by Bruce Webb
Articles this week by Kevin Hassett and Lori Montgomery sought to make the claim that deficits scheduled for 2017 had been moved up to today as a reason to immediately convene a bi-partisan Commission to Reform Social Security (with ‘bi-partisan’ meaning dominated by Blue Dog Dems and Conservative Republicans). Well first of all Social Security surpluses don’t disappear in 2017, at least ‘surplus’ as defined by CBO and second of all it really doesn’t matter that cash surpluses (Income excluding Interest) are projected to fall behind cost in that year. Understanding why requires some understanding of how Social Security manages the Trust Funds. The following shows assets in the OASI Trust Fund at year end 2006 and 2007 (the latest data available). Click on the image to get an enlarged version.
Discussion below the fold.
Generally when talking about Social Security we talk about THE Social Security Trust Fund having a current balance of $2.4 trillion and a shortfall date of 2017, a depletion date of 2041 and a payroll gap of 1.7%. And this includes the Trustees, most of the Tables in the Annual Reports talking about the OASDI Trust Fund. But in law and in actual operation there is no such thing as an OASDI Trust Fund, instead there is an OASI Trust Fund for the Old Age/Survivors Insurance program and a totally separate DI Trust Fund for the Disability Insurance program. Which means among other things that each Trust Fund has different dates for shortfall and depletion and different payroll gaps. Moreover they have different degrees of uncertainty, the range of outcomes for DI being much wider than that of OAS. Plus DI is much smaller being about 10% of the size of OAS. All of this has some implications for policy, which I will describe in a later post. (In fact until I took a closer look at the following figure I didn’t realize how big those implications were.)
But for now I want to concentrate on the assets in and the interest on the OAS Trust Fund. The Trustees explain it like this.
Another source of income to the trust funds is interest received on investments held by the trust funds. That portion of each trust fund which is not required to meet the current cost of benefits and administration is invested, on a daily basis, primarily in interest-bearing obligations of the U.S. Government (including special public-debt obligations described below). Investments may also be made in obligations guaranteed as to both principal and interest by the United States, including certain Federally sponsored agency obligations that are designated in the laws authorizing their issuance as lawful investments for fiduciary and trust funds under the control and authority of the United States or any officer of the United States. These obligations may be acquired on original issue at the issue price or by purchase of outstanding obligations at their market price.
The Social Security Act authorizes the issuance of special public-debt obligations for purchase exclusively by the trust funds. The Act provides that the interest rate on new special obligations will be the average market yield, as of the last business day of a month, on all of the outstanding marketable U.S. obligations that are due or callable more than 4 years in the future. The rate so calculated is rounded to the nearest one-eighth of one percent and applies to new issues in the following month. Beginning January 1999, in calculating the average market yield rate for this purpose, the Treasury incorporates the yield to the call date when a callable bond’s market price is above par.
Although the special issues cannot be bought or sold in the open market, they are nonetheless redeemable at any time at par value and thus bear no risk of fluctuations in principal value due to changes in market yield rates. Just as in the case of marketable Treasury securities held by the public, all of the investments held by the trust funds are backed by the full faith and credit of the U.S. Government.
The first thing to note is that the Trust Fund is not a unitary account subject to the current interest rate. Instead it is a portfolio of specific certificates and bonds with specific rates and maturity dates. As noted by the Trustees the bonds are not affected by current interest rates, the $27 billion in 2009 bonds yielding 7.25% are going to add $1.95 billion to the Trust Fund no matter what happens to rates this year. Meaning the total portfolio is going to add interest to the TF of around $110 billion. At year end the OAS Trust Fund will have a net increase in assets over the year, that people don’t want to call that increase a ‘surplus’ is their problem.
This doesn’t mean that the Trust Funds are immune to changes in current interest rates, the higher yields of the bonds maturing in 2009 will be replaced by new lower yielding ones, but this only effects a fraction of the portfolio. If we examine the Table we can see that the $290 billion in new issues in 2007 have yields of 5.0%. About a third of that represented by new issues due to interest, a third on rolling over maturing bonds earning less than 5.0% (though not by much) and a third rolling over maturing bonds earning more than 5.0%, the total effect was a near wash. The 2008 Report will likely see a bigger effect and the 2009 even more on projected balances but those $290 billion in 2007 issues will continue to yield 5% until their respective maturity dates.
Which leads us to 2017. Currently the dollar amounts of cash surpluses (tax revenue over total cost) are dropping even faster than projected and on the OAS side may be down by half or more. Which may move the date that those cash surpluses vanish up from 2017, but if OMB and CBO are right not by much, and in any case the total increase in assets that year will be more than $200 billion and probably pretty close to the $243 billion projected by CBO last August. Now under current projection none of this increase will be the result of excess contributions from FICA which will fall $24 billion below the amount needed to cover projected cost. All of it will be due to interest earned on the portfolio, which has led some to argue that the response of Congress will just be to punt the ball down the field by cutting benefits rather than having to cut spending or increase borrowing or taxes on the General Fund side. But this ignores the fact that there will be specific bond issues maturing in 2017 including $77.3 billion yielding 5.250%. Now some in Congress will take a lead from Hassett and AEI and claim that we shouldn’t count interest on the Trust Fund as real, but they are going to have a lot harder time saying the same thing about the maturing principal. Now not all of that $77.3 billion was the result of excess FICA taxes, up to half of it comprised of interest earned on the portfolio at the time of the issue of the bond. Meaning they are put in the position of issuing new Special Treasuries to roll over that $77.3 billion, insisting they are still fully backed by Full Faith and Credit while denying they have the cash to even pay down some of the maturing principal.
Sitting here in 2009 and thinking about the Trust Fund as a unified whole where 2017 marks the time you start drawing on interest and having exhausted that start drawing on principal in 2023 is to ignore the operational reality of the Trust Fund. In actual practice a portion of that principal comes to maturity every year and has to be either redeemed or rolled over. If we examine the right hand (2007) column we can see that every dash represents a bond issue that has been retired and replaced by a new Special Treasury, in doing so the Treasury has shown it still stands behind Full Faith and Credit. But the instant it refuses to recognize a valid claim for repayment of the portion of the bond that represents actual excess contributions they have effectively abrogated the entire Trust Fund. You can play fancy games with accrued interest and insist that it shouldn’t count, but you have a lot harder time doing that with maturing principal.
Only if you regard the Trust Fund as an undifferentiated lump where no principal payment is due until 2023 is it possible to sell the idea that the date that Income excluding Interest falling behind Total Cost is significant. But in practice we redeem some of the principal every year even now. Back in the late eighties there was hardly a suggestion that Shortfall was a significant date, main because Trust Fund depletion was reasonably close and coming closer. And Trust Fund depletion had real fall-out, it presented the direct tradeoff of either just taking a benefit cut or increasing taxes. And when and if we get fifteen years from projected depletion we will be able to make a good faith argument that something needs to be done, the looming benefit cut coming not from a failure to stand behind Full Faith and Credit, indeed the whole concept of Trust Fund Depletion assumes that Full Faith operates all the way down just as it did in the late seventies.
Table VI.A4.—Historical Operations of the Combined OASI and DI Trust Funds, Calendar Years 1957-2007 [Amounts in billions]. In 1971 Social Security went to Shortfall as tax revenues fell below cost. Our response? We honored the interest and maintained full benefits. In 1975 Social Security hit ‘peak’ as all accrued interest was needed to pay benefits and we were no longer able to just roll over the principal. Our response? We honored the remaining interest and the principal and maintained the necessary transfers from the General Fund needed to pay benefit. Plus we put in place the Carter 1977 Amendment which pushed the problem out by 50 years. Unfortunately it didn’t totally work and by 1982 Trust Fund depletion was imminent. Which led to the 1983 Reform which required a combination of tax increases and benefit cuts. But at no point did anyone suggest that somehow the interest and principal in the Trust Funds were not a real and binding obligation.
2017 doesn’t matter because the people who insist it does are making a dishonest argument that implicitly claims that Full Faith and Credit is something selective and elective, and that Congress can effectively wave it away by legislative action, and claim that a Supreme Court ruling that people don’t have ownership rights to the Trust Funds supports that. Well I don’t think that is what the Court really had in mind. That Congress is free to change the benefit schedule in response to Trust Fund balances dropping below acceptable ratios (i.e. current law sets a trigger at a TF ratio of 100) does not mean they are free to simply change it out of an attempt to avoid payment when that TF ratio is projected to be at 385 by effectively denying that legally owed interest falls under Full Faith and Credit. That claim may survive encounter with the Courts, but I don’t think it survives its encounter with the ballot box. I just don’t see “Yes you have $4.4 trillion in the bank, and yes it is earning interest, but no you are not allowed to actually draw on it” as a winning argument in the 2018 mid-terms.
Basically AEI is hoping to pull a fast one before anyone notices what their plan actually does.