It's also a good explanation of how there's already a de facto gun registration in place.
With all due respect, I can think of hardly anything less registration-like than boxes and filing cabinets full of hand-written, misspelled, yellowing logbooks and Form 4473s scattered through the back offices of every gun shop in our fair land.
Is the system remarkably efficient at turning up the whereabouts (or at least the origins) of a gun that has been manufactured recently and hasn't changed hands too many times in the private sector? Yes.
But as for it being used as a "registry" or some "master list" of gun owners, the Powers That Be would be better off using the phone book.
27 comments:
I didn't say it was an automated system. ;-)
But, like you said, it can be used to identify gun owners. So, to me, it's a form of registry. Not complete, not accurate, but something's there identifying gun owners.
I concur with Rust. I've been saying it for years. It is a defacto registry, no matter how much lipstick you put on it.
-SayUncle
"Registry" is a singular noun.
At best, you may claim that there are about 100,000 separate registries that could theoretically be collated (and where would they get the money, time, or manpower?) to find out who has bought guns from FFL dealers in the last twenty years. That may or may not still have them. That may or may not still even be alive.
As well to say that pdb is helping to maintain a national registry of video game console owners via his customer records.
Like I said, they'd be as well off using the goddam phone book.
Simple question: do they have a list of gun owners or could one be compiled?
Regarding money, time and manpower, the fed is the largest employer in the world.
No one said it was a good registry.
-SayUncle
But, like you said, it can be used to identify gun owners. So, to me, it's a form of registry. Not complete, not accurate, but something's there identifying gun owners.
it would be much easier and more efficient to track gun owners via any purchase of a "gun accesory" via credit card.
Nobody, not even the BATF, wants to deal with the large scale hassle of digging through 15K of 4473' forms or some pawn shops variance program to find out who is buying a single pheasant gun and who is buying an EBR once a month.
Could it be done? Absolutely. Will that be the "way it is done"? Absolutely not.
George at "Sportmansguide.com" has more valuable info for the BATF he can spreadsheet out to demographics and regions in a matter of hours than 3 months of digging through paper forms will glean. And may more easily succumb to the .gov pressure than Bubba at "Bubba's Guns and Bait" will. No matter how much lipstick is put on it.
think about it like so...
What would tell you more about "EBR owners"?
1. get me the last 5 years worth of 4473 forms from every FFL in Texas.
or...
2. Call "Cheaper than Dirt" and have them send over a list of anyone who has bought more than 500 rounds of any "mil-surp" ammo in one purchase. Have them hi-lite those that have bought more than 1K rounds a year.
extrapolate that into further possibilities.
As a software engineer who worked in the mailing industry, I know something about paper handling and OCR. 150,000 forms could be feed, OCRed, and in a database in a day. Handwriting recognition is getting pretty good these days, especially for forms. I hate to say it Tam, but paper format is not much of a hurdle anymore.
As a software engineer who worked in the mailing industry, I know something about paper handling and OCR. 150,000 forms could be feed, OCRed, and in a database in a day. Handwriting recognition is getting pretty good these days, especially for forms. I hate to say it Tam, but paper format is not much of a hurdle anymore.
you are right in the technology. I know that scanners that copy 2 sides at ~30 pages a minute can be had affordably by a small business. Doing something on a larger (.gov) scale is not going to be a huge issue mechanically.
The point, as Tam (I assume) is trying to make and I support, is not the mechanics or the technology, but the logistics.
I don't know how many 4473 forms are in existence, but assume you took just 3 major dealers in town and ran their 4473's, you are looking at accumulating and transporting a huge amount of paperwork to your site. You have to log it, account it, and store it, and then re-transport it. Then you take the data, which is all handwritten, and hope your recognition software picks it all up and out, and then try to collate all that data into some coherent stream that will tell you, in the end, that Charlie Jones bought an FN-FAL. It actually does not tell you where said FN is, mind you. Just that one was purchased.
Again, it can be done, but handling huge amounts of 4473 forms (say 900K per year or more, per BATF regional area as a wild estimation) is idiotic.
So many more ways to get better information cheaper and with more applicable accuracy.
The BATF (in some apocolyptic crackdown)doesn't want to spend time looking for guys buying high end Beretta over and under shotguns. And that is going to be more than half of what the 4473's will collate for them. They want specific products traced to end users. Finding those users based upon ammo sales, accessory sales, and the like is tons more efficient and easy, as the major retailers already store the info in a format well suited to "tracking one down".
as usual, I hit "reply" too soon.
How about this idea.
Assume the BATF goes hogwild on some "gun ban craze" or whatever. Would you be more worried about your 3rd or 4th owner AR-15 if:
1. The BATF was heard to be collecting millions of 4473 forms from dealers accross the States...
or:
2. The BATF had just seized all transaction records from Tapco, S.O.G., and AIM Surplus?
pc,
Thank you.
I remember a controversy about an attempt by the ATF to create an illegal registry during the Clinton administration. When ATF tripled the license fee, thousands of "kitchen table dealers" went out of business and turned in their 4473s. I can't find any substantiation of it on the web, but I did find this story about a federal court ordering a dealer to turn all of his records for 1999 over to ATF. Don't know how the appeal turned out.
I'm confident that the ATF has computerized as many firearms sales records as it could. Only we peasants are required to obey laws.
The first obvious step would be to prevent any future transactions between individuals. No transfers but through an FFL. Then put together a registration system for the FFL. Otherwise registration would be a waist of time.
I don't know how germane it is to this conversation, but an 03 C&R FFL holder is not required to either turn in nor retain his records after his permit expires.
Food for thought!
porta's cat said... ...you are looking at accumulating and transporting a huge amount of paperwork to your site. You have to log it, account it, and store it...
Three letters and four numbers for you: IRS 1040.
You vastly overestimate the amount of work involved in this. You have no idea how much paper a place like Chase bank generates and handles in a single day. All of the problems you're pointing out were solved years ago. And remember, ATF already takes and stores the 4473s from any dealer that goes out of business.
OCR of forms? This is a well established, huge market already. Go see http://www.parascript.com/solutions2/forms_processing.cfm. Quite frankly, their handwriting recognition kicks ass. There's little you could fill in on a 4473 that a human could read that it couldn't read, and read way faster than you ever could. And what it can't get, keyists can verify.
The database for this is trivial and doesn't need to be terribly accurate. It doesn't matter if you get four copies of the same name and address. That's REALLY easily ironed out with USPS address validation (http://zip4.usps.com/zip4/welcome.jsp).
The full system could be in place, and all of the back log of 4473s scanned in two years. This is mostly off the shelf components. Once the system was in place, do you have any idea how cheap and efficient it would be for the ATF to keep it updated by requiring that every gun dealer mail them every 4473 filled out that week?
Now, while it certainly might be easier for the feds to figure out who probably has an EBR by data mining places like Cheaper Than Dirt and Sportsmansguide.com, that's not what Chuck Schumer is looking to do. He doesn't care if you own a single shot shotgun or an AR-15, he wants to register every single gun and gun owner. If he could take this step, it's one big step in the direction he wants to go. He doesn't care about the efficacy of his measures. He cares about passing legislation to register and eventually ban guns. The next step after computerizing all 4473s would be to require 4473s for private sales and making the 4473 a web form instead of paper.
The problem with that scenario is that if a registration law like that passed, it would cross many people's Line in the Sand.
I personally (that means "not anecdotally") know plenty of FFLs who shouldn't store all those oily rags around the cabinets that hold alltheir 4473s.
The full system could be in place, and all of the back log of 4473s scanned in two years. This is mostly off the shelf components. Once the system was in place, do you have any idea how cheap and efficient it would be for the ATF to keep it updated by requiring that every gun dealer mail them every 4473 filled out that week?
you are obviously either an engineer or a clerk. You beleive that "your technology" or "your system" is of far more import than is reasonable to the task at hand.
If the Evil Empire wanted said information (4473's), they would lose a large % of it by default, as Tam suggests. There would be tons of unexplained fires. If the Evil Empire wants said information, it would be far better and easier to get it in real time, not by processing millions of tons of paper over two years. Chase Bank has a system in place for absorbing paper of that maginitude. BATF does not. The systems that would have to be insitutued prior to make it unefficient before it starts.
A "gun registry" would logically start at the NICS check. That system is already in place. No need for:
requiring that every gun dealer mail them every 4473 filled out that week?
To get an immediate grasp of "who owns what" would be impossible. So, the best route would not be finding the guns, but finding who buys "things for guns". Far easier to legislates or cow retailers than to try to track down paranoid Bubbas raised on Red Dawn. You would do two things in such a move. You would essentially cut of the food supply for firearms, and you would have a very good list of "active" gun owners.
Anyhow, it is a pretty ridiculous point you press. "They" could also knock on every door in the USA and do a search. They have the manpower and time to do so. In 2-5 years that could be done as well. That would turn up far more firearms than 4473's would, as 4473's only show a "point of purchase" for a specific class of sale.
Anyhow, I think that you are thinking too much about the application of technology and have gotten lost in possibilities over realities.
It's a registry. Granted it is so inaccurate and incomplete to be useful, but it's still a registry.
Well.
What I do for a living is build systems that turn paper into data. The organization I currently work for scans, on average, 40-42,000 pages per day, almost all of which go through OCR and ICR and specific fields are validated against databases to make sure we get things like names spelled correctly. The document images are linked to database records - which act as metadata - so one mouse click on the record brings back the image.
For the software we use - which is commercially available, but has quite a few customizations in it - that's a pretty trivial task. For hardware, we have a daily total capacity of a little under one million pages for a 24-hour day. We do a fair amount of remote scanning: we have 16 scanners ranging from 50ppm to 120 ppm, in six buildings spread out over a 25 mile area, all connected to a central repository which is managed by almost a dozen servers, and which gets backed up nightly to a second remote repository. Our total storage capacity is not quite three billion pages.
What gives us the most trouble is free form text, just words on a page, because there's no structure to it and the OCR software has to bust its butt. Give me a form, however, and it gets pretty easy. I'll place a form box with X-Y coordinates to get good reads, and since I know what's supposed to be in that box, the known universe of options gets pretty small, so the software can run faster. Plus, I can validate one form box against another.
Numbers are easy - there are only ten of 'em. If I read sloppy printing as "Murxhy" instead of "Murphy" I can extract the zip code from the address block, and the street number, and bounce it off a database with driver license info. There will be only so many people whose name starts with M, ends with Y, and who live at street number 1234 in zip code 12345, and I can use other info on the form to select which street number 1234 has the highest probability of being Murphy's. Give me access to another database - say, a utility billing database - and I can lock down 'ol Joe Murphy's geo location pretty tight, even given some room for error in my read probability algorithms.
That said, the labor required to feed that many pages through scanners is not trivial. We do only 40K a day, and we're up to our eyebrows in it. And, knowing Joe Murphy's address doesn't mean that Joe is there. Names like "Potrazeebie" are easy to dial in, but the Smith and Rodriguez families consume a lot of processor cycles to narrow down, and sometimes a list of three to five is the best I can do, from a total set of about two million. I may be able to knock on a few doors and find the John Smith I'm looking for, but expand that total set to 100-125 million and the task becomes impossible.
Not that it couldn't be done, but people in my industry would know that it's happening because it would suck up a large segment of a fairly small talent pool. And, a lot of us read blogs like Tam's and Kim's, so keeping it a secret for long would be pretty tough.
Forward Trace - BATF program several years back that went to FFLs and "forward traced" folks who purchased EBRs. AR-15s were a constant favorite.
Hey - I really enjoy reading your blog, and I wanted to post a link to a newer blog that I've been keeping...I'm trying to build a reader base. I do quote you from time to time, too. :-)
http://squeakywheelseeksgrease.blogspot.com
Thanks!
As I just can not resist throwing gasoline on the fire, remember that CCW is a de facto registry as well. While it does not catch ALL firearm owners, it does register a specific subset.
So is NRA membership, shooting club membership, etc.
So is NRA membership, shooting club membership, etc.
BINGO!
If you go for the pre-apocolyptic , circling black helicopter, tattoo of 666 on every hand or forehead type scenario, the Evil Empire of Satan is not going to put all their nest eggs into the compliance of FFL holders to supply them with names.
Instead, they will do what you always do when you start such a large scale investigation. You follow the money.
If I had "unlimted access to records", I could spend a few hours on the network and come up with a list of names of gun owners, and by most likley "type and caliber(s)" and never once type "FFL" into my search box, and that list would be more exhaustive and conclusive than a bunch of 4473's. I could relatively easily collect on 90% of shooters in a given area:
1. Names
2. Shipping addresses and physcial addresses.
3. Types of firearms potentially owned or with access to based upon calibers of ammo purchased and/or accesories (mags, furniture, etc.)
4. How those firearms were stored and what that implies about your number of firearms and type (safes and number of safes and type).
5. The rough amount of ammunition on hand, or at least potentially.
6. Where you might congregate or shoot. If "I" couldn't find you, a bit of excessive pressure on a few of your shooting buds would turn you up poste-haste.
Information that is far more important, in terms of a major jackboot crackdown, than "you purchased a shotgun back in 1999 and an AR-15 in 2003."
As many "unpapered" guns, and "grandpa always wanted me to have his old M1 Garand" type things out there, no sense in wasting an inordinate amount of time on a literal paper trail.
In fact, I bet a few guys in suits and shades could turn up 75% of the gun owners just by questioning women pushing baby carts in the park of your neighborhood.
Now, I am getting as weird as the rest of you.
As rustmeister states, there is a "gun ownership registry" that exists for all time electronically and it is literally at the finger tips of anyone with a marketing software program.
porta's cat, I think you're missing the point.
We're not talking about the logicalness or effectiveness of such a registry. I certainly don't think the feds would try to create a registry secretly via this method. I certainly don't think that it would be a very effective registry. I did not address the issue of civil disobedience or sabotage. I certainly believe some would happen, but do you really want to rely on that? My point was simply to show that having 4473s on paper and distributed at many locations is not much of a hurdle to digitization of those forms.
I think Chuck Schumer et al. would pass a law requiring that it be done, if they could, and would crow loudly about it, just like they did with the useless AWB.
Yeah, logic and gun control don't mix....
I don't know if I am so much "missing the point" as making a different point.
Certainly most anything can be legislated, and you are right that given the right political and social environment, every 4473 form in the USA could be digitized and then made into some mass registry. You could conceivably even build in some safety mechanism to "fires of unknown origin" by scaring the Beejezus out of retail FFL types with mass fines and imprisonment and all of that stuff. Some free spirits might test the resolve of the .gov to "do anything", but by and far most retailers are not going to do anything but comply. So yes, in tat, you are at least theroretically correct.
However, if the .gov was trying to take my gun out of my warm and alive fingers, I would hope they would resort to the long and tortuous method of collecting 4473's. Geez, what a convoluted method to do so.
As gun owners, most of us by and large on "in the matrix" in so many easier to find ways than that.
S.O.G. has gotten in trouble before with the BATF. Suppose they do it again, and Schumer and some AG decide to seize all of SOG's retail records for "investigation into conspiracy". How many "gun owners" will now be in the potential eyes of .gov in one fell, and well digitized, swoop?
What if .gov simply decided that from now on each firearm transaction needed to be electronically processed with make, model, and serial number and blipped to FBI-NICS for an immediate "red light -green light" done in real time?
So...yeah, 4473's could be used to trace firearms. They are, as Tam suggested, in individual cases. But as a mass .gov method of tracking and finding gun owners, it would be an abject blessing to anyone trying to skirt the law or remain hidden in such.
As an FYI, from a letter to Clinton from Reno describing some of the successes of the NICS check, etc.
Over 10 million checks were run through NICS between November 30, 1998 and December 31, 1999.
Assuming that is reasonably accurate, and with the surge of firearms purchases since that time, there are a LOT of 4473's being drawn up. Assume that 5% of those are CHL holders (these days) who do not get a NICS check, or 500K additional forms, and just to keep it simple, we can assume that it has since averaged out to ~13 million checks a year, and you are looking at about 104 million 4473 forms generated since 1999 or thereabouts.....
I wonder how many times that would stretch to the moon.....
http://www.usdoj.gov/opa/pr/1999/November/574ag.htm
Post a Comment