Penguin has pointed to the exact provision in SSA's Program Operations Manual that exposes SSIP's error. That is exactly what I needed to see. RS 00202.025: "A spouse entitled to a [Retirement Insurance Benefit] ... receives his or her own [Retirement Insurance Benefit] ... plus the difference between that benefit and the spouse's benefit." SSIP assumed my wife would be entitled to her own RIB (PIA x 1.32) plus the difference between her PIA and the spouse's benefit. That was, indeed, too good to be true and my hat is off to Penguin for diving deep enough into the POM to come up with the answer. And, as Penguin points out, the term "excess spousal benefit" is not to be found on the social security website or in the POM. It appeared in Kotlikoff's column on the PBS website (google "Social Security Secrets You Need to Know Now") where he explains:
Are there are two different formulas for spousal benefits depending on whether the spouse is collecting his/her own retirement benefit? It sure seems that way because when the spouse is collecting a retirement benefit, the excess spousal benefit (potentially reduced for taking spousal benefits early) comes into play. And when the spouse isn't collecting a retirement benefit, the spousal benefit equals half of the worker's full retirement benefit. (Note, the spouse has to collect a retirement benefit before full retirement age if she applies for her spousal benefit.) The answer, in fact, is no. There is only one formula. The formula for the spousal benefit is always the excess benefit formula. But here's what happens to the application of that formula if the spouse is not collecting a retirement benefit. In that case, the spouse's full retirement benefit (also called the Primary Insurance Amount) is set to zero in calculating the excess spousal benefit. The reason, according to Social Security, is that a worker's Primary Insurance does not exist (i.e., equals zero) if the worker has not applied for a retirement benefit (and either suspended its collection or started to receive it). In other words, your Primary Insurance Amount is viewed as non-existent until you apply for a retirement benefit. This construct - the primary insurance amount doesn't exist until it's triggered by a retirement benefit application -- lets Social Security claim to have one formula for spousal benefits. But there are, in effect, two spousal benefit formulas and which one you -- the person who will collect a spousal benefit -- faces will depend on whether or not you take your retirement benefit early.
Well, I'm not sure of all this, but Professor Kotlikoff's program (Maximize My Social Security) and Professor Reichenstein's program (Social Security Solutions) apparently got the right answer to our optimal claiming strategy and SSIP did not. Too bad, because we could have used the extra $300/month
and, in other ways, I like it better than the other two programs. It provides an analysis of breakeven dates for numerous strategies compared with the "optimal" one so you can easily grasp their relative merits. It also provides the clearest printout of projected payments, down to the month. On the other hand, the first obligation of any optimizer is to provide the correct answer. And, John, your are quite right that SSIP did not actually defend its projection.
I'm going to have another round with them and, hopefully, they will address this head on.
Attempting to answer my own question:
1. If wife files and suspends at age 66 (or, in the other scenarios, if the wife collects her benefit at 66), husband could collect spousal benefits from age 64 instead of waiting to 66. Instead of collecting $450 (half of wife's PIA of $900), if husband collects 24 months early, the reduction factor would be 24*25/36 of one percent = 16.66%, so he gets $374.98. Up to age 70, instead of collecting $450/month for 4 years = $21,600, he collects $374.98/month for 6 years = $26,998.56.
2. If wife files and suspends at 68, although delayed retirement credits are NOT applied to the calculation of spousal benefits, IF (and I realize this is a big if) she continues to work, wouldn't the extra 2 years of earnings potentially be used to increase her PIA? [If she earned enough for those two years to be counted as two of her highest 35 years of earnings.] So, the husband's spousal benefit could potentially be slightly higher for his years 66-70.
Again, seems screwy to me. Why have the wife file and suspend at 66 and do nothing for 2 years?
Englishgirl, yes, it would be pointless for wife to file and suspend at 66 as opposed to 68, but that doesn't go to the merits of SSIP's plan, just the timing of her visit to the SSA. I assumed that this was an artifact of computer-generated recommendations and took it to mean SSIP wanted her to file and before I reach 66. As for first hypothetical, if I were to claim any benefits at 64, they would be wiped out because of my continued earnings. This problem will go away when I reach 66, which is why all three programs said I should claim spousal benefits at 66. I don't really know the answer to your second question.
JW Nearly Retired: Option 1 is not as good as the plan MMSS and SSS came up with because my spousal benefit from age 66 to 70 will be $450/month, which is greater than the $300 differential between my wife's own benefit ($900) and her spousal benefit ($1200). So if she were to step up to her spousal benefit at age 68, we'd lose $150/mo for four years.