There are political discussions in the United States regarding whether Social Security is maintainable over time.

I want to know how the Social Security system works in United States, and what the basis is of the concerns about its maintainability.

  • This question seems too broad, and not necessarily (just) political. Maybe narrow it down to an aspect that can be answered reasonably in one question? Dec 6, 2012 at 11:51
  • 3
    @BrockAdams The politics influence several aspects like education, economy, etc... Saying that this question is not about politics is being a bit blind. Dec 6, 2012 at 12:19
  • @AlbertoBonsanto - It's more that there are plenty of non-political aspects to it. Frankly, I don't think it's too broad, but it does seem to be likely to be fully answered on Wikipedia. I won't VTC but am tempted.
    – user4012
    Dec 6, 2012 at 13:16
  • @BrockAdams - please have a look at both the answer provided and my edit. Hopefully you will judge it as reopenable
    – user4012
    Dec 9, 2012 at 16:20
  • 6
    "How does the Social Security Work in United States?" A: It doesn't! (cue rim shot) I'm here all night, please tip your wait staff!
    – user1530
    Mar 19, 2013 at 5:36

3 Answers 3


First, the formal basic information on how Social Security works:

Informally, it's very simple from the view of an individual:

  • A portion of your income is mandatorily taxes (FICA tax), with the proceeds used to fund the Social Security fund.

  • Once a person reaches eligibility to be paid Social Security (usually specific age), they receive payments from that fund.

  • The system was designed and tuned so that, on average, given how many people work, how much they make, how long they are expected to live after they retire, and how much the population grows, on average, more money is expected to be collected in FICA taxes than expected to be paid out to retirees.

Here's a bit more on how Social Security formally works - this seems to be fairly comprehensively answered on Wikipedia, especially Current Operation section

In the United States, Social Security refers to the Old-Age, Survivors, and Disability Insurance (OASDI) federal program.


Social Security is primarily funded through dedicated payroll taxes called Federal Insurance Contributions Act tax (FICA). ...

The FICA taxes are imposed on all workers and self-employed persons. Employers are required to report wages for covered employment to Social Security ...employers deduct these payroll taxes from workers' wages before they are paid. Generally, the payroll tax is imposed on everyone in employment earning "wages" ... and also taxes net earnings from self-employment.

Trust fund:

Social Security taxes are paid into the Social Security Trust Fund maintained by the U.S. Treasury...

Current year expenses are paid from current Social Security tax revenues. [ This is critical to understanding the problem - DVK]

When revenues exceed expenditures, as they did between 1983 and 2009, the excess is invested in special series, non-marketable U.S. Government bonds, thus the Social Security Trust Fund indirectly finances the federal government's general purpose deficit spending.

The Social Security Administration's authority to make benefit payments as granted by Congress extends only to its current revenues and existing Trust Fund balance, i.e., redemption of its holdings of Treasury securities. Therefore, Social Security's ability to make full payments once annual benefits exceed revenues depends in part on the federal government's ability to make good on the bonds that it has issued to the Social Security trust funds. As with any other federal obligation, the federal government's ability to repay Social Security is based on its power to tax and borrow and the commitment of Congress to meet its obligations.

Now, for the maintainability problem.

The short short version:

  • Social Security is a closed system. E.g., it can not pay out benefits above the money that it owns. Its ONLY revenue source is FICA taxes (as far as accounting, the bonds purchased for surplus collection are also considered revenue, but in reality they are simply prior years' unspent FICA taxes).

  • As noted at the very beginning, the FICA taxes and benefit eligibility was designed with very specific demographic and economic assumptions, to make the system sustainable.

  • Currently 2 of the main assumptions no longer hold true:

    1. the # of years that benefits are paid out per person increased dramatically due to average life span increase in USA

    2. the amount of the workforce compared to amount of people in retirement is starting to drop precipitously due to Baby Boomer generation unwisely having chosen to have significantly fewer children than SS models anticipated).

So, once we have "too many" Baby Boomers retired and drawing SS benefits, for "too many" years, with "too few" - relative to that - people paying FICA taxes - there eventually simply won't be enough money in the system to pay all of the promised benefits, without additional sources of revenue. (where "too" refers to assumptions made when original SS system was calibrated)

Somewhat controversially, Social Security can be envisioned as a giant Ponzi scheme (marketing pyramid). Here's how it works.

  • Let's assume, for ease of modeling, that a human being lives for 3 years:

    • 1 year childhood

    • 1 year working adulthood (Children are born when you turn 1 year old)

    • 1 year retirement.

  • We start with a population of 4 people called generation #1 (aka G#1) - split into married couples, aged 1 (they can start work), another 4 people called Generation #0, aged 2 (retired), and an empty fund.

  • Everyone working earns $100 a year, pays 50% FICA tax. Every retiree gets $50 Social Security a year. The math works out the same as real life, just easier to calculate.

    We use 50% because in real life, a lot more people are working than getting SS due to the difference between retirement longevity vs. work longevity.

  • Here's what happens:

    • Year 0.

      SSTF balance = $0

    • Year 1.

      4 people from G#1 work for a year.

      They pay $100 * 4 * 50% = $200

      They give birth to 2 children per couple (4 total) of G#2.

      4 retired people get their SS checks for $50 each, for $200 total.

      End of year SS Trust Fund balance: $0 + $200 - $200 = $0.

      Now, here's the interesting part to show the actual mechanics. Each bill we collected in taxes into the trust fund has a post-it not attached, with the year it was collected.

      Notice that G#1 people got their SS paid by bills collected from G#0.

    • Year 2.

      4 people from G#2 work for a year.

      They pay $100 * 4 * 50% = $200

      They give birth to 3 children per couple (6 total) of G#3. Yay for Baby Boomers!

      4 retired people from G#1 get their SS checks for $50 each, for $200 total.

      End of year SS Trust Fund balance: $0 + $200 - $200 = $0.

      Notice that G#1 people got their SS paid by bills collected from G#2!

      This is super-important! People in G#1, when they retire, are NOT paid by the money they paid into the system as a FICA! They are paid out of the money that the next generation pays right now, while they are retired!.

      Also note how everything is in perfect balance. Our 50% FICA tax was designed so that we have as much money coming in as paying out.

      Now, let's see what happens when you start modeling population changes...

    • Year 3.

      6 people from G#3 work for a year.

      They pay $100 * 6 * 50% = $300

      But the Baby Boomers like their creature comfort, and 2 of the couples only have 1 child each. G#4 has 4 people.

      4 retired people from G#2 get their SS checks for $50 each, for $200 total.

      End of year SS Trust Fund balance: $0 + $300 - $200 = $100. We are in surplus!

      Notice that G#2 people got their SS paid by bills collected from G#3! And since G#2 had more kids than 2, the SS fund now has extra money.

    • Year 4.

      4 people from G#4 work for a year.

      They pay $100 * 4 * 50% = $200

      They give birth to 1 child per couple (2 total) of G#5 .

      6 retired people from G#3 get their SS checks for $50 each, for $300 total.

      End of year SS Trust Fund balance: $100 + $200 - $300 = $0. We ATE the surplus!

      Notice that G#3 people got their SS paid by bills collected from G#4! And since G#3 had LESS kids than 2, the SS fund's extra money was spent right out.

    • Year 5.

      2 people from G#5 work for a year.

      They pay $100 * 2 * 50% = $100

      They give birth to 2 children per couple (2 total) of G#6.

      4 retired people from G#4 get their SS checks for $50 each, for $200 total.

      End of year SS Trust Fund balance: $0 + $100 - $200 = -$100.

      WE JUST RAN OUT OF MONEY, because the next level of Ponzi pyramid did not have enough people.

Your Social Security just went broke. Welcome to what will happen to 21st Century in USA, when all the Boomers who couldn't be bothered to raise enough children will all over the sudden retire, and live longer than SS-originally-planned in retirement years.

  • 8
    "Social Security fund" is an accounting fiction however. In fact, Social Security payments come to US Treasury, and Social Security payments come out of US Treasury. There's no some huge vault where Social Security funds are stored, to be paid later to the recipients - in fact, those funds are immediately converted to US government bonds, which means they are effectively spent by the government immediately after collecting, with the promise to pay later when the time comes.
    – StasM
    Dec 10, 2012 at 9:54
  • 6
    @StasM - What happens with debt ceiling is the same thing that happens with typical alcoholic... "lemme borrow ANOTHER $5 for the VERY EVER LAST drink EVER"
    – user4012
    Dec 11, 2012 at 0:58
  • 7
    @DA. - the only difference between lottery and illegal gambling is that one is run by the government. If you don't understand how the underlying finance of SS works, you shouldn't argue over it. The fact that government decided that SS is legal does NOT change the fact that it is run using EXACTLY 100% same rules as a ponzi scheme. Either explain how the financial details of SS is different from Ponzi, or quit your baseless argument.
    – user4012
    Mar 19, 2013 at 5:20
  • 4
    @DA. - my answer was pure math. You're the one who came in and started playing word games in comments without a shred of evidence.
    – user4012
    Mar 19, 2013 at 5:37
  • 6
    A ponzi scheme is generally marketed as some sort of investment fund where investors hope to get more money out than they put in. Social security, on the other hand, is designed to be a 0-sum game on average. It was designed such that if population was consistent(neither growing nor shrinking) to sustain itself. It's more comparable to insurance Mar 19, 2013 at 19:24

I won't answer the entire question, as it appears a lot has already been written, but will try and provide a really quick TL/DR response to your second part:

and what the basis is of the concerns about its maintainability

The concern is that Social Security benefits are paid by the current working generation, and benefits are taken by the current retired generation.

This creates two potential problems (both of which are happening now):

  1. If more people are retired than working, more money comes out than goes in and
  2. If people live longer, more people take more money out for a longer amount of time than originally planned.

Alas, the system wasn't designed to handle these fluctuating variables automatically very well...hence the need to debate how we're going to go forward with fixes.


Social Security wasn't set up to handle the fluctuations of the economy and demographics as well is it maybe could have been.

  • 2
    "If more people are retired than working, more money comes out than goes in" This isn't correct, you need about 3 workers for every retiree
    – user1873
    Mar 19, 2013 at 4:55
  • @user1873 huh? How does that link disagree with the statement of mine you are quoting?
    – user1530
    Mar 19, 2013 at 5:02
  • @user1873 if your link was meant to imply that things "aren't quite that simple" I'd agree, so added an even shorter summation that hopefully accommodates the massive amount of variables a bit more broadly.
    – user1530
    Mar 19, 2013 at 5:34
  • 2
    The contrapositive isn't true. If you say P => Q, then !Q => !P (ie P="more retiree than worker", Q="more money out than in" so, "More money in than out or net zero" => "more workers than retiree or same number") So, you see the problem is that equal numbers of retiree/workers results in more money going out.
    – user1873
    Mar 19, 2013 at 5:37
  • @user1873 Yes, you are technically correct. Though, if I were to sling the term contrapositive around (admittedly, it's not a term I use with any regularity), I might state: the contrapositive of my statement isn't necessarily true--specifically when there is is less that 2x the number of worker to retired people at this particular moment (ie, the current set of demographics and economic conditions we're dealing with). Hopefully my addendum is a better catch-all in the TL/DR vein.
    – user1530
    Mar 19, 2013 at 5:45

(These are rather additional thoughts to @DVK's answer which are too long for a comment.)
Also, it is from a German perspective, because I know the numbers here better. There quite some differences between the US and Germany, but I think they don't matter that much for this question.

edit: turns out that the US system is more different from the German system than I initially thought from @DVK's answer, because
in Germany the pay-as-you-go is set up in a way that it ultimately will never go bancrupt, but the contributions may explode or the paid out pensions may loose in value instead (or, cross-financing by general taxes, which is roughly 1/3 of the total volume already, increases).
In contrast, the US system is more closed, and if I understood correctly, guarantees a money value of payments, but OTOH does hold reserves and projects the sustainability of the funding for 75 years.

However, knowing this, I think that the scenarios I calculate below actually are more relevant to the original question than I thought: because the US social security actually does these projections, they should have been pointing out the problem early. A decline in births is something that can be spotted basically immediately. The decline in births could have been spotted latest by ca. 1965, probably earlier because it was linked to a solid reason.
In 1965 the oldest of the baby boomers were 20 years old - that would have left almost 50 years to accumulate more reserves, and the whole work life for most of the baby boomers. Obviously, the decision whether to increase FICA or reduce paid out pensions is a political one.

But in any case, this population pyramid of the USA does not suggest any serious decline in the working generation:

US population pyramid

particularly not compared to other countries:

Germany population pyramid

Due these social security reserve projections say that at the current funding scheme, the pension cass will run out of money within 10 or 20 years - still quite some time to take decisions.
So from a German perspective, I'd say with respect to the demographic change, the US system works quite well.

The calculation seems somewhat arbitrary to me (as does the major part of similar discussions I've heard): it assumes that children are raised at no cost.

In fact the 1 yr working adulthood has to pay for the pensions as well as for the children. A quick search on some numbers for Germany

  • average pension (≈ 800 €/month) * 18,5 yrs average duration of pension payment ≈ 175.000 €.
  • Raising a child to the age of 18 is quoted with ≈120.000€. But lots of people do not start into the working adulthood at age of 18 nowadays. This doesn't include the costs for education (paid by taxes which are of course paid by the working adults. My guess is that kindergarden + school + university is at the very least another 120.000€/person).

Bottomline: order of magnitude for raising a child and providing good education is at least the same as for the pension. With that the calculation changes:

  • first change we observe: Paying as much pension as is leftover for the working doesn't leave any money either for the adults or for the children. We'll have to lower the pension niveau to 1/3 of gross wage (or, stay with 50%, but live 4 years and work 2 of them). However, I'll pay them 300 €/a, and call it gross product.
yr children adults                 retired     
    require require  family gross   require pension private comment
                    savings product         cass    savings
1   4=400€   4=400€     ±0€   1200€  4=400€     ±0€     ±0€
2   6=600€   4=400€   -200€   1200€  4=400€     ±0€     ±0€ uups 
3   4=400€   6=600€     ±0€   1800€  4=400€   +200€     ±0€ debts payed off
4   2=200€   4=400€   +200€   1200€  6=600€     ±0€     ±0€
5   2=200€   2=200€     ±0€    600€  4=400€     ±0€     ±0€ fortunately there were savings
6   2=200€   2=200€     ±0€    600€  2=200€     ±0€     ±0€ 

Note that the paying off of the debts does not include any interest, and that bancrupcy of the pension is only avoided by the retirements savings of the large (6p.) generation who had to pay off the debts their parents made for raising them and had to accumulate savings for their own retirement. This was possible because they did have the economic strength, but had they not saved the surplus while working, the pension system would have been bankrupt.

A more "German" version would not have family savings, but a governmental education cass in addition to the pension cass, and cross-funding would happen between pension and education.

Here's a variant II, where the parents economize to raise their children.

yr children adults                 retired     
    require require  family gross   require pension private comment
                    savings product         cass    savings
1   4=400€   4=400€     ±0€   1200€  4=400€     ±0€     ±0€
2   6=500€   4=300€     ±0€   1200€  4=400€     ±0€     ±0€ families shorten expenses by 200€
3   4=400€   6=600€   +200€   1800€  4=400€   +200€     ±0€ 
4   2=200€   4=400€   +200€   1200€  6=600€     ±0€   +200€ adults inherit 200€
5   2=200€   2=200€     ±0€    600€  4=400€     ±0€   +200€ still something left to inherit
6   2=200€   2=200€     ±0€    600€  2=200€     ±0€   +200€ and that stays like this

The situation would be basically the same if the grandparents help saving the initial 200€. The "wealth" of 200 € that is left at the end was basically created by the savings of the 2nd year. Note that the pension cass is barely left with a ±0.

The bottomline is that with this system, a growth in total population is the problem unless the gross product increases accordingly. Note that the estimated costs for education mean that this problem is even bigger, because each child costs roughly 1 1/2 as much as a retired person (people also live of their savings in retirement, but in any case raising a child is nowhere near negligible compared to pension costs). Yet, often more children are seen as the solution to the retirement problem. That would work only if raising a child was cheap, i.e. cutting down education - not a good idea.
A decrease in population can in principle be buffered by a partially capital/savings based pension system, but that needs economic discipline (of which I do not see much in current politics - edit: yet the US social security reserves seem to me amazingly well preserved). I've often heard the argument that a capital-based retirement system doesn't work out as the retired needs someone to look after him and someone to grow the food when he is retired. Howevder, also that is only partially true: there could be more people to grow the food if they didn't need to build a house to live in because the retired had done that already. But note also, that in our scenario with a population growth first before the decrease in population, the capital based system would have been bancrupt immediately: during population growth, there was no possibilty to save for the retirement.

In reality, I guess one also needs to take into account that saved money "itches", and "wants" to get spent (maybe that's more a European problem - see edit above). Particularly if it isn't one's own savings, e.g. politicians deciding at the +200€ stage of the pension cass to cut the tax rate going into pensions. Or, the baby boomers in the first scenario may find it unfair that they should save after they have already paid the debts. So they spend the surplus, and then the system goes bancrupt.

  • The original question ONLY discussed a closed system (social security) - the amount of money in that system is 100% independent of whether it takes $0 to raise a child, or $100,000. Therefore, your answer - while possibly correct from the point of view of overall economy (it's too involved to render snap judgement either way) does NOT really address the issue of whether Social Security is solvent or not...
    – user4012
    Mar 19, 2013 at 2:21
  • ... People would have paid amount X into social security no matter how many kids they raise and how much they pay for them - and would get PAID by social security 100% independently of whether they saved $500,000 for retirement by NOT spending on children they didn't birth. Therefore, such childless people draw the same from Social Security system, BUT (by proxy of their never-born-kids not paying into it) contribute less on the net.
    – user4012
    Mar 19, 2013 at 2:22
  • 1
    I'm afraid I'll have to downvote this answer, since it doesn't address the actual question (see comments above), but with the caveat that this is a VERY well thought out answer, and at first glance correct one - just to a different question. ("how does birthrate affect net economic growth" would probably a good question for this answer).
    – user4012
    Mar 19, 2013 at 2:24
  • @DVK: wrt closed system: so you really have a closed system? Then we found a first important difference: in Germany the statutory pension insurance "claims" to be paid by the respective payroll deductions (called pension cass contributions, not termed a tax) - but in fact about 1/3 of the money comes from governmental sources (i.e. other taxes), so it is far from a closed system. Mar 20, 2013 at 17:43
  • @DVK: for generation 0 it is accepted that instead of paying money into the system, they provided generation 1. My scenarios also do the reverse: a generation can pay more (savings) instead of raising children. In Germany, this is actually done in some aspects (there are parts of social insurance where you have higher fees as long as you are without children, you will also receive pension for years without contribution because of raising children as if you had paid contributions, most of the education cost is tax financed, you get a payback for raising children, ...). Mar 20, 2013 at 17:50

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .