What's Going on with SSA's "Death Master File"? (Part 2)

Earlier this week, we wrote about the unusual patterns we've been seeing in the Social Security Administration's "Death Master File," or DMF, with millions of identities being backfilled into the file with inaccurate dates of death. 

Additional media reports suggested that the identities of more than 6,000 non-deceased individuals — described as "migrants whose legal status had just been revoked" — were also added to the DMF, which was renamed as the "ineligible master file."

What we're seeing in the file

SSA releases weekly updates to the DMF, and in a typical week, the age distribution of new additions looks something like this:

This matches what we would intuitively expect of a file tracking deaths, with relatively few young people dying, a ramp upwards around the age of 70, and then a tapering-off by the mid-90s.

But this is what the age distribution looks like in the SSA's latest update to the DMF:

The spike at age zero and the high number of deaths after age 120 are reflections of the patterns we described earlier this week, and are caused by data issues and the backfilling of long-dead identities, respectively. What's new is the first curve on the chart: a spike in "deaths" around age 30.

And according to the updated DMF file, almost all of this younger-than-average cohort died on the same day: April 8, 2025. Below, we've compared the age distribution for April 8 deaths (red) with the age distribution for deaths in early January (black):

Among identities listed as having died on April 8, 2025, the median age was just 36 years old — which is, by any objective measure, extremely atypical.

Are the people with "April 8" deaths actually dead?

SentiLink's Fraud Intelligence Team manually analyzed a random sample of more than 100 identities with a listed death date of April 8, and for more than 99% of them, we were unable to find obituaries or any other confirmation they were deceased.

This is quite unusual: online obituaries for deceased people are often posted within a few days of their death. Not everyone who dies has an obituary posted online, of course, but given the size of the group we checked, we would expect to find many more obituaries. For example, using the same methodology, we checked a random sample of identities from the DMF with dates of death in January 2025, and were able to find obituaries for 70% of them. The vast majority of these obituaries were posted within 48 hours of the person's date of death. 

SentiLink also found positive signs that at least some of the identities added to the DMF are still alive. One person added to the DMF with a date of death listed on April 8, for example, posted to his TikTok account on April 10.

Who is in this "April 8 group" added to the DMF?

When we analyzed the "April 8 group," we observed that the five most commonly-occurring surnames are:

• Rodriguez 
• Garcia 
• Perez 
• Perez Rodriguez 
• Harutunyan 

For reference: according to the US Census, the five most common surnames in the United States are Smith, Johnson, Williams, Jones, and Brown. More than 2.5% of the US population has one of these five surnames, and they are also the five most common surnames in the DMF historically.

Yet in the more than 6,500 identities comprising the April 8 group, these five surnames (Smith, Johnson, Williams, Jones, and Brown) occur just 12 times in total. The chances of this occurring in a random sampling of the US population are vanishingly small. 

What does this mean for FIs?

Although the DMF has never been a complete source for truth about deceased identities in the US, it is nevertheless an important one that is used for identity verification by a wide variety of FIs, and we will continue to monitor and evaluate the impact of the SSA's decisions.