Reviewed by Donna Fuscaldo. The publisher confirms ongoing editorial review. Prepared with AI-assisted research, writing and design.
Reporting and calculations checked October 5, 2026. No new interviews were conducted.
A 3.5% Social Security cost-of-living adjustment for 2027 requires a September price reading higher than August’s. If September’s prices simply held at August’s level, our reconstruction of the government’s calculation produces 3.4%. That tenth of a percentage point is worth about $2 a month on a $2,000 gross benefit.
The Senior Citizens League’s September 11 forecast is 3.5%. As of October 5, the Social Security Administration has not announced the official 2027 adjustment. The agency’s current COLA page still identifies 2.8% as the latest determination and says the next will come in October.
There is enough published information to test what that forecast requires. We used the two available monthly price readings, last year’s comparison period and SSA’s rounding rule to work backward from 3.5%. The unpublished September reading would need to be between 329.047 and 329.997 on the CPI-W index, roughly 0.17% to 0.46% above August. Those are calculated conditions for the forecast to come true, not a prediction of September’s prices.
The calculation also puts the argument over the last decimal into household terms. A separate, still-unsettled Medicare premium can absorb more than the $2 separating the 3.4% and 3.5% examples. To estimate the deposit, a beneficiary needs both the raise and the deductions that will come out of it.
For the next step, our Social Security payment increase guide follows a fixed retirement calculation through lower-dime and lower-dollar truncation, then explains how a concurrent SSI payment changes the combined result.
One missing month, with two already fixed
The law uses the Consumer Price Index for Urban Wage Earners and Clerical Workers, shortened to CPI-W. For this calculation, SSA compares the average for July, August and September 2026 with the corresponding average for 2025. That earlier three-month average was 317.265. The comparison is between price levels across the two quarters, rather than an addition of three monthly inflation percentages.
The Bureau of Labor Statistics reported a CPI-W level of 327.104 for July and 328.481 for August. Both are the national, all-items figures before seasonal adjustment. September remains blank. The BLS release calendar schedules that month’s report for October 14 at 8:30 a.m. Eastern.
Using all twelve months of 2026 would answer a different question. So would substituting the more prominent CPI-U inflation number at the top of the monthly news release. CPI-W is the series written into the COLA formula. Its name matters: BLS bases it on spending by qualifying wage-earner and clerical-worker households, rather than on a budget designed specifically for retirees. The agency explains the difference in its CPI questions and answers.
That helps explain why a national adjustment can feel unlike the change in one person’s bills. Someone whose rent or medical spending takes an unusually large share of income has a different mix of expenses. The COLA calculation does not ask each beneficiary for that mix; the same national index supplies the inflation measure.
The measurement window
Three months on each side. One blank.
Each month gets one-third of the weight. The brass-edged windows show the only months that enter this comparison.
Marks use the same CPI-W index scale, 315–335, in both years. The September mark appears only when you choose an assumption.
Choose an illustrative September
The outlined September space is an unpublished observation, not a zero. Choosing a scenario fills it with a hypothetical value.
Observed values: SSA CPI-W table and BLS July and August releases. Index: 1982–84 = 100, all items, not seasonally adjusted. Rounding: 20 CFR 404.275.
The two published months do not get to stand in for the third. Nor should a calculator carry every decimal through to the end. Under SSA’s regulation, the quarterly average is first rounded to the precision of the published index, currently three decimal places. The percentage increase is then rounded to the nearest tenth of a percentage point, with a halfway result rounded upward. That order changes the exact edge at which one displayed rate becomes the next.
The September reading needed for 3.5%
Start with the simplest hypothetical: September’s CPI-W stays at 328.481, exactly where August finished. Adding 327.104, 328.481 and 328.481, then dividing by three, produces a rounded quarterly average of 328.022. Against last year’s 317.265, that is a rise of approximately 3.39054%. The resulting COLA is 3.4%.
We then tested September values in the same one-thousandth increments BLS publishes. The first one that produces 3.5% is 329.047. Every published value through 329.997 gives that same answer. At 329.998, the result becomes 3.6%. The band is a consequence of the rounding rule; nothing in the calculation assigns an equal likelihood to the values inside it.
Put another way, 3.5% requires at least a 0.566-point increase from August. That is about 0.172% for the month. The upper end of the 3.5% band is about 0.462% above August. These are changes in the unadjusted CPI-W, so they should not be compared directly with a seasonally adjusted monthly headline for CPI-U.
Read the edge, not just the headline
A forecast has a whole ledge to stand on.
Moving across September’s index changes the average continuously. The displayed COLA moves in tenths. The 3.5% ledge begins at 329.047 and ends at 329.997.
Use 290.000–340.000, up to three decimals. This deliberately broad testing range is not a forecast. Drag the landscape or use the slider’s arrow keys for 0.001 steps.
The September observation is missing.
No scenario selected.
The adjacent inputs 329.046 and 329.047 land on opposite sides of the 3.5% edge. Their difference is just 0.001 index point.
Exact boundary checks and worked examples
| September CPI-W | Q3 mean | Modeled COLA |
|---|---|---|
| 328.481 Flat from August | 328.022 | 3.4% |
| 329.046 | 328.210 | 3.4% |
| 329.047 | 328.211 | 3.5% |
| 329.858 Repeat the rise | 328.481 | 3.5% |
| 329.997 | 328.527 | 3.5% |
| 329.998 | 328.528 | 3.6% |
The mean is rounded to three decimals before the increase against 317.265 is rounded to the nearest tenth of a percentage point. The interval says which inputs share a result; it says nothing about their likelihood.
Clayso calculations from the cited BLS observations and SSA rounding rule. The terrain is the calculation’s piecewise contour, not a probability distribution. SSA makes the official determination.
There is a useful second test. The index rose 1.377 points between July and August. If September repeated that absolute increase, its level would be 329.858. The quarter would average 328.481 and the adjustment would round to 3.5%. Thus the forecast does not require September’s index to accelerate beyond August’s point increase. It does require more than no change.
One month’s movement is diluted by the three-month average. An extra full index point in September adds only one-third of a point to the quarter, or about 0.105 percentage points to the unrounded annual increase. Much smaller changes matter when they happen to cross a rounding boundary. This is why a tiny difference between two September readings can move the published COLA by a tenth, while a larger difference elsewhere leaves it unchanged.
August’s 3.5% is a different number
The August BLS release also reports 3.5%. It describes the CPI-W increase over the preceding twelve months: August 2026 compared with August 2025. Last August’s index was 317.306. The COLA comparison instead uses three months on each side, including a September reading that has not yet been published. The matching percentages can make the forecast look settled when they refer to different periods.
The Senior Citizens League’s estimate has a different status again. It is the output of the group’s statistical model, which it says uses inflation, interest rates and unemployment. Its August 12 projection was 3.6%; the September release reduced it to 3.5% and called that the group’s final prediction. A dated model forecast can change as evidence arrives. It does not replace SSA’s determination.
Same print. Different comparison.
Two ways to see 3.5%.
One lens looks backward at two August readings. The other looks ahead to an adjustment built from two full quarters.
BLS: August’s 12-month change
317.306→August 2026
328.481
3.5% observed. One month compared with the same month a year earlier.
TSCL: September 11 forecast
317.265→Q3 2026
September missing
3.5% predicted. The future COLA requires the full July–September comparison.
Matching numerals do not make these the same claim. The annual figure is known; the 2027 adjustment still lacks one month’s observation.
No September assumption selected. The forecast refers to a range of possible September readings, not one exact input.
BLS August 2026 release; Senior Citizens League, September 11 forecast. Neither lens is an official 2027 COLA announcement.
By October 5, September itself has ended. What remains unknown to the public is its measured price level. A price rise beginning in October would fall outside this calculation’s measurement window, even if it became the biggest economic story before SSA’s announcement. It could affect purchasing power after the measurement period without adding anything to the 2027 COLA.
A September fall would not erase the year’s rise
The opposite misunderstanding is that a fall in September’s prices would mean no adjustment. A modest retreat from August would still leave the three-month average well above last year’s. For example, putting September back at July’s 327.104 yields a quarterly average of 327.563 and a calculated 3.2% COLA. That is an illustrative decline scenario, not our outlook.
There is a no-increase rule. SSA says that when the comparison produces no increase, or an increase that rounds to zero, there is no COLA for that year. The mechanism does not impose a negative COLA. Its comparison rule also retains the third quarter of the last year in which a COLA became effective as the base. After a zero year, prices must clear that earlier benchmark before another adjustment is generated; a recovery from a lower intervening year is not enough by itself.
The Medicare number is still a projection, too
For a current gross monthly Social Security benefit of $2,000, a simple 3.5% planning calculation adds $70, producing $2,070 before deductions. At 3.4%, the increase is $68. The difference is $24 across twelve months. These examples show scale; SSA’s actual benefit calculation applies the COLA to the underlying benefit amount and includes rounding and any applicable adjustments, so multiplying today’s deposit is not an exact benefit calculation.
For someone whose Medicare Part B premium is withheld from Social Security, the increase in that premium comes out before the money reaches the bank. CMS set the 2026 standard monthly Part B premium at $202.90. The Medicare trustees report issued June 9, 2026, in Table V.E2 on printed page 207, projects $209.50 for 2027. The report expressly treats future amounts as estimates. We found no final 2027 standard-premium announcement in the official sources checked on October 5.
The projected difference is $6.60 a month. Combine it with the hypothetical $70 gross increase and the deposit rises by about $63.40, assuming the person pays the standard premium through Social Security and all other deductions stay the same. The two uncertainties are separate: the COLA forecast must hold, and the final premium must match the trustees’ projection.
From a gross raise to a deposit
The deduction cuts into the extension.
The beam measures the change in monthly dollars. A gross increase extends it; a higher deduction cuts some of that extension away.
Fictional gross benefit: $2,000. Neither a September index nor a deduction change has been chosen. An unresolved deduction is not treated as zero.
The $6.60 scenario subtracts the official 2026 standard Part B premium, $202.90, from the trustees’ projected 2027 premium, $209.50. The latter is an estimate, not the final 2027 premium. All other deductions are held unchanged in that scenario.
The $2,000 example without any controls
A hypothetical 3.5% COLA gives an approximate $70 gross monthly increase. With unchanged deductions, the modeled deposit increase is $70. If the standard Part B premium instead rises by the projected $6.60 and other deductions stay unchanged, the increase is about $63.40.
These are planning multiplications, not SSA’s individual benefit computation. Hold-harmless conditions, income-related premiums, who pays the premium and other adjustments can change the answer. This Social Security example does not calculate SSI.
CMS: 2026 standard premium; June 9, 2026 Medicare trustees report, Table V.E2, printed page 207; SSA benefit application and rounding.
That premium scenario has more than three times the monthly dollar effect of choosing 3.4% rather than 3.5% on the same $2,000 benefit. It is a reason to watch the deduction notice alongside the COLA announcement. Treating a gross raise as spendable cash can create a larger budgeting error than being a tenth of a percentage point off on the forecast.
The example will not fit everyone. Higher-income beneficiaries may pay an income-related additional premium. Some people have Medicaid paying their Part B premium; others pay Medicare directly. A person without a Part B deduction should not subtract the standard premium from a Social Security deposit just because it appears in a national example. Anyone reviewing next year’s coverage also needs to check their plan’s costs during Medicare open enrollment for 2027.
For qualifying people, Medicare’s hold-harmless rule limits the Part B premium increase so it does not, by itself, make their Social Security payment smaller than before. Protection depends on benefit entitlement and premium-deduction timing; new enrollees and people paying income-related premiums are among the exclusions. It does not cover every deduction. SSA’s explanation of hold harmless sets out those conditions.
Two benefit programs do not mean two independent raises
The COLA also changes the federal Supplemental Security Income payment standard. SSI is means-tested, however, so the federal maximum is not a promise of the amount any particular recipient will receive. Countable income, living arrangements and any state supplement can change the payment. The official 2027 federal amounts must wait for the adjustment; multiplying a rounded 2026 maximum is not a substitute for SSA’s published figure.
This matters especially to people receiving both Social Security and SSI. Social Security is generally unearned income for SSI purposes, subject to applicable exclusions. When that income rises, the countable portion can reduce the SSI payment. Forecasting a percentage increase on each current deposit separately can therefore overstate the combined gain. SSA’s SSI income rules explain the deductions and exclusions, and the agency’s individual notice resolves the actual amounts.
A bigger benefit can change the tax calculation
Federal taxation of Social Security still depends on income and filing status. The calculation generally combines income before Social Security, tax-exempt interest and half of Social Security benefits, with adjustments described in the IRS worksheet. Under the IRS rules, benefits can become partly taxable above a $25,000 base for an individual or $32,000 for a couple filing jointly; married filing separately has special rules. Up to 85% of benefits can be included in taxable income. That does not mean an 85% tax rate. SSI payments themselves are not federally taxable.
A COLA can move someone closer to, or farther into, taxable-benefit territory when other income stays the same. Separately, the additional senior deduction under current law is available for 2025 through 2028, up to $6,000 per qualifying person age 65 or older, with income-based phaseouts. It can lower taxable income for eligible people. It does not repeal the rules determining how much Social Security is included in income, and it does not turn the COLA into an automatic tax-free deposit.
Some first raised payments arrive before January
Readers comparing a Social Security forecast with a veterans’ award can use the VA COLA increase guide to separate disability compensation from education-rate calendars and understand why a proposed effective date does not establish a payable rate.
The forecast does not change October’s Social Security payments. The announcement and the first raised payment operate on different calendars. The October 14 date on BLS’s calendar is the scheduled release of September inflation. SSA’s own page promises its COLA announcement in October, without specifying a day there. Readers should verify the rate in SSA’s determination rather than treat the data-release appointment, or a forecast’s announcement date, as the determination itself.
A year-end COLA applies to December’s Social Security benefit, normally paid in January. SSI’s new federal standard applies to January. SSA’s 2026 payment calendar shows January’s SSI arriving on Thursday, December 31, 2026, because January 1 is a holiday. The Social Security payment normally due January 3 also moves back to December 31, because the 3rd falls on a Sunday and the earlier Friday is New Year’s Day.
The 2027 calendar puts the three Wednesday Social Security groups on January 13, 20 and 27. An early December 31 payment retains its benefit month; it is not an extra annual benefit. Someone planning around the raise should use the scheduled date for their payment group, along with their individual new amount, rather than assume everyone receives it on January 1.
For now, the 3.5% forecast is a useful assumption to test against a budget. The last inflation reading will settle the missing input. SSA’s determination will settle the percentage, and the individual notice will show the benefit and deductions. On a $2,000 gross benefit, the current flat-price and forecast scenarios are only $2 a month apart. The money available to spend depends on following the calculation all the way to the deposit.
Reporting and calculations checked October 5, 2026. This article analyzes published BLS, SSA, CMS, IRS and Senior Citizens League documents; it does not include interviews or an agency determination for an individual. Index scenarios and benefit examples are illustrative. Quarterly means were rounded to three decimal places before calculating and rounding the percentage increase. The official COLA and final 2027 Medicare premium were not announced in the sources checked at the time of review.