Don’t Ignore One Large Wild Card.
A recent issue of the NBER Reporter, No. 3, September 2, reports the following:
In the fourth paper, David Altig, Laurence J. Kotlikoff, and Victor Yifan Ye calculate how retiring at different ages will affect Social Security benefit amounts, taking into account taxation and other benefits. They find that virtually all individuals aged 45 to 62 should wait until age 65 or later to maximize their Social Security benefits. Indeed, 90 percent would benefit from waiting until age 70, but only 10 percent do so.
I waited until age 67 and, for my friends who ask, I recommend the same strategy, even if they have good health and, therefore, a long life expectancy.
I looked at the NBER study underlying this chapter. It’s by David Altig, Laurence J. Kotlikoff, and Victor Yifan Ye and is titled “How Much Lifetime Security Benefits are Americans Leaving on the Table,” NBER Working Paper #30675, November 2022.
As far as I can tell, the authors assume no changes in the system over the next 20 or so years. Why is that a problem? Because the system will run out of money in about 10 years. We don’t know how politicians will adjust. We can be pretty sure that they will adjust.
Among the plausible candidates are giving, say, a 10% haircut to everyone; giving a 5% haircut to lower-income people and a 20% haircut to higher-income people; indexing Social Security benefits to the same index that’s now used (and it’s not the Consumer Price Index) to adjust federal tax brackets; and other measures. The first 3 measures, if anticipated now, give people an incentive, all else equal, to start taking benefits at age 67.
It’s possible, of course, that the likely adjustments won’t change the authors’ recommendations. But Larry Kotlikoff has made a lot of money with his software that tells people when to start taking their SS benefits. (I remember paying $40 back in 2017 when I was deciding. It gave me the answer I had come to on my own, but $40 was rounding error on the cost of a mistake.) It’s disappointing that he doesn’t consider any of these adjustment scenarios and run his numbers accordingly.
READER COMMENTS
TMC
Nov 7 2023 at 9:20am
SS is the third rail in the US, so I’d guess the likely fix will be an increase in taxes. I think they’ll raise the level at which they collect the SS tax. It’s $160k up from 147k last year.
johnson85
Nov 7 2023 at 9:37am
Maybe. But part of the reason it’s the third rail is that everybody can pretend there are no costs to the status quo. They just keep pushing the cost to future tax payers and/or beneficiaries. When people actually have to feel the costs, the political calculus will change. I think the baby boomers are ultimately going to successfully screw younger, poorer workers. But I doubt nothing will be done to social security benefits.
I would think slowing the rate of growth would basically be a given. Probably with something like chained CPI. They could also slow benefits by matching the lower of growth in productivity or chained cpi prior to receiving benefits. I think the only thing that prevents a cut to benefits on the higher end of the scale is that they still aren’t that high and it will be really hard to do asset based means testing effectively. Not sure the politics will let a high earner that suffered a late in career job loss or had his/her nest egg scrambled by a spouses illness that $36k a year is just too rich.
Mark Barbieri
Nov 7 2023 at 12:11pm
There are a lot of options on the table. Increase the tax rate from 12.4% to something higher. Increase the 6.2% that people see as being paid by their employer so that the tax increase is more palatable. Increase the cap on incomes subject to the tax (without also increasing benefits based on those higher contributions). They could expand what income gets taxed beyond earned income and include dividends and/or capital gains. They could increase the amount of social security income recaptured through the income tax. They could reduce benefits (once again calling the reduction a raise in the retirement age). They could reduce the rate of growth of benefits by using a different inflation index (C-CPI-U instead of CPI-U or maybe PCE). They could change the adjustment used when looking back at income earned in the past to reduce the benefits earned from that income. They could change the already very progressive formula for how they count SS taxes paid. Currently, the last dollars earned by a relatively higher earner are worth about 10% of their first dollars earned. They could cap payouts. They could means test payouts. They could commit to using general revenue funds to supplement the trust fund.
There are lots and lots of options with lots of different impacts. David’s point is a very good one. Using a static analysis to decide when to claim is to ignore a significant risk of impacts from future policy changes that are almost certain to happen.
robc
Nov 7 2023 at 10:22am
The obvious solution, at least to me, is the one that has already been used to extend the main retirement age from 65 to 67. They added 2 months per year to retirement age over a 12 year birth period.
It should have continued on.
https://www.ssa.gov/oact/progdata/nra.html
If we followed the same pattern, then 1960-1971 would have 67 retirement age, 1972-76 would ramp up, 1977-88 would be age 68, 1989-1993 would ramp up, 1994-2005 would be age 69, 2005-9 would ramp up, 2010-2021 would be age 70, and so on.
I think that is a little too slow, but it would still help out. If you ramp faster (by skipping the flat periods), you would eventually eliminate SS altogether, unless we hit the singularity.
Mark Barbieri
Nov 7 2023 at 12:01pm
I’ve always been fascinated by how they were able to cut benefits but frame it as a raise in the retirement age and have that framing stick. Before and after they “raised the retirement age”, you could claim your benefits as early as age 62 or delay as long as age 70, with your benefits increasing the longer you delayed. Raising the age didn’t change that. What it did was reduce the amount of your benefits at every age. What you would have received at age 65 is now what you will receive at age 67. By the numbers, raising the retirement age was no different than cutting benefits, but because of an excellent job of framing the cuts as an increase in retirement age, few people seem to see it that way.
robc
Nov 7 2023 at 12:32pm
You are correct, of course, and if they continued the trend on, they would have had to move the 62/70 numbers also.
I have thought it was totally realistic based on the original SS, in which 65 was chosen because it was above average life expectancy at the time. They should have indexed the retirement age to life expectancy.
vince
Nov 7 2023 at 3:52pm
Life expectancy is higher, but I doubt if that translates one-for-one into years of ability for productive work. It certainly won’t for jobs involving physical labor.
johnson85
Nov 9 2023 at 9:40am
Certainly the healthy working years have not increased as much as general life expectancy, but tying life expectancy to social security was always sort of a farce. Our increase in life expectancy mostly didn’t come from older people living longer. It came from reduction in mortality of younger people.
The reality is social security is a pretty abhorrent program in a lot of ways. People with lower life expectancies don’t get nearly the benefit that people with longer life expectancies get. Some of that is offset by the “progressive” nature of the benefits calculation but on an income adjusted basis, it’s brutal for people in jobs that tend to be harder on the body.
Also unfairly rewards married couples with one earner and oddly enough the childless (tell me a policy goal that is served by having both of those outcomes simultaneously?)
If they raise the retirement age, it will just be another way to take from the relatively poor workers and give to relatively affluent baby boomers.
Herb
Nov 7 2023 at 1:48pm
I took my retirement at 62. At that time the break-even was 77. I invested what I did not pay taxes on and believe I did well with those funds. Part of my decision was that SS has dire problems and would need to either increasingly tax payouts or progressively cut payouts for higher tax payers. When the pain becomes large enough to the SS system, I expect some recipients will be zeroed out. So far, the government indolence has amazed me.
Dennis
Nov 7 2023 at 4:01pm
I am responding to your comment that when you were 62 that your break-even age was 77. I’ve been trying to figure out what my breakeven age is, and as you already know, since SS can be complicated, determining break-even age can also be complicated. The simplest calculation is to compare (1) lower payments starting at 62, vs (2) full payments at FRA, vs (3) higher payments starting at age 70, and get a breakeven age around 77 to 80. But — if we earn other income while receiving SS, then up to 85% of our SS may be taxed which means our payback period may be a few years longer to get the same amount. And we always have to compare dollars adjusted to a specific date. I’ve seen people claim that payback was about 4 years!!! I don’t see how.
MarkW
Nov 7 2023 at 2:24pm
There is also the two-body problem if you’re married. If one spouse dies, the other collects only one SS payment (the higher of the two). If you’re both going to live past 78, the best strategy is to wait as long as possible. But if there’s a good chance of one spouse dying earlier, then having the lower-earning spouse start drawing at 62 might be best. It’s a complicated problem.
johnson85
Nov 9 2023 at 9:42am
I’m amazed at how couples with two working spouses are penalized compared to couples with one working spouse. That seems like such an obvious injustice and it’s amazing to me that it’s lasted this long.
dennis
Nov 7 2023 at 4:10pm
I have said for years that I would gladly give up all Soc Sec “benefits” if I could opt out before 40 years old. I wasn’t even asking for 100% opt out but maybe pay 30% of the standard payroll deduction (for the allegedly needy) and keep and invest the other 70% myself. A wise investor could almost always make enough to come out much further ahead and thus NOT NEED any govt benefits. And that would be the deal — opt out and you don’t get anything from the govt. You are free to control your own destiny. Now that I’ve hit retirement and started receiving SocSec, I want it ALL back. It strikes me odd that the govt confiscates a huge chunk of payroll, holds it for like 40 years, finally gives you your money back and has the nerve to call it a “benefit.”
Mike Burnson
Nov 7 2023 at 7:10pm
I have a hard time believing that anyone falls for these studies. Each of them is based on getting back only what was paid in by both employee and employer: the interest rate is 0%, failing to account for the lost value from inflation. Purchasing power payback would be a more valid measure, and likely no more than five years. An online inflation calculator puts the current value of $1,000 in 1977 (entering the workforce after college) as $5,000 today. Any payback calculation for Socialist InSecurity that doesn’t factor reduced buying power fails miserably.
Simple interest from a money market or bank passbook account would far surpass SS. Very conservative investments would yield much more; average stock market returns would be so much higher still: a $1,000 S&P investment in 1977 would be worth $145,000 today. With SS and this “payback” calculation, it’s still $1,000.
David Henderson
Nov 7 2023 at 7:44pm
You write:
I don’t know what you mean by “these studies.” But if you’re referring to the study by Kotlikoff et al, your criticism misses the mark. He is talking about how to maximize returns from SS. At no point does he advocate settling for what was paid in and he is too good an economist to fall into the trap of using a zero percent interest rate or of failing to consider the effects of inflation.
You’re absolutely right that most of us would have been better off taking that money and investing it heavily in stocks. Unfortunately that’s not an option we’re given and so it does make sense to ask how one can make the best of the crappy hand that FDR et al dealt us.
john hare
Nov 8 2023 at 4:09am
I know this is a side issue from the original article as David points out. But on a lower scale than investing in the stock market is buying a house with the money that would have gone to SS in an alternate universe. If one could have applied the same amount of money as SS into an increased payment on the same first house, it could have been paid off in about 7 years. What could you have done with a paid off house and no SS tax from age 30?
David S Edrich
Nov 8 2023 at 3:17pm
You are right when you say: I have a hard time believing that anyone falls for these studies. Each of them is based on getting back only what was paid in by both employee and employer: the interest rate is 0%, failing to account for the lost value from inflation.
Except people do fall for these studies. There’s also an overly cautious bias by any public figure. Anyway, if you can get 7.5% in the stock market the nest egg you get for starting 5 years early and taking the 30% cut can supplement at 67 so you get the full 100% at 67 and the nest egg lasts till age 100, at 8% it last forever. Its just math. I did a simple spreadsheet, with compounding annually. No need for a study. Of course, this assumes 7.5% on top of inflation for that supplemental portion since the SS payment is adjusted for inflation. That fact should not be ignored. And there is a risk of a kind involved in assuming 7.5%. Note tax on capital gains is generally less than that on SS. The who-knows-what-happens-by-2035 makes the decision easy for me to start at 62.
Nick Ronalds
Nov 7 2023 at 9:25pm
I thought I would wait until I was 70 until I looked at the effect of spousal benefits on the calculation. My spouse gets 50% of my level of benefits when I’m 65, but if I wait, hers don’t increase as mine would but stay fixed (with an inflation adjustment). In short, our joint benefits would increase around 4.6% for every year we wait rather than 7%. That means we’d have to live another 21 years, to at least 86, before seeing a return on investment (the investment of deferring benefits), rather than just 14 years, to 79, as would be the case if I only had to think about my own benefits. Our situation is probably not that uncommon.
Jim Glass
Nov 9 2023 at 6:27pm
Remember ‘Butch Cassidy and the Sundance Kid’, when they are about jump off the cliff into the river? Sundance says, “I can’t swim”, and Butch tells him, “The fall is going to kill you”.
Well, I don’t understand all this recent concern about financial planning for future Social Security. It’s Medicare that’s going to kill you! And me. Maybe literally.
It’s unfunded financial liability is multiples bigger than SS’s, so resulting benefit cuts/tax increases will be too. And while SS cuts/taxes will be relatively just some marginal bucks for high earners, cuts and restrictions and cost-induced inefficiencies in medical provision could be fatal. Which is more important?
Well, whichever, it’s coming. Might as well enjoy the jump!
Herb
Nov 10 2023 at 7:10pm
I had stated earlier that I took SS at 62. Another reason I took SS at at that date was that I had decided decades before that SS would not be there for us. I donated as much as possible to Traditional & Roth IRAs & both our 401ks. I find SS to be found money at this time, still expecting it to disappear, but, seeing our politicians in (in)action, probably not.
Vivian Darkbloom
Nov 15 2023 at 2:43pm
Late to this discussion; however, I have a hard time understanding the point Kotlikoff et al are trying to make here. Perhaps it’s because I have not read the full article or perhaps it is because they have not clearly explained their findings. Nevertheless, I find it difficult to accept the following, or at least it is highly misleading:
“Indeed, 90 percent would benefit from waiting until age 70, but only 10 percent do so.”
A quick check of the social security actuarial tables leads me to the conclusion that the chance of the average person dying between the age of 62 and 70 *is alone* greater than 10 percent!
https://www.ssa.gov/oact/STATS/table4c6.html
So, how could it be that 90 percent of people would benefit by waiting until age 70 if less than 90 percent would even make it from age 62 to age 70?
Comments are closed.