Fable
Which ledger truly counts as burden reduction
The estuary station installed a bronze cabinet that sorts letters. The manager soon received two sets of accounts: one said the bronze cabinet handled 70% of the work itself, while the other said the manual counter's queue had only dropped by 40%. The owner frowned: the same bronze cabinet, how can the labor savings differ so much?
The craftsman explained that the copper cabinet could independently complete many small tasks: recognizing streets and alleys, attaching sorting labels, and putting ordinary letters into the corresponding bag openings. By task count, it really did take away most of the fragmented work. But the shopkeeper looked at the counter queue and found that many people were still lining up to ask about lost letters, address changes, and supplementary credentials.
The owner first asked the copper cabinet to take on more tasks. It began to force-sort even uncertain addresses, which made the short-term books look better, but returned letters increased. The counter didn't get any relief; it had to handle more error correction. Counting only how many items the machine completed masked whether the human workload had actually decreased.
The old accountant kept two sets of books. The first recorded the proportion of all decomposable tasks that the bronze cabinet completed independently; the second recorded how many requests that would originally have gone to the human counter were safely intercepted by the bronze cabinet and no longer needed a person to take over.
They also discovered that some tasks were completed automatically but never occupied the counter; some requests were answered by the bronze cabinet but still left customers uneasy, so they returned to the manual window. Both numbers matter, but they answer different questions.
Later, the courier station no longer mixed the two when reporting. The former showed how many work steps the machine had absorbed, while the latter showed how much pressure had been reduced at the manual entry point. The owner could finally judge: was the copper cabinet improving internal efficiency, or was it truly shielding the counter from visitors?
For a while, to improve the first ledger, the craftsmen broke the copper cabinet into more small actions: stamping a seal counted as one item, changing a bag opening also counted as one item. The book replacement rate rose, but the counter did not hire fewer people. Only then did the owner require every number to state its denominator.
Later the post station looked at two curves together each month. The task substitution rate helped craftsmen optimize internal processes, while the deflection rate helped the shopkeeper judge window pressure and staffing arrangements. Only by looking at both together could they avoid mistaking "the machine is busy" for "the humans really have less work." Later, before the owner cut the night shift, he first checked whether the counter really served fewer people, then whether the bronze cabinet was merely doing more internal actions. Saving on labor budget depended on the second ledger; optimizing the process depended on the first.