Dr. Black Did It as a Shorter; Here’s the Data
The problem with waiting overnight to post is that other people figure out the same thing:
If the true CPI-E increases faster than CPI, then chained-CPI is worse, not better.
It’s actually worse than that. The measure by which Social Security is raised turns out not to be what we usually call CPI (Consumer Price Index for All Urban Consumers), but CPI-W (Consumer Price Index for Urban Wage Earners and Clerical Workers):
What George W. Bush referred to as “the miracle of compound interest” works both ways. Seniors see their Purchasing Power decline every year. So when Jared Bernstein, Sensible Centrist, says:
I support the change [to Chained CPI]—it’s a more accurate measure of price growth (though a chained index for the elderly would be better), and I’m sure it’s coming, so I want to get something for it. That ‘something’ is an offset from the benefit cut for poor, old elderly.
It would be nice to think he was Ernest Lee Sincere. But we know Jared Bernstein is not innumerate, so we have to assume he’s acting from malice aforethought, since this took me less than 20 minutes to put together from scratch, including the normalization:
The reason it took me twenty minutes: ten of those were spent trying—unsuccessfully—to find post-2007 CPI-E data. But unless Mister Bernstein and his cohorts are declaring that costs for the Elderly were actually deflationary from 2008 forward (at which point they would be correctly laughed out of polite society and relegated to Beltway Conversations…oh, wait…), note that CPI-W (the green line and, as noted above, the current, already substandard, measure) is below (that is, less than) the CPI-E line and above (that is, greater than) the Chained CPI line.
It takes three years–until 2010—for CPI-W to reach the 2007 actual CPI-E level. It takes four years, to 2011, for Chained CPI—the “more accurate measure,” per Mr. Bernstein’s blog post—to get to that level.
The man who says he “I want to get…an offset from the benefit cut for poor, old elderly” in exchange for going to the malicious Chained CPI is basically saying, “I’m going to create a larger and larger group of poor, old elderly in the future in exchange for a couple bowls of gruel now.”
We’ve seen other people make this mistake (most notably Victor Matheson—who, G-d help us, teaches economics at Holy Cross—in comments chez DeLong), but rarely are we so clearly reminded that not only is Barack Obama a poor negotiator, but (again: see Summers, Geithner) the people he had and has negotiating for him are venal.
“(at which point they would be correctly laughed out of polite society and relegated to Beltway Conversations…oh, wait…)”
lol when i read that, great line. screw bernstein
Bernstein’s comments and positions are more nuanced that this summation would have it. In fact I spent a good part of the last day debating the point with some major social security wonks/political blogging heavy hitters on a mail list I belong to.
One point (which I expand on in a post at my rejuvenated Social Security Defender blog) is that the current law scheduled initial benefit projects to rise at a real rate in advance of all of CPI-U, CPI-W, Chained CPI, and even CPI-E.
Note that I said ‘initial benefits’, to the extent that Chained-CPI actually trails a fully formed CPI-E (because the ‘E’ stands for ‘experimental’ and not ‘elderly’, BLS not having been funded to create a full index) that higher level of real initial benefits will begin to erode against the baseline. But where that benefit line would cross a theoretical fix that froze initial benefits in real terms at 2012 levels is not as easy a slam dunk question as some might think.
The truth is that under current law and current economic assumptions there is a discontinuity between ‘scheduled’ and ‘payable’ benefits. Now some of us, including the authors of the Northwest Plan would close that discontinuity totally on the revenue side and totally out of the pockets of workers. Others would close the gap on the revenue side by lifting caps and extending FICA to unearned income. And by all means we should push for that.
But the simple arithmetic reality that if you don’t backfill the entire gap with new revenue you can only avoid the sudden benefit discontinuity at Trust Fund Depletion by adjusting either the initial or continuing benefit formula. And there are solid reasons to prefer that the Hobson’s Choice of a ‘fix’ in that circumstance falls on the side of the continuing benefit formula than the initial benefit formula.
The the line of argumentation to support those reasons is a little complicated. And it is a lot easier to just conclude “Screw Bernstein”.
So I guess Jared will just have to take care of himself. At least today. Because football. Or something.
Moot? Huffington Post today: “UPDATE: 4:30 p.m. — Republican senators leaving a GOP conference meeting Sunday afternoon told reporters that they are dropping chained CPI from their fiscal cliff proposal.” http://www.huffingtonpost.com/2012/12/30/fiscal-cliff_n_2384726.html
And once again we may never know whether Obama wants it. I’m guessing Reid, who opposes the CPI-E for Social Security benefits, “explained” to McConnell behind closed doors that the Dems would pin the CPI-E prominently on McConnell’s chest, along with those of other GOP members of Congress who are up for reelection in 2014. Thus, Reid might say heads you lose, tails you lose, and you get to decide. When it’s engaging its brains a little bit, the GOP prizes winning elections over winning a minor policy objective.
I keep hearing two numbers, 40 cents per week and 80 cents per week. I assume these numbers are really two ways of saying the same thing — 40 cents per employee or 80 cents combining employee and employer contributions.
Correction requested if I am wrong.
“because the ‘E’ stands for ‘experimental’ and not ‘elderly'”
But the BLS actually says “Experimental Price Index for the Elderly (CPI-E)” which indicates that the “E” stands for elderly and that the index is experimental. At the very least you need to indicate that the “E” could mean either (although where it stands in the acronym strongly says “Elderly”); whenever someone says it stands of “Experimental” I have to think I’m reading someone who either doesn’t know what they’re talking about, has been snowed on the issue, or is deliberately trying to mislead people.
Over the seventy five year actuarial window it would average forty cents (today’s money) per week each year for the combined employer plus employee tax.
In order to avoid the “shock” in 2033 or so of a two percent raise (each) to avoid benefit cuts, it would be better to raise the tax eighty cents per week for each the employer and the employee each year until then.
or better still, just raise the tax eighty cents per week, on each, whenever the Trustees project short term actuarial insolvency… that is, an actual likely problem over a time period that means something.
i hope you (andothers) can see that this is not a material difference… that is, they wouldn’t feel it in any case.
but rarely are we so clearly reminded that not only is Barack Obama a poor negotiator
He’s actually very good, once you accept the fact that he’s not on our side.
Look how much he’s done for the big banks.
I hate to see liberals discussing the “nuances” of the chained CPI.
There are no nuances. It is simply a dishonest way to cut benefits for people who paid for their own benefits.
In any case it is NOT a “more accurate” measure of consumer price increases.
If the price of steak goes up, the price of steak goes up whether you substitute chicken or not.
What the chained CPI MIGHT be is a measure of “the cost of living,” as in staying alive. If you can’t afford steak, you will switch to chicken, and if you can’t afford chicken you will switch to cat food, and if you can’t afford cat food you will switch to turnips.
The motive of our leaders is to reduce us to a condition of barely surviving. All else is an “improvement in the standard of living” and we can’t have that.
Even if we pay for it ourselves.
You can substitute chicken for steak, I suppose, but what can you substitute for health care?
And that’s the bottom line. You have something that has become a lot more expensive (not because of medical sophistication, but because we are simply being charged more) that you really really really need. That’s the point for the elderly – they have a greater and more frequent need for this increasingly expensive thing as they get older. And they can’t just buy chicken instead.
I still don’t understand the compounding. How do I substitute chicken for beef each and every year? This must be a stupid question because no one ever helps me out with it. (If so, please respond with why it is stupid. I’m not sensitive.)
Anna, you substitute chicken for beef the same way you substitute a Chevy for a Buick. You use a formula assigning a predetermined value for the chevy, another one for the buick, and then plug in the presumed value of a bunch of other food and consumer items, etc. until you crank out a number which you then compare to the price of similar etc etc. In short, it isn’t really possible. It’s just one way of figuring the change in the cost of living from one year to the next. This “chained CPI” does not include the out of pocket cost of medical treatment, copays, uncovered items and so on. The main purpose of using this index is to reduce SS payments although it also affects other indexed figures like tax brackets. Why cut SS benefits when they are already inadequate? SS is the target of many conservative groups and has been for 77 years. That’s why. If the chained CPI is passed, GDP is reduced. Beats me how this is good. NancyO
Nancy, Well yea, that’s what I kind of already knew. What I am asking has to do with (probably) the way the calculation is done. I can’t figure out how compounding a human response to inflation is compounded in such a way that I only change from beef to chicken once and from chicken to chicken bones once and from chicken bones to sea gulls on my shed (bones to bullets) once. It could have something to do with the baseline index but I wanted someone to tell me or, preferably, show me that my question is either relevant or irrelevant. I think it would be one or the other and not somewhere in between. Maybe the question is so, so stupid that no one understands it (or because no idiot with half a brain would ask it).
Anna–“Maybe the question is so, so stupid that no one understands it….”
It’s not a stupid question. The chained CPI is a pretext to cut SS benefits. Period. You can’t do what they say this CPI does–nobody really knows what happens when the price of beef goes up when no one can afford it anyway.
It’s actually a very smart question. It shows that you understand that cutting SS benefits is unnecessary, cruel, and just plain stupid. But, the guys who work in this administration have decided to cut benefits just to be sure____(your favorite reason here.) Mine us because. NancyO
If there’s a convincing argument that the chained CPI is a more accurate measure of inflation, I can’t find it. Reading the report that apparently “started it all,” which I found at http://www.ssa.gov/history/reports/boskinrpt.html , I conclude that the C-CPI probably answers the wrong question, consequently measures the wrong concept, and conflates inflation with behavioral responses to inflation. But I’m all ears.
Meanwhile, the validity question isn’t relevant to our politicians. Conservatives want to cut Social Security benefits with this tool (or any tool!). Obama is willing to make the concession if he gets enough in trade and we apply the new metric across the board to provide a fig leaf for claiming that we’re simply moving to a better measure. Only the staunchest defenders of Social Security adamantly oppose C-CPI–let’s hope they’re lucky.
I am not sure I understand your question, but that does not make either of us stupid.
see if this works:
if the price of everything is going up, folks whose income is not going up will have to buy cheaper stuff…or maybe just less stuff.
the geniuses behind the chained cpi figure this means the cost of living has not gone up, because people are still living even if they have to cut back.
this can go on for some time… after all it’s only about ten percent after thirty years and by then most of them will be dead anyway.
or to put it another way… if you are living on one dollar more than you need…happiness. if you are trying to live on one dollar less than you need… well, you stop living. our leaders are trying to fine tune just where that point is. or where those of us still alive will seriously complain about it.
but note this has NOTHING to do with “inflation” or “consumer prices.” it’s about how much pain can we cause the poorer half of the working class.
because that makes us feel better about our own taxes. because, you see, taxes are theft. money is time and time is life. so taxes are killing us. and the only thing that can make us feel better is to kill someone else. or watch them suffer. yes, suffer is better.
i forgot to say, the workers could avoid the chained cpi benefit cut by just raising their own tax 13 cents per week each year.
(that’s one third of the forty cents it would take to solve the whole “Social Security crisis” altogether.)
but no one wants the workers to know that.
Thanks for all your effort Nancy and coberly.
Hey coberly, now don’t mad about this question. You know I’m with you on the Northwest Plan.
I was just thinking about all the “defenders” that think eliminating the cap is the answer. Now an orchestra started playing in my head about the trust fund and the cap. Has anyone ever produced a model for slowing the reduction of the trust fund while floating the cap to make up the difference? Or would this require too high a cap to even bother with?
can’t see why I should get mad.
The cap could be raised to about 150k. This would take care of one third of the actuarial insolvency (that is replace about 13 cents of the needed 40 cents per week increase in the tax to make SS “actuarially solvent” for the full 75 years.
The folks whose tax would be raised above the present cap, would get a benefit increase that would approximately equal the amount of the tax, plus enough interest to cover inflation and average growth in real wages. So not necessarily a bad deal for them. Depends if they see a guaranteed income of about 35k in retirement worth what they “might have gotten” if they had invested the extra 6k payroll tax on the market, which of course, they might have lost all of.
I don’t think it’s worth messing with the cap right now, certainly not behind a slogan like “scrap the cap” which just scares and angers the “rich” people who will have more to say about how SS is fixed than any of the people chanting scrap the cap.
The Trust Fund would be affected by all of this… but try to trust me: the Trust Fund is NOT IMPORTANT. Some day it will go back to “zero” (really a one year reserve as designed), and SS will go back to being fully (almost) pay as you go. And that will be better. Even if the Trust Fund were stolen by flying saucers today, it would not make a material difference to Social Security or be an injustice to anyone.
“Eliminate the cap” is the surest short way I know of to make sure they cut benefits instead. Or they might take the eliminate the cap hit for a few years to make it easy for them to implement means testing and practical “final solution” of the Social Security problem forever.
did i say that Jared Bernstein, who says
“I support the change [to Chained CPI]—it’s a more accurate measure of price growth “
shows thereby that he is an idiot.
the chained CPI is not a more accurate measure of price growth. it might be an “accurate” measure of people trying to live by switching to cheaper, less desired, products. but that is not a measure of price growth, accurate or not. it is simply a lie.
a better answer to your question
the chained CPI is supposed to run about 3 tenths of one percent lower than the current CPI.
This would mean that a person getting a thousand dollars a month SS check would get about three dollars less per month in “cost of living adjustment” each year. this doesn’t sound like much.
but on a thousand dollars a month you are probably not eating much steak, and three dollars less a month would mean maybe one less pound of chicken per month… maybe two or three dinners.
okay, you don’t need meet every day.
but next year it’s another three dollars… now a total of six dollars a month… so now four of five meatless dinners per month.
or maybe it’s six times 12… 72 dollars a year. could be an electric bill you can’t pay,
and after ten years it’s 30 dollars a month… now about twenty meatless dinners per month. or 360 dollars per year. begins to look like some medical care might get deferred, or just do without heat all winter.
or, here’s an idea… just buy a pellet stove. you can cut your heat bill, it says in the store, and the economists all agree. but of course you have to have the thousand bucks to buy the stove, and that’s hard to find on a thousand dollar a month Social Security check.
so you see it wasn’t just a question of a one time switch from steak to chicken. you weren’t eating steak in the first place. and after that it’s just a gradual giving up of chicken. beans and rice works okay though.
just don’t let the economists find out. when they learn you can live on beans and rice, they’ll cut your benefit check. [they’ll say] no sense paying for chicken if you can live without it. what do you think “cost of living” means anyway?
coberly, Yes, that is kind of like I am visualizing it in my head and why I brought up the question of compounding changes in behavior and what on earth that really meant. The insanity made me think that I wasn’t thinking right.
Oh well, I guess I’ll let it lie until it comes up again in a month or two.
There is definitely a “holier than thou” attitude in the postings here regarding “true inflation.”
It is quite true that conservatives want to cuts benefits but want to disguise their cuts with air of economic science so they advocate the use of a CPI measure that will generate a smaller COLA adjustment.
On the other hand, the liberals on this want more generous benefits for the elderly but want to cloak their request in an air of economic science so they call their increases a COLA and advocate the use of a CPI measure that will generate the largest possible COLA adjustment regardless of “truth”. Everyone here is trashing Bernstein and economic research that would make the CPI a better measure of changes in consumer welfare unless the research results in an larger COLA adjustment (like may happen with the CPI-E) in which case changes in the CPI are embraced.
The lack of understanding regarding the substitution effect or the occasional outright rejection that it exists, not to mention the other major factors (the new good effect and the quality effect) that economists identify as biases in the CPI, is not what we should expect from a reality based community.
And a follow up– I guess that now that I have been slimed by Ken, I need to do a bit of work on my own. Not sure if Ken is a stand up guy or not, so I don’t know if a detailed response is worth my time, but here goes.
The chained CPI (C-CPI-U) is designed to correct for flaws in the regular CPI-U (or CPI-W) measure that occur when consumers can switch to cheaper goods with no loss of utility. Estimates from the BLS in the late 90s (a bit dated, I realize) suggested that the regular CPI overstates the effect of inflation on consumer welfare by 0.4% per year. CPI-E is a measure designed to track changes in price for expenditures by persons over 62. Between 1982 and 2011, CPI-E grew roughly 0.2% faster per year than CPI-U or CPI-W. Thus, the elderly were undercompensated due to the use of CPI-W instead of CPI-E by about 0.2% per year but overcompensated by about 0.4% per year due to the CPI-W failing to account for substitution effects leaving them about 0.2% overcompensated.
A move to C-CPI-U would cut 0.3% or so from COLAs now leaving the elderly 0.1% undercompensated for the effects of inflation. So, the move to chained CPI clearly leaves the elderly worse off compared to the unchained CPI, but gets them closer in absolute terms to the COLA needed to make them exactly as well off after inflation. If you want your SS COLA to make the elderly absolutely better off each year, then CPI-E is your ticket. If you want your SS COLA to be welfare neutral, then C-CPI-U is probably the closest of the 4 measures over the 1982-2011 time period.
Mind you, the story is quite different if we concentrate only on the time period from 2006-2011, the latest data. To make the elderly exactly as well off as before the inflation that occurred over this period we need to use a CPI-E adjusted for substitution effects which is not an index that is currently available. CPI-W has gone up 2.4% per year, CPI-U has gone up 2.3% per year, CPI-E has gone up 2.3% per year, and C-CPI-U has gone up 2.0% per year. A chained CPI-E is likely to be roughly 0.3% lower than CPI-E (just like the difference between CPI-U and C-CPI-U) meaning a C-CPI-E would come in at around 2.0%, exactly what C-CPI-U has done.
Again, a move to chained CPI is clearly a cut in SS benefits compared to current policy and is a cut in well-being for the elderly compared to current policy. The question I asked, and was trashed for asking by Ken, is whether the use of chained CPI would be sufficient to maintain the welfare or well-being of SS recipients from year to year, neither making them better off or worse off over time. The answer appears to be that chained CPI would not quite have been sufficient to maintain the elderly’s well-being over the time period of 1982-2011 but would undercompensate them less than current policy now overcompensates them. From 2006-2011, chained CPI would essentially exactly compensate the elderly for inflation as opposed to CPI-W which has overcompensated them by about 0.4% per year.
Look, if you want to make the elderly better off, suggest an annual SS adjustment of 10% per year. Just don’t call it a Cost of Living Adjustment. It is a raise. If you want to use CPI measures like CPI-W that don’t account for the substitution effects, again that is fine, but it is a raise (albeit a small one) not a COLA. And the fact that we are not using a CPI-E is also bad economics. Bernstein is right suggesting a better COLA would be a chained CPI-E. If you want the elderly to get bigger increases than that, that is fine, but one shouldn’t hide behind an inaccurate inflation index to defend real increases in Social Security.
Victor — This is a bunch of crap. Nobody is talking seriously about moving the the CPI-E at this point, but simply are saying switching to a less benefit formula right now cannot reasonably be supported.
The CPI-U and W are bona fide attempts by BLS to track increases in the cost of living accurately. They are the primary official numbers. They are given greater predominance than the C-CPI-U. Where is the empirical evidence that C-CPI-U is more accurate, or actually more accurately considers substitution effect? You have no more idea whether C-CPI-U does it better or not. So far, BLS has not seemed fit to substitute Chained-CPI for the current official number. BLS is constantly refining its methodology. When BLS does make that substitution across the board, then we can have a civil conversation. In the absence of that, the C-CPI-U proposal is nothing more nor less than an attempt to reduce benefits.
Bruce Webb —
I’m not sure how “nuanced” a proposal to shift to C-CPI-U can be. Samne for a statement with no supporting argument whatsoever that it is “more accurate.”