What the Social Security 2100 Act's COLA Change Does in Law

Last verified Aug. 5, 2026. This analysis reads Section 102 of the introduced H.R. 9519 text as mirrored by GovTrack, with a cross-check against the 118th Congress version of H.R. 4583 where the drafting history matters. Senate text, an engrossed House version, or a later official PDF could change the reliance analysis and should be checked before citation in client work.

The operative target is 42 U.S.C. §415(i)(1)(D), the codified version of Social Security Act §215(i)(1)(D), which supplies the CPI reference for the title II cost-of-living adjustment calculation. Section 102 of the Social Security 2100 Act would replace that reference with a formula using “the Consumer Price Index for Urban Wage Earners and Clerical Workers” or “the Consumer Price Index for Elderly Consumers,” “whichever such index results in the higher percentage.” That is the first correction to make before any policy argument starts: the bill does not simply switch Social Security COLAs to CPI-E. It creates a higher-of rule, so CPI-E matters only in a measuring period when it produces the larger percentage increase under the statutory calculation. [1]

The first legal move: a higher-of formula, not an index substitution

The current statutory mechanism in Social Security Act §215(i), codified at 42 U.S.C. §415(i), does not ask generally whether older beneficiaries face different costs. It asks for a measured percentage increase from one computation quarter to another. Section 102 works inside that machinery. It amends the index reference in §215(i)(1)(D) rather than creating a separate benefit supplement or an independent annual adjustment outside the existing COLA architecture. [1]

That placement matters. A benefits counselor could fairly describe the proposal as a higher-of-CPI-W-or-CPI-E COLA provision. The safer phrasing is not “COLAs would be based on CPI-E,” because the proposed text preserves CPI-W as a competing measure and uses CPI-E only if it yields the higher percentage. In a year when CPI-W is higher, the formula would choose CPI-W. In a year when CPI-E is higher, it would choose CPI-E. The bill’s design is an upper-envelope calculation, not a permanent migration from one index to another. [1]

The distinction is more than editorial neatness. A pure CPI-E substitution would raise one implementation question: how to define and publish the elderly index. A higher-of formula raises two: how to publish the elderly index and how to compare the two indexes within the COLA computation. The statutory text answers the second question by selecting whichever index produces the higher percentage, but the first question is deferred to a separate Bureau of Labor Statistics mandate and an interim research-index rule. [1]

The decoupling rule prevents the higher COLA from doing other work

Section 102 then adds a new §215(i)(6). Its function is easy to miss in summaries because it does not change the headline benefit calculation. It says, in substance, that when the higher-of rule produces a larger increase than would have resulted from the CPI-W calculation alone, that excess increase is not to be used to determine amounts under provisions outside the relevant Social Security and SSI benefit adjustments. [1]

That is a legal firewall. Without it, a larger Social Security COLA could be argued to ripple into other amounts that cross-reference Social Security’s COLA mechanics. The proposed §215(i)(6) is drafted to prevent the larger adjustment from cascading automatically into separately indexed amounts. A summary that says only “CPI-E COLA” loses the rule that tells agencies and practitioners where the larger number stops.

This is also where the provision becomes less useful as a slogan and more useful as law. The bill is not merely choosing a measure of inflation. It is choosing a measure for a specified benefit calculation while blocking that choice from becoming a general index rule for other statutory amounts. That kind of decoupling provision is often where later administrative questions live.

The bill needs an official CPI-E before the formula can operate cleanly

Section 102 does not assume that the current elderly index is already the finished statutory instrument. Subsection 102(d) directs the Bureau of Labor Statistics to prepare and publish a Consumer Price Index for Elderly Consumers. That mandate is a separate legal operation from the amendment to §415(i)(1)(D). The COLA formula points to CPI-E; the BLS mandate is the mechanism meant to make that reference operational. [1]

The distinction is warranted because the elderly index now available from BLS is a research series, R-CPI-E. BLS describes R-CPI-E as an experimental index for Americans 62 years of age and older and states that it has limitations, including a smaller sample size than the official CPI population indexes and a method that uses CPI-U expenditure weights as a proxy rather than a separate elderly-consumer survey. [3]

SSA’s own policy literature has also treated the CPI-E/CPI-W comparison as more complicated than a simple senior-cost correction. A Social Security Bulletin article discusses historical CPI-E and CPI-W differences while emphasizing that the CPI-E population—persons age 62 or older—is not the same as the Social Security beneficiary population and that subgroup differences matter when interpreting the index. [4]

Those limitations do not make a CPI-E proposal legally defective. They do make the publication mandate important. If Congress wants a statutory COLA formula to depend on an elderly-consumer index, the operative question is not whether “CPI-E” sounds more beneficiary-specific. It is which index BLS must publish, under what methodology, and when the statutory reference stops depending on the research series.

The interim rule deems CPI-E to mean R-CPI-E

Section 102(e) supplies the transition. Until BLS publishes the new Consumer Price Index for Elderly Consumers required by the bill, references to CPI-E are deemed to refer to the Consumer Price Index for Americans 62 years of age and older, R-CPI-E. [1]

That is not just a housekeeping clause. It decides what happens if the benefit formula takes effect before the new official elderly index exists. The bill would not leave the Commissioner without an elderly measure; it would plug in R-CPI-E for the interim period. The price of that continuity is that the calculation would temporarily depend on the very research series whose limitations BLS itself flags. [1][3]

A practitioner explaining this point should keep the verbs separate. BLS does not currently publish the R-CPI-E as the same kind of official production index as CPI-W. Section 102 would require BLS to publish a new CPI-E and, until that happens, would deem the statutory reference to mean R-CPI-E. Adoption by bill text and empirical reliability are different questions.

The sunset and recomputation clause is its own benefit event

Section 102(f) then does something that many summaries omit: it limits the higher-of mechanism to a defined period and requires redetermination and recomputation afterward. In the introduced H.R. 9519 text, the provision refers to computation quarters “ending on December 31 of calendar years 2027 through 2036” and directs the Commissioner to redetermine and recompute title II and title XVI amounts as though the amendments had not applied. [1]

That matters for reliance. If enacted in that form, the provision would not merely start a higher-of COLA and leave all future benefit amounts permanently carrying that compounding effect. It would create a temporary statutory regime and then require a recomputation after the covered window. The recomputation language is not policy atmosphere; it is a benefit-administration instruction.

The same temporary structure appears elsewhere in the 2026 bill’s benefit provisions, including provisions using 2027 through 2036 windows for benefit increases. That shared window is one reason the COLA sunset cannot be brushed aside as a drafting afterthought. [1]

The computation-quarter wording needs verification before anyone states the first affected COLA

The introduced 2026 wording is also where the pencil should stop. Section 102(f)(1), as mirrored, refers to computation quarters ending on December 31 of calendar years 2027 through 2036. But the existing §215(i) framework defines the relevant computation quarter by reference to a quarter ending September 30. The mismatch is not something a careful reader should silently repair. [1]

The 118th Congress version helps explain why this is a version-control problem rather than a merits argument. In H.R. 4583, the corresponding sunset language referred to computation quarters “ending on September 30 of calendar years 2024 through 2033,” tracking the ordinary COLA computation-quarter structure more naturally. [2]

There are two wrong ways to handle that discrepancy. One is to pretend it is dispositive and treat the COLA section as unworkable. The other is to assume it is an obvious typo and publish a clean first-year answer. The safer conclusion is narrower: before stating which calendar year’s COLA is first affected, a practitioner should verify the official operative text, including any engrossed version and any Senate counterpart text.

That caution is especially important because COLA years are easy to misstate. The measurement period, announcement timing, benefit-payment month, and statutory effective language are not the same thing. For forecast discussion around the 2027 window, use the separate projection analysis rather than importing a forecast into this clause reading: What Will the 2027 Social Security COLA Be?.

What secondary coverage can and cannot prove

The bill’s broader return to Congress is useful context, but it should not do the work of statutory interpretation. Representative John Larson’s July 22, 2026 release described Larson and Senator Richard Blumenthal as introducing legislation to strengthen Social Security. [5] ThinkAdvisor reported on July 14, 2026 that the Social Security 2100 Act had returned and noted that repeal of the Windfall Elimination Provision and Government Pension Offset was dropped after those provisions were separately repealed. [6]

That status and political context belongs in a bill-wide overview, not inside the COLA formula itself. For the broader provision-by-provision treatment, see What the Social Security 2100 Act Would Change. The COLA question is narrower: what does Section 102 amend, what does it preserve, and what does it temporarily deem into existence?

CRS offers a useful example of how a higher-of rule would operate, but it should be handled as an illustration rather than a forecast. CRS product IF12675 uses a hypothetical comparison in which an annual COLA would be 3.0% under CPI-E and 2.8% under CPI-W for December 2025; under a higher-of formula, the 3.0% figure would be selected. Those figures should be rechecked against the PDF before being repeated in a legal memo, and the example should not be converted into a prediction about any enacted COLA. [7]

Solvency context explains the sunset design, but it does not interpret Section 102

The implementation details sit against a strained financing backdrop. The 2026 Trustees Report summary states that the Old-Age and Survivors Insurance Trust Fund is projected to become depleted in the fourth quarter of 2032, at which point continuing program income would be sufficient to pay 78% of scheduled OASI benefits. The same summary reports a 75-year actuarial deficit for OASDI of 4.42% of taxable payroll. [8]

Those numbers do not tell a court, agency, or benefits office what §415(i)(1)(D) would say after amendment. They do, however, make the sunset-and-recomputation design legally consequential. A temporary higher-of COLA with later recomputation is a different administrative object from a permanent index change, especially when the trust-fund calendar and benefit-window calendar are both part of the legislative design.

The reliance point

Section 102 is legally meaningful because it is layered. The first layer changes the §415(i)(1)(D) index reference into a higher-of-CPI-W-or-CPI-E formula. The second prevents the larger COLA from automatically changing other indexed amounts. The third orders BLS to publish a new elderly-consumer index. The fourth uses R-CPI-E during the transition. The fifth sunsets the mechanism after the stated window and requires redetermination and recomputation. [1]

A benefits lawyer can describe the proposal as a higher-of-CPI-W-or-CPI-E COLA change. Stopping there is not enough for reliance. Before citing the first affected computation quarter, the sunset window, or any Senate counterpart language, check the official operative text—not a press release, a news snippet, or even a mirrored introduced version.

References

1. H.R. 9519: Social Security 2100 Act, GovTrack, https://www.govtrack.us/congress/bills/119/hr9519/text
2. H.R. 4583: Social Security 2100 Act, GovTrack, https://www.govtrack.us/congress/bills/118/hr4583/text
3. Consumer Price Index for Americans 62 years of age and older (R-CPI-E), U.S. Bureau of Labor Statistics, https://www.bls.gov/cpi/research-series/r-cpi-e-home.htm
4. The Experimental Consumer Price Index for Elderly Americans (CPI-E): 1982–2007, Social Security Bulletin, Social Security Administration, https://www.ssa.gov/policy/docs/ssb/v67n3/v67n3p73.html
5. Larson, Blumenthal Introduce Bill Strengthening Social Security, Office of Congressman John B. Larson, July 22, 2026, https://larson.house.gov/media-center/press-releases/larson-blumenthal-introduce-bill-strengthening-social-security
6. Social Security 2100 Act Returns, ThinkAdvisor, July 14, 2026, https://www.thinkadvisor.com/2026/07/14/social-security-2100-act-returns/
7. Social Security Cost-of-Living Adjustments and the Consumer Price Index, Congressional Research Service, https://www.congress.gov/crs-product/IF12675
8. A Summary of the 2026 Annual Reports, Social Security Administration, https://www.ssa.gov/oact/trsum/