By David Enna, Tipswatch.com
The recent release of the June 2024 inflation report created a baseline number for next year’s cost of living adjustment for Social Security recipients. It’s all based on a needlessly complex formula the Social Security Administration uses to set the COLA each year.
- The index. The SSA does not use the standard measure of inflation that you see reported each month. Instead it uses CPI-W, the Consumer Price Index for Urban Wage Earners and Clerical Workers, which often runs slightly lower than the standard CPI-U.
- The time period. Instead of using an annual rate of inflation, the SSA looks at only the average of three months, July to September, and compares that to the average from a year earlier. Last year, for example, the three-month average was 301.236, an increase of 3.2% over the average for 2023. So the 2024 COLA was set at 3.2%, even though annual U.S. inflation was running at 3.7%.
- The summer months. Inflation can be notoriously volatile in the months of July to September. We could easily see a month of high-ish inflation, or an another month of deflation, as we saw in June. That means any Social Security COLA projection — including mine — is just an educated guess.
For June, the BLS set the CPI-W index at 308.054, an increase of 2.9% over the last year. So does that mean the Social Security COLA will end up being 2.9%? No, because only the next three months — July to September — matter in this equation.
In this chart, I have provided four potential monthly inflation scenarios for the July to September period — 0.0% per month to 0.3% per month — and then calculated the effect on the eventual Social Security COLA.
Most likely, none of these scenarios will end up being accurate. Anything can happen, including a bout of deflation. But I think the scenario with the highest probability is inflation averaging 0.2% a month over the three months, resulting in a Social Security COLA of 2.7%.
So my projection is 2.7%. That is a drop from 2024’s increase of 3.2% and is below the current rate of U.S. inflation, at 3.0% for the year ending in June.
As of May 2024, the average Social Security benefit check was $1,778.24, according to the SSA. An increase of 2.7% would raise the average to $1,826.25.
What others are saying
I wrote everything above before looking at any other projections for the Social Security COLA. In my experience, many of these forecasts are wildly off base.
One group I trust is the Senior Citizens League, which crunches a lot of data in developing its forecast. The group is projecting a COLA increase of 2.63% for 2025, which would effectively round to 2.6%.
I’ve seen other projections as high as 3.0%, which could happen but seem unlikely.
SSA COLA versus CPI
In general, the combination of using CPI-W and the smoothing effect of a three-month average results in the Social Security COLA being lower than annual CPI. The Senior Citizens League has lobbied for years to replace CPI-W with CPI-E, and index that more accurately reflects costs faced by older Americans.
For 2024 the COLA was 3.2% even though CPI-U increased 3.7% over the September 2022 to September 2023 period. In some years, however, the COLA has outpaced official inflation. That doesn’t look likely for 2025.
Also read: Does The Social Security COLA Shortchange Seniors?
* * *
Follow Tipswatch on X (Twitter) for updates on daily Treasury auctions and real yield trends (when I am not traveling).
Feel free to post comments or questions below. If it is your first-ever comment, it will have to wait for moderation. After that, your comments will automatically appear.Please stay on topic and avoid political tirades.
David Enna is a financial journalist, not a financial adviser. He is not selling or profiting from any investment discussed. I Bonds and TIPS are not “get rich” investments; they are best used for capital preservation and inflation protection. They can be purchased through the Treasury or other providers without fees, commissions or carrying charges. Please do your own research before investing.
Treasury Inflation-Protected Securities 
Pingback: U.S. annual inflation in August fell to 2.5%, lowest since February 2021 | Treasury Inflation-Protected Securities
Pingback: U.S. annual inflation fell to 2.9% in July, lowest rate since March 2021 | Treasury Inflation-Protected Securities
Very instructive and useful. Thank you David. Given the path government is on to run down the SS trust fund, it is no wonder they use such a formula to short seniors vs actual inflation. And if you look at shadow statistics.com you see that CPI has been doctored over the years also to understate true inflation.
Correction to my post: website is shadowstats.com.
The information on, and methodology of, the shadowstats website are by no means undisputed as a means of determining what the “true” percentage of inflation “ought” to be.
https://www.thestreet.com/economonitor/emerging-markets/deconstructing-shadowstats-why-is-it-so-loved-by-its-followers-but-scorned-by-economists
Thanks for the many interesting articles on TIPS and I-bonds.
It seems that the difference between Social Security cost-of-living changes (based on CPI-U) and CPI-W amounts to a net 0.2% difference in favor of Social Security over the 16 years listed in this article. This doesn’t seem to be a problem.
The difference with CPI-E (or R-CPI-E) would indeed be more significant. +2.2% over the last 10 years in favor of CPI-E according to the Senior Citzens League (https://seniorsleague.org/80-percent-of-retirees-want-better-inflation-protection/). That is not that much and actually less than I expected.
What am I missing? The index levels set for each month (like June 2024, e.g.) incorporate all the monthly changes going back to the prior year (and back to whenever the base for CPI-W was set). So over a multi-year period, it seems like the cumulative adjustment derived from comparing these baroque July-to-September periods should be no different than comparing, say, year-end index levels. Why does using the 3-month average create a consistent downward bias?
It’s not necessarily always a lower trend, but it isolates just three months and then averages the total. So it smooths the effect of, let’s say, a one month bump in September.
First, I think if you use “government” and “calculation” in the same sentence then “needlessly complex” is a tautology.
The averaging does indeed smooth out monthly bumps, but the effect of smoothing doesn’t change the COLA over the long term as “dips” this year (caused by averaging) are made up next year.
And maybe smoothing is a good thing to soften the affects of geopolitical events that have short duration but high impact?
If one looks back 7 years — to skip the 2015/2016 zero and make-up COLA — and compares annual changes from Jul-Jul, Aug-Aug, and Sep-Sep; 4 of the 7 years have a max/min difference of over 60 basis points, and last year was almost 100 basis points. So, again, maybe smoothing is a good thing?
But in the long run, it doesn’t matter if they used a 12 month average, a 3 month average, or a single month — the “dips” from one year are made up the next year.
As a writer on the topic I think you should embrace the 3-month averaging as it provides 3 times the opportunities for an article.
I’m curious about something: What do you make of the June personal consumption expenditures index–often referred to in the media as “the Fed’s preferred inflation gauge”–of 2.5% year over year (per Barron’s) in light of the CPI-U, CPI-W, and other varieties of CPI? Is it meaningful at all to the average investor/consumer as opposed to the Fed? Thanks.
PCE is definitely meaningful because it is the index the Fed watches, and it is almost always lower than CPI-U. But it has no effect on inflation accruals for TIPS or interest rates for I Bonds. I don’t track it carefully.
Very interesting and detailed article on Social Security and how the COLA is calculated. I am personally waiting until I am 70 to claim it. Thank you for all your efforts in providing this information. I am always learning from your site.
when the Social Security Administration provides you an estimate of your income at retirement, how should COLA’s be considered? For example, if the estimate is $3000 per month at age 67, and you are currently 60, will that $3000 actually be higher once you reach 67 because of cola adjustments?
Your base benefit gets the COLA increase when you reach 62, whether you start or delay taking SS payments.
just to make sure I understand what you are saying, if the social security tells me today my expected payment at 62 will be say $2500, will I actually get $2500 two years from now (age 60) or will I get $2500 that has been adjusted for two years of cola?
I think Social Security sets your full retirement amount in the year you turn 60, and then the COLA applies after that, whether you start at 62 or 70. Of course, the number could also go up if you keep working and will go up the longer you delay. If others have more complete info, please post it.
this is close to answering the question I’m asking but not quite. I’m trying to figure out how to make my question clear. I’ve already stopped working. Today social security tells me what I will get at age 62, 67 and 70. Will that number change due to cola adjustments or will it stay the same. That’s my question.
I had the same question you are asking now. I haven’t worked since 2021 and check my social security benefits once a year (around July). When I checked in earlier this month, my benefits at 62 had increased by 4.85% from last year’s stated amount. I have no additional earned income or work services since I haven’t worked since early 2021. So, yes the COLA adjustments do apply.
I don’t know where that 4.85% came from though as the COLA adjustment should have been lower than that.
In the simplest case, your SSA benefit is based on 3 factors: Your wage history, some wage “indexes”, and the “bend points” used in the formula for you Primary Insurance Amount (PIA)
The indexes are used to adjust your wages to account for wage growth and put everything into current year dollars from a wage growth perspective. The indexes stop changing (growing) the year you turn 62.
The bend points are calculated from the same data used to create the wage indexes, so they are also fixed the year you turn 62.
The formula uses the average of the highest 35 years of indexed wages (a.k.a Average Indexed Monthly Earnings (AIME)) and the bend points to calculate your PIA
For years 63 and beyond, your PIA will change for two reasons:
1. You have new wages that cause a change to your AIME — they are high enough to be in the top 35 and thus a lower wage drops out.
2. The current COLA. Your PIA is adjusted for the COLA every year from age 63 and beyond — even if you have not started drawing benefits.
Finally, the SSA website estimate assumes you will have wages until you draw benefits. You need to explicitly tell it to not do that.
So why did you see an increase in your estimate? Well that depends on your age and the estimate assumptions for future wages, but hopefully the above explanation will help you determine an accurate estimate of your PIA.
The simple answer is yes, the estimates will change each year (go up) basically in line with COLA.
The underlying calculation they use can be a bit complicated and varies depending on situation, but the net result comes out pretty close to same that estimated future benefits increase each year with COLA.
The estimate the SSA provides on their website assumes you will work until you draw benefits. You must explicitly select some options if you anticipate no future wages.
If you are 62 or over, and you make the appropriate selections about future wages, the estimates provided will be very accurate. If you are over 62 they will account for any prior COLA, but I am 99% confident they will NOT account for any future COLA — everything will be in current year dollars.
Having the estimate in current year dollars should be more useful as your expenses in retirement are also in current year dollars.
Thank you, this is the answer I was looking for. That makes sense to me, and what I was guessing. So I’m sitting here at 60, and if Social Security says my yearly pay is estimated at $3,000 at age 67, then once I reach age 67, it probably will actually be higher because of COLA adjustments between age 60 and age 67. The $3,000 represents “today’s” dollars, not inflated future dollars.
for your last post, see my response to Max. (it seems that wordpress only allows a fixed number of replies to replies in a string of comments so I cannot reply directly to your last one).
If you are not 62 or older the wage indexes and the bend points will change every year — even if you have no future wages. I believe the estimate the SSA presents in the online calculate will use the latest current value (and not an estimate of you specific year 62 values).
So your actual benefit when you start drawing should be higher than what is presented as a 60 year old. Both because the wage indexes should increase and the bend points should become more favorable until you’re 62. But also because you will receive COLAs beyond 62.
But (again) if you’re trying to relate your benefit amount to your expenses in retirement — you need both columns to be in the same year’s dollars. The SSA website estimate does that.
And it also DOES account for the changes to your benefit (negative) if you draw before 67 and (positive) if you draw after 67…as a 60 year old, your FRA is age 67.
Your earnings prior to 62 are ‘indexed’. At 62 and beyond you get the COLA as your adjustment. If you take before your FRA (full retirement age) your benefit is reduced. If you wait beyond your FRA your benefit is increased. Check out this site for a more detailed explanation and way of analyzing what your benefit might be. https://ssa.tools/Awesome calculator IMHO. Much better than SSA site though their info is useful too. Good luck.
Thanks for sharing your outlook. Barron’s, at least, agrees with you. On 7/11 an article by Elizabeth O’Brian stated “Cooling inflation has lowered the projected raise that Social Security recipients could receive next year to 2.7% from a projected 3% last month.”