Tag Archive for SSDI

Don’t Cry About the SSDI

Last spring we did a two part series on the Social Security Death Index and it progenitor, the Death Master File. The posts are here and here. The series was prompted in part by reports of errors in the Death Master file, especially those which reported living persons as deceased.

We learned that nearly 90% of the information in the DMF comes from “first-party” reporters, that is, family, friends and funeral homes. About 5% of the information comes from States and federal agencies and the remaining 5% from postal authorities and financial institutions.

In terms of errors, the Inspector General of the Social Security Administration reported in 2009 that 89% of the errors in the DMF were the fault of SSA staff input mistakes. First-party errors comprised between 3.4% and 4.1% of the errors. The lowest error rates were from state reports (0.7% for non-electronically transmitted records) and 1.5% for states enrolled in SSA’s Electr9nid data records program. [The VA also had an error rate of 0.7%].

Now, as of November 1, 2011, SSA will no longer include state records in the Death Master File. Much has been written about this, including its potential impact on genealogists.

To understand what has happened, it is useful to examine the history and the legislation on the topic. The Death Master File was made public in 1980 as a result of a lawsuit under the Freedom of Information Act (FOIA). We’ve talked about FOIA as a genealogical tool in this space before. See the posts Getting Info from the Government–FOIA 101, FOIA Spotlight: The U.S Deparrtment of State,  and My FOIA Request Update.

The purpose of FOIA is to make available federal government information to the public. There are numerous exemptions to FOIA. The 1978 lawsuit, Perholtz v. Ross, in the U.S. District Court in the District of Columbia was settled by a consent decree between the Government and plaintiff Perholtz, who apparently sought the SSA’s death information to sell to businesses that might have need to verify deaths.

The first point is, then, that the DMF hasn’t really been public very long. The second is that from 1980 to 2002, state information was not used in the DMF. In 2002, in part in response to a 1998 SSA IG report, the SSA started down the road of electronic reporting from states. By 2008, only 22 states participated in the SSA’s Electronic Data Reporting (EDR) project.

The third point concerns why the SSA has withdrawn state records from the public DMF. The law which authorized the SSA to compile death information in the first place was intended to prevent overpayments and fraud in federal benefit programs.  The statute is codified at Title 42, United States Code, section 405(r)[originally section 205r of the Social Security Act of 1935]. The law requires  the SSA to establish a program under which States voluntarily contract with the SSA to periodically furnish the SSA with information on the death certificates filed with them so that necessary corrections may be made to the beneficiary records maintained under the social security program. This provision was added as part of the Social Security Amendments of 1983, Public Law 98-21.

But part of the 1983 Amendments provided that:
Information furnished to the Commissioner of Social Security under this subsection may not be used for any purpose other than the purpose described in this subsection and is exempt from disclosure under section 552 of title 5 and from the requirements of section 552a of such title.

The reference to section 552 is to FOIA; section 552a is the Privacy Act. The 1983 amendments took state-provided information out of the realm of public record as an acknowledgement of differing state rules about the privacy of such information.

As much as I hate to see governmental restrictions on what ought to be public information, I can’t say that this restriction will have a significant impact on genealogy.  Recall that 90% of the data in the DMF comes from first-party reporters, not states. So at least 90% of the data is unaffected. Some genealogists may have to dig a bit deeper into other sources in some cases. Recall also that states differ concerning the restrictions on vital records. Thus, in some cases, what’s no longer easily gotten via the SSDI, may still be relatively available from state sources. And even though the SSA is going to pull existing state records from the DMF, that does not invalidate the data. Nor in my opinion should make that data non-citeable. The citation should include the date the information was obtained from the SSDI and the source of the SSDI (remembering the distinction between the SSDI and the DMF).

Research Note: The SSDI (Part II)

This article is about the Social Security Death Index, not Social Security Disability Insurance.

In our last post we learned a few things about the Social Security Death Index. First we found out that the government doesn’t use the term SSDI; this term is used by nongovernmental concerns to describe the product that they derive from the Social Security Administration’s Death Master File. These nongovernmental entities receive the DMF from the National Technical Information Service (NTIS), a bureau within the Department of Commerce. Individuals may purchase the DMF directly from NTIS in a variety of forms. And here’s a handy link to understanding the architecture of the DMF, if you are technically inclined to explore the infrastructure of the product.  You can use the information to construct your own custom queries of the DMF.

We also discovered that 90% of the information in the DMF, and thus in the SSDI, comes from “first party” reports; that is, from private individuals (usually family members or potential beneficiaries), funeral homes, nursing facilities, etc. The Social Security Administration adds these reports immediately into the DMF. The DMF has coding that indicates verification or proof of death. On some versions of the SSDI this comes out something like this: “V = verified (family member contact)” and “P = proved (death certificate observed).”

There is always the issue of accuracy of a product like the SSDI or the DMF. Numerous studies have been done on this issue over the years. The main concern is not that someone who is dead will be left off the SSDI; that happens all the time, usually because of the lack of a report of death. The greater public concern is that someone who is alive will be listed as dead on the DMF. Statistically the occurrences of this error are rare but they are frequent enough and the consequences severe enough to warrant close scrutiny by the SSA. The consequences include termination of Social Security benefits, loss of the ability to get credit, even the loss of a job.

And an even greater consequence of erroneous death reporting is identity theft. That’s because the SSDI or DMF contain personally identifiable information on individuals. That’s not such a big deal if you’re dead, but if your date of birth and Social Security number are published while you are alive, well, you can imagine what might happen.

In early 2009, the Inspector General of the Social Security Administration reported on the sources of DMF errors. See the table below.

Sources of Error in Death Reporting

Source: Social Security Administration, Office of the Inspector General, Sources of Erroneous Death Entries Input into the Death Master File, Audit Report No. A-06-09-29095, 4 Feb 2009

These data should relieve the anxiety of those who are concerned about first party reports. The fact is there simply no incentive to to erroneously report a death to the SSA, and only the worst of motives for doing so deliberately. In fact, one might suggest that in some cases there is a greater incentive not to report a death to the SSA. Doing away with first party reports is not among the solutions to the error problems with the DMF. As the SSA IG observed:

SSA would have to turn away widows or other family members who visit a field office to report a death until it receives “official” notification from the State before initiating or terminating related benefit claims. SSA would also have to stop processing death reports received directly from other non-State sources, such as funeral homes, postal authorities, etc. . . . .

Eliminating first-party reports of death information would delay the processing of claims, increase erroneous payments, and cause public relations problems.

So how does the SSA keep the DMF accurate and up-to-date? One way that’s taken on greater prominence is the use of the Electronic Death Registration system (EDR). The system allows states to report deaths directly to SSA. While states do that presently, EDR enhances the accuracy of the reporting by allowing the states to verify the Social Security number online before the report is sent. The preceding table seems to indicate that EDR reporting is significantly less error-prone than other forms of reporting. The SSA has undertaken other internal steps, including new technology and training, to reduce staff errors in death report inputs.

Another important fact is that subscribers to the DMF who publish the data are contractually obligated to correct and update the data as they get new data from NTIS.

How do you report an unreported death? Call the SSA toll-free at 1-800-772-1213 (TTY 1-800-325-0778), 7 a.m. to 7 p.m., Monday through Friday. You need to have the deceased person’s Social Security number when you call.

How do you correct an error in the SSDI? Go to the nearest local field office of the SSA and take proof of the correct information with you. If it is verified by the SSA, the correction will appear as soon as possible in a weekly or quarterly update to the DMF.

Research Note: A Bit of Info about SSDI

A poster on the APG public mailing list recently asserted that she had come across an error in the date of death listed for a person on the SSDI. She questioned how this could happen when the entry was supposedly verified by a family member.  In response another poster said that as far as she knew, the Social Security Administration does not accept death reports from private individuals.

Unfortunately, the second poster’s information that SSA does not accept death reports from private individuals is incorrect.

it is useful to understand what we refer to as the Social Security Death Index is and is not.  For the best understanding of just about anything, an old rule that we baby boomers learned in the 1960s applies: go right to the source and ask the horse!

As a result of a [1980] Freedom of Information Act lawsuit, SSA maintains a record of reported deaths known as the DMF.  The terms of the related consent judgment required that SSA make available to the [public], the SSN, surname, and date of death of deceased numberholders.  As of June 2007, the DMF database contained detailed information on more than 82 million numberholders.  SSA provides DMF data to the Department of Commerce’s National Technical Information Service (NTIS).  NTIS, in turn, sells the DMF data to customers we broadly categorize as follows: (1) Federal, State, and local government customers; (2) industry customers including financial, investigative, credit reporting, and medical research organizations; and (3) public customers, including genealogists, individuals, etc.  Customers can purchase the complete data file for $1,725 and subscribe to monthly electronic updates for another $2,600.  The electronic updates provide subscribers with DMF additions, corrections, and deletions.

[Source:Social Security Administration, Office of the Inspector General, Quick Response Evaluation:Sources of Erroneous Death Entries Input to the Death Master File, Audit Report Number A-06-09-29095, February 4, 2009 (available at http://www.ssa.gov/oig/ADOBEPDF/A-06-09-29095.pdf)]

The DMF is not available directly from SSA. That which we civilians refer to as the SSDI (Social Security Death Index) is one of the commercially repackaged versions of the DMF that a vendor has obtained from the NTIS.

The DMF contains nearly 89,000,000 records of Social Security recipients who have passed away since 1936. Most of these records date from 1962 to the present.

Each year SSA adds more than 2 million records to the file. The records include both beneficiary and non-beneficiary records and verified and non-verified data. Sources of death data are: family members; funeral homes; States, Federal agencies (Centers for Medicare & Medicaid Services, Department of Veteran’s Affairs, etc.), postal authorities, institutions and internal sources from SSA’s payment records. Ninety percent of the file includes reports from family members and funeral homes.

[Source: Ray Wessmiller, Using Audit Software and the Death Master File to Catch Crooks, Newsletter of the Information Systems And Audit Control Association – National Capital Area chapter, September 2002, available at http://tinyurl.com/3jajwhm.

As depicted on the chart below, SSA receives most death reports from funeral homes or friends/relatives of the deceased.  SSA considers such first party death reports to be verified and immediately posts them to the DMF.

Sources of DMF Info[Source: Social Security Administration, Office of the Inspector General, Benefit Payments in Instances Where the Social Security Administration Removed a Death Entry from the Beneficiary's Record,  Audit Report Number A-06-07-27156, June 19, 2008, available at http://www.ssa.gov/oig/ADOBEPDF/A-06-07-27156.pdf].

So indeed most death reports come from private individuals.  In our next post will talk about how the SSA verifies DMF information and how you can correct errors if you find them in the SSDI.

Two Quick Bits about Internet Research

Green Light (meaning “good”!):  I don’t know how I missed this one, but you can now order SS-5′s online from the Government using a credit card!  Go to the Social Security Administration’s “Electronic Freedom of Information Act” page at:

https://secure.ssa.gov/apps9/eFOIA-FEWeb/internet/main.jsp

The records requested will be mailed in hard copy to you.  This is a step in the right direction!

Yellow Light (caution):  When I was digging around the RootsWeb site tonight, I found a  search box that said, “GenealogyArchives:  Type in your name and get your entire family tree instantly.”  When you do so, you get a search page that claims to have been generated by “GenealogyArchives,” but looks suspiciously like the hundreds of mutations of “person searches.”   The results page seems to have been designed to evoke the image of the bad offspring of Ancestry.com and Intelius. I’m not saying these companies are involved in this, but somebody wants you to feel like you’re with an old and trusted (more or less) friend.

Be careful out there!

Detective Work: A Misplaced Headstone

I was trolling through Greenman Tim’s Cabinet of Curiosities [which I'll write about at a later date--for now suffice it to say, a day without Walking the Berkshires is no day at all] when I came across this curiosity:

Clarence Thomas

Delaware
PVT 52 CO
152 Depot Brigade
World War 1
December 13 1890
March 14 1956

It’s the inscription on a headstone that’s been “riding around” in the back of a pickup truck of Tim’s aunt’s handyman in Maryland. Tim reported that Aunt Peggy says:

In John’s opinion this was the stone of a black man, hence the Depot Brigade, as African Americans were not allowed into the regular army in the first world war. It is in good condition, and shows no sign of being hit by tractor or plow, as can happen around the Eastern Shore of Maryland.

I told Tim [in his comments] that it’s not the gravestone of a black man. How do I know? Well, the 152d Depot Brigade was stationed initially at Camp Upton, Yaphank, Long Island, New York, as part of the 77th Infantry Division. Like all Army units in WWI, the 152d was segregated and there were no black soldiers in it.

The early history of Camp Upton is told in a 1918 pamphlet by Roger Batchelder. He says that the 77th Division was known as the “Metropolitan Division” because “every man in the division was formerly a resident of Greater New York.” Batchelder notes that there were black soldiers at Camp Upton. But they were not in the 152d Depot Brigade. The black soldiers were in the 367th Infantry Regiment and the 351st Machine Gun Battalion. And these organizations, though barracked at Camp Upton, technically were not part of the 77th Division. Instead, they were part of the 92d Infantry Division (Colored), the so-called “Buffalo Soldiers” Division.

So how to find this Clarence Thomas? Notice first the style of the inscription on the headstone. This is basically the style of inscriptions on headstones provided by the Department of Veterans Affairs. See examples here. Given that, we should take a look at the VA’s Nationwide Gravesite Locator. This database includes veterans buried not only at National Cemeteries, but at private ones as well. [It's worth noting here that there are no National Cemeteries in Delaware; there are three in Maryland, one in Annapolis and two in Baltimore]. Of the 110 “Clarence Thomas” names in the VA database, none fit our dates of birth or death, although there is a Pvt Clarence L. Thomas, buried in New Jersey (DOB: 1/24/1890; DOD: 3/7/1952), who comes close. While the disparate birthdates may be of no significance, let us assume that a four year discrepancy in a death date isn’t close enough even for government work. Nonetheless, let’s keep this one in mind while we move on.

The next simplest search may be the Social Security Death Index. We should try several versions of the SSDI. I like the ones at Rootsweb and GenealogyBank best. In using the SSDI for this case, I would narrow the search to those named Clarence Thomas who died in 1956. I would consider a 20th century death date more reliable than a 19th century birthdate. I might also check a couple of years before 1956. Having done the SSDI, we don’t find any Clarence Thomas that comes close to our date parameters. (Why might that be? The data in the SSDI comes from the Social Security Administration’s Death Master File. The majority of deaths listed there are from 1962 or later. Our subject died in 1956. Second, not every person is in the Social Security system. This depends on the type of employment and pension arrangements a person has.)

The next tool I would use is Find-A-Grave. This site has 19 million user-contributed gravesites, and it’s easy to use. I’d follow the same procedures concerning dates as we did with the SSDI. There are a total of 168 individuals with the name “Clarence Thomas” in the Find-a-Grave database. Unfortunately, none of them appear to match our Clarence Thomas by dates or locations.

The tools we’ve used thus far are the “quick solution” tools. They’ve not given us an answer.
Before we move on to the more complicated tools, let’s consider all that we know, including the following:

  1. The headstone apparently was found in the area of Maryland known as the Eastern Shore.
  2. The person in possession of the headstone believes it may memorialize a black man.
  3. The deceased was in a military unit that was almost certainly segregated, limited to white soldiers.
  4. A chronicler of the military installation where the unit was stationed says that “every man in the division was formerly a resident of Greater New York.”
  5. The headstone indicates that the deceased either was born in, or entered military service from, Delaware.

With those things in mind, we’ll move on to the more sophisticated analysis next.

COMING: Detective Work II: Newspapers, The Census, and Other Tools