This post about the Social Security Trust Fund is a follow-up to this post about the bipartisan push to rollback Social Security. In that piece, I said:
For the owners of the country (and their paid national managers), the real emergency associated with Social Security isn’t the day the last dollar will leave the Trust Fund. It’s the day the first dollar will leave. That’s a whole different problem, and a whole different timeline, for them. How did I come to that conclusion? Read on.
That’s pretty straightforward. A little radical, but not complicated. Read the rest to see the logic that leads me to that conclusion. I also promised:
Next up in this series
This is theory, explanation, and needs to be fleshed out. I’m going to see if I can get some numbers. The size of the Trust Fund is known, but this is also about dates. The Bigs (who love you and want you to be happy) tell you that the care-about date is 2035 (or whenever), the day the Trust Fund runs dry. That deadline induces yawns, one of the problems that its eager promoters have.
But if I’m right, the real deadline is when the first dollar leaves. When is that? Right now, I don’t know. And when that occurs, what’s the rate of decay? What does that graph look like? Again, I don’t know, but it matters. That’s the rate at which the general fund has to make up for the loss, somehow.
I’d love to find out both data pieces, wouldn’t you? Especially the date when the Trust Fund starts cashing those Treasuries. I’d bet money that this is the cause that’s making the Bigs panic and sweat. Stay tuned.
This piece attempts to answer those questions. When does (or did) the first dollar leave the Trust Fund? When does the last dollar leave (based on current projections)?
I did some digging (with a lot of help from friends) in the latest Social Security Report (HTML version here; PDF version here). In it I found one graph and one table that appear to contain the data I was seeking about the Trust Fund. At least in the narrow world of these questions, we have answers.
(Before we go further, I want to acknowledge a huge debt of thanks to Bruce Webb, an expert in these matters, who walked me through this material. Bruce blogs at Angry Bear and he knows his stuff. Thanks!)
How does Social Security refer to the Trust Fund(s)?
Before we start though, a note. Actually there are two Social Security trust funds, but one is much smaller than the other. Since I’m going to quote from some SS literature, I want to give you the key terms:
▪ OASI — Old Age and Survivor Insurance fund (the main trust fund)
▪ DI — Disability Insurance fund (the far smaller one)
▪ OASDI — Both funds taken together
With that in mind, welcome to the world of Social Security documentation. In all cases, I’ll be quoting from the HTML version (cutting and pasting), but referencing page number from the PDF. Needless to say, the two versions are otherwise identical in content.
(The term “pdf page number” means the page number your PDF reader window shows. “Numbered page” means the page number printed on the document page itself. As you know, many documents have unnumbered introductory pages.)
The rise and fall of the Trust Fund under one set of assumptions
First let’s look at the figure I mentioned (Figure II.D2 in the document, shown below); then we’ll consider the table. Scroll down for the figure if you like, or read the following and then scroll down. The introductory text for the figure begins (my emphasis):
Figure II.D2 illustrates the year-by-year relationship among OASDI income (excluding interest), cost (including scheduled benefits), and expenditures (including payable benefits) for the full 75-year period.
Stop for a moment and notice that list:
Income excluding interest — earned interest is not accounted for below under income.
Cost including scheduled benefits — what SS owes to its beneficiaries.
Expenditures including payable benefits — what SS is able to pay.
Why the terms “cost” and “expenditures”? Welcome to the world of SS documentation. The figure introduction continues (my paragraphing):
The figure shows all values as percentages of taxable payroll. Under the intermediate assumptions [one of the assumption sets the document uses], demographic factors would by themselves cause the projected cost rate to rise rapidly for the next two decades before leveling off in about 2035.
However, the recent recession led to a reduction in the tax base and a surge in beneficiaries, which in turn sharply increased the cost rate. This recession effect obscures the underlying rising trend in the cost rate for the next 5 years. The projected income rate is stable at about 13 percent throughout the 75-year period.
You can read the above as wiggle room, appropriately taken. Now the figure itself (pdf page 19, numbered page 11):
Four things to notice:
■ The figure has three lines — a thin black solid line labeled Income; a black dotted line labeled Cost; and a thick black line labeled Expenditures, actual money going out. Starting at the year-2000 mark, the Cost line and the Expenditures line lie on top of each other (the thick black one dominates), while the Income line is visible separately. This reflects the fact that everything that is owed (Cost) is paid (Expended).
In 2033 however, the Cost and Expenditures line separate, and in 2034, the Expenditures line (the amount actually paid out) joins the Income line.
In other words, through 2033, Income is less that what’s owed (Cost), so the difference is made up by the Trust Fund. As soon as the Trust Fund is depleted (2033 by this projection), what’s owed (Cost) can only be met by Income, so Expenditure now equals Income. That’s the dropping-off-a-cliff effect you see in 2033.
■ That cliff is not terribly steep. As the chart points out, in 2033, the program can pay 75% of scheduled benefits from income alone. In 2086, the program can pay 73% of scheduled income. That’s not a huge shortfall to make up. Remove the salary cap and you’re most of the way there.
I’d personally remove the salary cap on all earnings, not just “income,” and make sure the hedge fund billionaires paid in as well. 6% of maybe $50 billion earned per year for just this industry is a lot of catfood-prevention. But that’s me; I’d like to fund the program.
■ Now look at the first part of the chart, near the left edge. Cost and Expenditure (the solid black line that lies on top of the invisible dotted line) are below the Income line. In this period, the Trust Fund grows and current Income pays the bills. In 2009 Cost crosses above Income. You might think that’s the point at which the Trust Fund has to cash in its chips.
Except that …
■ “Income” was defined as income minus earned interest. That’s normally a good idea when making projections like this; you never know what the interest rate environment will be, especially looking 75 years out. Including earned interest would make the 75-year projection almost worthless.
Which simply means that in 2009 the Income line would be higher if the interest it actually earned were added to it. Since this is mainly about projections, you can’t have one definition for Income in one part of the chart (the past) and one for the future (the forecast). They picked the definition of Income that makes the forecast make sense.
Bottom line here: 2009 is not the place where the Trust Fund stops growing.
So what did we learn? 2009 is not the high-water mark for the Trust Fund — we’ll need to look at Trust Fund Asset tables to determine that. But under the assumptions above (the “Intermediate” scenario as described below), the Trust Fund won’t be depleted until 2033 — two decades, folks.
And the Intermediate scenario is just one of three. There’s another scenario that shows an even better date.
To get the high-water mark for the Trust Fund, add back earned interest
To refine the dates, I’m reminded to look at Table VI.F8 (pdf page 214, numbered page 206), which adds back the interest income and shows year-by-year projected total assets. I’ll let you click through to look at it, but note the last column, “Assets at end of year”:
▪ Under the “Intermediate” scenario, the high-water mark for Trust Fund assets is 2020 ($3.06 trillion).
▪ Under the “Low-Cost” scenario, it’s 2030 ($4.6 trillion).
▪ Under the “High-Cost” scenario it’s 2013 (remember, this is an April 2012 report).
Notice that in the table under the Intermediate scenario, the Cost is higher than the non-interest income already, which is exactly what the graph showed. Because of interest, however, the Trust Fund keeps growing.
You can evaluate those scenarios here (pdf page 16; numbered page 8). Each is a blend of assumptions about fertility rate, death rate, immigration rate, productivity, wages, unemployment and so on. Clicking through will show you what those assumptions are. It’s the three-bears approach to what the program could owe — worst-case cost, best-case cost, and somewhere in between.
Clearly, which assumptions are correct matters a lot. The “Intermediate” scenario is generally cited, since it straddles the middle. The “High-Cost” scenario is tempting though, since it assumes a higher unemployment rate, lower productivity, and lower interest rates.
I am not optimistic about either jobs in general or good jobs going forward, since the current U.S. industrial policy is to ship our industries out of the country and hand the (untaxed) savings to billionaires, using corporations as a pass-through. (I’m not joking.) And I don’t expect that to change anytime soon.
On the other hand, I don’t expect the population to continue to grow for the next 20 years either, so that lowers cost, especially if it’s the elderly — the SS recipients — who are most vulnerable. In my view, there’s a surprise on the near horizon, and the frailer and poorer will be hit first and hardest. (More here.)
Taking the Intermediate scenario (or something a little more pessimistic), the Trust Fund will start cashing in (making net withdrawals) between now and 2020. I’m pessimistic about the economy, so my answer to the first question, “When do we start draining the Trust Fund?” is … soon.
And the answer to the second question — “When does the Trust Fund run dry?” — is anyone’s guess, but I’d be shocked if it was earlier than 2030 … two decades from now.
So back to the point in my earlier piece, about what’s driving the Bigs to reduce Social Security benefits — you tell me what’s motivating them. Is it the prospect of spending money in 2033, after most of them are dead? Or the prospect of replacing off-budget Treasuries with on-budget Treasuries … “soon,” while most of them are still in their own wage-earning (billionaire-financed) years?
As I look at our species, “soon” trumps “after I’m dead” for most of us, especially for the greedy. But do feel free to decide for yourselves. Perhaps our masters are beacons of unselfishness after all, and we just haven’t noticed it.
To follow or send links: @Gaius_Publius