Not So Fast

Ballistic fingerprinting won't work in response to the D.C. sniper.

By Dave Kopel & Paul H. Blackman

National Review Online. October 23, 2002 9:00 a.m. More by Kopel on "ballistic fingerprinting."

The Sniper has reinvigorated gun-prohibition groups. Gun control is the answer for them, of course, to a murder spree by someone using some kind of rifle or handgun in a caliber of approximately .223. The prohibition lobbies and their media dupes are calling the suspect .223 "high-powered," but it's really about as low powered as a centerfire rifle gets. As a hunting round, it's for varmints, not big game. While the Violence Policy Center has been making a big deal about "sniper rifles" and "sniper culture" (thereby denigrating the many decent people who serve as snipers for the police or the military), the .223 caliber — if that's what the shooter is using — is quite far from the long-range, high-power type of rifle used by police and military snipers.

Thus, the gun-control issue du jour is so-called "ballistic fingerprinting" or "ballistic DNA." The theory is that when a bullet goes through a rifled barrel — handgun or rifle — striations are made on the bullet that are unique to that particular barrel, just as DNA and fingerprints are unique. So, once you have a crime bullet, all you need to do is compare it to all other bullets, find your match, and go from there, just as you would do with a fingerprint or DNA sample.

Note the first obvious problem: You need something to compare it to. Some law-enforcement officials would probably like it if they had computers files of everyone's fingerprints and DNA, and such a database would certainly help solve many crimes. A scheme for collecting such private information from every household in the United States would be politically impossible currently — but once the privacy of half of all American households that own firearms is breached, then breaching the privacy of the other half of households may be politically easier.

Supposedly, the government can collect a barrel print from every gun, or at least every new gun, and then a crime bullet can be compared to the sample. Obviously, that also requires knowing who owns each gun associated with a particular bullet, which is one reason why the program requires universal gun registration. If it's okay to require that all rifles and handguns be sampled and clearly associated with the owners who own the barrel that can make such markings, it hard to see why it would not be okay to require fingerprints and DNA samples from everyone. There are far more crimes that can be solved with fingerprint or DNA analysis than with bullet-striation analysis.

Indeed, the case for fingerprint/DNA collection is stronger than the case for collection of bullet markings; fingerprints and DNA are immutable, and thus very useful for criminal investigations. But barrel markings change. They change over time, microscopically with each firing of the gun; for inexpensive guns with softer metal barrels, 50 or 100 rounds can change the striations. If 1,000 rounds are fired through the gun (as would be common during a long weekend of serious target practice), the changes are all the greater. In addition, scraping the inside of the barrel physically or chemically can deliberately alter barrel markings.

Finally, for many guns, the barrel can be changed entirely. Barrels are commonly available after market components. Gun owners often buy replacement barrels to improve a gun's performance, or to replace a barrel that has been worn out by heavy practice. For guns that have integral barrels that can't be switched out, the barrel can be rebored — as is often done for a gun that has been worn out by heavy use.

Even if we somehow register and collect data from every new rifle and handgun, from the many tens of millions of rifles and handguns currently in private hands (including guns owned by criminals), and from the unknown number of spare barrels currently in private hands, and even if there were some magical way to prevent gun barrels from being altered by scratching, there's the problem of shotguns. Unlike handguns and rifles, shotguns are not rifled; that is, they do not have grooves in their barrel designed to make a bullet spin. For this reason, shotguns aren't as accurate at longer ranges, especially compared to rifles. But shotguns loaded with slug or with large pellets are accurate enough at ranges of 100 yards or less (the probable range of the Maryland/Washington/Virginia murders — and of more than 99 percent of violent-crime guns) to kill a person. So a 100 percent perfect "ballistic fingerprinting" system would likely encourage criminals to switch to shotguns (sawed-off shotguns are almost as concealable as large handguns), but criminals would still have an easy way to avoid detection.

An additional problem is the difficulty of making the comparisons. Nowadays, bullets, fingerprints, and DNA are matched by taking the sample in question and comparing it to a relatively small number of other samples — sometimes just the DNA/fingerprints/bullets of one suspect, but, at most, several possible suspects. The wider the sampling, the better the chances of finding a match, but the greater the time and expense necessary to do so. The more resources that are spent on data hunts, the fewer resources that are available for other forms of criminal investigation.

For example, Maryland has mandatory sampling of all new handguns, at a cost to handgun buyers of about $20 per gun (the cost to collect the sample). For the state of Maryland, cost of the equipment and manpower to operate the equipment amounts of $5,000 per handgun sold. The system has thus far solved no crimes. Thus, so far, every dollar spent by Maryland for the sampling scheme has been wasted money, money that could — in a state currently suffering a budget crisis — have been spent on more detectives or in other ways that really do solve crimes.

Perhaps one day, there will be a handgun owner in Maryland who will be a lawful registered owner, who will not have intentionally or unintentionally changed the barrel print, and who will commit some crime that gets solved by the Maryland database. The crime might be an attempted murder, or it might be illegally celebrating the new year by firing a gun in the air; the cost of solving this crime might be approximately $500,000. Although registered guns are used in an almost infinitesimal percentage of violent gun crimes, if spending $500,000-per crime was actually the answer to solving a quarter of the nation's annual gun crimes, the spending would consume nation's entire criminal-justice budget. Wouldn't you rather have a court and maybe a prison, too? And perhaps a few detectives and other police officers?

But that cost figure is based on the hypothetical that the system would work. In fact, ballistic markings are not even remotely as reliable as fingerprints. Indeed, they are not even as reliable as tire tread analysis. If barrel markings were reliable, then an analyst ought to be able, at the least, to distinguish different types of guns. A tire-tread analyst can usually tell what kind of tire made a tire print, even if he can't be certain which individual tire made a mark.

As a bullet travels down a barrel, the rifling makes the bullet spin in a spiral motion. This is done by putting lands (lines where the barrel is narrower) and grooves (lines where the barrel is wider) in the barrel. The number of lands and grooves, and their widths, vary depending on the firearm model. The amount of spin put on the bullet also varies. For example, Colt Sporters now generally spin 360 degrees every seven inches, and Ruger Mini-14s every nine inches. The easy part of bullet analysis should be determining which barrels might have made the markings on the bullet, just as the easy part of tire-tread analysis is determining which models of tires have the same tread as those found at a crime scene.

It is thus telling that the authorities in the D.C.-area murders, the police are unwilling to exclude any of dozens of models of .223s from consideration. The police even suggested that bullets might have come from an AK-74 (an Eastern European firearm which is relatively rare in the United States), even though the AK-74 fires bullet with different dimensions from the .223.

One of the problems with any attempts at analysis is that bullets — especially higher-rifle-velocity bullets — are frequently deformed by whatever objects they strike. To use the DNA approach, imagine analyzing a DNA sample with 50-80 percent of the letters missing. Or a partial fingerprint where the majority of the print isn't there. The authorities in D.C. have announced that two of the bullets are worthless even for matching to the other bullets. The remaining ones appear not to be giving much information, even of the broadest kind.

The best that can be said about the rest of the ballistic "fingerprinting" — firing-pin and ejector marks — is that they're less unique than barrel striations. Firing-pin marks are also subject to natural and intentional alteration.

In short, so-called "ballistic fingerprinting" is vastly less useful than real fingerprinting or DNA analysis as a crime-fighting tool. It's far less useful than tire-tread analysis. By consuming immense resources from the criminal-justice system, the gun-registration system would seriously reduce criminal-justice effectiveness and cut the number of cops on the street — as Canada's simpler gun-registration scheme already has.

From the viewpoint of the prohibition lobbies, however, the misnamed "ballistic-fingerprinting" scheme does have advantages. The scheme amounts to partial gun registration today (in any form that could be politically viable in the legislature), setting the stage for more comprehensive gun registration in the future (without which the scheme would be useless). Since "gun registration" is a political loser almost everywhere, the gun-prohibition lobbies have the opportunity to push for registration under a new, high-tech name. And what is the purpose of gun registration? The former president of the group currently known as the Brady Campaign, the late Nelson Shields, explained registration's purpose in a 1976 New Yorker interview:

The first problem is to slow down the number of handguns being produced and sold in this country. The second problem is to get handguns registered. The final problem is to make possession of all handguns and all handgun ammunition — except for the military, police, licensed security guards, licensed sporting clubs, and licensed gun collectors — totally illegal.

Gun registration has been a very useful tool for gun confiscation in England, Australia, New York City, California, and many other places. Criminals don't register their guns, but many law-abiding citizens do, and when the government makes ownership of the registered gun illegal, the gun-owner, knowing that his gun is already on a government list, has little choice but to surrender his gun to the government furnace. This is a pleasing result for the gun prohibition groups, but if this is the public policy direction for America, we at least ought to acknowledge what is being done, rather than pretending that gun registration — cloaked in a high-tech euphemism — is going to solve crimes.

— Dave Kopel is a columnist for NRO. Kopel and Blackman are co-authors of No More Wacos: What's Wrong with Federal Law Enforcement and How to Fix It.