Yes because according to my link above, they have already used a few leading 9s. They’ll certainly go to the 9xx series before the whole thing breaks. 666 starting SSNs are excluded because obviously it would cause more trouble than it’s worth.
There appear to be around 450 million numbers left and 5.5 million given out a year. It’s not just births. It’s also anyone who comes to the US to work. If the rate of new numbers doesn’t change, we’re over 80 years away. Of course that’s not realistic but we probably have around 50 more years.
Examples of short-term thinking are all over the place. How much could it cost to add a few more digits to the SSN right at the start to avoid enormous conversion costs in the future? When encoding dates in computer databases, why not use just two digits – after all, the twentieth century will just go on forever, right? And when assigning internet addresses, how convenient that most computers have 32-bit words – there can never possibly be more individually addressable devices on the internet than can be expressed in the segmented fields of a 32-bit word, right? Also, “no one will ever need more than 640K of memory” – Bill Gates.
Doubtless they thought it was a problem for future people to solve and, since they had no conception of a computer, probably figured the solution would be simple like adding one more digit to the string. A human would easily see the difference in a 4-2-4 string from a 3-2-4 string. No need to future-proof at the outset. The problem is easily solved if that day ever comes.
It is easily solved now too but it is a Y2K problem…everyone needs to move to the new, bigger number at roughly the same same and that costs money and companies HATE spending money so will push back. It will not be solved until the last possible moment when it simply has to be done. That time is many years in the future though.
My grandfather was born in 1896. He had a SSN. I don’t know when he received it, but he drew SS benefits for many years until he passed away in 1991.
I’m trying to figure out if there is any document that still contains his SSN these days. Any bank account or other financial instrument with that number has long been closed. He owned some land when he died, but that was sold long ago as well.
My point is: why couldn’t his number be re-used today? Is there anything out there that might cause a conflict if a person is issued that number now?
It seems like it wouldn’t cause any problems - but my husband recently applied for SS, And there’s a letter on the website informing him of how much his benefit will be and when he will get the first payment. Further down, that letter makes reference to the payments that were stopped in 1985 - which is accurate, he was receiving benefits until then as a survivor. . But it might be confusing if some kid born today got your grandfather’s number and gets a letter about their benefits years from now referring to your grandfather’s benefits from 1991. Which doesn’t seem like too much of a problem - after all, anyone looking at a record that says a person born in 2025 was receiving benefits in 1991 should know there’s an explanation that doesn’t have someone getting benefits decades before they were born. But we just saw something similar become an issue.
Re-using SSNs has essentially the same problem as making them not be 9 digits. Namely that every computer system in the USA that deals with SSNs would need to be modified to account for reuse. Or account for new e.g. 14 digit numbers with a check digit. So if the IT effort has to be spent, spend it wisely.
The SSA can’t just start re-using them willy nilly. They’d have to notify the world of their intent to begin reusing them. And wait 10 or 15 years for everyone to update their systems. Then they could begin re-using.
So how about instead of that kludge, they do something smart and instead announce that starting in 10 or 15 years all new SSNs will be something completely different, e.g. 12 numbers and letters incorporating integrity checks. And cause everybody to have to spend about the same amount of IT effort updating systems to support those new numbers.
Interestingly, passport “numbers” are now alphanumeric. I’ve recently renewed some passports for members of my family, and the “number” now starts with the letter “A”.
Well, I wasn’t involved in that decision, but I just happened to be involved in IT consulting with a major bank just as the year 2000 was approaching. My work had nothing to do with Y2K, but being in the middle of Y2K paranoia was quite the experience. Essentially the whole gigantic IT department was in lockdown. No new software, no updates, nothing could be released without emergency approval from a high level of management, and had to be certified as Y2K-proven.
So what I can tell you with certainty is that the yokels who made the two-digit decision did not anticipate that whatever tiny cost savings they may have achieved would be much more than offset by many billions of dollars in direct and indirect Y2K costs, including the huge costs of lost productivity.
It doesn’t really matter whether he said it or not, the reality is that this really was the mindset at the time. I could certainly have believed it – I came into PCs from a background that included minicomputers where 4K 12-bit words of memory was plenty good enough to contain both a high-level language interpreter (FOCAL) and the actual FOCAL program itself. 640KB of memory? Why, that’s practically infinite! Who could possibly need more than that? What in the world would you do with it?
I was an engineer in the disk drive industry from 1988 - 1998. OK, a little later than the supposed Gates quote. The standard drive when I started was 80MB as I recall. We knew damn well (our business depended on it) that capacity would continue to increase similarly to Moore’s Law.
Just a rough back-of-envelope calculation - still 450M numbers (or eventually, 550M).
Population is 340M, and assuming life expectancy avg 76 years, to replace this population will need 350M numbers in the next 76 years. Add in immigration, temp workers, etc. - let’s say that 450M is roughly good until Y2.1K…
Alternatively using 5.5M in 450M gives 81 years. within the same margin.
This gives any future expansion plenty of time to warn and adapt
Keep in mind, in the early days both (a) memory an even disk space was expensive. and (b) RAM was made by hand, 1 bit at a time. I just missed the days of writing COBOL with program overlays because the full program did not fit in RAM. A co-worker learned 360 Assembler back in the day because his programs in COBOL were too big to fit in the 40K of RAM their mainframe of the day had. Even disk space was expensive. Making things work was higher priority than worring about Y2K - after all, it was 30 years away, who was still going to be using those programs then? Later they found that weaning off those old programs was the nightmare it turned out to be. (Notice our issues with Air Traffic Control and even the nuclear missile control.) New programs were built relying on those old programs for a back-end. (Especially CICS terminal access to what were once batch datanbases) People may have started adding programs that accounted for 4 digit year, but the trick was to be sure that nothing buried in the myriad of programs still assumed “19”. I remember diagnosing an RPG program to discover it did not know the decade, it simply had a number hard-coded that had to be recoded as “8” once it became the 1980’s.
The joke about RAM and ever-growing programs was “Intel giveth and Microsoft taketh away.”
.
