A census can be an unmatched resource for tracking families through time and space, providing an every-ten-years snapshot of ancestors no matter how often they changed address. Genealogists tend to be as familiar with historical censuses as the census departments themselves, poring over the often blurry and misspelled information contained in those pages. But where did the census come from, and why? This week we look at the origins of census-taking and the role of the census in the rise of Big Data.
The Big Count
The task of preparing, gathering the information, then processing the vast quantities of data produced by a national census are extraordinarily challenging. So why does the United States and nearly every other country in the modern world spend so much energy conducting them?
The reasons for creating a census haven’t changed much over time. The Roman Republic (509 BCE - 27 CE) was the first known society to conduct a census and the word census, unsurprisingly, is a Latin word, derived from censere, meaning to estimate. For the Romans, the purpose of the census was to create a list of men in the Republic who were eligible for military service. The census was and remains a way for the State to understand its own population.
In the winter of 1085 CE King William the Conqueror of England held a meeting with his advisors, “a very deep consultation with his council, about this land; how it was occupied, and by what sort of men. Then sent he his men over all England into each shire; commissioning them to find out ‘How many hundreds of hides [traditional measure of the landholdings of a household] were in the shire, what land the king himself had, and what stock upon the land; or, what dues he ought to have by the year from the shire.’” Thus was the Domesday (modern: Doomsday) Book created, England’s first attempt at a national census. As the The Anglo-Saxon Chronicle (a history written in the 9th century CE) narrative above reveals, the primary motivation for creating The Domesday Book was financial: how much in taxes did the subjects owe the king?
The Founders of the United States were well aware of the importance of an accurate census to governing a rapidly growing, geographically widespread population, so they mandated the production of a decennial census in Article I, Section 2 of the United States Constitution. Like William the Conqueror, taxes were one of their priorities. "Representatives and direct Taxes shall be apportioned among the several States... according to their respective Numbers,” the Constitution stipulates, highlighting one of the most important civic foundations of the newly-formed country: a representative government must have access to reliable information about the people it is representing.
The first elected official to oversee the U.S. Census in 1790 was Thomas Jefferson, then the Secretary of State to President George Washington (the first five censuses were the responsibility of the U.S. Secretary of State). In that first census, the task of gathering data was left to local census marshals; Jefferson merely certified the results. But when the numbers were tallied, both Jefferson and Washington were sure that the population estimate of 3,929,214 people was an undercount, perhaps by as many as 200,000 people. The growing nation could not afford to miss 19 percent of potentially voting and tax-paying residents, and so began the ongoing quest for a more accurate count.
Less than a century after achieving independence from Great Britain the population of the United States reached 23 million, close to double the population of its former colonial masters. The booming population was good for the American economy, but increasingly a nightmare for those enumerating the census. By the 1870s the sheer amount of census data required almost a decade to process--just in time for the taking of the next census. The federal government found itself awash in more information than it could use.
Part of the problem was the number and complexity of the census questions, which asked for information about the names and ages of people in every household and also for detailed information about landholdings (Arable? Hilly? Swampy?), among other time-consuming queries. The nature and number of census questions has changed through the years, reflecting the changing concerns of the country.
By the late 19th century the U.S. government sought outside help in solving its census problems, and held a contest rewarding anyone who could find a faster way of compiling the oceans of data. The winner was Herman Hollerith, the son of a German immigrant to Buffalo, New York, in 1888. He invented the electric tabulating machine, an early predecessor of the computer punch card. By automating the tabulation process, Hollerith reduced the time needed to process census data ten times faster than his competitors. Hollerith’s innovation continued to improve and he was hired to conduct a census of Russia as well as a multitude of companies whose business depended on accurate data processing.
Herman Hollerith had created a technological solution not just for governing the huge population of the United States, but also for harnessing the power of massive amounts of data, something that would become an absolute necessity in the 20th century. Eventually, Hollerith’s tabulation company took another name: International Business Machines, a.k.a., IBM.
Family historians have relied on the records of the U.S. Census for as long as they have been available, and they continue to be an invaluable resource. And since the advent of the Internet in the 1990s, home researchers have had the opportunity to draw not only on the hundreds of years of painstakingly tabulated census information, but also the incredible power of Big Data, something that the U.S. Census itself helped to create. So the next time you’re perusing census.gov or the deep archives of a family tree mega-web site, thank the hardworking census-takers and the clever solutions of Herman Hollerith, just two of the people who made all that data possible--and usable.
More Than Just Names, Ages and Locations of Residence: What You Should Know About Census Records
Census records are the cornerstone of most genealogical research. They help place ancestors in location and time, provide details about age, birth and occupations, add names of previously unknown relatives to the family tree, and so much more. Many make the mistake of stopping research with the obvious details included in the federal population schedules. If you are one of those people, you may very well be missing out on the opportunity to expand and enrich the details of your family history.
One of the first things you should do is to learn about the history of the area in which your ancestor lived. This includes boundary changes. Sometimes we believe that our ancestors have moved, when in reality the county or state boundaries changed. There are also instances where there has been more than one incarnation of a county, with the county being formed in location, resolved and then re-established later or in another part of the state. One of the best sources for county and state boundary changes is the Map Guide to the U.S. Federal Censuses, 1790-1920 by William Thorndale and William Dollarhide. Another valuable resource is this collection of interactive maps, which allow for toggling between years for a snapshot of boundaries for a particular state and the counties within.
Not all censuses were created equal, meaning that that did not all record the same information, or if they did, that information was recorded under different guidelines. Each year there were specific instructions provided to the enumerators, detailing what questions were to be asked and how the answers were to be recorded on the forms. These instructions might differ for like-questions for different census years. For example, the race classification sometimes changed dramatically from decade to decade and could be quite complex and confusing. Without understanding the instructions to the enumerator, improper conclusions can be drawn. In 1930, for instance, enumerators were instructed to no longer use the term “Mulatto” when classifying individuals of both White and Black lineage. Instead, the guidelines dictated that they be recorded as Black, no matter what percentage of each lineage. Similarly, a person of mixed Black and “American Indian” lineage was to be classified as Black, with a caveat that if they identified as and were accepted by the community as being “predominantly American Indian,” they could be recorded as “American Indian.” In contrast, if the person being enumerated was both White and “American Indian,” they were to be classified as “American Indian,” unless the percentage of Indian heritage was very small and they were accepted by the community as being White. As a matter of fact, any instance where the person was White and mixed with any other racial lineage, they were to be reported as the other race. If a person was of two minitory interracial linages, they were to be recorded as the race of the father. The instructions for each federal decennial census, from 1790 to 2000, are available for review on the Federal Census Bureau website.
When researchers refer to “the census,” they are normally referring to the federal census population schedules; however, there were many secondary schedules that recorded rich biographical information about our ancestors. Sometimes this information can provide more context to the lives of ancestors, but oftentimes, the details included in these secondary schedules can lead to records and information that may have never been otherwise known, allowing researchers to add new direct or collateral ancestors, extending their family trees. These secondary schedules include agriculture schedules, detailing acreage and other farm-related details; mortality schedules, which recorded individuals who had died in the year prior to the enumeration; veteran’s schedules, recorded both Revolutionary War and Union Civil War veterans; slave schedules, which were oftentimes the only documentation for the enslaved prior to Emancipation; manufacturing schedules, included information such as number of employees and other business-related details. One lesser known secondary schedule group which can often provide the most genealogically-rich information for those included, is the United States 1880 Defective, Dependent, and Delinquent Schedules. This group of schedules recorded information for the blind, “deaf and dumb,” “idiotic,” insane, “maimed, crippled, bedridden, or otherwise disabled,” or those who were enumerated in a prison, orphanage, or poorhouse, essentially enumerating the pertinent individuals twice in great detail within a single census year. The secondary schedules vary in completeness and the census years in which they were included, as well as online availability; however, all extant schedules are available through the National Archives.
For each census year, enumeration districts were established. Each district was created with the expectation that one census taker could enumerate the residents in that geographic area during the census period. As such, the divisions were based largely on population, resulting in some districts being only a few city blocks, while more rural districts could encompass an entire county. It was not altogether uncommon for there to be confusion or misunderstandings regarding the enumeration district boundaries, causing some individuals to be enumerated twice within one census year. Enumeration district maps can be helpful in locating misindexed or missing ancestors by comparing maps from previously examined years to determine the enumeration district in which they may have been enumerated in subsequent years. Enumeration District Maps are not available for all geographic locations or years; the 1940 Enumeration District maps are the most accessible and available for viewing online.
In addition to the Enumeration District Maps, the Enumeration District Descriptions can prove extremely useful in learning more about the neighborhood where ancestors lived. These textual descriptions describe the streets, institutions, apartment buildings, and the like.
Federal-level census records are not the only censuses that are beneficial to genealogy research. State conducted enumerations can provide similar information and place ancestors in geographic location and time, confirm familial relationships, as well as birth details, but they may also have recorded additional information that is not included in the United States censuses. In the instances where the federal census records have not survived, either in whole or in part, such as the 1790, 1800, 1810, and 1890 enumerations, the state census can prove invaluable, capturing information or ancestors who may have otherwise remained unknown. Many states conducted censuses and they vary by year and extant completeness. State-level enumerations can be found online, in state archives, or at the Family History Library. One state census record of special note is the 1925 Iowa State Census, which includes a plethora of information, including the actual names and birth locations of the parents of the enumerated.
Looking to the future, the 1950 United States Census will be released in April 2022 and may be the first time that many baby boomers are found in census records with their families. Learning how to utilize some of the enumeration district maps and descriptions may prove helpful, as when it is initially released to the public, there will likely be little-to-no indexing. If history repeats itself, it will take several years for the indexing to be complete, so take the time to expand your census knowledge and you just may learn more about the lives of your ancestors or even knock down a long-time brick wall. Happy sleuthing!
IBM (https://www.ibm.com/ : accessed 11 February 2021), “The Punched Card Tabulator.”
Project Gutenberg (http://www.gutenberg.org/ : accessed 11 February 2021), “The Anglo-Saxon Chronicle by J. A. Giles and J. Ingram.”
The New Yorker (https://www.newyorker.com/ : accessed 11 February 2021), “How the Census Changed America.”
Population Reference Bureau (https://www.prb.org/ : accessed 11 February 2021), “Milestones and Moments in Global Census History.”
United States Census Bureau (https://www.census.gov/ : accessed 11 February 2021), “Through the Decades.”
United States Census Bureau (https://www.census.gov/ : accessed 11 February 2021), “Director Biographies: Directors 1790-1810.”
United States Census Bureau (https://www.census.gov/ : accessed 11 February 2021), “Index of Questions: 1930.”
Wikimedia Commons, database with images (https://commons.wikimedia.org/wiki/File:1890 Census Hollerith Electrical Counting Machines Sci Amer.jpg : 11 February 2021), photo of original artwork “File:1890 Census Hollerith Electrical Counting Machines Sci Amer.jpg;” photograph uploaded by user Infrogmation.
Library of Congress, Prints and Photographs Division, digital images, Library of Congress (https://loc.gov : accessed 11 February 2021), digital image of original print from wood engraving, “Taking the census / after sketch by Thomas Wort..” Thomas Worth, artist, digital id cph 3b39850.
“U.S., Enumeration District Maps and Descriptions, 1940,” digital images, Ancestry.com (https://ancestry.com : accessed 11 February 2021), New York, New York (image 1); citing, “Enumeration District Maps, 1940,” United States Census Bureau, NARA, Records of the Bureau of the Census, 1790 - 2007, RG-29.
“Enumeration District Descriptions, 1850 - 1950,” digital images, National Archives Catalog (https://catalog.archives.gov/ : accessed 11 February 2021), 1930 San Francisco, California - ED 38-399, ED 38-400.