After another unplanned extended break - the vast majority of my time is spent chasing a tiny child with the frame of a linebacker and the speed of a running back around the house (on the plus side, my legs and arms look great) - we're starting another new series on the blog. We're going to do detailed profiles of many of the main records types you'll be using in your research to get you more comfortable with them, with the information they usually provide and the quality of that information. We're going to start with a mainstay of genealogical research: the United States Federal Census. The census, along with vital records, forms the skeleton of your American ancestors' life histories. So let's get started.
A page from the 1790 Federal Census population schedule for Tisbury, Dukes County, Massachusetts. This location was selected because Tisbury is a funny name.
What is it?
Article 1 Section 2 of the United States Constitution lays out the requirement to occasionally count the number of residents of the United States for the purposes of taxation and determining representation in the House of Representatives. It mandates that all free persons shall be counted, "excluding Indians not taxed, [but including] three fifths of all other persons." In other words, Native Americans weren't counted, and each slave was counted as three-fifths of a person. This is the famous Three-Fifths Compromise (later made obsolete by the 14th Amendment). Briefly: the South wanted to count its slaves for purposes of increased representation. The North did not want to count the slaves at all, because they were neither counted as taxpayers, citizens, or even people. This disagreement nearly caused the Constitutional Convention to break apart without an agreement. In order to save the Convention, the North and the South agreed to count each slave as three-fifths of a person, but not even this compromise would settle the animosity between these two sides for long. The first Federal Census was taken in 1790, seven years after the Constitution was ratified, and has been continued decennially (every ten years) since then.
At first, the Census Bureau stuck strictly to its mandate: early iterations of the census were little more than counting the number of people, categorized by age and gender, in a given household. As the decades passed, however, the content of the census became more and more complex. It evolved into a detailed snapshot of this country's families at a given moment in time, containing a wide range of genealogically valuable information. The census has become an essential tool for the American family historian, as it allows you to trace your family backwards in ten-year long chunks with significant reliability.
What is its availability?
With one exception, the U.S. Federal Census is readily available for the years 1790-1940. Due to privacy concerns, there is a 72-year lag in between the taking of a given census and its release to the general public. The 1950 census will become available, then, in 2022, and the 1960 census in 2032. Most people who are interested in family history, however, have easily confirmed recent ancestors who can be quickly located in the 1940 census, so while the lack of the more recent censuses for public consumption can be irritating, it's not usually a serious obstacle to the family historian. It can be a stumbling block for those who are attempting to do research on living people, such as younger adoptees searching for information on their birth parents.
The only iteration of the census that is not generally available to us is the 1890 census, whose original data was destroyed in a fire at the Commerce Department in 1921. This being well before the advent of the Google Cloud - or, apparently, the idea that having more than one copy of priceless and irreplaceable data in more than one location might be a grand idea - the 1890 census is considered a near total loss. Only around 1,200 pages containing 6,000 names survived the fire, and those few fragments are available for your perusal. Genealogists often use city directories to help bridge the resultant twenty-year gap between the 1900 census and the 1880 census, but that only helps if you know the name of any breadwinner(s) in the household at the time.
1880 Census, still Tisbury, Dukes County, Massachusetts. Notice among other additions, we have columns indicating birthplace and whether the person was literate. Also,"Tisbury" was still a funny name.
What information does it contain?
As previously stated, the early censuses contain a limited amount of information. Until the 1850 census, the population schedules (that is, the part of the census dedicated to counting the population) contained mainly the name of the head of the household, as well as the number of white males and females of different age categories, and the total number of slaves (not given ages or genders). Post-1790, the enumerators also tallied the number of household members in various categories, depending on the iteration of the census, such as the number of residents involved in agriculture, or the number currently attending school, or even the number of the insane.
The 1790-1840 population schedules are a bit more difficult to work with than the later ones. In order to make best use of them, you'll need to have a decent picture of the family make-up drawn from later censuses. For example, if the 1850 census gives you a family with three daughters aged 12, 16, and 18, and you want to confirm an 1840 census as being the same family, you'll need to make sure that they have the correct number of girls in the 0-5 years old and 5-10 years old categories. If you are convinced that it's the right family but the numbers aren't adding up, you'll have to consider other possibilities: was an extended family member staying with them at the time of the 1840 census? Or might there have been additional children who died prior to 1850?
Starting in 1850, the census became a lot more user-friendly. That was the first year that enumerators were instructed to list every member of the household separately. For the first time, we have the names and actual ages1 of each resident! That year also provided information such as (but not limited to) birthplace, marital status, and occupation. The 1860 and 1870 censuses provided much the same information. The 1880 census added a few very valuable categories to the repertoire: the relationship of each household member to the head of the house (wife, son, granddaughter, etc.), the street address of each urban household, and the birthplace of each resident's parents.
With the growing immigrant population, the next available iteration, the 1900 census, added categories related to an immigrant's year of immigration and citizenship status, as well as the number of years married to the current spouse. The 1900 and 1910 censuses also contained two categories not found in any other years: the number of children born to a mother, and the number of those children still alive at the time of the census-taking. The all-too-common discrepancy between those two numbers is a sobering testimony to the difficulty of life and the commonplaceness of loss for the vast majority of human history. The 1920 census added categories on each resident's native language and his/her parents' native language. 1930 had a one-time category asking whether the family had a radio set, and the 1940 census did something that they should have done from the beginning: it noted which member of the family was the respondent when the census-taker came to the door.2
And that brings us to a very important matter:
How reliable is it?
This is actually a more complicated question with regards to the census than it is for most other record types. The census is an original source; if you don't feel like going back and looking at the post specifically examining source types, that means that it was made at or at about the time of the event it documents, and that it isn't based on any prior sources. Original sources are always preferable to the alternative, which is derivative sources. In addition, since the census is one of our main sources of genealogical information for the past two hundred years, and since the census-taker stood at our ancestors' front door to ask them questions about themselves, we tend to think of the information in the census as trustworthy.
As usual, there's a catch. Until the 1940 census, where the actual respondent is noted for the record, the only information in the census that we can confirm as being primary information (that is, a first-hand account from a person in a position of knowledge on the subject) is the home address. That's right, out of all of the copious amounts of information contained therein, we can only implicitly trust the address because that's the only information where we know the source, which was the enumerator, standing at that aforementioned front door and noting the address for the record. Everything else must be considered secondary or indeterminate information, and here's why.
The enumerator would talk to whoever came to the door. Maybe it was the head of the household, who would be in possession of accurate information about himself and his children, sure, but he wouldn't be in possession of firsthand knowledge about his wife's birthplace or age, much less the birthplace of her parents. To make matters worse, if the nine-year-old son was home alone the day the census man came to town, then that family's enumeration would be based on the knowledge of a nine-year-old. If no one was home at all, the census-taker would ask the neighbors what they knew about the family. With the exception of the 1940 census, we have no idea who actually provided the information!
If the information about your family in the census lines up with other reliable sources about them, then you can probably safely assume that one of the adults in the family provided the information, that they did so honestly, and that the enumerator didn't make any mistakes in writing it down. If the information seems way off, then consider that the source of the information may not have been reliable.
1940 Census, West Tisbury, Dukes County, Massachusetts. Turns out Tisbury, original flavor, had almost no one left living there. Did the name finally get old?
Are there NON-population schedules?
Yes! There are four other types of census schedules available. The first of these is the Agricultural Schedule, which will be helpful if your ancestors were farmers who made a certain amount of money from their farm. This schedule is available for the years 1850-1880, and will provide information ranging from the size and cash value of the land, to the number and types of farm animals, to the amounts of various types of crops produced during the preceding year.
The next non-population schedule is the Manufacturing Schedule, which documents businesses that, like the Agricultural Schedule, were of a certain size/profitability. This schedule was produced in 1820, and in 1850-1880. Information provided includes the types of products, the types and amounts of raw materials used, the number of employees, and the monthly cost of labor. Neither the agricultural nor the manufacturing schedules have survived with the same completeness as the population schedules have, but are still worth checking to see if your ancestor's farm or business is present!
Mortality Schedules were taken in 1850-1880; these record deaths in the single year immediately preceding the census. A mortality schedule will list name, age, gender, race, marital status, birthplace, month of death, occupation, cause of death, and the length of the final illness. If you're lucky enough to have an ancestor who died in 1849, 1859, 1869, or 1879 (in between 1 June of that year and 31 May of the next), then this schedule may give you information you can't find anywhere else.
And finally, for the years 1850-1880 (are you sensing a pattern yet?), there is what's known as a Social Statistics Schedule. This one, with a single exception (more on that in a second), doesn't document specific individuals. Rather, it is a snapshot of the community. It contains information such as number of schools, students, and teachers, the names of local newspapers, the types of church denominations, the number of criminals in prison, and average wages for various occupations. The only information the social statistics schedule provides for specific individuals is in the 1880 schedule, where the names of the deaf, dumb, blind, and criminal persons were listed.
Where can I find it?
The population schedules are easy to come by. The 1790-1940 censuses, save 1890, are available for free on FamilySearch.org, and on National Archives and Records Administration computers. This full collection, plus what remains of the 1890 census, is available via subscription to Ancestry.com; the website also offers the 1880 and 1940 censuses for free. In addition, the 1940 census can be found on the Official 1940 Census Website. And finally, for those who don't care for searching via internet, both NARA and the Family History Library in Salt Lake City have the full run on microfilm.
Your best resource for finding a specific schedule, including the non-population schedules, as well as other useful resources (graphic organizers, additional articles, and fun facts) is the Search Census Records Records Online and Other Resources page offered by NARA.
In future posts for this series, we will examine other record types in detail, asking the same questions of them: what information do they provide? Are they reliable? And where can I find them?
1 When I say "actual age," please read, "the age of each person as it was assumed to be at that point in his/her life, because most people didn't actually know or care about their correct age."
2 Please note: the above descriptions are not a full accounting of each census's offerings. Such an undertaking wouldn't fit well in a single blog post.