The Dangers of Relying on Legal Fictions and Analogies (Social Security edition)

by Neil H. Buchanan

My latest Verdict column, "Social Security Will Be There When Today’s Young People Retire," was inspired by my continuing discussions about Social Security with people in their twenties.  With so much disinformation spewing from Republicans about Social Security, especially from the establishment candidates (with the front-runner's views being wrong in a different way), it is constantly necessary to debunk all of the hysterical warnings about "bankruptcy" and all that.  [Update: That column has now been republished by Newsweek under the title "No, Social Security Is Not, Repeat Not, Going Bust."]

The biggest takeaway from that column is that the worst-WORST-case forecasts do not show Social Security "running out of money," as even some journalists now describe it.  Even in that worst case, millennials would still receive modest but important benefits.  I show, for example, that a single person whose gross salary is $40,000 per year would receive about $13,400 per year in Social Security retirement benefits under current law.  This is by no means a fortune, but it would all be effectively tax free, so that a person with low-middle income during her working life would receive retirement benefits that are quite modest but would more than keep her from becoming homeless or destitute.  If the economy performs better than that absolutely worst-case forecast, benefits could be as high as $18,900 per year for a lifetime $40,000 earner, under current law.

One of the long-running confusions about Social Security is the question of whether a person's payroll taxes are being deposited in some kind of personal savings account.  Historical records show that President Roosevelt made a very deliberate decision to describe Social Security in exactly that way, so that each person would feel a sense of having earned her retirement benefits.

As a matter of political strategy, that was a brilliant move.  It has, however, created a political opening for ideologues who try to scare people by telling them that "your money" has already been spent.  Even though every bank would be a Ponzi scheme under this logic, because banks do not hold depositors' money in vaults (but instead lend it out at a profit), some people become scared and believe that there is something nefarious going on when an opportunistic politician tells them that there are no individual accounts holding their money.

Overall, I think that Roosevelt's move was a smart one.  To this day, people's reactions to proposals to decrease Social Security benefits are clearly based on a sense of personal ownership, so that the long-term political stability of the system has unquestionably been enhanced by giving people the (quite accurate) sense that their payroll taxes create an obligation for the system to pay them benefits in the future.

There is, however, another fiction about the system that is also being exploited for political gain.  The "running out of money" claim, after all, is a claim not about Social Security itself but about the system's trust fund, which might or might not reach a zero balance sometime within the next few decades.  (The absolute worst-case scenario has the trust fund reaching zero in 2028, would result in benefit payments for a $40,000-per-year worker of $13,400 every year thereafter, as I described above.)  Even if the trust fund goes to zero, therefore, the system will continue to operate and provide important benefits.

But the additional politically exploitable question is: What is "in" the trust fund, even before it reaches a zero balance?  George W. Bush's answer was that the trust fund was just a bunch of worthless pieces of paper.  Defenders of the system responded that the trust fund is "invested" in the safest asset known to man, Treasury securities, on which the government would default only if Republicans in Congress were ever to carry through on their threats not to increase the debt ceiling.  (Back in 2005, when Bush was pushing his partial privatization plan, the debt ceiling had not yet become a political weapon.  This meant that the safety of Treasuries was still seen as absolute.)

For years, I did not question the claim that Social Security's trust fund was, as an accounting matter, invested in Treasuries.  In part, I did not bother to inquire further because it frankly does not matter whether that is true, as I will explain momentarily.  But, at least as a matter of paper-shuffling, the picture that the system's defenders painted had Social Security buying a total of $2.7 trillion of Treasury securities during the Baby Boomers' working years, and then selling those securities on the open market during the Boomers' retirement, until the money is all gone.

It turns out that this is not true, even as a formal matter.  The securities into which the trust funds are supposedly deposited are "special issues," which means that the trust fund is not competing with the general public to buy Treasuries on the open market, nor will Social Security be dumping Treasuries on the market in the coming years as it draws down the trust fund.  The special issues, it turns out, are simply an accounting mechanism that Treasury uses to keep track of how much extra money Social Security has collected over time, to which it adds interest payments based on a legislated interest rate.

As I noted above, all of this is very much beside the point, as far as the actual underlying reality is concerned.  However, it is easy for people to become focused on a particular security or a particular interest rate and think that large purchases and sales by the Treasury would be important in some way.  In particular, during the build-up of the trust fund, the higher demand for Treasuries would seem to raise prices (which reduces rates), while Treasuries' prices would fall (and rates rise) when the trust fund is redeeming all of its holdings.

Is that bad or good?  It is neither, because Social Security is not in fact buying or selling anything.  The trust fund is a legal fiction, not a reality.  The legal fiction does, of course, create an essential legal obligation that we must continue to honor, but Social Security is not really in the business of buying and selling Treasury securities.

What is really happening?  Each year's Social Security surplus of payroll taxes over benefit payments has reduced total federal borrowing.  That means that the Treasury did not need to borrow as much money from the public as it otherwise would have, which means that there were fewer Treasuries sold to the public.  This held down interest rates in general, because the government's reduced borrowing needs left more money available for private borrowers, who did not have as much competition  from the government for loanable funds.  Now that Social Security is starting to spend more than it is taking in, overall federal borrowing will be higher for the next few decades than it otherwise would have been -- just as Reagan and Greenspan planned, back in 1983.

But wait, you say.  Is that not exactly what I said was not happening, only two paragraphs above?  As a matter of the system's effect on interest rates, yes it is.  The reasoning, however, is meaningfully different.  Based on the idea of "crowding out," which is about as orthodox an economic concept as one can find, the point of reducing federal borrowing is to leave more money on the table for private borrowers to finance productive investments.  Reducing the interest rate is the whole point, so that more such investments will be worth financing.  That is how the Baby Boomers pre-paid for our retirements, by building up a bigger pool of real investments that will allow the economy to support us during our golden years.  As we draw down the notional balance in the trust fund, we are in fact using up the excess investments that we put in place for exactly that purpose.

In other words, even though the trust fund is not really invested in anything, reducing Treasury's borrowing needs meant that the supply of Treasuries was reduced, which had the same effect as if the trust fund had actually demanded more Treasuries on the open market.  And during the draw-down phase, the effects will be again the same but in the opposite direction.

Would this have been different if we had never set up Social Security in the first place?  Not at all.  Imagine that we had in 1935 instead set up a system of private retirement accounts, with people making deposits into financial institutions during their working lives and then withdrawing money during their retirements.  The results would have been exactly the same as they are with a Social Security setup.  During Baby Boomers' working lives, funds available to banks would have risen, so that interest rates on loans to private investors would have gone down.  When Boomers started to draw down their savings, rates would rise, just as they would with a Social Security system.

It is possible, of course, that Baby Boomers would have built up less money in the aggregate in those private accounts than Social Security's trust fund has accumulated.  That would mean that living standards during retirement would be lower than with Social Security, which is one of the reasons that people defend Social Security.  But the devotees of private accounts insist that it is possible to set up such accounts so that people are no worse off than they would be with Social Security.  And because nothing in Social Security prevents people from saving more -- in fact, we actively encourage people to save via IRA's, 401(k) accounts, and so on -- an imagined world in which there are only private accounts would not have accumulated a greater net amount than the $2.7 trillion in the trust fund today.  In fact, if it had, then the decline in interest rates would have been even more pronounced during the Boomers' working lives, and the future increase in rates commensurately greater.

In other words, this yet another situation in which any supposed downside of the Social Security system -- in this case, the decline in interest rates followed by the rise in interest rates -- would have been fully replicated in a system of private accounts.  That is because the real issue is not how we are financing the retirements of Boomers, but simply that there was a Baby Boom at all.  When an outsized cohort moves through its life cycle, any system that tries to "save" for that cohort's retirement is going to have the same impact.

This does not, however, mean that there is no downside to switching now to a system of private accounts.  As I have noted before (and will surely write about again, many times), the transition to a system of private accounts would put a special burden on millennials, because they would have to honor the reliance interests of Social Security recipients while also building up their own savings accounts.  Moreover, all of the administrative costs of a system of private accounts would be a net drain on the system, compared to Social Security's incredibly low overhead.  And this says nothing about the difficult of dealing with people who would make bad investments and end up with nothing for their retirements.

Those issues, however, are separate from the fundamental point here.  The trust fund itself has no impact at all on the financial markets, because it is not "invested" in anything nor will it need to be "spent down."  Saving for the future, no matter how we do it, requires either having the government borrow less or private individuals save more.  The net result is the same, on interest rates and everything else.

All of this is far beyond the political debate, of course.  Still, it is fascinating to think through the many misconceptions that can arise when we rely on analogies -- payroll taxes are like bank deposits, the trust fund is investing in Treasuries -- rather than saying what is really happening.  The good news is that what is really happening with Social Security as it actually exists is no worse than, and is in many ways much better than, the alternative.