Spin, Spin, Spin

A couple of posts down is the story of Caroyln Lisle who shot an intruder in her Rancho Cordova home. Quoted in the story is one William Vizzard, described as "chair of the criminal justice department at California State University, Sacramento." One commenter called him "a gun control flunky" and suggested Googling to prove it. So I did.

I found this interesting transcript from PBS's Newshour from October 18, 2002 where Mr. Vizzard, described as an ex-employee of the BATF was one of the panel. The discussion was about "ballistic fingerprinting," and was inspired by the fact that the DC snipers were active at that time.

This will be kinda long, but I'm going to fisk it.

RAY SUAREZ: The recent sniper attacks in the Washington, DC, area have revived a debate over a technology that helps authorities trace ammunition found at crime scenes. The technology is called ballistic fingerprinting, and it's based on the idea that every gun leaves unique markings on its bullet casings.

Um, not quite so unique. Modern manufacturing methods and tooling mean that guns coming sequentially off a production line are very likely to have very similar tooling marks.

Gun makers would be required to register those fingerprints so a national database could be compiled. Until recently, crime labs relied solely on the human eye and a microscope to look at evidence from bullets, but now bullets, bullet fragments and shell casings are scanned into a computer and compared against thousands of other bullets or casings.

SPOKESMAN: When a barrel is produced or a firearm itself is produced, it's made by other tools. The metal is formed and moved around, scraped away, and those imperfections of each of those tools in the manufacturing process have accidental characteristics it imparts on the gun. It's still a needle in a haystack, but now we can get through the haystack faster. These comparisons from bullet to bullet are into tenths of seconds.

This assumes that there's enough left of the bullet to allow identification. On top of that, the bullets being compared must also be of similar composition. For example, a .45 caliber 230 grain full metal jacket bullet made by Speer will have much different markings than a 185 grain hollowpoint bullet manufactured by Federal when fired from the same gun. Especially if the first was fired into a water barrel and the second recovered from a corpse after impacting a major bone. A computer probably wouldn't be able to get a match. A human eyeball Mark I might. But the human eyeball takes a lot longer than tenths of a second.

RAY SUAREZ: Law enforcement officials back the idea of ballistic fingerprinting and so does the federal Bureau of Alcohol, tobacco and Firearms. Several lawmakers have called for legislation requiring gun makers to record the ballistic markings. The National Rifle Association and other gun rights advocates oppose legislation, saying the fingerprinting is an unproven science.

The Bush administration was also skeptical, saying earlier this week the technology might not be reliable and could infringe on privacy. But on Wednesday, Spokesman Ari Fleischer said the President does want to look into creating a national registry.

ARI FLEISCHER: The president wants this issue explored. And to that end, the Bureau of Alcohol, Tobacco and Firearms has been meeting, and met yesterday afternoon with White House staff to start to discuss the various issues: The technical issues, there are feasibility issues, the pros and cons about how this could possibly... may be effective, whether it could work or whether it would not be able to work.

Gee, thanks Mr. President.

RAY SUAREZ: While the national debate continues, two states, New York and Maryland, have already enacted laws requiring a ballistic fingerprinting for handguns.

Neither of which has yet to have a match that didn't identify a crime gun already in their possession, contemporary with the crimes, and firing ammunition that matched that found at the crime scene.

We pick up the debate with Joe Vince, the former chief of the crime guns analysis branch of the Bureau of Alcohol, Tobacco and Firearms. He now has a consulting company in the Washington area. And William Vizzard, a former ATF agent, now chair of the division of criminal justice at California State University at Sacramento.

Well, since the speculation, Joe Vince began with the Washington area sniper, let's take a look at how ballistic fingerprinting may have been useful in a case like this, in investigating a case like this?

JOE VINCE: In a case like this, Ray, this right now, the ballistic evidence is your best evidence. It's almost your only evidence. So it provides a lead for law enforcement. And any time you're working an investigation, law enforcement officials are looking for leads that take them to the next step.

You build your case incrementally. And for this, knowing that it was a .223 what type of firearm it could come from, is very useful.

RAY SUAREZ: But would you, if you had this database, have been able necessarily to narrow down what firearm?

Note, now, how Mr. Vince artfully dodges answering the question:

JOE VINCE: With a database like this, the possibilities multiply. And we have to remember that law enforcement today needs to rely on 21st century technology. In 1890, if you wanted to get in law enforcement, you received a badge, a gun and a club, and said go out there, enforce the law.

Well, that was towns of hundreds. Now we have metropolitan areas of millions. Law enforcement has to leverage technology in order to help them solve crimes -- as the gentleman said on your earlier piece, to take the haystack and eliminate as much hay as you can to find the needle.

Simple question, wasn't it? Would the technology help narrow down what firearm? Answer? Evasion, obfuscation, and haystacks. Is he called on this? Don't be silly.

RAY SUAREZ: William Vizzard, would this have been a useful tool in this investigation?

WILLIAM VIZZARD: Well, it conceivably could be although -- given the circumstances -- an individual who apparently has planned these shootings in advance, it's most likely not in the sense that we have about 200 to 250 million guns in circulation in the United States today.

And it's possible for an individual simply to acquire one of those and use it, knowing that it's not in the database. I mean, if we were to begin a database, for instance, today, or at whatever point Congress would cease debating it, presumably, it would start recording bullets and cartridge cases from that day forward.

Now one could of course try to collect the 250 million existing samples out there. But I don't hear anybody really advocating that because the mechanics of simply trying to track down those guns and get some sort of record on them is extremely difficult. So Joe is certainly right about the technology; it's extremely useful technology.

It's proven very useful in a number of crimes involving suspect firearms and bullets or cartridge cases recovered from crime scenes. And had we been doing this from the 1930's on, and of course in those days we didn't have in way of cataloging it, we might have some utility at this point.

Hey! An actual answer! No wonder he's no longer in the ATF.

RAY SUAREZ: But given the plans that are under consideration now this gun would have had to have been either used in a crime before or purchased and profiled at the time of purchase in order to get a hit in a database, is that right?

WILLIAM VIZZARD: It would have to be placed in the data base through scanning in either at time of manufacture or when it was picked up by the police.

Of course unlike fingerprints when you fingerprint an individual and they're subsequently released and you have their fingerprints, in the case of guns, normally when police get their hands on a gun, they don't release it and so it's usually useful only for checking against previous crimes as opposed to building a database for future crimes.

This, too is factually accurate. So far I'm impressed.

RAY SUAREZ: I'm sorry, Joe Vince, go ahead.

JOE VINCE: Well, a good comparison is over 100 years ago when we started fingerprinting. We had no database and we were doing everything in a card file. However, we said this is a good tool to use and it has been extremely useful. Now we have a computerized AFA system; that's a national system that has fingerprints computerized.

And we don't only put bad people into that system of fingerprinting. Every man and woman who enters our armed services is fingerprinted. Schoolteachers are fingerprinted.

My wife is a schoolteacher; she's in there. The reason for that in the military is to obviously check their background, check the teacher's background but also, God forbid, if they were injured or killed in the line of duty, we could identify them.

This is the same thing we have to do. We have to take incremental steps now and build our database up so we have the same capability that we have with fingerprints.

Yet no one suggests that we fingerprint and DNA scan every single individual so we can pick criminals out of the population from crime scene evidence.

RAY SUAREZ: William Vizzard notes that there are some 250 million guns already out there. How long would it take until you had a database that was actually useful, a body of profiles that was large enough to be useful compared to the number that's already out there?

And, once again, Mr. Vince dodges the very simple question: "How long?"

JOE VINCE: Well, I agree with Bill, there are a lot of firearms out there. But we have to take the next step. (And there is ALWAYS a "next step.") I was in Palm Beach, Florida, last week and I talked to the sheriff's office there. Six months ago they received the IBIS equipment and that has already linked seven or eight different homicides and shootings together that they did not know it was related.

So you can see, in a short period of time you can have some success. We have to start somewhere. Congress wisely already allocated the money. We've put the equipment everywhere in the United States. Now we have to effectively use it as a law enforcement tool.

Uh, Mr. Vince, you matched crime scene evidence. You did not identify the firearm or its possessor. And YOU DIDN'T ANSWER THE QUESTION.

RAY SUAREZ: Mr. Vizzard, you've used the fingerprints analogy. To carry it one step further, it's pretty hard to change your fingerprints. Is it hard to change the so-called fingerprint that a firearm puts on a shell casing?

WILLIAM VIZZARD: It's difficult. It's more difficult than the opponents have characterized. Firearms are made of extremely hard steel and it takes a long time to wear them enough to significantly alter them. But they are capable of being altered, unlike fingerprints and DNA.

Not exactly true, Mr. Vizzard. For example, take two identical Glock model 17 handguns manufactured three years apart, both of which had been ballistically fingerprinted at manufacture. Run 10,000 rounds through gun #1. Then replace the barrel with a new one you can buy - without a background check, via mailorder. You won't get a ballistic match on the bullet any more. You might be able to get a shell casing match, but after 10,000 rounds I'd imagine the breechface, the extractor, and the firing pin would be quite worn and the last two items might have been replaced. Add to that the fact that the hardness of the brass and the primer cup has a significant effect on the markings put on the case and you just decreased the possibility even more. Finally, swap the slides and barrels between gun #1 and gun #2. It's the frame of the pistol that's considered the "gun." But it's the slide and barrel that leave the ballistic markings. Your trail just went cold.

WILLIAM VIZZARD: I think the real issue probably here is that the devil is in the details. It's a question of cost/benefit analysis, not a question of whether it would be desirable to have this data. I think it would be. I'm not an apologist for the NRA. I'm not morally opposed to the idea.

I simply think that if you consider the cost and the benefits, for instance, we aren't currently, I believe, scanning into AFIS, any of the prints-- any of the non-criminal prints that Joe mentioned, either at the state or the federal level. Some local agencies do.

We are taking DNA only on a very small number of samples from serious offenders. It varies from state to state, depending on what the state law is. We would probably solve far more crimes collecting DNA from everybody in the United States than we would from collecting ballistics from every gun manufactured, so I think you just have to weigh what's the cost going to be, how is it going to work. (I stand corrected. Someone has suggested it.)

Is there going to be a chain of custody issue, which I haven't heard anybody discuss; you can get a lead without a chain of custody issue, but if you want to actually make the comparison and you don't recover the firearm, that's going to be a problem.

So I don't think it's a case of it being a bad program in the sense that it's evil. I think it's just simply a very difficult program. And before you rush into it, you sit down and you figure the cost and you figure the benefits. And you say what would we do with the money if we didn't spend it on this. That's my only point.

And a good one it is.

RAY SUAREZ: Well, Joe Vince, how would it work? A lot of the firearms sold in the United States are made overseas. There are domestic makers and sellers as well. At what point in the life cycle of a gun would we check the markings that it puts in the firearm?

JOE VINCE: It would have to be when the firearm sold. Right now in Maryland and New York, they're doing it with new handguns. And it really is not keeping a database of names. It refers back to a serial number of a gun and then back to the records of that dealer.

So the government really doesn't have the information. But we do it in a way that's very similar to the tracing of firearms that we do now for crime guns, which has also been very useful. But again I really think we have to look at integrating this, too with the various information systems we have in law enforcement.

Like, say a gun registration database? That would be the logical "next step" would it not?

JOE VINCE: The idea is that law enforcement collects enormous amounts of information. This is just one piece and DNA is another. But it's getting knowledge from all that information. That's what we have to look at. So it is integrating this so we can get those leads consistently and so that crimes like the sniper in Maryland can be swiftly apprehended.

RAY SUAREZ: How about that, Mr. Vizzard, the idea not being that it would provide absolute information, but when cross referenced, when overlaid with a lot of the other sources that police use, it might be useful?

WILLIAM VIZZARD: It would clearly be useful in some cases. My guess is that for sometime what you would get are rather poorly planned crimes, particularly among younger offenders who tend to acquire new guns more readily than older offenders.

I suspect-- I really would question Joe's characterization of collecting at the time of sale. Frankly collecting at the time of import or manufacture would make more sense. We're talking about a lot of guns here and I envision ATF being back where they were when they used to put personnel at the distilleries -- simply putting somebody at the factory and scanning the data in there, but without a national gun registration and licensing system, you've got real limits on the value. (Thank you for making my point, Mr. Vizzard.)

And of course that's why the NRA gets so exercised by it. I'm as not offended by a licensing and registration system as they are. But without that information, private sales very often result in guns just simply being swallowed up and disappearing.

And we do oftentimes trace guns to individuals. We oftentimes lose the track, also. So I think you just have to again analyze the worth of the system as it relates to the specific kind of information you're looking for. Nobody, I think at this point, can estimate the cost.

Every computer system ever built has turned out to be different than people expected and I realize we're running the system on a small scale today. But if we start running on a much larger scale, we'll probably gain some economy of scale and probably also run into problems we didn't know we would have. All of those things have to be addressed.

Thus endeth the transcript.

All in all, I thought Mr. Vizzard was quite fair, and Mr. Vince was the typical official-line-spewing, job-justifying government flunky.

For further reading on the efficacy of a ballistic fingerprinting database for identifying firearms in the general population, I strongly recommend the initial ballistic fingerprinting study report to the California legislature, Feasibility of a California Ballistics Identification System , the follow-on AB1717 report - Technical Evaluation: Feasibility of a Ballistics Imaging Database for All New Handgun Sales, and the Maryland State Police Forensics Division IBIS report (a 2.5Mb scanned document in PDF format. Maryland never officially released this report as far as I can tell.)

When Vizzard said "Nobody, I think at this point, can estimate the cost" he wasn't kidding. What he didn't say was nobody can estimate the effectiveness, either. Without those two crucial bits of information, it's damned hard to do a cost/benefit analysis, isn't it?

UPDATE, 2/12: Reader Kevin P., who was the commenter that characterized William Vizzard as a "gun control flunky" has withdrawn that comment, and instead states: "I withdraw that term unreservedly and apologize to Mr. Vizzard should he ever read this.

"However, I will stand by the assertion that he is a gun control advocate. He is a rarity, an informed and knowledgeable gun control advocate, probably because of his career in the ATF. His performance in the PBS ballistic fingerprinting debate was fair and accurate - but it is something that should be expected and demanded of everyone."

Yes, it should. Kevin P. also links to this quite interesting review of Mr. Vizzard's book Shots in the Dark: The Policy, Politics, and Symbolism of Gun Control by Dave Kopel. Give it a read.

Thank you, Kevin. Stuff like this makes my day.