Tuesday, December 28, 2004

Kinsley redux on Social Security

Michael Kinsley's restatement of his "logical proof" of why private investments in Social Security are a "certain failure" merely seems to repeat his first statement, rather than try to clarify or add anything to it.

As the gist of his argument seems unchanged I'll just link back to the prior answer given here (short and long versions) and ....

1) Repeat that it remains totally implausible that additional investment in stocks through private accounts of a mere $200 billion annually (as a theoretical top -- probably much less in reality) will "bid down" the yield on more than $33 trillion of stocks so much that anyone would even notice -- much less all the way down to the level of that on government bonds, as his "logical proof" dictates.

Can one really imagine that increasing investment by less than six tenths of one percent can have such a dramatic effect on stock prices?

2) Add something new to the discussion regarding bond prices and yields by pointing out that even purely logically it doesn't necessarily follow that borrowing today to fund future Social Security benefits will "bid up" bond yields at all, much less all the way to the level of that on stocks, as Kinsley's logical proof dictates.

This is because, as Alan Greenspan has stated, future promised Social Security benefits are a liability of the U.S. that already exist, and to the extent that the markets believe the U.S. will actually pay them, the cost of financing them logically should already be reflected in bond prices. In which case making these implicit obligations explicit by financing them today should make no difference at all. As the Chairman put it...

"A critical consideration for the privatization of social security is how financial markets are factoring in the implicit unfunded liability of the current system in setting long-term interest rates. ... If markets perceive that this liability has the same status as explicit federal debt, then one must presume that interest rates have already fully adjusted to the implicit contingent liability." [my emphasis]
This gets to the question of just how secure promised Social Security benefits really are. If they are truly as fully secure as many defenders of the status quo assure us they are, and the markets believe it, then funding those benefits today should have no effect at all on interest rates. After all, the government is already on the hook for all that money, so simply admitting it honestly shouldn't change anything.

On the other hand, if promised benefits aren't secure and are expected to be cut, then funding them today at promised levels would be an extra cost expected to bid up the current rate on bonds. But if promised benefits aren't secure, shouldn't the defenders of the status quo be honestly admitting that and openly addressing that as a problem? Yet they aren't.

This issue is discussed at more length in a recent post -- for here the point is that even strictly logically it is not necessarily so that borrowing today to fund future benefits will increase rates on bonds by anything like what Kinsley assumes must be the case.

(3) Repeat something important that Kinsley ignores. Even assuming that real investments only match the cost of paying down government bonds, a major benefit results from borrowing today to prefund Social Security benefits that will become due 30 years from now.

As noted before with data from GAO, by 2030 the government is expected to be under never-before-seen fiscal pressure with annual deficits rising towards 20% of GDP -- matching the size of the entire government today -- on the basis of current policies.

Now, say the government issues a $1,000 bond in 2005 that it will pay off over 30 years. It invests the money it receives from the bond in a portfolio of private investments worth $1,000. The return on the private investments just matches the cost of paying off the bonds, that is all. After 30 years the bond is paid off and the $1,000 in investments remain. Now there is $1,000 available to pay benefits in 2035 without the government having to increase borrowing or taxes during a huge fiscal crunch.

The cost of paying the benefit in 2035 remains $1,000 -- but the financing of it has been moved forward in time by 30 years to when it is much more affordable. Instead of the benefit being at risk in 2035 because Congress can't afford to raise taxes/borrowing yet more in the crunch to pay it, the $1,000 was raised when conditions were easier so the benefit it will pay is secure. This is not a good thing?

4) Note that Kinsley almost gets to a vital logical point that almost all defenders of the Social Security status quo dodge -- actually comparing options by considering the faults of the status quo as well as those they ascribe to whatever reform proposal they are criticizing -- but then he dodges like everyone else.

Many responders made a related point, "Compared to what?" Whatever its flaws, is privatization inferior to the current system, with its looming inability to keep its promises? ...

But privatization is not supposed to produce a net loss in anyone's retirement nest egg ...
... he proclaims, mistakenly deducing that it will do so -- but also, and more importantly, ignoring how the status quo is guaranteed to do exactly that.

The Social Security Administration's actuaries' say the average-wage individual who entered the work force in 1994 -- and so is in his/her late 20s or early 30s today -- will receive in benefits from Social Security 15% less than contributed to it through taxes over a lifetime. For high-wage individuals benefits will be as little as 50% of taxes.

And even these promised benefits are 30% underfunded. So if Social Security remains "paygo", with no investment returns supplementing it, then by the Iron Laws of Arithmetic the average-wage individual's return will drop to 40% less than taxes paid and the high-wage individual's return to 65% less. (In a paygo system a funding gap can be closed only by reducing benefits or increasing taxes, and both acts further reduce the ratio of benefits to taxes.)

Of course, this is the huge change in the nature of Social Security that defenders of the status quo -- including Kinsley, it seems -- never, ever want to mention.

Before the year 2000, all annual cohorts of retirees received from Social Security more than they contributed to it, with most receiving much more. After 2000 all annual cohorts of retirees will receive less than they contributed to Social Security, they will be made poorer by it on a lifetime basis, with this loss growing as the years pass.

So instead of saying ...

... privatization is not supposed to produce a net loss in anyone's retirement nest egg ...
... shouldn't Kinsley be saying...

... Social Security is not supposed to produce a net loss ...
... ? I'd think so.

And if he ever gets around to doing this, and admitting the size of the net losses Social Security is committed to produce, what will be his proposed remedy?

One remedy is private investment accounts. Even mere 2% accounts that earn the historical norm on private investments, when compounded for 40 years, can give most young workers positive net returns on Social Security -- and so save it as we know it. (And again, does anyone else really believe with Kinsley that increasing investment by less than 0.6% will drive such returns down?)

But without private investment accounts, how is this to be done, how will Social Security avoid making today's young workers poorer after making all prior generations richer?

This is the question Michael Kinsley didn't bother to consider, much less answer about the status quo when he said private accounts shouldn't reduce nest eggs. What about the status quo?

Even Sweden has 2.5% private accounts in Social Security for these reasons. It seems the Swedes haven't yet figured out the logical proof of how they can't work!

Swedish social policy -- too right-wing for Michael Kinsley and U.S. Democrats.