Handling US Census Conundrums, Inconsistencies And Discrepancies

To be sure, the US Census can breathe life into your ancestor's times. It can be so much more though. It all depends on how you handle errors, inconsistencies, and discrepancies, which are commonplace in the US Census. Handling them masterfully can solve Census conundrums.

This is particularly the case for your enumerated ancestor, who remains elusive despite advanced-level searches. Consider stepping into the shoes of your ancestor. Transforming your mindset and analysis of information from finding your ancestor to finding out about your ancestor can prove to be the key in, yes, finding your ancestor, listed in the US Census all along.

Limitations Of Database Search Engine Dependency

The "finding out" approach may seem obvious. However, the focus in many Census guides and articles on handling errors, inconsistencies, and discrepancies in Census records is less about recommending the possible viewpoint of the Census informant (perhaps, but not necessarily the enumerated ancestor) and more about presenting creative research methods. Moreover, when the informant's viewpoint is discussed, I have only noticed it either in the context of the enumerator's experience or as a brief aside. This limited treatment of the informant's perspective is unfortunate because the ancestor's real-life circumstances, street-starts, or desire for privacy are also very important to consider and understand. After all, such an understanding can make all the difference in finding your otherwise elusive ancestor, or at least make credible sense of errors, inconsistencies, and discrepancies that exist across Census years.

In "Fact Or Fiction? Understanding & Overcoming Errors In Genealogy Records," Archives.com contributor Peggy Patrick insightfully noted that social biases and stigmas apparently made its way into the Census record. She explained, "Inaccurate information was intentionally given to a census taker. There is a Native American Indian in my family, but every census record indicates she was white. The stigma of divorce sometimes caused a woman to indicate she was married, although there was no husband listed in the household." In order to ensure that such misinformation would not prevent a positive identification of ancestors, Patrick recommended that mastering (Census) databases "is primary."

Mastering a database or two may have been Patrick's secret to success, but a New York Times article entitled "Census Takers' Troubles" brings in numerous data-altering factors that show the limitations of the very database wildcard search options that Patrick recommends. The article carefully recounts the frustration and fear that an 1880 Census enumerator went through while attempting to collect information on a household in New York City. Besides detailing the enumerator's experience, though, the story also clearly paints another picture as well: one of a lady who appears to be fearful for her safety, concerned for her husband's wellbeing, and skeptical of governmental intrusion into her household's privacy. Although the lady's thoughts are purely my speculation, since the enumerator never inquired about the subject's reservations to answer the Census questionnaire, the record of her responses and actions clearly convey her thoughts.

Given these circumstances, including being threatened by the accompanying police officer with a $100 fine, how accurate (and complete) could we expect the information to be that the lady provided to the enumerator in the end? Would the lady care to ensure that the names of her husband and herself are correctly spelled and would she provide accurate information on their respective ages, places of birth, etc.? According to the enumerator, the lady only provided very basic information and was not asked if anyone else besides her husband lived at the same address. And what of the remaining 25 households, whose informants, the New York Times reported, apparently had similar reactions?

Mastering a database cannot substitute true historical analysis, which includes remembering that our enumerated ancestors (the informants?) were humans. Even if you have managed to master or even conquer a database search engine or are carefully scanning each page of a Census enumeration district, it is still all too easy to reject or disregard results that seem to conflict with everything you know or imagine being within the realm of possibilities. The (not-so-finite) realm of possibilities should remain in the forefront of your researcher mind, especially when dealing with the informant's (the enumerated ancestor's?) perspective.

The World Through The Eyes Of The Census Informant

From my experiences, both personal and for clients, the strategy that brought positive results to many cases of elusive ancestors was to rethink the Census, switching from a researcher-perspective (trying to "defeat" a database or forcing sense out of errors, inconsistencies, and discrepancies, or worse, declaring a "no find") to an "informant's perspective". Due to space restrictions, I shall discuss one very representative example.

A client commissioned me to investigate the case of an ancestral family of theirs, the Harsanyis, of Morris County, New Jersey. They had already found their immigrant ancestor Bertie Harsanyi with family in the 1930 Census (see Fig. 1 below), but could not locate him or his family in any earlier Census, even though it was known from his obituary that he had emigrated from Hungary in about 1903 and settled in Morris County. The obituary also mentioned a brother by the name of Charles, also of the area.

Many factors had prevented my client from finding Bertie (or Charles) in earlier Censuses. Not least was a difference in spelling of the last name. I found Bertie and family in the 1920 Census (Fig. 2) under the name "Horshine"; the different spelling could have resulted for numerous reasons including the enumerator's phonetic misspelling of the non-Anglicized surname, the enumerator's phonetic translation of the informant's Hungarian accent, and/or the enumerator's inattention to accurate spelling and/or struggle to spell correctly. These issues are well covered in Census guides and articles, which carefully study the errors in the Census caused by enumerators.

From the informant's perspective, though, a different story comes forward for the researcher. The informant (Bertie?) could have intentionally misspelled "Harsanyi" to Anglicize it at that very moment. The informant could have also simply confirmed the spelling as already written down, especially if as suspicious as the above-mentioned unnamed lady was of the Census enumerator.

Can these speculations be confirmed? That depends on how much we can learn about Bertie's personality, societal views, and view about government. In this case, the 1920 and 1930 Census record (Fig. 3 and 4) of Bertie's brother Charles strengthens the possibility of an "intentional misspelling"; like Bertie, he is also enumerated as "Horshine" in 1920 and "Harsanyi" in 1930. The consistency in discrepancy across the Censuses cannot be ignored.

Furthermore, Charles appears with family in the 1910 Census (Fig. 5), listed as "Hartshorn". Bertie has yet to be located, if he was in fact enumerated and his entry survives. Does this mean the Harsanyi were actually the Horshine or the Hartshorn family and that the brothers changed their last name once or even twice? That could be possible and could be discovered through further non-Census research.

In his obituary, Bertie is said to have been originally from Hungary. In the 1930 Census, his birth country is listed as "Czechoslovakia". Yet, "Gomor" with notation "Hun" (standing for "Hungary") is given in 1920 as his place of birth. Is one an error? Did Bertie not remember in which country he was born, did he actually choose to be misleading or to provide misinformation, or was he in fact telling the truth--both times? Could he have been born in Gomor, Hungary and Czechoslovakia?

We can determine the likely truth behind the information on him by analyzing his ages across Census years. He was listed as 33 in 1920 and 41 in 1930. This means he was born between 1886 and 1889. Czechoslovakia did not yet exist at that time, but it did in 1930. Hungary existed in 1930, in 1920, and for centuries earlier; its borders changed over those centuries and between 1886 and 1889, Hungary was part of the Empire of Austria-Hungary.

I do not have the maps of historical Hungary or Czechoslovakia memorized. Without doing anything else, in this instance it might be a safe bet to say that Bertie inadvertently aided our research by really telling the truth about his origins in terms of the then (1920, 1930) borders. Gomor (Gömör, Gemer) happens to be a province, which changed hands a few times.

What of his age? Ages are reasonable details to be concerned with, until we realize that informants were sometimes reserved, if not hostile, to that Census questionnaire item. An enumerator recorded in her diary: "Some people surely are afraid other folks will find out their ages."

Besides being perhaps just as guarded, imagine the challenge of an immigrant informant like Bertie remembering the word for their age in a language not their mother tongue, and then those for the rest of the household. Or maybe he mispronounced his age. On the other hand, perhaps he really did not know his age or could not immediately recollect it. Or maybe he replied one way as a joke and it was recorded as his answer. His obituary states his birth year at 1887.

Discovering Your Elusive, Enumerated Ancestor

The Harsanyi case speaks volumes about the importance of researchers to think creatively about the informant's circumstances when trying to make sense of errors, inconsistencies, and discrepancies that appear in (and across) US Census records. Real care must be taken that the information in the Census is not taken as absolute fact. Such an approach can be easily and seriously misled, creating unnecessary conundrums that even advanced searches of databases cannot untangle.

Really imagine the circumstances of the informant who provided the answers to the Census questionnaire. You might just suddenly locate your enumerated ancestor lurking in the Census, just waiting to be found.

Figures 1-5

References and Further Reading

"Census Takers' Troubles," New York Times, 5 June 1880, 10:1, abstract in Kathleen W. Hinckley, Your Guide to the Federal Census for genealogists, researchers, and family historians (Cincinnati: Betterway Books, 2002), pgs. 48-49.

"Diary of 1910 Census Enumerator Fleta M. Myers, Mt. Zion Township, Macon County, Illinois," abstract in Kathleen W. Hinckley, Your Guide to the Federal Census for genealogists, researchers, and family historians (Cincinnati: Betterway Books, 2002), 57-59.

"Genealogy.com - Name and Word Spelling", Genealogy.com, accessed 5 August 2011.

Hyman Alterman, Counting People: The Census in History (New York: Harcourt, Brace & World, Inc., 1969), 232-233.

Kathleen W. Hinckley, Your Guide to the Federal Census for genealogists, researchers, and family historians (Cincinnati: Betterway Books, 2002).

George Joynson, "You Found Them Where? Getting Around Misspelled Names", Archives.com, 2 July 2009, accessed 5 August 2011.

Peggy Patrick, "Fact Or Fiction? Understanding & Overcoming Errors In Genealogy Records", Archives.com, 24 June 2011, accessed 5 August 2011.

Craig Rice, "Even a Beginner Can Search Census Records Like an Expert", Archives.com, 27 August 2009, accessed 5 August 2011.

William Thorndale, Map Guide to the U.S. Censuses, 1790-1920 (Baltimore: Genealogical Publishing Company, 1987).