In the mid 1980s, a research buddy and friend, Sue Hawkes Cook told me about a Cambridge, Ohio firm that tapped into a database, the Social Security Administration’s Death Master File, now known as the social security death benefits index; the social security death index; or ssdi. For around six dollars you could send away for a specific name. I no longer remember the exact price or the firm’s name. In the late 1980s, probably 1988 or 1989, the LDS church added it to their in-house searchable databases with the newest International Genealogical Index a.k.a. IGI and Ancestral File searches. This database has grown monumentally from about 25 years of records, to nearly 50 years of data, roughly 1962 to 2011. It is the nearest thing to countrywide death index as exists for the United States, since deaths are reported to county health departments for each state. Genealogists use it constantly, so much we forget it isn’t a source, just an index….
It is available through FamilySearch, RootsWeb, Ancestry, Heritage Quest and others. Looking back through my research, I have used versions from the LDS Church, Automated Archives and GenRef. I have two versions from the mid 1990s, both CD rom collections. Then RootsWeb put it up on the web, interactive and the rest is history. It is a research mainstay, I check it constantly for questions involving births from 1870s forward. One of the simplest features is to check to see where a certain surname is situated geographically, this can be done by state, county, town and down to specific zip codes. Zip codes were invented to make delivering mail faster and easier, not to help genealogists, and over the decades some have changed. On the other hand, many have remained the same.
Search by surname, by unusual first name and birth dates, death dates and towns, by zip codes, by social security number. I search this data base every which way I can. When I come across a new surname, one that seems to be unique, I check soundex variants to be certain I have the correct spelling. It is possible the name I have is misspelled and a quick check in the ssdi will put me to the right spelling. I use it to plot migration patterns by plotting identifying issue states and dates. If there is a difference in the last known residence zip code compared to the benefits zip code, this could be a clue as the location of descendants or burial places that don’t match the last known residence.
It’s an index, not a source. Because it is an index, it has built-in all the inherent problems of an index, missing entries, typos, handwriting analysis issues ie illegibility, how to deal with suffixes, i.e. Jr., Sr., vs. I, II, III with or without commas, computerization issues, how to deal with long surnames, where to truncate names, hyphenated names, Spanish surnames and their truncation issues, Dutch, French, Italian, Irish, Scottish names with prefixes of Van, Vander, Mc, Mac, De, Da, De La, D’and O’, the spaces, apostrophes all create variants for computerized searching. Then there are people who use their middle name instead of their first name, Loyd Burns aka B. Loyd Burns, aka Bruce Loyd Burns might be very hard to find if you don’t know Loyd isn’t his first name. Then there is the helpful bureaucrat along the way who “fixed” Loyd’s name to Lloyd. Some of the death dates are really update batch dates or “when the benefits stopped” dates, so sometimes April 1 and April 15th or 16th might be dates to search for the last days of March or the first half of April. That doesn’t cover the possibilities of spelling Eisenhower any of ten different ways up to and including Icenhauer.
Some web sites including Rootsweb and Ancestry have a link to print a letter requesting the ss5 application form for $27 through the FOIA (Freedom of Information Act). I have written two letters to obtain this information, but the price point is generally too high. Death certificates from some states are less, then again some states don’t consider death certificates public records. The advantage to SS5 is that is places the subject in a time and place in his or her own writing about parentage and birth, though nowadays the form is generally filled out by a parent at the birth of a child. Some informants on death certificates really knew very little. NOTE to the federal government deficit team: if the price of this FOIA request, was five or ten dollars a person, I might use it more and I expect other researchers feel the same. At $27, I save it for brick wall research, when other less expensive options fail.
Different ssdi files have various current dates. The information available on FamilySearch has their file current til 28 Feb. 2011. In order to verifty the current date of the file you are searching, use Smith or Jones and last month’s date. If no one shows up, it isn’t updated, go back month by month until you get some hits. That is the last update. That method is faster than reading through all the file description information posted by Ancestry and Family Search. Check the other files to see if one of them has a more current update. For a while Rootsweb.com prided itself on being the most up-to-date. I don’t know if that’s true or not today. Ancestry claims to update weekly, but other sources say the government updates the DMF monthly.
Reasons a person may not show up:
- He or she isn’t dead.
- He or she never claimed benefits.
- Someone is still claiming benefits on that number.
- He or she died prior to the computerization of the records. The social security administration may have data on this person if he or she was alive in the late 1930s, but they are not “in the file.”
- Typographical errors, yours or the ss administration’s.
- Day and month transpositions.
An aside about dates: When I first wandered into the Western Reserve Historical Society Library in University Circle in Cleveland, I was in my twenties and the true and serious “old guard” were in their seventies and eighties, Betsy Eastwood and Bill Johnson come to mind, fantastic researchers. These researchers had documented lines back to the 1620s in America and written and published books about those lines (before computers!!!). They used the dd/mm/yyyy notation, a European style system which seemed fine for them, but pretentious for me. I was used to the American style of mm/dd/yyyy, so I kept that notation way in my research. They gently suggested the other style. I found out less than two years into researching that they were right, completely right, the 1 June 1800 style was less apt to be confused than the June 1, 1800 method. It didn’t matter if it felt pretentious, because it kept the problems to a minimum. I received information from a cousin of a cousin and the dates were tangled, day and month transposed in about half the possible dates. It is a steep learning curve and I didn’t really want to change. I can still remember thinking, oh sure, I could use the European style but I’m careful, so it wouldn’t matter and that was for “serious researchers” anyway. I was just doing this all for fun. I was wrong. I use dd/mm/yyyy all the time now. Day/month transpositions should be checked in any date sensitive index like the ssdi and of course all linked genealogy sites too.
I have seen birth dates switched between husband and wife files. Don’t forget the Jan. vs. June possible errors. Remember social security numbers were issued beginning in 1938, a time when some offices used handwritten documents and some offices had typewriters. When computerization occurred another layer of possible errors were added in.
While new social security numbers can now be issued at birth, previous numbers were issued on an “as needed” basis. A first job, or a first bank account are two occasions that come to mind. Now a social security number is also a tax identification number and has multiple other uses, but for almost thirty years that wasn’t so. There are instances where a father and son had numbers one digit apart as well as siblings also one digit apart, a consequence of applying the same day at the same office.
The numbers connected to issuing state, are not always the birth state, but it isn’t a bad guess to start with. There was a busker in Boulder who had memorized the zip codes in the United States and also the social security numbers issuance provenance. He would amaze the crowds by suggesting home towns from zip codes and birth states from the first digits of a social security number. It didn’t work all the time; in my immediate family, it’s true 3 out of 8 times, but in a group of 30 people it’s true about 3/4 of the time. It’s a good clue and that’s why indexes are great, for the clues.