The 2027 ICHRA decision should be built from several current inputs: the correct affordability percentage, current enrollment instructions, local carrier availability and a contribution model for the workforce. A national market headline cannot replace any of them.
What is established in the reviewed sources
| Input | Reviewed position | Employer consequence |
|---|---|---|
| 2027 affordability | IRS required contribution percentage: 10.22%. | Use the correct plan-year calculation. |
| Federal open enrollment | CMS guidance: November 1, 2026–January 15, 2027. | Separately confirm the deadline for the intended coverage start. |
| Cigna individual/family medical | Company notice says no 2027 offering. | Identify affected employees and local replacement options. |
| Reviewed CMS premium dictionary | Describes 2026 data. | Do not label those premiums as verified 2027 rates. |
Sources for these planning inputs: source source source source
New state-market signals for 2027
Four September sources show why evidence type and approval status have to travel with every number. An approved weighted market average, a preliminary plan inventory and a final employee-level premium answer different questions. They cannot be averaged into a national ICHRA trend or one state score.
| Source and plan year | Scope and status | Observed metric | What it cannot establish |
|---|---|---|---|
| Washington · 2027 Decision September 9 | Exchange individual market; approved average | 22.2% average increase | Each plan, county or employee's lowest-cost Silver change |
| Connecticut · 2027 Decision September 11 | Individual and small group, on and off exchange; approved averages | 11.3% individual; 15.1% small group | Which arrangement costs less for the same workforce |
| Pennsylvania · 2027 Decision September 18 | Individual and small group; approved weighted averages | 15.97% individual; 10.28% small group | Actual local plan prices or an ICHRA-versus-group result |
| Maryland · 2027 Presentation September 21 | Exchange individual-plan inventory; rate and form approval pending | 43 plans in 2026 and 43 in 2027, with mix changes | Final approval, premiums, networks or affordability |
Sources: source source source source source
Read the underlying updates: Washington's individual-market decision, Connecticut's paired decision, Pennsylvania's individual and small-group rates and Maryland's preliminary plan lineup. Use them as prompts to refresh local inputs, not as substitutes for plan-level evidence.
Carrier participation: an updated signal, not a local plan count
In narrative dated September 15, KFF reports nine carrier groups announcing exits from some or all of their state ACA Marketplaces for 2027, and six announcing entry into new state Marketplaces. Its notes group insurers by parent company. These counts describe different geographic moves; subtracting six from nine would not measure a national net loss of three insurers or any employee's available plans. source
The same page retains a table-source note dated July 2, 2026. We preserve that inconsistency rather than treating the table as a fully reconciled September county inventory. We have not identified a specific additional entrant from a documented revision history. State-specific conclusions require current exchange, regulator or carrier records. source
| Check for the employee's location | Why it is a separate question |
|---|---|
| Participating carriers and exact plans | A state announcement does not establish every county's plan inventory. |
| Provider networks and formularies | A familiar insurer name does not establish access to a particular doctor or prescription. |
| The employee's selected-plan premium | This helps describe the employee's actual premium cost after the contribution. |
| Applicable lowest-cost self-only Silver premium | The affordability benchmark is a separate input; it need not be the plan the employee chooses. |
Use the network and prescription checklist, employee contribution examples and 2027 affordability guide. The latter separates the employee premium-tax-credit test from employer-shared-responsibility analysis and applicable safe harbors. Do not substitute a general market average for the required input. source source
The Market Explorer remains a scoped 2026 public-data resource, not a live 2027 quote engine. A new employer record, such as Alliance Entertainment's disclosed transition, also does not establish the quality of insurance in any employee's location.
The local evidence still needed
This review does not establish final 2027 premiums for every county, each employee’s eligible plan menu, provider-network continuity or the outcome of every state filing. A requested rate change is not necessarily an approved rate, and an insurer’s national announcement is not a substitute for an employee-specific plan comparison.
Build the analysis so an updated premium or carrier record can replace an assumption without rewriting the entire proposal. Record the plan identifier, geography, year, source and date alongside the amount. Mark preliminary inputs visibly.
Compare costs from both sides
The employer’s model should include contributions, administration, implementation and any related service costs. The employee’s model should include the premium remaining after the allowance, the policy’s cost-sharing structure and relevant coverage needs.
Use several scenarios rather than one average employee. A young employee in one county and an older employee in another may face different premiums and networks. An apparently stable employer budget can coexist with a substantial change in an employee’s experience.
The employee’s premium-tax-credit position requires its own evaluation. An ICHRA offer is not interchangeable with a cash raise that leaves Marketplace eligibility untouched. source
Translate the model into an operating schedule
Once the model is credible, assign responsibility for the plan documents, notices, employee education, enrollment, payment confirmation and issue resolution. HealthCare.gov’s guidance explains the ICHRA framework, including notice requirements; the administrator should explain how those requirements map to the actual launch date. source
Then test the workflow before launch with representative cases. Follow an employee who needs help choosing a policy, an employee who opts out and an employee whose first payment encounters a problem. These exercises reveal operational gaps that a contribution spreadsheet will not show.
What would make the decision ready?
A reviewable decision packet should contain the workforce assumptions, dated local plan evidence, contribution scenarios, employee impact, provider responsibilities and unresolved questions. Use the free evaluation brief to organize that packet.
This is a planning synthesis of external evidence, not a prediction that ICHRA will be cheaper or better for every employer in 2027. New rates, guidance or carrier information should trigger updates to the affected input and the conclusions that depend on it.
Sources and evidence
Review dates are recorded for each source above. Company pages are useful for confirming how a product is described, but they do not prove service quality or customer results.