Example of what changing from CPI-W to C-CPI-U would mean

[Yo_Mama]

The source data for this example is here:
http://www.bls.gov/news.release/cpi.t05.htm
and here:
http://www.ssa.gov/cola/automatic-cola.htm

If one began in 2000 with a $1,000 benefit, in 10 years using CPI-W a recipient would have ended up with a $1,314.12 monthly benefit.

If one began in 2000 with the same $1,00 benefit, in 10 years using C-CPI-U a recipient would have ended up with a $1,247.44 benefit - a five percent reduction. Of course it is cumulative - each year the lower inflation adjustment occurs to a lower benefit, so in 20 years one would be down probably somewhere around 10%. After 25 years it would definitely be more than 12%.

For example, if one uses the same sequence for another 10 years (repeating the 10 year sequence of relative adjustments), the result is:
C-CPI-U: $1,556.10
CPI-W: $1,726.90 (-$170.80)

The comparison is not perfect because if we were using C-CPI-U, we would not be using the December version - we would be using the Q3 version, which has not been published.

Nonetheless, this amounts to a very significant cut in benefits, and it will be much worse for the poorest elderly, and it will hurt the oldest beneficiaries the most. To put it in current dollar terms, it is as if we are taking that $1,000 monthly benefit check and cutting it to $900 for older retirees.

CPI-W is constructed using the buying patterns of a poorer group of wage earners. Using C-CPI-U instead of CPI-U not only increases substitutions, but also compares the average wage-earner's purchasing patterns to those of a group of people with different purchasing patterns.

To understand how dangerous this can be, consider the Consumer Expenditure Survey broken down by age bracket:
ftp://ftp.bls.gov/pub/special.requests/ce/standard/2011/age.txt

Consumer units over 75 years (age of reference person):
Spends 13.9% of income (average) on food, compared with a share of 10.5% of income for all consumer units.
Spends a total of $2,755 per person annually, or $230 a month. Thus the estimated 20 year gap of $170 dollars is a significant fraction of expected food expenses for an elderly retiree.

All of this implies that this change will not in fact save the money that it is theoretically expected to save, because by reducing the SS incomes of elderly people by 10%, you will be pushing many more of them into qualification for Medicaid and/or food stamps. Medicaid and food stamps would correct this problem by essentially adjusting away the 20 year lower inflation adjustment, but it won't save the federal government any money for those poorer households, will it? Currently Medicaid and food stamps for an elderly person cost the federal government more than the $170 savings - in most states it is at least $200 in subsidy a month.

So this is a farce, but a rather cruel one. Older people with low incomes spend higher amounts relatively for the basics. When they are already eating eggs and pasta, figuring the substitution of chicken for beef is irrelevant. That occurs in higher income households, but not for these people. When the cost of pasta goes up by 30% in 5 years, they just have to turn out the lights or eat less, or go to the food bank, or apply for additional benefits.

Average monthly SS checks are not high at all - see the Statistical Snapshot for October 2012:
http://www.ssa.gov/policy/docs/quickfacts/stat_snapshot/

The all-beneficiaries average check is $1,132.80, and the average retirement check is $1,239. Fooling around with benefit cuts of this magnitude simply isn't going to produce the savings expected, because a great deal of the "savings" will be compensated for by increased benefits to the elderly in other poverty programs! A three-hots-and-a-cot benefit program doesn't allow for these types of cuts to save money unless we also cut other benefits for the elderly.

I believe we should be a "reality-based" party, and I'm not seeing the realism here.

[newfie11]

1. Thank you for making this clearer

It still pisses me off royally

[Yo_Mama]

2. Personally I think it is inhumane and idiotic.

It's not going to save that much money, is it? Not over the long term. I guess we can get our jollies by selectively torturing the poor elderly, but I don't think we are going to save money.

I think the real idea is to pretend to save money? So we're going to take it out of some 80 year old's hide?

What the hell is this country becoming?

[audio_inside]

3. Slightly over-dramatized...

Won't quibble with the exact figures you have cited because I can't see all the math, except to point out that the BLS release you linked to gives figures for CPI-U, and you have (correctly) referenced CPI-W in your example. So I don't know if you had actually found the relevant CPI-W figures or resulting SSI COLA values elsewhere (eg here) or if you are incorrectly substituting CPI-U for CPI-W values.

I would take issue with a couple of your qualitative conclusions, however:

To put it in current dollar terms, it is as if we are taking that $1,000 monthly benefit check and cutting it to $900 for older retirees.

I don't think the situation is quite as dramatic as you've expressed it; really, "it is as if" we are decreasing COLAs by $10 or <1% more out of the monthly benefit each year for 10 successive years.

Consumer units over 75 years...<have a food expenditure of> $230 a month. Thus the estimated 20 year gap of $170 dollars is a significant fraction of expected food expenses for an elderly retiree.

Here you're comparing a senior's food budget today against a projected benefit reduction 20 years from now. An realistic comparison would use the expected food expenses for an elderly retiree projected forward 20 years( eg $340, assuming a 2% C-CPI-U) when that $170 is going to seem like a less dramatic though obviously still significant fractional cut.

[Yo_Mama]

4. In answer

The first link is to a table giving CPI-U & C-CPI-U, so certainly can check the math.

The second link is to the SS COLAs, which are figured off CPI-W, but earlier than the December version. So you can get all the figures used from the links, which you did not read.

You cannot use the C-CPI-U to figure retiree food expenses 20 years in the future, because food expenses, especially for basic foods, are rising much faster than the overall consumer price index. And they will continue to do so, because energy costs are rising and food expenses are very dependent on energy costs over the long haul. Further, basic foods are Giffen goods
http://en.wikipedia.org/wiki/Giffen_good

In other words, the cost of basic foods are not constrained by the normal law of supply and demand. Any retiree living on SS can explain this principle quite eloquently.

In no way did I exaggerate the effect. Based on the 10 available years of history, after 20 years the retiree would be about 10% short in income. That does indeed equate to about $100 out of a $1,000 monthly check. CPI-W (please go to the link and click the companion file for CPI-W) currently counts food as 14.9% of the monthly consumer expenses. Out of a $1,000 monthly check, it would of course be higher - more like 19%. But let's use the 14.9%. If a retiree were spending $149 each month on food out of their monthly check, and the monthly check were cut by $100, what would you call that? That's the current-dollar comparison.

Of course the retiree is not going to survive by cutting monthly food expenses to $49.

I underdramatized it. Indeed the problem is well understood, which is why one version of this proposal wants to add back a 5% increase for an 80 year old or something like that.

http://www.bls.gov/web/cpi.supp.toc.htm

The entire point of this proposal is that it is supposed to save money later, but it won't save that much and it will literally torture the vulnerable. Instead we could merely raise taxes a bit and really count on getting the money.

The moment you get into cutting SS, you deal with the problem of the mass of poorer elderly who promptly become eligible for other social programs. SS is a subsistence level program, not a living-wage type program.