by Bruce Webb
Each year the Trustees of Social Security provide a graphic representation of the results of the three alternative projections: Low Cost, Intermediate Cost, and High Cost represented by results I, II, III respectively. And the overall shape of the graph didn’t vary much from 1997 to 2007 (other selected years can be seen at Alternative Shapes). While the date of Trust Fund Depletion (on the figure a TF ratio of 0) moved out some for Intermediate Cost (II) the result for Low Cost never really varied, instead it yielded a more or less flat Trust Fund ratio. For most purposes a flat Trust Fund ratio means that Social Security is ‘fixed’ in that income including interest on the Trust Funds continues to meet projected cost. But you can get too much of a good thing, for example the interest on a Trust Fund with a ratio of 2000 at a nominal 5% rate would fully cover cost, which would operationally transform Social Security into a General Fund welfare plan. But this seemed to be just a pie in the sky idea, instead even the most optimistic projection showed Trust Funds ratios under Low Cost somewhere between 300 and 500 meaning that Interest would only be called upon to meet 15-25% of ongoing cost after 2029 or so. Nice steady state system.
The whole question seemed rather academic in that Low Cost was always seen as an optimistic economic model. If the best we could possibly accept was a steady state system why really even think about Low Cost, if it happened it happened.
But the narrative changed drastically with the 2008 Report, the story line got a lot more twisted, it is just that nobody noticed. The 2009 Report is due out in a couple of weeks and who knows exactly what story it tells, but for those interested you can take a look at the Future of the Shape per the 2008 Report below the fold.
The first thing to note is that this version of Low Cost is not a good outcome, there is no good reason to have the Trust Fund long term at anything but flat. In fact I argued that we should actually target a TF ratio of 100 in Low Cost and the 100/100 Target. But rather than rehash that I want to examine another aspect.
In the Summary of the Annual Report the Trustees supply three numbers for Intermediate Cost: cost of an immediate fix, cost of a fix in 2041, cost of a fix at the end of the 75 year window, and then two more: amount of benefit cut needed in 2041 and cost at the end of the window. In years past those first three numbers are implicitly the equivalent of Low Cost in that they projected the same outcome of fully funded system with a presumedly flat TF ratio. In 2008 it is different. Now instead of needing numbers right at the probability limit (which is what Low Cost is presented to be) instead economic outcomes in between Intermediate Cost and Low Cost could end up producing that steady state result.
Let’s take some pairs of numbers for ultimate years (ten years out) from Intermediate Cost and Low Cost:
Unemployment: 5.5%, 4.5%
Real Wage: 1.1%, 1.5%
Productivity: 1.7%, 2.0%
Real GDP: 2.2%, 2.9%
Which raises a series of questions. One, even if we conclude that Low Cost is on the whole too optimistic, what is the economic argument against outcomes that split the difference? For example 2006 was not considered to be a really great economic year but the respective numbers turned out to be: 4.6%, 1.8%, 1.0%, 2.9%, that is while productivity lagged the other numbers came in pretty much at Low Cost levels. Granted 2008 numbers are going to look terrible and 2009’s worse but is there someone out there with a serious argument about why these number series can’t revert to 2006 levels on average?
Two, what if we embarked on an economic policy that targeted the 100/100 outcome? I posed a form of this question back in 2006 in Goldilocks and the Three Social Security Bears but would reformulate it into something like the Four Musketeers. First let the Trustees continue to present the three current alternatives based on what they claim is best possible analysis, i.e. High Cost, Intermediate Cost, and Low Cost. But supplement that with a fourth economic model that would explicitly target a 100/100 solution (100% of scheduled benefit with a 100 TF ratio). And then let us have a debate whether we can reach that through policy means.
Because frankly laissez-faire is pretty much yesterday, the people who claimed that everything could be explained by EMH and that we would just end up where we would have been proven dead wrong. Specific choices by policy makers and investors combined with some Animal Spirits put us in the place we are and only die-hards are clinging to the idea that any of this was just the result of the Invisible Hand operating in a market of fully-informed participants. (Ask Bernie M’s clients how informed they were in retrospect). The dominant view in current Administration thinking is that we should strive for specific economic outcomes through deliberate policy actions. Why not include Social Security solvency via economic growth as one of those outcomes?
Because if we just slice the differences we are looking at something like 5.0% unemployment, 1.3% real wage, 1.9% productivity and 2.6% real GDP in the out years to mostly or entirely solve the current projected gap. Those are not mad, crazy numbers and we don’t just have to accept what the economy throws at us. I don’t think we are likely to even get back at any point to 1999 numbers that had us at 4.2% unemployment, 4.7% real wage, 2.4% productivity, and 4.4% real GDP and certainly not as an average. But it seems kind of un-American to insist that the future can never even approach the past. These are just not unrealistic policy targets, why NOT shoot for them.