The Severed Tail

Goodhart's Law from Colonial Hanoi to the Qumran Caves

An urgent administrative ledger compiled by the French municipal health service in Hanoi on June 12, 1902, recorded a puzzling spike in sanitation operations: in a single twenty-four-hour period, local bounty hunters had delivered 15,041 severed rat tails to city collection stations. Over the preceding five years, Governor-General Paul Doumer had transformed the capital of French Indochina by laying nine miles of subterranean stone sewers beneath the European Quarter. Designed to showcase European hygiene, the cool brickwork pipes instead created a predator-free breeding superhighway for Rattus norvegicus. By the spring of 1902, brown rats were swarming up through drainpipes into colonial villas, and when cases of bubonic plague surfaced, colonial administrators mobilized to eradicate the rodents.

Initially, the administration hired Vietnamese catchers to descend into the sewer manholes, paying a fixed piece rate per rat killed. By late April 1902, kills reached 4,000 a day, but the sewer crews could not keep pace with rodent reproduction. Seeking to harness civilian labor, municipal officials announced a public bounty: any resident who killed a rat would receive 1 centime—roughly the price of a bowl of rice—for every dead rodent brought to a municipal depot.

Within a week, sanitation clerks at the depots were buried under mountains of rotting carcasses in the tropical heat. To eliminate the biohazard and speed up accounting, the municipal council revised the rule: citizens no longer needed to carry the whole rat across town. They only needed to hand over the severed tail as physical proof of a kill.

Deliveries soared immediately. During June 1902, daily tail counts climbed from 4,000 to 10,000 and then past 20,000 on June 21. Yet French health inspectors walking through Hanoi noticed a strange sight: hundreds of healthy rats scurrying across the alleys completely devoid of tails. Rather than killing the animals and destroying their future income stream, catchers were trapping live rats, docking their tails with knives to claim the 1-centime bounty, and releasing the fertile rodents back into the sewers to breed another generation of profitable tails. Worse still, when inspectors followed merchants into the suburban villages surrounding Hanoi, they discovered farming compounds where families were breeding thousands of rats in cages strictly for their tails. Out of funds, the colonial administration abolished the bounty, prompting the breeders to dump their worthless livestock into the city ditches—leaving Hanoi with a far larger rat population and an outbreak of 263 fatal plague cases in 1906.

Forty-five years later, the same proxy inversion destroyed the physical integrity of the oldest biblical manuscripts on Earth. In the spring of 1947, after Ta'amireh Bedouin shepherds discovered seven intact parchment scrolls inside pottery jars above the Wadi Qumran near the Dead Sea, archaeologists in Jerusalem sought to recover every remaining scroll from the desert caves. To incentivize Bedouin finders during the excavations of Cave 4 in 1952, Father Roland de Vaux of the École Biblique and Gerald Lankester Harding, Director of the Department of Antiquities of Jordan, established a purchasing tariff through the Bethlehem antiquities dealer Khalil Iskandar Shahin, known as Kando: the museum would pay 1 Jordanian dinar (roughly $2.80) per square centimeter of inscribed leather or papyrus.

Within months, instead of receiving intact scroll rolls as had emerged from Cave 1 in 1947, de Vaux and his team at the Palestine Archaeological Museum were handed shoeboxes containing more than 15,000 fingernail-sized scraps of parchment. Realizing that an intact scroll sold as one item fetched less money than fifty fragments sold sequentially by the square centimeter, intermediaries had taken knives and torn two-thousand-year-old Hebrew and Aramaic texts into confetti, condemning scholars to spend five decades reassembling the puzzle.

In June 1975, Charles Goodhart, a monetary economist at the Bank of England and the London School of Economics, presented a paper at the Reserve Bank of Australia in Sydney that diagnosed the flaw underlying both colonial Hanoi and Qumran Cave 4. Analyzing why British money-supply aggregates (M1 and M3) ceased to correlate with inflation the moment the Bank of England adopted them as policy targets in 1971, Goodhart identified three phases of metric collapse:

• * The Baseline Statistical Regularity: In an unmanipulated environment, a cheap, countable proxy metric (rat tails, parchment fragments, or M3 bank deposits) correlates reliably with a complex underlying goal (plague eradication, manuscript preservation, or price stability).
• * The Control Shift: An institution attaches financial rewards or penalties directly to the proxy number.
• * The Adversarial Arbitrage: Rational agents discover that maximizing the proxy metric directly is far cheaper than producing the underlying outcome, severing the link between the indicator and the goal.

As anthropologist Marilyn Strathern summarized Goodhart's 1975 finding in her 1997 audit of British university research metrics: when a measure becomes a target, it ceases to be a good measure. Once the French colonial treasury priced rat tails at 1 centime apiece in June 1902, the metric stopped measuring rodent mortality and began measuring the agricultural yield of Hanoi's illicit rat farms.