 Welcome to the National Archives and Records Administration's 2022 Genealogy Series. We are pleased to present this educational series of lectures on how to conduct family research using the 1950 Census. My name is Andrea Matney, and as the program's coordinator, please allow me to do a quick introduction and provide instructions on how to participate. These lectures will demonstrate how to use records from the 1950 Census and other federal resources for genealogical research. Our presenters include experts from the National Archives and Records Administration and the U.S. Census Bureau. Sessions are intended for beginners to experienced family historians. All are welcome. We invite you to join the conversation. Please participate with the presenters and other family historians during each session's premiere. You can ask questions via chat by first logging into YouTube. Keep your eye on the chat during the broadcast because the speaker will answer your questions there in the chat. Type your questions in at any time, but please keep your questions on today's topic. In addition, find live captioning, handouts, and the events evaluation form under the video box by clicking on Show More. Note on our schedule that all sessions are broadcast on Wednesdays at 1 p.m. Eastern. And yes, all of the recorded videos will remain available for viewing later and at any time. Welcome to today's program from Parchments to Printouts, History of the Census from 1790 to 1950. I'm so pleased to introduce today's speaker, Sharon Tossi-Lacy. Dr. Tossi-Lacy has served as the Chief Historian for the U.S. Census Bureau since June 2015. Previously, she spent 25 years as an officer in the U.S. Army and U.S. Army Reserves before retiring in 2015 as a Lieutenant Colonel. She holds a B.S. in Engineering, Psychology from the United States Military Academy, an M.S. in Education from Long Island University, and a Ph.D. in History from the University of Leeds. And now turning the program over to Dr. Tossi-Lacy. Thank you. And thank you for having me here today. I'm delighted to talk to you all about just a brief history of the Census Bureau as we've evolved and changed along with our nation. We can divide Census Bureau history into several distinct errors. And as we go along, you will see how these are parallels to what was going on in the United States at the time. As you know, when dealing with historical material is always best to understand that what you will see and read must be taken in the historical context in which it was written. So why do we even take a census? Well, first of all, it's to ensure representative and proportional representation in Congress and we've done a reapportionment after every census except 1920. The House is capped at 435 members since 1911, and that was actually put into law via the Reapportionment Act of 1929. We did have a temporary increase in 1959 when Alaska and Hawaii were admitted when we increased the House of Representatives to 437. But after the 1960 census, it went back to 435. And there's nothing mysterious or mythical about that number. It is merely the number of seats that can comfortably fit on the House floor. In 1790, only the names of the heads of households were recorded and the demographics were divided into free white males over the age of 16, free white males under the age of 16, free white females, all other free people and slaves with no age or gender changes. Indians not tax were counted and those were simply American Indians who lived in tribal relations or on tribal lands, which were considered their own sovereign nations. Indentured servants were counted under free persons and only aggregate data was compiled. There was no statistical analysis. The reason that we counted free white males over 16 and free white males under 16 was simply so that the federal government knew the number of men available for militias. In 1800, the questionnaire provided space to tally free white males and females in several age categories, under 10, 10 but under 15, 16 but under 25, 25 but under 45, and over 45. American Indians, slaves and free blacks were listed in single categories, undivided into age groups. And here we started seeing that the federal government realized that the census could provide the much more data than simply a count of people. In 1810, we began counting a census of manufacturers. However, that data was so erratic, it was generally considered useless, except to identify broad industrial trends. Unfortunately, many of these early census schedules were destroyed by the British during the war of 1812. And interestingly, the only president for whom we do not have an original census schedule is George Washington, since his census record in Virginia was destroyed. And he died before the 1800 census. In 1820, we added questions on the number of household members working in agriculture, manufacturing and commerce. And this was the first time all races were tallyed by age groups. However, the manufacturing data was still not as useful as hoped. So in 1830, we did not collect that. However, we did begin collecting information on the blind, which was the first time we saw the census used for social statistics and determination of where money would go. In 1840, we once again started collecting data on manufacturing and industry employment data. And because the government needed to be able to estimate the cost of paying pensioners their pensions, we did collect data on revolutionary war pensioners. We had separate agricultural schedules and a schedule is simply another word for questionnaires. We also collected data on schooling, literacy, and on the deaf community. In 1840, we started moving closer to statistical analysis, although we were still just aggregate data. And aggregate data just simply means we counted up and reported the numbers. In 1850, the biggest change was that the individual became the unit of data collection as opposed to the household. And so for the first time, the names of all free people in any household was collected. However, for slaves, they still only collected age and sex. We also added new questions on occupation, literacy, birthplace, property ownership and other vital and social statistics. And this was the beginning of statistical revolution where statistics started becoming a science. So we started seeing some rudimentary statistical analysis. In 1860, looked very much like the 1850 census, using the same legislation actually from the 1850 census, only adding American Indian as to the list of races. 1870 is when we start seeing the shift towards a more modern census. Francis and Moser Walker, who was a statistician, and one of the founding fathers of the statistical sciences was the superintendent of the census. And he understood the power of the data we collected and how it could be used to tell a story. We began seeing graphics appearing in our final reports. We saw the addition of Chinese as a racial category. We saw the elimination of the slave schedule. And we also saw the introduction of the seat and tabulating machine, which would allow the clerks back at census headquarters to speed up how they counted the numbers. And as we added more questions, the time it took to tabulate that data got longer and longer to where it almost took to the next census to finish compiling the data from the previous census. And at this point there was no permanent census bureau. So they had to do something to start speeding it up. Francis and Moser Walker had also been an officer colonel in the Union Army, and he started realizing that we needed to have the workers at the census bureau reflect communities particularly out in the field up until 1870. The data in the field had been collected by U.S. marshals who would appoint temporary deputy marshals, mostly because they were the only federal officers in the field that could access every part of the nation. However, as the federal government grew, we also started having more representatives out in the communities. And Francis and Moser Walker thought that we should start using people from the communities to count the people there. And so he made a concerted effort to hire freed slaves, disabled veterans and the widows of slaves. In 1880, when we began using professional enumerators, instead of U.S. marshals, we started seeing African Americans and women in the field force and back at census headquarters. In 1880, we also for the first time did a census of the Indians quote, not tax, that is those living on reservations or tribal relations. We added Alaska to the census, we had just bought Alaska from Russia. In 1890 was actually our most comprehensive schedule. In between 1880 and 1890, the census bureau held a contest to see if we could speed up the compilation of the material. And it was won by a former census employee Herman Hollerith, who created the first electric tabulating machine, the Hollerith machine. The company would actually eventually go on to become IBM. So this was the first error of computers, and we would use variations of this Hollerith counting machine until 1951. We also saw the addition of new racial categories and 1890 was the only time we used most of these categories, Quadroon, Octoroon and Japanese. And it was very precisely defined. We collected data on Civil War veterans and their widows with a separate questionnaire just for Union veterans. We collected a myriad of social statistics on chronic illness, home ownership and deadness, the number of children naturalization papers. We declared the end of the frontier in 1890, which meant we had two people per square mile living everywhere in the continental United States. Unfortunately, this wonderful comprehensive census that we had was almost completely lost in a fire in 1921. Because for some reason, no one realized it probably wasn't a good idea to let workers smoke around all these fragile old papers. And actually it wasn't the fire that destroyed everything, it was the water they used to put out the fire. And I think 98 or 99% of those records unfortunately have been lost. In 1900, we started using experts in a variety of our areas. For example, when we did the census of American Indians in 1900, we used sociologists to go out and help collect this data. Sociologists and anthropologists and for some of the smaller tribes, this was the only information that we had gathered on those tribes. We did enumerate all American Indians on the general schedule. We used experts like Alexander Graham Bell to who oversaw and actually designed the schedule for the blind and deaf. Because you looked at the data we had collected in 1890 and said, not only are you collecting the data incorrectly, you're not asking the right questions, and you're not asking them in the right way to get the information that you would like. So we had a myriad of these special agents who would come in as experts in their field to oversee that particular area. We saw a lot in the manufacturing side too. We enumerated soldiers and sailors by departments of war and Navy and even might be out on ships or overseas. Hawaii was included for the first time. We collected information on utilities, transportation, juvenile delinquency, and we also eliminated many racial categories such as mulatto, quadrant and octarune. In 1902 was a time of big shift. We had been advocating for permanent Census Bureau for a long time. We started having other federal agencies asking us to collect information for them outside of the decennial census. And so they finally established a permanent Census Bureau in 1902. In 1905, we separated the economic census from the decennial census so that they were two different sets of collections and in different years. In 1910, we began enumerating people where they regularly slept, not where they worked, which is how we still do it today. If you look at some of the older censuses, you'll see sometimes people were enumerated in two places. Many Presidents were enumerated both at the White House and at their permanent home. There's an example in 1880 of Lauren goes Wilders, husband Almanza Wilder is enumerated both where he had his claim and then where he was actually working. And so therefore, temporarily absent family members for the school or the military were now enumerated with their family, not where they might be just temporarily. We also added the racial category for other for people who didn't fit neatly into the available categories. And we use this 1910 census to mobilize the economy and the population for World War One. In 1920, we had to take into account the boundary changes from World War One. So you would see people who identified where they were born in 1910 say hungry because where they were born was the Austrian Hungarian Empire. And then in 1920, they would identify where they were born as Chuckles of Acacia. And we had to take those changes into account. We asked questions about the ability to read and write English. As I said earlier, we did not do apportionment after 1920 because there were questions and pushback from the congressman who represented rural areas versus those who represented urban areas. Because more and more people had moved away from the rural area for the first time lived in urban areas in larger numbers than rural areas. And obviously there was a shift of political struggle, political power that was settled in the 1929 reapportionment act, which was the act that we still are under today. In 1930, we enumerated 48 states, the District of Columbia, Alaska, Hawaii, Guam, American Samoa, the Virgin Islands, Panama Canal Zone. In the height of the depression, we also did separate unemployment censuses in 1931 and 1937 where we mailed out postcards for people to fill out. But to be honest, the data collected from there just wasn't useful because a lot of people didn't want to admit they were unemployed and didn't participate. Some of the racial categories we had, we started using the term Negro for black. We added Mexican, Filipino, Hindu, and Korean. And actually the addition of Mexican, and it was only in the 1930 census and then it was not there for 1940, was the first time we collected data on Hispanic population specifically. In the past, we had done final reports on people with Hispanic last names, but as you know, that doesn't take into account people who may not have Hispanic descent but may not have a Hispanic last name. And 1940 was really the beginning of the modernization of the census as we recognize it today. We started using statistical sampling, which is where we asked extra questions of a certain percentage of the population. This would eventually become the long form. This included probability sampling, which had only been used on experimental basis before. And sampling had been tested before on that trial census of unemployment carried out by the Civil Works Administration and surveys of retail stores in the same decade. Where you could take just a percentage of households and ask them a question and extrapolate that data to the larger population. And the sampling in 1940 census allowed the additions of a number of demographic questions of these enumerated persons without asking, adding an undue burden on either the enumerator or the person answering the question. Sampling also allowed the Census Bureau to increase the number of detailed tables published and review the quality of data processing with more efficiency. We also saw the use of imputation, which was meant if data for one household didn't make sense in the context of their neighbors, they could or data was missing. They could fill in the data based on the data of their neighbors because neighborhoods tend to be fairly homogeneous and population and housing was collected at the same time but on different questionnaires. And this information was very important for World War Two. If anyone's ever heard of the victory plan, which was the plan that laid out exactly how many people, how much equipment and how many manufacturers and literally down to the number of nuts and bolts and how much steel we needed to win the war was all used with information that had been collected by the census. Not only demographic but economic are minerals, data we collected on minerals and everything like that. It was also used for a verification of birth information by people trying to prove citizenship to participate in war work, trying to prove their ages. And also for people who may have been stuck overseas at the break, you know, at the beginning of the war, they were able to use the census to prove that they were American citizens to be able to come home. And within our library, our librarian actually had a whole staff that probably dealt with up to 10,000 inquiries a day about about the census and because it wasn't available to the public, they had to look it up for people. The census data has also been used to settle wills. There was one famous case that took 10 years to settle it. A millionaire had left no will and the Census Bureau librarian spent 10 years establishing family connections using census data. And as you know, the National Archives just released the 1950 census. And let's look at 1950 looked very much like 1940 with a few important changes. So let's take a look of what the world looked like in 1950. Well, there were enormous social shifts post World War II. People had worked, moved for war work. And between the Depression and the war, we had seen the migration of so many millions of peoples, both North and to the West, and particularly the African American community that moved west and north and then stayed after the war was over. We saw education opportunities via the GI Bill. We saw economic and building booms, places out West like Arizona and Nevada that previously were not considered habitable. Now with rudimentary air conditioning, they can start building there and actually that's where we saw some of our biggest growth. Approximately 57% of the population now lived in metropolitan areas. And the percentage of people, increasing people living in metropolitan areas as opposed to rural areas was actually twice as great for African Americans as compared to white. And in fact, the number of blacks living in non-farm areas increased 33.7%. And we saw the beginning of the baby boom population growth of 19 million more, mostly fueled by that early baby boom which started in 1946. The census showed 16,196,536 children under the age of five, which made it the largest of any five year age band. And by comparison in 1940, there were only about 5.5 million in that age group under one to zero to five. So 90% of the growth was fueled simply by natural growth. Eventually the baby boomers would grow to 76 million and they would be the first demographic cohort that was named by the Census Bureau. The other 10% of growth we saw were immigrants mostly from war tour in Europe who came in under the Warbrides Act of 1945, the Fiancé Act of 1946, the Displaced Persons Act of 1948. We also saw the birth of suburbs. For the first time, people did not have to live in the city to work in the city. And we also saw that half of all homes were owner occupied, a huge shift from 1940. The number of dwelling units actually increased from 37.3 million to 46 million, a 23% increase, which is the greatest numerical growth of housing between any two censuses. The fastest growing states in terms of percentage increases were California, Arizona, Florida, and Nevada. And the West as a whole was the fastest growing region at a rate of three to four times as great as other regions. We also saw technological advances where the proportion of homes with electric lighting increased from 79% in 1940 to 94% in 1950. 96% of households had radios as opposed to 83% in 1940. TVs were new, but we did collect data on television sets, and about 5 million sets were reportedly owned, mostly people in metropolitan urban areas. Americans were living abroad in large numbers for the first time, mostly military families and federal workers and their families. In the past, when we had had military members living overseas, it was just that the particular soldier, they didn't have their families. Germany and Japan were the places where most of them were living. Statistical sampling now became a more exact science and could be used to collect a wider variety of data without imposing an undue burden. And the world was moving to the computer age. In 1951, we introduced the Univac system, which replaced the hollard tabulating machine, and part of the 1950 census was tabulated with that. Let's just take another quick look at how 1940 and 1950 differed. In 1940, we asked 50 possible questions. In 1950, we had 38 possible questions. In 1940, only 16 were asked of everyone, and then the rest were supplemental or only asked of a small percentage. In 1950, there were 38 possible questions, only 20 of everyone, and the rest were supplemental. We had a larger number of final reports, although in 1940, the compilation of the final reports, to be fair, was cut short by the war. We moved the census and manufacturing minerals industry away from the decennial census year, which helped free up enumerators and made the enumeration much more efficient. And for the first time, the majority of enumerators and crew chiefs were women out in the field. The census bureau had been employing women at headquarters since 1870 and in the field since 1880, but now you had a lot of housewives who chose to go out and be enumerators. And you could see just by the sheer number of enumeration districts how, you know, what a bigger number. An enumeration district was generally how many homes a single enumerator could visit in one day. So let's take a quick look at how did we come up with the questions for 1950. The questions were compiled after meetings with consultants, technical advisory committees, and also by identifying the data needs of other federal agencies, which is pretty much the way we do it now. We had some pre-tests that we did in 1940, but it was much more widespread in 1950 to make sure the wording was understandable and arranged properly on the questionnaires. And we also, for the first time, started experimenting and testing out different ways of collecting the data. For example, in several places in Ohio and Michigan, we had tests during the census where we tested self enumeration where people would fill out the forms on their own versus the enumerator visiting the home and collecting the data. We often used existing surveys to test our questions, such as survey of current population. And this was the first move towards removing enumerator bias. One of the concerns that they had is that enumerators sometimes would fill out the questionnaires based on their own observations or beliefs. And I have a family that we use as an example. It was a white man married to a black woman. And in one census, the entire family is listed as white because obviously the father was home. So an answer met with the enumerator. So the enumerator made the assumption that the entire family was white. And in the next census, obviously the mother was home. So the enumerator made the assumption the entire family was black. The Bureau of Indian Affairs requested a separate form for the census of federal Indian reservations. And because of the baby booms and because we also know that one of the historically undercounted populations are children under the age of one. They had infant cards for those who had been born since December 1949 and up until census day of April 1st. Unfortunately, those were not preserved and are not available. We conducted a census of housing agriculture at the same time. Again, these schedules were not microfilm, but these, and so they're not really covered here, but they included questions on irrigation, drainage, horticulture. And again, we worked with self enumeration. For example, the agricultural schedule was mailed to the household for them to fill out. And then the enumerator picked it up when they visited others such as the survey of residential financing were actually mail out mail back because people were often reluctant to give those type of personal details to an enumerator. So they had the option to mail that directly to the Census Bureau. In fact, the enumerator bags were pretty full. They could have up to seven forms to fill out at one household. And in 1950, again, was a time of innovation and changes were at the beginning of the computer age. One of the big things we had to do because we weren't the only ones using the census data. We had state and local and even private entities wanting to use this data to make their own assumptions, but everyone had a different measure of what a metropolitan district was. Some places used election districts. Some places use the town and someplace just determine boundaries based on population density, which is what we did. It was how much an enumerator can count in one day. So we actually teamed up with the Office of Budget Management and they came up with some standard metropolitan areas. And in that way, the data we were giving them could be used by anybody at any time because we were all looking at the same population and the same territory within that population. So in 1949, these standard definitions were issued by the Bureau of the Budget, which is the predecessor of OMB, the Office of Management Budget, and given the designation standard metropolitan area. And for the first time, we had a large number of American soldiers and federal employees and their families living overseas. So we did special overseas census reports, which were sent and distributed by the Department of Defense and the Department of State to each family. They collected these and sent back. And in all, in the 1950 census, we found a little more than 481,000 Americans living and working abroad for the federal government or on a merchant ship. And you have to compare that in 1940. That same survey would have found about 118, 119,000. Now, these people were not included in the importionment count unless their ship was docked in the U.S. And the simple reason was there was no way to know what locality to attach them to. For historically undercounted populations, we made an effort to do missing mispersons forms and newspapers and had a specific night to count transient locations such as motels, tourist courts and shelters. As I said before, we use variations of the Hollerith machine from 1890 until 1951. And actually the Univac one was the first non military commercial computer. And I think the serial number was 00001, the one that was brought to the Census Bureau. We experimented for the first time in those 10 districts in Ohio and Michigan. And some of the experiments were not only mail out where they would get the form in the mail and hold it for the enumerator, but we also tested maybe moving back to the household as a sampling unit versus the individual self enumeration versus direct enumeration. And actually the numbers on self enumeration versus direct enumeration were significant enough that we tested it even further with bigger populations in 1960. And after that, after 1970 went completely to self enumeration, which we still do today. The counting of college students prior to 1940, actually prior to 1950, college students were generally counted at their parents' homes, not at their college, because it was considered that on their breaks they would go home and live with their parents. So that would be their usual place of residence. However, the number of GIs returning to school with their families meant when school was not in session, they didn't live with their parents. And so to go more in line with the usual place of residence, college students were counted where their college was. And actually that remains a standard to today, which presented a bit of a challenge in 2020, as many students were sent home because of COVID to do school remotely, but we still counted them at their college. And finally, in 1950 was the first time we did a post enumeration survey, which meant we went back to a certain number of households and double checked the information we collected from them to make sure we hadn't missed anyone to make sure that the information we got was correct that had been written down correctly that the people understood the questions correctly to check individual data to check agricultural data and most significantly double check the data from the experimental areas. So we did post enumeration surveys and 3500 small areas to identify any households we may have missed, and approximately 22,000 households were re interviewed to check accuracy of data. And again, we still do a post enumeration survey until today. So 1950 really set the stage for the census to move forward to where we are today. The agriculture and economic censuses were separated from the decennial census, the housing census was integrated into the demographic questionnaire, although it was eventually pulled out and made its own census. We did not conduct any more censuses of reservations for the Bureau of Indian Affairs. The experimentation and self enumeration continued, and we started seeing long forms. So instead of asking supplemental questions, what the enumerators would do is we had mailed out the forms for people to fill out the enumerators when they picked them up a certain percentage of the households. They dropped the enumerators dropped off long forms to those households, who then would fill those out mail them back and actually that sample size was about 25%. We saw the development of FOSDIC, which was developed here at the Census Bureau with in conjunction with a couple of other agencies where you would fill out a circle, and it can be read the light could be read as it went through the machine anyone who's taken a standardized test where you to fill in the little bubble. That's FOSDIC. We added questions about employment and transportation because as the suburbs grew, people started living further and further away from their work. And so we wanted to see how that affected how people were getting there you had more households with two cars and things like that. And one of the most significant developments after 1950 is the opening of the National Processing Center in Jeffersonville, Indiana, which became the center for packaging and shipping equipment supplies, preparing maps, and the preliminary processing reports forms and the distribution of printed reports. So as you can see 1950 really set us up for the census of 2020. Thank you for your time. If you have any questions at all, please feel free to contact me and my myself my historians are always available to answer questions and to help guide people through census data. Thank you for joining us for today's presentation. We are wrapping up the lecture portion of the broadcast, but we'll continue to take your questions about today's topic in the chat section. Please take a minute to complete our short evaluation form so we can plan future programs based on your feedback. Find the link to it under show more. If we did not get to your question today, please send us an email. Note that the presentation's video recording and handout will remain available on this YouTube page and our website. If you enjoyed this video, check out our Know Your Records program. We have many more educational videos on how to do research with us. Thank you to the genealogy series team who contributed to the success of this program. We are grateful for your work. Again, please stay if you have questions, although we are concluding the video portion of the broadcast. We will continue to take your questions in chat for another 10 minutes.