Arne's Blog

Scheduled Social Security Benefits

2010-12-09T12:47:00.000-08:00

According to the SSA annual report, the Trust Fund will be depleted in 2037. If that happens what will happen to people's benefits?

Currently, initial benefits for new retirees go up every year by an amount indexed to the increase in average wages. Since average wages go up faster than inflation, the value of benefits is going up. The chart (with data from suplemetary table VI.F10 of the report) shows this increase. It has gone up in the past. It will continue to go up according to "scheduled benefits" (the blue line, as long as the TF is not depleted. If the TF is depleted as projected by the report, then available funds would be enough for intial benefits to follow the red line.

Useless Comments

2010-07-11T06:48:00.000-07:00

Useless comments go here.

Wealth Distribution - Mean and Median by Age

2010-03-01T10:36:00.001-08:00

From 2007 Survey of Consumer Finances
4. Family net worth, by selected characteristics of families, 1989–2007 surveys

Mean Net Wealth is higher than median, so the distribution is skewed toward the higher net wealth.

The ratio of mean to median has increased for every group except 75 or more, so the distribution has become more skewed.

Median wealth begins decreasing for people entering the 65+ age group. Mean welath increases even after age 65 in 2007. This changed between 2004 and 2007. The 2004 data (which is also in Table 4), shows mean wealth decreasing after 65.

More Misleading

2010-02-28T20:37:00.000-08:00

Charts of the other two data sets.

Misleading?

2010-02-28T19:59:00.000-08:00

This chart shows data from the 2008 BLS Consumer Expenditure Survey
http://www.bls.gov/cex/#data
The data is calculated by dividing household income by the number in the household.

Part of this data was included in a comment on EconomistMom, but only the red highlighted part. The poster said, "We again see that per capita income, expenditures, and “food&fun” are highest for the oldest group ..."

This has got to be one of the grossest examples of misleading use of data that I have ever seen.

SS Projected Income/Cost #1

2009-04-06T18:38:00.000-07:00

Another chart for a blog comment at Angry Bear.

The Income/Cost values are calculated from 2006 report Table IV.B1 for combined OASDI. I failed to update to 2008 because I was just looking for a conceptual impact. The magenta line is projected according to Intermediate Cost and the Red is projected according to Low Cost. The idea was to determine what would happen if we chose to limit haw fast we allowed the SS Trust Fund to be used up. The green line is the calculated tax rate required to assure that tax receipts provide a minimum of 85 percent of each years costs. The would start going up in 2025 and would reach 6.7 percent in 2040 (which is the last year IC costs and income are provided).

Notes:
The scale is exagerated since they do not go to zero, but it makes it more readable to match the left and right scales.
I assumed that costs do not change as the tax rate changes because I have only the data from the table, no underlying models.

SS Projection Uncertainty and Recessions

2009-04-01T17:38:00.000-07:00

In a post at Angry Bear Bruce Webb counters someone's claim that the recession is going to force us to deal with changes to SS sooner. My thoughts (on accuracy of projections) require looking at this chart which is drawn from several different years reports.

You can see that the 1990-1991 recession caused actuals to diverge from the 1985 projection (green line). Employment was very slow to recover, taking about 5 years to return to pre-recession growth rates. By 1995 the projection assumed that the low growth rates had a permanent effect. Five years of growth brought things back up, but the 2001 recession turned it down again. Even with the slow recovery from 2002 to 2006, the 5-10 year projection from 2000 was beginning to look too pessimistic.

Social Security: Capacity of a PAYGO Program

2008-05-22T10:20:00.000-07:00

How many beneficiaries can be supported for each current worker? The annual report does not directly address this question, but it is fairly easy to use the numbers to estimate it.

Workers and beneficiaries data from:
http://www.ssa.gov/OACT/TR/TR08/lr4b2.html
Income and Cost data from:
http://www.ssa.gov/OACT/TR/TR08/lr4b1.html

For 2007, divide beneficiaries (163K) by workers (49.6K) to get 0.30 'demographic' ratio
Then multiply by income rate (12.75%) and divide by cost rate (11.26%) to get 0.34 'capacity'

This gives you real demographic data to plot along with a calculated value to give a graphic view of how large the annual surplus or deficit is. It tells you how many beneficiaries you can have if the average income per worker and average cost per beneficiary stays the same. The above plot uses actual data from 1970 to 2007 and projected values from the 2008 report for Low Cost (LC) model and for Intermediate Cost (IC) model.

Some observations:
1. LC projection can support more than IC as expected.
2. For both LC and IC the capacity trend is upwards. The growing economy can support a greater ratio of beneficiaries even though they are living longer.
3. There was a rapid increase in capacity and in workers (decrease in beneficiaries per worker) in the late 90s.
4. The graph shows the 2017 date for IC and 2021 for LC when we no longer have a surplus.
5. The difference in capacity between LC and IC is much less than the difference in demographic projections.
6. The math does not account for interest income, so there is really a surplus even when the solid line is slightly above the dotted line.
7. LC demographics will lead to surpluses even with IC capacity, but IC demographics will be in deficit even with LC capacity.
8. We can grow the economy to get out of a problem with people living longer, but not out of a problem not enough workers.

Rocket Science

2007-10-25T15:40:00.001-07:00

I am not an economist, but economics is not rocket science is it?

I performed the reliability analysis of the GOES attitude control system, so I am a rocket scientist.

Real Data was available

2007-02-12T09:45:00.000-08:00

.
.
.
.
.
.
.
.
.
Why did Mark Thoma at Economist's View not use real data in his post about how wage and price growth affect social security benefits?

SS is not overfunded

2006-11-01T15:09:00.000-08:00

I posted a comment at Economist's View that Bruce Webb is wrong about SS being overfunded.

I can agree that it is a great program, that the SSA preductions are overly pessimistic, that people are systematically trying to destroy it, and that under current rules (and the pessimistic forecast) beneficiaries will be better off in 2060 than they are now, without agreeing that it is overfunded.

My analysis as shown in previously posted charts indicates we have not yet seen the impact of demographics.

Increasing Income Inequality

2006-09-21T17:15:00.000-07:00

I have seen a lot of blogging about inequality lately. The above chart is derived from data at http://elsa.berkeley.edu/~saez/TabFig2004prel.xls. By subtracting the top one percent (bottom line) from the top ten percent (top line), I get roughly the professional class. ($91K to $282K) . I observe that something happened in 1987 and whatever it was only included the elite top 1 percent. If the Bush tax cuts have a significant affect, it does not show in this chart because of the volatility back to 1987.

Employment Drives Trust Fund Growth?

2006-07-20T12:00:00.000-07:00

.

.

.

.

.

.

I posted a comment at Economist's View that I believe employment is more critical than productivity to growth of the SS Trust Fund. When employment (as observed in the ratio of workers per beneficiary) changes, the growth of the trust fund appears to follow.

SSA Long Range Model - Life Expectancy

2006-07-13T22:19:00.000-07:00

.

.

.

.

.

.

Even the LC seems over conservative to me. The numbers for females seem to be already in diminishing returns. The numbers for males are catching up to females. There is greater than 25 percent difference in how long people could expect to draw benefits, which seems like it would create a impact all by itself.

SSA Long Range Model - Productivity

2006-07-13T22:08:00.000-07:00

Recent productivity numbers have been higher than the assumed values. When compared to the last 25 years perhaps not so bad. Then again, the trust fund growth beat out the estimates even during times when productivity was less than the 25 year average. How good is the SSA model?

Solvency of Program with Trust Fund

2006-07-08T20:31:00.000-07:00

The chart in the previous post does not represent reality because it ignores the interest that the Trust Fund is earning. With the interest included the lines for the Low Cost assumptions never cross - the combination of PAYGO and Trust Fund creates a fully solvent program.

In my opinion, while it seems likely that an adjustment will need to be made to account for changing demographics, I think the decision point is 10 to 15 years from now when we have a better idea of which curves we are following.

Solvency of a PAYGO Program

2006-07-08T19:17:00.000-07:00

Here is a different way to look at Social Security solvency. When the solid line drops through the dotted line, the contributions are not enough to cover the costs. This correspondes to the date when we will need to start using the interest from the trust fund. The chart helps make it clear that a pay as you go (PAYGO) program [which would not have any interest to draw on] that works for one demographic is not balanced when the demographic is changed.

Low Cost Assumptions for SS Trust Fund Solvency

2006-07-06T22:54:00.000-07:00

Bruce Webb has been posting about why the shape of curve I does not change from year to year, even as the Trust Fund exhaustion date reported by the trustees (where curve II crosses zero) has been moving to the right almost every year. http://bruceweb.blogspot.com/2004/11/what-is-low-cost-alternative-what-does.html
It seems clear to me that curve I (per the 'Low Cost' assumptions) is not an estimation so much as a calculation of a condition under which the trust fund is not overfunded. As Bruce observes, this condition has required more and more pessimistic assumptions each year. If he is correct, we will never reach exhaustion.

Social Security Actuarial Modeling

2006-07-03T15:43:00.000-07:00

Is productivity really the number one factor in modeling trust fund depletion, or does it look that way because we have not yet seen the uptick in population of retired baby-boomers?

Social Security Trust Fund Results - Post 2

2006-06-30T11:58:00.000-07:00

A little more research and I can (to some extent) answer my own question. Forecasts from 2000, 2003, and 2006 all agree that 2003 would have the highest growth rate [as a ratio of fund to costs].

Still, why did the growth rate drop faster than predicted?

Social Security Trust Fund Results - Post 1

2006-06-29T14:51:00.001-07:00

The economy is growing. The SSA predicts the TF will continue to grow for another decade.

So, why did the trust fund growth rate drop in 2004 and 2005?

First Post

2006-06-03T22:45:00.000-07:00

Given that there are not ads, where does blogger.com pay for the service?