May not be able to put old Humpty Dumpty back together again. I've been doing some serious thinking, prodded on by discussions with various other PERS "experts" over at the OPDG "bar" and helped along by some correspondence with some better-informed sources about how the City of Eugene settlement would be implemented if the Supreme Court upholds virtually all of the City of Eugene case, and under the known circumstances of the Strunk decision announced in March. The answer for actives and inactives has been clear for some time, but for retirees there are no definitive answers. In searching for an answer to this vexing question, I've been forced to read through a fair amount of documents that were relevant at the time HB 2003 was being discussed, at the time the "settlement" was circulating, and the various legal briefs filed in the Strunk case, the White case, and at least two other cases. I've reviewed PERS Board meeting minutes, reviewed various pieces of correspondence with informed sources and have reached a point where I *think* I have a sense of where this might be going. All of this work is guided by one overriding goal - to put together a calculator that will help retirees gauge the impact of a negative ruling in the City of Eugene appeal. A fair number of people entrusted me with their PERS annual statements for the period starting in 1999 so that I could have some test examples to work with.
In this post I'd like to explain what MY understanding of what MAY happen if City of Eugene is upheld by the Supreme Court. In a later post, I'll provide a more elaborate example of how I *think* it will work.
Bear with me. For this to make sense, we have to suspend all expectations of a miracle. Let's face potential reality for the next few minutes. If the topic is relevant to you, you'll undoubtedly find the news mixed at best. I can generalize slightly and say that the severest impact will be on "window" retirees who had *all* of their account balance in the Tier 1 regular, regardless of when you retired. Similarly, anyone who retired before the market began its rebound in 2003 and early 2004 *and* who had money in variable until the date of retirement will be losers. However, the combination of the later retirements (late 2003 - early 2004) may be the best off, unless you retired under the "lookback", in which case the news is quite mixed. As with everything PERS-related, there is NO simple, single answer for anyone. This doesn't mean I can't write a calculator; it just means that it will require an incredible amount of user-input directly from member statements spanning 1999 - 2003/4, and some solid grasp of the PERS earnings postings on their web site. The program can do all the hard calculations, but they are complicated. I now have a far better grasp of why it will take so long for programmers at PERS to unwind the 0% crediting, unwind the COLA freeze and recalculate benefits under the scenario envisioned by Lipscomb. It is an incredibly complicated set of calculations that involves the "lookback" for post-June 2003 retirees, old and new actuarial tables. In short, the entire process will involve essentially starting over with the 1999 earnings crediting and recalculate EVERYTHING that has happened since earnings were credited in 1999. Add to that the Lipscomb variable adjustment and the whole process is one giant programming mess.
Today I'm going to talk only about the simplest case - a July 1, 2003 retirement. This eliminates the "lookback" problem and requires me to explain this only in terms of the "old" mortality tables. It also eliminates the problem of unrolling the 2003 earnings crediting because July 1, 2003 retirees got 4% (approximately) for 6 months of 2003. They weren't subject to the 0% crediting. The Strunk ruling *only* applies in this case to the COLA freeze, while Lipscomb *only* applies to the 1999 earnings crediting decision. The Supreme Count only said that the COLA freeze was a breach of contract and couldn't be used as a means of "recovery". The court went out of its way to point out (see yesterday's post) that this did NOT prevent PERS from "recovery" using other statutory methods (i.e. ORS 238.715, although they did not mention that statute specifically).
With this in mind, imagine our hypothetical 7/1/03 retiree receiving a Notice of Entitlement that indicates a Money Match "winning" method and a monthly Option 1 benefit of $2500 per month, which has remained "fixed" due to the COLA freeze since then. Now further suppose that by refiguring the 1999 earnings credit at 11.33% (Lipscomb and settlement) the member's "true" retirement benefit after correcting for the "error" (assume Lipscomb upheld) should have been $2350 per month. That becomes the "fixed benefit" to which the COLA should be added. That member WOULD HAVE received 0.77% effective 7/1/03, another 1.17% effective 7/1/04, and 2% (expected) effective 7/1/05. By 7/1/05, the member's benefit should be $2443.71 ($6.29 LESS THAN the current benefit). So, in this very simple case PERS would adjust the current $2500 monthly benefit DOWN to $2443.71, and then seek recovery of $6.71 x 24 months = $161.04**, for the amount the member had been overpaid because of the "erroneous" crediting of 1999 regular earnings.
So, for members in the "window" period, the crucial detail will be the amount by which the current benefit would have been reduced had the 1999 credit been 11.33% instead of 20%. The larger the difference, the bigger the potential reduction in current benefit and the higher the amount of the payback. Just keep in mind that this approach is distinctly different than the interpretation typically offered for the Strunk decision vis-a-vis the COLA freeze. The methodology described above accomplishes TWO things, not just one. It collects for the present value of the "overcredit" as well as for the future value of the "overcredit". The standard (?mis)interpretation would have only recovered the present value of the overcredit, while continuing to leave the future value untouched. Unlikely that it would happen that way.
**If PERS really did this properly, the "payback" would be more than what is illustrated here. A correct calculation would adjust the first 12 months by more (because the differential is greater); the second 12 months by slightly less (because the differential is getting smaller). The 12 month period that would include the latest (not yet given) COLA would finally get the member down to the $6.71 differential, but that would be moot because the adjustment would have already taken place. In the end, the new benefit as of 7/1/05 would be only $6.71 less than the benefit determined at the actual time of retirement. And here you thought this would be a simple explanation.
In this post I'd like to explain what MY understanding of what MAY happen if City of Eugene is upheld by the Supreme Court. In a later post, I'll provide a more elaborate example of how I *think* it will work.
Bear with me. For this to make sense, we have to suspend all expectations of a miracle. Let's face potential reality for the next few minutes. If the topic is relevant to you, you'll undoubtedly find the news mixed at best. I can generalize slightly and say that the severest impact will be on "window" retirees who had *all* of their account balance in the Tier 1 regular, regardless of when you retired. Similarly, anyone who retired before the market began its rebound in 2003 and early 2004 *and* who had money in variable until the date of retirement will be losers. However, the combination of the later retirements (late 2003 - early 2004) may be the best off, unless you retired under the "lookback", in which case the news is quite mixed. As with everything PERS-related, there is NO simple, single answer for anyone. This doesn't mean I can't write a calculator; it just means that it will require an incredible amount of user-input directly from member statements spanning 1999 - 2003/4, and some solid grasp of the PERS earnings postings on their web site. The program can do all the hard calculations, but they are complicated. I now have a far better grasp of why it will take so long for programmers at PERS to unwind the 0% crediting, unwind the COLA freeze and recalculate benefits under the scenario envisioned by Lipscomb. It is an incredibly complicated set of calculations that involves the "lookback" for post-June 2003 retirees, old and new actuarial tables. In short, the entire process will involve essentially starting over with the 1999 earnings crediting and recalculate EVERYTHING that has happened since earnings were credited in 1999. Add to that the Lipscomb variable adjustment and the whole process is one giant programming mess.
Today I'm going to talk only about the simplest case - a July 1, 2003 retirement. This eliminates the "lookback" problem and requires me to explain this only in terms of the "old" mortality tables. It also eliminates the problem of unrolling the 2003 earnings crediting because July 1, 2003 retirees got 4% (approximately) for 6 months of 2003. They weren't subject to the 0% crediting. The Strunk ruling *only* applies in this case to the COLA freeze, while Lipscomb *only* applies to the 1999 earnings crediting decision. The Supreme Count only said that the COLA freeze was a breach of contract and couldn't be used as a means of "recovery". The court went out of its way to point out (see yesterday's post) that this did NOT prevent PERS from "recovery" using other statutory methods (i.e. ORS 238.715, although they did not mention that statute specifically).
With this in mind, imagine our hypothetical 7/1/03 retiree receiving a Notice of Entitlement that indicates a Money Match "winning" method and a monthly Option 1 benefit of $2500 per month, which has remained "fixed" due to the COLA freeze since then. Now further suppose that by refiguring the 1999 earnings credit at 11.33% (Lipscomb and settlement) the member's "true" retirement benefit after correcting for the "error" (assume Lipscomb upheld) should have been $2350 per month. That becomes the "fixed benefit" to which the COLA should be added. That member WOULD HAVE received 0.77% effective 7/1/03, another 1.17% effective 7/1/04, and 2% (expected) effective 7/1/05. By 7/1/05, the member's benefit should be $2443.71 ($6.29 LESS THAN the current benefit). So, in this very simple case PERS would adjust the current $2500 monthly benefit DOWN to $2443.71, and then seek recovery of $6.71 x 24 months = $161.04**, for the amount the member had been overpaid because of the "erroneous" crediting of 1999 regular earnings.
So, for members in the "window" period, the crucial detail will be the amount by which the current benefit would have been reduced had the 1999 credit been 11.33% instead of 20%. The larger the difference, the bigger the potential reduction in current benefit and the higher the amount of the payback. Just keep in mind that this approach is distinctly different than the interpretation typically offered for the Strunk decision vis-a-vis the COLA freeze. The methodology described above accomplishes TWO things, not just one. It collects for the present value of the "overcredit" as well as for the future value of the "overcredit". The standard (?mis)interpretation would have only recovered the present value of the overcredit, while continuing to leave the future value untouched. Unlikely that it would happen that way.
**If PERS really did this properly, the "payback" would be more than what is illustrated here. A correct calculation would adjust the first 12 months by more (because the differential is greater); the second 12 months by slightly less (because the differential is getting smaller). The 12 month period that would include the latest (not yet given) COLA would finally get the member down to the $6.71 differential, but that would be moot because the adjustment would have already taken place. In the end, the new benefit as of 7/1/05 would be only $6.71 less than the benefit determined at the actual time of retirement. And here you thought this would be a simple explanation.
No comments:
Post a Comment