Relevant and even prescient commentary on news, politics and the economy.

BW on Soc Sec IV: A History Lesson

A blast from the past. The first Report I downloaded was the 1997. It didn’t report on Productivity, instead the lead number was Real GDP. So we will run with that. What were the numbers under Intermediate Cost?:
1997: 2.5%
1998-2006 2.0%
2010 1.8%
2020 1.3%
Hecuva job! A median prediction of a permanent 48% slowdown in growth rate by 2020. Not exactly a confidence vote in America. What happens if we turn to Low Cost?
1997: 3.2%
1998-2001: 2.7%
2002-2003: 2.6%
2004-2006: 2.5%
2010: 2.3%
2020: 1.8%
Ultimate: 2.2%
These are optimistic numbers?
Because this is coming off a data series as follows:
1991: -1.0%
1992: 2.7%
1993: 2.3%
1994: 3.5%
1995: 2.0%
1996: 2.5%
So we have a supposedly median projection that suggested a rapid and permanent dropoff to 2.0% in the very next year (1998), or a 20% decline and a supposedly optimistic projection suggesting that a 10% dropoff from the proven 1994 result was the best possible result for the current year of 1997, and a 19% drop for 1998 and following was too. You didn’t have to be Rebecca of Sunnybrook Farm to think that maybe the outlook wasn’t quite so bleak as these guys suggested. 1996’s 2.5% was considered a pretty good number but it was hard to think that it had established some sort of ceiling. If you just extended that 2.5% through the future most of the gap would be closed. Because then as in all years to 2007 Low Cost produced a fully funded Social Security system with flat Trust Fund ratio (see comment for 2008). (Figure I.H3, page 26 of above link).

Okay what were the actual numbers and the new projections from the 2008 Report: Table V.B2?
1997: 4.5%
1998: 4.2%
1999: 4.4%
2000: 3.7%
At which point I was pretty God damn confident we would pull this off. But along came Bush.
2001: .8%
2002: 1.6%
2003: 2.5%
2004: 3.6%
2005: 3.1%
2006: 2.9%
2007: 2.2%
But still on balance smashing 1997 Low Cost. And what do we need from Low Cost going forward?
2008: 3.4%
2009: 3.5%
2010: 3.5%
2011: 3.3%
2012: 3.1%
2013: 2.9%
2014: 2.7%
Ultimate: 2.9%
Few people would probably predict 3.4% for this year. On the other hand who really believes that 2004 numbers will never, ever be repeated? In your optimistic model? Discuss. And don’t forget to show your work.

BW on Soc Sec III: The Numbers of Intermediate Cost

For today let’s have less meta and more data.

The key economic and demographic tables are V.A1 ‘Principal Demographic Assumptions’, V.B1 ‘Principal Economic Assumptions’ and V.B2 ‘Additional Economic Factors’ all of which are available in the 2008 Social Security Report’s List of Tables as are the tables showing numeric outcomes in both inflation adjusted Constant dollars in VI.F7 and unadjusted Current dollars in VI.F8. In each case you get projected numbers for each of the next ten years and then for each five year interval to 2085 for each of the three models Low Cost, Intermediate Cost, and High Cost. In each case the models settle out to a steady state ‘Ultimate’ number in about 2014 and stay more or less steady through the rest of the 75 year window.

So we are faced with two different questions. First of all nobody has a crystal ball, economists are disputing what is happening this quarter, in fact there is not consensus on what happened last quarter, the idea that we can pin down Real GDP and Real Wages for any particular future year is in some respects nonsense. But what we can do is come to some sort of conclusion about where the economy is likely to go over the next ten years and what the long term potential will be. So question one is ‘How likely are we to end up in 2012 with the numbers we see?’ and two ‘What is America’s fundamental bottom line going forward?’ Or more polemically ‘Is America’s future bleaker than its past, or are the best days yet to come?’

So I would urge everyone interested in this topic to add the List of Tables to your Favorites or Bookmarks and visit the numbers early and often. As a teaser lets take some selected numbers for 2008, 2012, and 2016 from the standard Intermediate Cost alternative.

Productivity 1.9%, 1.8%, 1.7%
Real Wage Differential 1.3%, 1.2%, 1.0%
Real GDP 2.3%, 2.5%, 2.3%. Note here that they are figuring in some period of slow growth near term. If we took the series 2009, 2013, 2017 we see 2.8%, 2.5%, 2.3%
Unemployment 4.8%, 5.3%, 5.4%

This pattern has been seen in Intermediate cost for a decade, the official projection is that across the board over what can only be seen as the short term that the median outcome is a pretty sharp decline, and more importantly a permanent one, the future will never be better than 2006. No one I think would claim these outcomes are not possible, but are they really the median? As a bonus question, if they do represent a median under status quo economic policy, what could we do to boost them towards Low Cost? Over to you.

BW on Soc Sec II: The Shape of Low Cost

This figure shows in graphic form the outcomes of Intermediate Cost (II) vs High Cost (III) vs Low Cost (I)

Outcome II shows the standard narrative. A Trust Fund Ratio rising to a peak in 2017 then a more or less rapid falloff to zero as the first the interest is tapped and then the principal is redeemed with the result of total Trust Fund Depletion in 2041.
Outcome III or High Cost shows the same process only accelerated.
But Outcome I or Low Cost shows a much different picture. The Trust Fund ratio peaks at about 450 about 2020 then dips to 390 by around 2040 as a portion of the interest is tapped but after a period of plateau sharply increases through the remainder of the 75 year actuarial window. This is quite literally the picture of a potentially overfunded Social Security system. Is a Low Cost outcome guaranteed? Well no. Is it possible? Certainly the numbers are not at all outlandish and perfectly in line with economic performance over the last fifteen years.
Can we at least talk about the implications of this?

Well I just copied this over from the Bruce Web in response to a comment from Jim A on the previous post. Some explanation of terminology is in order here. First the three curves represent the three different alternatives supplied by the Trustees. The only one generally reported in outcome II or Intermediate Cost, with the other two (High Cost and Low Cost) supposedly representing the outward bounds. In point of fact for the years 1997 to 2004 the deviation from Intermediate Cost was always in the direction of Low Cost and often exceeded its numbers.

Social Security solvency is measured in two ways. One is by payroll gap which is to say the amount that the payroll tax would have to be raised immediately or contrawise at Trust Fund Depletion to fully fund the current schedule of benefits. Per the 2008 Report those figures are 1.7% and 3.54% respectively, numbers that have trended down significantly since the 1997 Report. The second measure is Trust Fund Ratio which is to say Trust Fund assets measured as a function of time with 1 year = 100. Under current law (or perhaps simply practice) Actuarial Balance is defined as having a Trust Fund Ratio of 100 for each of the next 10 years (Short Term Actuarial Balance) or the next 75 years (Long Term Actuarial Balance). In 2003 a new measure was added which would have us evaluate actuarial balance over the ‘Infinite Future Horizon’. This was in my opinion a simple gimmick to allow opponents of Social Security to use really scary numbers rather than focus on the traditional and reasonable 75 year planning window, a window that pretty much covers the interests of most peoples children and grandchildren.

Now in examining the graph we can see Low Cost (outcome I) we see a sharp acceleration of the TF Ratio after 2060 with it exiting the 75 year window at 650. This is dangerously high, that is if we consider the Trust Fund and more significantly the interest it earns as being a real asset. And we should because otherwise we would be arguing for what pgl calls a backdoor employment tax that called on one sub-set of Americans to subsidize General Fund spending. And theoretically there is an argument to be had there but it sure puts ‘paid’ to the ‘poor people don’t pay taxes’ narrative the Right is so fond of. But in point of fact flattening the tail down to where the Trust Fund is at its nominal target of 100 is difficult, the only way of doing that is to starve the beast by cutting payroll tax rates and using General Fund dollars now rather than later to pay down the Trust Fund. Because the longer you wait the higher the proportion of Cost has to be born by the General Fund as opposed to FICA to keep interest on interest from bloating the TF. This is incidentally why cutting benefits or increasing revenues by such measures as raising the cap are counterproductive, they simply increase the amount of accrued interest without actually benefitting anyone today.

Of course this whole discussion is moot if we actually get outcome II or Intermediate Cost. Our choice of policy going forward is entirely dependent on whether future outcomes are closer to II or I which in turn requires an examination of the assumptions underlying Intermediate Cost and Low Cost in light of the fact that most of the bias over the last 12 years has been strongly to the upside. Yet all of the policy discussion to date simply clings to Intermediate Cost assumptions and considers the problem to be one of deficit in 2041 when it is not clear that is at all the most likely outcome.

Next post or so we’ll stick in some actual projected growth numbers. In the meanwhile you can ponder the implications of a $117 trillion Trust Fund in 2085 (Low Cost projection) Table VI.F8.—Operations of the Combined OASI and DI Trust Funds, in Current Dollars, Calendar Years 2008-85 [In billions]

Bruce Webb’s Take on Social Security: I

Well Reader Dan invited me to submit some posts on Social Security and so here I am. Depending on reader interest this is the first in a series trying to get us to a deeper understanding of what Social Security actually is and where it is likely heading. And that may be a much different place than most people currently think.

Social Security finance is a really odd beast, it has a strong up is down quality that tends to baffle people. For example I will be making the case in a later post that cutting Social Security benefits actually harms the long range financial stability of the system. Why this seeming paradox holds requires us to examine some mechanical peculiarities of a PayGo system given the details of the current state of its financing. But before diving into the deep end I want to establish a conceptual framework, that is to show you Social Security through my eyes.

First it all starts, and mostly stops, with the Trustees’ Reports, more formally known as the ‘ANNUAL REPORT OF THE BOARD OF TRUSTEES OF THE FEDERAL OLD-AGE AND SURVIVORS INSURANCE AND FEDERAL DISABILITY INSURANCE TRUST FUNDS’. Almost every fact, factoid and date cited in this debate comes directly or indirectly from this Report. (Certainly CBO has a slightly different interpretation but they don’t really challenge the fundamental assumptions of the Trustees). For the most part in this game all of the players are playing out of the same playbook and more importantly out of a particular subset of that book. More on that point later. The first lesson people need to learn is that ‘It is already built into the model’. This is particularly important in relation to the demographics, because the first response from those hearing that Social Security is not particularly in crisis is: ‘What about the Boomer Retirement?’ ‘More retirees, less workers, what don’t you get about that?’ or ‘People are living longer!” Well I know all those things mainly because the demographers working in the Office of the Chief Actuary know them full well and have already accounted for them in their models. Which are published in the Report. Which I read.

Second point. Social Security really and truly is Insurance. Endless efforts have been made to blur this fact, and I expect/hope to see some in comments, but they all fail in the face of the actual operation of the system. Social Security collects premiums from all wage workers up to a determined wage level and disperses benefits to qualified beneficiaries. Which is pretty much the same as what New York Life does, cash flows in one door and out the other. That varying things happen with excess cash in between are not really the concern of the buyer and future beneficiary of the policy, as long as they get the promised benefits when the time comes then all’s fair. That is there is no functional distinction between drawing a benefit dollar from the existing investment pool as opposed to simply paying it out of current premiums, PayGo does not equate to Ponzi. Banks and Insurance Companies don’t have current cash reserves equivalent to all of their obligations, instead that money is locked up in what may be long term investments, instead they rely on current cash flow to carry the load. Of course banks are required to hold some cash reserves with the Federal Reserve and Insurance companies themselves carry reinsurance, you can’t get too close to the edge. And this is where Social Security Insurance departs from New York Life, instead of reinsurance it has Trust Funds.

Which gets us to the third point. Social Security is not the Trust Funds anymore than New York Life is its reinsurance policy. The Trust Funds are simply the numeric measure of how much Social Security has in long term reserves. This amount can over any given year or series of years trend up or down without directly effecting benefit payments. In an up year Income Excluding Interest exceeds Cost and so surpluses flow to the Trust Fund, in a down year Income Excluding Interest lags Cost and the Trust Fund has to be drawn down, but typically not by much, current income is always carrying much of the load and would even if the Trust Fund went to zero.

Last points for today. Trust Fund ‘crisis’ is normally set at two possible points. One is crisis at Shortfall, that point where Current Income is projected to fall behind Cost for an extended period. This would require a sustained drawdown on the Trust Fund as more and more interest is diverted to pay current benefits. Whether this really equates to ‘crisis’ requires an analysis of the actual dollars in context. That rarely happens, instead people jump right to ‘Raise taxes to unendurable levels! or Slash benefits to the bone! We got to act now!!’ Well this hysteria whether real or fake (and much of it is fake) is simply that. Absent an examination of the numbers it is simply as Shakespeare put it ‘A tale told by an idiot, full of sound and fury, signifying nothing’, In this particular case if you can’t say it in numbers you really shouldn’t be saying it at all.

The second ‘crisis’ point, and the more traditional one, is crisis at Depletion, that point where the Trust Fund has been drawn down to nothing. Once again the temptation is to jump to “Oh My God we’ll have to raise taxes’ or worse ‘Its bankrupt!! Dead broke!!! Checks will stop!!’ Well that too is mostly nonsense, by law we need do nothing at Trust Fund Depletion, instead benefits would simply be reduced to whatever level was supported by current income. Whether that adds up to ‘crisis’ is once again a matter of comparing the numbers to the actual schedule of benefits while comparing it to benefit levels today. In the event the relevent equation from the 2008 Report is 78% of 160% = 120%. More on that later.

So lessons attempted here?
1. Numbers matter and they are to be found in the Reports. Including all the relevant demographic ones.
2. Social Security is in fact a Pay as you Go Insurance plan with a certain level of reserves. It is not a Ponzi scheme.
3. That reserve is called the Trust Fund, but is not itself the whole of Social Security any more than your bank account is the whole of your finances.
4. No one likes to draw down their reserves as we would at Shortfall, but you have to put that in context of your entire financial position. Which means looking at the numbers, see point 1
5. You plan for your retirement and ideally meet 100% or more of your goal. Hitting 78% may or may not be some sort of crisis, it might just mean buying a smaller boat. We need to examine Depletion in the context of both absolute level of benefit cuts and probability that the current projection will come to be.

I’ll be glad to answer questions and meet challenges in comments. In future posts I will be explaining some of the odd terminology used: Low Cost alternative, Trust Fund Ratio, Covered Worker Ratio etc. For those who just can’t stand to wait you might want to browse the Nov. 2004 archives of the Bruce Web (