The actual records of civilization’s first population counts have apparently not survived, but it is known that in early Babylonia, Egypt, and China the inhabitants were counted on a regular basis. There are ancient written accounts of the Greeks and Romans having taken censuses, but those tallies, too, seem to have been lost over the centuries. On the North American continent, the Spaniards led the way in census-taking, counting heads in what was then Mexico in 1577.
Since 1790, the U.S. government has taken a nationwide population count every ten years. Though never intended for genealogical purposes, the federal censuses are among the most frequently sought records for those looking for links with the past. Unique in scope and often surprisingly detailed, the census population schedules created from 1790 to 1920 are among the most used of federally-created records. Over the course of two centuries, the United States has changed significantly, and so has the census. From the six basic questions asked in the 1790 census, the scope and categories of information have changed and expanded dramatically.
Article I, Section 2, of the U.S. Constitution required that an enumeration of the people be made within three years after the first meeting of the Congress. In March 1790, after President Washington signed the first census act, Secretary of State Thomas Jefferson sent a copy of the law to each of the seventeen U.S. marshals and instructed them to appoint as many assistants as they needed to take the census.
From 1790 to 1880, census districts were aligned with existing civil divisions. The district marshals were authorized to subdivide each district into reasonable geographical segments to facilitate supervision of the enumeration. Enumeration districts were limited in size to 10,000 individuals by the Census Act of 1850, but final tallies show that the number was usually less than 6,000. In 1880, the Census Office appointed supervisors to further subdivide the districts. In that year, the average population of each of the 28,000 enumeration districts was less than 2,000.
Early censuses were essentially basic counts of inhabitants, but as the nation grew, so did the need for statistics that would reflect the characteristics of the people. The logical means for obtaining a clearer picture of the American populace was to solicit more information about individuals. In 1850, the focus of the census was radically broadened. Going far beyond the vague questions previously asked heads of households, the 1850 census enumerators were instructed to ask the age, sex, color, occupation, birthplace, and other questions regarding every individual in every household. Succeeding enumerations solicited more information; by 1920, census enumerators asked twenty-nine questions of every head of household and almost as many questions of everyone else in the residence. As W.S. Rossiter, chief clerk of the Bureau of the Census around the first part of the twentieth century, stated, “The modern census is thus the result of evolution.”
Although the Constitution, ratified in 1787, called for a census every ten years, there was no special government agency to conduct and tabulate the results of this massive survey. Until 1840, federal marshals managed the process as best they could. In 1850, the first Census Office was opened in Washington, D.C. However, it was disbanded after the 1850 census and only reestablished in time to take the census and tally the results in 1860, 1870, 1880, 1890, and 1900.
Not until 1902 was the Bureau of the Census established as a permanent bureau in the Department of the Interior. In 1903 the bureau was transferred to the Department of Commerce. The Bureau of the Census is responsible for providing statistics about the population and economy of the nation and collecting, tabulating, and publishing a wide variety of statistical data for government and private users.
Few, if any, records reveal as many details about individuals and families as do the U.S. federal censuses. The population schedules are successive “snapshots” of Americans that depict where and how they were living at particular periods in the past. Census records since 1850 suggest dates and places of birth, relationships, family origins, changes in residence, schooling, occupations, economic and citizenship status, and more.
Once home sources have been exhausted, the census is often the best starting point for genealogical research. The availability of statewide indexes for almost every census year makes them logical tools to locate individuals whose precise residence is unknown. While some inaccuracies are to be expected in census records, they still provide some of the most fascinating and useful pieces of personal history to be found in any source. If nothing else, census records are important sources for placing individuals in specific places at specific times. Additionally, information found in the census will often point to other sources critical to complete research, such as court, land, military, immigration, naturalization, and vital records.
The importance of census records does not diminish over time in any given research project. It is always wise to return to these records as discoveries are made in other sources because, as new evidence about individuals is found, some data that seemed unrelated or unimportant in a first look at the census may take on new importance.
When family, vital, or religious records are missing, census records may be the only means of documenting the events of a person’s life. Vital registration did not begin until around 1920 in many areas of the United States, and fires, floods, and other disasters have destroyed some official government records. When other documentation is missing, census records are frequently used by individuals who must prove their age or citizenship status (or that of their parents) for Social Security benefits, insurance, passports, and other important reasons.
Problems Created at the Time the Census Was Taken
When evaluating any source, it is always wise to consider how, when, and under what conditions the record was made. By examining how enumerations were conducted, it becomes easy to understand why some individuals cannot be found in the census schedules or in the indexes to them.
From the first enumeration in 1790 to the most recent in 2000, the government has experienced difficulties in gathering the precise information it desired for a number of reasons. At least one of the problems experienced in extracting information from individuals for the first census continues to vex officials today: There were and still are many people who simply do not trust the government’s motives. Many citizens have worried that their answers to census questions might be used against them, particularly regarding issues related to taxation, military service, and immigration. Some have simply refused to answer enumerators’ questions; others have lied.
Boundaries
In the days before regular mail service, government representatives conducted door-to-door canvasses of their appointed districts. Supervisors subdivided districts using existing local boundaries. The town, township, military district, ward, and precinct most often constituted one or more enumeration districts. Boundaries of towns and other minor civil divisions, and in some cases of counties, were ill defined, so enumerators were frequently uncertain whether a family resided in their or an adjoining district. For this reason, it is not unusual to find individuals and families listed twice in the census and others missed entirely.
Over the years, state, county, township, and city ward boundaries have changed. Any census search can be thrown off by these changes and inconsistencies. Maps of boundaries changes,
For a better sense of how census takers carried out their duties in a given year, it is useful to imagine the landscape and the modes of travel available in the specific time period. In the earliest census years, travel was obviously more difficult and, sometimes, very dangerous—conditions that did not improve for decades in the more rural states and territories of the “Wild West.”
To complicate the situation further, a large portion of the young nation’s population lived in small villages and isolated farms that were dispersed over a large area. It was not uncommon for a census enumerator to make a long trip to a remote farm, only to find no one at home. In these instances, he was left to make a decision—whether to try again on another day, or question farm or household help, neighbors, or even young children. The latter appears to have been an option taken by many. In some situations, enumerators probably found it easier to guess themselves.
Obtaining answers directly from the head of household or an adult in the house was no guarantee of accuracy. For a number of reasons, ages are always suspect in census records. Many people tend to be secretive about their age; women may have been particularly sensitive about revealing the truth. One woman tracked in the census taken in New York from 1850 to 1880 claimed to have aged only twelve years in the thirty-year period. According to the 1850 and 1860 censuses of Springfield, Illinois, Mary, wife of Abraham Lincoln, aged only seven years in the ten-year period. She, or someone reporting for her, claimed that she was twenty-eight in 1850 and only thirty-five in 1860. Dozens of cases have been similarly noted; undoubtedly, some honestly could not remember how old they were. If a person’s age was not exactly known, it was frequently rounded off to the closest decade, making ages reported as thirty, forty, fifty, and so on somewhat suspect. Therefore, unless an age reported in the census can be corroborated with another source, it should not be considered totally reliable.
When questions were answered by someone other than the subject of the inquiry, the likelihood of error increased. A husband or wife might not always know the birthplaces of a spouse’s parents. A child being quizzed might easily be unsure of the birthplaces of his or her parents. Census schedules do not tell us who may have answered the enumerator’s questions.
An important point to remember is that enumerators simply wrote down the responses given to them. They were not authorized to request any kind of proof, such as birth, marriage or property ownership records. However, every individual contacted by a government representative was required by law to answer truthfully. Anyone refusing to answer or willfully providing false information was guilty of a misdemeanor and subject to a fine. As early as 1790, offenders were fined twenty dollars, to be split between the marshals’ assistants and the government. But relatively few individuals were hauled into court for refusing to answer or for not answering truthfully. It would have been an impossible task for the government to follow through and to investigate everyone’s answers.
It was not until 1830 that the census office supplied printed questionnaires or “schedules.” The enumerators of the 1790, 1800, 1810, and 1820 censuses returned the results of their canvassing on whatever paper they had. Each also had to post copies of their censuses in two public places in their assigned areas. Presumably, people who could read would see discrepancies or omissions and call them to the attention of officials. Unfortunately few, if any, of these duplicates have survived.
Another factor that comes into play in the accuracy of every census record is the competency of the enumerator who recorded the information. Individuals were not necessarily well educated or qualified for the job, and anyone who has studied census records knows that good penmanship was not a requirement. Census takers were political appointees who were frequently chosen because they were of the correct political affiliation in a particular time and place, or just knew the right people.
Wages were definitely not an incentive for would-be census takers. In 1790, even the highest pay rate, one dollar for fifty persons, barely covered an enumerator’s expenses. In 1920, payment was on a per-capita or per-diem basis—sometimes a combination of the two. An enumerator was paid between one and four cents per person, depending on the urban or rural setting of the district to be counted.
The United States has always been home to a large number of immigrants, and those who did not speak English well presented still another problem for the census taker. Often, enumerators could hardly understand the information given to them by people with foreign accents. Names were frequently misunderstood and misspelled by enumerators to the extent that they may not even begin with the correct letters, making them hard to find in census schedules and almost impossible to find in indexes. The German name Pfeiffer could easily be heard and committed to paper as Fifer, for example. An Irish census taker in Cleveland recorded the Polish name Menkalski as McKalsky in the 1920 census. Places of birth may have been equally difficult to translate into English.
Whether recording information from a foreign-born or American-born individual, some enumerators took the quickest way to get the job done. Some used initials rather than given names, some used nicknames, and some omitted places of birth, value of real estate, occupations, and other details. In boarding houses, hotels, and clusters of workers’ cottages, enumerators could easily overlook entire families.
While enumerators were given basic instructions as early as 1820, it was not until 1850 that the Census Office printed uniform instructions for the enumerators, explaining their responsibilities, procedures, the specifics of completing the schedules, and the intent of each question asked. In 1850, the Census Office also provided enumerators with a large portfolio to accommodate the oversize forms (which measured twelve by eighteen inches), pens, a portable ink stand, ink, and blotting paper. Enumerators were instructed not to fold the pages and not to allow anyone to “meddle with [their] papers.” Pages were numbered consecutively as they were completed. Each page was dated on the day it was begun, even if it was not completed until another day. Every page was to include the enumerator’s signature, the name of the civil division, county, and state and, after 1870, the local post office.
According to the 1850 census instructions, the enumerator, on completing the entry for each family, farm or shop, was to read the information back to the person interrogated so that errors could be corrected immediately. But if an informant was unclear or incorrect in giving information in the first place, this procedure did little to correct errors. A significant portion of the American population could not read nor write in the last century, so, if an enumerator misspelled the family surname, it could easily have stayed that way, whether or not it was repeated by the enumerator.
As the enumeration of each subdistrict was completed, the enumerator was to make two copies that were to be carefully compared to the original for accuracy. Hand copying, of course, brings the very strong possibility of mistakes that are often unknown to the creator of the copy. Experience with the various copies of the census shows that most copies were not error free. It was cumbersome and tedious to copy names and endless columns of personal information. It is unlikely that enumerators envisioned the copies ever being read again once the statistical tabulations were completed, so it is easy to believe that many became careless as the job wore on.
As the process was completed, the enumerator was to signeach page of the census and, at the end of each set of copies, to certify that the census had been taken and copied according to instructions. One set was to be filed with the clerk of the county court, and the other two were to be forwarded to the supervisor. As the supervisor received the completed schedules, it was his or her duty to see that every part of the district had been visited and that the copies were in good order. One set was then sent to the state or territory, and the other was forwarded to the U.S. Census Office for statistical analysis. Unfortunately, it is almost impossible to distinguish the original census taken by the enumerator—the one likely to be most accurate—from the copies, which were prone to additional inaccuracies due to mistakes in the copying process. While it is usually not possible to know if the original census or a copy was sent, it is relatively easy to recognize the census that was sent to the Census Office. “Researchers can distinguish the latter set from the other two because the Census Office made tabulations directly on the schedules; consequently, the central office copy bears pencil, crayon, and red ink markings on virtually every page.”
In 1880, the procedure of making three sets of returns was abandoned. Enumerators forwarded the originals to the Census Office and did not make any copies. In an attempt to correct errors, however, the Census Act of 1880 called for “public exhibition of the population returns,” and for this purpose it authorized enumerators to make a list of names with age, sex, and color of all persons enumerated in the district, to be filed with the clerk of the county court. If any of these lists have survived, they will be found at the county level.
In most enumerations, census takers were instructed to number each dwelling consecutively in the order of visit, though it was not always clear how the instructions may have changed or been interpreted from year to year. It should be emphasized that there was no connection between the household numbers (usually the number listed in the first column to the left of the census page) and the locality or address.
Census instructions defined a dwelling as any structure in which a person was living, including a room above a store, warehouse, or a factory or a wigwam on the outskirts of a settlement. Institutions, such as hospitals, orphanages, poorhouses, garrisons, asylums, and jails, were counted as single-dwelling houses. It was not until the 1880 census that the character and name of the institution were required to be written in the margin. The 1880 census was also the first to include street addresses in cities.
In most years, census instructions stated that all persons temporarily absent on a journey or visit were to be counted with the rest of their family in their usual abode. However, children away at school and living near the school or college were to be enumerated with that family or institution. According to the instructions, “seafaring men” were to be reported at their homes on land, no matter how long their absence, if they were believed to be still alive. Sailors residing in boardinghouses were not to be counted there but rather at their permanent residences, if they had any. Expressmen, canalmen, railroad employees, and others engaged in transportation were to be enumerated with their families if they returned to their homes at regular intervals.
According to most authorities, the 1790 census schedules for Delaware,Georgia,Kentucky,New Jersey,Tennessee, and Virginia were burned during the War of 1812. Some records, such as the 1790 records for Virginia, have been reconstructed from state enumerations and tax lists. In later enumerations, city blocks, neighborhoods, townships and sections of townships, and even entire counties are known to be missing from the census schedules, simply because no census was taken in the particular area in a given year or because they were lost before they reached Washington, D.C.
Probably the most noted loss of the federal enumerations is that of the 1890 census. Most of the 1890 schedules were destroyed in a fire in the Commerce Department in 1921.
Another confusing situation can arise in census research when names show up in a district where they do not belong—sometimes more than once! According to Arlene Eakle, Ph.D., “padding the totes,” or adjusting the census for political reasons, was not uncommon. “Frontier areas, anxious for statehood, often added bogus names. In 1857, seven counties in Minnesota had wild population totals, complete with fake names on the schedules. Jurisdictions facing increased taxes might also understate their populations to keep overall per capita taxes lower. The 1880 Utah census juggled households to disguise polygamy at a time when federal officials were seeking evidence for the prosecution of those convicted of unlawful cohabitation.”
Bogus entries may have been a frustration in some times and places, but a far greater problem in every census year has been that of undercounting. Whether families or individuals were not counted because they lived in remote areas or because they would not tolerate an enumerator’s personal questions, millions have been missed since official government census-taking began. While no stone should remain unturned in the search for an individual in the census, the unfortunate truth is that a significant portion of the population has been missed entirely.
Probably no other factor causes more frustration for a researcher than finding the general area in which an individual or family should be found in the census and then not being able to read the page or pages of interest. Often, worn and torn pages, faded or smeared ink, and the disintegrating paper of the original census are to blame. Most frequently, however, poor microfilming techniques caused unfocused and blurred sections, overexposed and underexposed pages, and words to be obscured because of tightly bound volumes or mending tape.
Microfilming of federal census records took place in the 1940s, when the technology was in its infancy and techniques had not yet been perfected. Because of the poor quality of the original microfilming, some of the 1850, 1860, and 1870 schedules were microfilmed a second time. The versions can be distinguished because the earlier microfilming included two pages to a frame, the newer having only one census page per frame. Unfortunately, the original census schedules for 1900, 1910, and 1920 have been destroyed by acts of Congress, so records that are not legible cannot be re-microfilmed.
The quality of microfilms may vary from one copy to another. Generally, the original microfilm will be better than later generations of the same. Census microfilms have been duplicated a number of times in order to make the records available to as many researchers as possible. In some cases, the National Archives in Washington, D.C., may have the best copy. Images online are made from these microfilms.
Poor penmanship, archaic handwriting styles, and symbols are other leading causes of researchers’ inability to find or read specific names or information in census records. Many letters can be misinterpreted unless a study is made of the enumerator’s handwriting style. For example, uppercase letters L and S are frequently difficult to distinguish. In one district of the 1850 census, the word “lawyer” looks more like “sawyer.” Likewise, a birthplace of Missouri might look more like “Mifouri” or “Mipouri” to someone unfamiliar with the long s character that took the place of a double s in some manuscripts.
Despite the many imperfections of the census, it should again be emphasized that census records are one of the first sources used in almost every genealogical project. They are invaluable for placing an individual in a particular time and place and for connecting the individual to other sources. The foregoing descriptions make it fairly easy to see why census records are not perfect or entirely reliable. But, as noted author Val Greenwood suggests, “no research on an American genealogical problem after the beginning of census is complete until all pertinent census schedules have been searched.”
Because most censuses must be searched on microfilm or Online, below are some suggestion to save time.
Become familiar with the surnames in your area so that you can recognize them with only half of their letters distinct. Study a county history, a tax list, or a landowner’s atlas.
Create a “pony” from the actual entries in the census. How does the writer make a, h, s, p, j, and other letters which could be misinterpreted? Draft an alphabet with uppercase and lowercase letters for comparison. An easy way is to slip a piece of plain paper onto the viewing surface and trace the letters from the page.
Use a reader in a darkened room, with a strong light to project the image. Slip a colored piece of paper—pink, yellow, and green are effective—onto the viewing surface.
Copy the microfilmed page, enlarging or reducing the image to make it clearer and sharper. Many microfilm copiers have interchangeable lenses.
Review the whole schedule so that you don’t miss important entries that appear out of place. Record all columns for each entry, even if the information seems unimportant, and record all members of the household whether they are familiar or not. In multiple-family dwellings, record all family units living in the building. These families are often related, especially in immigrant settlement areas.
Copy the data exactly as it appears in the record. If the given name is abbreviated, copy the abbreviated form. Do not expand it. If the entry is crossed through or changed, copy the entry, the cross-through line, and the changes. Note carefully the last entry on each page. Family units may be split between two pages without a repeat of the surname.
Use finding tools and indexes to get into the census quickly, then search the census carefully to get all the data it contains. If all the data is available, it is possible to block out the pedigree for several generations from this source alone. Then, proof can be sought in other records to insure that names in the pedigree really belong there. If you are researching a common name like Brown or Foster, the censuses can help eliminate those that do not fit, making searches in other sources less time consuming.
Professional researchers recommend that photocopies be made of census pages (microfilm) or save images to your computer (online) whenever possible. The advantages of photocopies over transcriptions are several: the possibility of mistakes being made in the transcription process are eliminated; a photocopy will include neighbors and provide an overview of the population makeup of the area; and a photocopy makes it easy to go back and reevaluate information as new discoveries are made in the research process.
While information in the census may be quite accurate, at times the order in which data has been entered can be misleading. For example, a head of household recorded in 1820, 1830, or 1840 may not be the oldest person in the house. With only age ranges to distinguish, it is impossible to know who may be a grandparent, a younger brother, or a man with both parents still living at home. Individuals listed in early censuses in any age grouping could be servants, visitors, or boarders not related to the family. Even in 1880 and later, the relationships noted apply to heads of household only. Children listed as sons and daughters of the head of household may be unrelated to the wife.
To protect the privacy of living individuals, access to population schedules is restricted for seventy-two years after the census is taken, so they are not available to researchers during that time. The Personal Service Branch, Bureau of the Census, P.O. Box 1545, Jeffersonville, IN 47131, will provide, for a fee, official transcripts of census records from 1940 to 2000. Access is restricted to whomever the information is about, their authorized representatives, or, in the case of deceased persons, their heirs or administrators. Use Form BC-600 to request information.
(Baltimore: Genealogical Publishing Co., 1987).
