My uncle Rod started a conversation with me on Facebook about the proposal to change benefits to a “chained CPI” basis. Along the way, our conversation ran into what I believe is a core, intractable issue with fixing our entitlement system – it is the Zelig of political rhetoric. Rod asks the following:
You observed I was “making a welfare argument rather than a savings plan argument,” and I ask your help in providing me with an explanation I can more easily comprehend (based on my generational and cultural bias). Most folks I associate with who are my age – no matter their personal wealth or lack of same – regard Social Security as earned and owed them. I know intellectually (as many elderly folks understand), it’s much more complicated; that money (especially money channeled through the government) is fungible and now used at a loss to support a system with significant consequences for younger generations. However, I also know the reality that politics (which involves the huge misapprehensions of the voting public and the misdirection of political marketing) has such a significant impact on the distribution of wealth (wherever it occurs) prevents the resolution of this issue the way you most eloquently expressed it: “which is why it should be a welfare program or true insurance program rather than a savings program. “ My father (your grandfather) once told me “most of politics is the sorting out of other people”s money.” I wonder if we are really “doomed to wrap up the subsistence-level folks with people who have options,” in such circumstances or if that forecast may possibly be off the mark. It’s an honest question and not meant confrontationally
Both fair to ask, as I throw around my own shorthand too easily, and a really meaty question in my mind, since it goes to the program’s manifold nature: It has been sold as a savings plan (and emphatically not a welfare plan), it functions as a scarily flexible pay-as-you-go transfer scheme with an insurance component, we tend to justify it as an insurance program or defined benefit plan while arguing increases or decreases in benefits as if it were a welfare plan or ‘safety net’. Unfortunately, basing plan-wide decisions on the living conditions of the neediest fractions (maybe 20%) of the broad beneficiary base may render the program unsustainable, which is obviously hugely unfair to future beneficiaries. I think we need to disentangle its parts, and return to the concept of a true safety net. I believe it is the best avenue to meeting both our moral obligations to serve current retirees and balance that with a sustainable program and economy for future beneficiaries.
Let me define terms:
A Savings Plan puts money aside, earns a rate of return, and returns the contributions plus the return to plan members. You get exactly what you put in out, more or less depending on the return on investment. Also called a ‘defined contribution plan’, it is the form of 401ks and cash savings plans.
Insurance pools risk. You pay a premium. If a rare event happens to you, you get a payment – substantially more than you put in, paid for by people who did not suffer the event. Insurance plans typically build reserves against losses, which are variable (otherwise it isn’t really insurance..)
Defined Benefit Plans combine savings plans with an income insurance feature. Typically, the risk of a low return over the savings period is insured by protecting a certain level of nominal income. The plan provider assumes some of the return risk or lays it off on an insurance company. Nonetheless, reserves are set aside from other liabilities and expenses to fund the future expenses of the plan. This is required by law of all private plans, and similar, but horribly porous, rules apply to government plans.
Welfare Progams take taxes from one set of citizens and transfer them as payments to another set of citizens. Typically those taxes would be taken from the wealthier and given to the poorer. In the US, we use this year’s income as a proxy for wealth and make a certain amount of pure transfer payments.
Social Security has been described to the public more as a Savings Plan, Defined Benefit Plan or “old age insurance” (see this history, and Paul Samuelson’s characterization of it) than welfare, depending on the audience. In the budget it is called “Old Age, Survivor and Disability Insurance”, or OASDI). Indeed, life expectancy was much lower at its inception, so living past 65 was an insurable event rather than the majority’s experience. In any event, it was more salable as a savings plan, with people “getting out what they put in”. In 1935 it was almost entirely saving plan, with only the payer getting benefits. In 1939 they added survivor benefits and in 1956 they added a pure insurance feature – disability benefits.
In truth, the mechanics never quite worked like a retirement plan of any sort. The functional design is “pay-as-you-go”. The government collects payments from current workers and pays out benefits to retirees. It depends on the population of contributors to grow relative to those withdrawing to remain solvent. These are the precise mechanics of a Ponzi Scheme (although that doesn’t make it morally equivalent to Bernie Madoff, try as some might). Ponzi mechanics probably seemed quite plausible at the time. Our current situation of decreasing workers: retirees and the associated burden on earners must have seemed very unlikely. Happily, politicans developed some foresight later and began to build up a “Trust Fund” to prepare for that day. Unfortunately, Congress then decided to include those revenues, without the associated future expenses they were raised to offset, in the current budget and used it to pay non-plan current expenditures. This pissed away the “savings plan reserves” and locked in a dramatic need for future borrowing to support the future pool of retirees.
Here’s why it was never a savings plan, despite the rhetoric – there is no legal requirement that the system pay out what was put in. It is nice to say people “regard Social Security as earned and owed them” and I agree in principle. But that doesn’t actually define the amount they are owed, we amend that as we go along. Politicians define it, and they’ve changed the definition over cohorts and time to skew returns in favor of people who earned less (albeit they had to have had a paycheck). You can try out changing incomes and birth years here, but you’ll find everyone gets a very low return, and younger contributors near the FICA ceiling are getting negative returns (or look at the graphs in this PDF). Even the much-demonized reform proposed by President Bush suggested “progressive indexing” to significantly accelerate the progressive nature of social security. I think this trend is generally a good thing, but it leaves the debate in a confused place. Social security is a hybrid insurance, defined benefit and wealth transfer plan. We call it an “entitlement” because it is too broadly applicable and too frequent an expense to be insurance, but it is restrictively defined to those who paid in, therefore it isn’t welfare.
There’s nothing intrinsically wrong with a hybrid program to combat multiple risks. It is certainly serves political rhetoric because the “OASDI” can be characterized conveniently in any context or with any bias. On the other hand, a muddled blob is usually a sub-optimal way to solve a problem. Since we call it a savings plan, we stick with the current somewhat regressive means of financing it, and mix up people with means and people in hardship. Since it is nominally “insurance” the savings plan doesn’t have to show a reasonable return. Because it is part wealth transfer plan, we can always argue on a welfare basis that some the beneficiaries deserve more, or that moral hazards exist. Because the pay out is defined flexibly, we can argue, one way or another, on how features like inflation indexing might work and balance the interests of future beneficiaries against present beneficiaries. Because no formal reserving is established, we can argue the interests of future generations against present in a completely unstructured and subjective manner.
For these reasons, it is difficult to have a rigorous discussion about it, and far too easy to fall into demonization. Arguments can come from any of these perspectives on what Social Security is or is intended to be. One thing, however, is for sure. Our current system of blobby “entitlements” is going to take up a huge chunk of GDP in the future, which will absolutely dampen growth. This is driven by 1) demographics 2) healthcare inflation and 3) under-saving. We have a chance of making this burden smaller by increasing productivity. Paradoxically, it would be easier to grow productivity if government transfers didn’t absorb more and more of the economy (they generally do not contribute to increased productivity as much as the private sector).
Thus I come to the conclusion that we cannot afford “entitlements”. However, we have a moral obligation to take care of the elderly and poor. We need to start means-restricting social security and medicare, making them true wealth transfer programs. The chief problem with this is that welfare programs have not been politically sustainable. It would also make it much easier to argue that we should support higher living standards for those who depend entirely on the program, because the long-run impacts to the program would not be as large.
Disability and catastrophic healthcare are easier. These are still low frequency events, and can be handled with an actual insurance program. Both ought to be government provided catastrophe programs. The rest, the so-called “working layers” can and should be left to markets, to give productivity and market forces a chance, with means-tested government assistance.
One of the other problems with social security is the disincentive it creates to work on the books. Far too many domestic and temporary workers stay off the tax rolls (the IRS thinks the revenue gap is about $345 Billion!). These people will retire as well, and they won’t have benefits at all. But the current taxes on entry level earnings are punitive enough that many will refuse low wage work if it has to be on the book (some, also, are illegal, and can’t go on the books). We aren’t helping these people.
So none of this answers the chained-vs-conventional CPI question. In previous comments Rod cited how elderly he knew in New Hampshire would be impacted by even a small change in the CPI calculation. This is no doubt true, but it is a ‘welfare’-type argument, pertaining to only about 20% of the beneficiary pool. The CPI change, however, would apply to the entire beneficiary pool and therefore have a significant effect on the program, potentially worsening generational inequities. He points out that current beneficiaries believed “they earned it”, although the idea that they earned non-chained CPI specifically is clearly not required by the program. It is natural to want to help or protect those 20%, but perpetuate a system that brings in even more unpleasant long-term tradeoffs and future difficult moral decisions?
The trouble with our current situation, and one of the clearest problems facing western civilization, is that selective safety nets aren’t politically sustainable, but broad entitlements don’t seem to be economically sustainable.
ASIDE: I don’t take the question as confrontational at all. One thing about me, though, is that process matters – in actions and rhetoric. I get frustrated with arguments from mixed premises, anecdotes or logical leaps – even if I agree with the conclusion. So, as you can see, I spend a lot of time obsessing on definitions and premises. All in all, I think it’s a good thing (and a VERY good thing professionally), but it is an extreme INTP trait and pisses a lot of people off. Rod’s argument seems to conflate the vague idea of “they earned it” (a savings plan argument) with “we should improve their standard of living” which is a valid moral argument, but one that is needs-based. Pursuing the latter in the least expensive way possible is the right approach. TANSTAAFL.
Another interlocutor suggests that excluding moral arguments is the best way to dismantle the program. To be clear, I do not exclude them, I only wish to restrict them to the applicable population and measure the cost/benefit equation transparently. On the other hand, he may be right that OASDI and Medicare would become politically unsustainable if we reduced them to their safety net components. I said as much above. Nonetheless, it is just as bad to drive an unstable program into unsustainability. OASDI’s sustainability is debatable. Medicare is clearly a massive problem.