Social Security Cannot Run Out of Money

No, really.

Michael McDonald tweets the following in response to a Greg Sargent column that is mostly about demographic change and immigration.

This is a common sentiment, especially among think-tankies, “serious” pundits, and journalists. (For the record, I do not believe Michael McDonald is any of these.) It also happens to be:

• Wrong in its important elements.

• A nice, simple topic for me to finally get this Substack started.

The reality is that Social Security is not a “Ponzi scheme” and cannot “run out of money” in any meaningful way. While demographic change might force us to cut benefits at some point in the future, it cannot result in an abrupt collapse, and even a reduction in benefits is not guaranteed. Plus, if a reduction in benefits is ultimately required, there is no particular reason to do it now instead of just taking a wait-and-see approach.

Before I get started, a disclaimer: some of you may object to what I have written here on the basis of some legal details about Social Security’s structure. Note that I am economy-knower, not a law-knower, and economists are generally concerned with real, physical constraints, not legal ones. Which is my nice way of saying: I don’t care what the law is, I care what it ought to be, and these are very much not the same thing. If we let Social Security die because of a legal technicality when there is no particular economic reason to plunge millions and millions of people into poverty, that would be Very Bad. So please shut up.

Impending Doom?

The claim that Social Security is “doomed” comes in two forms:

1. The Social Security Trust Fund is projected to run out of money in (fill in a date of your choosing), at which point Social Security will no longer be able to make payments.

2. Our aging population means that fewer and fewer workers will be providing for more and more old people. This is unsustainable, so the program will eventually collapse.

Both claims are wrong, though in different ways. The former is completely and hopelessly wrong, outside of maybe some legal sense (see above). The latter has an element of truth to it regarding the ability to maintain a given benefits level, but there is no reason to expect a sudden collapse, nor is there any particular urgency in addressing the issue since it may not even pan out.

I will take each argument in turn.

The Diminishing Social Security Trust Fund

At its core, Social Security is a fairly simple program. Young workers pay a flat percentage of their wages in the form of taxes while they are working.¹ Old people collect a fixed amount of monthly income that is (very loosely) tied to how much money they paid in Social Security taxes over the course of their lives.

Sometimes, in a given year, Social Security tax revenues will be higher than Social Security payments. When that happens, the government takes the extra money and “saves” it. Specifically, the government buys a bunch of government bonds with the left over tax money and sticks those bonds in a warehouse. This warehouse full of government bonds has a fancy name, which is the “Social Security Trust Fund”.² When Social Security tax revenues in a given year are not high enough to cover current payments, the government then draws on the Trust Fund to pay out benefits. What this means, in reality, is that it takes those government bonds, converts them to cash, and uses the money to pay for benefits.

This all sounds well and good, but the whole notion of the Trust Fund is built on a core absurdity — the government is borrowing money from itself. Some of you may have noticed this when reading the above. When Social Security tax revenues are higher than benefit payments, the government uses the extra tax money to buy government bonds. When Social Security tax revenues are lower than benefit payments, the government cashes in those government bonds to pay the difference.³

I am now going to do something that is usually an extremely bad idea: I am going to make an analogy between the government budget and a household budget. So before you get too invested in the following analogy, please note that such analogies are to be handled with extreme care.

Suppose that I decide to budget my household expenses. Specifically, I decide to put $300 each month toward food food, while all of the rest of my income is budgeted for other stuff.

Inspired as I am by the Social Security trust fund, I decide to do something pretty ridiculous: I start the JBS Nutritional Trust Fund, a drawer full of IOUs that I write to myself, or cash in, based on my spending. If I ever use less than $300 in a given month to buy food, still spend the leftover on whatever I want, but before I do so I write an IOU for the remainder and stick it in the drawer. When I spend more than $300, I still buy the food I want, but before I do so I take out some IOUs, exchange them for money (with myself), and use that money to buy food.

Over the years, the number of IOUs in the drawer goes up and down, but always remains positive. One month, however, I find that I have blown through my entire $300 food budget by the 15th. Worse, when I check the JBS Nutritional Trust Fund, I find that there are no more IOUs left in the drawer.

Did the trust fund go bankrupt? I suppose so. Does this mean I have to starve for two weeks? Of course not. Why should I starve when I still have money to pay for food? In fact, if I were to just spend some of my other money on food, it would not represent a change from the way I had always done things, since the JBS Nutritional Trust Fund never really imposed any constraints on my spending. I was always just covering present expenses with present income, plus any borrowing. If my total obligations, food or otherwise, were greater than my income, then maybe I made purchases on credit — actual credit, not money I “borrowed” from myself. If my total obligations were less than my income, then I saved — actual money, not IOUs. The JBS Nutritional Trust Fund had no impact on any of that.⁴

The government operates exactly the same way: Social Security benefits are paid out of current income — full stop. If the tax revenues that we have decided to call “Social Security taxes” are not high enough to cover this year’s benefits, the government pays those benefits out of the treasury, which may or may not include new borrowing. If Social Security taxes are higher than the level of benefits required, the government uses that extra social security tax money elsewhere (though it writes itself an IOU first). Perhaps Social Security has a “net worth” in some legal sense, but it does not have a net worth in any real sense. As a matter of real resource constraints, Social Security cannot go bankrupt any more than my food budget can.

So no, Social Security is not a “ponzi scheme” on the verge of “collapse”. If the trust fund falls to zero and benefits need to be paid, the government can just pay them out of present income. This is, in fact, what it has been doing this entire time.

The Coming Demographic Disaster

The United States population is aging, like the populations of most developed countries. Ultimately, Social Security benefits to seniors today are paid for by workers today (if this is not clear, see the above section). Will our aging population cause Social Security to collapse at some point in the future?

No.

Demographic change may force us to cut benefits at some point in the future, but it will not cause a sudden collapse. In fact, even the mere fact that population is aging does not guarantee that cuts will be needed, and so there is really no harm in a wait-and-see approach to the problem.

Let us begin with a couple of “facts” about the economy that, hopefully, we can all agree on.⁵

• Total aggregate production (think GDP) increases with the number of prime age workers. (More workers means we can make more stuff.)

• Total aggregate production increases over time, simply due to improvements in technology, and maybe an increase in the amount of capital each worker has. (Over time, we will just generally get better at making stuff, and maybe have more capital.)

The argument goes like this: Consider the benefits that seniors collect every year, and note that they are tied to inflation. If the number of seniors increases while everything else remains the same, then clearly seniors will take a larger and larger share of aggregate production. If this continues indefinitely, then eventually their share of the economy will become so large that it becomes politically unjustifiable, at which point the young might force some cuts.

That’s it. That’s the problem with demographic change. Note the lack of a sudden, cataclysmic collapse.

However, even this phenomenon is not guaranteed merely by an aging population, and this point relies on the second bullet above. In the coming years, we might find ourselves with more capital per worker than we have previously had, which could offset any declines in productivity due to an aging population. Similarly (and more certainly) we can expect technology to cause the economy to grow, all else equal, and this can likewise potentially offset any decreases in senior standards of living.

Given that we do not know which of the two will occur, and that even if we end up in the bad world where we are forced to cut benefits there is no reason to expect a sudden collapse, why do we need to do something about Social Security benefits now? Should we not just wait and see which world we end up in, and how significant of a cut we actually need, before we start plunging seniors into poverty?

1. There is, famously, a cap on Social Security taxes, making it one of our very few regressive taxes — the rich pay a lower portion of their income in Social Security taxes than the poor. But for most people, it’s basically a flat tax.

2. There are actually two trust funds. I am referring to the Old-Age and Survivors Insurance (OASI) trust fund.

3. Strictly speaking it is not correct to say that the government cashes in the bonds with itself, because government bonds have fixed terms. In reality, the government sells those bonds on the open market. This is not a particularly important detail, though, since those bond sales just offset other bond sales that would have occurred otherwise if the government needs to borrow. It does not matter whether the bonds were printed by the treasury or came out of the Trust Fund — the government is just borrowing from the public to meet present obligations.

4. Would it make a difference if the government saved actual currency instead of bonds? This is where the household analogy breaks down. (They’re dangerous!) The answer is no, but that is because the money supply is decided by the Federal Reserve and set to achieve certain macroeconomic goals (full employment, low inflation), and the government taking money out of circulation would be offset by the Federal Reserve to the degree that it affects anything.

5. This would all be much easier to explain with a bit of high school math, but unfortunately Substack does not make that easy.