A Modest Gun Registry Proposal


This silly idea came to me last night while I was trying to sleep. I throw it out there for your potential amusement.

RAMPART: A gun registry that works the right way

One problem with some popular DLTs is lack of write bandwidth; reads are cheap and infinitely scalable, but writes directly to the blockchain are bottlenecked by a number of fundamental technical limitations.

What if there was a way to turn this bug into a feature, though?

Imagine the inverse of the bitcoin scalability problem – writes are scalable, but reads are greatly bandwidth constrained. A database that’s easy to write to but very difficult and expensive to read from sounds like a terrible technology, but it’s an excellent fit for one peculiar problem: a centralized firearm registry, where each record stores a chain of custody for a single firearm and is indexed by serial number (probably combined with some other metadata, like make and model).

Think about it. The primary rationale that gun control advocates offer for a gun registry is ease of tracing crime guns. But gun rights advocates are worried that a centralized registry of guns will inevitably lead to confiscation.

A a gun registry that can only be read from a few hundred times per day would address both of these concerns. Law enforcement would have the ability to trace a few hundred crime guns per day, but at such a low read rate there would be no way to compile a big enough unencrypted list of the nation’s hundreds of millions of guns or tens of millions of gun owners.

For now, I’m calling this database rampart but if you have a better name then hit me up.

Possible implementations

A fairly simple and naive implementation of rampart would be one in which a record of a firearm transaction between two parties is created and encrypted locally, and the keys are discarded before the encrypted transaction is sent off for storage in a central database. Because the keys have been thrown away, anyone wanting to read the record would have to throw compute power at the problem of decrypting it, and this would make reads expensive and slow.

Depending on the encryption algorithm selected and the key size, we could tune the difficulty of cracking the records so as to keep the practical read bandwidth below some desired threshold.

This scheme has a number of problems, though, the main one being that if the encryption algorithm we select ever falls to some computationally cheap attack, then the entire database is compromised.

Another possible implementation might be to require a block to be written to a public, write-constrained blockchain like bitcoin before a read can be carried out in the database. I’m still thinking about how this might work, though.

A rhetorical intervention disguised as a technological invention

Gun controllers will complain that rampart doesn’t enable law enforcement to look up all the guns held by a specific owner; they’ll claim they need this ability so they can take weapons from people who are deemed unfit by virtue of a mental illness diagnosis, suicide attempt, domestic violence conviction, etc.

This complaint is correct; this database does not support any sort of confiscatory use case by design. Rather, the point of the database is to neutralize the primary “tough on crime” rationale that gun controllers give for wanting a central gun registry, and to take that off the table entirely, leaving them with “confiscation” as their only argument for a standard gun registry.

Note that this database doesn’t actually have to exist as anything other than a proposal and a working codebase for gun rights advocates to get most of the benefit of it. If all it succeeds in doing is to take tracing crime guns off the table as the stated rational for a registry, then it will be a success.

Note: Some will be offended at the apparent lack of good faith in this proposal, but I take it as a given that gun registries are dumb and don’t work. A read-constrained database like the one proposed here might be an acceptable way to prove that point here in the US, while keeping confiscation off the table.


Problems with this:

  1. it would be a near impossibility to get the existing millions of firearms into this database (those opposed to it on principle aren’t going to volunteer the information and obtaining the information by force is hardly an option)
  2. firearms possessed by criminals won’t be in this database severely restricting it’s utility
  3. a technological solution such as this is subject to a design shortcoming that is overcome by future technology (IOW, this is nearly guaranteed to fail at some point in the future)

The problem with most proposals such as this is that they are directed at law-abiding firearm owners who, by and large, are NOT the problem. The problem is with firearms possessed by criminals or those with mental limitations.

My contention is that law-abiding firearm owners possess firearms for a multitude of reasons but the majority possess them, at least in part, for reasons of self defense. Who are they defending themselves against? Armed criminals, of course.

That being the case, it is clear that law-abiding firearm owners do NOT want to knowingly transfer a firearm to someone who is either a criminal (“prohibited possessor” if you will) or has one or more traits shown to be indicative of a tendency towards physical violence (e.g., mental illness of some kind).

Therefore, the task at hand is to find one or more ways to incentivize and/or remove hurdles for law-abiding firearms owners to prevent the knowing transfer of a firearm to either a criminal (“prohibited possessor”) or someone likely to use a firearm in the commission of a crime.

The current impediments are that it costs money and time to have a 4473 done at a FFL. Additionally, you MUST physically present yourself at a FFL to have a 4473 performed.

In states that have concealed carry licenses, some private sellers demand that a state issued ID along with the CCW license be produced to demonstrate that the buyer is a “proper person”. That works to a degree but it is likely trivially easy to create a fake state issued ID or CCW permit that would fool an ordinary citizen not intimately familiar with recognizing counterfeit documents.

I would argue that the solution is to remove both the cost and inconvenience barriers to getting a 4473 background check performed as part of a private sale.

My proposal:

  1. make the 4473 (or equivalent) accessible online to anyone
  2. impose no fee for performing a 4473 (or equivalent)
  3. collect no information concerning the firearm being transferred (no reason to collect that information and it would be a disincentive to those worried about a national gun registry)
  4. to maintain security of private information, it is likely best that the BUYER submit the online 4473 which would then produce some sort of authorization code with an expiration of a few days. The buyer would give the seller this authorization code which could be checked online with a simple response of yes/no to whether the person associated with the authorization code was a “proper person”

My understanding is that this is, at least in general terms, how a firearms transfer occurs in Switzerland.


It’s a modest proposal. I’ll let Jon clarify, but he doesn’t expect gun owners to support this necessarily. It’s a way of “letting the garden weed itself” so to speak with regards to guns controllers’ actual aims. If they get on board with this, then they could be convinced to support what you propose as well (I concur that the Swiss model is the best).

Far more likely is an outright rejection because it doesn’t serve the aims they’ve expressed in other venues (facilitating confiscation).


Although this is something I would oppose, being more of an absolutist on the Second Amendment, I understand the benefit of discussing these type of proposals, and also get a bit of a smile anticipating the reaction from the anti-gun crowd.

One thing I would be concerned about would be how improvements in computing technology would affect cracking the encryption on the records. If it takes 10 minutes to open up a record today, how long would it take 5 years from now? Admittedly I am not very familiar with blockchain so I may be just talking out of my butt here.


My perspective is that registration for the purpose of tracing guns used in crimes is not going to move the needle, as guns used in crimes are often stolen or transfer ownership explicitly through off record means. So any registration process whose integrity is dependent upon buyers and sellers complying is not going to be effective. But I might be missing a larger point if a blockchain registry has the benefit of not relying on compliant participants.

If we’re looking at a creative way to accommodate a high value ask from the anti-gun, I’d agree that going after an effective background check process is the problem to solve. I don’t have a feel for how effective this would actually be, because again, bad guys don’t go through standard channels to obtain guns. But if it helps a little and squashes the narrative that there are easily-exploitable loopholes, then I’m all for it.

Something I never understood: why are private party transfers (which I understand is the “gunshow loophole”) not required to undergo a record of transfer and/or a background check? California requires this on a state level, with the main downsides being a 10 waiting period and having to pay a FFL fee (the time & money that @gopher mentions). But if I were going to sell a firearm directly to another private party (even a friend or relative), I would absolutely want to have the reassurance that the buyer is in the clear to own one.

I personally don’t feel that the “Swiss model” suggested above infringes on my rights. But from a “deal” perspective, I’m skeptical that the anti-gun side would recognize this as a significant give, and insist on a pile on of other requirements including registration.


@OldGamer: you just made the argument against mandatory background checks for firearm sales: “…guns used in crimes are often stolen or transfer ownership explicitly through off record means.”

IOW, criminals are going to get the majority of their firearms via illicit means regardless of what types of mandatory background checks we might impose. The only ones that obey those kind of laws are the law-abiding firearm owners and they aren’t the problem.

Again, I go back to the underlying question of why the majority (all?) of the proposed solutions to firearms getting into the wrong hands involve the stick (i.e., punitive laws) rather than the carrot (i.e., removal of barriers to law-abiding owners doing the “right thing”).

Law-abiding firearm owners don’t want their firearm to get into the wrong hands. Give them the tools to do so with minimal (preferably zero) cost and maximum accessibility without imposing gun registries or disclosure of private information. The proposal I put forth above satisfies those requirements and I haven’t seen any other proposal that even comes close.


This is also why in my Politico article I proposed a license for simple possession. The proposal presumes that criminals can and will get guns regardless of any background check regime, so why bother? License people for simple possession of a semi-auto, which would give law enforcement the ability to stop and frisk and bust unlicensed folks for possessing.

(Obviously this assumes that the licensing regime is strict enough that folks who are involved in crime won’t pass is.)


How is your proposed licensing scheme any different than the current CCW license?


It’s shall issue, which is not how CCW licenses work in many states. Onus on the state to find a reason to deny, probably based on the agreed upon definition of “prohibited person”.


“Shall issue” is BY FAR more common than “may issue” and the trend is towards “shall issue” rather than the opposite (at least that’s my perception). Basically, the “may issue” states are CA plus the northeast (RI, CT, NJ, MA, DE, MD and NY) which are all notoriously unfriendly to firearms.

See here: https://www.concealedcarry.com/law/the-difference-between-shall-issue-may-issue-and-constitutional-carry/

Still not seeing where this “semi-auto license” has any value…


Yes, you’re correct that numerically more states have shall issue. By population, more people may live in may/no issue counties than shall issue. I’d need to look into that.

I’ll be honest, I wouldn’t want to put myself on this registry, but I still think writing about it was a creative way to direct the conversation on the restrictionists side about what the actual goals are vs what’s commonly stated.


Yep, that’s right. Background checks and the vast majority of gun control laws are wholly about making people “feel safe” vs. “being safe.” Here’s an idea: what if we required all laws & regulation to be tied to a performance-based metric? So require any new law to explicitly state the intended result and define a quantified minimum metric. Give the law x years to measure the effects, and if the metrics are not met, it auto-terminates unless lawmakers go back and amend it. While we’re at it, create a publicly available dashboard rating public officials based on the performance of laws they were a part of authoring.

I believe it’s to accommodate the perception from most of the anti-gun that “fully” semi-auto is the main attribute that gives a firearm mass shooting lethality. The proposal is a deal of gives & takes where the greatest concerns from the anti-gun are addressed while actually giving back entitlements to the gun owners. What I especially like about Jon’s proposal is it leverages the ignorance of the anti-gun against them.

We already see direct conflict of rulings on the federal district court level on this. The the 7th and DC Circuit courts upheld lawsuits that CCW may issue laws were unconstitutional, while 1st, 2nd, 3rd, 4th, and 9th Circuit denied similar lawsuits and upheld may-issue laws. Seems this is the exact type of condition that would motivate the SCOTUS to take up a case and provide clarity.


I could definitely get behind this kind of results based legislation. A good many laws on the books would be eliminated if they actually had to live up to the “promises” made when they were enacted.

I see this as having a snowball’s chance in hell of ever being enacted, though. Big government doesn’t want accountability…

I guess I need to read the proposal in full; does anyone have a link?

Agree that legal clarity on whether “may issue” is even constitutional would be of great benefit. It would seem to be a difficult argument to make that “may issue” is constitutional given that we have the 2A (assertion of a positive right to bear arms) but who knows what legal contortions the Supreme Court might go through if they felt “may issue” needed to be defended.


Just a thought:

This sounds like something that should definitely be promoted in countries that already have strict licensing or records keeping requirements, as a way to scale back how those existing records can be abused by law enforcement.

Though I do understand it could be a useful compromise politically in some parts of the US too (hopefully won’t be necessary for the country as a whole).