Orszag (and Diamond) Flunk Reading & ‘Rithmetic: SS Legacy Debt

Or maybe I did. Given that Peter Orszag is incoming head of OMB and I am, well, me, odds are I am the one in error. But I can’t find it. Maybe someone here can help.
This started as an exchange in comments on Is Obama Echoing Bush Andrew Biggs claims the following

As for the backwards transfer, this is pretty standard stuff now. Google “social security” “legacy debt” and you’ll find info on it — written by Peter Orszag and Peter Diamond, who last I checked weren’t part of the right wing conspiracy. As it happens, I haven’t argued whether the backward transfer was fair or not, in part because it doesn’t make any difference — what’s done is done. What I will say is that had early generations only gotten out what they paid in, plus interest, the trust fund today would be worth around $20 trillion and we wouldn’t face the prospect of tax increases or benefit cuts in the future. I’ll let others judge whether that makes sense or not.

and explains that with this

The current effect on the trust fund is equal to the difference between early participants’ taxes (plus interest to today) minus their benefits (plus interest to today) — in other words, the present value.

For those who haven’t been following along “backwards transfer” is a reference to some posts here from last August. $17 trillion Backwards Transfer & Backwards Transfer: Biggs Responds and the reference to Orszag and Diamond seems to be related to Saving Social Security: the Diamond-Orzsag Plan The link is to the intro, if you download the PDF you will find the relevant passage as follows:

Third, and finally, our plan addresses the burden of the legacy debt. Benefits paid to almost all current and past cohorts of beneficiaries exceeded what could have been financed with the revenue they contributed. This difference is what we call the legacy debt. Without this debt—that is, if earlier cohorts had received only the benefits that could be financed by their contributions plus interest, the trust fund’s assets today would be much greater. Those assets would earn interest, which could be used to finance benefits.

Well none of this sounds quite right to me for reasons I’ll outline below the fold.

First how much was actually paid out to these earlier cohorts and when? Well I went back and took a look at the tables, in particular Table IV.A2 which tracks income and cost from program inception right up to introduction of DI in 1957 and then Table IV.A4 which shows combined OASDI from 1957 to 2007.

Tracking just benefits paid:
OAS 1941-1956 $25.6 billion
OASDI 1957-1968 $181 billion
OASDI 1969-1980 $793.1 billion
OASDI 1981-1992 $2.476.5 trillion
OASDI 1993-2000 $2.829.3 trillion
OASDI 2001-2007 $6.330,9 trillion
For a grand total of $12.636.4 trillion. Of particular note in context is the total paid out from 1941 to 1980: $999.7 billion. Why is 1980 significant? Because that is the last year that a retiree could have started drawing benefits WITHOUT paying in his entire work career. That is someone 65 in 1980 would have been 21 in 1936. So we have this cohort of ‘deadbeats’ (which BTW includes most of the ‘Greatest Generation’) dragging down almost exactly a trillion dollars in benefits. But all of those benefits had been paid for, at the end of 1980 the Trust Fund was in very poor shape but still had $26.5 billion in accumulated surplus. Table VI.A4.—Historical Operations of the Combined OASI and DI Trust Funds, Calendar Years 1957-2007 [Amounts in billions]. And if we fast forward to today the youngest survivor of this group would be 93 years old and us sitting on a total surplus of $2.4 trillion.

So how does a fully paid for $1 trillion dollars in benefits to 1980, plus an additional $11.6 trillion in benefits fully paid for by the end of 2008 leaving a surplus of $2.4 trillion work out to impose trillions of dollars of ‘legacy’ debt, still less the $17 trillion amount suggested by Biggs? I don’t see that his formulation helps us any. Benefits paid in 1980 using 1980 dollars are off the books, calculating their present value doesn’t make much sense to me. In any event someone needs to show their work. Because from the initial discussion we know the source of the $17 trillion claim. It comes from misunderstanding the following table:

(I’ll try to get a bigger version in comments. Or maybe someone can give me an assist and blow this up a little)(Click on the image here or the image in the first comment for a larger version).
The problem arises from not quite getting the definitions of ‘past’ ‘current’ and ‘future’ participants as used by the Trustees and if I had to pinpoint the source of the error it would be in the following passage (bolding mine)

The first line of table IV.B7 shows that the present value of future cost less future taxes over the next 100 years for all current participants equals $17.4trillion. For this purpose, current participants are defined as individuals who attain age 15 or older in 2008. Subtracting the current value of the trust fund (the accumulated value of past OASDI taxes less cost) gives a closed group (excluding all future participants) unfunded obligation of $15.2trillion. This value represents the shortfall of lifetime contributions for all past and current participants relative to the lifetime costs associated with their generations. For a fully-advance-funded program this value would be equal to zero.

This does not bear the meaning that Biggs, and perhaps Orszag and Diamond, place on it. That A + B = $C trillion does not mean that A > 0. Still less that A = $17 trillion. Or even $15 trillion. Instead you have to understand these ‘unfunded liabilities’ include all projected gaps between revenue and cost for everyone born before 1994 all of the way out to 2108. Moreover all but $4.3 trillion actually occurs AFTER the 75 year window

The values in table IV.B6 indicate that extending the calculations beyond 2082 adds $9.3 ($13.6 – $4.3) trillion in present value to the amount of the unfunded obligation estimated through 2082. That is, over the infinite horizon, the OASDI open group unfunded obligation is projected to be $13.6trillion. The $9.3trillion increment reflects a significant financing gap projected for OASDI over the infinite future period after 2082. Of course, the degree of uncertainty associated with estimates beyond 2082 is substantial.

(BW: no shit). On my reading Diamond-Orzsag propose to start charging us in 2023 for a ‘legacy debt’ that mostly won’t be incurred until after 2082. Yet blaming it on past participants (i.e. dead people) getting trillions in extra benefits. Sorry I just don’t see how the numbers support that. Past participants clearly paid for everything they got, the Trust Funds never actually went to zero in their lifetimes (though it got close in 1983). And that portion of current participants who are already retired will have paid for almost all their benefits, given that a relatively small portion of such retirees are likely to be alive in 2041 or 2048. For that matter that portion of current participants within 15 years or so of retirement are mostly paid up.

Perhaps someone can come up with some explanation of why it makes sense calculating bills already paid off in then current dollars adding up to future liabilities. On a cash flow basis it seems nonsense, as long as your income rate exceeds your cost rate it is hard to see how you could realistically be said to be in any kind of debt situation. Table IV.B1.—Estimated Annual Income Rates, Cost Rates, and Balances, Calendar Years 1990-2085 [As a percentage of taxable payroll] It is hard to believe that two sentences could lead to so much confusion, but it looks like the culprit is right here:

This value represents the shortfall of lifetime contributions for all past and current participants relative to the lifetime costs associated with their generations. For a fully-advance-funded program this value would be equal to zero.

Someone has simply put the fully loaded Gen-X cart in front of the Boomer horse and telling us it is our duty to push. Well no we have been pulling our share of the load all along.