Thursday, September 08, 2011

Is Social Security a Ponzi scheme?

Since Rick Perry keeps calling Social Security a Ponzi scheme, let's examine the accuracy of this characterization.

A true Ponzi scheme has two main elements. First, new investors' contributions are used to pay old investors' benefits. Second, an exploding or unsustainable growth rate is needed to keep the promised or expected benefits coming. (A chain letter where you ask six people to send you a dollar, and they then each ask six people to send them a dollar, is a classic example.)

Social Security has the first of these two elements, reflecting program cash flows and the initial decision, made when the program was started during the Great Depression while millions of seniors faced ineradicable poverty, to start paying benefits to retirees who had not significantly contributed to the program.

This alone, however, is not problematic in the way that the term "Ponzi scheme" inevitably suggests. If you think it is problematic and thus justifies the label, here's another non-exploding and seemingly Ponzi-like scheme for you. My family, like many others, has for countless generations been running this incredibly Ponzi-like scheme called "parenting." As a baby you get these free benefits, paid by your parents via their labor in raising you. Then when you grow up you pay in to the plan (if you have children) by raising your kids.

Surely this is even worse than Social Security. After all, at least in Social Security you pay your taxes before you get your benefits. Here, you get your benefits first! But there's still the key feature that you don't self-finance, i.e., raise yourself from infancy. Instead, each cohort relies on an adjoining one to pay it. Yet somehow this audacious scheme has proved sustainable over time.

So there really is no Ponzi scheme unless an exploding or unsustainable growth rate is needed to keep the promised or expected benefits coming. How does Social Security rank in this regard?

As it happens, the program does not have a well-defined relationship between taxes paid in and benefits that are promised. This depends on how the payroll tax rate and base on the one hand, and the benefit formula on the other hand (each subject to statutory modification at any time), happen to play out given birth rates, life expectancies, wage growth, employment levels, etcetera.

But in Paul Samuelson's famous conceptualization of Social Security in a classic 1958 article, we can think of the program as one in which everyone would get back exactly the amount they paid in if, among other simplifying abstractions, wages and population were constant over time and everyone's life consisted of a fixed "work period" followed by a "retirement period." Throw in population growth and rising real wage levels, and Samuelson foresaw a positive rate of return to Social Security retirees (possibly exceeding the real interest rate, in a Peter Diamond extension of the model) that depended on those two factors. That is, if we imagine payroll taxes being handed over to retirees in a strict pay-as-you-go system, the amount available to be paid over rises if the workforce grows along with wages that are subject to the payroll tax.

There's nothing Ponzi-like about that; it's entirely sustainable. But in actual Social Security, two things went "wrong." One was the baby bust after the baby boom, while the second, more important change was rising post-retirement life expectancies. Having retirees live longer was equivalent to having more of them, and had adverse effects on worker to retiree ratios even if each demographic cohort was larger than the one that came before.

Does this mean that Social Security became a Ponzi scheme after all? Strictly speaking, absolutely not. If we think of Social Security retirement benefits as being adjusted to reflect changing payroll tax revenue levels (even though this is only true, if at all, over the long run), then we'd say that retirees hold an implicit financial instrument, the payoff on which depends on wage and demographic trends. So all that rising life expectancies does in this scenario is cause the payoff to be lower rather than higher than it would otherwise have been. But that's in the nature of the implicit financial instrument - a built-in feature that does not cause Ponzi-like collapse, but merely affects the actual payouts from a program that can keep on running anyway.

This brings us to the point that comes closest to justifying Perry's angry braying about Social Security. The program's retirement benefits do not automatically adjust for these demographic changes. Instead, barring Congressional legislation, they proceed on statutory autopilot (albeit depending in practice on demographic and macroeconomic outcomes). So if adverse demographic changes do not lead to immediate tax or benefit changes, you get a program shortfall that emerges over time, and currently promised benefits become eventually unsustainable without new financing.

That, however, is merely lag in adjusting the actual rules on the books to keep on track with the Samuelson structure that is implied by having a largely pay-as-you-go scheme. And it does not mean that the program will collapse - merely that benefits will need at some point to be adversely adjusted (say, to the tune of 20 or 30 percent) in the absence of increased financing.

Perry is right (words that I must confess I hate typing) insofar as his point is that the current scheme requires adjustment, and that people won't get the full benefits promised by present law unless there is extra financing. But he is very substantially wrong in comparing this to a Ponzi scheme in which, as we well know, the investors (except for the lucky ones who got out fast) end up with nothing.

It would be more accurate to compare Social Security to an investment that has historically produced one rate of return, but which in fact appears likely to offer you a lower rate of return. E.g., suppose stock prices have historically appreciated, over a long period of time, at about 2% annually in real terms. You now hold stocks and are hoping to earn that rate of return. But suppose we can see the future well enough to anticipate that, in fact, you will end up learning less than that (and perhaps even a zero or modestly negative rate of return). That's too bad, but it doesn't make stock market investment a Ponzi scheme.

The negative adjustment to expected returns also obviously does not make mandatory retirement saving (a key feature of Social Security) a bad idea, given that lifetime consumption smoothing is necessary unless you're happy to have two houses and two dinners a day now, followed by none of either when you're old.

In sum, "Ponzi scheme" is really out of place as a description of Social Security, even though the program has financing problems. It's true that there have been some negative shocks to expected returns from the program, and that since it doesn't automatically self-adjust there will be a rising expected long-term program deficit until Congress gets around to adjusting things. But this is way out of Ponzi territory.

What about Medicare? That program is best described as Samuelson-plus. Program participants "bet" not only on economic and demographic trends, but also on the trend in healthcare expenditure relative to GDP. (When that rises, the program becomes costlier under constant benefit design.) What makes Medicare far more fiscally unsustainable than Social Security is the fact that healthcare expenditure levels have been significantly rising relative to GDP - and unsustainably so whether there is a government financing role or not.

Medicare is closer to Ponzi territory if we posit an implicit commitment, not just to the current statutory design, but to covering so large a percentage of seniors' healthcare outlays under a system where they so frequently can get the best procedures that are technologically available at a given moment. The built-in unsustainability comes from purporting to guarantee something that is itself growing at an exploding rate. But that is a piece of the broader healthcare problem that we face, more than an aspect of Medicare as such. So calling Medicare a Ponzi scheme (which even Perry does not seem inclined to do) is less illuminating than saying that healthcare generally is on an unsustainable growth path in our society, that we will need to address in one way or another.

UPDATE: Looking at Perry's statement more closely (although in this post I was more interested in the pervasive "Ponzi scheme" meme than in his murky mental processes), I see that his operating definition appears to be that, if the system is going to go kaput and pay you zero, then from your standpoint it is a Ponzi scheme. That is certainly a reasonable way to define Ponzi schemes. But of course it is wildly inaccurate as applied to Social Security, which is projected to be able (based on expected future Trust Funds) to pay about 75 percent of future retirees' benefits. So he is using what we will charitably call his own facts, rather than the actual ones, in calling Social Security a Ponzi scheme.

Interesting point about Perry: While he obviously feels entitled to his own facts, rather than the actual ones, on global warming, evolution, Keynesian stimulus, Social Security, etcetera, he apparently caused his last two gubernatorial campaigns to run very serious empirical tests regarding how alternative types of campaign expenditures and activities actually contribute to electoral success. See the NY times blog article here. In other words, this would appear to be a guy who (contrary to so much evidence from his cheap talk) actually knows and cares about expertise and empirical proof, in cases where there is something in it for him. But if there is no direct personal benefit to him, then he evidently doesn't care.


Ellen Livingston said...

Excellent points. Perry, of course, doesn't care that it's not a Ponzi scheme - it's just the sort of simplistic slogan-like language that works exceptionally well with "low-information voters," and he knows it. In a sense, SS and Medicare are victims of their own success - increased life expectancy is due at least in part to a higher standard of living and better medical care these programs have promoted. Even the AARP admits that they will need some adjusting. It doesn't take a rocket scientist to see that 65 isn't the same age in 2011 as it was in 1935, when life expectancy in the US was actually below 65. You do have to wonder, though, how effective Perry's SS-bashing is as a political tactic: it may play very well to his base, but last time I checked seniors like to vote on Election Day.

JPB said...

This is a very useful summary of SS and more or less rebuts the characterization of SS as a pure Ponzi scheme. But what about one other Ponzi element, which is that the last people in don't get paid anything. What if we decided that the population of the earth was unstustainable and certain countries, such as the US, volunteered to stop reproducing? Wouldn't the last generation to contribute to SS get nothing in that case? Maybe too farfetched, I'll grant, but that seems like a significant ponzi like feature.

Anthony said...

I had read somewhere (can't quite remember where now) that the "increased life expectancy" problem is over-stated because the big mover here is not that people are continuing to live much later into old age than they used to; while this is happening, the bigger factor is fewer outliers amongst those who die very young (and hence, never joined the work force to begin with, and thus have a neutral effect on the system). I don't really know too much of the actual economics behind Social Security, so I don't know if this is true, but it sounded plausible when I read it. But thanks for an excellent break-down of the "Ponzi scheme" analogy.