Why Is My White Ancestor Listed as Black in the Census?

Tracing Your Roots: A family photo backs up lore about an ancestor’s race, but the paper trail does not.

Eddie and John Dow
Eddie and John Dow Courtesy of Amber Taylor

Dear Professor Gates:

Growing up, I saw a picture of my great-great-grandparents, and it has fascinated me. My great-great-grandfather John Dow was a white man born in Arkansas. He left Arkansas around age 11 in search of his father, Samuel Ward, in Mississippi. He eventually married Eddie Smith, a black woman, and had 10 children. I can’t find a record of them or their kids until the 1940 census, where he’s listed as Negro. He was clearly a white man, as his picture shows. We have come up with a theory that my great-great-grandmother may have answered the door and the census worker assumed that he was black, but we really don’t know. Their marriage was not a secret to anyone in Jackson, Miss. In fact, he opened an RC Cola stand for an African-American neighborhood.

Do you know why he would have been listed as “Negro”? I’m trying to trace my roots and found that John Dow was born around Jonesboro, Ark., and that we’re distant cousins to the family that owns Dow Chemical Co. What is the best way to get started? —Amber Taylor

It’s not uncommon for information to be documented incorrectly in census records, particularly when it comes to race. As Professor Gates and researcher Anna Todd of the New England Historic Genealogical Society wrote in a previous column, “The status of a person listed in the federal census (black, white or mulatto) was ultimately the personal interpretation of the census taker, based on assumptions made regarding skin color and other aspects of an individual’s appearance, regardless of what the occupant of the home told her or him. Therefore, one can’t necessarily infer parentage, complexion or much else based on that designation in a census record.”

Alternately, someone other than John Dow in the household might have provided information that was not completely accurate; or it is possible that the person providing the information did not mention that he was white and the census taker just assumed that everyone in the household was black.

Of course, there is the possibility that he did, in fact, have African ancestry and chose not to “pass,” or identify as white, despite his European appearance. During much of the slavery era, the law of hypodescent, or the infamous “one-drop rule,” meant that any amount of African ancestry, no matter how small, was enough for an individual to be identified as black. That law was really racist pseudoscience designed to maintain children born to black enslaved women and fathered by white men (often their masters) as property. Typically, the condition of the child of a slave mother followed the condition of the mother, no matter what the child looked like phenotypically. Even after slavery, this tradition of hypodescent persisted.

Any of these scenarios could be what happened in the 1940 U.S. census, where your ancestor John Dow, whom you believe to have been white, was recorded as “Negro.”

So Which Scenario Is the Correct One?

We searched for earlier records of the family to see how John Dow’s race is recorded over time, and we saw a pattern. We noted via Ancestry.com (subscription required) that John and Eddie Dow were recorded in the 1927 city directory for Jackson, Miss. In the directory, the couple were recorded as colored, marked by a “(c),” and were living at 1815 W. Lynch. The city directory provided a clue for how to find even more records of the family: They were recorded at the same address in the 1927 city directory and in the 1940 census, meaning that they would have been living in the same place during the enumeration of the 1930 U.S. census.

In that census, the family’s surname was recorded as “Daw” and John’s age was recorded as 52, placing his birth about 1872. Once again, John Dow’s race was recorded as “Negro,” as were those of the rest of the family. In addition to Willie, James, Joseph, Catherine and Robert, who were also listed in the 1940 census, this earlier census includes three other children: Margaret, born about 1914; John, born about 1916; and Laura L., born about 1920. If you are having trouble working even further back in time searching for John Dow, you could try searching for his older children, who should be recorded in the 1920 U.S. census.