I know there are several threads about the Social Security report below, but they’re getting long so I’ll my question here. In the Economic Assumptions page, it says “Actual economic data were available through the third quarter of 2014 at the time the Trustees set the assumptions for this report.” (Why does the report use data that is 9 months old?). Anyway, the projected CPI in the table on that page for 2015 is 0.2%, which I wouldn’t think anyone would have predicted based on data through the Q3 2014. The Program-Specific Assumptions page says “More recent volatility in oil prices has again affected the CPI.” This occurred in the 4th quarter of 2014 (and Jan. 2015), so they obviously took this into account. But this drop in CPI affected the 2014 CPI, which was 0.3% (CPI-W from Dec. 2013-2014), while Table V.B1 gives it as 1.75% (so this was probably estimated with only data through Q3 2014).
So I guess they used data through Q3 2014 for Table V.B1, and had some reason for the 0.2% CPI for 2015 based on that data, but that they used more recent CPI data for Table V.C1, which predicts the COLA (and it is that page that mentions the volatility in oil prices). Also, maybe this is the reason the reports have always been so late – they might be waiting for recent CPI data to have a better idea of what the COLA will be. Anyway, if they’re using different data for different tables, I think they should mention that.
This won’t affect the long-term projections, of course, but I’m also interested in the short-term (particularly the AWI series, since that determines benefits).
Interesting development in the stock market yesterday. For the first time ever the value of Amazon’s shares (market capitalization) exceeded that of Walmart. Ten years ago Amazon was a fraction of Walmart.
Walmart has 2.2m employees (largest in America) while Amazon has only 154,000. Amazon’s revenue per employee is 2.6Xs that of Walmart’s employees.
Amazon and Walmart have very different business models. This is one measure of the rise of the robots.
Does this matter? In the short run it doesn’t, but in ten years the number of big box stores (and the number of employees) will probably fall.
not a good reason to delay a report due in April in order to include data from June or whenever. the report should be based on the data available at the time the report was made(or was due. otherwise why not wait until the 2016 data is available?)
as far as i know the SS adjustment for cost of living is made after the fact. so never mind the prediction.
Mike B. The process by which the OACT actually produces the Report is fairly murky, but I am in contact with an economist who used to head up the part of the Commissioner’s office that oversaw it and he has given me some hints.
One it would appear that the Report is produced in phases: one the straight number crunching and actuarial work that produces the Tables, then some interpretive layering of what those numbers mean, and finally a review process by all the Trustees’ staff. Which since the Trustees include the Secretaries of Treasury, Labor and HHS means whole groups of folks with their own technical expertise. How and to what extent this feedback gets incorporated into the final Report is what is murkiest about the whole operation, in part because it raises the spectre of politics being injected via what are after all Presidential Cabinet members with vested interests and a desire to promote their bosses agendas.
Also by law the Report is due out by April 1. Not that this deadline has been meet over the last seven years, but it does mean some pressure to get the analytical work done early enough to allow the multiple reviews necessary. Which I think largely explains your “why nine months?” question. It’s not supposed to be nine months.
Which all maybe hints at the differences between V.B1 which is designed to project ultimate (out year) numbers as opposed to V.C1 which on your account (I haven’t reviewed it) which is designed to guide current year CPI. This may be particularly true because SSA has been slammed hard in recent years over COLA dating back to a very unfortunate incident in 2008 (I think). What happened there is that a spike in gas prices that was projected to be persistent through the next year and so jacked up the CPI projection turned out to be transient. But by then the COLA figure had been set and turned out to be several points above the actual CPI. Now Social Security has a hold harmless provision so even though the effect was retirees getting too much COLA by the numbers there was no clawback or reduction. Instead the new benefit level was just allowed to stay in place until actual CPI caught up. Which took two years but which also meant no COLA increases. On paper all retirees had actually gotten their inflation adjustment early but in the real world it meant them seeing two years of price increases while having no COLA at all. The whole thing was a mess and a big black eye and it doesn’t surprise me that there is an effort to use as up to date numbers for the purpose of setting COLA even if that means some divergences from the set of Tables used to project ultimate numbers.
But as I say the whole process is less than transparent and all we can do is exactly what you are doing, which is to pick at the numbers and highlight discrepancies even if no actual explanation will be forth coming from OACT or the Trustees.
Thanks for the replies. Coberly – You are right that the COLA is set by data after the fact (it is announced the day in October that the CPI for September is announced, since that is the last value needed for the COLA), so the projected COLA doesn’t really matter. However, the news stories I read about the Trustees Report mentioned the projected COLA, the Trust Fund exhaustion years, and that’s about all. So it is a number that the press picks up on (I’m actually more interested in the projected AWI, and what data went into that, but I have no way to judge that, unlike the COLA).
Bruce – Thanks for the comments about the process. You misunderstood what I wrote about V.C1. It is a table giving the values of COLA, AWI (average wages), and the cap for past years and projections for future years. So there is a current-year COLA prediction in there, which as I wrote above is one of the few numbers I saw reported. However, the COLA is based on changes in the actual data (this year’s COLA will be the change in average CPI-W from the 3rd quarter of 2014 to 2015, or 0 if this is negative), so the large COLA announced in 2008 was not based on a projection (the projection was actually lower), but a spike in the CPI data. The CPI then went down, so there was deflation and no COLA in 2009. In 2010, there was inflation but no COLA since the CPI hadn’t recovered to the 2008 Q3 level (which was the baseline for 2010 and 2011). So, yes, retirees had actually gotten more than they should have based on the CPI numbers but they were upset anyway about the 2 no-COLA years. I think those 2 years are what probably accounts for the heightened interest in the COLA – before that, people just assumed there would always be something. (No-COLA years also affect Medicare, since they can’t increase Part B premiums for current, non-high-income recipients if there is no COLA, so next year there is likely to be a steep Part B increase for new recipients and high-earners.)
Mike thanks for clearing up some confusion for us.
I do have to wonder if they had known that the gas price hike that drove the 2008 Q3 CPI was going to be transitory whether they could have kept the COLA down. In general I like decisions to be driven by data and limiting discretion – because politics but this was a case where some foreknowledge would have allowed some policy smoothing that would have saved a lot of grief.
On an anecdotal level. A lot of my friends/drinking buddies were already collecting Social Security at the time and after a few attempts I just gave up explaining what happened and started agreeing “Man you guys got screwed! Damn Government!” Because nobody wanted to hear that in reality there were dollars ahead. Because it was not like those extra dollars in 2009 put anybody on easy street, spending had already adjusted to account for the small amount of “extra” dollars and those two years of no new COLA really hit hard psychologically.
Plus as you note in passing even when COLAs started again lower income retirees got smacked with a big incremental change in Part B premiums that gobbled up most or all of that year three COLA.
Any thanks again for the explanation of the mechanics.
Rjs – On the first story you link to, the amendment to the highway bill to strip Social Security for people with outstanding warrants has been removed from the bill.
People why not solve many of the governments problems and the need of ever increasing taxes on everything the public needs or owns.
End fractional reserve banking and set up public banks. This still lets Wall Street and all others do with there funds what ever they will in the private market.
I know there are several threads about the Social Security report below, but they’re getting long so I’ll my question here. In the Economic Assumptions page, it says “Actual economic data were available through the third quarter of 2014 at the time the Trustees set the assumptions for this report.” (Why does the report use data that is 9 months old?). Anyway, the projected CPI in the table on that page for 2015 is 0.2%, which I wouldn’t think anyone would have predicted based on data through the Q3 2014. The Program-Specific Assumptions page says “More recent volatility in oil prices has again affected the CPI.” This occurred in the 4th quarter of 2014 (and Jan. 2015), so they obviously took this into account. But this drop in CPI affected the 2014 CPI, which was 0.3% (CPI-W from Dec. 2013-2014), while Table V.B1 gives it as 1.75% (so this was probably estimated with only data through Q3 2014).
So I guess they used data through Q3 2014 for Table V.B1, and had some reason for the 0.2% CPI for 2015 based on that data, but that they used more recent CPI data for Table V.C1, which predicts the COLA (and it is that page that mentions the volatility in oil prices). Also, maybe this is the reason the reports have always been so late – they might be waiting for recent CPI data to have a better idea of what the COLA will be. Anyway, if they’re using different data for different tables, I think they should mention that.
This won’t affect the long-term projections, of course, but I’m also interested in the short-term (particularly the AWI series, since that determines benefits).
Interesting development in the stock market yesterday. For the first time ever the value of Amazon’s shares (market capitalization) exceeded that of Walmart. Ten years ago Amazon was a fraction of Walmart.
Walmart has 2.2m employees (largest in America) while Amazon has only 154,000. Amazon’s revenue per employee is 2.6Xs that of Walmart’s employees.
Amazon and Walmart have very different business models. This is one measure of the rise of the robots.
Does this matter? In the short run it doesn’t, but in ten years the number of big box stores (and the number of employees) will probably fall.
Mike B
i offer my non expert thoughts:
it doesn’t matter.
not a good reason to delay a report due in April in order to include data from June or whenever. the report should be based on the data available at the time the report was made(or was due. otherwise why not wait until the 2016 data is available?)
as far as i know the SS adjustment for cost of living is made after the fact. so never mind the prediction.
Mike B. The process by which the OACT actually produces the Report is fairly murky, but I am in contact with an economist who used to head up the part of the Commissioner’s office that oversaw it and he has given me some hints.
One it would appear that the Report is produced in phases: one the straight number crunching and actuarial work that produces the Tables, then some interpretive layering of what those numbers mean, and finally a review process by all the Trustees’ staff. Which since the Trustees include the Secretaries of Treasury, Labor and HHS means whole groups of folks with their own technical expertise. How and to what extent this feedback gets incorporated into the final Report is what is murkiest about the whole operation, in part because it raises the spectre of politics being injected via what are after all Presidential Cabinet members with vested interests and a desire to promote their bosses agendas.
Also by law the Report is due out by April 1. Not that this deadline has been meet over the last seven years, but it does mean some pressure to get the analytical work done early enough to allow the multiple reviews necessary. Which I think largely explains your “why nine months?” question. It’s not supposed to be nine months.
Which all maybe hints at the differences between V.B1 which is designed to project ultimate (out year) numbers as opposed to V.C1 which on your account (I haven’t reviewed it) which is designed to guide current year CPI. This may be particularly true because SSA has been slammed hard in recent years over COLA dating back to a very unfortunate incident in 2008 (I think). What happened there is that a spike in gas prices that was projected to be persistent through the next year and so jacked up the CPI projection turned out to be transient. But by then the COLA figure had been set and turned out to be several points above the actual CPI. Now Social Security has a hold harmless provision so even though the effect was retirees getting too much COLA by the numbers there was no clawback or reduction. Instead the new benefit level was just allowed to stay in place until actual CPI caught up. Which took two years but which also meant no COLA increases. On paper all retirees had actually gotten their inflation adjustment early but in the real world it meant them seeing two years of price increases while having no COLA at all. The whole thing was a mess and a big black eye and it doesn’t surprise me that there is an effort to use as up to date numbers for the purpose of setting COLA even if that means some divergences from the set of Tables used to project ultimate numbers.
But as I say the whole process is less than transparent and all we can do is exactly what you are doing, which is to pick at the numbers and highlight discrepancies even if no actual explanation will be forth coming from OACT or the Trustees.
Thanks for the replies. Coberly – You are right that the COLA is set by data after the fact (it is announced the day in October that the CPI for September is announced, since that is the last value needed for the COLA), so the projected COLA doesn’t really matter. However, the news stories I read about the Trustees Report mentioned the projected COLA, the Trust Fund exhaustion years, and that’s about all. So it is a number that the press picks up on (I’m actually more interested in the projected AWI, and what data went into that, but I have no way to judge that, unlike the COLA).
Bruce – Thanks for the comments about the process. You misunderstood what I wrote about V.C1. It is a table giving the values of COLA, AWI (average wages), and the cap for past years and projections for future years. So there is a current-year COLA prediction in there, which as I wrote above is one of the few numbers I saw reported. However, the COLA is based on changes in the actual data (this year’s COLA will be the change in average CPI-W from the 3rd quarter of 2014 to 2015, or 0 if this is negative), so the large COLA announced in 2008 was not based on a projection (the projection was actually lower), but a spike in the CPI data. The CPI then went down, so there was deflation and no COLA in 2009. In 2010, there was inflation but no COLA since the CPI hadn’t recovered to the 2008 Q3 level (which was the baseline for 2010 and 2011). So, yes, retirees had actually gotten more than they should have based on the CPI numbers but they were upset anyway about the 2 no-COLA years. I think those 2 years are what probably accounts for the heightened interest in the COLA – before that, people just assumed there would always be something. (No-COLA years also affect Medicare, since they can’t increase Part B premiums for current, non-high-income recipients if there is no COLA, so next year there is likely to be a steep Part B increase for new recipients and high-earners.)
Mike thanks for clearing up some confusion for us.
I do have to wonder if they had known that the gas price hike that drove the 2008 Q3 CPI was going to be transitory whether they could have kept the COLA down. In general I like decisions to be driven by data and limiting discretion – because politics but this was a case where some foreknowledge would have allowed some policy smoothing that would have saved a lot of grief.
On an anecdotal level. A lot of my friends/drinking buddies were already collecting Social Security at the time and after a few attempts I just gave up explaining what happened and started agreeing “Man you guys got screwed! Damn Government!” Because nobody wanted to hear that in reality there were dollars ahead. Because it was not like those extra dollars in 2009 put anybody on easy street, spending had already adjusted to account for the small amount of “extra” dollars and those two years of no new COLA really hit hard psychologically.
Plus as you note in passing even when COLAs started again lower income retirees got smacked with a big incremental change in Part B premiums that gobbled up most or all of that year three COLA.
Any thanks again for the explanation of the mechanics.
on social security: Senate Bill May Kick 200,000 Off Social Security If They Have Arrest Warrant
here’s what i came over here with:
2016 presidential election: How do your beliefs align with the potential candidates? — Take a quiz here and find out.
i’ll try pasting the text of my results:
Candidates you side with…
95%
Bernie Sanders Democrat
on environmental, healthcare, foreign policy, education, economic, immigration, and domestic policy issues.
compare answers
58%
Hillary Clinton Democrat
on healthcare, foreign policy, and domestic policy issues.
compare answers
52%
Chris Christie Republican
on education issues.
compare answers
37%
Martin O’Malley Democrat
on healthcare, immigration, and domestic policy issues.
compare answers
31%
Rand Paul Republican
on foreign policy and education issues.
compare answers
25%
Mike Huckabee Republican
on education issues.
compare answers
20%
Bobby Jindal Republican
on education issues.
compare answers
18%
Donald Trump Republican
on education issues.
compare answers
15%
Rick Perry Republican
on education issues.
compare answers
15%
Ben Carson Republican
on foreign policy and education issues.
compare answers
10%
John Kasich Republican
no major issues.
compare answers
10%
Marco Rubio Republican
on education issues.
compare answers
7%
Scott Walker Republican
on education issues.
compare answers
7%
Lindsey Graham Republican
on education issues.
compare answers
7%
Carly Fiorina Republican
on education issues.
compare answers
5%
Rick Santorum Republican
on education issues.
compare answers
3%
Ted Cruz Republican
on education issues.
compare answers
1%
Jeb Bush Republican
no major issues.
compare answers
Rjs – On the first story you link to, the amendment to the highway bill to strip Social Security for people with outstanding warrants has been removed from the bill.
People why not solve many of the governments problems and the need of ever increasing taxes on everything the public needs or owns.
End fractional reserve banking and set up public banks. This still lets Wall Street and all others do with there funds what ever they will in the private market.
Good piece on Naked Capitalism today about Billionaire hedgie John Arnold’s scheme to privatize retirement savings via GRAs… http://www.nakedcapitalism.com/2015/07/will-hillary-clinton-adopt-hedgie-billionaire-john-arnolds-schemes-for-retirement-insecurity.html
More evidence that mainstream Dems like Madame Secretary are probably more dangerous to retirement security than the GOP. In case you needed it.
am soc
thanks. i probably didn’t need it. but it helps to keep me a worried man.
Krasting
in case you see this, Arne posted a spreadsheet of his own that appears to confirm mine. Perhaps he’ll take your bet.
On the other thread.