The 2027 Social Security COLA, Computed From the Actual CPI-W Data — Not Forecast
July's real CPI-W print is in, and it puts the 2027 COLA near 3.1%, below the 3.6%–3.8% summer forecasts. Here's the SSA's own formula run on published data, back-tested against the 2026 COLA it already announced.
The Social Security Administration announces it in October 2026, based on CPI-W data for July, August, and September 2026. Everything before that is an estimate.
Get new posts in your inbox
No spam, unsubscribe anytime.
About the Author
Sujit KarkiFinance Researcher & Market Analyst
Independent finance researcher and market analyst with expertise in macroeconomics, equity markets, and personal finance. I help regular investors make better-informed decisions through rigorous, data-driven analysis.
If you or a parent lives on Social Security, the number that matters most each year isn't the stock market — it's the cost-of-living adjustment. This post originally ran in July with the forecasters' estimates. July's actual inflation data has since been published, so I have replaced the forecast with the arithmetic.
Pro Tip
Updated 7 September 2026. When this published in July, the best available
information was a set of forecasts clustered at 3.6%–3.8%. The July CPI-W print
has since landed, and running the SSA's own formula on it points closer to
3.1%. The original forecast section is kept below, because how those
estimates moved is part of the story. The headline number has changed.
Key Takeaways
Running the SSA's formula on published CPI-W data puts the 2027 COLA near 3.1%, below the 3.6%–3.8% that summer forecasters projected.
The method reproduces the 2026 COLA of 2.8% exactly, which is the reason to trust it.
Two of the three months that decide the number are still unpublished, so the honest range is 2.7%–3.7%.
The COLA rests on only three months of data, which is why these estimates swing all summer — unlike tax brackets, which average twelve months and barely move.
A Medicare Part B premium increase can quietly eat part of whatever the raise turns out to be.
What the actual data says
The COLA is not a forecast, a negotiation, or a policy choice. Since 1975 it has
been one piece of arithmetic: take the average CPI-W for July, August and
September, compare it with the same three months a year earlier, round to the
nearest tenth of a percent. That is the entire rule.
So rather than repeat someone else's estimate, here is that formula run against
the published series — CPI-W, BLS series CWUR0000SA0.
First, does the method work? Before trusting any projection, it should
reproduce a COLA the SSA has already announced:
Back-test: the 2026 COLA, already announced
Input
Value
Q3 2024 average CPI-W
308.7290
Q3 2025 average CPI-W
317.2647
Computed increase
2.7648% → 2.8%
What the SSA actually announced
2.8%
It matches exactly. That is the only reason the next table is worth reading.
Now 2027. The comparison basis is the Q3 2025 average of 317.2647. Of the
three months that decide it, only July 2026 has been published: 327.104.
August and September are still to come.
If August and September simply hold flat at July's level, the 2027 COLA comes to
3.1%. Reaching the 3.8% the summer forecasts pointed at would need both
months to run about 0.65% hotter, every month — a real possibility, but not the
base case the published data supports.
Warning
Notice how wide that spread is: a plausible range of monthly outcomes moves the
COLA a full point, from 2.7% to 3.7%. That is not sloppiness in the estimates —
it is arithmetic. The COLA averages three months, so each unpublished month
carries a third of the weight. Compare that with the 2027 tax
brackets, which average twelve months: there,
the same uncertainty in one month moves the answer by less than a tenth of a
percent. Same kind of formula, completely different sensitivity — and it is why
COLA predictions lurch around all summer while bracket projections sit still.
The script that produces these numbers is
scripts/project-cola.mjs and re-runs
against live BLS data, so this section updates as August and September land
rather than going stale.
What the forecasters said in July
Three trackers have weighed in as of mid-July 2026, all pointing the same direction but landing on slightly different numbers:
Source
2027 COLA estimate
As of
The Senior Citizens League (TSCL)
3.8%
July 14, 2026
AARP
3.6%
July 2026
Mary Johnson (independent analyst)
3.7%
July 2026 (down from 4.7% the month before)
Actual 2026 COLA (for comparison)
2.8%
January 2026
The estimates cooled from spring highs — some had reached 4.2%–4.7% — because June's CPI-W came in softer than expected. This is a rare case where lower inflation is a mixed blessing: better for your grocery bill, but it shrinks next year's raise.
How the number actually gets calculated
None of these estimates are guesses in the "vibes" sense — they're forecasts of a specific, mechanical formula the SSA has used since 1975:
How the number actually gets set
From CPI-W to your check, in 5 steps
1
CPI-W collected for Jul, Aug, Sep
The Bureau of Labor Statistics measures inflation for wage earners over the third quarter
2
Averaged and compared to the prior year
This year’s Q3 average vs. last year’s Q3 average
3
The % increase becomes the COLA
Rounded to the nearest 0.1% — this is the only input, by law
4
SSA announces the official number
Always in October, for the following January
5
Applied to benefits starting January
Automatic — no application needed
Every estimate you see before October — 3.6%, 3.7%, 3.8% — is a forecast of step 3, made before steps 1–2 are even complete for Aug/Sep.
What that's worth in dollars
Percentages are abstract; dollars aren't. Against an average retirement benefit
of $1,937.53, here is what each end of the range actually delivers:
Monthly increase on an average $1,937.53 benefit
COLA
Monthly raise
New benefit
2.7% (cool August/September)
$52.31
$1,989.84
3.1% (flat — what July's data implies)
$60.06
$1,997.59
3.7% (hot August/September)
$71.69
$2,009.22
Monthly raise at 3.1%$0Average benefit $1,937.53 → $1,997.59
The gap between the ends of that range is about $19 a month. That is the
practical cost of the two missing data points.
Interactive · estimate your 2027 raise
What your COLA is actually worth
Your current monthly benefit$1,937.53
2027 COLA scenario
Expected Medicare Part B increase$6.60/mo
At a 3.8% COLA, your benefit would rise to $0.00/mo — a raise of $73.63/mo, or $0.00/mo net of the Medicare Part B increase you set above.
Illustrative only — the SSA doesn't announce the official 2027 COLA until October 2026, and the 2027 Medicare Part B premium isn't finalized until around November 2026. Not financial advice.
Why this number carries so much weight
This isn't a rounding error for most recipients. Per Congressional Research Service data cited by AARP and SSA reliance studies, most beneficiaries lean on Social Security for a large share of their household income:
Three in five beneficiaries get at least half their income from that check, and roughly one in seven get nearly all of it. That's why the full-point spread still on the table — 2.7% vs. 3.7% — isn't academic: for the 90%+ group in particular, it's close to their entire raise for the year, full stop.
The Medicare Part B catch
Here's the part the headline number hides. For most retirees, the Medicare Part B premium is deducted straight from the Social Security check. The 2026 Medicare Trustees Report projects the standard 2027 Part B premium at around $209.50/month, up from $202.90 in 2026 — a projection, not a final number. CMS doesn't announce the actual 2027 premium until around November 2026, and past projections have missed (the 2022 premium, for example, was projected at $158.50 but finalized at $170.10).
Warning
Don't budget the full 3.1%. Budget the raise minus your expected Medicare premium increase. For lower benefit amounts, the Part B bump can eat a meaningful slice of the COLA — and the real number won't be locked in until November.
The uncomfortable trade-off
A larger COLA helps recipients now but strains the system. The nonpartisan Committee for a Responsible Federal Budget (CRFB) estimated in May that a 3.8% COLA in 2027 would add roughly $300 billion to Social Security's fiscal shortfall over the next decade and advance a key trust fund's projected depletion date by about three months, from late 2032. CRFB separately estimates that hitting that insolvency date without Congressional action would force an automatic ~25% across-the-board benefit cut. That's not a reason to root against your own raise — but it's the honest context.
Here's What I'd Actually Do
Treat the estimate as a range, not a promise. Plan around 3.1%, treat 2.7% as the downside, and treat anything above 3.5% as upside. The official number is October.
Net out Medicare. Calculate raise minus expected Part B increase for your real spendable change — and don't assume $209.50 is locked in.
If you're not yet claiming, don't let COLA math rush you. Delaying benefits still raises your base far more than any single year's COLA.
Revisit your withdrawal rate. A higher COLA slightly reduces how much you must pull from savings — recalculate once the official number lands.
No. It is what the SSA formula returns if August and September 2026 hold flat at July levels. Those two months are a third of the calculation each, so the outcome can still land anywhere from about 2.7% to 3.7%. The SSA sets the official figure in October.
Sujit Karki|||6 min read