Did missing/corrupt dates in COBOL default to 1875-05-20? | Hacker News Reader

Did missing/corrupt dates in COBOL default to 1875-05-20?

retrocomputing.stackexchange.com

277 pointsSeenNotHeard499 comments

The SSA's master records, the Numerical Identification (NUMIDENT) files, store dates in text as either CCYYMMDD or MMDDCCYY strings according to their archived versions.

https://www.openicpsr.org/openicpsr/project/207202/

I have a hard time believing the DB2 systems would convert it to days/seconds/whatever since 1875. It's not impossible, but I think whoever came up with the 1875 thing was simply wrong.

That's not to say there are 150 year olds collecting social security either. Dates of birth are sometimes missing or entered wrong and sometimes death records don't get entered. It's also clear DOGE didn't understand that social security numbers can't be used as a unique identifiers (nor why it's unnecessary) which can lead to all sorts of issues when processing.

Edit: It also seems the SSA presumes anyone over 115 is has died and halts payments which makes it even more unlikely there are 150 year old beneficiaries: https://secure.ssa.gov/poms.nsf/lnx/0202602578

Social Security started paying out in 1940: https://www.proquest.com/docview/146227490?sourcetype=Histor...

With anyone 65 years or older eligible to start receiving checks.

The 1875 date almost certainly comes from that. I wouldn't be surprised if someone set it as a default for anyone they didn't have a birth date for because you could safely assume someone was older than that if they were receiving social security when it first started _or_ because there are some of those initial payments that truly were never discontinued.

In 1875, the American Express Company established the first private pension plan in the United States.

I can also see a situation where the SSA recipient in 1940 was born in 1875. The recipient age 65 could have married someone very young for 10 years. That younger person could be on survivors benefits and continue receiving payments.

Here is an example: https://www.ssa.gov/history/idapayroll.html

On January 31, 1940, the first monthly retirement check was issued to Ida May Fuller of Ludlow, Vermont, in the amount of $22.54. Miss Fuller, a Legal Secretary, retired in November 1939. She started collecting benefits in January 1940 at age 65 and lived to be 100 years old, dying in 1975.

https://www.ssa.gov/pubs/EN-05-10084.pdf

https://www.ssa.gov/survivor/eligibility

So if in 1965 this retired person married someone age 18... Then that person at 18 never remarried. Could they end up in a situation where they receive survivors benefits from the original applicant?

The recent spouse would be born in 1947 and marry in 1965 for 10 years until their spouse's death... not have remarried or worked, then at age 60 in 2007, derive survivors benefits from original applicant?

> So if in 1965 this retired person married someone age 18... Then that person at 18 never remarried. Could they end up in a situation where they receive survivors benefits from the original applicant?

"In order to receive this (spousal) benefit, you must have been married for at least 10 years, and both you and your ex must be at least 62 years of age".

Basically, you could but you wouldn't get much to make it worthwhile. I do know that veterans benefits did (at one time?) payout to spouses in the way you ask, though. IIRC benefits from the American Civil War were paid out into the 21st century due to a handful of cases where old veterans married young women. The last known "civil war widow" died in 2008 having married an 86 year old at 19. A daughter received her father's benefits until 2020. There was also another spouse who died in 2020, but she never collected the pension.

https://en.wikipedia.org/wiki/American_Civil_War_widows_who_...

The ten year rule is more recent than the SSA.

There is a case like that with civil war benefits, the last survivor married a young girl that was his caretaker when he was like 90, and she finally died a few years ago, basically a lifetime after the war had ended.

Also factor in there were civil war widows legitimately eligible for pensions from their dead soldier husbands until the last one died just a couple of years ago:

https://en.wikipedia.org/wiki/American_Civil_War_widows_who_...

Also can be via your father, this lady was collecting a civil war pension until 2020:

https://en.wikipedia.org/wiki/Irene_Triplett

Can’t wait for Musk’s team to finally peel back these layers, realize that the code actually implemented the laws, and have to admit they are the idiots and apologize for wasting the government’s money on a poorly-run audit.

> Can’t wait for Musk’s team to finally peel back these layers, realize that the code actually implemented the laws […]

Bloomberg's Odd Lots podcast had an episode on this: the hard part in replacing/updating government system is not the coding part. The hard part is understanding the policies that have been changed and modified over the decades.

See "This Is What Happens When Governments Build Software" (Jun 2023):

> There's a lot of frustration about the government's ability to build things in the US. Subways. Bridges. High-speed rail. Electricity transmission. But there's another crucial area where the public sector often struggles, and that is software. We saw it with the infamous rollout of Obamacare. We see it in the UX of the Treasury Direct website. And we saw it in the way state unemployment insurance systems broke during the pandemic. So why is it so hard for the public sector to build and maintain software? On this episode we speak with Jennifer Pahlka, the founder and former executive director of Code for America and author of the new book Recoding America: Why Government Is Failing in the Digital Age and How We Can Do Better, as well as Dave Guarino, who recently left the Department of Labor after working on upgrading the unemployment insurance system. Both have a long history of working on public sector software systems and they explain why the problem is so tricky.

* https://www.youtube.com/watch?v=nMtOv6DFn1U

One large component is that a lot of business rules and policies have been encoded into the software logic, and (re-)translating that into code in a new(er) language is part of the challenge.

Related, "Why COBOL isn't the problem":

* https://news.ycombinator.com/item?id=41420217

> Bloomberg's Odd Lots podcast had an episode on this: the hard part in replacing/updating government system is not the coding part. The hard part is understanding the policies that have been changed and modified over the decades.

That is the hard part in any system, not just government. Especially government, maybe, but not uncharacteristically and certainly not exclusively!

They wont admit anything

At least we'll get a good example for the Chesterton's Fence wiki page.

Musks team wont be able to make any sweeping changes here that impact eligibility or payouts - thats the kinda thing where angry pensioners form a mob outside the white house and get front page coverage on every news outlet.

The code doesn't necessa8ry implement the laws. There could easily be a flag on each record that indicates "current recipent" with a value of "n" or "y". The code, in that case could be something like: if current recipient = y, then print check.

I suspect any admission would be made with far less fanfare than what accompanied the accusations made today.

No one is wasting money. These audits must be done. I do think however that the chances of anyone over 115, or deceased, collecting money is extremely low.

What makes you have the impression that the government doesn't do audits, with actual professional auditors? Because it does, and does so regularly

If an audit ultimately finds that the computer code is perfectly aligned with the law, how does that make it a poorly-run audit, or a waste?

Do you normally tweet your half baked “discoveries” as propaganda in an effort to undermine the government when performing an audit?

Edit: I really don’t think the way this audit is being conducted is out of a genuine love for America, or the American citizens, or even out of just wanting to do good work. This is very nakedly a dishonest, petty, and malicious investigation.

"Fact-checking Elon Musk's claims in the Oval Office" - https://www.bbc.com/news/articles/cwyjz24ne85o

This definitely warrants an /s even if it's obvious :)

You'll be waiting a long time.

Rule #1 of the new American oligarch: never admit a mistake

> Rule #1 of the new American oligarch: never admit a mistake

Is this really a new rule? By my understanding of history, this has been the standard operating procedure for governments since... well, since they became a thing. And when mistakes are admitted, they're done so very slowly. For example, it took the U.S. government nearly half a century to apologize for the Japanese internment camps.

Musk admitted in the White House interview that he'd made a mistake on the $50m for condoms to Gaza.

Wait if you like, just dont hold your breath.

They didn't use computers in 1940 for it though. All this was stored on paper where using days since 1875 wouldn't have made much sense. That would lead to most of the beneficiaries having a negative number.

They didn't get their first computer until 1956 and by that point, the 1875 date would have made little sense.

It simply makes more sense there was data entry errors. Indeed, we know there are since the SSA makes records available for dead people. They show people supposedly born as early as 1800:

https://aad.archives.gov/aad/fielded-search.jsp?dt=3059&tf=F...

> All this was stored on paper

Rather, as also mentioned in the answers, on Hollerith cards, 5-digit encoded (2 for the year, 3 for the day of year). Punch-card appliances were very much a thing. (WWII logistics were run by punch-cards.)

They used IBM cards, not Hollerith cards (they used IBM 077 collators and 405 accounting machines).

They were coded with 6 digits: 2 for month, 2 day and 2 year according to the SSA.

https://www.ssa.gov/history/ssa/yourss55.html

Notably, this is from December 1955, already on computers. There's no mention of how data was handled before this. (Also, pre-WWII, people used do talk of "Hollerith cards" in a general way – and the system started in the 1930s.)

There is so much more info on their site.

https://www.ssa.gov/history/hfaq.html

https://www.ssa.gov/history/index.html

https://www.ssa.gov/history/operations.html

Early Operations Social Security Bulletin

Volume 1 - June 1938 - Number 6

ACCOUNTING OPERATIONS OF THE BUREAU OF OLD-AGE INSURANCE

Joseph L. Fay and MaxJ. Wasserman

Bureau of Old-Age Insurance. Mr. Fay is Chief of the Baltimore Accounting Operations Section and Mr. Wasserman is Chief of the Statistics Section. They were assisted by Edward O. Manning of the Statistics Section. https://www.ssa.gov/history/fay638.html

> The 1875 date almost certainly comes from that.

Here's 1875 the claim that's been all over Twitter these days[1]:

"Social security runs on COBOL, which does not use a date or time type. So the > date is stored as a number using the ISO 8601 standard. The epoch for this is 150 years ago (1875) - aka the metre standard."

"So if you don’t know the date of something, it will be a 0 value, which in COBOL will default to 1875 - 150 years ago."

Weirdly this post itself didn't get a lot of attention (only 2.2k likes), but a ton of people took a screenshot of it and passed it around as fact, which got a lot more attention. For instance, post like this[2] with 56 thousand likes (there are a bunch of them on Twitter right now, all with a screenshot of this original post).

It's a good example of how misinformation spreads like wildfire, and how extremely few people bother to check to see where or not its accurate. I'm not sure it's useful trying to figure out reasons for a 1875 date when we don't actually have any indication that this was actually occurring, and the discussion started because some made an erroneous claim.

[1] https://x.com/toshiHQ/status/1889928670887739902 [2] https://x.com/enkday/status/1890808249663869204

The 1875 date is a made up thing. There's millions of records more than 100+ years old, some as old as 229 years old by birth date. And, it's still very much a problem if the first people to receive SS are still receiving it, or if the system is still cutting checks because their death wasn't ever recorded.

> or if the system is still cutting checks because their death wasn't ever recorded.

My speculation: there are a small handful of long dead people, for whom the SSA has never received proof of death, and they fall through some bureaucratic loophole in the "assume people over 115 are dead" rules, so the SSA is still printing them checks. But these checks haven't been deposited in a bank account in many decades, and the SSA doesn't even have a current address to send them, so it just sticks them in a filing cabinet in their basement. No fraud, the government isn't really losing any money since the checks aren't being deposited or cashed, just a little bit of bureaucratic stupidity that no doubt adds up to a rounding error in the federal budget. But you can see how Musk/Trump/et al could spin this to sound a lot worse than it really is.

The general problem is that what plays well in the media and what makes the most difference aren't necessarily the same thing.

If you want to find large-scale government waste, it's going to be things like Medicare approving unnecessary treatments, or corrupt officials subtly shaping the bidding process for government contracts to make sure they're awarded to the expected cronies for the expected exorbitant costs, or subsidies that are supposed to go to the poor but are shaped in ways that really cause them to go to landlords or drug companies.

But "we streamlined the bidding process for cloud services contracts to make it feasible for smaller entities to make viable bids" is a non-headline even if it would save billions of dollars. "We're eliminating housing assistance and turning it into a refundable tax credit in the same amount for the same people" would improve government efficiency (and outcomes), but it's not viscerally satisfying and makes people feel bored or uncertain rather than happy or angry.

What sells is simple overt fraud. If you can find some 30 year old cashing their dead grandmother's social security checks, that's a headline that will get people worked up. Which means they want to go find some of that stuff to throw meat to the base, while hopefully also doing those other things that are actually more important but don't get the mob excited.

The opposition then plays politics either way. If you actually reduce Medicare spending on unnecessary treatments you're "cutting Medicare" but if you go find a few clear instances of outrageous fraud then it's "just a drop in the bucket" and you're not making any real difference.

> But these checks haven't been deposited in a bank account in many decades, and the SSA doesn't even have a current address to send them, ...

While this is likely true for a majority of cases. I am confident that there are both bank accounts and addresses that have remained valid since the 1940s. For example, ABA routing/account numbers on checks go back to 1910.

Presuming checks are being printed, it would not surprise me if they are being sent through the U.S. postal system. What happens to them after that, though, is anyone's guess. Banks generally stay on top of deceased customers. But even then, a bank account can be transferred to beneficiaries and retain the same account number.

i would expect that banks would flag a check being deposited in the name of a known deceased former account holder. But there have already been cases mentioned in this thread where someone can receive SSA benefits on behalf of a deceased spouse/parent. As such, it is conceivable that those checks could be deposited.

I guess the difference is: if you see a social security check for a 150 year old being deposited, that would make you suspect fraud, and so you would investigate to determine if it is fraud, or some obscure case where it is actually legitimate. Whereas, if the checks never get deposited, where is the fraud? Which is why I think that the undeposited check case is the most likely to “fall through the cracks”

The very first recipient, Ida May Fuller, would be 150 years old currently.

That makes a lot of sense.

Also they do a heck of a lot more than retirement benefits. And really, it could actually be a survivor benefit and the deceased person could have been born 150 years ago and the survivor still living.

After much searching, I found documentation on SSA's Master Beneficiary Records format from after the Y2K update.

Date of Birth in some of these records is stored as days since January 1, 1800. There's no reference to 1875 anywhere. See section 2.3 Date formats:

https://omb.report/icr/200704-0960-012/doc/2576801

But it was common to encode missing two-digits data as something in the high-90s, maybe specifying a reason in the least significant digit. (I was taught to do so at university and did so myself for a project.) So, if the start date is indeed 1880, the code for missing date could still have been 95.

I’m still kind of surprised they would store days since 1800, but it makes a lot more sense than 1875. At least with 1800 the last two digits are the same, so debugging it is somewhat sane.

Let's say, someone grew up in an orphanage in the 1940s with an unknown date of birth. This was before any computer systems and the date would have been recorded on punch-card, probably in 2+3 digits, here defaulting to `00` for the year for a missing value. It's plausible that this would have been transitioned as-is to any more modern date encodings.

My understanding of the claim is that the mainframe data is stored in a high fidelity format but some records simply lack a full year. At the time social security was instituted it was common for people to not know their own age precisely. Beyond memory issues arising from old age, most people just didn't have a pressing reason to track it. So social security benefits were granted to people who seemed old enough even though they didn't have a birth certificate or similar. The claim, as I've heard it, is that the DOGE dolts transferred this data to a more modern system where they blindly passed incomplete or placeholder values into an 8601 library implementation that uses the 2004 standard's reference date of 1875 by default.

That theory could track since DOGE did seek approval to use SQL visual query software, the software may have parsed those birth dates in that regard.

Again why it is so irresponsible of Musk to have addressed the public in the way he did without (presumably) all of the facts and without nuance

How can anyone on the outside make all these guesses about the code and databases? It feels like a lot of speculation to try to prove a claim wrong, but without evidence.

Indeed, how can an outsider like Musk and his DOGE compatriots make any claim at all? I think it is far more likely that Elon was mistaken, or lied, than the SSA is paying out to multiple non-existent 150 year-olds. In fact, that there are so many plausible explanations that don't involve fraud, it really behooves us to be skeptical. Elon made the claim: he is responsible for supplying the evidence.

Note that the doge.gov website has a section called "savings," which originally said "Receipts coming by Valentine's Day." Now it says "Receipts coming over the weekend!", and seeing as it's 11:50pm on the East Coast of the US right now, that seems unlikely, too.

I’ve gone rushing to my boss in the past with similarly hastily researched claims and conclusions to what Musk is tweeting. In my experience, they were usually the result of me having performed insufficient research, and did not have nearly the impact I thought they would. However, these kind of junior level mistakes make for great propaganda. Anyone can read his conclusion and the chart and with zero critical thought, think “duh! The government is so stupid! I knew I couldn’t trust them to run a benefits program correctly!” And that’s exactly what Musk wants us to conclude. As you note, however, we cannot conclude that - there is a much, much higher burden of proof we should be demanding.

We all know that Musk is not someone with a strong commitment to the truth. We all know he's a grifter. How long has FSD been coming now? What was the deal with the diver?

But so much of these DOGE criticisms are simply reflexive "Musk bad." There is no a priori reason why it's unlikely that a small amount of people are being paid benefits out of a byzantine bureaucratic system that's almost a hundred years old.

Just a few years ago, I was listening to a story on NPR about SSI fraud (also administered by the SSA) and how common it was. Now, apparently we're all convinced all government agencies contain approximately zero fraud, or at least detectable-by-Musk fraud, simply because it's the cause du jour. It's totally bizarre to read these HN stories (I flag most of them) and people are making totally unfounded claims on the processes, controls, and operation of government IT systems, apparently assuming they are just as good as what exists at Meta or Google. Well, I have personal experience, and maybe my program was an outlier, but these people seem to have extremely rosy expectations, and often have no idea how the government works even in theory ("there's an IG to find things like this!").

It's really exhausting and disappointing. DOGE can be a misguided operation run by a grifter that misreports findings and will never save the amount of money promised and a lot of stuff in the government can be awful, missing even basic controls.

I started out with a similar position, but so far everything I've heard out of DOGE is either:

1. Misrepresenting congressionally approved expenditures he royally disapproves of as "fraud"

2. Misleading, under-researched, unsupported statistics as "waste"

The longer this goes on, the more impressed I am with our government. I had no idea it would be this hard to find anything legitimate to attack.

> 1. Misrepresenting congressionally approved expenditures he royally disapproves of as "fraud"

This seems dubious to me. In almost all congressional expenditures the funds allocated are done so at a department level with broad statutes applied. It is largely left up to that department to decide what to actually do with those funds specifically. It seems like a stretch to characterize this as "congressionally approved" and more like an abdication by congress on large swaths of government spending.

> But so much of these DOGE criticisms are simply reflexive "Musk bad."

I mean, these are your own words:

> DOGE can be a misguided operation run by a grifter that misreports findings

And that's the main issue and criticism, not that everyone believes that the government is perfect

But people aren’t saying that. Instead, they’re just inventing reasons why Musk is wrong, with no evidence or invented evidence, like the subject of this article.

I think the issue is a mix of "the boy who cried wolf" and really smart programmers with almost no real world experience who have been told to find "fraud". I am confident the government is rife with it. Issue is we have unleashed these young developers who I am sure are all excellent programmers who can code me under the table. Issue becomes developers with experience will question requirements, very good developers without experience will say "this is easy" and then just implement it. They were told to find fraud so we are presented with the classic "to a hammer, everything looks like a nail" .

On top of that, Musk is so eager to announce progress that he just repeats what he is told as he knows that these are very smart programmers.

I found myself working in an heavily regulated industry recently and I can tell you if you dropped a young really smart programmer into the code base they would find a million things that are inefficient and wrong. In reality they were carefully coded that way to support some byzantine law or legacy program or one off. An experienced dev will ask "anyone know why we do X in module Y"? and he will get an answer saying "yes, its because of...". The young hotshot dev just deletes and rewrites it or yells "fraud"

So Musk is shooting himself in the foot when if he just waited and verified he could release valid and I am sure damning evidence of fraud but currently he is just "crying wolf"

I've personally worked in such a system. I can tell you -- as someone who asked the question often -- that often, the answer is "because we wrote it that way in 1975 and no one has bothered to spend the time to learn how to make it better since then.

It is not necessarily true something written forty years ago is right. It is not necessarily true that something written forty years ago has some secret logic in it that cannot be improved upon. It is not necessarily true that something written forty years ago cannot be improved by fresh eyes.

Obviously one must not change things recklessly -- but there are often improvements that can be made.

And don't forget that the programmers in question are specifically selected based on their ideological adherence and loyalty. Which is to say, they are already convinced that waste is there, and that it is welfare and entitlements mostly.

Given that other systems do have problems with paying out to dead people or people who are alive but faking their age, it'd be impressive if the US system had zero such cases. The claim isn't unlikely on its face, especially given what he said about Treasury never denying payments for any reason in order to minimise complaints.

As to "how can he make claims at all", it's because civil servants have been told to work for him, which means he is supposed to have access to more information than actual outsiders do (everyone on this thread).

I'm sure the US system has a few cases of dead claimants. I mean, it's a country of 330m people.

Dead and 150? Bit harder to believe. It's an extraordinary claim, if you get my drift.

Why? Japan found over 200k ghost centenarians in their database. A significant amount were fraud related, relatives that keep receiving checks for example. So not hard to believe for me.

Yeah, exactly, and no-one can see Musk is tweeting about either.

But we're generally assuming that he's clinging to anything that justifies a pre-existing narrative.

As a person to whom all of this is "clear", and who "understands" everything, could you please explain how a system whose entire function depends on knowing the person's birth date would function without knowing the birth date?

Social security office knows that there are records in that database without confirmed deaths.

They use multiple techniques and data sources to determine who to send benefits to.

This is not news to the SSA.

https://oig.ssa.gov/assets/uploads/a-06-21-51022.pdf

"AGENCY COMMENTS SSA disagreed with our recommendations. Agency officials stated that most of the records discussed in the report involve numberholders who do not currently receive SSA payment"

So, they can do better, but sure, they are sending some relatively small number of checks out to dead people. That doesn't mean Musk needs to lie about the the program as an excuse to cut the whole thing, which is actually what we see playing out.

> they are sending some relatively small number of checks out to dead people.

This isn't even necessarily true from the report. SSA said 44k of the 18.9 million 100+ were receiving payments, but the US had something like 90k centenarians in 2021. It's probable that a huge majority of the 44k were still alive at the time of the audit.

at this point I just assume it's a given that he's tweeting misleading information. it seems like he removes all attempts removing community notes on himself as well. I just don't think the lay person on twitter, which is dumber than a an average layperson, can learn these things in 1 graph or 200 chars.

51% is most. 49% could still be > $100bil

In this particular case 'most' means 99.77%:

> At the time of our review, approximately 44,000 of the 18.9 million numberholders were receiving SSA payments.

There were something like 90k centenarians in the US at the time so the amount of fraud from this category is likely some small fraction of that 0.27%.

Let’s see what the revised fraud report discovers.

I have been working on COBOL systems for quite a while now. Currently and for most of my career, we mostly always use DB2 compatible dates ("CCYY-MM-DD").

Pre-Y2k a lot of dates were created with:

ACCEPT WS-DATE FROM DATE.

The above WS-DATE was in YYMMDD format, which is why there was a Y2K issue and needed to be resolved with windowing code. However, windowing code wouldn't work for somebody that was over 100 years old...

Doing a little research there is also (which I have never used since we just use DB2 and date parameter input files for CCYYMMDD dates):

ACCEPT WS-CENTURY-DATE FROM CENTURY-DATE.

This date is in CCYYMMDD format. According to google, the epoch for this date is January 1st, 1601.

Using an epoch of jan 1 1601 is ... interesting since it means you would have to deal with the switch from julian to gregorian calendar in september 1752. This included dropping 11 days.

Calendars and historic records is a pain.

Different countries switched to the Gregorian calendar at different times. Pope Gregory XIII papal bull went into effect October 1582, see: https://en.wikipedia.org/wiki/Gregorian_calendar

And so 15 October 1582 used to be the 0 for some COBOL date functions.

Later that was changed to Jan 1 1600. In IBM's systems you can control what you prefer by a switch: https://www.ibm.com/docs/en/zos/2.4.0?topic=services-date-li...

No joke, I actually hit this condition in a test suite and ended up stumbling across the October 1582 date in a Ruby library. It wasn't until I searched "October 10 1582" on the Web that I learned the significance. https://gist.github.com/abachman/f97806e1c0fe8e4e1849e5f8412...

tl;dr - MySQL uses 1000-01-01 as the minimum value for a datetime field. Different Ruby libraries use different methods to represent dates, which can lead to situations that appear to claim that 1000-01-01 != 1000-01-01.

I assume they just use the Gregorian calendar throughout - roughly what is called "proleptic Gregorian" [1], although since the Gregorian calendar started on 15 October 1582, which is before 1601, it's not really proleptic. Just barely retroleptic.

[1] https://en.wikipedia.org/wiki/Proleptic_Gregorian_calendar

The Windows epoch is January 1st, 1601. The rationale is:

"The Gregorian calendar operates on a 400-year cycle, and 1601 is the first year of the cycle that was active at the time Windows NT was being designed. In other words, it was chosen to make the math come out nicely."

Still more logical than January 1st, 1970.

Unix guys weren't overthinking it on systems running on 16 kB of RAM, or whatever. Just get a damn number, sheesh.

I'm not sure how a date in the format YYYYMMDD can have an epoch of 1601.

Epochs are used when storing the offset in some unit like seconds or days to a reference date. The epoch for a year YYYY is 1* BCE (since there is no year 0).

It doesn't make sense at all though for YYYYMMDD where there are multiple units and 0 is invalid for two of them.

* You're not really supposed to use dates < 1582 in ISO 8601 though without prior agreement though it's meaning isn't really defined.

Edit: There do exist standards that default to 1875-05-20 when there's a null date however (GIS-related), but I've not seen anything that suggests the SSA uses it.

> I'm not sure how a date in the format YYYYMMDD can have an epoch of 1601.

IBM mainframe Cobol contains builtin functions which convert between YYYYMMDD strings and integers - and 1601 is the epoch they use for that integer representation. I assume the person you were replying to was talking about this fact, just stating it somewhat confusingly

> The epoch for a year YYYY is 1* BCE (since there is no year 0).

Well, there is a year 0 in astronomical year numbering. One can say that 0 CE = 1 BCE and -1 CE = 2 BCE and more generally n BCE = -(n-1) CE for all integers n > 0. Maybe “0 CE” is incorrect per a strict definition of “CE”, but it is correct if we define it less strictly

> * You're not really supposed to use dates < 1582 in ISO 8601 though without prior agreement though it's meaning isn't really defined.

There’s really only two choices: proleptic Gregorian, or Julian. I don’t know why ISO 8601 doesn’t just mandate proleptic Gregorian with astronomical year numbering. But I suppose in the rare cases a computer system needs to represent pre-1582 dates (such as historiography), Julian is the norm. I suppose they could extend the syntax to specify which calendar is being used

> Edit: There do exist standards that default to 1875-05-20 when there's a null date however (GIS-related), but I've not seen anything that suggests the SSA uses it.

Do you know which ones specifically? Would be interested to know this

The COBOL days since 1601 epoch makes sense. Now that I think about it, if those functions don't work with negative numbers, then I imagine even the YYYY strings couldn't have years before 1601.

> Do you know which ones specifically? Would be interested to know this

This is the only one I've found:

https://docs.ogc.org/is/18-010r7/18-010r7.html#100

Temporal datum with Calendar ... and with TimeOrigin omitted so should be assumed to be 1875-05-20.

> The COBOL days since 1601 epoch makes sense. Now that I think about it, if those functions don't work with negative numbers, then I imagine even the YYYY strings couldn't have years before 1601.

IBM mainframe COBOL has two operation modes, ANSI (where day 1 is 1601-01-01) and “Lilian” (where day 1 is 1582-10-14, the day on which the Gregorian calendar was first introduced). ANSI mode complies with ANSI/ISO standards for COBOL, Lilian is an IBM-only legacy standard. So, in Lilian mode, the earliest value for YYYY is 1582. I don’t believe negative numbers are accepted.

>I'm not sure how a date in the format YYYYMMDD can have an epoch of 1601.

Remove the CMOS battery from a computer with no internet connection running Windows XP and it will default to the 1600's

https://en.wikipedia.org/wiki/Epoch_(computing) (CTRL-F 1601) Also notice, no mention of 1875.

Also just search "CENTURY-DATE COBOL Epoch"

If you search "CENTURY-DATE COBOL Epoch 1875", it is mostly just recent entries debunking 1875 as an epoch.

It was posted in the comments that ISO 8601, at one point, mentioned 1875-05-20 as a reference date. According to Wikipedia, it was later omitted[0]. I guess it's possible that the social security system (EDIT: that we know today) was initially designed with that date as a sort of epoch. Either way, it seems nuanced and no one has the full story (including Elon)

[0]: https://en.wikipedia.org/wiki/ISO_8601

Wikipedia says "ISO 8601:2004 established a reference calendar date of 20 May 1875 (the date the Metre Convention was signed), later omitted from ISO 8601-1:2019." I was curious what "reference calendar date" is supposed to mean.

Thanks to links in the SE thread, I found the relevant actual text in ISO 8601:2000 (I don't know how different it might be, if at all, in the 2004 document):

> The Gregorian calendar provides a reference system consisting of a, potentially infinite, series of contiguous calendar years. Consecutive calendar years are identified by sequentially assigned year numbers. A reference point is used which assigns the year number 1875 to the calendar year in which the “Convention du mètre” was signed at Paris.

This last sentence is simply an obtuse way to say "this year right now, as I [jomar] write this -- we call this 2025". Apparently the ISO committee did not want to refer to what was going on around 1 AD or felt that the missing 0 between 1 BC and 1 AD would lead to confusion or something, so instead used the birth year of the metre to state the bleeding obvious.

Nobody actually knows what year Jesus was born, so setting "1 AD = Jesus' first full year of life" wouldn't be accurate. It's more or less the same puzzle as COBOL's alleged default.

https://en.wikipedia.org/wiki/Date_of_the_birth_of_Jesus#Acc...

The current year numbering is because a monk called Dionysius Exiguus thought the existing system of numbering the years since Emperor Diocletian was stupid as Diocletian persecuted Christians. Dionysius decided the year in which he invented his calendar was 525 years since the birth of Jesus and created the Anno Domini system.

Dionysius didn't really explain why he thought Jesus was born 525 years ago (in December of 1 BC). Many historiographers have tried to understand his logic. Thankfully, unlike COBOL programmers, Dionysius documented his reasoning for picking 1 BC as the reference year, so we only have to argue about whether he was correct.

The original post on X completely misinterpreted this to mean that ISO 8601 established days since 1875 as a standard date storage format

Isn't it possible that 1875 was chosen, because of compatibility with US legacy data encoding and that then an arbitrary, but notable date in this year was chosen as the reference date (i.e., the Convention du Mètre)? I guess, the USA had quite a stake in this…

I don’t think there’s anything in ISO 8601 pertaining to integers that represent the number of years since some specific year though.

ISO 8601 prescribes string representations, not integers, and it requires at least four digits for the year, and the year the Convention du Mètre was signed is expressed as 1875, not 0.

> was first published in 1988

No, the social security system and the COBOL that powers it predates that standard by quite a lot.

That makes a lot of assumptions, especially given there was most likely Y2K problems that needed addressing and system upgrades taking place since the initial system rollout. This page[0] states that there was changes to the system in the 1960s but I can't find anything more recent so... maybe you're right. Who knows.

[0]: https://www.ssa.gov/history/ibm.html

Y2K problems were related to attempting to represent years with 2 digit numbers. By no means would setting a date to 1875 fix a Y2K problem, it would just exacerbate it. Even if the problem originated from a Y2K "fix", it would still be a major bug, resulting in millions to billions in excessive payouts going to who knows where.

I think parent post is saying that when they fixed the Y2K problem, they likely also modernized the date representation to comport with international standards. It wouldn't really be any more difficult to do both of those at the same time than just fixing Y2K bugs.

Whether or not Musk is right about 150 year olds getting Social Security doesn't matter, he just wants to destroy the administrative state. So long as that happens, they'll be perfectly happy with the outcome even if they get proven wrong on some technicalities later.

Prove Musk wrong on this and he'll just go about his day as normal; 20 minutes later there'll be someone tweeting an unhinged screed about how the US government is spending 10 trillion this year changing the name of the Department of Homeland Security to the Department of Homeland Inclusivity to which he'll quote tweet "Interesting" and then he'll set his little band of freaks to cause mayhem somewhere else.

I'm sure Fox news will investigate his claims and let their listeners know that he lied about this.

Y'all are trying to be very specific about this 150 year old thing when there are vast number of people with ages above 100 that are in the database:

https://x.com/elonmusk/status/1891350795452654076

   ...
   100-109 4,734,407
   110-119 3,627,007
   120-129 3,472,849
   130-139 3,936,311
   140-149 3,542,044
   150-159 1,345,083
   160-169 121,807
   170-179 6,087
   180-189 695
   190-199 448
   200-209 879
   210-219 866
   220-229 1,039
   240-249 1
   360-369 1

And that means what?

Data values are nothing without the exact context in which they were created and the exact context in which they are used. That's like level 1 data analysis.

Publishing such data without context is deceitful.

How is it published without context? We know that this is the age field from the social security system. And that the query omits records that are recorded as dead. Therefore, the social security system in America has records that claimed to be for people who are alive for whom the date of birth field is incompatible with that status. That seems like quite a lot of context, actually.

I don't know why some people are finding it so hard to accept that there is likely to be fraud in this system. Look into the determined origins of the so called blue zones to see that every country has problems with this, albeit some more than others. It's the government giving out free money, so naturally it attracts very sophisticated fraud schemes and civil servants are rarely motivated to track it down and investigate properly.

It does seem to me that logically, you have two choices:

1) Take the numbers at face value. In that case, you are predicting millions of accusations of fraud and an enormous number of prosecutions in the next year or two.

2) the situation is somehow more complicated, and most of those millions of records with 140+ ages do not represent fraudulent activity.

P.S. Mentioning the blue zones is incredibly silly. Those regions have modest numbers of individuals being reported in the 100-120 age range, which probably are fraudulent. None of those areas have millions being reported to be 140+. For instance, Sardinia had 13 reported centenarians per 100,000 population, which would be equivalent to ~39,000 centenarians in the US. So it's orders of magnitude less than this database shows.

Or the third choice:

(3) The data quality is so bad that there is no way to reliably know if a case is fraudulent or not without an investigation so expensive the ROI is negative, so everything remains unclear and unresolved forever.

That's usually how it goes with cases like these. Sometimes there are gangs who organize large scale fraud against the system and those might attract the attention of prosecutors, but people not reporting a death or double payment or similar isn't worth it. There might never be a clear answer to how the database got into this state. But the basic point stands that data quality for a critical dataset is really low in obvious ways, so what about all the non-obvious ways?

I mentioned blue zones because there are a lot of people on this thread who are really having a hard time believing there can be problems as obvious as people who have died but continue receiving payments in social security schemes. Blue zones is just an easily searchable keyword to learn more about other times when it's happened at scale, e.g.

In 2010, the Japanese government announced that 82 percent of its citizens reported to be over 100 had already died.

In 2012, Greece announced that it had discovered that 72 percent of its centenarians claiming pensions – some 9,000 people – were already dead.

Puerto Rico’s government said in 2010 that it would replace all existing birth certificates due to concerns about widespread fraud and identity theft.

https://www.aljazeera.com/news/2024/9/26/the-secret-of-blue-...

Obviously nobody is claiming the database reflects reality. Governments often have multiple data sources that are badly out of alignment. A census can give more accurate data, but that doesn't mean SS is synced to it. For instance, in the UK during COVID, more people came forward in some age ranges for a COVID vaccine than theoretically existed in the country at all. The UK's population data is so badly screwed up that people started using the quantity of NHS numbers issued instead to try and estimate it.

It all boils down to what you consider more likely. Someone not quite understanding the data they're looking at or that nobody noticed several tens of millions of dead people receiving social security.

Whichever it is I would be very careful before making any grand public statements about it. And as far as fraud goes this doesn't sound anywhere near sophisticated.

Those aren't the only two possibilities.

I guess it depends on how much prior knowledge someone has of such problems. The correct explanation isn't "nobody noticed", it's "people noticed but could not/would not do anything about it". But how much you know about these problems will depend where you get your news from. If you never read right wing sources, you aren't exposed to stories about benefits/grants fraud as places like the Guardian or CNN just don't cover it, so all this will appear fantastical and absurd. If you're used to such stories then it all seems plausible and normal.

These types of things usually have some aspects in common. One is that front line workers, when asked in a safe environment, will give enormous estimates for how much of their payout is fraudulent. The people who are in denial about it are usually the people at the top, not the bottom. They don't ask, nobody tells, they don't want to know because they'll be held accountable for it and they fundamentally, ideologically, do not care about waste or inefficiency, viewing it as preferable to even one innocent person not getting their money.

I can't find it now but Musk has alleged similar things; he got private estimates from civil servants that 50% of the payouts from their department were probably fraud of some kind (it might not have been SS, I don't remember), and that the staff in charge of blocking payments at the Treasury were told to never do so. This kind of thing is what you'd expect. Often the front line civil servants are frustrated by this situation, as it's their taxes being wasted too. It's the people at the top who cover it up.

How many of those are test data, simple clerical errors (and how many of those are already in the process of being rectified), and how many of those are in actual use (e.g. how many actually use those SSNs in the wild)?

This is the important missing context. Musk can and does claim a lot. He rarely, if ever, provides any evidence or context. And none of his or his team's actions can be verified or monitored.

He does. That's what started this thread: he provided evidence and context to back up his claims of widespread government waste and fraud. What that got him was a bunch of smears from people spreading misinformation about COBOL.

Anyone else in the world would just stop sharing stuff publicly given this kind of public abuse. There's no requirement to do so. But Musk just shared evidence and context again, by showing the output of the equivalent of SELECT GROUP(age), COUNT(*) FROM SSNS WHERE DEAD = FALSE:

https://x.com/elonmusk/status/1891350795452654076

It shows approx 1.3 million SSNs in the 150-159 age bucket, along with many millions more that are actually over that age, and over 1000 SSNs allocated to people over 200 years old, including one that is for someone marked as 360-369 years old. There are over 20M people listed as 100+ years old. Total sum is around 395M people. Therefore there is no 1875 epoch and that claim was simply misinformation.

So here's what we know given his statements:

1. There are a lot of data entry errors in the SSN database.

2. These aren't test data, misinterpretations of an epoch, etc.

3. Such errors are extremely common and not being rectified.

4. Many SSNs aren't unique.

5. There are far more records in the database than people in America.

If we combine those statements with prior knowledge of other related topics, we can infer:

• The data quality issues are much more extensive than just age and number of records.

• Whatever processes are meant to ensure data quality don't work.

• This was not previously known to the public.

• Civil servants know all this but are often unable to do anything.

We also know that this situation is expected. It would be much more crazy if Musk announced his team couldn't find fraud in SS. Just look at the graph for payouts from the American disability benefits system - it tracks general economic performance. Other countries don't see such a thing in their payouts, where improving economic conditions magically make long term disabilities disappear, but they also tend to be more aggressive at cracking down on benefits fraud. The UK did a big purge some years ago where every single person claiming disability benefits was re-assessed.

Anytime someone comes in and does basic checks of government finance systems they always find lots of very basic stuff. At one point it was discovered, again in Britain, that a Labour council was regularly paying invoices multiple times and nobody had noticed for years. Paying money out to dead or non-existent people is a common problem in all such systems.

Context would be for example knowing how the code around the data handles those values. Some numbers in a database do not in any way imply real world effects.

Another thing is that COBOL records commonly have complicated unions (to save space) where a separate code affects how you interpret the fields. You need to understand all the business logic to make sure you are reading the data correctly.

This list adds up to 20million

60 million receive social security from age https://www.ssa.gov/oact/STATS/OASDIbenies.html

1/3 of recipients are fake death fraud?

There's about 60million from cenus data between the age of 60-85...

Does no one think critically?

Nah, many people willfully and gleefully put blinders on when presented a thin veneer of a statement if it supports what they want to be true.

As you point out, Elon’s stats don’t even pass a cursory snuff test in regards to what he’s trying to imply, but his champions don’t bat an eye.

With "death" flag set to false - sure. But are they collecting benefits / are they active in any way? Otherwise it's just that the death may not have been recorded.

If Elon Musk wanted to be honest, he should've published a statistic like that, but with SSNs that seem still active.

There were around ~1M (if not more) unidentified bodies found in the States over the last 100 years. On top of that, before digitisation, you've had mistakes in filling the data, lost documents and a ton of other possible causes for not marking someone as dead in SSN. As long as they are not collecting benefits, it's more hassle to fix than it's worth.

In a bit similar spirit - in Poland we're considering turning public health insurance into just a tax, and doing blanket assumption that every citizen has health insurance. Since we have less than 1% uninsured, the costs of tracking and verification are comparable/higher than just giving "free" insurance to the remaining 1%.

> If Elon Musk wanted to be honest,

Elon and his fans don't think this way. They don't consider it dishonest to publish misleading information (it's "just the fact").

If Elon wanted to be truthful and thorough, he would actually do some analysis. It's quite clear that he doesn't care about either of those things though and just immediately tweets whatever the 20 years olds with laptops send him.

This is good info. The histogram definitively disproves the COBOL theory.

It’s always good to respond to odd things with curiosity rather than cynicism.

It also seems clear that the vast majority of these old records are not collecting benefits, and even the few that are may have valid reasons (ex living younger spouses)

https://xcancel.com/ThatsMauvelous/status/189135619250239902...

> Y'all are trying to be very specific about this 150 year old thing when there are vast number of people with ages above 100 that are in the database:

You may wish to read the SSA Inspector General audit report, "Numberholders Age 100 or Older Who Did Not Have Death Information on the Numident.":

* https://oig.ssa.gov/assets/uploads/a-06-21-51022.pdf

> It's a gripping read. It tells me, for instance, that 98 percent of these folks have received no payments.

* https://twitter.com/justinwolfers/status/1891516202612060438

The report states:

> SSA determined the estimated $5.5 to $9.7 million in expenditures to correct these errors was too costly to implement and that the effort would have limited benefit to the administration of SSA programs. We acknowledge that almost none of the numberholders discussed in the report currently receive SSA payments. However, SSA issued each of these individuals a valid SSN and these SSNs could allow for a wide range of potential abuse.

So it would cost several million dollars to correct the database to prevent less than several million from going out.

Once again, perhaps the government knows what it's doing, these "discoveries" are not new or surprisingly, and that a cost-benefit analysis has already been done.

It seems that for a lot of people, that source isn't reputable (e.g. they would want to see more than a screenshot a guy posts on his website to know the situation is as you described).

It's gonna be hilarious if: this is a data entry error (1957?), the payments are to a surviving widow because the dates just fit [0][1], simply the query was wrong, or more probably Musk just lied.

[0]. https://en.wikipedia.org/wiki/American_Civil_War_widows_who_... [1]. https://en.wikipedia.org/wiki/Ida_May_Fuller

This is the best debunk I could find: https://iter.ca/post/1875-epoch/

I like this comment:

----

There was a time when data structures were made to fit purpose, not compilers. Having a look at the subject, shows clearly the constrains for valid dates:

    Social Security was introduced in 1935.
    To be eligible for benefits one had to
        pay in at least 40 quarters, that's 10 years
        be at least 65 years old

This means the first regular beneficiaries of social security payments were 65 in 1945, aka of the 1880 cohort. Virtually noone participating in this system can be born before 1880. Anyone older will not most likely not be a beneficiary, and anyone younger (aka still paying in) will be, well, younger.
So add another 5 years for wiggle room and we end at a nice round 1875 as earliest year for any birthday to be recorded.
A perfect rational base for a date entry, isn't it?

----

The "COBOL doesn't work like that" comments are missing the forest for the trees: This is a very old system with bespoke coding to match legislation, not legislation to match compiler default behaviour.

Fundamentally, unless a government employee that has worked directly on this codebase speaks up, we're all just guessing.

This is why I like hn. Many times I read about all of the horrible things that the current administration does. Due to so many cases where I know the news is wrong I stopped trusting anything. However, I feel like here I can let my guard down a bit and be more certain that a specific criticism is true.

>that a specific criticism is true

Or false, in this case.

I love how everyone is talking technical details whilst ignoring where it came from. This eliding of context and focusing on a technical question is a great way for people trying to cope with either stress if you're against it or with criticism if you're for it.

I worked with COBOL long before DB2. I also worked on the COBOL compiler in the 70s. Unless things changed drastically if a value wasn't initialized the compiler left garbage in the variable when the program was loaded. If you were lucky this caused the program to ABEND when you tried to use it, if no ABEND you got strange results. It was up to the programmer to set a default date if one was required by the application. Many applications started with VSAM which did not care if a date in a record was invalid. This caused many systems to set default dates when converting to DB2. DB2 does not allow garbage to be loaded into a date field. COBOL does not initialized data unless instructed to do so.

Here’s a tweet from Elon with a table of ages of all people marked not deceased who are collecting social security

(hint: there is a smooth trend of people of all ages up to 199, so the 1875 thing was pure misdirection)

https://x.com/elonmusk/status/1891350795452654076?s=46&t=NN3...

This comment definitely needs to be bumped up. It adds extra data to the mystery.

Here's the raw data:

  Age Range,Count
  0-9,38825456
  10-19,44326480
  20-29,47995478
  30-39,52106915
  40-49,47626581
  50-59,45740805
  60-69,46381281
  70-79,33404412
  80-89,15165127
  90-99,6054154
  100-109,4734407
  110-119,3627007
  120-129,3472849
  130-139,3936311
  140-149,3542044
  150-159,1345083
  160-169,121807
  170-179,6087
  180-189,695
  190-199,448
  200-209,879
  210-219,866
  220-229,1039
  240-249,1
  360-369,1

Bar plot of data:

https://imgur.com/a/34Y5BSv

If you assume all the non-deceased (as marked in the DB) are still getting a SS check, then simply tallying up 65+ amounts to a 8 trillion dollar social security payout annually. We are definitely not paying that.

My guess is the deceased flag is different from the “eligible for payout” flag, or that flag is determined on another join. They spoke too quickly, which is a common human error.

Someone else smarter run some experiments.

This is the smoking gun that Musk has no understanding of what he's looking at. A competent engineer would look at these numbers, recognize that they don't match payments, and shut up until they figure out the bigger picture that can explain everything they see.

Instead, he raced to tweet it. This is just dumb.

Agree.

Some professional auditors on Reddit were saying how often they go into a business and find a bunch of stuff that immediately looks like fraud, then they go interview knowledgeable employees and find a ton of weird accounting edge cases which explains all the out-of-bounds stuff.

But.. it sounds much better to the population if you scream that you found $8Tn a year in fraud.

Oh, do not for one second think that this behavior isn't intentional.

Elon knows it is BS. But that's not the point, the point is to create outrage, and legitimize it all to the MAGA crowd. It is all intentional misleading.

I love how these people are absolutely prone to even basic "common sense". You need to blame someone for your misery.

One of the oldest tricks. We are still just unga bunga tribal apes some of us smarter that the other but it's still not enough.

Quick Google search says 1.6 trillion will be paid this year.

From this: https://x.com/elonmusk/status/1891350795452654076

The sum of groups age 60-89: 94,950,820

The sum of all groups age 60-369: 121,794,498

From this: https://www.ssa.gov/oact/STATS/OASDIbenies.html

Number of beneficiaries receiving benefits on Dec 31, 2024: 68,455,973

If I understand the post, these are persons who are marked not deceased--it is not stated that they are receiving money, though it may be implied. For example, there are some persons under ten years of age who are receiving Social Security benefits, as having lost one or two parents (survivor benefits). The number certainly does not amount to 39 million.

In the early 1980s my job at Army Materiel Command (DARCOM in those days) was to report energy consumption trends for all Depots, Arsenals, RDT&E, etc. in our Defense Energy Information System (DEIS) database. I was a 30-yr old newbie to DoD, but the numbers were not adding up. Nobody questioned what was being reported, because frankly, they only looked at trends not what the outrageously huge energy consumption actually was. Some installations were reporting almost no energy usage and others reporting waay too much. It took me a while to identify the problem, and much much longer to correct it. In a nutshell, only some installations were reporting correctly: Everything was supposed to be converted to MBTUs. But I found some reporting in 1,000s; others reporting in gallons of fuel instead of converting to MBTUs; others using wrong conversion factors. Elon's team found data that doesn't make sense. There's nothing wrong with publicizing the absurdity of what he found. Explanations will come, but the fact remains - somebody has been asleep at the wheel, and nothing good comes from bad data. Are we overpaying? Maybe so. Is it fraud? Maybe. Is it innocent error. Could be. Could it be that fraudulent SSNs are getting used without detection?? A real possibility.

Don't gripe about Elon. He's doing a job that nobody before him has done. It's about time. Thank you Pres Trump!! Thanks Elon!!

The OIG most recently did the job in 2023, as multiple HN members have posted:

https://news.ycombinator.com/item?id=43083362

Related document for how we used to handle this before the "oligarchs tweeting slander" approach gained support:

https://oig.ssa.gov/assets/uploads/a-06-21-51022.pdf

Numberholders Age 100 or Older Who Did Not Have Death Information on the Numident

> The attached final report presents the results of the Office of Audit’s review. The objective was to determine whether the Social Security Administration had effective controls to annotate death information on the Numident records of numberholders who exceeded maximum reasonable life expectancies. Please provide within 60 days a corrective action plan that addresses each recommendation. If you wish to discuss the final report, please call me or have your staff contact Michelle L. Anderson, Assistant Inspector General for Audit

They decided not to do anything about it (e.g. add a "presumed dead" field) because they thought it would be a waste of money!

> In response to our 2015 report, SSA considered multiple options, including adding presumed death information to these Numident records. SSA ultimately decided not to proceed because the “. . . options would be costly to implement, would be of little benefit to the agency, would largely duplicate information already available to data exchange consumers and would create cost for the states and other data exchange partners.”16 SSA also believed a regulation would be required to allow it to add death information to these records, and adding presumed death information to the Numident would increase the risk of inadvertent release of living individuals’ personal information in the DMF.

Submitted here in case anyone wants to discuss the SHOCKINGLY boring REVELATIONS contained within:

https://news.ycombinator.com/item?id=43077199

I think it's worth reading both the 2015 OIG report on the topic ("Title: Numberholders Age 112 or Older Who Did Not Have a Death Entry on the Numident", A-06-14-34030) and also the 2023 followup you submitted. I left a comment over on that submission after reading both.

It's nice that the hard work of investigating government inefficiency is being noticed and celebrated -- you can really see the tensions between providing reliable services and fighting fraud risk in the 2015 & 2023 reports.

If you care about finding waste, it seems like a really strange choice to summarily fire the inspectors general who have worked hard on this sort of investigation.

They aren't willing to provide the right lies or spin to justify cutting SSA altogether. That is the end goal.

Thanks for sharing. This is what I expected -- of course somebody already looked at it, did the math and decided to not do anything, which was the most reasonable thing.

However, the MUMPS programming language (which is still commonly used in various medical stuff) does use December 31, 1840 as its epoch. (It doesn't have a separate date type, but it does have date-handling functions which operate on numbers and use this as the epoch.)

I worked on a COBOL system in the early 90s that stored a one-digit year :)

However, the records were never stored for more than 4 years, so this was never a problem.

Who needs just one Y2K when you can have one every 10 years!

I keep getting gobsmacked by how much misinformation and straight up lies there are on the internet these days. And what's worse is that I keep falling for it like everyone else, even though I pride myself on being so skeptical about everything. I remember reading that last week and thinking "oh, interesting" and now I'm angry at myself for not questioning that more, especially since I worked at a bank.

With so much manipulated information, AI-generated content, and straight up lying, I really can't tell what's real and fake anymore.

I distinctly remember finally not being able to tell the difference between fake and real info during the Allen Texas shopping mall shooting. I went on Twitter to get more info and I couldn't tell what was real and what was fake for the first time because everything was so convincing. That feels like ages ago now because things are so much more sophisticated.

Can you give us a misinformation example related to this post. This sounds like an empty complaint with no vector.

There's a snarky thread going around attempting to dunk on DOGE employees who revealed that some social security recipients have 150 year old birth dates. The claim is that this is just the default start date for COBOL (like the UNIX epoch), but it's not even true, and even if it was, it's still a major problem if we have social security recipients with no birth date in the system.

I think you have to read this one generously. The claim isn't about how COBOL works, except incidentally. It's more along the following lines:

COBOL doesn't have a default date/time type

As such implementation decisions are left to the implementor

The implementors* of the SS system chose 1875 as the epoch date for reasons

*I made a lot of money in 1999. The original implementors of SS probably used something else ("it'll be rewritten before this is a problem" was essentially the whole raison de etre of Y2K). The 1875 thing, if it's a thing, was probably the result of Y2K work. But I have no direct knowledge of these matters.

> The implementors* of the SS system chose 1875 as the epoch date for reasons

The problem is we have no solid evidence that is actually true. The claim appears to originate in an anonymous DailyKos comment which contains so many factual errors (e.g. claiming this is due to COBOL), it is unclear why any of it should be believed. For all we really know, the SSA code doesn’t treat 1875 specially at all. And even if it actually does, are these social media claims that it does based on inside knowledge of how it works, or just a lucky guess?

That’s not to say DOGE’s claims about 150 year old social security recipients are right - for all I know, they could be wrong - but, if they are wrong, it could be for some reason which is completely unrelated to “1875 as an epoch”

I did a lot of COBOL Y2K work in 1998-2000 and I have never once heard of any 1875 epoch related to COBOL until the last few days. We either modified code to use windowing functions or we just full converted to 4 digit years depending on severity. For instance, windowing code for front end, 4 digit years for DB, VSAM and flat files.

"The implementors* of the SS system chose 1875 as the epoch date for reasons"

So, they invented their own date system instead of just using the standard CENTURY-DATE or an ISO-8601 date from their DB. Highly doubt...

1875 being the result of Y2K work makes no sense to me.

Also the social media claim was that "this is how COBOL works", not "this is how the SS system works". It does not seem that the person who made the original social media claim has any insider knowledge of the SS COBOL system.

Definitely seems like misinformation to me.

Did DOGE reveal anything, or did Musk just make an unverified claim?

If the claim is true and it's really a case of fraud, isn't that a case for the courts and DOJ to handle?

Or he is really trying to undermine trust in the government, like his false Gaza condom claim?

The latter. He's trying to build an "Iraq has WMD" case for cutting SSA, by falsely claiming all the payments are fraudulent.

The problem is that DOGE public statements like those of Trump are a fountain of lies so there is little point in digging into them because the alternative is the equivilent of looking for some edible corn in a bucket of human waste.

Logical theories include connected records include an age, its chaff unconnected to any money being moved retained for legal reasons, and them just making it up.

Since they won't substantiate this you can choose your own adventure whilst waiting for the boring truth which is probably on balance that all receipients have a known age but they were unable to look it up correctly and something somewhere returns default when not available i in that code path.

> it's still a major problem if we have social security recipients with no birth date in the system.

It's not a problem, since there are other ways of determining eligibility. If a person doesn't have proof of a birth date, what are you supposed to do? Make one up?

And the claim is that it's fraud, which requires evidence, not some anomaly which can be several things. Musk and DOGE deserve the "dunk" since they're spreading unsubstantiated BS.

If a person doesn't have proof of birthdate, you could make up a realistic one. Pick 1935. Pick 1942. Pick anything but longer than the oldest human to ever live ago. And you know what? Since centenarians are rare enough, maybe do annual checkups on those folks to see if they're still alive before cutting checks.

Musk and DOGE do not deserve a dunk that is a lie just because they deserve some comeuppance in general.

I’m guessing that when Social Security started being paid out in 1937 that many US citizens were born around 1875 and also did not have a birth certificate. So when the system was computerized in the ‘70’s there was no way to correctly input data so they simply continued that method.

The trick is that while this is a database in Social Security, it is not the database of social security recipients.

Thank, you. All this whining about missing death dates totally misses the point. 150 year olds are not receiving checks!

It's sad that there are so many people that don't realize what they're saying is absolutely illogical. I agree with you, the technical details are irrelevant if there are people getting payments with those attributes it's a problem.

Nope, you're totally wrong. it's entirely a technical issue if it's an old system that handles exceptional circumstances in odd ways. The facts of the case(s) are not known.

You don't know what the recipient's details are, you're just saying it's a problem without knowing any of the facts around the individuals circumstances.

You're simply thinking in the most shallow way possible.

Isnt this related to an unverified claim of fraud, so it dpesnt really make a useful distinction because the misinformation is countering the same.

The stack exchange question is the example. It specifically talks about a bit of information making the rounds that turns out to be wrong. You read the linked page, and didn't just jump to the comments, right?

Except the link doesn't refute anything at all.

It just says that basically each COBOL system will implement dates their own way.

The thread says any such implementation has nothing to do with it being COBOL and further that COBOL doesn't even have a concept of dates, which absolutely does refute the original claim of "in COBOL, if a date is missing...".

It's exhausting. Each side of every argument is full of misinformation, intentionally or not.

Do not both sides any of this. One cannot claim that both sides present misinformation and then not acknowledge that one side is doing so intentionally and the other is not.

Elon tweeted that there was a lot of 150-year-old recipients. That's all he said. [1] So there was a rush to point out why, if this 150 year old number is the only information he's providing of fraud, it is not a prima facie case of fraud. That was a good faith response to a bad faith, selective release of information.

[1] https://www.foxbusiness.com/politics/musk-claims-150-year-ol...

So then Musk provides more data, but again, not enough data to provide all the context. What he leaves out is that there have been multiple, prior good faith attempts to investigate these data entries, identify whether there's any fraud, and address any problems. This was the work of inspectors general whose job is to work in good faith to try to resolve these issues.

There is one side acting only in bad faith. If they were acting in good faith, they would raise these issues through legal channels (inspectors general) and then have an orderly, legal process to address them. That is how it has always been done, for a reason. They are not operating legally because they know that what they are doing is in bad faith and would be found out as such.

What we are witnessing is a dismantling of the rule of law. It's important to recognize that and to not to be complicit in it.

You're replying to a thread on an article about how someone just made up the 1875 thing to own Elon.

There's definitely a "both sides" problem here. Many more commenters are making totally unfounded assertions about how these systems actually work for the same reason.

You yourself are pontificating about the inspector general's report at a level of expertise beyond what you likely have. I have some familiarity with IGs, though not in the SSA. It has been eye-opening to see people crawling out of the woodwork to talk about their role, their effectiveness, their "good faith attempts", etc. They don't actually know any of this: it's just ammo they found online to "get" Musk since the controversy started.

Why is it so hard to just suspend judgement about these claims, rather than attack them with little basis? Or at least go after it for solid philosophical reasons? I can't understand why the level of discourse on this subject on Hacker News, of all places, is so bad.

>You're replying to a thread on an article about how someone just made up the 1875 thing to own Elon.

That commenter on DailyKos wasn't put in charge of dismantling parts of the government.

Musk Was.

The way Musk has acted in the past 4 weeks has caused people here to feel it's not in good faith. Suspending judgement when people are lying to your face is a recipe for getting the wool pulled over your eyes. We have ample reason to think Musk isn't acting in good faith, mainly because of the drive-by posting he is doing and the way he's not working with Congress. He wants to make incredible claims without credible evidence... which just makes his claims incredible.

If Musk were doing what FDR did 100 years ago -- affecting major change by working through Congress -- there would be a different response.

I don't approve of Musk for varying reasons, but FDR's relationship with Congress was a little more complicated than that. If anything, Musk and Trump have not yet come close to Roosevelt's excesses, which are now thankfully mostly forgotten and not used as examples. Roosevelt did not so much as "work" with Congress as directly control it, particularly in his first term. Even in his later terms, he directly embedded his executive staff in Congressional committees. He also had no problem ignoring legislation that did not suit him, and had no problem using the FBI and IRS to harass and destroy his political enemies, which helped cooperation quite a bit. Today we find even a hint of this unacceptable. (We've also now mostly forgotten FDR's war against the media.)

From the distance of almost 100 years, it's difficult to see that FDR was an extreme radical when it came to executive power and probably the most powerful US President of all time. It's a good thing he was mostly a good one too.

> We have ample reason to think Musk isn't acting in good faith, mainly because of the drive-by posting he is doing and the way he's not working with Congress.

I don't think the "drive-by posting" or not working with Congress indicate anything of the kind. Congress has for decades done it's best to do absolutely nothing. Even when just a few years ago they busted the CIA spying on Congress, Congress did nothing. Congress has passed basically no major legislation since the ACA (itself pathetically watered down and passed narrowly) except in extremis. The best you get is one party or another grandstanding in committees about some nakedly partisan "investigation." But if they don't even do anything about a level of corruption and abuse that includes spying on Congress itself, why on earth do you think they'd care about low-to-mid-level fraud/waste?

As for drive-by posting, I think we all know Musk is a fundamentally unserious person, not a deep thinker, and prone to exaggeration or outright lying. That doesn't mean that DOGE personnel aren't looking into and finding things that perhaps we'd be better off without and which the executive has the legitimate power to correct or terminate. I don't see any problem with looking more deeply into Social Security payments. As I said in another post, we've had years of reporting on mass SSDI fraud, and other countries do have small amounts of pensioner fraud. Probably we have some too, and this strange belief that "the IG and Congress would have discovered any such problems" seems bizarrely naive to many of us who have worked in the government.

At any rate if Musk is making false claims, then making our own false claims does not help us or any good cause in any way. Lying because someone else is lying does not make anything better. Almost as bad is silly tendentious "well, actuallys" that dress up as a "fact check" and word everything very carefully as if Musk is totally wrong about something when the substance is correct. Consider the Reuters situation. Here's what Musk tweeted:

> “Reuters was paid millions of dollars by the US government for ‘large scale social deception,’” Musk tweeted the night before. “That is literally what it says on the purchase order! They’re a total scam. Just wow.”

Here's an example of a fact check: https://leadstories.com/hoax-alert/2025/02/fact-check-u-s-go...

The entire story is written as if Musk and Trump are deranged conspiracy theorists...when in fact what Musk tweeted was correct. Yes, OK, actually "Thomson Reuters Special Services" was the one with the contract, not "Reuters", and yeah sure they're all actually ultimately part of the same organization, they own Reuters, but there's a firewall between them, we promise, so...and yeah, it was to study "large scale deception" and "social engineering defense" but actually it's a good thing and-

Long story short, Reuters did have some dumb consulting contract with the government regarding social engineering "defense", which likely was a huge waste of money. Of course, it also wouldn't be surprising if the government was not only paying for "defense" (just as the Department of Defense was not "defending" American from Iraq) but Musk didn't make that claim in the Tweet.

This is just an example. I find it just as bad as what Musk does. If we're trying to educate people telling our own lies and bending the truth to fit our narrative doesn't help anything.

>Each side of every argument is full of misinformation, intentionally or not.

This is just silly information, and is used to sow resignation and demean actual and valid arguments.

There are plenty of arguments and forums where misinformation is treated disdainfully, as it should be. HN used to be one of them, but everything anti trump and musk seems to get brigaded, rather than debated.

I will point out that it is predominantly one side in American politics using misinformation as a weapon. Its the side that brought us "teach the controverry" instead of accepting the scientific reality of evolution. Its the side that made it illegal for the CDC and ATF to do studies of firearms. Its the side that claims to be anti- politics in science while at the very same time politicizing science.

Your statement benefits the side that doesnt have truth on its side, and is therefore harmful.

You're saying the party of "the President is fine don't trust the cheap fakes" is to be trusted?

One party agreed their old candidate was losing it, the other did not so your both sides argument already falls apart

I am making two claims:

1. Not all arguments suffer from misinformation on both sides.

2. The political right uses misinformation as a weapon at a scale that dwarfs that of the political left. Your single example does not counter that argument.

That always was possible and always required critical thinking and further evidence to resolve.

People just don't have the skills, or are being brainwashed into not using their skills.

In this case the misinformation isn't even the attack the mis-informers think it is.

Anybody getting payments with this erroneous data is committing fraud because the choices with that mis-information are that we don't know how old they are. This means they have been 150 years old since they got into the system. That scenario is actually worse than they have been getting payments for some decades after their qualification age.

> This means they have been 150 years old since they got into the system.

No, the theory being made was that they appear to be 150 years old because the marker for "unknown DOB" -when queried- comes out as 1875 by default and that wasn't accounted for / known about by the Special Boys given their lack of knowledge around these systems. It's entirely plausible that they look 150 years old (with a badly formed query done in a rush) but there's an "eligibility" section of fields which says they're eligible because they verified their citizenship[0] in 1952 with their military record[1].

Basically doing a query on DOB/age instead of eligibility. Easily done! Just not something to brag about in a press conference before triple-checking that you weren't about to make a massive arsehole out of yourself to the world.

[0] Or however it works, I'm from the UK, I don't know.

[1] Or equivalent validating document.

In the attention economy, what you're saying being true has no relation to you getting paid (and reality is often more boring).

This is what you get for wanting everything to be "free as in beer" and "free as in speech" at the same time in a market economy. Systems optimized to profit from free information have reached great efficiency.

Ignorant. When I learned COBOL, which would have been contemporary to many of the original systems, data types were numbers of various formats, characters, primitive fixed length strings, and bits. There was no data type for dates. It would have been roll your own. Pre UNIX being mainstream we used '7-4-5' dates in assembly based from 1900 (this was a financial business but not the IRS).

This isn't a COBOL issue (if it's even an issue at all), it's a data design issue. As many have pointed out, there are reasons for this possible origin date.

Not on the System (Wang VS) I worked on in the 80s. Plus in many cases dates used a 2 digit year. So 1875 would be seen as 1975.

Also I never head of this default.

As I just posted in a thread https://dev.to/mdchaney/cobol-dates-may-20-1875-and-disinfor...

Nobody in this HN thread has used the word "sentinel" - see another HN about the concept https://news.ycombinator.com/item?id=36195425

People got hung up on: - "COBOL defaults to..." rather than "banking practices are..." - epoch start dates - many pointing out COBOL didn't use epochs or counts, just much-damned YYDDD or YYMMDD actual strings.

Also, Elon loves to stir with partial misinfo hence his tweet https://x.com/elonmusk/status/1891350795452654076 with the breakdowns by age bracket. "Death set to FALSE" means "Death date not known" but that's not clickbaity enough.

That long tail looks awfully like data entered from historical records lacking death dates - there have been a few discussions of the cost of finding death dates and the decision to avoid spending $millions on it, as this is not data used to make payments.

You would expect, in a system that's pulling data from many sources, to see historical jumps in data cleanup like this. Imagine a few large states finally get around to digital records of deaths, so their data is easily aggregated - you get a sudden flushing of people who would previously have been left on the list. However, this will only apply from a certain age onwards as those sources in turn don't have the time/budget/interest to digitise really old records.

I'm late to the party but want to add that "reference date 1875-05-20 defaults to 150 years" explanation makes no sense to me.

Assuming the reference date is correct and unknown birth date is stored as timestamp 0. Since we are before 2025-05-20 at the moment the reported age has to be 149 years, not 150. What I'm missing? Would be very unusual to round up age.

Perhaps there is a default date of 1875 if the date of birth isn’t known. When slavery ended just before 1875, a large number of them never received an actual birth certificate while enslaved, and likely didn’t know their date of birth… so it would go with the system default.

Related, "Why COBOL isn't the problem":

* https://lucid.co/techblog/2020/11/13/why-cobol-isnt-the-prob...

* https://news.ycombinator.com/item?id=41420217

No! It does not - Someone coded it to default or entered the data that way, and no edit prevented it - and it sounds like no regular maintenance (even a reporting process) is present to indicate ages > 100 (which may raise a flag to be looked into).

The question of unknown and how to default this value in databases. Since null is not usually possible, there have been hacks on how to do it. 1-1-9999 anyone?

Why are we taking Musk at his word when his current MO is to cast doubt and mistrust on government spending?

Musk, Trump and the admin have already been pushing so many outright lies via their propaganda channels. Despite being deunked the lies are repeated nonstop.

The lie that USAID spent $50 million on condoms in Gaza (the money was for running hospitals) for example.

Or the lie about funding an opera about a transgender woman (the money was for a university in Columbia, unrelated to a performance put on at the school).

They are spewing lies left and right. Lies go twice around the block before the truth has put its pants on.

Why should we believe him on this?

> Or the lie about funding an opera about a transgender woman (the money was for a university in Columbia, unrelated to a performance put on at the school).

You've got that wrong. The grant to fund the opera exists: https://www.usaspending.gov/award/ASST_NON_SCO20021GR3086_19.... Specifically the grant was given "to raise awareness and increase the transgender representation through the opera As One, by American composer Laura Kaminsky".

The lie in this case would be that USAID awarded the grant, it was awarded by the Department of State instead.

Probably will go unnoticed or downvoted because of twitter link but here is the main goal of spewing such a bullshit:

https://x.com/KariLake/status/1891841704703013067

Look at the comments. They know that COBOL defaults to 1875, point is that propaganda pundits look for any reason to spew toxic misinformation and rile 99.999% of uniformed sheeple that voted for orange felon.

It's horrifying to see even the Associated Press repeating this claim a truth. AP links to the Wired article, which links to the same old ExTwitter post that originated the claim. Neither did any fact checking ffs

Elon Musk is a liar with no credibility whatsoever. Don't believe anything he says.

Musk said: “Crazy things like just cursory examination of Social Security and we’ve got people in there that are 150 years old.”

He qualified this claim as a “cursory examination”. It’s clearly a comment about the quality of the data and systems. That this is the kind of thing that would be prone to fraud.

Before you hit downvote, please provide evidence that you didn’t hallucinate Musk’s claims here.

But then he goes around saying he actually found massive evidence of fraud. It’s not like he tweets: “Guys, my admittedly cursory examination gave me a feeling this old system which I’ve never seen before could be prone to fraud.”

What proof does he (or you) even have that a COBOL system is particularly prone to fraud? The world’s most important banks still run many things on COBOL. Are you saying that bank mainframes are full of IT fraud?

Who said a COBOL system is prone to fraud?

Incomplete and inaccurate data is prone to fraud. Talking about COBOL is missing the point.

The point is that a date set to 1875 is actually a null in COBOL.

It doesn't mean the data is missing or inaccurate, any more than a null reference in Java means your program is going to segfault. It simply indicates that the data is not present in this scope.

If Musk was posting: "Guys I just discovered they have tons of null checks in their code here, that's obviously an indication of fraud!" — would that make any sense to you?

The part of all of this which I have a problem with from the outside, is that it seems extremely irresponsible to address the White House and the public to say you have found people that are 150 years old receiving social security payments in such a provocative way. As if you have so simply stumbled across this error that has gone unnoticed.

There’s been no formal inquiry, there was no mention of any of the checks and balances that may have been occurring, there’s been no nuance to the argument from him at all. Okay maybe there are people that look to be 150 years old, is there a reason why? Were they actually being paid the social security or was there a legitimate exception? Maybe people were but it was so few and far between and was an internal controls issue which all governments and companies have globally or was none of it true and your team simply assumed they were being paid.

After reading an article where Gov security experts were worried because they were having to give Elon’s team access to Putty and SQL tools. It seems like people are going through this data and making inferences that may not be completely true or vetted.

> … it seems extremely irresponsible to address the White House and the public to say …

It’s simple as:

1. They don’t care;

2. It fits their agenda;

Exactly. Statements like this have been used to approve new programs and legislation that in at least one case led to sweeping profiling and the incorrect retraction of government benefits, causing tens of thousands to end up in poverty and debt and the consequent issues of divorce, children removed from parents, even suicides: https://en.wikipedia.org/wiki/Dutch_childcare_benefits_scand...

This all started when a group of Bulgarian migrants took advantage of a loophole, registering as residents for a small amount of time and defrauding the government for about €4 million. The price tag to set it right again was €7.4 billion as of last year: https://nos.nl/artikel/2503966-de-toeslagenaffaire-van-een-m...

What I'm trying to say is that right-wing / populist talking points can and will lead to sweeping reform with long term and far reaching consequences. For a few years we had an era of fact checking where these statements were immediately marked, but all the services have removed it again. I'm honestly surprised Twitter still has community notes.

And there's no punishment for being wrong. Even a newspaper has to publish official apologies, not that anyone bothers to read them. He could say the sky is red on that platform and there's nothing anyone could do to force him to change it or admit he's wrong.

Problem with this is, if or when they are proven wrong, it will be very privately, no one will ever hear about it. It will be lost in the noise of “we found lots of fraud”. And no one will be able to get as straight an answer as the accusation. So no actual consequences for difamation. If feel like this is an issue with all politicians. They accuse very loudly, admit wrongdoing very quietly or never.

Reminder that Musk once wrote an email to buzzfeed stating "“I suggest that you call people you know in Thailand, find out what’s actually going on and stop defending child rapists, you fucking asshole", about Vernon Unsworth, the man who saved 12 children from dying in the Tham Luang cave. He had no proof and went on to accuse him of not even being a diver.

https://www.buzzfeednews.com/article/ryanmac/elon-musk-thai-... https://www.ccn.com/5-saucy-elon-musk-revelations-in-explosi...

And then won a legal case by arguing that it was just a generic insult not a specific allegation (and having expensive lawyers).

Have they revealed how many people actually fall in that category? Agreed that more nuance is necessary.

He recently shared the usual X style cropped screenshot which implied millions of records "where death is false" and the rows age appeared to be 110+

As someone who uses databases everyday, there's always nuance to data, especially when you're trying to simplify data for a small report. Hard to believe there are millions of records so obviously wrong.

With US not having centralized vital records registry and also not having demographyc registry and not issuing id cards out of principle -- it's pretty easy to have consistency issues on the scale of millions.

Add the fact that people also move to other countries and US government may not even know somebody died, because death was not reigistered in US.

I wouldn't put it past our government. Who ever even looks at this stuff? Nobody cares to look when they're spending other people's money. Traditionally the play has been: don't rock the boat and you keep your job for life.

Office of the Inspector General looks at this stuff.

Trump fired 17 of them:

https://campaignlegal.org/update/significance-firing-inspect...

> This blatantly illegal and incredibly concerning mass firing removes the only independent offices within agencies designed to protect taxpayer money and root out corruption, fraud, waste and mismanagement.

I also posted the 2022 report the office generated elsewhere in the thread, a follow up to their 2015 report.

At the very least Social Security should try to get the correct birth dates and verify these folks are alive. Because it sounds like no one is checking and fixing this. If you were running social security and you see people have no date of birth and they come up as 150 years old, you fix it. Apparently that basic thing is not happening and Elon musk is calling this out. I see no problem with fixing it.

There's opportunity cost in fixing it. The problem is we don't know if it's actual fraud. The SSA might, or they might not. It might not actually be a problem, or a problem that is an acceptable cost of running the agency. It might even be a cost that's acceptable to the country, given all the complexities involved (that we don't know about)

That's the problem with ALL the things Musk is coming out with. We just don't know! We don't know how much we are "losing" to them, we don't know know how serious these problems are. We don't know the context.

We're just getting thrown little bites of information. They sound ridiculous! They are! A lot of things can sound ridiculous if you say them out of context. He's saying a lot of stuff where the listener can go "well, yeah, obviously we should fix that". It's the "common sense" play of the right wing. Yes, it's hard to disagree with. It might be right, and we might need to fix it. It also might not be the best use of our time, energy and money to fix. We don't know what the opportunity cost is.

What are you saying or referring to? The Stackexchange posting you're replying to does not refer to the U.S. government at all as far as I can tell. How do you connect the link you're responding to to the White House at all?

https://nypost.com/video/elon-musk-claims-some-recipients-of...

Musk doesn't care about

- correctness of information

- completeness of information

- nounces of formulation

- accidental misleading people

he has more then once verbatim repeated or re-posted conspiracy theories, including trivially disprovable which also don't pass the common sense check (i.e things you only "fall" for if you either act extremely reckless and negligent or intentionally fall for it as it pushes your ideology no matter how wrong it is)

„Social Security isn’t paying out to 150-year-olds, it’s paying out to people whose birthdays we don’t even know“ is not exactly reassuring.

How is it possible that the db contains records with such a vital datum missing? How is it possible that SSNs aren’t unique? I can’t come up with an explanation that doesn’t boil down to either negligence or fraud.

You ask in effect: 'How is it that a data collection

- whose cardinality is well into the hundreds of millions, likely close to half a billion (500,000,000);

- which has been maintained for now just about exactly a hundred years;

- which predates birth certificates and birth records being common across large swaths of the country (not predates their systematization or encoding, but predates their reliable existence);

- which for a lot of its history has been maintained by hand,

should come to have occasional inconsistencies?'

So framed in knowledge of the Social Security Administration's history, I confide the question may reveal its own answer.

Inconsistencies make sense.

But it also seems to fine to say "we aren't paying out any more money for any inconsistent records till the person comes forward and gives us the info to fix the records".

Obviously, if such a thing is done at scale, you need to have the staff to handle all the phone calls etc in a timely manner.

Until what? The person produces records that don't exist?

Someone born in a cabin in the middle of the woods 90 years ago is not neccesarily going to have "info" to fix their birth records (especially if non religious, so there are no church records). They could give you a date, but they aren't going to be in a position to prove it.

Are we optimizing for nice and clean database table constraints over paying out welfare to people who need it?

back in the 1920s and 30s it wasn't uncommon for people to be born at home and social security numbers wouldn't have been obtained at birth because there were no social security numbers yet. In some cases data of birth is uncertain e.g. summer of 1934 or 35. IN such cases any written record is likely to be at least unofficial and more than likely non-existent.

It's not the fault of the person, there is no legal basis to do so and it simply doesn't matter. Eventually everyobe just does anyway.

I'm confident they don't just pay someone if their data is inconsistent; the database is not the only source of truth.

The next question then becomes, how much does it cost to fix these inconsistencies? I'm also 99% confident that there's a small army of people that work to do so already, that is, fix and work with and around the inconsistencies.

> which predates birth certificates and birth records being common across large swaths of the country

All true, but you underestimate severity of the problem by a large margin if you assume that it is confined to the USA, a fairly affluent and bureaucratically stable country.

People in USA databases aren't necessarily born in the USA. Refugees by definition don't come from situations with stability.

I can't tell what argument you're trying to advance here.

That's not what he's asking. He's asking why there apparently isn't a standard process to locate and clean missing data through investigation, assuming that's the issue.

Bear in mind that social security fraud is a major problem in many countries and impossibly old people is a usual indicator. The famous "blue zones" that were once studied for their long lived people are now believed to mostly be an artifact of undetected pensions fraud.

> That's not what he's asking. He's asking why there apparently isn't a standard process to locate and clean missing data through investigation, assuming that's the issue.

If that's what he was trying to ask, then I have, as detailed above, some notes on the attempt. But the myth that Social Security numbers were ever designed to work even half-assedly well as a single national ID number will never die, I suppose.

> Bear in mind that social security fraud is a major problem in many countries and impossibly old people is a usual indicator. The famous "blue zones" that were once studied for their long lived people are now believed to mostly be an artifact of undetected pensions fraud.

I can't really evaluate Newman's preprint [1] with much confidence, but it's only been out since September, and the Ig Nobel isn't peer review. Given also that the US state with the largest count of "supercentennarians" in Newman's data, Iowa, has a whopping 37 [2] of them, you're stretching this piece of un-peer-reviewed research far past anything it could reasonably be asked to support - hence, I suspect, the "are now believed" passive-voice weaseling in your claim quoted above.

[1] https://www.biorxiv.org/content/10.1101/704080v3

[2] As given "1.25e-05" (p. 5, ibid.) or 12.5 per million, for a population just over three million. I've taken the liberty of rounding down on the assumption we are not meant to be counting fractional supercentennarians.

>a standard process to locate and clean missing data through investigation

I understood GP as saying that this data is the best, cleanest data available. If nothing better exists, then record what you have. If that means some fields are blank, well then they're blank - you'll have to deal with this for country-sized databases, one way or another.

>now believed to mostly be an artifact of undetected pensions fraud

Im genuinely curious about this, would you mind sharing references?

But mostly given that the primary condition for receiving social security is age, it doesn't inspire confidence that they are making payment without having any age data.

> social security fraud is a major problem in many countries

It is not a major problem, it is a minor problem blown completely out of proportions by right-wing propaganda to cover much bigger problems caused by power-grabbing, tax-evading plutocracy.

> a standard process to locate and clean missing data through investigation

What would that process look like assuming you are e.g. trying to establish a DOB of a person born long time ago in a country that does not exist anymore under wartime conditions? Would that be cost effective?

According to the Social Security database there are 20,789,589 living people over the age of 100.

According to the last US census that number was about 90,00.

Are you ok with sending social security checks to 20 million dead people?

Do we send Social Security checks to everyone in the Social Security database?

> How is it possible that the db contains records with such a vital datum missing?

It might not be missing, it also can be unknown and unknowable. There are people with their date of birth not recorded or recorded only approximately. It happens when the date is recorded on a piece of a paper in a building that then burns down or during natural or human-made catastrophes. Or it might not be recorded at all, or only some parts are missing (like the day of the month). Working for a different public administration, it's so common we have a special date type for that.

Or a definitely wrong value was removed but there's nothing to correct it with.

For example, if someone was registered with a birth date of 02/18/1458, or has a birth certificate with that on it. Bear in mind that humans make mistakes, and the people who issue birth certificates or type them up are humans. 02/18/1458 is patently wrong, but what do you do? Just guess that someone misread a 9 as a 4? What if the person in front of you is clearly too young (or old) to have been born in 1958?

You can't just revoke someone's social security because someone screwed up.

> How is it possible that the db contains records with such a vital datum missing?

This is a weird question. How would you even go about finding out an unknown date of birth? Especially when it happened > 50 years ago and half the world away.

For whom is it "vital" ? Will it prevent me from going about my daily business?

See also: Why are so many people born on the 1st of January? It's a statistically impossible number.

A: Because they know their approximate year of birth, only.

I know a few people that don't have an official birth date. My brother in law is Papuan and no one in his village has an official birth year or date since they just don't keep track of it, so when he got his first official government ID at a much later stage of his life, he chose an arbitrary date cause it means nothing to him. He's probably around 35-40, but his birthday says he's in his 60s.

It's not all that rare even in the West, especially for older people. Records get lost or never get collected in the first place.

You think it is not possible that people born in homesteads in the 1920s and 1930s in rural and isolated communities don’t have birth records? Or immigrants from countries that didn’t have paper birth records?

My grandmother was born in the 1930s in Chicago (second most populous US city at the time) and she never even knew exactly what year she was born in. It’s fairly common.

Looks like you never touched any system with real life data

Edit: the link below gives more information on ssns. SSNs so far been unique, but there are some issues

SSNs are not reused: https://www.ssa.gov/history/hfaq.html

I don't know how it happens in the US, but in France, during several years, I had an error-code SSN. My papers had been misfiled at birth, due to the unreadable writing of the cleric who recorded my birth, and it took 2-3 years to fix this.

This error-code SSN was full of 0s, so obviously shared with others.

I wouldn't be surprised if some migrants waiting for confirmation of paperwork also had some shared temporary special-code SSN.

Thanks, very informative link

Thats not how SSNs are supposed to work at all.

Yeah, who's ever heard of an incomplete database or a database that barely holds together and has had to have weird adjustments done to it in order to avoid fucking up other bits of data?

We can tell you've never touched a database or worked in tech.

Not that there is any shame in that. But the same way I keep my opinion to myself when my doctor gives me my diagnosis, people who have never handled a database really should follow that advice.

A local library can barely keep its database straight and it's very slightly easier than a database containing millions of people's information. Dozens of millions.

How is it possible that your code contains bugs? I think that question may have the same answer.

So the bottom line is that a left-wing activist put out some lies in order to discredit Musk, and every single reader of the original misinformation piece swallowed it without any attempt to analyze it critically. And they say that right-wingers are the ones who are more amenable to propaganda?

Yet everyone's conclusion is, as usual, Musk bad. Nobody cares about left-wing fakes news if they are useful to the agenda being pushed forward.

One cannot even begin to address the amount of misinformation that has come from the left since Trump has been elected.

This is why, however, people in general act with decorum. Trump and Musk are making themselves easy targets for this vitriol. Leaving people with absolutely no will to critically think about any of the claims levied against them.

In fact, any claim levied against them is welcomed ammunition. Any attempt to reduce this ammunition is considered support for the enemy.

During Trump's first administration people decried that we had begun living in a post-truth society. That accusation was made against the Trump supporters.

Now, during his second term, we truly live in a post truth society. Where in most cases the truth simply does not matter to most people. This division seems to run so deep I am afraid at what it means for our society.

One thing I've noticed is when Trump says something the left starts by denying it before using it themselves. Fake news was coined by Trump and now it's adopted by the left but they call it misinformation, and they even started the Trusted News Initiative during Covid to weaponize it. It's not lies, it's "misinformation", true, but wrong?

Confusing times and the only conclusion I can draw is that those in power are losing ground thanks to the internet and it makes them behave like this. Everyone in willful ignorance. Even otherwise clever people are acting bat-shit crazy as soon as Musk is mentioned, but a few years ago they would've bragged about getting a Tesla using their NFT profts.

Imagine all the wasted cycles dealing with Trump's/Elon's/DOGE misadventures.

I assume this was in response to Elons claim of finding a number of people aged 150 years old and the response that this was due to COBOL default date and how this was an example of the people doing the datamining were incompetent.

Someone would have to be really incompetent to find a spike at precisely 150 years old and not investigate it further. Elon tweeted ~ 10 hours ago the age breakdown and there does not appear to be a spike at 150 so if that information is correct then this is no longer evidence of incompetence.

I fail to see why anybody would believe anything coming from Musk's mouth.

Are you trying to tell me that, eg

> $10M for "Mozambique voluntary medical male circumcision"

is a complete and utter fabrication, designed to make the federal government look stupid, and has little/no basis in reality?

I'm shocked, I tell you. Shocked!

https://x.com/doge/status/1890849405932077378?s=46

I can think of two good reasons such a thing may be a good idea in such a place, if you can’t, you’re not qualified to second guess the decision.

Wait, I thought they sent 200 million condoms over there.

He was making fun of someone for suggesting that the federal government uses SQL.

Dude has been burning credibility at an amazing pace.

probably because despite all of his negative aspects, he has transformed the online payments industry, space industry, and auto industry.

All that while playing Diablo and posting 200 racist memes a day....

lol I think the Diablo Rakki was determined to be a hoax but I did find this interesting paste of Things Elon Musk launched or unveiled in less than 10 years:

Boring Co. Crew Dragon Cybercab Cybertruck DOGE Falcon Heavy Falcon 9 FSD Grok 1,2,3 Megapack Model 3 Model X Model Y Neuralink OpenAI Optimus Powerwall Roadster 2.0 Robovan Solar Roof Starlink Starship Tesla Semi

And yet millions still do. People believe what the WANT to believe, and that frequently is something other than facts.

People believe stories. It's ridiculous to suggest otherwise.

As competing narratives become more compelling, people switch the stories they believe.

So if you want people to start believing in the set of facts you hold to be true, tell them a compelling story about those facts, don't just tell at them for not fact checking.

"People believe stories. It's ridiculous to suggest otherwise."

No one suggested otherwise, in fact, that's virtually verbatim what I said.

Additionally, the data is comparable to the census [1] which suggests there is not a lack of data context or translation from DB date format to actual date/age.

Then again, these are just posts of screenshots on a site filled with AI-generated content.

[1] https://x.com/jonatanpallesen/status/1891406665242546383

Do they receive social security though? His tweet doesn’t explicitly say so, might be willfully omitted.

I don’t know, incomplete information all round.

I consider it plausible but also consider Elon to be a rather unreliable source. Part of this seems like JJ Abram's mystery box storytelling where new mysteries are opened before the old mysteries are resolved - and when finally resolved the resolution is deeply unsatisfying.

My experience with government is that where it’s not recklessly incompetent it is flagrantly corrupt, and increasingly so. To an extent I would not have believed had I not seen it myself. I completely understand why others may not share that belief. While I am in general support of taking a chainsaw to government, as opposed to a scalpel, I do wish it was better people doing it. That way when such crazy statements are made I could consider them likely true because the person stating them would not have done so without it being true. Elon has told too many falsehoods for me to give him the same benefit of the doubt.

When you say "government", do you mean the USA civil service, the subject of Musk's ire?

>"where it’s not recklessly incompetent it is flagrantly corrupt" //

Care to give a couple of the examples from your direct experience?

Just run of the mill stuff, pay-to-play to such an extent that everyone in a public service department that I sell software into all have jobs with vendors lined up for after they retire. They are very open about it. Most other vendors have been investigated for bribery by SFO, PNF and or FBI. Only a few stuck, I think they just bribe more people to make the other bribery charges go away. One that did stick was KBR, a subsidiary of Halliburton. Most vendors in my space seem to operate on a pyramid scheme where the money they take in from new deals has to pay for the consulting fees of the predecessors. I’m more interested in making good technology than dealing with the stress of all that so I try to compete on technical merits - which is tough.

It’s a pretty pathetic argument that you support wielding the chainsaw, but preferably from a more desirable tyrant. Nobody who would agree to wielding the chainsaw in a democratic society is worthy of the privilege.

People who share your nihilist perspective combined with solutions contrived from ignorance are why we’re here. And where we are going will redefine corruption.

It's absolutely willfully omitted. I guarantee 0 evidence of anything will ever come out, just shock headlines to make it seem like they're doing something.

An audit was completed in 2023 (using June 2021 data) where they found that of the 18.9 million aged 100+, approximately 44,000 were receiving payouts. UN data says there were ~90,000 centenarians in the US in 2021.

https://oig.ssa.gov/assets/uploads/a-06-21-51022.pdf

https://www.bumc.bu.edu/centenarian/statistics/

Social Security is incredibly complicated. There’s a canonical answer to every question. Someone like Elon doesn’t know or care about the details.

There’s also a number of programs. SSI, Disabilty, Death Benefits, retirement, etc. Blind people are eligible for a benefit by law.

Is there, somewhere in the United States, a disabled, blind child whose parents are deceased and whose birthdate is unknown? Almost certainly yes.

Is that fraud? Only if you’re a “genius” who targets victims who can’t fight back to demonstrate his intellect.

They might, but that's irrelevant in a way. If they're 150 years old, they're dead, and if you claim the OASDI benefit of someone who is dead, you're committing a crime.

A brief search suggests it's inheritable in limited circumstances. Are you saying that's false?

I'm recalling the posts about the last person claiming civil war pensions dying in 2020 (155 years after the war) and wondering if there is anything similar going on here.

War pension claims like that were common and not fraud.

The pension is claimed by the wife of the veteran , it was a practice to marry a very young girl very late in life partly because of the pension.

https://en.m.wikipedia.org/wiki/American_Civil_War_widows_wh...

The person you are referring to was eligible as a disabled child of veteran https://en.m.wikipedia.org/wiki/Irene_Triplett

There are many age curiosities out there like the fact the grandson of John Tyler the 10th president who was born in 1790 is still alive, I.e his grandfather was born 235 years ago!

they are just that curiosities, not really fraud and even when it is one, it never significant just like voter fraud never turns out to be one .

If it's designated by law to be inheritable then it's a moot point.

It's also worth noting that OASDI is not the same as a Civil War pension. One was set up to support men who fought in service of the Federal government, the other was set up to keep the Old Aged, Survivors, and Disabled Insured. They're legally separate programs and thus have different statutes regulating who can draw benefit from them.

Even if that is the case, there should be an easy way to query that. even If they update the systems, and discover zero fraud or bad data it’s still a net win

I find the stack exchange interesting, but I still think its post-hoc reasoning. I would prefer one of the people who worked on these systems to say it, than have a plausible, but none-the-less third hand take published as "the answer"

I didn't wok on one of the US government systems, but much of my career was in COBOL compilers (writing run time systems, and checking against the ANSI standards). COBOL would not default to 1875 for a corrupt date.

> COBOL would not default to 1875 for a corrupt date

But COBOL doesn't have a date type.

So isn't its behaviour defined by how the system has been implemented ?

We already have "the answer", it's obviously bad or exceptional data. The question "is it fraud" can't be answered by looking narrowly at that data.

But the idea that no one has noticed the problem and investigated it (including possible fraud) is very unlikely.

Agreed. And, the persistence of bad data like this isn't uncommon. We had a bunch of fields in the SQL registry system I work in set to the literal 4 character ascii string "null" because of a coding error over field == null by somebody. It did no harm to the attributes of interest at the time. It only had consequences later on when the codebase changed.

Reply on news.ycombinator.com