truly is "The way things ought to be" out there. Their Utopia, our Distopia. Posted by: dasboat at July 30, 2003 10:53 AM | PERMALINKThanks, Howard. Yes, that's exactly what I'm saying. It's not that people can game the system or that it will motivate people to actually go out and shoot someone, it's just a question of whether the government should encourage people to be happy about horrible events. That would get very unpopular very fast. (Try to imagine someone getting a big check on September 12, 2001, based on a futures contract on the World Trade Center....) I'd also encourage everyone on this thread to be very careful about arguments suggesting the plan wouldn't work. It might, and it might not, and I'll bet that Robin Hanson and his colleagues are pretty knowledgable about everything that's been brought up here. The only way to find out if *it actually works* would be to try it. Posted by: Kevin Drum at July 30, 2003 10:53 AM | PERMALINKSince a free market is the best arbiter of everything, I think we should start using it for directing scientific research. Within a year we can have warp-drive and travel to the stars! Posted by: Tripp at July 30, 2003 10:54 AM | PERMALINKif there were people not directly involved in 9/11, but who had prior knowledge of 9/11, who shorted the u.s. market, isn't that a perfect example of what a futures market could lead to? how would these leads have come up except through a financial investigation stemming from the existence of a market? if we didn't follow on good leads, that's a fault of the intelligence, not the market, and since there was no evidence the terrorists themselves used the market to their advantage, it's certainly not a reason to fear the manipulation of this terrorism futures market... Posted by: Scott at July 30, 2003 10:59 AM | PERMALINKKevin writes: "I'd also encourage everyone on this thread to be very careful about arguments suggesting the plan wouldn't work." Why shouldn't we discuss whether the plan would work? It's fair game. We might not know any more or less about anyone else on the topic (hell, I'm a molecular biologist -why should I know anything about any of this?), but this is after all, a discussion. As I've written, I don't think it's a plan that will work. I don't think markets are in fact predictive in the short-term, and I think the short-term is where the action is on this. I've learned a lot on your blog about other issues, why not this one? I'm sure that someone will come back at me with why the plan will work. I'd like to know. Hopefully they'll educate me in a way that's not too painful! ;) Posted by: dasboat at July 30, 2003 11:01 AM | PERMALINKAny way you look at it, this market is a lose-lose situation. If you run it honestly then you're adding a new level of financial incentives to terrorist attacks, which terrorists can then cash in on, and this is just plain stupid. It'd also be destined to fail, since as others have pointed out, investing in an event would provoke the Feds to increase security and thus make it actually less likely to happen. Poor investment! On the other hand, it's perfectly reasonable to assume that if you buy $1000 in Dick Cheney's death and he catches a bullet the next day, you're going to get an interview with a CIA agent instead of a big fat check. And I don't need to explain how this would cause the market to give flawed data. I don't see any possible way that DARPA could hedge against all three of these logistical flaws. The plan is a big waste of money, something Insty should be more sensitive to. I have to laugh whenever I see a right-wing blogger complaining that all Democrats think anything out of the Bush administration is automatically bad and stupid. Seems like in this case, there's a lot more evidence that Insty thinks it's a good idea for no other reason than that it came out of the Bush administration. Speaking of which, here's my prediction for the unintended consequences of this market: The security budget on Bill and Hill gets doubled since every right-wing blog reader logs in to slap down a hundred bucks on his dream of a Clinton being assassinated. (Check out the Amish Tech Support Deadpool if you don't believe me.) Posted by: neil at July 30, 2003 11:04 AM | PERMALINKThe biggest problem with this is that I simply can't understand how it would possibly be useful for anything. Posted by: John at July 30, 2003 11:05 AM | PERMALINK....Did the Instapundit know about any of this stuff? Fat chance. Does he care that he was responding lamely and out of ignorance? Nope. Does it matter that this sends exactly the wrong signal to the blog/hacksphere about the consequences of efforts to find original ways to fight the war on critical thinking? Yes. Will the Instapundit take any responsibility for that? Nope. Posted by: Instahack at July 30, 2003 11:24 AM | PERMALINKKevin: You can try something like it at https://www.tradesports.com/ . And now that the idea is out there, no doubt other gambling venues will accommodate the interest. And intelligence organizations will keep an eye on them of rising threats. Posted by: Bill Woods at July 30, 2003 11:36 AM | PERMALINKLet's all please take ten seconds to remember that this futures plan
was under the auspices of the Orwellian Total Information Awareness
program that the GOP-led Congress, for God's sake, killed funding for
when the public got wind of its agenda. I said then they weren't dead,
and lo! they aren't. And yes, they're still run by convicted Iran-Contra
felon John Poindexter, chosen by Bush Jr. to spy on Americans. Kevin D. writes: "I'd also encourage everyone on this thread to be very careful about arguments suggesting the plan wouldn't work. It might, and it might not, and I'll bet that Robin Hanson and his colleagues are pretty knowledgable about everything that's been brought up here. The only way to find out if *it actually works* would be to try it." The Iowa Political Stock Market has been cited here and elsewhere as an example of a similar sort of futures market, but the IPSM's record of political predictions has been pretty mixed. Markets are not an ideal mechanism for making predictions in the short term, they are basically ways of trading goods and services. Even with the best information available, insurers for instance can't predict which _particular_ ship may go down at sea. This is why you have something like group insurance in the first place, which spreads the risk those in a given pool may be subject to. That's why terrorism insurance is generally expensive because insurers have a hard time accurately assessing the risk terrorism poses to a given client. It should be noted that Jewish organizations had their terrorism insurance rates skyrocket *after* 9/11, not before. Posted by: David W. at July 30, 2003 11:42 AM | PERMALINK"(Try to imagine someone getting a big check on September 12, 2001, based on a futures contract on the World Trade Center....)" Try to imagine that said someone was in Egypt or Indonesia because their government had set up such a program. We would not be amused, to say the least. Would you want a hospital to set up a market for doctors to invest based on which cancer patients would die? Tripp makes a good point too. Maybe this could work, maybe not, but it has enough downsides that the reasoning behind it needs to be more than the neocon perception that the free market answers all of life's problems. "The biggest problem with this is that I simply can't understand how it would possibly be useful for anything." Another good point, John. Who's going to have to figure out the import of and act on any data coming out of such a market? Our security and intelligence agencies. If we're so distrustful of their abilities now, would we feel more comfortable if such a market had predicted bombings of shopping malls on 10/31/01 (big email rumor, for those who don't recall)? Posted by: denise at July 30, 2003 11:45 AM | PERMALINKThe way money flows in and out of international markets is already perverse and destructive (speaking as a broker); the winds of change are controlled by a relatively few mega-investors (organizations, corporations, LPs, governments) who can literally break markets and ruin economies by simply moving their money around. The asian financial crisis was more or less people simply pulling their money out. Not you, not me, maybe a tiny bit of your dough in a pension fund or something, basically just people and organizations that make money by trading money. It's really, really perverse when you consider the power people wield by simply trading slips of paper. None of these international money-giants do it for any other reason than to make money. They don't move billions and trillions to feed the starving and house the homeless, or rebuild war-torn countries, or lift a nation out of disease and poverty and debt... no concern for anything but money. "The markets" of any country, or the economies of nations and regions are nothing more than prospects. It's simple wheeling-dealing on a global scale. When I first read about the terror market I got physically ill. See, "the market" isn't about much more than momentum and feeling. "Value" orienated analysts will tell you we're still in a bubble- everything is overvalued, pretty much, and the only reason anything goes up or down anymore is momentum. Markets become self-serving, I would guess they neccessarily become self-serving, it's just part of the structure. The ol' "lifts all boats" meme is a load of horse crap- the only boats a market will lift are the boats already in the market. What the hell would we have on our hands when the terror market goes self-serving? Not existing for any other reason than to make money? Any statistical purpose would be worthless, #1, and #2 we'd have 800 pound finance gorillas making money off of death and destruction. Rather than just make money off the companies that produce the weapons of war they'd be able to benefit directly from war itself- conflict would be a money-making prospect. And it's not just the moral repugnancy of all of it, it's the danger, the very real danger that the market would start to drive itself. What the fuck kind of corrupt, hedonistic, orgy of sin would a terror-market bubble be? Good lord. The guys behind Iran-Contra ran "the enterprise" as a business, they intended to make money off their activities... it would take about 2 seconds for the entire terror market enterprise to be hijacked by purely monied interests. If there's any better sign of the civic and moral decline of the US than the terror market idea I'd hate to see it. Posted by: Tim at July 30, 2003 11:51 AM | PERMALINK"If there's any better sign of the civic and moral decline of the US than the terror market idea I'd hate to see it." I definitely think it's a sign that we've truly made a fetish out of the ideal of the free market. Posted by: David W. at July 30, 2003 12:06 PM | PERMALINKIt occurs to me that this could be used in a purely intelligence based format. Might work if we allowed lower level analysts, investigators, etc, from CIA, NSA, FBI and others participate for small amounts of money in this. After all, the main objection here is that markets are good at testing aggregate knowledge. Actual intelligence and law enforcement groups are likely to know very little individually, but this might help build up a good "big picture" of info coming into the intelligence community. Posted by: ElPresidente at July 30, 2003 12:11 PM | PERMALINKCouple points here. The projected 10,000 investors would have included a large number of brokerages and big investors, all of which have advantageous information -- consultants who are former Mossad agents, buddies in the Saudi Royal family, etc. Maybe not inside information, but good information about trends - e.g. Saudi security services are very nervous, etc. Any one of these investors is irrelevant, but the aggregate price of futures focused on the occurrence or non-occurrence of an event in the middle east (breakup of OPEC, independence of Palestine) would mean that the market's aggregate knowledge is superior to what a handful of intelligence assets can come up with. A rise or drop in the futures price would signal likely upheaval. Second, the terror for profit motive is unlikely, but if someone tried it, helpful in TWOT. The SEC has mastered reading stock trends, and when a party moves to make a killing (e.g. 'shorting' U.S. insurance stocks the day before the 9/11 attacks, it is immediately evident to the analysts. Not all statistical blips result in arrests, but the SEC takes a close look. Say a suspect moslem charity with $5million in the market takes it out a week before a big attack on a U.S. embassy in Yemen, a major oil producer - well, the investigators may have a place to look. Third, as for the morbidity of it all, it's nothing that insurance companies don't do, all the time. In some states it's against public policy and law to take out insurance on an unrelated third party, but in other states it's just fine. For that matter, futures contracts - which are in effect a form of insurance - are taken out all the time and encouraged, as well as regulated and taxed. Fourth and finally, such a market would have legitimate investment uses as a risk arbitrage device. Say an oil company was nervous about the prospect of a major upheaval in Saudi, or its Egyptian sources predicted a big attack on U.S. interests by Saudi citizens. This would cause problems for OPEC, lead to decreased U.S. demand for Saudi oil, and so forth. The price of oil would likely drop, and the oil company's profits would drop too. The company could purchase a futures contract either based on the ideas market - "if there is a terror attack and demand for Saudi oil drops, grantor pays"; or a futures contract to sell oil at the higher, present day price to a U.S. wholesaler, also in the event of an attack and a drop in demand. The contract wouldn't be some morbid incentive to terror; rather it would be an insurance policy that would maintain oil company profits regardless of which direction the market moves. At the same time, the price for such futures contracts would rise as the pool of would be buyers increased, and would-be-sellers decreased. This in turn would drive up the aggregate price of futures contracts in this market, which would then be interpreted as an indicator that something might be up. I don't know how well this would have worked in practice. The short selling just prior to 9/11 though, shows that either some actors in the market have good information about when terror attacks will occur; or the market itself is smarter, in the aggregate, than we give it credit for. Either rationale would justify attempting DARPA's proposed market. It sure beats cutting back on civil liberties, anyhow. Posted by: Omnibus Bill at July 30, 2003 12:11 PM | PERMALINKFor the sake of spreading more light than heat now, here's a link to a DARPA document which was released back in may that mentions the Futures Market Applied to Prediction (FutureMAP) http://www.darpa.mil/body/tia/TIA%20DI.pdf (Go to Page B-8 to get the skinny.) Posted by: David W. at July 30, 2003 12:18 PM | PERMALINKJohn Cole, I'm sure you realize the enormous difference between a terrorist attack, or even just the fear of a terrorist attack, and a Steelers win. Moving on... Posted by: PG at July 30, 2003 12:21 PM | PERMALINKMaybe I should call myself a defense contractor, sell my Dead Pool to the Pentagon, and retire with the proceeds as the Official Cabo San Lucas Blogger. Posted by: Laurence Simon at July 30, 2003 12:25 PM | PERMALINKHedging on the terror market doesn't sit any better with me than any other aspect of it. I'm not suggesting terror would be manipulated deliberately to cash in, I think that would actually be very unlikely. Rather, I don't think anyone could predict what would come of a terror futures market once it became a momentum market. I'm not talking about simple, buying "A" because you're going to cause "B", I mean I can't imagine what could come of it, and there's no reason to find out. There are other ways to test aggregate knowledge other than a for profit market; filling this market with millions and billions it was 10,000 investors initially, they would open it up after that) would do nothing except to entrench it, and institutionalize it in a way that eventually it would be an integral part of the economy. It simply could not exist without becoming nothing moe than another arena in which to move and make money- and it's an incredibly bad idea to move and make money directly related to violent conflict. Posted by: Tim at July 30, 2003 12:28 PM | PERMALINK"Since a free market is the best arbiter of everything, I think we should start using it for directing scientific research. Within a year we can have warp-drive and travel to the stars!" Beautiful! Sign me up. What's appalling (and appallingly stupid) about this idea is that it would give people a financial stake in more terrorist atrocities. Gee, too bad about the dead people, but this tidy sum will help put my kid through college. Posted by: Invisible Adjunct at July 30, 2003 12:35 PM | PERMALINKElPresidente writes: "Might work if we allowed lower level analysts, investigators, etc, from CIA, NSA, FBI and others participate for small amounts of money in this" I was thinking that, in such a system, a good currency might be minutes or hours of paid vacation time. Posted by: Jon H at July 30, 2003 12:41 PM | PERMALINKHere's an excerpt from the DARPA .pdf I gave a link to earlier: TECHNICAL APPROACH: FutureMAP's innovation is to use markets to
replace today's approach of discussion and consensus among experts. The
new approach is to set up, as it were, a "market" in two kinds of
futures contracts: One pays $1 if an attack takes place; the other pays
$1 if there is no attack. Market participants trade the issued
contracts freely. Prices and spreads signal probabilities and
confidence. Since markets provide incentives for good judgment and
self-selection, the market will effectively aggregate information among
knowledgeable participants. This approach has proven successful in
predictions concerning elections, monetary policy decisions, and movie
box office receipts; DARPA research is investigating its success in
Defense-related areas. The short selling just prior to 9/11 though, shows that either some actors in the market have good information about when terror attacks will occur; or the market itself is smarter, in the aggregate, than we give it credit for. Or, it could be a fluke. Certainly a fluke makes more sense than the market somehow being smart in ways we can't possibly understand. If nobody knew anything individually, then it doesn't make sense that somehow, in the aggregate, "the market" realized that something bad was going to happen with airplanes and insurance companies, and realized that short-selling those stocks made sense. So it seems to be me that the reasonable choices are a) there were people who knew or suspected an attack was coming, and used that knowledge to make a bit of cash on the market; or b) the short-selling (to whatever extent there was short-selling) was a random fluke. Posted by: John at July 30, 2003 12:59 PM | PERMALINKWell, when Nancy Reagen used psychics to predict the future, it lead that administration to sell arms to Iran, WHO WERE FIGHTING IRAQ, OUR CURRENT ENEMIES so it can't be too bad to accept info from non traditional sources. :) Posted by: JIM at July 30, 2003 01:25 PM | PERMALINKTim- The concept fueling this idea isn't ostensibly free-marketism. Scientists and statisticians have noticed a weird, inexplicable phoneomenon....ask any large, random sampling of unqualified people a specific question that few if any would have any reason to know anything about, average out the answers and what you get is almost always close to the truth. Ask 400,000 people how much grain was produced in New Zealand in 1943, average out the answers and the result will be a hairs-breadth away from the figure listed in the farmer's almanac. Nobody knows why this is. Posted by: theperegrine at July 30, 2003 01:27 PM | PERMALINKtheperegrine -- Then why don't insurance companies send out big surveys to predict where the next earthquake will be? It just seems there's a limit to what kind of information can be acquired this way. Posted by: denise at July 30, 2003 01:33 PM | PERMALINKdenise- Because nobody is sure if the phenomenon can be applied to predicting the future. That's what experiments like the terrorism futures market are trying to explore. Posted by: theperegrine at July 30, 2003 01:43 PM | PERMALINKtheperegrine writes: "Scientists and statisticians have noticed a weird, inexplicable phoneomenon....ask any large, random sampling of unqualified people a specific question that few if any would have any reason to know anything about, average out the answers and what you get is almost always close to the truth." But in this case, there's a party acutely interested in making sure the result is wrong, because if they do what the market thinks they'll do, they're more likely to fail. Markets may work for forcasting the weather, to a point, but that's because the atmosphere doesn't look at the market, see that the market predicts a heavy tornado season, and decide to ignore tornado alley and instead crank up some floods and catastrophic hailstorms. Posted by: Jon H at July 30, 2003 01:59 PM | PERMALINKJon H- You're preaching to the choir. I'm just saying that the fundemental principle at work, though it may be spooky and quasi-religious in nature, has nothing to do with market-worship. There are many reasons why this model could never work in the same way that surveying random guesses from disinterested parties would, most already discussed in this thread... Posted by: theperegrine at July 30, 2003 02:07 PM | PERMALINKWelcome to psychohistory 101, I'm your proffessor Dr. Hari Seldon. Now look at example 1 from the ancient era on Earth. Can anyone tell me the primary flaw in the market-based model for predicting terrorism? Yes, that's right Suzi, predicting the behaviour of individuals is as manifestly impossible as predicting the motions of particles in a thermodynamic system. In fact, given the variable of consciousness, it is harder to predict the actions of individual sapient beings than it is to predict the motion of non-sentient particles. That such market based predictive systems functioned fairly well at predicting the activities of masses of people (wether it be who their political choice would be or when printer sales would go up) gave these primitive psychohistorians the erroneous notion that they could also use such methods for predicting the activities of individuals and small groups. In retrospect it should have been obvious to these people that what they were attempting was doomed to failure. Aside from the moral aspects of gambling on life and death, the basic tenets of both quantum indeterminancy and chaos physics were well established at the time and their applicability to the situation fairly obvious. But ideology blinded them to these simple mathematical truths and only through the moral indignation of their foes was this misguided attempt discarded. Now, turn to page 23 and we'll discuss the mass consensus internet mind that did finally correctly predict the fall of the Bush dynasty in 2118 under the mind-boggilingly insane Jeb IV thus ushering in the true origin of psychohistory... Posted by: Hari Seldon at July 30, 2003 02:11 PM | PERMALINKThe biggest problem with this is that I simply can't understand how it would possibly be useful for anything. Could you please post a list of things that you don't understand so we can all immediately petition the government to stop all those activities. I, for one, do not understand cancer or AIDS research. Clearly we should stop funding that. Posted by: John Cole at July 30, 2003 02:34 PM | PERMALINKElPresidente writes: "Might work if we allowed lower level analysts, investigators, etc, from CIA, NSA, FBI and others participate for small amounts of money in this" It seems like these are exactly the guys who shouldn't participate: Day 1: Lower level analyst buys future in (for example) Brooklyn bridge being attacked, at $1. Day 2: Lower level analyst leaks report that he has information Brooklyn bridge might be attacked. Day 3: Future price rises to $100, lower level analyst sells. Posted by: hdf at July 30, 2003 02:44 PM | PERMALINK1. theperegrine - It would be nice to see a citation for that New Zealand grain production anecdote other than "The Shockwave Rider." It's not such a "weird, inexplicable phenomenon" if it's fictional. 2. To everyone who mentioned short-selling prior to 9/11... assuming that that did happen and wasn't a coincidence, it happened because the sellers thought they could make money that way. And if they thought the feds were using the stock market to predict and prevent terrorist attacks, or that the feds would show up at their doorstep to ask what they were up to, what incentive would they have had? In other words, any conceivable past benefit of such an intelligence strategy is negated once anyone knows about the strategy. Posted by: Eli Bishop at July 30, 2003 02:53 PM | PERMALINKHari Seldon posts: Aside from the moral aspects of gambling on life and death,... And Hari's the only poster so far to even MENTION the moral aspects. Thanks, Hari. I don't care if it "works" or not; this is a shameful idea that totally deserves to have been shot down. But Poindexter et al. would probably put up their own grandmothers as stakes if they thought they could make money on it. Posted by: Temperance at July 30, 2003 03:38 PM | PERMALINKJohn: "If nobody knew anything individually, then it doesn't make
sense that somehow, in the aggregate, "the market" realized that
something bad was going to happen." Kevin D writes: "It's not that people can game the system or that it will motivate people to actually go out and shoot someone, it's just a question of whether the government should encourage people to be happy about horrible events." Do you have life insurance? Would your wife (or other beneficiary) be happy about getting that fat check upon your death? Posted by: Al at July 30, 2003 05:36 PM | PERMALINKI, for one, do not understand cancer or AIDS research. Clearly we should stop funding that.-- You don't understand what cancer or AIDS research would be useful for? Damn. ;) Posted by: jesse at July 30, 2003 05:48 PM | PERMALINKOn the question of airline stocks prior to 9/11,we sem to have a winner. From Snopes: Claim: In the days just prior to the September 11 terrorist
attacks, the stocks of United and American Airlines were shorted by
parties unknown. ...In the month prior to the 11 September 2001 attacks on the World Trade Center and the Pentagon, highly unusual trading activity involving American and United Airlines stock was noted by market analysts who at the time had no idea what to make of it. Wildly unusual discrepancies in the put and call ratio — 25 to 100 times normal — were observed in stock options of the two airlines. In one case, Bloomberg's Trade Book electronic trading system identified option volume in UAL (parent of United Airlines) on 16 August 2001 that was 36 times higher than usual. ...But it was during the final few trading days (the market closes on weekends) that the most unusual variances in activity occurred. Bloomberg data show that on 6 September, the Thursday before that black Tuesday, put-option volume in UAL stock was nearly 100 times higher than normal — 2,000 versus 27 on the previous day. On 6 and 7 September, the Chicago Board Options Exchange handled 4,744 put options for United Airlines' stock, translating into 474,000 shares, compared with just 396 call options, or 39,600 shares. On a day that the put-to-call ratio should have been roughly 1:1 (no negative news stories about United had broken), it was instead 12:1. On 10 September, another uneventful news day, American Airlines' option volume was 4,516 puts and 748 calls, a ratio of 6:1 on yet another day when by rights these options should have been trading even. No other airline stocks were affected — only United and American were shorted in this fashion. Accelerated investments speculating a downturn in the value of Morgan Stanley and Merrill Lynch (two New York investment firms severely damaged by the World Trade Center attack) were also observed. The Chicago Board Options Exchange is investigating each of these trades and at this time is declining to offer comment on its progress. The volume traded and the one-sidedness of the trades, however, make it clear that those who had knowledge of the details of the attacks (which airlines would be involved and that the World Trade Center was a target) were behind them and stood to profit mightily from them. Posted by: Tom Maguire at July 30, 2003 07:25 PM | PERMALINKShouldn't we be spending taxpayer money on improving our domestic and foreign intelligence services rather than on things like this? I personally find it perverse but if private individuals wanted to do it, that's their business. It's when the government wants to do this that I really can't take it. Posted by: Al-Muhajabah at July 30, 2003 07:28 PM | PERMALINKI understand what it is for- I just don;t understand the science. Presumably, the previous uthor understands what this future markets is for, he just doesn't understand how it works- you get my point. Heh. Posted by: John Cole at July 30, 2003 07:40 PM | PERMALINKPicking up on Kevin's comment about "I'd also encourage everyone on this thread to be very careful about arguments suggesting the plan wouldn't work. It might, and it might not, and I'll bet that Robin Hanson and his colleagues are pretty knowledgable about everything that's been brought up here." If I am taking his point, I am in complete agreement. Look, I think we can all agree that a contract that pays off if Chirac of France is assassinated is a terrible idea, for a lot of obvious reasons. We can also agree, I expect, that driving a car at 80 mph in a residential neighborhood is a bad idea. However, I would not conclude that, since sometimes driving a car at high speed is foolish, it is always foolish to drive at high speed, or that it is always foolish to drive at all. Similarly, we agree that the "Wack Chirac" contract is stupid. I expect that the folks working on this may have had a similar lightbulb flicker over their heads. The real challenge would be to think of contracts that might yield useful info, and are not facially stupid. The NY Times suggested some: "The overview of the plan said the market would focus on the economic, civil and military futures of Egypt, Jordan, Iran, Iraq, Israel, Saudi Arabia, Syria and Turkey and the consequences of United States involvement with those nations." I do not know what that means. A futures on coup attempts flunks the PR test. How about a futures on Egyptian GDP? Is that a number worth attempting to predict? Would such a contract fly in people's faces as morally repugnant? It is a bit of a challenge. There is betting on US elections - how about a market on Turkish elections (but the PR could be bad)? My point, which I think was Kevin's point, is that it is easy to think of dumb contracts. Just because you can think of 100 stupid examples does not prove the idea is stupid, however. And the critics in Congress were basically throwing out crazy examples to generate LOL headlines and embarrass the program sponsor. This, for example, came from a critic: "For instance," Mr. Wyden said, "you may think early on that Prime Minister X is going to be assassinated. So you buy the futures contracts for 5 cents each... " And the NY Times intro was: "...It is an online futures trading market, **disclosed today by critics**, in which anonymous speculators would bet on forecasting terrorist attacks, assassinations and coups."
Following David W's suggestion (12:18), I checked the DARPA document describing this. It is on p. 68-69 of the .pdf file, and is definitely not the work of droolers, on this issue at least. http://www.darpa.mil/body/tia/TIA%20DI.pdf Posted by: Tom Maguire at July 30, 2003 08:16 PM | PERMALINKPeregrine- The point it it doesn't matter if it could work or not, eventually it would be self-serving and become a momentum market just like the stock market. Yeah, futures are different than stock, but no more insulated from greed and, good god, greed would be the engine of this enterprise. Why would someone enter the terror market? Because they got some hot info? Hell no. If they had hot info, and weren't part of the scheme, they could just call someone. They'd be in the market to make money. It would take about 10 minutes for it to change from aggregate statistical tool to financial tool. It doesn't matter if it would work or not- it would necessarily become corrupted. And to argue for it, just on the basis that it could work (not that you're doing that neccessarily), is so ridiculously sick... I mean, did we not just live through a huge stock bubble? Why on earth does this thing have to be profit driven? All those elections and box office receipts that get predicted, are they predicted in a market for chissakes? Of course it could work, but is there any reason whatsoever to believe that it wouldn't become a self-serving entity? Because the people involved would be doing it for the benefit of mankind? Yeah, just like the stock market. What we got here is an Adam Smith moment: the best way to order an aggregate terror knowledge network is by a free market- just as long as everyone involved is primarily focused on building that knowledge, and not simply making bank. It doesn't work that way in any other market, why would it work that way in a terror market? Posted by: Tim at July 31, 2003 12:58 PM | PERMALINKPeregrine, I don't get it - you just posted another link to a quote from The Shockwave Rider. The reason for my question is that I keep reading similar anecdotes about how a bunch of people with *no* knowledge of a fact will average out to a correct guess - which isn't at all the same as things like the Iowa political market or Oscar betting pools that are based on people paying attention to publicly available knowledge. In the absence of any citation, that sounds to me like an urban legend, one of those things like "we only use 10% of our brain." Come to think of it, that describes many of this administration's beliefs about the world. It doesn't really matter, though - I agree with all of the other criticisms of this insane idea, anyway. Posted by: Eli Bishop at July 31, 2003 01:16 PM | PERMALINKIf anything the "terror market" is the latest effort to further politicize intelligence. If this monstrosity had existed prior to the Iraq war, I have no doubt that the administration would have found ways to manipulate the "market" to support its justifications. The very hyping of the threat by the government would have moved the market toward its precooked conclusions. And since the government would have controlled membership of the pool of traders, who's to say they wouldn't have stacked its membership to those who matched its beliefs and would trade accordingly. The "market" would have provided one more faith-based data point completely decoupled from actual empirical intelligence. It's simply a device to hide intelligence manipulation since the CIA hasn't been nearly as compliant as administration hawks prefer. Posted by: fastback at July 31, 2003 05:47 PM | PERMALINKIt is wise to apply the oil of refined politeness to the mechanisms of friendship. Posted by: Figueroa Yolanda at January 20, 2004 03:26 PM | PERMALINKNothing's far when one wants to get there. Posted by: Spingarn Evan at May 2, 2004 12:06 PM | PERMALINKIt's not all lies - not all of it. That's the age-old dilemma. Posted by: StoweBerns Lindsey at May 3, 2004 12:18 AM | PERMALINKMisfortune shows those who are not really friends. online casinos | casino bonus | casino directory | high roller casinos | casinos Posted by: doi at May 23, 2004 01:55 PM | PERMALINKGratitude is the most exquisite form of courtesy. There is no great genius without some touch of madness. Posted by: Saposnik Andrea at June 30, 2004 05:47 AM | PERMALINK
Best XXX Sites - Is it true or not? Could the pill work for me? Get more information! |
|
|
Powered by Movable Type 2.63 
|
||||
