After I published my recent three-part series of posts about the workings of the Federal Reserve (here, here, and here), I received an email from a reader, who asked the following questions:
Whenever I talk about so-called entitlements with my conservative friends, they respond by saying that Social Security is unsustainable. They say that there are too many old people, compared to young workers. When Social Security began, there were fewer old people and a lot more young workers. Now, the pyramid is reversed, they say. The only solution is to cut benefits. A common statistic they quote is that the government spends 4 dollars on every adult over 75 for every dollar they spend on someone under 18. So, they promise to keep benefits as they are for people 50 or up, and cut the benefits for future beneficiaries like me.The emailer's questions raise two separate sets of issues: (1) How does Social Security really work, and is it doomed? and (2) Are we spending too much money on benefits for older people, and too little for younger people (especially children)? (The answer to the emailer's last question is, of course, that we are spending far too little on children, but not necessarily too much on older people.) Long-time readers of Dorf on Law will recognize these as issues on which I have focused quite a bit of my writing, which made the invitation to revisit those policy questions quite welcome.
Is that really the only choice? Are these numbers cited in a misleading way?
I liked those questions so much, in fact, that I decided to write my Verdict column last week on the first subject, that is, the sustainability of Social Security. (I plan to address the second set of issues in a future column and blog post.) Here, I will explain a bit further the details of the argument, and then comment on two related issues that I raised in the column.
My bottom line (which is hardly unique among those who actually study Social Security, rather than scream about it) continues to be that the system is completely sustainable, and that the only thing we have to fear is political fear itself. It might even end up not being necessary to change anything at all to keep the system going, so that all of the current panic over the system will have been entirely unnecessary. If changes are ultimately necessary, they can be made later, and in a progressive way, as I will explain momentarily.
The emailer's first question should not have surprised me, but it did. The decline in the workers-to-retiree ratio has been well documented for decades, of course, but I thought that it had been well-established that the demographics raised no reason for panic. I often forget, however, that reality is often powerless against politically convenient half-truths; so it was helpful to be reminded that this idea keeps coming back from the dead.
It is a half-truth, because the number of workers per retiree is meaningless until we know how productive those workers are. As I pointed out in my column, worker productivity has risen so rapidly that it is possible for each current worker to produce goods and services in amounts that increased the living standards of current workers (and their non-working offspring) as well as that of current retirees. That will continue to be true as the Baby Boom works its way through retirement.
The only question, therefore, is whether the system is currently set up in a way that divides the economy's output such that both workers and retirees can enjoy the benefits of economic growth. It is possible, after all, that retirement benefits will rise too fast even relative to the economy's projected growth. The good news is that we either are already in the sweet spot, or that there is a simple way to get there later that does not harm workers or retirees. There is simply no need for panic (or even immediate change, no matter how calmly undertaken.)
This leads to the first of my observations about the way I framed the issues in the Verdict column. Because I had so recently written about the way the Fed works, and how it creates money, I found myself analogizing the current politically-contrived panic over Social Security to a "bank run." That is, even though there is definitely (by design) never enough money in banks' vaults to even come close to meeting depositors' needs (if all depositors were to decide to withdraw their money at once), that does not make the system unsustainable. It does, however, make the banking system vulnerable, because if people become convinced that their money is "gone" (recall the anecdote from The Beverly Hillbillies that I mentioned in my most recent Fed-related post), then they will try to withdraw it, which will create a self-reinforcing crisis. Although the analogy is admittedly imperfect, my point was that Social Security can be vulnerable to the political equivalent of a run, by which a system that is completely sustainable is dismantled by people who have been able to convince a trusting public that the system is unsustainable.
What surprised me most about that analogy is that it caused me to revisit one of the framing issues that Social Security experts have long struggled to explain. The standard story is that Social Security is not really a set of bank accounts, but that it is in fact just a pay-as-you-go system that FDR decided to describe as if it were a set of individual bank accounts. The historical records do, indeed, show that FDR decided to use account-like language on purpose, worrying that people would be worried if they did not think that their money was sitting somewhere, waiting to be withdrawn upon retirement. (This deception is then supposedly worsened by the language of the trust funds, which are not held in cash, either.)
I now think, however, that we have typically misread the significance of Roosevelt's rhetorical move. Rather than saying, "Well, FDR shaded the truth in the service of garnering public confidence in the new system, telling them that a system with no real money in it was like a set of bank accounts," we should instead say, "Bank accounts don't act the way people think they do, either, because there is no more real money in them than there is in Social Security."
Any lie, in other words, is not in describing Social Security and its trust funds incorrectly, but in describing bank accounts incorrectly. Social Security "accounts" are empty, but they are empty in the same way that bank accounts are empty, because both involve flows of funds that use legal rules to draw in taxes/deposits and determine benefits/withdrawals. Those rules are not simply principal-plus-interest in the Social Security system, but many individual retirement accounts are also quite complicated. Conceding the idea that being a pay-as-you-go system makes Social Security somehow less "real" or safe, therefore, does needless and inappropriate damage to Social Security's support. I have argued that "Money is Magic," and this simply says that there is no need to concede that Social Security is somehow less real, simply because the money is not in a vault (or a lock-box).
The second point at which I surprised myself in my Verdict column was in how I described the distributive consequences of financing Social Security, should we ultimately need to adjust taxes or benefits to stabilize the system's finances. As I have argued many times, the "mid-range" estimates of when the trust funds will reach zero are actually based on rather pessimistic economic assumptions. If they turn out to be accurate, however, then the system will reach a point in twenty years or so when legislation will be needed to raise funds to supplement payroll taxes, or else existing law will automatically reduce benefits (by approximately 25% from projected levels).
Leaving aside my usual argument about the growth of benefits, and how 75% of those benefits is still better than today's benefits for most retirees, what happens if we increase taxes instead? One thought is that we can increase future payroll taxes, to keep the system's status as a self-financed entity. Another possibility is that we could permanently supplement the payroll taxes with another revenue source. What would be the logical basis for taxing any other source to sustain Social Security?
Recent estimates indicate that, if wages for the last generation had continued to constitute the same share of the economy as they had previously, not only would workers be much better off today, but Social Security would be (even in the less optimistic forecasts) fully funded. That is, because of wage stagnation, the base of the Social Security tax is smaller than it would have been, which reduces revenues.
This, I argued briefly in my column, provides the basis for a distributional argument to levy a progressive tax (if needed) on upper incomes, to supplement Social Security taxes. We have experienced an entire generation of what has been called "The Great Widening," wherein workers' increased productivity has been impressive, but it has almost all gone to the people at the top of the economic heap.
Even if we cannot get ourselves to address the underlying problem, therefore, we can at least address one of the collateral effects. Social Security might not need additional funding, but if it does, we ought to consider finding that funding by taxing the people whose gains have undermined Social Security (and so much else). As always, the real conflict is not between generations. It is a matter of policy choices over the degree of progressivity in our fiscal system.