To be more specific, who talked to the census enumerator? In the U.S. Census for 1790 to 1930, the answer is: we don’t know. The enumerators were not instructed to indicate who furnished the information they recorded, so they didn’t bother to do so.
In 1940, however, the enumerators were instructed to put an X in a circle to the right of the name of the informant (or write in the name of the informant to the left of that household if the informant was not a household member). How can knowing the identity of the informant help the present-day genealogist? For one thing, it can help us assess how knowledgeable the informant likely was. If the head of household or his wife provided the information: probably fairly knowledgeable. If the no account son-in-law or a 9 year-old child was the informant: information provided is likely suspect.
So who were the informants in the 1940 census? Well, in St. Louis City, Ward 1, the informant was overwhelmingly the lady of the house: the wife is nearly always the informant. When there was no wife in the household, head of household was nearly always the informant.
I wanted to see who the informants were in rural households, so I also looked at Jefferson County, Herculaneum (populated place). And guess what? Here, at least, the informant is head of household about half the time, and wife the other half. Why, we may ask? Probably because a farmer (or other self-employed agrarian) was more likely to be home when the census taker came to call than was a city man who was likely employed in a factory, warehouse, retail shop, office building, or government office.
Moral of this story: the more you know about the U.S. Census, the more interesting it becomes!
Note: You can search or browse the 1940 census using Ancestry Library Edition (Ancestry LE) or HeritageQuest Online (HQ Online). You can use Ancestry LE at Central Library or any of our branches.
HQ Online (and many of our other reference databases) can be used from home if you have a valid Library Card and PIN.