William J. Polley

(Boy this would have looked a lot more readable if the HTML worked. The relevant links are;
2007 Social Security Report List of Tables
http://www.ssa.gov/OACT/TR/TR07/trLOT.html
2003 http://www.ssa.gov/OACT/TR/TR03/trLOT.html
2002 http://www.ssa.gov/OACT/TR/TR02/trLOT.html
LMS Nonpartisan Social Security Reform Plan
http://www.ksg.harvard.edu/jeffreyliebman/lms_nonpartisan_plan_description.pdf )

Sorry this from the Issue Brief this is kinda dishonest:

"Social Security faces a shortfall over the indefinite future of $13.6 trillion in present-value terms, an amount equal to 3.5 percent of future taxable payrolls."

First of all the term of art is not 'indefinite future', if you examine the title of Table IV.B6 you see that the real term is 'Infinite Horizon' (cue your favorite SF music theme here)
(2007 SS Report: List of Tables) . Moreover this measure is new, it is not to be seen in the 2002 Report but does spring to the light of day in the 2003 Report as Table IV.B7.

Now up to 2002 the standard long range actuarial window was the 75 year one and that is still the official measure and it seems reasonable enough. If we assume that a worker enters the work force part time at 16 then 75 years would have him 91. So why would we put in a new measure that extrapolated 1.7% productivity until the heat death of the sun?

Well the answer came clear to me in 2004 when Samwick first posted a version of his plan on DeLong and stated that the payroll gap was 3.5% (without qualification). When I replied "Whence 3.5% Samwick?" (since the more or less official 75 year gap was then 1.92%) I was directed to this brand new table and sure enough there it was. But who put it there? And why? Well the official answer in 2003 is here;

"Consistent with practice since 1965, this report focuses on the 75-year period from 2003 to 2077 for the evaluation of the long-run financial status of the OASDI program on an open group basis (i.e., including both current and future participants). Table IV.B7 shows that the present value of open group unfunded obligations for the program over that period is $3.5 trillion. Some experts, however, have described the limitations of using a 75-year period. Overemphasis of summary measures (such as the actuarial balance and open group unfunded obligations) for the 75-year period can lead to incorrect perceptions and policy that fails to address sustainability.

In order to provide a fuller description of long-run unfunded obligations of the OASDI program, this section presents estimates of obligations that extend to the infinite horizon. The extension assumes that the current-law OASDI program and the demographic and economic trends used for the 75-year projection continue indefinitely. Table IV.B7 shows that extending the calculations beyond 2077 adds $7.0 trillion to estimated unfunded obligations, making the total open group unfunded obligation $10.5 trillion. The $7.0 trillion increment reflects a significant financing gap for OASDI after 2077."

Who were "some experts"? Well as it happens it is right around this time that the LMS: Nonpartisan Social Security Reform Plan was under development and one of its authors was in a perch with a little influence over government economic reporting. And how did author 'S' describe himself in his own plan?

"Andrew Samwick is Professor of Economics and Director of the Nelson A.Rockefeller Center for Public Policy at Dartmouth College. From 2003 to 2004, he was Chief Economist on the staff of President Bush’s Council of Economic Advisers, where his responsibilities included Social Security"

Is 'Infinite Horizon' Samwick's direct work product? Who can say? But institutionally he was in the right place at the right time.

As to a possible motive you have to look back at the LMS plan and see what 'Reform' means in the here and now. This data is from table 2 on page 7;
Contributions to 75-year Actuarial Balance
Policy Improvement in Actuarial Balance
(% of payroll)
Raise Taxable Maximum 1.00
Change in Benefit Formula 2.08
Increase in Full Benefit Age 0.62
Diversion of Trust Fund Revenue to PRAs -1.56
Total 2.14

Now this does not include the 1.5% increase in payroll tax that goes right to the PRA (Personal Retirement Account). So what LMS actually proposes is increasing tax by a total of 2.5% and reduce benefits by 2.7% to 'fix' a program whose total current payroll gap is 1.95% (up from 1.92% when the plan was drafted.)

Does LMS provide a better result for the average retiree than simply letting raising the payroll tax by 1.92%? Or by some mix of 1.35% payroll tax increase and LMS's change in retirement age effect of .62? Well no, not according to Table I, in particular the Single Earner couples actually come out behind at all income levels.

How do you get people to buy into a plan that proposes a benefit cut equal to 2.7% of payroll AND a 2.5% tax increase in face of a 1.95% gap, and moreover has 1.0% of that increase come by lifting the cap and taxing people making between $97,950 to ultimately $167,000? Well it helps a lot if you can tell them that the 'real' payroll gap is 3.5% (without adding the 'over the Infinite Future' piece).

Now I am sure there is a reasonable defence to be made for Infinite Future. On the other hand it would be nice to have the economic assumptions under which the 'Expected Yield on Mixed Portfolio' is derived. Because that 3.5% payroll gap assumes a specific set of economic projections that include ultimate Productivity at 1.7% and ultimate Real GDP at 2.0% (2007 Report; Tables V.B1 and V.B2) any amount of economic growth over that shrinks that 1.95% gap all on its own.

Any proposal to current workers that starts with a 3.5% payroll gap without equally mentioning the 75 year 1.95% gap is an effort to pull a fast one in my view. Current workers simply don't benifit from LMS in any significant way and over the short run are seriously disadvantaged by the immediate tax increase.

Well I sent a humongous comment focusing on the numbers of LMS, hope it survives the cut. Now for the political component.

The LMS Plan has some merits. In the short run it deprives future Administrations and Congresses of the cover of current Social Security surpluses when making spending increases or tax cuts, and that's good. In the long run it partially solves the intergenerational obligation piece, of current workers having to rely on future ones to fund an ever more generous benefit. Personally I don't see this as a problem but I can understand why the authors do.

But politically it just doesn't fly. First the cap raise, described in the plan as follows;
"A bit more than two-thirds of the 1.5 percent of payroll diverted from the trust fund is replaced by gradually restoring the percentage of earnings that is subject to the OASDI payroll tax to 90 percent (where it was in 1982) and maintaining it at that level thereafter. This increase in the taxable maximum would be phased in between 2008 and 2017. The new long run taxable maximum is equivalent to $171,600 at today’s wage level.5 The maximum level of earnings to be included in benefit calculations remains the maximum taxable earnings in current law—workers receive no incremental benefits from the increase in maximum taxable earnings."

Now I am all for returning taxes to 1982 levels but near as I can see LMS adds effectively adds 6.2 points to the 28% bracket and the lower reaches of the 33%. Good luck selling this to the Faculty Club, still less to Congressmen and senior Hill and Administration Staff. They just got smacked in the wallet and hard. Moreover the net effect is to reverse much of the Bush tax cut. In a stroke you have pissed off the entire public and private policy elite, and in return they get? Nothing.

Moreover you have just blown a huge hole in the payroll budget of any firm with relatively high income wage workers. Lets just say that Bill Gates won't be pleased.

And workers can do the math, and if they can't we can help them. The difference between LMS and a straight out payroll fix is .45% of payroll which is by the way down from .52% of payroll last year. The decade long trend has been for the payroll gap to shrink and there is good reason to believe that is going to continue. I don't know of anyone outside the SSA Actuaries office that is committed to long term GDP settling in at 2.0%, and any amount of growth over that acts to shrink the current gap.

Finally older workers will not be pleased with the increase in eligibility for early retirement from 62 to 65. The majority of workers currently take early retirement. Because they are tired and want to spend time with their grandkids while they are still reasonably active. This plan like all plans that rely on mortality tables and actuarial fairness and ignore the reality of what it means to work manual labor into your 60's is Ivory Tower thinking. Shoot give me tenure and an office in Dwinelle Hall and you would have to drag me shouting into retirement, it is a little different when you are working moving apples down at the port. Shoot if you lowered Medicare eligibility to 62 you would likely have a stampede. I personally know people in that age bracket who would walk tomorrow if they had medical coverage.

Others can read through the advantages of the LMS Plan on pages 7-9. Frankly they seem bloodless and unlikely to fire up any current legislator or voter, the advantages of this plan over others or in doing nothing are not such as to overcome the short to medium term pain. All of the benefit of the plan such as it is accrues to people who are still in the workforce after say 2055, few current workers would see any financial advantage at all. And as they say 'Money talks'.

http://www.ksg.harvard.edu/jeffreyliebman/lms_nonpartisan_plan_description.pdf

Much shorter version. Infinite Future Horizon is a gimmick introduced into the 2003 Report to allow privatizers to use 3.5% as their payroll gap rather than the 1.92% it was then over the 75 year horizon or the 1.95% it is now.

It might be true that the gap cannot be entirely eliminated by productivity growth, although the authors of the Issue Brief assert that rather than demonstrate it. I don't agree, I have been reading the Reports as they came out since 1997 and I see a payroll gap that has shrank from 2.23% to 1.95% all without any action at all. Moreover that shrinkage largely tracks productivity improvements over the projections.
http://www.epinet.org/content.cfm/issueguides_socialsecurity_changes

The authors further claim that inactivity simply increases the cost. Well that simply has not been true in nine of the last eleven years. In particular we saw a rather dramatic shrinkage in the payroll gap from 2.02% in 2006 to 1.95% in 2007 and this despite some pretty dismal productivity and GDP numbers. Will the payroll gap drop at that rate on average for the next 28 years and so go to zero? Hard to say. But it did drop on average by .085 points annually between 1997 and 2000, that is we know it can happen because it has happened.

The Treasury Department is claiming that tax cuts will produce growth but will only commit to 2.0% GDP ultimate when it comes to Social Security calculations. Per the Social Security Trustees (which BTW include the Secretary of Treasury) there has not been a 5 year period of sub 2.0% annual GDP growth back to 1960 with the lowest of all being 2000-2005 at an average of 2.4%.
http://www.ssa.gov/OACT/TR/TR07/V_economic.html#wp192078

I have yet to see a credible defence of the economic numbers of Intermediate Cost in light of the claims made by tax cutters. Because 2.0% GDP doesn't cut it. For that matter 2.4% Real GDP is not exactly proof of the magic of tax cuts.

Maybe someone can explain why 2.0% GDP is the likely mid-point for growth past 2020. Me I suggest turning that frown upside down.

Let's focus on real problems like healthcare.

My original post may have violated both the bulk and collegiality thresholds of the Econoblogs. I got a private e-mail from DeLong in 2004 defending his professional colleague on much the same grounds.

But it is no longer good enough to simply assert "We can not grow or tax our way out of this". Intermediate Cost gives the answer to "What would it take to tax our way out of this?" Answer an immediate 1.95 point increase in FICA. Low Cost gives the answer to the question "What would it take to grow our way out of this?" and the answer is 2.8% GDP growth. Now it would be perfectly legitimate to argue why 1.89% tax increase is unreasonable or why 2.8% GDP growth is unreachable but to simply refuse to discuss the models is unacceptable. If economists won't discuss these issues with relevant numbers then what is the point?

Put your models on the table and let us compare them to the ones presented by three Cabinet Secretaries. Low Cost is out there, simply ignoring it is not going to keep this yapper from biting ankles.

I don't mind listening to myself speak. Issue Brief no 1 makes for claims;
1) Social Security faces a $13.6 trillion dollar funding gap over the 'indefinite future' which translates to a 3.5% payroll gap
2) The only way out is some combination of tax increases and benefit cuts
3) Delay makes the fix more costly for future generations and is such unfair
4) There is no realistic way to grow out of this

The last three points are debateable while the first is misleading in context.

First the payroll gap for the next three generations is only 1.95% and has a trend line that is declining. I can see financial planning for my grandkids but the LMS plan calls for me to sacrifice current income and future retirement benefits for my great-great grandkids's benefit. We are expected to believe that it is too expensive to pay for heating oil for seniors or health care for children today but by God we need to worry about taxation levels in the 24th century. But clearly the bigger risk to those future generations is handing on the General Fund deficit, which of course could be largely addressed by applying LMS rates directly to the top marginal rates. So think we can discount compassion for the future as a motive here to zero. So much for points 1 & 3.

Points 2 & 4 are simply the flip sides of the same assertion. The infinite future 3.5% gap or the 75 year 1.95% gap are not absolute. They are contingent on a specific economic model. Using either implicitly endorses that model and any solution has to come from within it. The No Economist Left Behind challenge showed pretty conclusively that you can not get historic rates of returns under Intermediate Cost assumptions. The challenge to LMS is two-fold, either meet the NELB challenge directly, or reveal your own economic model. If the latter be sure to rescore Social Security solvency under it. Maybe this has been done, but I read through Professor Samwick's Social Security archive on VoxBaby and didn't find anything of the sort. If I missed it then please supply a link. It doesn't seem a lot to ask before hiking my taxes and cutting my benefits.

Last shot. Nobody is talking about this. I just took a quick run through among the Center-Left Econoblogs, searched the NYT archive and via CNN Googled the web and came up with nothing. Certainly Google picked up both the Issue Brief itself (result 1) and the Press Release but other wise nothing. With one exception. The WaPo did run a story on Tuesday on page D 01 obviously based on the Press Release.

If Secretary Paulson and Professor Samwick thought this would get the discussion jumpstarted, well lets just say they still have some work to do.

Look at this thread. You got me. And Lord asking if we could just talk about something real. From all appearances this issue is not only on the back burner but maybe in the deep freezer somewhere under the frozen peas.

Sorry about the delay, Bruce. I've been swamped. Your first comment was sent to the moderation bin because of all the links. I did find it among the spam. I'm hoping that Movable Type 4 will handle spam and comment moderation better. Stay tuned.

Time is scarce again tonight, but let me just make a couple of remarks.

*I don't know about anyone else, but I wasn't expecting this to jumpstart a discussion. Any serious talk of SS reform is dead until, oh, probably 2016 or 2020. Nevertheless, I think there is value in continuing the debate in academic circles. You make a number of valid points, and I am happy to give them a forum.

*I concede in my post above (and two year ago) that you can't expect any old privatization scheme to produce a Pareto improvement. I would not support any SS reform that diminishes the benefits or raises the taxes of people close to retirement. My ideal reform would be phased in over a very long time...as long as politically feasible. My cutoff age for implementation of any reform would be quite low, perhaps as far as to say that anyone over 30 today would see no change at all. The more gradual the better.

*Your comment concerning the political difficulty is well taken. I started to write something about that in my post, but deleted it in the interest of keeping the post brief and focused. I'm glad you brought it up. I agree that it's a problem, but should that cause us to throw up our hands? Why wouldn't a plan that held harmless anyone over 30 and offered improvement to those under 30 be politically workable? Again, I'm pulling 30 out of the air. (In their papers, Altig and Gokhale originally said 43, then 32.) Have we no statesmen willing to get behind a creative, gradual, and Pareto improving solution? If not, I guess I should just slink back to the corner.

*I'm sympathetic to your argument about the Low Cost solution. Very. As my old post makes clear, I've never been one to think that this is an urgent problem, and I do not reject out of hand the possibility that we could grow out of it. The urgency implied by large estimates of the shortfall are not what motivates me. Instead, I resonate with what you said above:

"The LMS Plan has some merits. In the short run it deprives future Administrations and Congresses of the cover of current Social Security surpluses when making spending increases or tax cuts, and that's good. In the long run it partially solves the intergenerational obligation piece, of current workers having to rely on future ones to fund an ever more generous benefit. Personally I don't see this as a problem but I can understand why the authors do."

Yes.

So why not take a bold but gradual approach to offer future generations a SS system that relies less on intergenerational transfers and takes the SS surplus out of the hands of politicians? The last attempt wasn't it, and I'd rather keep the status quo than do a bad job of reform. We can both agree that it's not worth it.

I may be more optimistic that a "good" reform is possible. So be it. While I haven't constructed a complete plan yet, maybe I will someday. Like I said, we've probably got another 10 years before the pendulum swings back around. But it will.

All the advantages of LMS can be achieved by simply diverting the current surpluses into different asset classes managed much like the current federal Thrift Savings Plan or Calpers (retirement system that covers most California public employees). Given that LMS proposes mandatory annuitization and nobody suggests individual workers will have day to day control of their accounts I can only conclude that calling them IRAs is strictly motivated by ideology.

The interesting thing is that LMS does not propose a 'fix' at all as conventionally defined. Its combination of a 1.5% payroll increase and lifting the cap (for another 1% of payroll) would be more than enough to fund the current 1.95% payroll gap and then some. Instead it piles on with a combination of benefit cuts and increase in retirement age simply to wipe an obligation off the General Fund off the books at the expense of current workers under cover of 'infinite future horizon' projections.

It's not going to take until 2016 or 2020, the economic models strongly diverge about 2012. By that point we will know. If we can just punt the ball past 2008 and put in a Dem Pres the danger zone will have passed.

I haven't been talking about LMS. I think we can do better. For one thing, I would be inclined to leave the tax rates alone and introduce private accounts only for workers young enough to be able to build substantial wealth in them. Anyone over that age would see no change. Period. The rest is just the details.

A good stretch of productivity and GDP growth could well push the day of reckoning past 2012.

2012 is not paticularly a 'day of reckoning' as much as the point by which it will be abundently clear what the outcome will be.

Private accounts as an add on to the current system would be fine. The simplest way to achieve that would simply be to open the federal Thrift Savings Plan to Americans generally.

But nobody has really shown how they could or should replace any part of Social Security as currently configured. I used LMS because it is Professor Samwick's work product, but Posen had the same defects. For lower income workers they simply don't work.

The drive for private accounts is fundamentally one of ideology and not of economics, the plans all tend to shatter when run up against real world numbers, which is why most privatizers try to avoid discussions of such things as productivity to start with.

People are generally shocked at how low the numbers have to be for the system to balance as is (i.e. Low Cost)
Table V.B1 Principal Economic Assumptions
http://www.ssa.gov/OACT/TR/TR07/V_economic.html#wp188118
Here we see ultimate productivity (after 2013) only needing to be 2.0%. On the other hand CPI of 1.8% seems pretty optimistic.
Table V.B2 Additional Economic Assumptions
http://www.ssa.gov/OACT/TR/TR07/V_economic.html#wp192078
Here we have Real GDP at 3.0% in 2013 and unemployment at 4.7%. Fairly strong numbers but consistent with recent economis performance.
Table V.A1 Principal Economic Assumptions
http://www.ssa.gov/OACT/TR/TR07/V_demographic.html#wp185380
Fertility seems fairly high. On the other hand the immigration numbers seem more realistic for Low Cost 1.4 million in 2015 as opposed to 1 million for Intermediate Cost.

Low Cost is not a guarantee and some outcome between Intermediate and Low clearly possible, I just can't see every variable and most particularly Productivity and Immigration slumping to the levels needed.

Sorry about the delay again. My day job has been keeping me busy.

Your points are well taken. I have previously noted that an important part of Social Security is its function as an insurance policy. But it is also a source of retirement income. The result is a good type of insurance for those with low incomes but a lousy retirement plan.

But you can't separate the two. A person can't say, "I don't need that insurance because I can buy better insurance elsewhere, so count me out of your lousy retirement/insurance plan." Don't get me wrong. I'm not suggesting that we should allow that. Not at all. Just pointing out that it is a combination of the two.

Whole life insurance is similar. You pay more for whole life than term because you're also building up cash value. But the growth of that cash value is less than what you can get elsewhere because you're also paying for the insurance.

But the government doesn't have to sell Social Security like a life insurance salesman has to sell whole life policies by trumpeting the investment aspect of it. The government could separate the investment from the insurance if it wanted to. The fact is they don't want to. It's not politically smart. I can accept that as a practical matter even if it bugs the daylights out of me.

I would like to see a portion of the FICA tax kept as a social insurance premium (let the actuaries figure out how much it should be to support the insurance goals of the program) and allow people to opt into something like the Thrift Savings Plan with the remainder if they wanted to.

Even if it was just 1% it would be an improvement. And of course people could put in more than the required amount if they choose.

But again, as always in my ideal plan, it should be phased in so that anyone over a certain age (the exact age could be debated) would be kept on the current system and held harmless with respect to benefits.

The devil is in the details. Social Security is not the ideal retirement plan, on the other hand it has been getting better every decade. At this point the pure legacy costs are mostly gone, a person entering the work force before 1935 and collecting benefits today would be 90 years old, and even then would have over 45 years of contributions. Now we are simply tasked with supplementing current retirees checks out of 2007 productivity.

On the current schedule of benefits a retiree in 2041 will have a benefit 160% in real terms compared to a current retiree with
comparable work history. The progressivity of the system across income levels is fairly mild, the progressivity of the system across time is pretty strong. Of course under current projections the system can only yield 75% of that or 120%, still not a bad deal. The task for privatizers is too show better results using the same numbers. To date they have failed.

James Buchanan put the problem like this in the first article of the Fall 1983 Social Security issue of the Cato Journal.
http://www.cato.org/pubs/journal/cj3n2/cj3n2-1.pdf

"Economists are professionally trained to search out and to discover alleged inefficiencies in institutional arrangements, public and/or private. Further, once alleged inefficiencies are located, economists have an impulsive proclivity to propose reform. I am no different from my peers in these respects; on several occasions, I have joined the ranks of those who have advanced proposals for reform in the Social Security system.’ As I have emphasized in earlier papers, however, the task of designing reforms that meet the test for Pareto superiority, even when considerations of practicability are totally neglected, is by no means an easy one. But, of course, unless some such reforms can be demonstrated to be possible, the existing system
must he judged to be Pareto efficient."

That is a nutshell has been the problem, the plans on the table have not met the test for Pareto superiority at least for workers at or under the median. Now if the discussion were not so infused with ideology we could examine this with open eyes. For example privatizers tend to blur the differences between a system of private accounts and a system whose assets are more widely diversified and attribute the efficiencies gained by the latter to the former. But there is no reason that the Trust Fund could not capture whatever premium stocks have over bonds long term without having the huge overhead of private accounts. We only have to look at CALPERS or other large public retirement plans to see how that works. But fundamentally the argument is not being driven by efficiency, instead it is being driven by a Reaganite "Government is not the solution, Government is the problem" stance. Improving Social Security while keeping it under government control is not an option for most of these people, it is private accounts for everybody or nothing, half measures are not welcome.

I wrote a couple of posts and diaries back in 2005 and 2006 that discussed the implications of Social Security Solvency including this one from Jan 2006 http://bruceweb.blogspot.com/2006/01/invest-or-divert.html . In it I propose what might be seen as the governmental version of LMS, the end result is a Social Security system that is totally delinked from the General Fund with the Trust Fund fully invested in non-federal asset classes. The plan would accomplish most of the policy goals of LMS at a lower cost, it just wouldn't have private accounts.

But really the whole debate has to start from honest disclosure and discussion of the numbers. The question of whether Social Security could be improved by separating out the insurance and retirement components or by reexaming the cap to see if it may in fact be set too high is fruitless without an honest scoring of the economic projections underlying the various plans. We are not getting that. Under trend growth Social Security is fully funded as is. Which doesn't make it optimal. But until supporters of private plans admit the fundamental economic realities supporters of traditional Social Security are not going to give an inch.

We could call this the Show Me Your Pareto stance. Instead the other side wants us to concede that any plan based on private accounts must be assumed Pareto Superior. That was the Bush plan. Assert that we couldn't tax our way out, equally assert that we can't grow our way out, and force everyone to agree that something had to be done. Well not so fast. We know what it would take to tax ourselves out. 1.95% of payroll. We know what it would take to grow our way out. 2.1% productivity and 2.8% GDP. Or some combination like an additional 1% of payroll and a concerted effort to maintain average Real GDP at 2.4%. There hasn't been a five year period back to 1960 with growth lower than that and in fact only three with Real GDP averaging less than 2.8%. Who decided that these numbers are impossible going forward? Where is the honestly brokered discussion of future productivity trends?

Chirp.