Thursday, November 1, 2007

Is Josh Marshall An Amateur on the Social Security Issue?

Josh writes:

When it comes to the policy and number-crunching nitty-gritty of Social Security I'm definitely an amateur.


This is the only line in his post that I disagree with. Permit me to quote a few of his best insights.



We need to remember that now and for at least a decade into the future Social Security is actually subsidizing the rest of the federal budget. The program brings in much more than it pays out. As we all remember from the voluble debates two years ago, the surplus is being used to buy US government bonds which go into the Trust Fund. And that socked away money will keep the program solvent through the middle of this century as the baby boomers retire, and revenues in no longer cover promised payments out. We've been doing that for about a quarter of a century. The problem on the political side of the equation is that the enemies of Social Security have spent a couple decades arguing that the Trust Fund doesn't exist or that it is simply a bookkeeping device with no true financial meaning. If that's true, it means that Americans workers have spent the last twenty-five years using their payroll taxes to subsidize general revenues and make it easier to float big tax cuts for upper-income earners without getting anything in return.


Well said! Read the rest as he has some real insights on the politics as well.


28 comments:

BruceMcF said...

Except there is no "socked away money" ... that dragon's hoard vision of the trust fund obscures the fact that when the trust fund starts "drawing out the money", what will really happen is that it will have to come out of general revenues from those years.

Which is why the Washington Post and David Broder and the usual suspects are clamoring for recognition of a crisis ... since general revenues are more progressive than payroll taxes, that represents more money from the top end of the income ladder going to fund the system that does so much to promote the economic stability that lays the foundation for their wealth.

And it seems that the Washington Post, David Broder and the usual suspects are opposed to the top one percent, or even top ten percent, contributing something more like a fair share of keeping the economic engine ticking over.

Robert D Feinman said...

The question is why does Josh understand this better than the Dem candidates who have all started talking about the SS "crisis"?

It's seems it is now blame the boomer time. The changes made to SS twenty years ago were a scam as Josh implies, but the formulas used took into account the boomer bulge. This didn't just sneak up on the planners last month when the first boomer applied for benefits.

There is a problem in Medicare, but fixing it would mean going after the big insurance, hospital and drug firms. Something that the Dems seem unwilling to do since they have been feeding at the same trough as the GOP. In fact the balance of donations has been shifting to the Dems, thus demonstrating, once again, that business doesn't have an ideology - they just buy who ever is in power.

Bruce Webb said...

Actually while Josh got this right early and often I don't think he has really examined the numbers in any kind of depth. Few people have. At one point I was astonished to see that Social Security threads rarely had an actual numeric component, everything remained at the conceptual level, with generalities that didn't get to the specifics.

And I think the reason is pretty clear, the other guys don't have the numbers to make their case and they know it. There is a saying among attorneys that perfectly captures the current state of play.

"If you have the facts, argue the facts.
If you don't have the facts, argue the law.
If you don't have the law, pound on the table"

I hear a lot of table pounding but practically zero data analysis. Privatizers don't want to discuss the assumptions of Table V.B2 or the dollar figures of Tables VI.F7 and VI.F8, not because they are unaware of them. They just don't like the story they tell.

rosserjb@jmu.edu said...

Let me just follow up on Bruce Webb, as I often do, which means that Josh is not so hot, and brucemcf is needlessly worried.

So, for the one hundred million umpteenth time. During the past ten years the economy and demography have performed better than the "optimistic projection" that is never reported in the media or discussed by the politicians, or even by people like Ball or Marshall. According to that projection, which has so far been worse than reality, the fund runs a surplus forever. No one ever has to worry about whether there is a trust fund or not. The fund just keeps lending money to the rest of the government forever. This is a problem?

Now, it is almost certainly going to be the case that eventually, some projections have it starting next year as the baby boomers really do start to retire in serious numbers, that surplus, now nearly $200 billion per year, will start to decline. This is a crisis? According to the still too-pessimistic "optimistic (or "low cost") projection" the surplus will decline until about 2030, the point at which the final baby boomers finally retire, although there will still be a surplus (I count on the indefatigable and up-to-date on the data Bruce Webb to tell us what that low cost projection surplus will be exactly in 203). After that time, the surplus just turns around and heads back up into the stratosphere.

It had a lot of flaws, but the Greenspan Commission, supported by Tip O'Neill and Ronald Reagan, did fix social security. Those changes were made to "pay for the baby boomers." They worked. The baby boomers will very likely be fully paid for if no changes are made, without ever even having to worry about whether or not the bonds held by the SSA will actually get paid for (as the ones for medicare are right now).

So, Marshall and Ball are better than most of the Dem candidates and their advisers, but still not good enough.

Barkley

Bruce Webb said...

$290 billion surplus leaving a year end total of $9.4 trillion in the Trust Fund for 2030 (adjusted for inflation $180 billion and $6.2 trillion)
Table VI.F8.-Operations of the Combined OASI and DI Trust Funds, in Current Dollars, Calendar Years 2007-85

Anonymous said...

Marshall's comments should certainly look familiar to all of us, though I don't remember writing to him with the details. So there are so many of us that agree that this entire issue is just so much bull. So what's to come of so much general knowledge that doesn't seem to be understood in the hall of Congress or, for that mater, in the general media?

As has been previously noted, one fly in the ointment is how the Trust Fund gets to redeem those government bonds. Since the monies have been utilized to supplement general tax revenues currently and in the recent past, it would appear that general taxes will have to be raised in order to pay down the debt. Given that there is so much debt in the form of Government Bonds, and given that in this specific case the debt occurs between different agencies of the govenment, guess which debt will be most dubject to default? Of course a reasonably balanced budget, exclusive of what is owed to the Trust Fund, would go a long way to reassuring that the Trust Fund bonds will actually have face validity. Right now there are too many con artists pushing the tale about the bonds in the Trust Fund being "just so many IOUs."

Absent the sudden appearence of a more trust worthy political class, one with some semblence of concern for their working class constituents, I think we're in big trouble.

rosserjb@jmu.edu said...

Jack,

Only if you believe the b.s. "intermediate projections," which there is little reason to believe. If the more likely outcomes occur, nobody will ever be cashing in any of those bonds. Just not a problem. Nothing to lose sleep over.

The only thing to lose sleep over is the continued efforts of politicians by both parties to screw around with s.s. It is fine. It ain't broke. Just leave it alone, everybody.

Anonymous said...

Come to think of it, why do Marshall's words read as though this is such a big surprise. I refer in particualr to his last sentence. "If that's tru, it means that the American workers have spent the last twenty-five years using their payroll taxes to subsidize general revenues and make it easier to float big tax cuts for upper-income earners without getting anything in return."

So what else is new? When was it that that .05% was not more highly regarded than the average American worker? For that matter, when did Congress or the White House occupant ever show a sincere concern for the average American worker? This is the point that America has to digest, that their elected representatives don't give a crap about anyone below the upper crust.

Anonymous said...

Barkley,
I'm not at all questioning the fact that the scare tactics are B.S., nor that the system will be well able to fund itself for the long term. But just think about what you're saying about the unlikely need to repay the Trust Fund. If the Trust Fund will continue to grow and be solvent with no need to cash in the IOUs then why are we paying the excess amount in payroll taxes? It is then all the more deplorable that FICA is being used as a general tax to fund the general operating budget and/or allow for the absurd tax cuts awarded to the top fringe of the economy.

Anonymous said...

Josh makes a good argument that so long as the Republicans insist that the Trust Fund is meaningless, it makes little sense to raise regressive taxes or benefit cuts to extend solvency. The R's could just pocket the money and continue pressing for big income tax breaks for the wealthy (and, later, more benefit cuts for everyone else).

As to the legitimacy of SSA's Intermediate Assumptions, the Social Security Advisory Board analyzed those assumptions a few years ago and basically backed them up (minor changes in immigration and mortality, which offset each other). My recollection is that the main reason why the projected GDP growth rate is low by recent historical standards is that the actuaries assume a relative decline in the number of working Americans, which in turn is based largely on things like fertility, mortality and immigration rates. They use productivity numbers generally in line with the last few decades, and the projected GDP growth flows from that -- it's just derivative. So, Bruce, the burden shifts back to you: which specific assumptions do you disagree with?

rosserjb@jmu.edu said...

Jack,

The dark secret only occasionally mentioned is that Ronald Reagan really was not that much of a tax cutter. The real bottom line of his tax policy was to cut income taxes on the rich while increasing the sharply regressive fica. We actually probably could have a fica tax cut and be fine (especially one balanced by an income or some other tax increase). But obviously we are having a hell of a time convincing most people that nothing needs to be done at all, much less that we could get away with cutting fica taxes. Heck, the game a lot of Dems are playing (see my Frankenstein Monste posting) is to use an Entitlements Commission to justify raising them while cutting benefits.

anonymous,

Well, no. The intermediate projection assumes a sharp decline in productivity, just as it did ten years ago, wildly pessimistically. The projections are almost identical to what they were ten years ago to now, which were way off on the pessimistic side. Please keep in mind that while there are honest technicians in the outfit, the SSA trustees themselves were Bush Treasury Sec appointees and totally for privatization. These numbers are being cooked, and very seriously so.

Barkley

Bruce Webb said...

"the Social Security Advisory Board analyzed those assumptions a few years ago and basically backed them up"

The Social Security Advisory Board is about as independent as the Defense Advisory Board which had such independent minds as Perle, Gingrich and I believe Ollie North on board. The SSA Board is loaded with outright advocates for privatization and near as I could tell pulled those mortality estimates out of nowhere, simply assuming that the US would catch up with the French on this front while ignoring the actual health outcomes.

There is a verbal sleight of hand going on here. That the Social Security Trustees are called 'Trustees' implies that their fiduciary responsibility is to the public, a fact that is amplified by having two 'Public Trustees' among them. Nope there is nothing to keep the President from loading both the two Public Trustees posts and the SSA Advisory Board with outright advocates of his positions and so it has proved.

In any event it really doesn't matter what the SSA Advisory Board or the Stochastic Analysis say, the proof is in the pudding.

Bruce Webb said...

Jack the answer to your question is complicated. Under Low Cost assumptions the Trust Fund is not irrelevant, interest on it is still needed to make up for the gap between Income and Cost. Under those assumptions is the Trust Fund projected to be larger than it needs to be? Well I suggest the answer is yes, overall contributions to the Trust Fund should be managed and monitored with realistic growth models in place, we don't need Trust Fund Ratios at 400 or higher, something like 200 would be perfectly prudent and allow time for adjustments to be made.

Now it is possible that the economy will return some version of Low Cost Plus, where current income never lags current cost at which time we would have some serious equity questions about paying down the Trust Fund in the face of the fact that the people who overpaid are no longer around to draw benefits. But first things first, we need to drive a stake through the heart of 'Crisis', then we can discuss FICA rates going forwards.

rosserjb@jmu.edu said...

I do not have the latest numbers, although I think Bruce Webb does, and he can correct me if I goof up here, but back in 2005 the story was roughly this.

So, the intermediate projection from 1997 forward had real GDP growing at 1.8% per year. Actual growth was 3.6% per year, higher even than the "low cost" optimistic projection (forget its exact number). The cutoff for avoiding a deficit on the fund was 2.2% (assuming intermediate demographic projectiosn). Above that one does not get a deficit ever (this does assume interest income from the fund is coming in, as Bruce has noted). These numbers may have shifted a bit in the last two years, but not by much.

Anyway, pretty obviously the intermediate projection is a cooked number, quite close to a cutoff beyond which there is no problem at all, less than 1/2 % higher projected growth rate, still way below what we have seen, even though these numbers have barely been adjusted at all since 1997, despite the massive failure of those projections.

Growth in the US in the 3rd quarter of this year was 3.9%.

So, one can say, oh, but what about the scary future when there will be only two workers per retiree rather than the current three, say aroundn 2030? Well, that is about what Germany has right now, with basically zero population growth. However, they are successfully paying pensions that are considerably higher than those in the US, and their economy is still growing (2nd quarter) at 2.5%, above that magic cutoff that says, "no deficit ever."

So, it is numbers like these that have people like me just absolutely pissed off when any semi-intelligent, semi-informed person even remotely begins to take seriously all this drivel about social security. It is just utterly worthless garbage and crap from top to bottom. That the Dem candidates are getting into this game is disgusting. They are just pawns of Goldman Sachs and other big Wall Street firms, scamming the public big time on this matter.

Barkley

PS: I count on Bruce correcting any mistakes I have made, or updating the numbers perhaps.

Bruce Webb said...

Well as to updates all anyone has to do is to link to my website. It essentially exists to give a pass through to the actual Social Security Reports. For example there is nothing about me or by me on the following page:
http://bruceweb.blogspot.com/2007/04/2007-report.html
Just links to the tables and figures of the 2007 Report. As an example clicking on the link 'Economic Assumptions' should take you right to Table V.B1 'Principal Economic Assumptions'. (Just now it took me to the correct section but not the table itself. Weird but scroll down a bit and there you are.) Table V.B1 gives you figures for Productivity (1.7% 2007, 2.1% 2008 then to 1.7% ultimate). Table V.B2 gives you figures for GDP (2.6% 2007, 3.0% 2008, 2.0% ultimate).

These are certainly possible outcomes but would seem more suited to the explicitly pessimistic High Cost alternative than the median projection that Intermediate Cost is presumed to be.

As for Low Cost. Well as a result of a horrific Q1 it probably will be a tad optimistic. On the other hand a repeat of Q3 numbers would put Real GDP at 3.1% for the year. Which is quite bit closer to Low Costs 3.4% than IC's 2.6%. And it was vaguely absurd to assert that 2007 Productivity of 2.0% was actually at the top of a range of possible outcomes. That is while we may not hit it, suggesting that it was an optimistic number is just odd. Q3 productivity is not yet reported but with a 3.9% GDP number we can expect something like Q2 2.6%, which would put us on track to meet or beat Low Cost even given a rocky Q1.

As to the two worker per retiree thing, that assumes that legal and illegal immigration combined will decline in both relative and absolute numbers. Whereas Low Cost assumes it will level out in absolute terms but still decline in real. Why a clamp down on legal immigration in the face of supposed labor shortages would be the selected policy option simply goes unexplained. Numbers in Table V.A1.

Anonymous said...

Bruce/Barkley, please help me understand your position on the productivity assumptions.

Here's what the 2007 trustees report says:

"For the 40 years from 1965 to 2005, annual increases in total productivity averaged 1.7 percent, the result of average annual increases of 2.0, 1.3, 1.2, and 2.3 percent for the 10-year periods 1965-75, 1975-85, 1985-95, and 1995-2005, respectively. However, it should be noted that this growth rate of 1.7 percent reflects a shift of employment from low (farm) to high (nonfarm) productivity sectors that is not expected to continue in the future.

Because productivity growth can vary substantially within economic cycles, it is more useful to consider historical average growth rates for complete economic cycles. The annual increase in total productivity averaged 1.6 percent over the last four complete economic cycles (measured from peak to peak), covering the 34-year period from 1966 to 2000. The annual increase in total productivity averaged 2.2, 1.2, 1.2, and 1.6 percent over the economic cycles 1966-73, 1973-78, 1978-89, 1989-2000, respectively.

..."For the intermediate assumptions, the annual change in productivity is assumed to average about 1.9 percent over the 2006 to 2010 period, then gradually decline to the ultimate assumed level of 1.7 percent by 2013."

What's wrong with this analysis? My impression is that both of you have focused much of your critique on the numbers over just the past decade or so. But SSA argues that you need to take a broader, long-term look at what's happened over several decades, and several economic cycles. It's not self-evident to me that they're wrong, or that this analysis reflects anti-SS bias. But I look forward to reading your response.

[Incidentally, your comments seem to refer to GDP growth and productivity growth interchangably, though they're obviously different statistics. Just something to watch out for.]

Anonymous said...

Bruce,
If "...interest on it is still needed to make up the gap..."
That is an interesting comment in the context of "those are just so many IOUs" as some people seem to be trying to say. Of course those same people don't seem to address the sanctity of debt payment for govt. issued bonds, but if tha be the case how does said interest get credited to the Trust Fund? If there's no intent to honeor the debt, why would there be any intent to pay the interest?

Frankly, these questions are more of the rhetorical kind. I raise them only with the intent of demonstrating that the entire issue is not based upon the reality of the numbers, but, instead, the political exigency of creating the aura of crisis with the goal diverting the funds collected in the form of FICA taxes and/or the monies in the Trust Fund as a means of supplementing the income tax revenues that are demonstrably inadequate given the budgetary habits of our conservative, ie aristocratic elements of the political and business community.
It's a political issue. It's not amenable to facts. It's based upon ideological argument with a significant hidden agenda. How many average citizens, remember who the silent majority are, read any of these facts, or know any more than they hear on the TV "news" or commentary programs?

Bruce Webb said...

Anonymous I am always careful to distinguish between Productivity and Real GDP in my comments. That may not always be clear if you just skim them.

The key words in the Trustee's analysis is "peak to peak". Why use that measure rather than total productivity? I think we all understand that growth is cyclical, that is why we use trend lines to smooth out the variation. And if you examine the actual data tables presented by the Trustees you will see that since 1960 we have had three five year periods below 1.7% productivity, one at 1.7% and five above it. And if you examine individual years since 1996 you will see only one below 1.7% and nine above 2.0% Table V.B1: Principal Economic Assumptions

If you read through the Reports in sequence you find similar language justifying the low numbers each time followed by a new Report showing better than expected performance. I read the Reports as they come out each year and have for a decade and for whatever reason my private judgement has been a better predictor than the conclusions of the Trustees. Well even a blind pig finds an acorn once in a while, so who knows.
"What's wrong with this analysis?" History. The Trustees projected the Trust Fund to go to depletion in 2029 with the release of the 1997 Report. They now project it to go deplete in 2041. Per you I should have just ignored my better judgement back in 1997 and accepted the analysis of the Trustees. Well I didn't and am not particularly inclined to start now.
EPI: Changes in Trustees Projections over Time

rosserjb@jmu.edu said...

Jack,

I confess to having been sloppy about GDP and productivity numbers. The latter are the ones that are indeed the most important.

The argument about ending movement from the farms is nonsense. That ended a long time ago, well before the acceleration of productivity that took place after 1997 (actually starting a bit before then). As Bruce notes, somehow this acceleration simply escapes the SSA Trustees.

The number I would like to see, and I do not know if Bruce has it, is the cutoff between the system running a surplus forever versus eventually running deficits (although even if that happens, the situation is expected to improve again naturally afte 2030 once the baby boomers stop retiring, so a deficit could appear and then disappear without the dreaded bankruptcy ever happening). It has been my understanding that this crucial cutoff number is only a bit above the intermediate projection and well below the low cost projection, which has fed my suspicion of cooking in the reports.

Anonymous said...

Barkley,
You addressed your comment to me, but I don't know why that is. It doesn't address the point I'm trying to make regarding the irrelevancy of the statistical minutia in the debate concerning the "crisis" that social security funding is facing down the road. Given that there is no real crisis, one plays into the hands of the dissemblers by continuing the debate. There are four different threads on this site currently dealing with this problem. The liars don't care about the facts. They simply want to continue the debate as though there is an argument to be made. Debating with a liar is giving credence to the lie. Don't debate. Just call it out for what it is, lies, deception and deceit.

PGL makes the point very well in his post concerning the National Review "pundit" Ramesh Ponnuru. The only debate is whether Ponnuru is plain stupid or dishonest. He's a Princeton grad, summa cum laude. So we know he's not stupid. That leaves only one alternative. He's a dishonest propagandist. He doesn't even have any reasonable credentials allowing him to argue an economic issue. He doesn't have to understand economic data because it is irrelevant. He only wants to propagandize. He's a deceitful ass, as are all those so-called pundits who simply regurgitate the false assumptions and the faulty statistical interpretations that they are based upon.

Lies can only be combated by identifying them as such, and dealing with liars requires the same tactic. It's not a debate. No one on the ground wants to be inundated with statistical data. The public just wants to know who's telling the lies. Th public appreciates the identification of the untrustworthy. That's the only real solution to any falsification. Just point it out clearly and loudly. Don't be shy about it.

rosserjb@jmu.edu said...

Jack,

Sorry. Guess it was anonymous I was addressing. I agree that lies should be called what they are. There have been so many lies in this by so many people that it is hard to convince people. Some basic numbers do need to made clear to realize what a pack of liars we are dealing with. The most telling piece of evidence has been the ongoing refusal of the SSA Trustees to adjust anything noticeably since 1997, despite the utter failure of their projections since then. But the media and commentators just keep on repeating their lies and going on about how "something must be done," to the point that we have this ridiculous spectacle of Obama joining the crew.

Anonymous said...

Bruce/Barkley --

First, let me be explicit about the fact that my interest in exploring your critique about the assumptions in the trustee reports does not reflect hostility to you or SS -- to the contrary. I'm not arguing that the SSA actuaries are right, I just want to understand your perspective and the facts.

Bruce, your latest post says that "The key words in the Trustee's analysis is 'peak to peak.' Why use that measure rather than total productivity?"

I don't understand this point. The Trustees took the average of all the economic cycles between 1966 and 2000, covering all the intervening years, and found that the average increase in total productivity was 1.6 percent. Since all the years were covered, I'm not sure why it matters whether the cycles were determined "peak to peak" (as opposed to, say, trough to trough).

In any case, forget cycles and this language from the report. Go up a few sentences in the report and you get this:

"For the 40 years from 1965 to 2005, annual increases in total productivity averaged 1.7 percent, the result of average annual increases of 2.0, 1.3, 1.2, and 2.3 percent for the 10-year periods 1965-75, 1975-85, 1985-95, and 1995-2005"

So there you have it: if you include 1965 and go all the way through 2005, total productivity went up an average of 1.7 percent. And the question for you is: why isn't that a reasonable basis for assuming productivity growth going forward for 75 years?

If your only counterargument is that productivity has been higher over the past decade, that seems unpersuasive to me on its face. First, why focus only on the last ten years when we have 40 year data available? What reason is there to believe that the last ten years is a more accurate predictor of future productivity growth than the four decades from the '60's on?

At first blush, at least, there would seem to be an argument that the last ten years is likely to prove an aberration in terms of productivity growth because it coincided with a dramatic improvement in information technology. While obviously there will be continued innovations in the future, it could well be hard to replicate the impact of computerizing so a large swath of the economy, as happened over the past ten years.

In any case, regardless of the relative impact of technology over the past ten years, and regardless of the short-term accuracy of SSA's predictions over that period, the broader point remains the same: when it comes to developing a long-term productivity assumption, why is SSA wrong to use 40 year numbers?

Bruce Webb said...

Anonymous mathematically that would seem to be impossible, It is difficult to tell what they did do but the numbers are as follows

1960-65 3.2
1965-70 2.0
1970-75 2.1
1975-80 1.0
1980-85 1.7
1985-90 1.3
1990-95 1.1
1995-00 2.1
2000-05 2.5

Which averages to 1.88%. Now it maybe that there is some effect that would explain how averaging five year periods should get a different result than averaging each individual year, I wouldn't think the difference would be .28, But the real point is we have had exactly one year since 1996 with productivity below 1.7% and only two years below 2.0%. Whatever you think of the Trustees methodology, it is absurd to posit 2.0% as the highest conceivable sustainable rate of growth going forward (Low Cost). Which is not to guarantee we will have high growth decades like the 60's and the 90's all the time but come on someone needs to have some faith in America. To suggest that the best is all behind us is like giving up.

Finally the proof is in the finer grains of the analysis. If you study the numbers in V.B1 closely in successive Reports you see something quite interesting. Despite current year growth coming in well ahead of projections the consistent assumption is that that is a new top limit is somewhere below that line. Moreover there are some really odd changes in assumptions for second year numbers.
Table V.B1 Economic Assumptions
2003 Report. Actual 2002 3.6% Projected 2003 1.9% Projected 2004 2.3% Projected 2005 2.1%

2004 Report. Actual 2003 3.4% Projected 2004 2.7%
Projected 2005 1.8%

Okay maybe you can explain why in the face of numbers coming in above 3.0% three years in a row that 2.7% would be a reasonable first year revision for 2004, but what is the possible justification for adjusting 2005 DOWN from 2.1% to 1.8%. What exactly motivated the Trustees to assume that a 50% slowdown was likely for the second year? For that matter given the actual numbers for 2002 and 2003 why would you select 2.8% as your top end Low Cost number. Was it actually impossible for 2004 to do any better than decline by 20%?

Sorry a careful reading of the Reports in sequence from 1997 to 2007 shows that Low Cost is being fit to a curve and Intermediate Cost artificially set below it. I made a prediction about the shape of outcome 1 in Figure II.D7 back in 2004. That prediction passed the test in each of the next three Reports: Low Cost (I) consistently produces a fully funded system with a flat trust fund ratio. Perhaps someone could demonstrate the mathematical likely hood that this could happen eleven years in a row. Be my guest.

BTW an engineer that posts at DeLong thought I had probably screwed things up and went and checked the data and came to the same conclusion. For whatever reason, perhaps just actuarial prudence, the economic models are lowballing probable near term growth. The next test of my theory is scheduled for sometime after March 31st. I am willing to bet on a 2007 Productivity number above 1.7% (IC) and am calling for something a lot closer to 2.0% (Low Cost).

SSA is not wrong to use 40 year numbers, but it is wrong to tinker with the near term numbers to make the results support a particular policy. Which is where the evidence is leading us.

Bruce Webb said...

Anon. Shorter version. Close reading of the Reports has shown a constant readjustment in the numbers in the out years to sustain some notion of 'Crisis'. This is particularly true since no one else seems to be using these numbers. In particular the President's budget doesn't use numbers this low, and I doubt that the people calculating future effects of tax cuts are either.

Who else out there is using 3.2% Real GDP as a ceiling for outcomes in 2012? and 2.2% as a Median? Or 2.1% productivity as a ceiling over that same period with 1.8% as a median? Has anyone told the bond market about this?

Anonymous said...

Thanks for that response, Bruce.

The explanation for the difference between SSA's total productivity number and your 1.88 figure seems to be that you included 45 years worth of data going back to 1960, whereas SSA went back only to 1965, which produces the 1.7 percent figure.

I haven't been able to find comprehensive productivity figures for years before 1960, but according to BLS (http://www.bls.gov/lpc/prodybar.htm), productivity between 1947 and 1973 averaged 2.8 percent. Which raises the question of why SSA stops at 40 years when looking back further probably would produce a higher average rate.

As for SSA's low-balling of short-term estimates, might it not be reasonable for SSA to use its long-term assumption after just a brief transition period?

In any case, assuming that SSA's long-term numbers are reasonable, but that they've been low-balling their short-term numbers, how big an impact does that have on 75-year solvency? That is, if the transition from recent history to the long-term historical norm were assumed to run, say several years, instead of about 2, how big an impact would that have on the shortfall/insolvency date?

Finally, one more point: the SS Advisory Board has a "Panel on Assumptions and Methods" which was supposed to release its report in July, but apparently has not. Have you been in touch with the panel to raise your concerns? It's one thing to post on blogs, but if you really want to have an impact it would seem to me that you should be in contact with the panel. You can find information about it here: http://www.ssab.gov/documents/2007Technical_Panel_announcement.pdf.

Thanks again for being willing to engage on this.

rosserjb@jmu.edu said...

anonymous,

A policy bottom line here is that given these differences in method, and the fact that based on what I have seen before, the cutoff productivity increase number is probably about 1.8 or at most 1.9%, given that 2.0 is the "low cost" projection that gives nowhere near a deficit ever, it would seem reasonable to wait a few years and see what happens. If, as seems awfully likely, the SSA projections continue to be as far off, or even nearly as far off, as they have over the last decade, nothing will need to be done. If we suddenly and permanently drop down to something mroe like their intermediate cost projections, well, despite all the whooping and hollering, there is plenty of time to make the relatively small adjustments of any sort that would fix it.

Barkley

Bruce Webb said...

"The explanation for the difference between SSA's total productivity number and your 1.88 figure seems to be that you included 45 years worth of data going back to 1960, whereas SSA went back only to 1965, which produces the 1.7 percent figure"

Well that would make some sort of sense (and I'll have to admit you are one of the few honest actors I have encountered on this) but I was using numbers from the exact same Report as the one you were citing. If the SSA Trustees did not want to have critics using numbers back to 1960 maybe they shouldn't have included them in Table V.B1: Principal Economic Assumptions Barkley on balance gives me way too much credit. I am not a number cruncher, I am a number pointer. Certainly it helped the Trustees to use a 40 year number by ignoring the 3.2% GDP number from 1960-1965, but it is not like any of this was my own work product. They gave me 45 year numbers, I used them. Present an explanation why excluding that five year period for the data set was justified and we can talk.

Bruce Webb said...

Shoot I almost always make that same mistake. Productivity is in Table V.B1, Real GDP in Table V.B2. And after you warned me. Per the Trustees productivity from 1960-1965 was 3.2%. Real GDP was 5.0%.