Are social security numbers recycled? Something to throw into the mix is that there are probably a lot of people like me who use us multiple numbers.
I only have one now, so they tell me, but I had an earlier unused one.
When in high school I wanted a summer job with the city parks department. They made me get an SSN.
The next summer they couldn’t find my old paperwork and just had me get a new number.
Eventually I thought to ask about this and the credits were combined. But the old number was not reused, presumably.
There must be a lot of people who don’t recall their number when changing jobs and just get a new one.
Well, no. Most people properly follow the Social Security regulations and just ask for a replacement SSN card in this situation. (Which would be faster, anyway.)
You signed the application, including the part where you stated that you had not previously opened an SS account. So, technically, what you did could be called fraud. They won’t do that, of course. Because you yourself informed them of this, and had the accounts combined. And because they don’t search for this situation – they have no incentive to; if left uncorrected, it increases the value of the Social Security Trust Fund; you opened & paid into 2 SS accounts, only 1 of which will ever pay out benefits.
Cecil’s idea of just “look the thing up in the Big Book” really isn’t feasible.
First, until 1973, there was no such one Big Book; there were individual sections of this book at the hundred or so Social Security offices scattered across the country. So to verify a number, you would have to make a long-distance call to the appropriate office, somewhere across the country. For every single number. That’s a lot more effort & expense.
Secondly, even after 1973, the technology to “look the thing up in the Big Book” was not very practical. This would have involved an online database of all Social Security numbers; most data processing at that time was done with batch tape files. And to do the realtime lookup to verify the numbers would have required an online network of inquiry terminals, a very expensive proposition at the time. And with no increased income to the SSA to cover this expense. (To say nothing about the security & privacy risks involved.)
Just working thru the algorithm to verify that this SSN is a valid one, and reasonably likely to have been issued in that state at that time is easy & cheap to do. Oh, and it isn’t just “some sharp-eyed sleuth at SS HQ” – every bank & financial institution, and every large employer does this regularly.
Not generally – not that late – though it wouldn’t surprise me to learn that government work still was. Similarly, there were a lot of in-house on-line systems by then, but I won’t swear to that when it came to government systems.
This is a false aspersion on your part. You are probably a lot younger than I am, so of course you don’t know what you are talking about, but I can assure you that in those days they didn’t ask if you had a prior number.
There appears to be at least one typo in the state assignment list in Cecil’s column:
…
486-477 MN
478-485 IA
486-500 MO
…
Unless 486 was used for both MN and MO, that line is probably wrong. (I know for a fact that 486 was once assigned in Missouri.) Also, the 486-477 range is the only one that is reversed in sequence (hi-lo instead of lo-hi) in the list. My thought is the MN line should be “468-477”, with the 6 and the 8 transposed.
Well, he did say that a few numbers overlapped states.
NO, it’s a typo. Minnesota is 468-477.
OK, so what did you answer on question #10 on the SS-5 Social Security Application Form: “Has the applicant or anyone acting on his/her behalf ever filed for or received a Social Security number card before?”
(That question has been on the application form for a long time, since at least the 1950’s.)
Years ago, my dad read an article in Scientific American that said that it was impossible to get consecutive SS numbers… the author went on for pages explaining the algorithim involved and sounded quite reasonable. Too bad my dad had proof that it was possible…the three of us children got our numbers at the same time (there was apparently a big push to have kids get their numbers and they had sign-ups at the shopping center…this would have been sometime after 1957) and our numbers are all identical except for the last digits, which are 5, 6, and 7 for my brother, my sister and I. Someday when I have a lot of time I will search for and find that article.
I think the number (360MM) is a gross underestimate of the number of SSN’s issued to date.
The current population of the US is 300MM. We can assume that all these people have SSN’s. Now, how many people with SSN’s have died since SSN’s were established? I think this brings it to well over 360MM.
Now, add the number of people who came to the US from overseas and were issued SSN’s to work in the US. This has to be hundreds of millions and this number includes students with summer jobs, etc. I know this has been in place since at least 1994 because we hosted two friends from Ireland who worked here for a summer.
In the past (late 80’s-90’s), I was a full time DBA (Oracle, Informix, MS-SQL). I also did SQL training in the Philadelphia and NYC regions. As a personal privacy guy, but also as a DBA best practice, I never, ever recommended using SSN’s as database keys. Why? Because they are not GUARANTEED TO BE UNIQUE. In each class, at least one student had a story about duplicate SSN’s. Not all these were legit (typo, data entry error, etc), but several had found that people had duplicate SSN’s.
Anyone else encounter duplicate SSN’s and have additional details?
-Paul
Here is the official statement of the Social Security Administration on the subject: Are numbers re-issued after death?
As you can see, you have grossly over-estimated the number issued. It should also be noted that the column referenced in the OP was discussing the situation as of 1992, 14 years ago. In that 14 years, only an added 60 million numbers have been issued. Numbers are issued at the rate of roughly 5 million a year.
Claims of duplication are just that, unsubstantiated claims. If there is a claim of duplication, it is either an error of fact (the numbers aren’t duplicate) or the result of fraud (someone is using another’s number, perhaps having simply made one up similar to one they knew about).
When you make a statement about something like this, it helps to have actual facts, provable ones, rather than anecdotes. Which is not to say it absolutely could not happen; government being what it is, anything is possible. But unless and until you can point to person A and person B, and see their cards with duplicated numbers, and are able to authenticate that the cards were actually issued by the SSA, it’s just a story. 
Sorry. Not even a good bluff.
This is incredibly naive for one simple reason: Computer errors have been happening since the days of Ada Lovelace. To say that the SSN assignment system is the world’s first and only massive database without error is, well, erroneous.
Then it should be easy to give us an example, right?
Nobody is saying that. The Social Security System does have massive databases, and no doubt they contain errors in the data in them.
But the SSN is the primary key* field of most of those databases, and the DBMS system will automatically enforce uniqueness of the primary key. Since this is generally done as an index into a storage location, uniqueness is inherent: you can only have 1 record in a storage location.
*If you look up primary key, you find that they happen to use SSN as an example: “Examples of unique keys are Social Security numbers (associated to a specific person)”.
Your posts increasingly display an unwillingness to accept facts or to grapple with opposing ideas through sensible discourse. When you are told that the question has been on the application since at least 1950, unless you have evidence to the contrary, your only choice is to request evidence in support of that assertion, or accept the truth of the statement, if you want anyone to listen to your own assertions. You asserted you had received a second social security number, which required you apply for it. You’ve been told the application has required you to assert that you have not ever previously been issued a SSN, and that has been true for the last 50+ years. Now you have three choices: accept this as fact and realize that, despite the fact you may not have realized it at the time, you made a declaration that was false on your application, request proof that the assertion made by t-bonham@scc.net is true, or provide evidence that it is not true. A simple dismissal with the statement “not even a good bluff” simply shows that rational discourse with you on the subject isn’t possible.
Similarly, your assertion about database error is nothing more than talking out of your <anatomical description of some place other than the oral orafice deleted>. Yes, errors are made, but so what? No one said there weren’t errors. What I asserted was that the likelihood of error regarding the duplication of numbers was extremely low, and as such, in the absence of evidence that it has actually occurred (there’s that evidence thing again!), should be regarded extremely skeptically. Again, you can offer proof that numbers are duplicated, state that you yourself consider it likely that they are duplicated, despite lack of evidence to support that conjecture, or accept that it is less likely than posited earlier in the thread. But to simply say, “computer errors happen” as a statement of why duplicate numbers must exists, without any review of how the numbers are assigned, how the database of numbers is built, how errors can be made and how they would apply to the issuance of duplicate SSN’s, etc., is an example of simply saying that what you believe to be true is true, and the hell with the evidence or logic.
Neither of these strategies will win you many converts around here. :dubious:
Having worked for SSA for 30 years (recently retired) I can speak from personal experience on both of these issues.
Every case of a person having two SSNs that I have seen was the result of they (or someone, usually a parent or guardian, acting on their behalf) applying for a SNN without acknowledging that they were already issued one. Sometimes this was an innocent error, such as a person not realizing that they had been assigned a number as a child (these were mostly older cases from before the system was computerized) or forgetting that they had previously been assigned a number, and sometimes the person had some reason for wanting to have a second number. The worst case of the latter was a person who claimed when he filed for benefits that he had used four different SSNs in the (IIRC) 50s and 60s, using different names for each, in order to circumvent some sort of restriction on the amount of work he could do as a trucker…the details are a little vague in my mind now.
Cases of more than one person using the same number were always the result of clerical errors or fraud. Scrambled earnings cases were a major pain in the ass, since they involved trying to establish who had or had not worked for each employer, often with the employer long out of business. Fraud cases sometimes involved illegal immigrants passing numbers around, or people wishing to hide income; I had a case of a person using his father’s SSN to avoid having income reported while he was receiving disability benefits.
Clerical errors by whom, I guess is the issue at hand. Clerical errors by the SSA? That is, are two different people issued the same number? If so, how often does it happen?
Always great to get some real evidence to work with! 
If there is indeed a plan to generate non-sequential numbers, that seems like a crude attempt at a checksum, or some way to prevent a single digit mistake from matching a valid number. Unlike most serial numbers issued for other purposes today like bank account or credit card numbers, there is no built-in validity digit to SSNs.
Not to change the subject, but since we have a SSN expert here … are there really people who do not have a SSN? The reason why I’m asking is that I once developed a database application for a small life insurance company that had some very, very old members. Naturally I chose the SSN as the primary key, but soon realized that some members did not have a SSN (note that these people were in their 80s or 90s). I got around the problem by generating unique IDs based on permutations of their birthdates, but did they really not have SSNs? If so, how did that come about? I would guess it is possible if someone was long retired and was being privately supported.