Northwest Plan for Social Security: Policy Proposal? or Benchmark?

by Bruce Webb

Yes. Both. But first a recap. The semi-final version of the basic NW Plan was introduced in this post: NW Plan for a Real Social Security Fix. The spreadsheets linked to that post show that under current projections you can fix Social Security by implementing a 0.20% FICA increase in 2010, another of 0.10% in 2011, followed by a series of 0.20% increases in 2026-2036. The NW Plan doesn’t set those future increases in stone, instead they are tied to a specific trigger, the date that either fund that comprise Social Security falls out of the Trustees primary official test, that of Short Term Actuarial Balance. As it happens DI already failed that test last year which explains the immediate increases for 2010 and 2011. But OAS retains a significant degree of uncertainty. For example even though CBO starts from the same set of demographic assumptions as SSA, their different treatments of the economic numbers result in wide variations between their median outcomes. The CBO projections from Aug 2008 can be seen at the following post Probability and Social Security. The most current SSA projections can be seen in the following figure:

(Update: the above figure is from the 2009 Report, the original post included the version from the 2008 (below). Sorry about that.)

The differences between CBO and SSA are explained in large part by different assumptions and methodology but are also effected by being different snapshots in time, SSA building in data for Q3 and Q4 of calender 2008 and Q1 2009 not available to CBO last summer.

So the Northwest Plan is being worked up as a Policy Plan to be presented to policy analysts and policy makers. But like all batttle plans it is unlikely to escape all effects of contact with the enemy, which in this case is the uncertainty of the projections. What the Northwest Plan does is to close both the Short Term and Long Term actuarial gap under the Trustees most current set of assumptions and definitions and provides a mechanism for adjusting the tax schedule each year in anticipation of events as yet fifteen to twenty years down the road. Rather than revisiting the issue every decade or two it provides for a permanent yet flexible fix that in turn allows us to focus if we wish on other measures that might move that future trigger point farther out in time.

For example if we examine the above graph we see that DI (light gray) has vastly more variation than OAS. Under the NW Plan the light gray line that current intersects with zero in 2025 instead would exit the 75 year window right at 180. But if we examine outcome I we see that line exiting the window at above 2000. Oddly it is politically hazardess for Social Security if its Trust Fund balances remain permanently north of a ratio of 300 or so. If over the next few years it looks like DI is on a course approaching outcome I we would want to divert a portion of the 2010 and 2011 increases, which are under the NW Plan devoted to DI, over to OAS. This in turn would serve to move the trigger point for OAS, perhaps imperceptably, perhaps noticeably. But we don’t have to rely on chance events, lowering future costs for DI is a matter of lowering the incidence of disability or of tightening eligibility. I’ll let Nancy Ortiz speak to the latter, from what I hear if anything the rules are too tight and the administration too disfunctional already. On the other hand it would be useful to focus on the differences between Intermediate Cost (II) assumptions on incidence and Low Cost (I) and then work on ways to achieve improved outcomes. For example it is possible that better enforcement of existing workplace safety law might directly serve to drive down disability generally and hence need for DI and in turn reducing DIs impact on OAS.

In this way the OAS trigger set by the NW Plan for 2026 can be an explicit target for economic policy. Once you change your model from the deterministic one typical of today, “Social Security will go to depletion in year X”, instead we should be thinking “What can we do in 2010 to push that trigger point back, or use it to increase benefits”. In that way the NW Plan frees us to approach Social Security pro-actively rather than passively.

The NW Plan provides a Social Security fix that is at the same time permanent, flexible in the face of events, and modifiable by direct action by policy. It is certainly worth a try.

But the title of the post also says it is a benchmark. How so? Answer below the fold.

On the merits ‘Nothing’ has been shown to be a perfectly sound plan. That is instead of rasing taxes by $1/week as the NW Plan does we could reallocate current FICA between DI and OAS in a way that made their future shortfall and depletion dates congruent and then start trying to move the joint trigger point outwards via economic policy. This would mirror past practice. But ‘Nothing’ is a tough sell, it implies that the people pushing it are in some sort of denial rather than what is the case, that examination of the data shows ‘Nothing’ to be on net a historical proven plan since 1997. But rather than continually rebattling that war we propose the implementation of the NW Plan to at least serve as a benchmark.

Because the NW Plan is scorable, it has numbers that can be checked and against which conclusions can be drawn such as “Yes under Intermediate Cost assumptions the NW Plan would put Social Security into Short and Long Term Actuarial balance at a cost of X dollars in 2010, Y dollars in 2011 and Zetc dollars in years 2026 and following.” Which then allows us to challenge any other plan to match up dollar for dollar. What does their plan cost in the form of foregone benefits or increased taxes in any future year? If it doesn’t actually provide better retirement benefits at a lower cost what are its offsetting advantages that would induce workers to take the deal? Put your numbers on the table and lets compare.

It is one of the striking features of the overall Social Security debate that future impacts for any given year are rarely if ever spelled out. Yes under Intermediate Cost assumptions cash income from taxation falls behind cost in 2017. What does that mean for 2017? or 2023? Yes if we do nothing on Social Security the system will only have projected resources to pay out 75% of benefits in 2037? What does that mean in real terms? Well you never know because the argument always moves form ‘Crisis’ to ‘Benefit cuts’ without anyone pausing to quantify the effects or reflect on alternatives.

Well the Northwest Plan is such an alternative. It provides 100% of the scheduled benefit while maintaining the Trust Funds in actuarial balance through the current 75 year window. Plus it offers a mechanism that allows Social Security to maintain that state over the Infinite Future if you like. Want to wipe $15.3 trillion dollars in unfunded liability off the U.S.’s books tomorrow? Enact the NW Plan.

Of course it will cost you a $1.50 a week by year two. Maybe that is a deal breaker, but at least people need to explain why that shoud be. Over to you all.