While criminals typically do not buy their guns at a store, all but a tiny fraction of those in circulation in the United States are first sold at retail by a gun dealer – including the guns that eventually end up in the hands of criminals.
That first retail sale was most likely legal, in that the clerk followed federal and state requirements for documentation, a background check and record-keeping. While there are scofflaw dealers who sometimes make under-the-counter deals, that is by no means the norm.
If a gun ends up in criminal use, it is usually after several more transactions. The average age of guns taken from Chicago gangs is over 11 years.
The gun at that point has been diverted from legal commerce. In this respect, the supply chain for guns is similar to that for other products that have a large legal market but are subject to diversion.
In the case of guns, diversion from licit possession and exchange can occur in a variety of ways: theft, purchase at a gun show by an interstate trafficker, private sales where no questions are asked, straw purchases by girlfriends and so forth.
What appears to be true is that there are few big operators in this domain. The typical trafficker or underground broker is not making a living that way but rather just making a few dollars on the side. The supply chain for guns used in crime bears little relationship to the supply chain for heroin or cocaine and is much more akin to that for cigarettes and beer that are diverted to underage teenagers.
In essence the criminal market for guns is crowd sourced. Which means it is far more challenging for law enforcement to “close the loopholes” than if were were organized criminal organizations. The allocation of scarce resources to investigate, prosecute, and incarcerate people who sold a single “crime gun” is far better spent on going after violent criminals instead of those involved in victimless crimes of straw man purchases.
There are about 500,000 violent crimes committed with a gun each year. If the average number of times that an offender commits a robbery or assault with a particular gun is twice, then (assuming patterns of criminal gun use remain constant) the total number of transactions of concern is 250,000 per year.
Actually, no one knows the average number of times a specific gun is used by an offender who uses it at least once. If it is more than twice, then there are even fewer relevant transactions.
That compares with total sales volume by licensed dealers, which is upwards of 20 million per year.
And this doesn’t include private transaction in the used gun market. But it does tell us that nearly 99%, if not far more, of each year’s gun transfers between owners don’t involve violent criminal activity.
Why spend resources on investigation of activities where 99+% of the actions are harmless, and there are no motivated witnesses to the “crime”, when you could spend those same resources investigating known activities which have a 99+% chance of there being a true victim and people motived to cooperate with law enforcement?
Update: See also this post about other aspects of background checks.