Recalculating ‘Nothing’: Social Security in a Time of Recession
by Bruce Webb (first three sentences edited since original post a few minutes ago)
The 2005 Report was a mild disappointment. The 2006 Report somewhat worse. 2007 showed some mild improvement but didn’t even get us back to 2005 numbers. Now the 2008 on the surface looked a lot brighter but most of that turned out to be the result of a one time change in assumptions about the relations between immigrant earnings and actual payouts (contrary to myth immigrants get a crappy deal from SS). And now the preliminary numbers for 2008 that will go into the 2009 Report are pretty sad and you have to be really, really on the Pollyanna side of life to believe that 2009 employment and wage numbers are going to be so robust as to make up all the difference by the 2010 Report. Really we don’t seem to have made much progress since the 2005 Report’s Conclusion which projected Trust Fund depletion in 2041, a payout of 76% after that, and an actuarial deficit of 1.92%.
All of which might induce some to think that we need to rethink ‘Nothing’ as a plan for Social Security. Well not so fast, instead we need to first need replicate the calculation I first suggested back in 2006 in my post The Cost of Inactivity: ‘Nothing’ as a Plan for Social Security.
Plenty of numbers below the fold. Be sure to bring a calculator
I am not here to make the case against PRAs because really I don’t need to. There is a case to be made that if you had to transition to a system of Private Accounts the time to do it is when the markets are at a low. The problem is convincing anyone that we in fact are at a low or that the markets will ever recover in a way that offsets losses to date in any reasonable time frame. So I’ll leave that one for Andrew Biggs.
In 2005 we were 36 years from Trust Fund Depletion then set for 2041. Today we are 32 years from TF Depletion still set for 2041. If we don’t see progress then at some point we will need to make some decision. On the other hand we don’t need to rashly throw today’s dollars at a problem that may not happen at all, or happen in a smaller than currently projected form. On what should we base this decision? Well on something I like to call ‘arithmetic’.
The discussion starts with some simplifying assumptions. One is that we can act based only on our knowledge of the past year and our prediction for the next one or two years.( Which is fair enough, anyone who can predict with precision what numbers will look like in 2011 is practicing witchcraft). Two that we are going to treat wages in constant dollar terms and moreover three, assume flat real wages for our four test cases. And four that increases in payroll tax ultimately bear on wages even though in cash flow terms half comes from the employer. Anyone who wants to present some more sophisticated calcs can do so in comments.
Okay take four individuals aged 25, 35, 45, and 55. The latter will reach full retirement age at 66, the others at 67, meaning that remaining work life is 42, 32, 22, and 11 years respectively. We will assume that each makes $50,000 a year and in constant dollar terms will continue at that wage rate. The payroll gap, meaning the amount FICA would have had to be increased immediately to restore 75 year actuarial balance, in the 2008 Report was 1.7%. That is other things being equal an increase in that amount starting a year ago would have ‘fixed’ Social Security. Instead we did ‘Nothing’. Now avoiding a tax increase should have about the same effect as having an equivalent tax cut. So doing nothing left $850 dollars in all four pockets. Under our four assumptions how much would the payroll gap have to increase by the next Report to make ‘Nothing’ a bad bet? Well it varies by age of the subject.
Our 55 year old has 11 years until retirement. As long as the increased annual tax of a higher payroll gap is less than 850/11 he is better off with ‘Nothing’ for 2008. Or $77. Which translates to .15% of payroll. So if the new number comes in anywhere higher than 1.85% he is on paper a loser. Can the number move .15%? Well yes but typically the movements come in at about a fifth of that.
Our 45 year old has 22 years until retirement. 850/22=$38.50. Or .075% of payroll meaning that his losing number is 1.78%
Our 35 year old has 32 years until retirement. 850/32=$26.50. Or .053% of payroll meaning that his losing number is 1.75%
Our 25 year old has 42 years until retirement. 850/42=$20.23. Or .047% of payroll meaning that his losing number is 1.74%
So would some of our players have turned out to be nominal losers by not paying $850 last year. Well maybe, even probably. But that serves to discount the current utility of that $850 over the last year or conversely the amount he could earn in interest over the 22 or 32 or 42 years of work life.
Now that is the result if you take the narrowest view. What if you get a little more historical? The gap in 2005 was projected to be 1.92% which assuming a constant dollar $50,000 wage means $960 dollars not taxed or a total of $4800 left in our players pockets since the heights of the There is No Crisis battle of 2005. How much would the gap have to go up in a single year to have made ‘Nothing’ a bad cumulative bet? 9.6% in a single year loses everybody money. But anything less than an increase of .79% makes our 55 year old a winner over the rest of his career and even our 25 year old (who probably started working in 2005) is okay as long as the increase does not go up more than .22% or a total of 1.92%. Which is to say that doing ‘Nothing’ since 2005 was a worst an even bet for even the youngest worker even if we simply get back to 2005 levels of payroll gap. And the numbers get kind of staggering if you back a little further out and look at 1997 and that 2.23% gap. Now you are looking at $13,800 in foregone taxes. It would take a huge jump in the payroll gap over the medium term to offset that.
As you start adding in real world variables the case only gets stronger. For example that 1.7% number in the 2008 Report assumed a specific performance for 2008: productivity 1.9%, real wage 1.3%, real GDP 2.3%, average unemployment 4.8%. The very failures that serve to make the payroll gap bigger for the next Report equally prove that a 1.7% increase a year ago would not have been enough to fix the problem anyhow. And this doesn’t even touch the effects of such a tax increase in what we now know was already an economy in recession.
Will there come a time when we are kicking our collective selves because we clung too long to ‘Nothing’? Well almost certainly not for me, I am just 15 years from full retirement. If you are 25 your mileage will vary. But in my view the case is still very strong for ‘Nothing’ until or unless the Trust Funds fall out of Short Term Actuarial Balance (TF ratio less than 100 in any of the next 10 years), which is the legal test for solvency. Which under current projections is about 2028.
But for those who insist that ‘We can’t afford to wait’ there are two responses, one polite, one maybe not so. Polite: ‘bet me’, (the less polite one alliterates). And then make them show their work on the back of their paper.