Cunning Observation: US Agricultural Statistics in the Time of Laissez-Faire

What does it mean for a small bunch of civil service statisticians to observe something as huge as the whole American agriculture? What are the eyes that can see this? We answer this question focusing on the actual practice of statistical observation at the beginning of the twentieth century. We show that aggregating heterogeneous pieces of observation is necessary— even though it might seem paradoxical—and it requires the statistician be a very cunning observer.