Soc Sec XI: Seven good questions by Reader Arne

Reader Arne, a long time and well informed commenter submitted to Tuesday’s Social Security post a list of seven deceptively simple in form questions to get my take on them. Which I gave in comments. Unfortunately what HaloScan gives it can take away and a time this morning both his questions and my answers had vanished. Well they came back and I thought it useful to front page them and invite readers either to attempt their own answers or comment on mine.

Remember the rules. Anyone can ask for a response to a question or argument. No one has the right to demand one. Silence in this case does not equate to consent.That said what is on your mind? Anything related to Social Security is fair game. But first over to Arne.

Here are some questions I think are interesting, but I would prefer to see Bruce post them with his thoughts and for others to try to stay tightly on-topic.

1) What should we do if the SSA intermedicate forecast is right? When and how would we change?
2) Which is really more important, productivity or demographics?
3) Given that the TF lasts at least 20 more years, is moving (some of) it from US Treasuries to equities a good idea? How does that change with LC instead of IC?
4) Is solvency relevent? {mild edit to remove a side irrelevant side issue}
5) Are the stochastic projections (Appendix E) meaningful?
6) What do the 1984 and 1985 reports show us about intent?
7) Why did the TF ratio outperform expectations in the late 90’s? Can it do so again?

1. Well there are two answers to that. One would rely on what I call Rosser’s Equation and point out that ‘crisis’ means a 78% result of a benefit scheduled at 160% relative to what similarly situated retirees get today. Since 78% of 160% = 125% and the event if and when will come after I am gone, I don’t see why I should get exercised about this one way or another. Particularly if the plan to avoid sticker shock is just to phase it in faster. So one perfectly good response is Nothing.

The second answer would be to point out with Coberly that a tax based solution is in fact pretty cheap. The current payroll gap is down to 1.7% of payroll which for a household of $50,000 would require an $850 a year fix, or half that if you examine that from an employer/employee split. But as I argue in Cost of Inactivity the trend short term is good and mathematically calculates out to Nothing for now as well.

The legal answer is to wait until the Social Security system falls out of Short Term Actuarial Balance. Under Intermediate Cost assumptions the TF ratio falls below 100 in around 2036, meaning the system would fail the test some nine years earlier or in 2027 which would leave 14 years to 2041 to come to a policy decision, which would be plenty of time and a much more informed information environment. Of course if depletion continues to move out in time then so would this decision point. So once again the answer to your question in the here and now is Nothing. Ask me again in 2027.

2. For decision making? Productivity. Because the Demographics are already built into the model and do not vary much over the short term that will be determinative, which is to say the 2009 to 2012 time period. Fluctuations in fertility obviously only manifest themselves 21 years or so out of phase. Whereas year over year productivity can vary greatly.

3. Diversification of the Trust Fund is in my view a good thing, though equities would not be my choice of asset class. The question is to some degree moot under IC. Surpluses start shrinking after 2017 and vanish after 2023. Given that fairly short investment window the difference in yield probably wouldn’t offset the risk of markets crashing at just the wrong time. We would be a little embarrassed if the Great Crash of 2023 wiped out most of our portfolio just as we needed to cash parts of it in. Under LC I would argue that diversification is almost an absolute necessity in that it avoids the Interest on Interest problem. If we assume Low Cost and took the current surplus and invested it in outside assets and then reinvested the earnings we would have a chance to really build a portfolio. Under Low Cost the time to tap the Trust Fund portfolio does not start until 2023 and at a much lower rate amounting to 25% of the interest accrued from a portfolio of Special Treasuries, presumedly something less with better yielding outside assets. You can do quite a bit with a reinvestment rate of say 80% of earnings. As to asset class I would pick school and transportation bonds which would serve to create jobs, lower borrowing costs to states and local jurisdictions, yet maintain the predictability of Treasuries.

4. Solvency is relevant because opponents of Social Security have framed the argument in those terms. They chose not to fight this out on straight out ideological, pro market terms. Instead they sold the message of ‘bankrupt’. I can only fight on the battlefield in play and for the most part that is one of Solvency.

5. I don’t find the Stochastic Projections compelling. I am even less a statistician than I am an economist but the following passage was troubling. Perhaps you could explain it:

“Each time-series equation is designed such that, in the absence of random variation, the value of the variable would equal the value assumed under the intermediate set of assumptions.”

The language in Appendix E is rather opaque, perhaps deliberately so, but it seems to build in the assumption that is actually in challenge, that Intermediate Cost is overall a reasonable median. But really the question of the Stochastics are above my pay grade.

6. Well the unkind answer would be to say ‘Look’ and link to but I went ahead and checked the 1985 Summary. They essentially gave the system a clean bill of health with the actuarial deficit of Intermediate II-B only being .41% while the 1985 definition of ‘close actuarial balance’ was ‘estimated average annual income rate between 95% and 105% of the average cost rate’. So I would sum their intent as ‘maybe we fixed it, but lets keep a close eye on it’. Because they were only .09% of payroll of falling outside their preferred band.

7. Well that one is easy. Whether you take as your key variable Real GDP, Real Wage, population in employment all were way ahead of expectations. Whether the late nineties could happen again depends of what you think caused it to happen in the first place, a question which has economists of all ideological types coming up with different answers. Personally I am a little more optimistic that those times could come again. Then again they are not needed to achieve Low Cost results. Low Cost only assumes 2.8% ultimate GDP and 2.0% productivity both well below the rates we were seeing in the late 90s. Which is probably the least understood thing about Low Cost. In historical context its numbers seem to be pretty low hanging fruit. To me anyway, I am still waiting for someone to explain why no year in the future could possibly be as good as 2006. I don’t need to appeal back to 1999.

Fire away.