The Atlanta Fed's macroblog provides commentary and analysis on economic topics including monetary policy, macroeconomic developments, inflation, labor economics, and financial issues.

Authors for macroblog are Dave Altig, John Robertson, and other Atlanta Fed economists and researchers.

« Maybe They Read Vox Baby At The White House? | Main | The Heads Start Talking »

February 06, 2005

Social Security Once More

Speaking of Vox Baby, Andrew's most recent entry (as of the time I write) has his take on the Social Security part of the State of the Union address.  It is characteristically thoughtful, and I commend it to you.  But there are a few points on which I differ, or am not fully convinced.  But first, the points that I agree with (which make up the bulk of the post):

[The President] excluded the possibility of raising the payroll tax to restore solvency. That will make it more difficult...

I did not like the statement, "By the year 2042, the entire system would be exhausted and bankrupt."...

The President was clear that everyone 55 and over is in the traditional system with benefits as specified by current law. Good. He was also clear about a gradual phase-in. Good. He explicitly mentions the Federal Thrift Savings Program as a guide for the sort of investment options available. Very good...

An item not discussed--exactly how progressive the reductions in future benefits from the current system would be. I expect this to be another area for compromise--make the traditional system that continues even more progressive, to make sure that the lowest-income people, who are less able to bear financial risk, are depending less on the personal accounts than other workers.

So what are my quibbles?  They are minor, but here they are.  Andrew writes:

There is a vagueness about the "greater than anything the current system can deliver" phrase that is troubling. It doesn't mention the risk associated with obtaining those returns.

The whole issue of risk seems, to me, vastly overblown.  First, I've yet to see any compelling case made against Jeremy Siegel's observation that, for the long-run investor, equity does not look like a particularly risky bet.  (Although Professor Siegel concedes the possibility of lower-than-historical equity returns going forward, that would presumably go a long with lower-than-historical bond returns. This was pointed out in the comment section of this Vox Baby entry.)  The option for moving to a "safe" fixed-income account should, of course, be made available in a private account system -- maybe even mandatory at some point near retirement.  But most of the people I know under the age of, say 55, have the great bulk of their saving in equity funds.  I think revealed preference counts for something.

There is certainly the question of what happens if their is a gigantic market meltdown of the magnitude that eventually led to the creation of our current social insurance system.  It could happen, I guess.  But if it does, nothing like the revisions contemplated by those who favor the status quo would be sufficient.  In that case, the solution is a set of intergenerational transfers that look like the picture in this post.   Nothing on the table comes even close to that (nor should it).

Which brings me to the second point.  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.

This has led some to suggest that we implement reforms that include some set of automatic adjustments -- in retirement ages, benefit formulas, and the like -- to guarantee the program is always in actuarial balance.  In other words, they would make the net returns in the system explicitly state-contingent -- which they already implicitly are.  Which exactly makes my point.  It would be appropriate for the opponents of private accounts to at least acknowledge this uncertainty whenever they invoke the dreaded "risk" dragon.

I'd make a related point regarding this comment from Andrew:

It is a stretch to say that the government "cannot take these accounts away." I agree that it would (probably) be more difficult for the government to impose a surtax on the accounts than it would be to cut benefits coming from the pay-as-you-go system, but "cannot" is too strong.

The point is, there is substantially less reason to "take these accounts away" in a defined contribution system.  The problems inherent in the pay-as-you-go, defined benefit scheme are the reason we are talking about "take aways" today.

And finally, Andrew says:

The issue of bequests is completely unnecessary. Social Security exists to provide insurance against outliving one's means. Nothing prevents people from leaving bequests currently if they so desire. I am not aware of any failures in the life insurance market that need government attention.

I disagree.  Even if I don't plan to leave a bequest, I can still receive residual utility from unintended bequests.  And that possibility may cause me to provide more "self insurance" against outliving my means than I might otherwise,  potentially (at least) "bringing new money into the system," as Andrew desires.

Other than that, I completely agree with him.

February 6, 2005 in Social Security | Permalink


TrackBack URL for this entry:

Listed below are links to blogs that reference Social Security Once More :


re: risk

Seems to me that the biggest risk in the current system is the one you refer to later on. Assuming two similar individuals retire at 65, the one who dies at 66 gets 1X and the one who dies at 85 gets a discounted 20X, maybe 14X or so. That seems a far greater risk than anything that could happen to the equity markets.


Posted by: quietstorm | February 06, 2005 at 11:07 AM

The variance of the average return is not the long-run issue as was noted by Samuelson's 1963 paper. The relevant risk is the unpredictability of one's wealth upon retirement. Why do folks get these two concepts so utterly confused? Thanks to quietstorm for understanding the point Samuelson made clearly 42 years ago.

Posted by: pgl | February 06, 2005 at 03:48 PM

pgl --

I'm not sure that's exactly what quietstorm had in mind. In any event, I agree that short-run volatility is an issue as "time T" approaches, but that issue is manageable. The long-run averages seem to me exactly the right thing to focus for most of the life-cycle portfolio decision during working years. But maybe I'm missing something? (I confess: Samuelson 1963 isn't ringing a bell.)


Posted by: Dave | February 08, 2005 at 06:16 AM

Post a comment

Comments are moderated and will not appear until the moderator has approved them.

If you have a TypeKey or TypePad account, please Sign in

Google Search

Recent Posts



Powered by TypePad