Beauty of Gray

Friday, October 18, 2002

Embracing Defeat

That’s the name of an excellent book on the American occupation of Japan after World War II. Given the talk of a potential American occupation of Iraq, and the many analogies that are drawn to our experiences in Japan and Germany, I think the book is very timely. It’s good enough to be of interest at any time (It won the Pulitzer Prize last year, although that’s not always a guarantee of top quality), but is particularly relevant now, and I’d highly recommend it to everyone.

I bought it and read it after the September 11 attacks, anticipating an American occupation of Afghanistan, although the more I read about both countries, the less relevant I though the book was. And as events have turned out, the US has had a relatively hands-off approach in the rebuilding of Afghanistan. But Iraq looks to be different, and the book can give some idea of the problems that are likely to face any occupation, and the magnitude of the difficulties it would face.

Issues like the unavoidable paradox of trying to impose democracy through military force and fiat, for example. “You’re going to think for and govern yourself! And you better think for yourself the way I tell you to, or I’m going to censor your newspaper!” These inevitable tensions, as well as culture shock between occupying forces and the local population. The interplay between resentment, admiration, and envy. There are a lot of general features that would be present in any occupation that are here described in the particular case of postwar Japan.

There are certainly many significant differences between the case of Iraq and that of Japan, among them the existence of the emperor and a civil service that the US could take over and govern through, which made the US job both easier and less fruitful. But read the book and judge for yourself.

A word of caution—the book is not a narrative history. Rather, it is a sociological analysis. It’s organized by topic, rather than following the thread of events through time. I don’t think the book could be done any other way, and you don’t need to know the period in great detail to follow the book, but that could be off-putting for some readers. But if you’re curious about the possibilities of a US occupation of Iraq: what it might look like, the problems it might face, the success it might have, Embracing Defeat by John Dover is an excellent place to start.

Link posted by Doug Turnbull at 2:00 PM

North Korean Nukes

There has been a lot of speculation about the recent announcement of North Korea’s ongoing nuclear program, but I think a lot of it is missing the point. Instapundit links to this piece by tacitus, which considers the possibility of military action. Instapundit's post, on the other hand, is similar a posts by Vodkapundit and a couple of others that present this announcement as a victory for Bush and his tough foreign policy. They ask why North Korea announced this now, and answer that North Korea has become scared by the Bush doctrine and the imminent invasion of Iraq, and don’t want to end up suffering the same fate.

I think these interpretations are badly flawed, for a couple of reasons. First, North Korea didn’t, out of the blue, make a public announcement that they had a nuclear program. They didn’t really come clean, of their own accord. What happened was that the American negotiator presented clear evidence that they were pursuing a Uranium enrichment program. At first they denied it, and then admitted privately to the US diplomats that, yes, you are right. The recent public announcement was not by the North Koreans, it was by the Bush Administration, revealing what the North Koreans told us in private.

So it wasn’t a confession. It was more an admission after getting caught in the act. North Korea had been pursuing this program in secret for several years, presumably trying to develop a weapon before the US became aware of it. But they also knew there was a chance the US would discover their efforts. And once that happened, there was no point denying it, so they didn’t.

Second, there is no chance the US will invade North Korea. They may deserve their spot on the axis of evil, rather than just being put there for affirmative action reasons, and it may be the case that we’d like to invade if we could. But we can’t, or at least doing so would require paying an unacceptable cost, so those saying “first we squash Iraq, then it’s North Korea's turn” need a reality check.

Why won’t we attack? Well, there are several very important differences between North Korea and Afghanistan or Iraq. The most important is the existence of South Korea, sitting right there over the border from the North. North Korea essentially holds millions of South Koreans hostage. If we start bombing them, they’ll start shelling Seoul, or launching missiles at Japan. Afghanistan was completely impotent to strike at US friendly targets, and Iraq has a minimal capability to do so. North Korea can easily inflict thousands, tens or even hundreds of thousands of casualties on South Korea. They can do more than just sit around absorbing punishment, waiting for some US soldiers to attack—they can counter-attack the South.

Second, they have a powerful big brother in China. They are a Chinese client state and, as messed up as the country is, it’s still a real chit in international relations for the Chinese. They like having this crazy little brother that only they can control; it gives them power in the region. They would be a lot more opposed to an invasion of North Korea than anyone is to an invasion of Iraq or Afghanistan. And they’re an important enough country that we need to take their desires under consideration, especially in their sphere of influence. And you always tread softly when it comes to conflicts with nuclear states.

Finally, the geography and fortifications of the Korean peninsula make it just about the worst place in the world to attempt to fight an offensive war. First, the country is very mountainous, which channelizes troop advancements along a few north-south roads. We can attack Iraq across an entire desert, from any of 4 directions. We can attack North Korea from 1 direction, along only a few roads. For Civil War buffs, think Burnside Bridge at Antietam.

But it’s even worse than that. Since the truce ending the Korean War, both sides have been spending the last 50 years fortifying their country. Artillery bunkers built into the sides of hills and caves, booby trapped roads and bridges, pre-targeted rockets and artillery, supply depots, marshalling points, minefields, airfields, machine gun nests and bunkers, etc. Everything that man can do to make a position difficult to attack has been done by the two Koreas.

And all this isn’t even considering the fact that North Korea may have already developed weapons of mass destruction, and unlike Iraq they have no shortage of targets to use them on if they feel threatened. So while they may talk about the military option, rest assured it isn’t going to happen, and wouldn't happen even if we weren't weren't focused on Iraq. Instead, diplomatic and economic approaches will be tried. They may or may not work, but you have to play the hand you’ve been dealt, and our hand in the North Korean game just isn’t that strong.

Link posted by Doug Turnbull at 8:41 AM

Final shot, in the other direction

I've received several e-mail critiques of my posts on ballistic fingerprinting, which together have pretty much changed my mind. I now no longer think that the program would be worth the cost and effort it would involve. What has changed my mind?

First, Rob Lyman and another reader, Andy, clarified that the amount of scratching needed to alter a guns ballistic fingerprint beyong recognition is so little that the gun could be altered with no degradation to its performance. Andy writes:

Gun barrel dimensions are accurate to at best 0.001 inches and most
are far less precise. The markings are caused by
flaws that are 20-100x smaller.

So, one could easily remove 2-5x as much material as would be required to change
the markings without even taking a very good gun
out of "as good as new" condition.

That's a key point, and since the features are that much smaller, it wouldn't require any serious machining to alter the marking. (The objection that gun hobbyists can build a gun or re-bore their barrels is unconvincing to me, since it seems like most criminals would have neither the equipment nor the inclination to do much re-machining of their guns.) But you really could do it with a file or some sandpaper--no lathes or drill presses required. Lyman also mentions that the fingerprint left on the brass is relatively easy to alter too, as well as the important fact that two guns manufactured close together in time at the same factory will have very similar "fingerprints." They diverge when used over time.

This points up the problem of false positives, which would grow more severe with a larger database. The possibility of coincidence between bullets found at the crime scene and the gun of a suspect captured via other means is relativeley small. But if you just have the ballistic evidence, together with a very large database to look at, the false positive problem goes way up. So the ballistic fingerprints are less like DNA or fingerprint evidence and more like chips of car paint found at the scene. So it's most useful when used in just the way it is now--IDing the guns used in crimes after the fact, and linking different crimes together

This divergence over time is another potential objection to the system which has been raised. I'm still not sure how serious of a problem this is, but given the small feature sizes on the barrel, they could very easily alter on a time scale of a few years of use rather than a few decades.

And speaking of cars, both Stephen Commiskey and a fourth reader, David, point out that you actually don't have to register cars to own one, you just have to register them if you want to drive them on public roads. So the analogy I made was incorrect. When I'm wrong, apparently, I'm really wrong. David also seconded (thirded?) the points about the relative ease of modifying a gun's fingerprints, also pointing out that a national ID program would be a strong incentive for such modifications, while the spotty system in place now makes it less of an issue.

The final objection is the old standby that much of the crime in the country is committed with stolen or smuggled guns. But that, to me, is an argument to crack down more on gun smuggling and theft, rather than to throw up your hands and give up worrying about any guns at all, legal or illegal. Just as the argument that a suitcase bomb is possible does not, in and of itself, make a ballistic missle defense program a bad idea. If your windows are open to let a thief in, the correct response is not to leave the front door unlocked as well, but to lock the front door and close and lock the window as well.

On a final note, Vodkapundit linked to my original post and there are several comments there also criticizing the idea of ballistic fingerprinting, if you want more.

Link posted by Doug Turnbull at 7:36 AM

Thursday, October 17, 2002
Kicking them while they’re down

Since September 11th, al Qaeda and other terror organizations have almost certainly been trying their hardest to carry off additional attacks against the US, but have been unsuccessful. Johnny Walker reported in interrogations than there were two successive waves of attacks planned to follow on the September 11 assault, but the vigorous US counterattack, combined with a stronger and more attentive US defense, have prevented any more major attacks, at least until now.

But, if the recent attacks, along with the synagogue bombing in North Africa, can be extrapolated to a trend, al Qaeda may, at least in part, be adjusting their strategy to attack lower value but less well defended targets. And unfortunately, the defenses are lowest in those countries that are poor, and so can’t afford the police and intelligence forces to prevent them. And not only are these countries the least well defended, they are also those that can least afford the cost of a terrorist attack.

It’s likely that the attack in Bali will eviscerate the tourist industry there and cause severe economic damage to Bali and to Indonesia in general. And if this strategy become more widespread, it will end up inflicting the worst damage on precisely those who are already suffering the most—those in the poor and developing world. There are Western interests there, which can be attacked away from the defenses of the continental US, but the real damage that will end up being done is to those host countries themselves.

Australia can survive with little adjustment without Bali; unfortunately the converse is not true—Bali cannot survive easily without Australians. As with AIDS in Africa, those hardest hit by calamity are, unfortunately, often those who were already the most miserable.

Update: Two bombs have exploded in the Phillipines, killing 6 and wounding 144.

Link posted by Doug Turnbull at 2:08 PM

Second Volley

As I said, guns aren’t really a major issue for me, but Instapundit has another post at least partially responding to my last one, which deserves a reply. In addition to seconding the concerns about the usefulness of ballistic fingerprinting, he thinks that I was too dismissive of the fears of confiscation.

Perhaps--I was certainly too glib. But on the whole, I'm unconvinced. While it's true that registration would be a logical precursor to confiscation, I don't see that it's a necessary or even likely outcome. To me, registration is no more likely (and in fact, less likely) to lead to the confiscation of private guns than a law against partial birth abortions is to lead to the outlawing of all abortions. (Although I’m curious about Instapundit’s quote that “In fact, gun registration has consistently led to confiscation.” In the US, or in other countries? If it’s other countries, then it seems like the US Constitution makes those a very imperfect parallel.)

But in both the gun and the abortion cases, I think the slippery slope argument is used too easily, attempting to tar moderate measures with the taint of extreme actions. (A necessary line of argument for advocacy groups, though, since a majority in both cases support some limitations, while opposing outright bans.) Slippery slope arguments aren't always wrong, but taken to the extreme they render almost any action at all unacceptable, since most things taken to the extreme become evils. So they can be used to attack almost any position, which is one of the reasons I'm highly skeptical of them.

In this case, I don't think the slippery slope is all that steep, or that slippery. The political climate is not such that widespread confiscation could even be conceivable, and if it changed enough so that guns could be outlawed (ignoring the Constitution, for now), Congress would have no problem passing a registration law long before it got to that point.

To put it another way, I don’t think the passing of these restrictions now would significantly increase the chance of future, more severe restrictions being passed later. Nor is the current situation a clear and logical “breakpoint,” where adding a restriction becomes a change of kind rather than of degree. With neither of these two conditions met, I don’t think the slippery slope argument really holds water, so to speak.

Link posted by Doug Turnbull at 1:51 PM

Wednesday, October 16, 2002
Returning target

Well, I am back from a wonderful honeymoon in Italy, and while I didn’t want to leave, it wasn’t because I expected to come home to a shooting range. But that’s the situation here in Northern Virginia. I don’t have much to add to the excellent coverage that other bloggers like Jim Henley have had of the ongoing case. At this point, my speculation is worth just as much as anyone else’s, which is to say not too much.

However, I’m a bit surprised that the related issue of ballistic fingerprinting has gotten so little attention in the blogosphere (although admittedly I’ve been trying to catch up and very well might have missed some discussions on it.) The only reference I saw was one or two dismissive comments on Instapundit.

Unfortunately, there seems to be remarkably little on the issue out there. The BATF which runs a national database to assist police forces has a pretty web page, but there’s little of substance there. A perfunctory search on the NRA webpage turned up no info on the technology at all. So, at the risk of venturing into an area where there are many others more knowledgeable than I, let me fire this first salvo out and see if the experts will come out of the webwork to clear things up for me.

Ballistic fingerprinting is a way to match up a bullet fired from a gun and/or the spent casing with the gun that fired it, based on distinctive marks that the gun makes on the bullet and casing. Pretty simple, in theory. This is the technology that has allowed the police in the DC area to definitively link all the sniper shootings together, as they all used the same gun. The idea behind a national ballistic registry is that all new guns should have a test bullet fired and the characteristics measured, so that, if the gun is later used in a crime, that gun can be identified and then tracked. Seems simple enough to me, and a common-sense proposal. Yet the President and, presumably, most gun-rights supporters oppose the idea. Why?

Ari Fleischer’s Oct. 15 briefing gives two main reasons, if you can pull them out from the evasions and irrelevant babbling that are any press secretary’s stock in trade:

Q There's a policy issue that emerges from the sniper attacks, and that's some gun control advocates are calling for a national ballistic fingerprinting system where every gun, before it was sold, would be test-fired, the ballistic fingerprint would be entered into a database, and law enforcement like Montgomery County and Northern Virginia and federal law enforcement could call on that. Does the President support that?

MR. FLEISCHER: There are a variety of technical issues involving the reliability and the accuracy of that program that bear looking into, and those issues will be explored. That would also, of course, involve an act of Congress and a determination of the will of Congress to make that happen.
But there are a series of steps that the President has taken that he believes can be very helpful and should be helpful on the federal level, principally involving giving local law enforcement communities and the prosecutors enhanced resources and more prosecutors so they can more quickly bring people who commit crimes with guns to account, hold them accountable, bring them to justice and try them before a court of law.

Q But on this issue, on ballistic evidence, the President has doubts about its reliability and accuracy?

MR. FLEISCHER: Well, there are experts who have questions that have been raised about its accuracy and reliability. And those questions are rather -- or those issues are rather straightforward. And that involves, over time, whether or not this tracing technology remains accurate. The more a gun is used, the less accurate the tracing can become. The ability of somebody who obviously is in the business of committing crimes, and therefore wants to figure out ways to protect his ability to commit a crime without being caught, to alter the barrel of a gun -- such things as a simple nail file put down the barrel of a gun can alter the amount of tracing that's on a bullet, and therefore change the accuracy of the fingerprinting, very unlike any fingerprinting of human beings. A nail file cannot alter the fingerprint of a human. A nail file can alter the fingerprinting of a weapon.

Q These are arguments generally raised by defense lawyers. Prosecutors rely on this evidence. And I went to the ATF website today, after you mentioned these concerns the President has about reliability and accuracy, and the AFT, on its website for the National Integrated Ballistic Information Network, where guns that have been used in crimes are finger-printed, says, as each fingerprint is different, a firearm leaves unique, identifiable characteristics on expelled ammunition. Is the ATF wrong?

MR. FLEISCHER: If you keep reading on the same webpage, I think you left something out. That same webpage continues to say, though no investigative tool is perfect or will be effective in every situation, the availability of an open case file of many thousands of exhibits searchable in minutes instead of lifetimes that would be required for an entirely manual search provides invaluable information. And that's what it does, it provides information that is crime specific.

Q About guns that have been used in crimes. So guns that commit crimes are in this database, but the President doesn't want all guns in that database.

MR. FLEISCHER: Well, again, that gets into the same issue as I got into with fingerprints. There is an issue about fingerprints, of course, as a very effective way to catch people who engage in robbery or theft. Is that to say that every citizen in the United States should be finger-printed in order to catch robbers and thieves. And these same issues are raised here. The President does believe in the right of law-abiding citizens to own weapons.

Q Fair enough, so it's about liberty and privacy.

MR. FLEISCHER: Well, it's also about the other issues I raised, in terms of accuracy and reliability. These are all various factors of why this is not a simple solution or a simple matter. And certainly, in the case of the sniper, the real issue is values. And that's what is at stake here. The real issue is values. These are the acts of a depraved killer, who has broken and will continue to break laws. And so the question is not new laws; the question is the actions here represent the values in our society.

Basically, the President’s objection boils down to the concerns that (a)—the technology is unreliable, that you can modify a gun barrel; (b)—that doing so is an invasion of privacy. Instapundit referenced a similar argument to (b), absent the mind-numbingly idiotic “fingerprinting everyone in the US” analogy, deriding the ballistic fingerprinting idea as a backdoor way to national gun registry. Well, that isn’t really an argument for me, since I’m all for a front door gun registry, and in fact am appalled that such a system doesn’t already exist. The privacy concerns here seem to me to be minimal. Apart from the paranoid fantasies of imminent tyranny starting with national gun seizures, a national gun registry will be no more of an invasion of privacy than registering your car, which is already required.

The first objection is potentially more serious, but as I said, I was unable to find any substantive discussion of the issue. The claim is that filing down your gun barrel will change its fingerprint. OK, but will it make it unrecognizable? (A similar question applies to the gradual alteration of the ballistic fingerprint over time.) My guess is that messing with the inside of the barrel would be more like writing on top of existing text—you’d need to do a lot of scratching before the initial pattern was no longer recognizable. And will doing this much damage to the inside of the gun barrel degrade the weapon’s performance? If it makes the weapon less useful, then at least there’s some net gain to it.

This objection also assumes that criminals will be smart enough to do this. They aren’t smart enough to do it now, with guns they’ve used in a crime, so what makes people think they will do so in the future? Criminals can also wear gloves to prevent them from leaving fingerprints, but they don't all do that either.

It seems to me that the only downside to the proposal would be its cost—is the cost of fingerprinting each gun worth the benefit it would have in crimefighting? That I’m not sure of. But the BATF already has a database system set up, so there’s little start-up cost. And the marginal cost for checking each gun shouldn’t be more than a few dollars, which could either be passed on to the purchaser as an indirect gun tax, or directly funded out of the existing FBI or other budget.

Update: I should have thought to check Volokh for information on a gun-related issue. He links (via Clayton Cramer) to a California DoJ study about the issue which came to apessimistic conclusion, based on the difficulty with multiple positives. Basically, the computer sorting would produce so many possible matches, which then have to be sorted by hand, that a large database would become practically useless for finding a true match. (On a blast from the past side note, one of the listed major contributors to the study was the LAPD's own Dennis Fung, everyone's favorite criminologist.)

This could indeed be a deal-killer for the proposal, although it should be pointed out that the study, by my reading, was looking at shell casings rather than bullets, both of which can be "fingerprinted." It's also not clear whether this is a really tough problem or just due to poor software pattern recognition programs. Pattern recognition software is increasing in capability by leaps and bounds--something that doesn't work today could easily be practical in a year or two.

But the real bottom line is whether investing in a ballistics database is cost-effective. How much would it cost, and how much would it help? If I didn't think they'd already made up their mind and were just stalling for time, I'd say the new Presidential proposal to study the issue is the correct step. My guess, though, is that the "study" has predetermined conclusions, and the facts will be marshalled to support those known conclusions. But without some real analysis of the issue, right now what you've got are two sides with vague concerns, each of which can make a plausible case either for or against a database. Which tells you nothing except what their preconceived notions were.

Link posted by Doug Turnbull at 2:54 PM

A Renaissance blog: Politics, sports, literature, history, and whatever else strikes my fancy.


Sneaking Suspicions
Crooked Timber
Unqualified Offerings
Asymmetrical Information
Cranky Professor
The Poorman
newsrack blog

Powered by Blogger