Incorrect U.S. Census Information—When the Census Taker Gets It Wrong

October 9, 2018  - by 

By Jan Mayer

On the first U.S. census day (August 2, 1790), 17 United States marshals and around 650 assistants1 began the task of finding and recording the population of the United States. These were the first U.S. census takers.

In 1880, specially trained census enumerators (census takers) were hired to replace the federal marshals in counting the population. A national census is taken every 10 years in the United States, and the information is then used to allocate congressional seats, electoral votes and funding for government programs. Census information is also used by businesses, community organizations, historians—and genealogists.

From the very first census, incorrect census information has been a concern. Thomas Jefferson and George Washington both speculated that the population of the country was higher than the 3.9 million counted in the 1790 census.2 Although early censuses recorded comparatively few names and gave a basic population count, more recent census records have much more information about individuals and households. As you search U.S. census records, understanding census errors can help you with your family history.

census records, census mistakes

Why Are There Mistakes in Censuses?

Most census mistakes are simply human error. Census takers risk severe penalties if they disregard confidentiality or deliberately misrepresent data. In fact, Census Bureau employees have always been required to take a nondisclosure oath and are sworn to protect the confidentiality of census data for life.

Nonetheless, inaccuracies do occur. Some of the most frequent reasons for incorrect census information include the following:

census records, census mistakes

  • A focus on counting. Counting the population has always been the main focus of the United States census, not keeping perfect historical records. In fact, census questions from past censuses may have been answered by any member of a household, a boarder, or even a neighbor who agreed to be truthful.
  • Spelling errors. Looking for ancestors, genealogists may be stumped by name spellings that vary from census to census. Some of this variation comes because many U.S. schools taught spelling by phonics (by sound) in the 1800s. Also, in 1790 only about 65 percent of the United States population could read at all, so spelling a name was up to the census taker, according to Bill Dollarhide, author and census genealogy expert. Thinking of different ways to spell or misspell a name can help you identify your ancestors despite spelling variations in the census data.3
  • Copying errors. Each set of census records has a different history of copies. Sometimes the copying process resulted in the county, state and federal governments holding separate copies, all of which may have slight variations. Genealogists usually view the copy from the National Archives and Records Administration and may not realize they can also check state and county records to see if the forms contain copy errors.
  • Missing or false information. Citizens are sometimes wary that the U.S. census is for tax collection or may dislike answering census questions. Misunderstanding can also arise from language barriers between a census taker and the person being interviewed. Especially in older censuses, people responding may also not have had precise answers for some questions. For example, Dollarhide notes that birthdays weren’t widely celebrated in the United States until the 1880s, and even parents may not have remembered exact ages for each family member.3

Today, the U.S. census is conducted initially with mailed questionnaires, which prevents many recording errors. Census records have also been partially or fully processed by machine since as early as 1872. The Census Bureau is always working on improving the enumeration and processing of future U.S. censuses.

navajo, native american census,

Can Incorrect Census Information Be Changed?

According to the United States Census Bureau, it isn’t possible to correct an error in a census record. The census records are historical documents, and historical documents are not perfect. The Census Bureau recommends the following, “Our advice to genealogists who find inaccuracies is to make a note in their family history that the census record may contain errors.”

The Census Bureau also points out that some of these errors can actually teach us about our family members. Families sometimes provided alternate or “Americanized” names, left illegitimate children out of their household count, or misidentified their racial heritage when answering census questions. These intentional differences teach us about the culture surrounding our ancestors and may help us identify missing or interesting stories in our family history.

The Value of Census Records

While U.S. Census Records are not the only resource for tracing ancestors, they are freely accessible at FamilySearch.org and also available on other genealogy sites.

Finding an ancestor in a census record can be a great start to building or extending a family tree. Although census data may not have the same level of accuracy as other genealogical records, censuses can help you discover family stories. They also contain vital clues for locating other records. With the information from one or more census records, you may be able to locate a birth, marriage, or death record for your ancestor. You also might be able to track down naturalization papers or learn where ancestors lived and traveled within the United States.

Read more about United States census records and how to use census records on the FamilySearch blog.

End Notes:

  1. “Heads of Families at the First Census,” United States Census Bureau, accessed September 27, 2018, https://www2.census.gov/prod2/decennial/documents/1790j-02.pdf
  2. “1790 Overview,” United States Census Bureau, accessed September 27, 2018, https://www.census.gov/history/www/through_the_decades/overview/1790.html
  3. Bill Dollarhide, “Census Mistakes,” Genealogy Blog, last modified April 13, 2012, https://www.genealogyblog.com/?p=18199

Leave a Reply

Your email address will not be published. Required fields are marked *

Comments

  1. I live in Mexico, in Puebla, municipio Tepexi de Rodriguez, my wife’s native village. I encounter indexing errors, because we know the families and in some cases, the individuals were or are known to my wife who is 76, and is a descendant of Emperor Moctezuma I and apparently also Cortes. (His natural son was comedore of Oaxaca and was known to do business here. He introduced silkworms here, though the King of Spain stopped it when he heard about it.) I don’t think there is much to be done about indexing errors, but the problem is for those in the future who will not have access to live witnesses as we do.

    But, within the municipio is a ‘rancho’, Moralillo. In the 1930 census, whoever collected the hand-out sheets for door to door usage, messed up the order when they came back.

    The first image (of 13 for the ranch) was 074a_004107727_00665jpg. The first persons listed are kids from the Moctezuma family, no head of household listed, which for ‘search’ puts those kids in a totally different place with a totally different family.

    At that time, Zenarina was still alive and a friend went and talked to her. Her sister was Altagracia, who is listed on image 079b_004107727_00676.jpg with the rest of the family.

    The next image after 079b is 080a_004107727_00677.jpg. The first people on this page were also kids with no head of family Comacho/Rosas listed. There is a probability this page should come after image 075b_004107727_00668.jpg but that is only a calculated guess.

    I can see how this mess occurred. Being a ranch, there were no house numbers and no streets listed so there was no way to sort them out correctly for anyone who didn’t know the ranch personally.

        1. This happened to my mother-in-law in the 1930 Census; she was in Everett, Massachusetts.

          My mother’s household is a complete mess in the 1940 Census. My widowed grandfather and his two kids lived with his never-married oldest sister, and another divorced sister and her son. The oldest sister was listed as the mother of her two siblings, and the grandmother of her three nephews and nieces. To top it off, the surname was misspelled, missing the last syllable.

  2. in 1910 census with head of household Martha Kenyon, Gussie August Kenyon age 2 6/12 years girl. this child was Not a girl, but a boy. remember in 1910 all children were usually dressed in gowns. boys might not be pants till 3 1/2 to 4 years old. told to me by an woman who had children during those days. “more likely in the backwoods.” according to her.

  3. In researching the 1940 census for Klickitat County, Washington. For the town of Glenwood, I grew up there and I found several errors on this census. My parents were listed correctly and so were my siblings. But my grand parents were listed as HATEN. The reason I knew was I am still alive and I recogniized my Aunt Dorothy, & Uncles: Henry,,Perry, & Harold. My Grandma’s name was misspelled..The other members of the family later were mentioned correctly, Roy, Genevieve, and Gerald were correct, so was Forrest and Mildred Harter and their children. the Lane family was listed as JONES.

  4. My great grandparents were listed as last name Carbenul on the 1910 census, not Cavanaugh. Fortunately, a close relative lived down the street, so I found them by accident. If you know where they may have lived it is helpful

  5. My uncle was listed as a girl in the 1930 census. Everything else was accurate. as far as my own personal knowledge. This uncle was 9 years old, so it’s not like he was an infant. And from what I know of him, he was all boy. It was just funny to see that.

  6. The first census I find my Great Grandfather, Enoc/Enoch/Enos North on was 1840 and it says he was born in Virginia..In 30 yrs, I have never found a clue to where in Virginia. He married in 1836 in Harrison County, Kentucky to Mary Ward, my Great Grandmother. I have never found any parents for either, (no proof listed anywhere for Mary Ward either.); in over 30 years since I starting searching for my Enoch North; I know of his life from his marriage to his death in 1872; I have his children’s names, their spouses, their deaths, his Pension File, his Civil War enlistment, he got “Camp Diarrhea” after being at the Camp for a few months and was never Mustered in again and suffered with it until he eventually died from it according to affadavits in his pension file; (can’t find any death records, only year, or any clue as to where he is buried). If only I could find what county in Virginia he was supposed to be born in ???. If his wife received a pension, wouldn’t Washington, DC want to know where he was buried ??

  7. In the 1900 census for Wilson County KS, my great grandmother is listed as Alice V Babcock. Her correct name was Olive Ellen Babcock who died of cancer in 1906. Unfortunately other descendants who dont know the correct name have made up a whole new person, which would mean my great grandfather had committed bigamy.

  8. The 1920 and 1930 Alabama Census are totally false. My great-grandmother is listed at ages where is was deceased. All the ancestor’s sites and Census records recorded are fabricated and you can’t go by them. No truth…