Wednesday, September 3, 2014

Better no Social Security replacement rates than wrong Social Security replacement rates

My reply to Alicia Munnell’s NASI blog post, over at the AEIdeas blog.


Arne said...

In a way, Andrew and Alicia Munnel are talking past each other.

Removing the table from the report does remove meaningful trend information.

The annual report was changed to remove the reference to "replacement rate", but the annual report did not compare its "replacement rate" , which is in quotes to the "Most financial advisors say you’ll need about 70 percent of your pre-retirement earnings" within the report. The SSA does still make the comparison on its retirement planner web page (although Andrew's link is not quite to that page).

So the correctly qualified use of "replacement rate" in the report as been removed, but the not qualified comparison is still there.

In the July 15 piece Andrew says "When the calculations are consistent, the replacement rate paid by Social Security comes closer to 60%". It appears to me that Alicia is taking exception to this. Andrew's calculations are not consistent because you cannot get consistent numbers from aggregated data. Aggregate data is skewed by zero income years, which would not be used by financial advisors.

It is interesting that on a wage indexed basis, the highest earning years tend to be in the middle of a career rather than at the end.

Andrew G. Biggs said...

Arne, I'd have to think about some of what you said, but a couple quick thoughts: 1. Trend information does matter, though you'd also need to take into account life expectancies; theory says that as life spans rise your target replacement rate should decline. Probably not as fast as replacement rates at any given age are falling due to the higher NRA, but lower nevertheless. Another way to think about it would be to track the ratio of lifetime benefits to lifetime earnings, which won't decline as fast as replacement rates as any given age. 2. The 60% replacement rate I cited wasn't higher than the Trustees Report figures because of zero years; it comes from a paper by Schieber and Pang that doesn't count zero years. My 2008 paper at SSA did include zero years and found higher rates, around 69%. I don't know if Munnell realized this, as I don't believe her sources at SSA were aware of the Schieber-Pang paper until recently (they were aware in the sense of having been sent the paper, but I assume didn't link it to the WSJ op-ed).

WilliamLarsen said...

Just a bit of information on the OASI benefit formula of 1977. It uses the highest 35 year indexed years. Any more than 35 years work history may no increase the initial OASI benefit. Based on my own case, my Initial OASI benefi will no rise no matter if I work until full retirement age. My wage increases as mos in their 50's 60's are not keeping pace with the younger workers. This means my indexed wages are higher just a few years ago than current wages. My sister and brothers also have the same problem. We found that working past 58 does no change our initial OASI benefit because we already have maxed it out due to indexing.

Even a few years worth of zeros or low wage years really does not affect the initial OASI benefit.

The bend points and the replacement rates of 90%, 32% and 15% truly limit any actual reduction.

My guess is that as more older workers have left the work force, their overall initial OASI benefit has not suffered nearly as much as many might think.

However, for the theoretically inclined, it is a simple calculation for each cohort to determine maximum replacement.

Does any of this change Social Security's future financial trend, No?

WilliamLarsen said...

I am stuck on Compuer 13 at the library where the "t" key does not always work.

75seconds Smm said...

its really nice and informative blog .a person who always want to reveal the suspense can play it and enjoyed it a lot.
With the Secugo app installed on with your Smartphone, you are now part of a very unique community...
Phone: +32 477 65 21 65