Politics

07 June 2008

Predicting Armageddon

Sadly, In light of this, this seems undervalued [emphasis added].

"If Iran continues its nuclear weapons programme, we will attack it," said [Deputy Prime Minister] Shaul Mofaz... "Other options are disappearing. The sanctions are not effective. There will be no alternative but to attack Iran in order to stop the Iranian nuclear programme."

Intrade on Iran airstrike by Dec '08

One thing notable about this graphic is how little reaction there has been to this news. To my reading (and I've followed this closely the last several years) it marks a significant step up in rhetoric by the top echelons of the Israeli government. The second thing I find remarkable is how the market has trended steadily upwards over the past four months or so.

This is but one of many 'quarterly' prediction markets Intrade has been running on an airstrike on Iran. All of them, obviously (at least so far) have rewarded the skeptics, however most have trended downward as the market close approached. This one isn't doing that.

I'm of two minds on how much stock to put in this market. (Sorry, pun intended).

On the one hand, military missions are, or at least ought to be, closely guarded secrets. And this would be no ordinary military mission. It is truly existential. Sure a few thousand people might be privvy to some kind of knowledge or direct insight into what was going on, but I doubt any of them would risk being shot in order to make a little extra on the side. So, on that basis, this has many of the characteristics of silly/failed markets in things like papal elections and supreme court nominations.

On the other hand, this is not as surprising as a terrorist attack (a dumb thing in which to make highly specific markets IMHO, even though there are many related indicators that might be useful).

Most nations don't want to go to war. (Iran may be an exception, but we're talking here about an attack on them, not by them). Most nations 'do the (diplomatic) dance'. They do relatively rote, predictable things in the lead-up to any conflict. And those things tend to follow patterns that are not random.

They may not be easily predictable, but there is information out there, held by people not constrained by an oath of secrecy, that is meaningful in discerning whether this will go down or not. Sadly, if I were to bet on it (and I don't -- at least not on real money markets) I would have to say that this market is way undervalued.


UPDATE (July 1, 2008):
Almost a month after I first made this post, the trend continues (see below). Despite Israeli moves that make such a strike seem more likely (whether for real or for bluff we can't know), the market hasn't had a very high 'beta' (volatility). This despite (or perhaps because of) significantly higher trading volumes in June. A good summary of recent events around this can be found here.

Airstrike on Iran by Dec31-08 as of 7-1-08
Overt Airstrike on Iran on or before Dec. 31, 2008

31 August 2007

What Can and Cannot Be Predicted (and Thoughts On Telling the Difference)

I just ran across this post ("Debating the Viability of Terrorism-Prediction Markets") over at the 'Footnoted' blog at the Chronicle of Higher Education discussing the risks of terrorism how best to assess them and whether they can be known at all. (If you've grown weary hearing about war, politics and terrorism, just skip to my non-terrorist business conclusions in the last paragraph.

In a recent interview [PDF] with the Federal Reserve Bank of Richmond, W. Kip Viscusi is asked about the public-policy response to the threat of terrorism and whether we are weighing the costs and benefits in a generally rational way. "The reason this is tricky is we don’t have very good numbers on what these risks are," Viscusi says. "The estimates of the probability of a terrorist attack or the number of people who are going to die in the coming year are all over the map. So if you can’t assess the likelihood of a terrorist attack or how deadly it is going to be, it is really hard to say how much you should spend to try to prevent it." [emphasis added]

Vscusi is characterized later in the article (and not without reason) as "one of the foremost academic experts on risk". His comment is in response to a persistent school of thought that claims, as Bruce Schneier does, that from an actuarial perspective, "terrorism doesn't happen".

Without the qualifier, the idea sounds terribly cold. With all reverence for the families of those who have lost loved ones in terrorist incidents however, it's not. Businesses and governments--not to mention non-profit institutions, individuals and families--all need to assess risks as rationally as possible and take measures to hedge them. Like it or not, the value of a human life can be defined--at least partially--in dollar terms. Anyone who's ever taken a course in economics or statistics (or balanced a checkbook for that matter) has figured out that infinite spending to protect against risk is infinitely foolish and that infinite spending on risk 'A' (and no spending at all on risks 'B', 'C', and 'D') is only a variant on the same wrong-headed assumption.

That's not what's really at issue. What is at issue is the degree to which the political leanings of some lead them to believe that we can know how much to spend combating terrorism and that the 'right' number is obviously much less than we are spending today (e.g., tactical domestic measures, overhead for business, strategic overseas measures, etc.) To which my response is: really? Show me your pre-9-11 white paper predicting the order-of-magnitude sea-change that occurred in that 'industry' on 9-11.

Bryan Caplan is one of Viscussi's critics. The CHE post notes:

"I am frankly puzzled," Caplan writes at EconLog. Citing the work of John Mueller,we have a long experience with terrorism, which has "shown it to be an extremely small problem in the broad scheme of things. How much longer does Viscusi want to wait before he'll conclude that the risk is very low?"

Unfortunately, long experience in and of itself is not sufficient for prediction, even at a macro level. Hold that thought for a quick diversion.

Here's where it gets weird. In criticizing Viscussi, Caplan, an econ prof at famously free-market GMU, ends up in league with Schneier, who as far as I can tell, tends towards the opposite end of the political spectrum. Both conclude, for entirely different reasons, that the future threat of terrorism can be known and condensed to a dollar figure and that rational budgets (both public and private) can be set accordingly. Oh that it were so.

Caplan, in particular "favors the establishment of a prediction market to help assess the likelihood of a terrorist attack". That's something I can conditionally applaud. If the results are used to "help assess", then we're fine. The possibility that a prediction market might help draw in and roll up marginal, highly distributed, even intuitive information that can supplement traditional (and sadly inadequate) intelligence-gathering mechanisms is certainly a good thing.  I've been a huge fan of prediction markets for years. They should be used for more things than they are today. To my delight, more and more are catching on.

But...

...as longtime readers know, I've also concluded that there are some (and arguably many) problems for which prediction markets are not only silly but grossly misleading. I don't have the space to review them all of them here. Thinking that they can precisely predict and quantify particular terrorist threats or even the threat of terrorism generally is a notion that falls into that category. Almost by definition, a successful terrorist venture is compartmentalized, secret and surprising.

In short, Caplan, Schneier and others appear to have an ideologically-induced blind spot that leads them to declare certainty where it does not exist. Let me explain.

Schneier and Caplan draw their essential argument from a backward-looking, actuary-style catalog of terrorist incidents. This many people died. This much property was lost. Productivity was reduced by this much for this long. Etc. Etc. It all sounds very rational. If we had reason to believe that terrorism were a natural, forecastable, perhaps even cyclical phenomenon, that approach would be absolutely correct. We don't.

The main problem, as I've noted before, is that:

The  unpredictability of terrorism renders any backwards-looking, purely quantitative, actuarial mode of analysis inappropriate and ineffective. That is, future deaths due to terrorism are something that neither Schneier nor anyone else can possibly predict with any degree of confidence.

Until 2001, the biggest single terrorist incident had caused around 300 deaths. Then in the space of a few hours, that number went up by an order of magnitude. There was no consensus (or even a significant plurality) of expert opinion predicting that that would happen - much less when, where and how. [emphasis in original]

And that's the problem.

Sudden, step-function, order-of-magnitude change, precipitated by a small group that has every reason to keep its plans secret is inherently unpredictable. It can only be imagined. If that sounds familiar outside of the terrorist context, it should. Businesses face this kind of challenge all the time; it is the very nature of business, in fact: snowboards vs. skis; digital photography vs. wet-process; PCs vs. mainframes; VoIP vs. legacy telephony; biotech vs. big pharma. The list is very long. It's harder to name an industry that hasn't been touched by this kind of change at some point (often precipitating the re-invention and re-definition of the former "industry") than it is to come up with a long list of examples of industries that have be altered in this way. Bottom line: it's important to differentiate between problems that lend themselves to forecasting and those that can only be dealt with via imaginative scenarios.

09 February 2007

China's "Authoritarian Resilience"

I note an interesting, seemingly oxymoronic concept in this piece: authoritarian resilience.

Beijing has pressed on, doing what Washington believed was impossible: compartmentalizing economic gain from political challenges. This does not mean that the market forces and various liberal instruments trumpeted by the United States should be dismissed or abandoned, but it does mean that as Beijing strengthens the resilience of its authoritarianism, Washington should cease basking in its delusions for inevitable democratic change. [emphasis added]

One key factor in the government's resilience, the author goes on to explain, is the supression of what are called "coordination goods", including:

...political rights, such as free speech and the right to organize and protest; general human rights, such as freedom from arbitrary arrest; and press freedom... the availability of coordination goods affects democratization because they drastically influence the ability of political opponents to coordinate and mobilize but have little impact on the continued economic growth that is crucial for sustaining an authoritarian regime’s legitimacy. [emphasis added]

It's a strange combination, to be sure, and in fact 'robustness' rather than resilience may be a better term for the Chinese government. The former connotes sheer strength and durability; the latter is more characteristic of a system that's able to bounce back seamlessly (or nearly so) from a wide array of unanticipated shocks and challenges. It's a distinction many large organizations should take to heart--and many have.

One could get balled up in a long semantic debate about the two 'r' words, but the point is this: Resilience tends to be associated with exactly the kind of nimble, distributed coordination the Chinese government is working to suppress in politics and promote in economics. (The limits of central planning--in any arena--having been amply proven by historical experiment.)

Durable, top-down organizations may be confused with resilient ones for a time simply because they're strong--and unafraid to wield their accumulated power. Longer-term, I would assert however (and contrary comments are welcome!) that the true resilience of highly distributed systems tends to triumph due to: 1) greater adaptability (they can deform in extreme ways without disintegrating altogether) and 2) the speed with which they can route around 'failure' (to borrow a familiar axiom about the architecture of the Internet).

Exactly how long that battle between resilience and robustness will take to play out in the case of China is anyone's guess. (Long term can mean very long indeed: the co-option of Rome by Christianity took several centuries--easy to see in hindsight but much harder to predict in advance.) The ray of light here may be the philosophical impossibility--at the margins anyway--of differentiating between coordination that is essential to economic growth and that which fuels political/religious change. The Chinese government faces the unenviable and ultimately fruitless task of teasing out the content and intent of each and every expression and drawing a clear line between these natural and overlapping forms of social cooperation.

So long as the tools for coordination in general are the same in business as in politics (laptops, cell phones, networks, fax machines, conversations between individuals, etc.), some political 'stuff' will sneak through--never having been formally identified as such. And so long as the tools are inherently empowering of the individual, their force in changing obedient, collectivist minds into questioning, impatient ones will be inexorable. At least we can hope. Timing is everything.

30 January 2007

Those Pesky Planners

An oldie-but-goodie that explains in pictures why I detest the term "planning" (prededed by any number of modifiers, e.g., 'strategic', 'scenario', 'long-term', etc.) in favor of the term "thinking".

The former implies rigidity and omniscience. The latter implies flexibility and humility.

H/T: Say Anything

05 January 2007

China's Long-Term Strategy

Short post under a big headline. My attention has been turned East after the Taiwan earthquake and submarine cable cuts. Two good references with perspective on Beijing's recently released Defense White Paper 2006: W$J editorial (subs only) and this morning's China Daily.

Official party line here ("purely defensive in nature")--a reminder that some kinds of transparency look good only because we've become accustomed to having none at all. A report out of the Pentagon that was this thin on data would be greeted by howls of protest.

Most interesting (though hardly new) factoid with a resiliency angle: "China receives about 63% of its oil through the Straits of Malacca -- which are protected by the U.S."

Surprisingly, nothing from Barnett on the paper (yet)--though plenty of other good stuff to chew on at his blog (as always), including the destabilizing effects of religion. (He means that in a positive sense.)
 

29 December 2006

Politics and Networks: Taiwan vs. the PRC

This piece light-heartedly captures a not-so-lighthearted subject: the massive political, military and economic implications of what the Taiwan earthquake has revealed [emphasis added]:

...China’s access to the Internet at large runs through Taiwan… man was that some s***ty planning - and a helluva chip the ROC can play should things get nasty between the two siblings.

PRC: “Um, we’d really like you to come back and be a part of your motherland.”
ROC: “But we believe in a multi-party system...”
PRC: “Fine, have it your way. We’re going to attack you tomorrow.”
ROC: “If you do, you’ll only be able to access Mainland-produced porn.”
PRC: “Please disregard, have a nice day.”

H/T: China Law Blog

04 December 2006

Whither China?

Wide-ranging, informative article on Mao's legacy, speculating on a range of futures for China.

Every society changes from one day to the next. But the economic and social transformation in China, especially since the beginning of the reform era in December 1978, has been particularly startling. Mao regimented the Chinese people, oppressed them, clothed them in totalitarian garb, and denied them their individuality. Today, they may not be free, but they are assertive, dynamic, and sassy. A mall-shopping, Internet-connected, trend-crazy people, they are remaking their country at breakneck speed. Deprived for decades, they do not only want more, they want everything.

Change of this sort is inherently destabilizing, especially in a one-party state. Social unrest, writes Samuel Huntington, becomes especially dangerous when political institutions fail to keep up with the forces unleashed by economic change. That is the dilemma of the Chinese Communist party, which, even as it has sponsored uninterrupted economic progress, has itself changed remarkably little from Mao’s days, and still stands in the way of meaningful political reformation.

As Tocqueville observed, “steadily increasing prosperity” does not tranquilize citizens; on the contrary, it promotes “a spirit of unrest.” [emphasis added]

... It would be difficult to underestimate the role played by wireless communications and the Internet in this phenomenon. Societies change—or reach a “tipping point,” to use the contemporary term—when enough people begin to think simultaneously in a new way. These days, Chinese thoughts and emotions travel through optical fiber at the speed of light—there are 123 million “netizens” in China, and 34 million of them are bloggers—and the Chinese are holding nationwide conversations for the first time in their history. Ideas—like, for instance, the idea of representative government—start out small and spread rapidly via countless chatrooms and online forums...

...so far, China has been more successful than any other country in regulating its Internet community; but this is a battle in which it will never be able to claim final victory... Today, the regime survives because no single cause has united the Chinese people and impelled them to march en masse to Tiananmen Square.

09 March 2006

The International Intelligence Summit - Rehearsing Divergent Futures

Life has interfered with blogging these past few weeks. (My family and colleagues might tell you that's good.) With this post, I'll report on a talk I was grateful to have the opportunity to give  last month at the International Intelligence Summit (IIS) in Crystal City, Virginia (just across the river from Washington, DC) - less on the summit itself. Apologies for the delay.

The IIS is an interesting affair - non-profit, non-partisan, truly international (thus the name) and (ironically given the subject matter) very much open to the press and pre-registered public. This was my first time attending. It was also my first time at a professional gathering that included bomb-sniffing dogs and large men in flak jackets with earpieces carrying sub-machine guns in the lobby. A client who'd helped organize part of the proceedings felt - upon seeing and participating in one of our scenario workshops - that people (and a few large organizations) he knew would benefit from hearing more about the methodology. He was right. More on that in a moment.

As many readers may already know, the IIS was made even more 'interesting' this year by the release on day one of material out of Saddam's archive of tapes that had been freshly translated from the Arabic by one Bill Tierney - a U.S. army vet and former UN weapons inspector.

I'll refrain from repeating a story that's been well covered already, or opining on the veracity of the material. I will observe however, that it represents just a fraction (estimates I've seen range from two to four percent) of what will in time be known about a regime, a place and a period in history that some think we understand well already. We don't now and that's frightening.

It is exactly this kind of gradual and highly distributed accumulation of evidence and insight, on such complex and open-ended  questions as the pre- (and even post-) war state of WMD in Iraq that lends itself to processing by distributed collective intelligence 'engines' - blogs, prediction markets and wikis to name just a few. Centralized, highly 'siloed' legacy approaches may work in the short term but they're ill suited to the sheer volume of material that eventually comes out. Or to put it another way: More eyes and more minds make for more understanding.

At peak, the conference drew somewhere in the vicinity of four hundred people. (I did a quick count of the heads at the keynote.) It included a number parallel 'tracks', all of which were fascinating. I attended talks on Iran, North Korea, China, Syria and others that seemed desperately obscure out of context. High-end falconry as a money-laundering mechanism for Middle Eastern sheiks. Threats to maritime assets and transport grids as a result of flotillas of small fast suicide boats e.g., in the Persian Gulf. Fascinating if frequently scary stuff.

One comment I heard repeatedly was to the effect that "each of these sessions is highly educational but I can't easily knit it all together". Indeed. Thus my talk on scenarios.

I had the slight misfortune of being positioned head-to-head on the schedule with author and analyst Richard Miniter next door - a talk I'd like to have seen myself. Nonetheless, enough seats were filled at my talk that it made sense to conduct it formally. Ultimately we ran out of time to take every question at the end.

The focus and title of the talk: 'Rehearsing the Future' - an apparent but highly deliberate oxymoron. How does one rehearse what one doesn't know? To boardroom ears that seems nonsensical. Management teams plan for the future. They speculate about the future. They project the future. They budget for the future. They seldom if ever rehearse the future - not in the way that other teams do anyway - not to the point of being able to recognize emerging patterns as a team and respond to them instinctively. The heavy burden of coordination and the inevitably flawed, hasty thinking that occurs in the heat of 'battle' impose a heavy cost on those who have not rehearsed a range of future 'plays'. In other words, scenarios.

My main point in the talk: Other kinds of organizations practice all the time. Sports teams. Pit crews. Orchestras. Armies. Police units. Most teams spend the vast majority of their time rehearsing. Not management teams.

Interesting factoid. I spoke with a colleague yesterday by phone. Until recently, he'd headed up strategic planning for a major U.S. city orchestra. The standard in that industry? 75% rehearsal time: eighteen hours out of a standard musical work week of twenty four - the rest being the performance. And that doesn't count the months and years of extracurricular  rehearsal time that individuals must engage in. Seventy five percent. What's the corresponding figure for a corporate management team? I doubt it's as much as seven percent in the most tight-knit strategically oriented organization.

The other major point I made - that seemed to get heads nodding - was that there are two types of uncertainty: actuarial and creative. Confusing the two (and applying the wrong planning system to deal it) can lead to ruin.

The first type of uncertainty - actuarial - involves events for which there is some known history that's highly likely to continue: the growth path of a stable industry, the frequency and severity of earthquakes or tsunami, the incidence of death from certain classes of disease or accident. Such things can be planned for, or at least insured-against. The risks are well understood. Probability dictates that such events come in clusters more often than not (leading to panic that present trends might continue). But they're inevitably punctuated by long periods of calm - periods that lull people into complacency. (Sure, lets build more condos on the beach at Cape Hatteras or below sea level in New Orleans!)

The second type of uncertainty - creative - is more difficult for organizations to wrap their arms around and impossible to insure against in the classic, monetary sense. For example, what was the 'probability' circa 1970 that snowboards (and the as-yet nonexistent entrepreneurial providers who would invent them) would cut into the ski industry?  The term 'probability' doesn't even make sense in that context. One could only imagine an innovation that might challenge the traditional franchise.

Or the 'probability' circa 1980 that digital photography would into silver halide (wet process) film sales? Again, nonsensical to ask the question in that way - at least at that time... when there was still time to do plenty about it... when many strategic options were still open to the established players. There was no meaningful 'probability' before the advent of computers and powerful memory chips. Only a creative scenario process could have helped Kodak to rehearse a future in which ribs might start showing on their once fat cash cow.

Or imagining circa 1990 that small committed bands of people might commandeer large airplanes and fly them into tall buildings on purpose, taking out much of the communications grid for the financial industry (among other nasty, evil things). As we noted sixteen months ago: "the four-day shutdown of U.S. exchanges after 9-11 raised pointed questions for the industry in New York".

It is only with hindsight that some are claiming we could have predicted and prevented the events of that tragic day. Yet there were signs, including this insightful documentary on PBS' American Experience chronicling the simultaneous hijacking of four large airliners by Middle Eastern terrorists in early September... of 1970. That was not enough to bet on at the time, but enough to warrant imagining: what would we do if...? ...the essence of scenarios... the essence of rehearsal for a future that sometimes stubbornly refuses to be predicted long enough in advance to do anything useful about it.

26 November 2005

State Resilience

Still catching up from several weeks off-grid, we note this excellent and richly linked thought-piece by Zenpundit from earlier this month. ZP (welcome to the blogroll!) riffs off Tom Barnett (among others) on the topic of state resilience (the political not the chemical kind), listing eight prerequisite links as recommended reading. Order up a double espresso and clear your schedule for an hour. You won't be disappointed. Excerpt:

State Resilience, as the term implies is a state having the quality of adapting and continuing to function despite severe trauma or losses. When under attack, Non-resilient States lash out stupidly, retreat or collapse. Resilient States adjust and hit back from an unexpected direction. The term indicates a fusion of political will with executive competence and material means.

Nation-states are at root simply very large, very complex, networks with the capacity to determine the rule-sets that govern the behavior of all the smaller, internal, subnetworks they contain or the external networks with which they come in to contact. The greater the legitimacy of the state, the less frequently it need employ physical force to assure compliance.

Broadband connectivity style State-building is a positive endeavor, a useful prophylactic in weak States before trouble begins and a vital support where resilient states are effectively combatting [sic] 4GW attackers. Resiliency however is critical to state survival - it is the foundation that will support the range of State-building programs and will be reinforced by them.

08 July 2005

Prediction Markets on Judges

Professor Bainbridge notes this reference to a recent spike in the Alberto Gonzales confirmation contract at Tradesports. If politics is what it takes to build popular awareness of the efficacy of prediction markets, I'm fine with that. It's important to recognize however - as I've said before - that a contract on the actions of a small group is inherently more risky and less liable to be correct than one for a general election in which actions and actors are widely distributed. It's also worth noting that Congressional swing votes are an even smaller group than the Council of Cardinals that elected Pope Benedict. Having thrown that bucket of cold water on prediction markets for the purpose of predicting judicial nominations, I heartily agree with Bainbridge's bottom line: that such markets are a heck of a lot better than the media talking heads.

My Photo

July 2008

Sun Mon Tue Wed Thu Fri Sat
    1 2 3 4 5
6 7 8 9 10 11 12
13 14 15 16 17 18 19
20 21 22 23 24 25 26
27 28 29 30 31    

Sites, People, Blogs