Your transcriptions on NARA’s Official 1950 Census Website at https://1950census.archives.gov/ are generally indexed within 24 hours of submission. By transcribing names, you help to push those spellings – and therefore those records – to the top of search results when other users search for those names. However, your transcription will not appear in the list of “Machine Learning (AI) Extracted Names” at this time. NARA is working on an enhancement to add the transcribed names to the search results display.
The 1950 Census release launched today at https://1950census.archives.gov. It includes a partial name index: primarily first names plus surnames for heads of households and persons in the household with a different surname. The index has a lot of inaccuracies due to optical character recognition (OCR) attempting to decipher the handwriting of 140,000 census enumerators. However, having at least a partial index online on Day 1 is wonderful. Most of the people I was interested in looking for I could find using the name index and some common sense, and if that didn’t work, I knew where to find ED descriptions and maps. I also spent time using the transcription tool to improve the discoverability of whole pages in order to help other researchers; I didn’t limit myself to just the people in which I was I was interested. Please help NARA improve the index by transcribing. Your transcriptions become discoverable by others about 24 hours after you input them.
Having spent virtually all day answering reference questions from Twitter, emails, and posts on the History Hub, it’s clear that there’s a few things that the researcher community needs to understand a little better. Among these, in no particular order:
(1) Census schedules exist for overseas American military and civilian personnel in Alaska, Hawaii, American Samoa, Canton [Kanton] Island, Guam, Johnston Island, Midway Island, Panama Canal Zone, Puerto Rico, U.S. Virgin Islands, and Wake Island. That’s it. Not in Germany, not in Japan, not anywhere else.
(2) If the census schedule says the “Not at home” household is on page 71 or higher, do not give up because there are “only 20” pages (or whatever) for the Enumeration District. Go to the last image, look at the Sheet Number in the upper right corner of the page. See if it is Sheet 71 or higher, and then work your way back a few pages until you get to page 71 or whatever page your “Not at home” household is on. Read my blog post for more information on what Page 71 and up.
(3) Yes, you can download individual images. Look for the three dots below the blue box that says “Help Us Transcribe the Names” and click on those three dots. You will then be given an opportunity to download: Click on the word download. Choose the level of quality you want (more pixels are better).
(4) You can share a link to an entire ED. You can also share a link to a single page if you’ve searched by name. Being old school, I copy and paste the link my browser shows, but there is a share feature that can do that, too.
(5) Learn by doing. If you’re not sure what one of the features in https://1950census.archives.gov does, click on it. Words in blue or highlighted in blue are clickable links. Gray features also often clickable. Play with it.
(6) https://stevemorse.org has a lot of great tools for census research – and more. Become familiar with them. Use them.
(7) NARA has provided a lot of useful resources with identical content at https://1950census.archives.gov/howto and https://www.archives.gov/research/census/1950. Read pertinent 1950 census blog posts: https://twelvekey.com/blog-posts-on-other-sites.
(8) Only one side of the census form was microfilmed in 1952. The original paper records were destroyed in 1961-63. Side 2 does not exist.
Ok, thanks for reading! Let’s go transcribe some more!
Get ready now with a sneak-peek from Michael Knight in NARA’s Genealogy Series: The 1950 Census Website: Design, Development, and Features to Expect, that premiered today, 30 March 2022. Researchers will be able to search by state, county, Enumeration District, name, and Indian Reservation.
The name search feature will not return perfect results on Day 1, however, for reasons outlined in the blog post, 1950 Census: Please Help NARA Refine the Draft Name Index! Researchers can help themselves — and everyone — by using the easy-to-use transcription tool to add complete names.
The U.S. Census Bureau hosted a webinar on Monday, March 14, 2022, at 1 p.m., Eastern Daylight Time, to give the media and data users an overview of 1950 Census records set to be released from the National Archives and Records Administration (NARA) on April 1.
The webinar highlighted statistics from the 1950 Census, the historical context to how the 1950 Census was conducted, and provided information from the NARA on how to access these records when they become available to the public and what resources are available now. The presenters are:
- Sharon Tosi Lacey, U.S. Census Bureau, chief historian
- Marc Perry, U.S. Census Bureau, Population Division, senior demographer
- Claire Kluskens, National Archives and Records Administration, genealogy/census subject-matter expert and digital projects archivist
- Jewel Jordan, U.S. Census Bureau, Public Information Office, public affairs specialist (moderator)
Resources: 1950 Census Records Release Press Kit
This post was updated 30 March 2022.
NARA’s 2022 Genealogy Series begins March 2, 2022, and it focuses on the 1950 census:
- March 2: Overview of What’s on the 1950 Census
- March 16: Mapping the 1950 Census: Census Enumeration District Maps at the National Archives
- March 30: The 1950 Census Website: Design, Development, and Features to Expect
- April 27: The Story of the 1950 Census Form P8, Indian Reservation Schedule
- May 11: From Parchments to Printouts: History of the Census from 1790 to 1950
- May 18: History of Census Records at the National Archives
- May 25: Historic Census Bureau Sources for Filipino, Guamanian and Chamorro, American Samoan, and Native Hawaiian Research
You’ll definitely want to watch NARA staffer Michael Knight’s March 30 program on the 1950 Census website — which will allow — on Day 1 — name searches AND allow users to help NARA improve this “first draft” name index.
Additional lectures may be scheduled – so check back at https://www.archives.gov/calendar/genealogy-series/2022.
All lectures premiere at 1 p.m. Eastern time – and remain online afterwards on NARA’s Youtube Channel. See you at the program!
This week the National Archives launched several webpages devoted to the 1950 census, including the main page, 1950 Census Records, at https://www.archives.gov/research/census/1950. Information on these pages will undoubtedly evolve over time. The supporting informational pages include:
- Questions Asked on the 1950 Census
- FAQs about the 1950 Census
- Finding Aids for the 1950 Census
- Instructions for Enumerators and the Public
- Census Forms in the 1950 Census Dataset
- 1950 Census Informational and Training Videos
- 1950 Census Published Statistical Data
- Native Americans in the 1950 Census
- 1950 Census: External Resources
As noted on 1950 Census Records, you will be able to search the 1950 Census website by name and location beginning on Day 1 — April 1, 2022. To develop the initial name index, NARA used Amazon Web Services’ artificial intelligence / optical character recognition (AI/OCR) Textract tool to extract the handwritten names from the digitized 1950 Census population schedules. Because the initial name index is built on optical character recognition (OCR) technology, it will not be 100-percent accurate. The National Archives is asking for your help in submitting name updates to the index using a transcription tool that will be available on the 1950 Census website. You can help us improve the accuracy of the name index and make the records more accessible for everyone. More information on this volunteer opportunity will be forthcoming.
When the 1950 Census becomes available on a NARA website on April 1, 2022, there will be a name search function powered by an Artificial Intelligence/Machine Learning (AI/ML) and Optical Character Recognition (OCR) technology tool. Granted, it will be imperfect on opening day – but that just means that all of us will have the opportunity to make it better through a transcription tool that will also be available. Exciting times! Read more about it at 1950 Census Release Will Offer Enhanced Digital Access, Public Collaboration Opportunity, a December 14, 2021 press release.
The National Archives recently digitized a previously unappreciated fragment of the 1890 census that contains information about 13 families in Alaska. This piece of the (mostly) destroyed 1890 census has essentially been “hiding in plain sight” since the National Archives published the Preliminary Inventory for Record Group 29, Records of the Bureau of the Census, in 1964. You can read more about it in “An 1890 Census Fragment for Alaska is Rediscovered.” Digital images are in the National Archives Catalog as “Logbook of Frank Lowell, Special Agent, Alaska District No. 2, 1890 Census” (National Archives Identifier 202288465).
NARA Cartographic Records staff member Amanda Pritchard brings you up to speed in “Census Enumeration District Maps for 1940 and 1950 Available Digitally in Our Catalog.”
In 1810, Salem, Massachusetts, was the 9th largest city in the United States, with 12,613 people. New York City was first with 96,373, and nearby Boston was fourth with 33,787.
From 1790 to 1870, U.S. Marshals and their assistants conducted the census. Preliminary Inventory 161, Records of the Bureau of the Census, on page 94, states: “Under the provisions of the decennial census acts, 1790-1820, the population schedules were to be deposited with the district court clerks, ‘who were to receive and carefully preserve the same.’ … A resolution of May 28, 1830 (4 Stat. 430), directed the clerks of the district courts to forward the population schedules for the first four censuses to the Secretary of State. … It is known that the 1790 schedules for Rhode Island were forwarded to Washington on June 22, 1830, as a result of the May 28 resolution. Presumably other extant population schedules, 1790-1820, were forwarded at about the same time, but no documentation of such action has been found.”
The Bureau of the Census bound the extant 1810 census schedules into volumes sometime between 1902 and 1934, but the volume that included Essex County lacked the town of Salem. Decades later these same records were microfilmed as National Archives Microfilm Publication M252, Third Census of the United States, 1810, which can be found digitally on popular genealogy websites. It’s likely that many people looking through the 1810 census schedules for Essex County have wondered why Salem was omitted. The answer finally came to light this year.
For unknown reasons, the 1810 census schedules for Salem, Essex County, Massachusetts, became aliened (separated) from Federal custody. Somehow, they eventually came into the custody of the Peabody Essex Museum Library in Salem, Massachusetts. A National Archives staff member noticed a reference to these records on Instagram in February 2021, which set NARA’s Permanent Records Capture team on a mission to return these important records to federal custody. Read more about it here: “Instagram Post Leads to Recovery of 1810 Census Rolls.”