Loading...
In the late 1800s, British colonial administrators in Delhi faced a genuine public health crisis: venomous cobras slithered through streets, gardens, and homes, killing dozens of residents each year. The colonial government devised what seemed like an elegant solution — a bounty program. Any person who brought a dead cobra to a government office would receive a cash reward of several annas per snake. At first, the program worked beautifully. Hunters fanned out across Delhi, killing cobras and collecting payments. Dead snake tallies climbed. Officials congratulated themselves on a policy success. But within months, something unexpected emerged. Enterprising residents of Delhi realized that hunting wild cobras was dangerous and unpredictable. Breeding them, however, was simple and safe. S...
Popular framing: A well-meaning government program was outwitted by clever locals who found a loophole, illustrating that people will always game incentives if given the chance.
Structural analysis: The program failed because it rewarded an output (dead snakes) that could be produced independently of the problem it was meant to solve (wild cobra density), while leaving no feedback mechanism to detect or correct the divergence. Canceling the program without a managed exit then converted a stock of captive cobras into a pulse of wild cobras, making the original problem worse. The two errors compound: bad entry design and bad exit design. The 'second-order error' of the British—canceling the program abruptly turned 'assets' (bred snakes) into 'liabilities,' forcing breeders to release them and making the territory worse than the starting state.
The popular framing locates agency and blame in individuals, which implies the fix is better policing of actors. The structural framing locates the failure in the system's feedback architecture, which implies the fix requires redesigning the measurement-reward loop and modeling exit dynamics before launch. Missing the structural gap leads policymakers to repeat the same class of error with new actors and new metrics.
