Debunking the claim of sudden death of future Social Security. 

For printable pdf file, click here.

By Ralph Hiesey   Revised March 2024.

Reference for this essay: From “Social Security Trust Fund Cash Flows and Reserves” from the Social Security Administration website. A very detailed, long, thorough and close to unintelligible description of the SS accounting process. In the first paragraph of their reference it says: ” Although some observers view the trust fund reserves and interest income as accounting fictions, a careful tracing of the cash flows reveals that the reserves and their interest earnings are, for all practical purposes, as real as those of any bank account.”  That is true!!–but they don’t explain very clearly what “true bank account” means–which I hope this site will make more clear.  When you “save” money in a bank, they don’t leave it in a vault; the bank  loans it to someone else to spend, to make interest money for the bank. Later, when you want it back they give it back with interest, from some of the small amount of cash they happen to keep as their “reserves” just in case someone wants it right away. Social security works similarly.

This describes with some precision exactly how Social Security accounting works, and why a lot of worries about Social Security’s future are overblown.  For this kind of funding, it is mistaken to believe there will be a future date upon which we suddenly run out of money.

What most people do not understand is that a problem with future financing of Social Security cannot be solved by spending a lot of money now, before the money is actually needed.  Money that is paid to social security recipients must always be collected very near the same time it is needed. I will soon explain why the economy requires that.

Here’s a fantasy example which I hope illustrates the problem.
In an economy suppose that in 2023 half the people decide to not work (temporarily take a one year retirement) in 2025.  Assume GDP for this economy in 2023 is $20T. In order to afford to do this the future retirees plan to work in 2024 twice as hard so they can save $10T to pay for their short term retirement they’re planning in 2025.  The total GDP in 2024 is therefore $20T + $10T = $30T.  They put $10T in a lock box to use in 2025.   Everyone is very happy, especially the Republicans who know they are being responsible by prudently saving the money they will need for their retirement.

The 2024 surprise just before retirement.
Or maybe not so happy after all: In 2024 $30T of goods were produced to provide that year’s $20T consumption plus $10T saving for the following year. But only $20T was available to pay for goods/services in 2024.  $30T worth of goods/services were produced while  $10T was placed in a lockbox.  So only $20T of goods/services were consumed. Lots of goods piled up in inventory.  Lots of food rotted.  Service providers couldn’t get enough customers, because 33% of the money they expected in pay went into a lockbox instead of their pay. This stuff could have been really helpful in 2025, but it was all produced too soon.  Unhappy people. Deflation was rampant.

Then 2025 had an opposite surprise when they tried to retire.
Half the work force retires. They open the lockbox to get their $10T to distribute it. To their dismay the entire GDP production was now only $10T, because only half as many were working, but demand for goods and services was $20T, of which $10T was from  workers and $10T from saved money of retirees. Now there are people demanding $20T of goods/services, but only $10T of goods/services were produced. Instead of a surfeit of goods the year before, now a terrible shortage. Prices double because of lack of supply.  People again very unhappy. Inflation was rampant.

That didn’t work out too well; but if only one person decided to do this, everything would have been (almost completely) fine.  Suppose in 2023 only ONE family decided to temporarily retire (for a year) in 2025. In 2024 they would save the relatively small money amount, $80K for themselves needed in 2025, and put this in a lockbox, which would have only a very slight effect on the economy—though it still would have the effect that $80K worth of goods/services would not be sold—somewhere that much would be left unsold in inventory in an economy of $20T.  Then they could retire in 2025 and spend their money, which would hardly be noticed—except that there would be a little more demand for that year than was produced. Maybe prices would go up by .00001% as a result. No one would notice this tiny inflation.  This shows that for very few people, it could make sense to save money in this way for retirement.  But if the number were large,  there would be lack of economic demand in the year it was saved, and lack of supply in the year it was needed for retirement.

Realistic historical solution actually applied: Some smart economists realized this problem in the 1930’s when they came up with social security.   They realized that they must not hold any actual cash for an extended time, because if they did necessary demand in the economy would be lacking.  They did not want to remove cash (as taxes, for example) from the economy at a greater rate than they disbursed it for government expenses.  Money must recirculate to keep products and services flowing from producer to consumer.  If over time, like months, cash is removed by taxation at a higher rate than spent, that will slow an economy, tending to deflation and unemployment because lack of transaction money will tend to cause economic demand to be reduced.  This is what I attempted to demonstrate with the example I described above of the economic disaster that would occur if large number of people that held a lot of saved money in a box—then tried to spend it all quickly much later.

Here’s how they avoided this problem.  The government took money in form of a payroll tax which most people thought was going to sit in a box somewhere.  That made people feel OK—people could feel that it was just their money the government was keeping safe for them somewhere in a lockbox for the future. But their money did not go into a box. They did not put it in a box because trapping huge quantities of retirement money for long periods would cause the economy to slow. Instead it was disbursed to retired persons immediately as a social security payment.  This solved the economic demand problem when the tax was paid.  If that cash were put in a lockbox there would be that amount of missing dollar demand in the economy to purchase the goods/services that the taxpayers produced to earn that tax money.  Demand was sustained by sending it to retired people who DID provide the necessary cash demand to consume those goods.  That kept money flow in balance.  This assumes that the total payroll tax amount collected was about equal to what was disbursed to retirees. But suppose it was not.

Treasury bonds enter the picture.  But there was a problem: suppose the total payroll tax paid was more than what was necessary to pay the retired people.  The government did NOT return the money directly the taxpayer! But indirectly they did exactly that. They realized that they must turn it back into the economy to provide extra demand to consume the extra goods/services that were produced by the surplus of payroll tax collected. Their method was to purchase Treasury bonds which were held by the U.S. Treasury, that were created and “paid” for by the money received as extra payroll tax.  Like all bonds that the Treasury sells, this money surplus from the payroll tax immediately went into the general fund that the government spends, like pay for government employees, wars, and other things that government taxes normally pay for. That’s always what happens with the money when Treasury bonds are sold.  So this is what happened with the payroll tax money —except that these bonds stayed on a shelf inside the government somewhere.  So the excess money from the payroll tax took a short trip into the Treasury bonds, then immediately out into the general government disbursement fund to get spent.  This kept the money flowing, and avoided accumulating cash, that if not disbursed would have caused a glut of unsold goods/services in the economy.

Some who found this out complained that the government was stealing their retirement money to spend it on other government services—not holding it in a box. But one benefit was that they were paying interest on the money.  Considering the huge amount of money that was held, if it had been held in cash it would have badly slowed the economy.

At some point, there came a day when the amount coming in from payroll taxes was LESS than the amount of required to pay retirees. Now instead of net buying of Treasuries, there was net selling. Most recently that happened in 2017. Did that mean they paid the retirees less?  NO. Where did they get the extra money to pay retirees? Simple.  The treasury department then cashed in the “Treasuries” they had accumulated that were “owed” to social security recipients, which included some interest. When redeeming the bond, money flowed the other way—that is from the general tax fund to the SS recipients.  That’s how any treasury bond previously sold to the public is redeemed at the end of its term. The money is then used  to make up for the additional amount needed for retirees which gradually reduced the supply of these treasuries. So in effect, at that time retirees began to be financed not only by the payroll tax, but also by some money which came from the general tax fund,  from your income taxes.  It’s just the opposite money flow that happened when payroll taxes were higher than needed to pay recipients.

The people pushing the social security panic button say “we can’t afford to provide SS—it’s too expensive—we will be bankrupt in 2035 (or whenever.)”  What proves this in their minds is that on that fatal day when the fund runs dry– “no more money will be left to pay social security.”  But no, it does not mean that.  Money will still come in from payroll tax— but since the trust fund is now zero, money from the payroll tax will only be able to pay a portion of the amount owed to the recipients—now estimated to be about 75% of what’s owed. But the additional 25% amount that was before coming from the general tax revenue—from worker’s income taxes–which took a quick trip through the bond was no longer available to make up the social security deficiency.   The BAD part after this fatal day is that now they could not legally pay the full amount of benefits expected by SS recipients. They would only be able to pay what was currently being generated by the payroll tax.   But there is a GOOD part exactly EQUAL to the bad.  Now the money from the general tax fund would no longer need to go to social security payments, so the seniors’ loss would be equal to the income tax payers’ gain. Non payroll tax taxpayers could get a windfall tax cut.  But if the Congress so legislated, the situation could be fiscally exactly restored as what it was the day before if they continued the same practice of making up the balance with the general tax fund—which would mean no windfall tax cut—and no tax increase–and no SS benefit cut.  The only difference is that the rapid trip through the bond route would not happen.

For professional social security finance worriers: There is a real concern about social security costs, and illusory concerns.

• The real concern:  Eventually and gradually over time the financial burden on taxpayers might be too great to pay money to retirees—primarily because the population ratio of older retired people to younger working people will be much greater than it is today. This will need to be properly addressed by increasing taxes or reducing benefits if and when that happens, not before.
• The illusory concern:  Eventually we will run out of money for social security.  If we don’t act now to pay more tax, or reduce benefits, we will be doomed in 2035;  it will be too late to do anything about it then. Depending on the worrier: social security will then pay nothing (completely wrong!)  Or more realistic: payouts will have to be reduced to 77% of what they were just before. But, if that happened, the general tax burden for taxpayers will instantly also be reduced by the 23% amount no longer paid to SS recipients—suddenly reducing the general tax expense requirement from that day onward.  So no catastrophe would happen. Social security payments will be reduced, but other taxes could go down by an equal amount with no other effect.  It could be avoided with essentially no fiscal change by continuing to fund the difference with general taxation—just as was happening one day before the “disastrous” event.

Solutions for the real problem:
The real problem could be solved by some law changes—for example raising the maximum income on which the payroll tax applies.  The payroll tax is now a severely regressive flat tax of 12.4 % only on those with income below $137,700 in year 2020.  (Note 1, below)   Making the payroll tax a flat tax on all income could potentially  provide considerably more money, and it would be a fairer flat tax, no longer be a regressive tax on lower earners. The payroll tax percentage could be lowered, and  would be shifted to higher earners, with cash total decreased to those with lower pay.

What would be the fiscal effect of that solution?  Answer: When implemented, no change in total taxes. It would be a big tax increase on those making above $137,700 per year in the form of the payroll tax.  However this would reduce the payroll tax rate for everyone with income  below $137,700.

Another alternative legislative route that could be taken could be to change the law so that general tax revenue would always be allowed to make up the difference if the payroll tax revenues at the time were not sufficient to deliver full promised benefits.  This would not mean any more taxes than were required the day before the treasury fund ran out—since the money to make up the difference had always come from there—even  before the SS fund ran out of money.    It would just be a minor accounting change that would eliminate the treasury intervening, and also shift the tax rate up on income tax, and equivalently down on the payroll tax.

Large overview for what this essay is really about: Social security benefits paid to the older population at one point in time must always come from taxes that are collected from younger workers at the same time.  The reason is that the purpose and effect of Social Security is to divert some of the GDP, meaning actual goods and services, from the younger working population at one point in time to the older group at the same point in time.  That means diverting some quantity of money at that time from the younger workers to the older workers—a money transfer being the method that is used to allocate such goods.

To generalize this result to any pension system, public or private: In any economy, the benefits in terms of products and services given to pensioners—benefits always come from those working during the same time period as pension checks are paid.   There is no problem of saving money, it won’t rot (except by inflation.)  But goods/services are for the most part perishable, and can’t be produced when people are 35, and spent when they are 70.  Advance money is not the tool that can solve this problem.  It can only be solved by taking the view that it is right that a larger quantity of people of working age should support those who are beyond working age.  You support your parents when they were older. Then your children will support you in old age. And so forth in time.

Note 1: The tax is supposedly divided half and half—6.2% employer and 6.2% employee.  So some might think the tax on the employee is only 6.2%.  Another political gimmick.  Realistically the tax falls completely on the employee—as any employer knows, when the employee is hired his labor must justify paying the “employer’s” portion from his labor to justify hiring the employee.  If the tax were completely eliminated the employee would no longer pay his 6.2%, and in addition the employer could raise the employee’s pay by 6.2% for no additional expense. So the employee would have an effective12.4% more to spend.

file: SocialSecurity.   Revised April 30, 2020.