Science1 publisher3 min readPublished
In a game-theory model, stronger art-crime policing leads museums to cut their own guards
Frederick Chen and Rebecca Regan model museums, thieves and police as players in one game where better public recovery of stolen art lets museums economize on security by enough to raise how often heists happen.
The Scientist · Science desk
What happened
- Frederick Chen of Wake Forest and Rebecca Regan of the Federal Reserve Bank of San Francisco model museum administrators, art thieves and law enforcement as players responding to each other, in the Journal of Cultural Economics.
- Governments that invest heavily in specialized law enforcement units for art crime get higher public recovery rates, the deterrent effect the model takes as its starting point.
- Museums that expect police to track down and retrieve stolen work feel less pressure to protect it themselves, and dial back their own private security spending to save money.
- A museum may also settle on lower security if it judges that thieves face a low black-market payoff, or that a ransom after a rare theft costs less than years of expensive systems.
Compiled by The ScientistSomething wrong?How this is made
Why it matters
- decision A funder choosing between a recovery squad and prosecuting illicit buyers gets a clear ranking out of this model: buyer-side enforcement lowers the thief's expected payout and leaves the museum's own calculation untouched.
- cost The paper's fix keeps the bill with the institutions: mandatory security standards attached to a police appropriation mean museums go on paying for the guards they would otherwise have cut.
- contradiction Anyone reading the recent European thefts as evidence for the crowd-out claim has the chronology backwards, because the paper's own publication details put it years before either case.
The crowd-out result rests on one assumption about the thief's calculation: that a museum's own guards and the police's ability to get a stolen canvas back are substitutes. Fund the specialist unit, and recovery rates go up [2]. The museum, watching the same thing, buys less protection of its own [3]. Whether heists become more frequent then depends on which of those two moves is larger.
Chen said, "Because the museum lowers its guard, the ease of committing the crime can outweigh the increased threat of police detection" [4]. He also said, "Strengthening public law enforcement may perversely increase the overall frequency of heists" [5].
Both sentences are conditional. The model identifies circumstances in which the frequency rises. What decides whether those circumstances are common is the substitution rate: dollars of museum security withdrawn per dollar added to a police budget. The account of the paper does not report one [17].
The paper's own policy list separates the levers. Longer prison sentences for art thieves raise the cost of the crime directly and do not prompt museums to cut security budgets [7]. Direct enforcement against illicit buyers reduces the expected payout for thieves, again without encouraging museums to lower their guards [8]. In the model as described, the backfire belongs to the recovery channel alone [19].
The second result is about the objects. Security spending does not rise step for step with an artwork's price tag, because a famous piece is virtually impossible to fence on the open black market and its illicit resale value drops drastically relative to its worth [10]. Thieves who cannot find a private buyer steal to demand a ransom or negotiate a deal instead [11]. A museum that expects a low black-market payoff, and reckons an occasional ransom cheaper than a multimillion-dollar system maintained year after year, picks the lower security level [12].
The work is older than the thefts used to introduce it. The publication details list Frederick Chen et al, "Arts and craftiness: an economic analysis of art heists", Journal of Cultural Economics (2016), DOI 10.1007/s10824-016-9269-6 [14]. The illustrations are recent: the French Crown Jewels taken from the Louvre in late 2025, and two paintings taken in the September 2026 Renoir Museum heist [13]. Nine and ten years separate the model from those cases [16].
Chen and Regan pair the warning with a fix, combining increases in police resources with mandatory security standards so institutions cannot rely entirely on the public safety net [9]. I would treat that as a condition on how the money is written, given that the model supplies a direction and not a magnitude. Chen said, "Art heists in movies look like high-tech battles between clever thieves and unbreakable security," and called the reality "an economic game of incentives, risk management and budget trade-offs" [6].
What to watch
- An empirical test: theft counts and museum security budgets before and after a country funds a dedicated art crime unit.
- Whether the Louvre and Renoir Museum investigations turn up ransom demands, the payoff channel the model leans on.
- Any appropriation for art crime enforcement that ties the money to minimum museum security standards.