(Originally published in Greene Speak, January 2007. Updated February 2023.)


The decade-by-decade details that have been cataloged by United States Census takers since 1790, culminate in one of the most research-rich and personally insightful record sets regarding the everyday existence of our ancestors and communities. Available for public perusal for years 1790 to 1950 (excepting the damaged 1890 entries), census records indicate: where and with whom our relatives lived; when and where they were born; how they earned a living; the languages spoken at home; the values of their real and personal property; and, all of this for each of their neighbors too. As incredible as this information is, the thing that is really exceptional about the Census, is that it goes a step further – a step taken when the Census taker walked through our ancestors’ doors and into their homes.

1870 Census

1870 U.S. census, Jackson Township, Greene County, Pennsylvania, population schedule, Rogersville post office, page 260 (stamped)/2 (written), dwelling 17, family 17, William Staggers household; digital images, Ancestry.com (https://www.ancestry.com/search/collections/7163/ : accessed 4 June 2013); National Archives microfilm publication M593, roll 1348.


The Census asks for a particularly family-oriented set of data unlike most other records. By viewing household members and their individual details we gain insight into the day-to-day home operations, such as: which son or daughter cared for the aging parents; which aunt or uncle took in orphaned nieces and nephews; how many children could attend school vs. how many were needed to work the farm, and so on. Additionally, the Census has a certain lack of formality that allows a more personal element of everyday family life to present itself. For example, the misspelling of a surname may be a clue to how our ancestor actually pronounced it when he or she spoke it to the Census taker who wrote it down. Likewise, an unusual variation of a common first name may indicate a nickname used by the family at home. Furthermore, educational and occupational information let us know who in the home could read and write, and what role each person played in household labor.


One challenge to Census research can be simply finding who you are looking for. That personal touch and lack of formality that makes the Census so unique can also make it more challenging to search. Aside from the frequent spelling errors, there are always those ancestors who decided not to live where you expected them to be, leaving you with a whole nation of possibilities.


Advice to conducting a successful search of the Census can be reiterated to apply to almost any record genealogists consult. Most particularly, you have to keep an open mind and you must look at the original record. Be prepared to accept entries that are not exact and be sure to consider every possibility. Don’t ever assume that “this is not my family” because a name, age, birthplace, etc. is not just right. Many ancestors could not read or write and even if they could, someone else often did the paperwork, such as the courthouse clerk, lawyer or, in this case, the Census taker.


Consider the source. There is no way to know who answered the Census taker’s questions when each house call was made. It could have been one of the eight children, Grandma who spoke broken English, or worse, a neighbor down the road. How many members of your household today could provide correctly spelled names and accurate birthdays and birthplaces of everyone living in your home? Not to mention, how many children Mom had given birth to and how many were still living, as was asked in 1900 and 1910? Or how old Dad and Mom were when each was first married, as was asked in 1930? Even if the head of the house was home to do the interview, it is impossible to know how meticulously he or she responded. Perhaps the family Bible was pulled for accuracy, or maybe the cow was milked between rushed answers.


Researchers seem particularly negligent about viewing the original document when it comes to the Census because so many records have been transcribed and published for us. If you click on an Ancestry.com hint to a Census record, for example, you will see a typed list of household members. Do you follow through and click on the original image of the actual record? If you don’t, you are missing columns of valuable details not included in the typed abstract and you are trusting the abstractor’s interpretation of the original record, which is a thing you need to decide for yourself. While transcripts and abstracts are excellent aides and provide many research leads, they are not a replacement for the real thing.


Consider a researcher who was frustrated that George Cumberledge could not be located in 1810. A closer look revealed that Cumberledge had been misspelled and then mis-transcribed. The original document shows “George Combridge” enumerated in Wayne Township, Greene County, Pennsylvania on page 15 (stamped).[1] A popular 1976 publication, Pennsylvania 1810 Census Index, by Accelerated Indexing Systems, Inc. (AIS), shows him as “George Cornbridge.” Though otherwise correctly referenced by AIS, this combination of misspelling and mis-transcription made George challenging to find. It was necessary to take a more liberal look at the spelling possibilities and return to the real document to make an independent assessment. Sometimes a familiarity with local names and families can help to read a sloppy or unusual entry.


Similarly, I was having no luck exploring a name index for the family of Harriet Eldora “Dora” (Mitchener) Baily in 1920. I tried searches for every potential member of the household with no result: Soundex, various spellings, nicknames – nothing. Finally, I went to the original Census and scrolled page by page through Cumberland Township, Greene County, Pennsylvania, until a suspiciously similar family appeared under the Gwyne surname. Ultimately, the census taker made an error when recording this Baily household. Instead of writing Baily into the surname field, he continued to use his mark quoting the Gwyne surname from the previous neighbor’s home. Consequently, the Baily family members are listed as Gwynes.[2] To make matters worse the Gwyne surname, a misspelling of Gwynne to begin with, was poorly written and has been transcribed in at least one interpretation as “Guyme.”


Know that the Census record, like any other piece of evidence in your search, does not stand alone to establish proof of your family history. Each fact should be compared with other records for support or contradiction. In its own right, however, the Census establishes a foundation as you rebuild the life story of an ancestor or reconstruct the social landscape of a community. There are few records that so reliably follow our ancestors from birth to death at regular intervals throughout their lives, while capturing so much detail about their lifestyles.




[1] George Combridge household, 1810 U.S. census, population schedule, Wayne Township, Greene County, Pennsylvania, page 15 (stamped); National Archives microfilm publication M252, roll 49.

[2] Dora Gwyne household, 1920 U.S. census, population schedule, Cumberland Township, Greene County, Pennsylvania, ED 125, SD 21, page 56B (stamped)/5 (written), dwelling 120, family 121; National Archives microfilm publication T625, roll 1573.