Hi! I’m new to IPUMS. I recently downloaded what I thought was the complete 1850 U.S. census microdata from IPUMS. Based on another source (Wikipedia) I expected this file to contain ~20M records (i.e., 23,191,876 total U.S. population including enslaved people; minus 3,204,313 enslaved people). Instead, the file I downloaded only contains ~5.9M records. This makes me wonder if what I downloaded is a sample of some kind. Is there a source for microdata control totals (i.e., expected full record counts) somewhere on the IPUMS site; or some other way I can determine what portion of the 1850 census my dataset reflects? Thanks for your help!
It sounds like you are receiving a data extract for 1850 that does not have as many persons as expected. I had a look at your account and found the culprit. On the Extract Request page, there is a button for Select Cases. It appears that you have selected LINK1850. The Select Cases menu is intended for users who want to limit their data extracts to certain cases to decrease the size and complexity of the dataset based on their selections. LINK1850 is the Historical identification key flag for the 1850 census which indicates whether or not the respondent can be linked to another full count census. In the 1850 census, 5,997,363 persons can be linked to other full count censuses. These are the persons who are in your data file. If you would like the entire 1850 census, please resubmit your extract after you have unchecked LINK1850 from the Select Cases menu.
Dan - Thanks so much for your very helpful response!! I’ll rerun and let you know if I have any follow-up questions. Actually, a follow-up question I have now is what the process is for getting access to personal identifiers (name/address) on the complete count datasets. I’m very interested in trying to link census data out to the broader historical record; and this additional information is critical to that ability. If you can please let me know, I’d appreciate it. Thanks again for your response to my previous question. Best, Jim
You can select and add linking key variables to your extract. Previously, you were using Select Cases feature which is designed to limit your extract to ONLY those cases with a linking key, but if you just add the linking key variables to your extract you will get the full extract including those with and without linking keys.
Edit: I didn’t answer your question! For access to restricted use data, read the linked webpage and reach out the ipums@umn.edu for more information.
Great - thanks again for your help - I’ve been able to download the complete dataset.