After a person dies, how long is their SSN kept inactive before it can be used again?

This is a spinoff of the safety-deposit box question. I know that Social Security numbers are rendered invalid for that person upon their deaths, but I’m curious - how long is the number kept dormant, if you will, before it’s used for a new person, presumably a baby?

I know I could get the answer from Google, but there has to be someone here who knows it off the top of their heads.

They are not reused at this point.

Thanks! I’ve heard that they were, although the sources may not have been the most reliable.

But it will be a problem some day. Unless the Feds turn off the system first.

If they use all possible numbers there are a billion possibilities. I’d think they are a bit under half used by now, and they could always add a tenth digit or make it alphanumeric.

My own guess is that when they have to start reusing numbers, they’ll do their best to select numbers that have been out of service for, say, 50 years or more.

I do know that they are not randomly assigned, and people In The Know can figure out pretty accurately where and when the number was issued.

They could, but it would present a Y2K-type issue. There are, undoubtedly, many thousands, if not millions, of computer programs which use or require SSNs, and are very likely all currently coded to require exactly nine numeric characters.

Yup. At least historically, the leading digits were a function of the state (and probably the office) where the original application was filed. My sister and I got our SSNs in 1975 (back before they started the policy of issuing them to newborns), at the same SSA office, at the same time, and ours are one digit apart from each other.

The first 3 digits used to be regional. But not for ~15 years now; it’s random.

See

Under Structure.

Not sure about SSN’s, but Canada has SIN’s (Social Insurance Numbers). I recall an article back in the 1980’s about someone who had a real problem that someone else was consistently using his SIN. Revenue Canada would get on his case for things like unreported income due to the other person’s antics. It was in hte news because he wanted a new SIN and Revenue Canada said they simply could not do that - issue a second, number, and transfer (or attach) previous data to that number. It was brought up in Computer Science class as a classic “the computers won’t let us do that” design problem.

I assume the same applies to SSN’s? You get one for life? What happens to those people who assume someone else’s SSN, let’s say because they are in the country illegally? How closely does the IRS track bad SSN’s?

I assume like the SIN there’s some sort of check digit in the mix, so not every 9-digit number is a valid number? Or can you randomly make up an SSN and it passes the test?

Maybe the OP is thinking of phone numbers? I think those do get recycled, maybe because for some area codes they do run out every few years.

The article I cited talks about check digits: there aren’t any.

And about the issuance of substitute numbers. Which SSA can do, but which might be a real PITA to change everywhere else.

Phone numbers in the North American Numbering Plan (NANP) are supposed to be held inactive for 6 months before reissue.

The problem with this is that America’s population is growing, so new SSNs are needed at a faster rate than old numbers used by past populations become available for recycling.

Yeah. Once they run out of initial issue numbers, pretty quickly the pool of reusable ones will be exhausted too. Which suggests that’s not the smart way to solve the problem of eventual exhaustion of unused numbers.

The USA survived switching the ID value for everyone on Medicare just a few years ago. For those who don’t have one or work with them they’re now 11-character alphanumerics. They used to be your own SSN plus a letter.

Good bet we’ll survive the process of starting to issue 9-character Social Security ID values with letters in at least some of the 9 slots. That’s probably the least disruptive change.

Allowing any digit or letter in any of the 9 positions produces ~100 trillion possible ID values. Even saving one position out as a check value gives ~2 trillion values.

But if there’s a billion SSN numbers, even if the USA reaches 500 million people, unlikely by 2100, that still leaves half the numbers out of service.

Assuma a steady-state 500M people with a 80-year life expectancy. That means 6.25M born and die each year. Before that means a number would be 80 years old before it has to be reused.

Kenobi points to the SS statistics of 453M SSN’s and 5.5M assigned a year. Even if we assume 6M/yr because of growing population, to use the remaining 547M would be 166 years. So basically it’s like the Y2K problem was in the 1970’s… someone else will deal with it later.

Does someone really have SSN 000-000-001 or do we eliminate anything starting with 0?

ETA - consider demographic collapse - the USA population only grows because of immigration. So the excessive increase of the last century will no longer be happening.

My cite addresses the details. One of which is that 000 is not a valid first 3, 00 is not valid for positions 4-5, and 0000 is not valid in 6-9.

Also TINs and ITINs look like SSNs but are not, and have their own carve-outs to prevent collisions between the 3 systems.

Not only is the American population growing, but you need an SSN to work on the books in the U.S., although I’m sure there are some exceptions.

You can look up your yearly income on the https://www.ssa.gov/ website, so if there’s any funny stuff going on, you can catch it. I’ve definitely heard of people who were dodging creditors stealing, or at least making up, other SSNs so their trail wouldn’t reflect their true income.

The new Medicare numbering system has been done to reduce identity theft, and also because in recent decades, using the Social Security number as identification is more and more discouraged. My state has not put it on driver’s licenses for a long time, I do know that.

And here I always thought that Shadowrun made up that term for the sake of the anagram.