Arnold Kling  

Social Security Risk

PRINT
The Economics of the Israeli-P... Health Care Celebrity Death-Ma...

I did not take away much from the first day of Cato's Social Security Bash. I guess my views on the subject are pretty well settled, so that for me the session served to illuminate what other folks are worried about than anything else. Unfortunately, I have to miss tomorrow's session, where Ed Prescott talks about labor supply elasticity.

A question that keeps coming up is, "How big is the shortfall?" Basically, I think that the present value of the future gap between promised benefits and tax revenues under current law is likely to be large. The only way to make it small is to (a) apply a high discount rate and (b) start a policy to reduce the shortfall tomorrow. Starting a policy tomorrow (such as raising payroll taxes by 1 percent of payroll) allows you to accumulate revenues and earn interest on those revenues to pay off future shortfalls. (This assumes, of course, that Congress "locks up" the increased revenues.)

Here is a paradox: If the shortfall is not large, then it appears to make sense to do nothing. However, if we do nothing, then the shortfall will be large. Only if we do something soon will the shortfall be small.

Here is another paradox: The people who want to apply the highest discount rate (and who want to discount the shortfall in the latter portion of this century to zero because it is "so far off in the future") argue that Social Security is low risk compared with personal accounts. But if it is low risk, then a low discount rate should apply.

I think that the risk issue actually is quite important. Critics portray personal accounts for young workers as an exchange of their "risk-free" social security benefits for risky stock market assets. In fact, given the magnitude of projected shortfalls, I think it is open to debate whether long-term promised Social Security benefits or stock market returns constitute the riskier asset.

For Discussion. Given the political economy involved, is there any way to estimate the risk to someone aged 35 that they will not receive Social Security benefits as large as currently promised?


Comments and Sharing


CATEGORIES: Social Security



TRACKBACKS (1 to date)
TrackBack URL: http://econlog.econlib.org/mt/mt-tb.cgi/196
The author at Right Mind in a related article titled Carnival Of The Capitalists writes:
    TITLE: Carnival Of The Capitalists URL: http://right-mind.us/archive/2005/02/14/1739.aspx IP: 198.206.162.134 BLOG NAME: Right Mind DATE: 02/14/2005 09:48:46 PM [Tracked on February 14, 2005 9:48 PM]
COMMENTS (23 to date)
jamesl writes:

This sounds like the beginning of a bad joke: How many economists does it take to calculate a shortfall?

It is important to present the problem to the American Public in a way that is correct, easy to understand and makes sense. One method would calculate the shortfall (if nothing is done) each year for 30 years and show it as a table. This will honestly show today's small present value as well as the large cost of waiting.

The table will demonstrate how both the "today we have only a small problem" proponents and the "sky is falling and we'll run out of money if we do nothing" proponents can be correct.

Franco writes:

Sorry if this is off the "for discussion" topic.

If social security is a bad deal actuarially, how is it that the shortfall is getting worse over time? I have 2 theories, though both seem unlikely to me:

1. Longevity increases are such that SS is no longer a bad deal (or will cease to be at some point soonish). That clearly would be unsustainable, of course, but I'm pretty sure that's far in the future, if at all.

2. If we imagine that a portion of payroll taxes go to pay for an actuarially fair benefit, and the balance go to paying off the implicit debt used to fund the benefits of the earliest recipients (who did get a good deal), then maybe we're not keeping up with that debt service (or will cease to keep up at some point). If this is the case, it seems that determining how much (if any) forced savings for social insurance there should be, and how paying (once and for all) for the early recipients should be divided within and across generations, are separable questions.

3. A third theory: It's not really getting worse, it's just that the way it's measured makes it look like it's getting worse.

Basically, it seems that without one of the first 2 things being true, if each cohort is getting a bad implicit rate of return on their payroll taxes then the program must be benefiting economically. But then why are things getting worse?

A related question is: what is the present (actuarially fair) value of already accrued benefits? In other words, what would the cost be of ending SS now, but continuing to pay benefits already earned? Ideally I guess this would include an estimate of what it would cost to allow people over a certain age to opt-in to keep paying payroll taxes and keep accruing benefits.

Edge writes:

I think the biggest medium term problem is simply assuring that any surplus generated with payroll taxes is not subject to political risk, and is put into an asset that has relatively low risk with a good risk/return ratio and with appropriate liquidity in 2025 and beyond (no, diverting funds to private accounts doesn't qualify, if only because it requires massive transfers from general revenues well after the TF assets are exhausted).

It makes no sense to raise additional payroll tax revenues unless that problem is solved, given the large on-budget deficit and the clear political risk that SS bonds might be defaulted or held political hostage.

It makes sense for general revenues to adjust now in anticipation of SS cash flows from 2020 forward, perhaps with a coincident change to SS bylaws that stipulates indexing changes for benefits with about a twenty year lead time to assure the trust fund ratio and cash flows smoothly handle non-action by congress. There is already a clause that would adjust indexing any years when the trust fund ratio reaches 0.20; that clause was written around 1984, when the ratio was practically zero. It should be revised now, to assure benefits are smoothly adjusted and the trust fund does not ever go negative.

Then, suppose the SS trust could be put into some other assets. What assets are appropriate? Or, would conservatives stop gaming the trust fund, if the cash flow issues from 2028 onward are addressed, so that treasury bonds can be held?

Edge writes:

Franco, I believe present value of accrued benefits is something like $8.5 to $10 Trillion. I ran a net present value calculation using 4% of GDP now, up to 6% of GDP of GDP by 2030 and beyond, with 100% for the first 10 years, scaling back over the next 30 years after that, and got to $8.5 Trillion. That's a really rough edged guess, but I've seen others use $10 T.

Take out the $1.5 Trillion in the trust fund, and assume that is due back from general revenue as a starting point.

Also recall that about 20% of SS revenues are for disability benefits, though that's accounted for in the 4% and 6%.

Randy writes:

The risk in the current system has 2 major components; the risk of economic conditions resulting in a reduced ability to tax, and second, political risks, including changing national priorities and the possibility of war.

I believe it is important that 100% of this risk is born by future generations. This is clearly seen today as risk to the current generation of retirees (or those over 55) is considered off-limits. The nature of the major components of risk makes them impossible to calculate for "any specific generation". But they are precisely calculable for "some future generation" - 100%.

The idea of private accounts does not eliminate risk, or necessarily even reduce it. But it does put risk and return back together.

Boonton writes:
The risk in the current system has 2 major components; the risk of economic conditions resulting in a reduced ability to tax, and second, political risks, including changing national priorities and the possibility of war.

By definition these two risks also exist for any 'private' account invested in any type of security. By definition such an account would also have the additional risk associated with the market and the specific investments it makes. Unless you can show that those risks are zero social security by definition has to be less risky.

Boonton writes:

BTW, where is this 'huge shortfall'? Projections either show Social Security being financially stable for the next 75+ years or having a deficit of 2% of GDP a half century from now...this in comparision to a 4% of GDP deficit we are running today.

Randy writes:

Boonton,

By definition these two risks also exist for any 'private' account invested in any type of security. By definition such an account would also have the additional risk associated with the market and the specific investments it makes. Unless you can show that those risks are zero social security by definition has to be less risky.

Good point. But my point is not to say that private accounts have no risk, just to point out that the risk taken in private accounts is by the same individual who will receive the returns. This is not true of the current system. Due to the nature of the risk components (particularly the political), this is in my opinion, a fatal flaw.

Boonton writes:

Which is part of the point of Social Security. It assumes the risk on the individuals behalf & its risk is tied to the underlying economy which is a lot more stable than a particular sector or stock in the economy.

It seems to me a diversified retirement portfolio should seek to diversify the risks. Social Security's risk are often in the opposite direction than private accounts making it a good counter weight in my opinion.

Since political risk and economic risk exist for both private accounts & SS, let's exclude them for the time beign. Look at a few types of risk and how SS and Private Accounts serve to balance each other.

SS* You won't live long after retirement therefore you'll only get to enjoy a handful of SS checks.

PA* You'll live a long time after retirement hence your nest egg may not last. If you purchase an annuity it will likely be quite expensive if your expected lifespan is large.


PA* Your returns are unpredictable. The guy who retires the year before you may have 10% more in his account just because he hit a lucky streak while you may have a lot less! To a degree this can be reduced by intelligent investing but it can never be entirely eliminated.

SS* Your income is predictable while your return varies only by your lifespan.

PA* An unexpectedly long lifespan means you will likely be faced with a crunch when you are the least likely to be able to return to work.

SS* An unexpectedly long lifespan only increases your return.

Starting a policy tomorrow (such as raising payroll taxes by 1 percent of payroll) allows you to accumulate revenues and earn interest on those revenues to pay off future shortfalls. (This assumes, of course, that Congress "locks up" the increased revenues.)

And, we know how that turned out from 1983.

Randy writes:

Boonton,

Which is part of the point of Social Security. It assumes the risk on the individuals behalf...

A system cannot assume risk. Only individuals can. And the individuals that are assuming the risk are all part of future generations. You are correct though - Social Security is a great deal for those in the current generation who get a guaranteed return with zero risk.

It is often said that Social Security is 70 years old and has been a great deal. That is true, but misleading. A more encompassing statement is that Social Security is only 3 generations old and already the bill is coming due.

Lawrance George Lux writes:

The greatest risk to Private Accounts is the one no one considers: the total period of a Worker's life that he is unemployed. Private Accounts disappear if they can be drafted, reduce in initial investment if they cannot. Score one for the current system.

There is actually only a 30% chance of a Shortfall in the current system, under conditions of relative full employment--this means putting another 6 million Workers back to work. Risk--current system and Private Accounts break even.

Longevity may have peaked, so there is only a 12% chance of a major Shortfall--Score one for the current system.

The Workers to Retiree ratio may turn around with more Workers in relation to Retirees. If this starts to happen before 2035, then there will be no Shortfall unless more Benefits are guaranteed under the current system. Break even again between the Two.

Final point: The 35-year-old Worker may see as many economic downturns as have Baby Boomers. These Downturns may last as long as those of the Baby Boomers with as much economic impact. Under such a Case: Current system 8, Private Accounts -8, because the Shortfall will be eight times as great for Private Accounts. lgl

Boonton writes:
It is often said that Social Security is 70 years old and has been a great deal. That is true, but misleading. A more encompassing statement is that Social Security is only 3 generations old and already the bill is coming due.

Let's review my Very Simple SS Model(tm):

Assume a constant tax rate on GDP (T) and a constant rate of GDP growth (G). Finally assume a simple pay as you go system where benefits = revenues.

Generation 1 Tax = T * GDP

Generation 2 Tax = Benefit = T* GDP * (1+G)

'Return' to Generation 1

= Benefit / Cost = [T*GDP (1+G)] / (T*GDP)

Which cancels out to 1+G...in other words the growth of the economy. This is about as risk free as you can get because you are not relying on a particular company, industry or even part of the economy. You are relying on the entire economy itself.

Randy writes:

Boonton,

If the current system is so risk free, why have I already received multiple payroll tax increases? Why can't I retire until I'm 67? According to the calculators, my life expectancy is 71 years. Those extra 2 years constitute a 33% benefit cut. And why am I now faced with an additional 23% benefit cut, or more payroll tax increases, or not retiring until I'm 70?

And why are you so bloody sure that you and the government knows what is best for me? You know full well that if the system were made voluntary, I and many others would opt out immediately. I have to assume, therefore, that you have a stake in this. What is it?

Edge writes:

Randy, the payroll tax increases up until 1975 were planned from the start, from what I've read. oth benefits and the tax rate were phased in.

We used to have ad hoc increases in benefits, but the 1983 reform fairly well set the actuary 75 year actuary standard and stopped the practice of increasing benefits without reflection on the out year consequences.

Boonton, yes you get the growth of the economy as the benefit, but you also get that smoothed out over many years. Your benefits are more predictable than the growth of the economy. You also get a likely bailout if something really nasty happens and equity simply evaporates. The government can continue to tax, and can find work for people to do, if that is necessary to keep payroll taxes flowing (and keep people working). As occured in Germany, when there were three occaisions last century when equity evaporated, but benefits kept flowing comparatively smoothly.

Boonton writes:

If you received multiple tax increases you will probably be of age to retire before the 67 cap is put into place (it's being phased in quite slowly). Over the history of the program lifespan has grown faster than the belated increase in the retirement age so if anything years should be added to the expected payout period.

In the Simplified Model, benefits = taxes which is what it should be in the long run. In the 80's Greenspan's commission increased taxes in order to build up reserves to pay for the boomers retirement. If you look at the graphs you'll see that SS revenues are running above SS benefits until sometime around 2018 (or later depending on growth assumptions) when the lines cross and benefits outstip revenue for a while. In the long run both benefits and revenues should return to where they are roughly equal.

Who ever said I claimed gov't knows best? Bush just handed you a nice tax cut plus even before Bush IRA/401K accounts have enjoyed a generous tax subsidy. No one is stopping you from saving for retirement.

Randy writes:

Boonton;

No one is stopping you from saving for retirement.

You are absolutely right. I've known that personal savings was my only hope since the time of the Greenspan commission. The estimate I get from the Social Security administration says I will get about $700/mth. I'm giving that to my daughters, they'll need it more than I will. And of course, there's the matter of refusing to participate in the looting of the next generation.

Boonton writes:

Assuming you live 7 years after you start collecting $700 a month to your daughter will total $58,800. She better remember to call on Father's Day!

Randy writes:

Boonton,

Three daughters. I hope they do call. But its nothing special. I'm just returning money that should never have been taken from them in the first place.

BTW, I went to your website - very informative.

Boonton writes:

Thanks Randy! Hope you clicked an ad :)

If only I didn't spend so much time commenting on other people's sites I could write more on my own!

Dewey Munson writes:

All SS payments spent would have otherwise come from assets which ultimately become your estate.

SS benefits effectively go to surviving generation.

Dave Altig writes:

Arnold --

I have the same skepticism about the persistent treatment of social security benefits as "riskless" (as long as the current administration is held at bay, that is). You can find my full argument at the link below, but here's the short version:

In what sense is the present social security system riskless? I'd say right now I'm pretty uncertain about what my true returns are likely to be. If I was 20, I'd be even more uncertain. Commentary in the blogosphere is chock full of discussions about productivity trends, demographic assumptions, and the like, all of which are highly uncertain. There seems to me very little probability that, twenty years from now, the fixes we might put on the current system today will reveal themselves to have been exactly the right medicine.

http://macroblog.typepad.com/macroblog/2005/02/social_security_1.html#comments

Randy writes:

It occurs to me that social security is simply debt. And the best kind of debt too. One that the lender has no right to refuse, and that the borrower has no legal obligation to repay.

Comments for this entry have been closed
Return to top