This figure shows graphically the outcomes for the OAS (black line) and DI (gray line) under the three alternative scenarios Low Cost (I), Intermediate Cost (II), and High Cost (III) in the 2008 Report. Almost universally reporting on Social Security revolves around Intermediate Cost and if we examine its lines we see the standard reported outcomes, the DI Trust Fund going to depletion in 2023 and OAS in 2041. But that was then, the 2009 Report shortened up the intervals to 2017 and 2037 for Intermediate Cost, and the 2010 Report is certainly going to see DI in a state of near term collapse. Okay Intermediate Cost shows us a program in need of some fixes, immediately for DI, and in the medium term for OAS. Familiar story.
And High Cost tells the same story, only more so.
But Low Cost is different. Under Low Cost Social Security OASI is projected to be fully funded through the 75 year window and even have balances growing faster than costs after 2050 and Trust Fund Ratios never falling below 350. Would this be a good outcome. Well the fully funded part is nice, and a more or less flat tail after 2050 is nice, but a constant TF ratio above 350? Well no that is not a good thing, and the only thing worse would have Low Cost OAS look like Low Cost DI and rocketing through a TF ratio of 1000 on way to 2000. There is a point where the Trust Fund is literally too big for Social Security’s good. To understand why follow me below the fold.
What would be the ideal shape and level for this graph. Well in shape it would look like Low Cost OAS, a near term bulge to build the TF to the point that interest on it meets the challenge pending from 2026 to 2050, and then flattening out. But at a lower level. This spreadsheet represents the NW Plan http://spreadsheets.google.com/pub?key=r49_nOHQG4QdHuwcbMGmP0Q The column fourth from the Right is labeled Trust Fund Ratio and if graphed would show a perfect shape, a Trust Fund topping out with a ratio of 375 and then gradually adjusting down to ultimate levels between 123 and 130. But not that the total Trust Fund Balance only goes down during a ten year period and then by not enough to matter, under the NW Plan the principal in the Trust Fund never gets paid back. And that is a good thing.
So when supporters of Social Security claim that the tax increases starting in 1983 were designed to pre-pay Boomer Retirement they are right in spirit but wrong as to detail, in an ideally balance system those dollars are never called on directly. And by that same token it is wrong to say the Trust Fund was designed to run out, because it wasn’t, instead the expectation was that changes to either revenue or benefits would be made in a way that keeps that tail from dropping below 100. Meaning that the Trust Fund simply serves as a buffer for bad times, as it was between 1972 and real crisis in 1982 and as it is doing today.
But even given that a TF ratio of 100 is a designated minimum why is a flat 130 better than say the flat 400 seen under Low Cost OASI? Well because of the effort needed to keep a TF ratio flat. A flat TF ratio assumes that Benefits are requiring enough resources above FICA and Tax on Benefits to take up enough of Interest so that the TF grows only enough to mirror overall program growth. So if we assume ultimate interest earnings at 4% keeping a TF level steady at 400 reguires that some 12-15% of total program costs be paid by General Fund transfers and this over and above the 12.4% of payroll directly devoted to it. Which sets up three potentially contending groups. First you have beneficiaries who have an equitable claim that the Trust Fund is collectively an earned obligation to retired participants, even though most of the principal was actually contributed by previous generations. Second you have non-participants, particularly those with income from capital, who point out legitimately enough that they are picking up to 15% of program costs and getting nothing for it. In between you have current workers who are in effect paying twice, once in FICA and a second time in income taxes.
In contrast maintaining a TF ratio of 130 requires much less societal strain. Rather than an 85/15 split between payroll tax & tax on benefits/debt service the split is 95/5 where that 5% is simply the cost of recession insurance.
Now some Social Democrats would claim that not only is 85/15 okay but they would take that to 50/50 or even 0/100. Well right on brothers and sisters! But quite apart from genuine philosophical beliefs in the virtues of thrift and self-reliance there are many practical and political strengths to having worker retirement financed by workers supplemented only by a small amount of interest on a reserve fund that prevents every downturn in employment from turning into outright class war. Instead Social Democrats could turn their attention to other societal goals like health care, paid leave, real wage increases, public education all of which directly or indirectly contribute to the general welfare, including that of wage workers.
Personally I would be okay with a ratio as high as 85/15 if that was driven simply by interest on an existing Trust Fund. Because once established a TF ratio of 400 is hard to drive down, the only way to do that is to artificially drive the second term up enough to actually start redeeming principal which in turn forces the whole system more in the direction of subsidized welfare. But levels much above 400 are actively dangerous.
If we return to the graph we see that Low Cost DI shoots up to a TF ratio of 2000 within the 75 year planning period. Under that scenario and at assumed 5% the Trust Fund would be throwing off interest equal to program costs. This would have the effect of changing a worker financed ‘entitlement’ program into a pure welfare program that increasingly would be justified simply on the basis of past contributions by past generations, at some point the concept of this being an earned benefit being totally lost. Now in the particular case of Disability this would not bother me much, I have no problem sticking a big part of the societal bill onto capitalists who may have created the conditions that led to that disability to start with, and besides I think the overall duty of society to aid the helpless is pretty clear to start with. But on purely practical and political grounds I would not like to see OAS going anywhere near the 1000 level, they system loses its insulation.
Which gets me back to my title. Social Security is under current assault under the belief that costs to finance payment of the interest owed are already too high. When pressed those opponents tend to reluctantly admit that they owe the money, it is just that it would societally harmful to do so. While I could and do argue that this is just nonsensical special pleading that the transfers at point of the greatest projected strain are no more than those of wars of choice, in the end the cost of paying down a $5 trillion debt is big while the projected cost a paying down a $2.5 trillion dollar debt is with subsequent interest effects less than half as high. Freezing the Trust Fund at its current levels would eliminate, at least it seems to me, this argument that not repaying the TF is not driven from ill-will but instead from practicality. Not that I believed in the argument to start with, just that I recognize its power. Hence my suggestion of a short term cut in FICA taxes sufficient to freeze the fund combined with a schedule of future increases to stabilize it.
Interestingly it looks like the economy is kind of doing my job for me. Under my formulation a vanishing Social Security surplus is not necessarily a bad thing, though certain that which is driving it, high unemployment, certainly is. Opponents are using this period of short term crisis to make long term changes to Social Security’s benefit formula, changes which ultimately worsen the situation by swelling the Trust Fund over the medium term. If we can just stave that off Social Security might emerge from this job recession actually better situated to achieve long-term health, freed from the burden of a burgeoning Trust Fund going forward.