The more successful a government is at catching illegal arms, the more the data suggests the problem is solved.
April 26, 2026
Original Paper
Enforcement-Generated Selection in Arms Trafficking Data: A Measurement Diagnostic for Compliance Monitoring
SSRN · 6620840
The Takeaway
Enforcement data on smuggling creates a selection effect that makes heavily policed items seem like a smaller threat. When police catch more shipments, the remaining traffickers become more secretive and sophisticated. This leads to a false impression that policy intervention is working and the market is shrinking. Security agencies end up with a dangerous blind spot because their own success hides the evolution of the problem. This means that the data used to judge safety is often telling the exact opposite of the truth.
From the abstract
When enforcement pressures shape who gets caught, and not just how many, enforcement data may reverse rather than attenuate the observed relationship between policy and outcome. Using the iTrace weapons database, this paper documents this risk in an Iraq-centered setting where all within-country identifying variation comes from Iraq. Embargo-exposed items show greater supply-chain complexity (+0.684 waypoints) but lower observed opacity (−0.140). Restricting to well-traced items (pooled weapons