Marshals finish counting the nation: 3,929,214 Americans, give or take
The Constitution demanded an "actual Enumeration" to divide seats in the new House of Representatives, and no government had ever attempted a count of this scale. Federal marshals rode out across the thirteen states (with Maine enumerated as part of Massachusetts) and the Southwest Territory, visiting every household as of August 2, 1790, and posting the tallies in two of the most public places in each jurisdiction for neighbors to check. Nine months later the count stood at 3,929,214 people. Both President Washington and Secretary of State Jefferson suspected the true figure ran higher.
This was the first working proof of the Constitution's apportionment clause, the number that would decide how many seats each state got in Congress. It set the template every decennial census since has followed. It also recorded a 17.8 percent slave-to-free ratio, the highest of any census the nation would ever take.
QWhat exactly did a marshal write down for each household in 1790?
Six pieces of information, no more. A marshal recorded the name of the head of family, then tallied free white males 16 and over, free white males under 16, free white females of any age, all other free persons, and slaves. That was the entire form. There was no question about occupation, birthplace, literacy, or income. Those categories appeared only in later censuses. The two age brackets for white males existed for a specific reason: Congress wanted a rough count of men old enough for militia duty. Everyone else in a household got sorted into just three broad buckets. Compared to the modern census, with its dozens of questions, the first count was almost skeletal, built to answer one question only: how many seats does each state get.
QWhy did Thomas Jefferson's own department end up publishing numbers he didn't trust?
As Secretary of State, Jefferson was the official who released the preliminary tables in 1791, meaning the same man who suspected an undercount was also the government's messenger for the count he doubted. His skepticism wasn't idle grumbling either; Jefferson had his own rough estimates of American population growth from correspondence and prior reasoning, and the marshals' final figure fell short of what he expected. He aired those doubts even as his department kept processing and forwarding the schedules for publication. Washington shared the suspicion but signed off on the same numbers for apportionment purposes anyway. The finished report didn't appear until 1793, two years after Jefferson's preliminary tables went out.
QHow did the 1790 census treat free Black Americans and Native Americans differently from slaves?
The form lumped free Black Americans in with a single catch-all category, 'all other free persons,' separate entirely from the slave count, meaning free people of color were tallied but never named or broken out on their own line the way slaveholders' households were. Native Americans fared differently still: those living on tribal lands outside state jurisdiction weren't enumerated at all, since the census only reached households the marshals could physically visit within organized districts. The six-question form simply had no mechanism for populations living beyond federal or state administrative reach. That gap wasn't an oversight so much as a boundary built into the law itself, one that later censuses redrew repeatedly as the country's borders and definitions of citizenship shifted.
