ISTM that if the gov’t wants to reduce SocSec spending without attracting much attention all they have to do is mess with the COLAs.
When I retired (at 60) I ran simulations through the Fidelity Retirement Calculator both ways. It turned out that for me it didn’t make any significant difference. So I started withdrawals at 62. Like @Tamerlane , it paid my mortgage.
How does that calculator work?
ISTM that the value partly depends on life expectancy, but also on the personal utility of money now vs the insurance that you won’t completely run out and will have enough to live off of if you have greater longevity than actuarially expected.
Those are calculable I guess but placing numbers on them seem hard to me?
Even social security is not completely risk free. Yes COLA could get reduced. It just seems to me to be the best risk adjusted option available? Decent return, lower risk than almost any other option, and betting on the outcome I desire, assuming good health.
I think that would attract a HUGE amount of attention. What they could do, quite easily, is increase the maximum amount of wage that is subject to SS tax. Currently, the max is 184,500. Anything a person earns over that in a year is not subject to SS tax. Removing that cap is an easy act, and would improve revenue immensely.
I set up a spreadsheet when I was trying to decide when to start drawing SS. I based the monthly amounts on what the SS website told me I would receive if I started drawing at age 65, 66, and so on.
In my case, I calculated the total payout if I started at 66, 68, and 70. Turned out that if I started drawing at 66 and got x dollars a month, I would draw y dollars at age 81. If I started drawing at 70 and got z dollars, I would draw the same y dollars at age 81. And of course, z was greater than x. So I live to be age 81, I will come out ahead. If I don’t, I won’t much care, although my wife will get the increased monthly amount if she outlives me.
Not sure if that is how the calculator works though.
Honestly to me that breakeven point is basic. It’s going to break even at average age to die for people my age.
It’s the utility curve bits that get so complicated.
Theoretically if I am an aggressive investor and the market performs along its historical norms for the next two decades, I could do even better investing the money. Even if I live a long time. So if I’m that person I should take it and invest it aggressively.
And a different utility if I expect to live at least as long as average. And especially if the long tail end of my portfolio in case of major market crash concerns me.
I guess I think of it as bond like? Exactly as a best possible annuity.
How does a calculator capture those things?
There are a number of online calculators. All of them that I have seen will tell you what your monthly payout will be when you retire, based on your DOB, latest year’s earnings, and expected retirement date.
None of them that I have seen would be able to do what you propose. If you’re proficient in Excel, you could probably make those calculations, but it wouldn’t be easy.
Oh those I know. And the Monte Carlo ones that give the fraction of runs you drop to zero by age X or alternatively leave an estate bigger than you started with.
I don’t think those are much use to me regarding the social security decision.
I may or may not live a long time. I don’t need the money now. I may never need it but I would be really pissed if I end up living a very long healthy life and am short on resources to enjoy it. A portion in a very attractively priced inflation adjusted lifetime annuity that starts paying out at 70 is a very attractive product to me, one that allows me to be less afraid of being more aggressively positioned otherwise. I’m going to max out on it.
It’s pretty obvious that removing the earnings cap is the right and rational thing to do. But that would attract the attention of influential high-earners who vote R (and might even piss off some who vote D), so it won’t happen. Rearranging the COLA calculation is easier to conceal/obfuscate.
That is exactly my thinking. And is that of several pros I respect. You cannot buy a better deal at any price.
The only shortcoming to SS is that high earners are capped on putting in, don’t get a great deal on getting out, and are taxed at a high marginal rate on the smallish (percentagewise) benefits they do get.
And of course the political risk that Washington could simply renege on the “deal” for our cohort at any moment. The lack of this particular risk is probably the only argument for commercial annuity-like products of any description.
They’d be very hard pressed (under conventional politics) to remove the cap on earnings without also removing the cap on payouts. The whole reason SS has been, and still is, politically viable is the semi-fiction that “everybody pays in; everybody gets out”. IOW, it’s not “welfare” where the undeserving freeloaders get anything.
The instant you tax bigger wages and don’t also pay out on those wages taxed, you break that bargain. And fiscal conservatives’ support for SS will evaporate overnight. Even though their Moms depend on it. OTOH, if you do tax the higher wages and also pay out on them to keep the political bargain intact, the fiscal benefit of the change turns into a mere rounding error.
Recall also that SS only taxes wages, not income. An awful lot of the money made by the high income set from employment (or business ownership) is not wages. It’s benefits, stock options, profit sharing, dividends, various BS fig-leafed payments, etc. And for most high income folks, they have a lot of personal influence on how much of their income is labeled “wages” and how much is not.
This is one of those headline changes that sounds good that will be 90+% offset by (fully legal) changes in taxpayer behavior. The only people left holding the bag will be those who are influenceless upper-middle management superdrones at big corps earning ~$150-300K. Who collectively are a drop in the bucket of the currently SS-untaxed income of the fatcat set. Which income is the prize folks keep salivating over taxing to “save SS”.
Best I can figure for the price to get the closest commercially available:
Take a hypothetical max earner: difference between taking at 66 and delaying until 70 if they can manage it -
Foregone benefits are ~$158K (4 years × ~$39,600)
Incremental annual gain at 70 are ~$21,700/year for life with CPI COLA.
Closest annuity product commercially available would run over $400K for that. With more risk, no survivor benefit, and not the same at least partially tax protected.
Quite a discount for a far superior product. Yes it runs positive at actuarially predicted age of death. So thinking only in terms of return that is “break even”. But the utility of having “longevity insurance” is the huge bonus.
A secure inflation adjusted additional income nearly $22K a year. That’s about the same as having $1M in TIPS, which is maybe the closest comparison?
Damn. Okay not quite the same as that because TIPS taking just the yield would leave estate. But again knowing the SS income is there allows someone to be that much more volatility tolerant with the remaining funds, which usually works out well for heirs.
Anyway. I see the most probable fixes being a bit of this and that. More of it counting as taxable income as the one that will impact me the most. But I see that as fair.
And that is assuming you can even get a true COLA annuity with annual increases tied to the CPI, rather than just one with a fixed (e.g. 3%) annual increase. I’ve haven’t done extensive searching, but my understanding is that annuities with “true” COLA riders are very rare or non-existent at this point.
Oh high degree of confidence you cannot. It is just the closest comparable commercial product.
Having made that claim, anyone here with the chops to help me support it?
Let’s use just waiting age 66 to 70. That’s going to increase payments by about 8% a year compounded, so 36% more I think.
Keep our hypothetical simple. From I can find taking SS at 66 typically provides something like 45% of expected retirement expenses for an upper middle earner.
So by waiting instead of needing the funds provide 55% of the income we are down to 40%.
It terms of buckets, that means how much less in the safest buckets, and how more reasonable to keep in the higher volatility higher reward bucket.
That marginal amount - how much more, average historical anyway, do we expect it to return over 20 years than if it was in the safest bucket?
This is my favorite expert’s take on the topic:
Note this article is 11 years old. Some of the detailed numbers have shifted, but the big picture approach to the analysis is as valid as ever.
Here’s a related take on the commercial equivalent of the delayed SS strategy:
Again a few years old now, but easy enough to update the numbers to current.
Both good articles and the first alludes to the asset allocation implications I am trying to quantify - but implies that most will just ignore them and instead “glide” into it.
I suspect there is quantification possible though?
Take the extreme case - a household has all expected expenses handled by income from real estate investments and SS. This stream will never run out. They need some bucket for unexpected expenses, health disasters for example, which needs to be protected from volatility, and somewhat liquid. But the other bucket? It can be very volatility tolerant and very probably higher long term returns. That should be able to calculated using historical data anyway.
Few will be in the position that SS completely covers expenses (although many have no choice but to be) but the closer one is to it the more volatility should be tolerable and thus the higher the long term return.
Is that calculable? Probably?
It’s certainly calculable. But the problem is the calculation relies on a large number of assumed numbers about alternate returns, actual longevity, future spending profile, etc.
So it doesn’t yield an answer. It yields a multi-dimensional cloud of answers. Which is something the industry struggles to simplify and struggles to depict in a way ordinary customers can grasp it.
As a metaphor: Monte Carlo analysis produces a whold cloud of possible outcomes. but somehow to sell that, the industry needs to collapse the cloud of possibilities down to one binary outcome: This e.g. asset allocation may, or won’t, run you out of money.
And trying to investigate a bit (Claude helped) I find some nice ways of framing it, but little actual numbers run. This is where a spreadsheet with varying numbers entered probably would be useful! 
The framings I have read mainly quantitize it as “income coverage ratio” - guaranteed income relative to spending needs.
Probably the most on point is the idea that there is more ability to take risk the higher that ratio is, but there still may be little willingness?
Per Claude there are these bits out there:
Meyer/Reichenstein: guaranteed income displaces bonds; every dollar of annual guaranteed income is roughly equivalent to holding $20–$25 in bonds (at a 4–5% capitalization rate), so higher income coverage should mechanically shift the portfolio toward equity
Pfau rule of thumb: each 10 percentage point increase in income coverage ratio supports roughly 5–10 additional percentage points of equity allocation in the residual portfolio, depending on exact numbers assumed.
I guess from there one could plug each of those different residual portfolios and residual uncovered expense needs into Monte Carlo calculators and see the clouds of answers that result in 20 plus years?
Why for all that is sane in the world are the markets going up this week?
Probably for the same reason that crude oil prices are down almost 8% today?
Damned if I know.