Community engagement requirements for Medicaid expansion adults go into effect January 1, 2027. For most states, time has run out to address this through a new procurement: Notices about the requirement have already gone to enrollees, eligibility system contracts were renewed months ago, the integration backlogs of those systems run through go-live, and fiscal year budgets are solidified.
When community service is identified not only as the most broadly available pathway but also the most unstandardized one to track in a state’s community engagement build, the more useful response is to work through three practical questions: what will support it, how will it be paid for, and when can it be live?
All three questions have answers that do not require a new procurement.
The contracted layer is the one every state already has. Eligibility determination, case management, notices, renewals, document intake and caseworker workflow all sit within the eligibility and enrollment system. That layer is contracted, staffed, and not moving because those systems are built to make determinations and keep records. While they can evaluate the information submitted to them, they were not designed to confirm that an activity took place.
The missing layer is everything that has to happen before a claim exists. Making opportunities available, matching participants to them, supporting service delivery and capturing hours at the point of service are a different kind of work. Until H.R. 1, no state program required an agency to know whether a person physically showed up somewhere to perform unpaid work.
That distinction is central. Asking an eligibility system vendor to stand up a statewide inventory of volunteer opportunities, match participants to them, and capture attendance in the field is asking a capable partner to build outside its competence, at custom-build cost, against a fixed date. The alternative is not replacing that vendor. It is adding a layer that the system was never scoped to hold.
Managed care plans come up as an alternative here. They are a genuine asset for outreach and member engagement, and states should use them that way, but plans cannot make compliance determinations.3 They also cannot create the volunteer opportunities themselves. A plan can tell a member the requirement exists, but it cannot create the shift, capture the hours, or write a verified record into the eligibility system.
The added layer needs five capabilities, each tied to a specific obligation in the interim final rule rather than to a feature list.
| Capability | The Obligation It Serves |
|---|---|
| Live inventory of qualifying opportunities from public and nonprofit organizations | The rule requires a host organization to exist, to provide oversight, and to run a process that tracks activity type, dates and hours with a point of contact who can confirm them. Making opportunities available and tracking them properly go hand in hand, for both in-person and virtual activities |
| Personalization by location, ability and credential | A state that mandates 80 hours without reachable opportunities in a given county is generating procedural disenrollments rather than engagement |
| Automatic capture at the point of service | Ex parte verification requires data the agency can act on without additional action from the applicant. However, hours reported after engagement are not that |
| Identity verification and credentialing on the record | Determinations are appealed and programs are audited. Attestation and scanned documents are thin evidence in both settings |
| Real-time bidirectional write back to the system of record | Verification is required at application and at renewal, and with variation across states. Each cycle that cannot clear automatically becomes a notice and a 30-calendar day response window |
Note what is not on that list: no new eligibility engine, no parallel case management, and no migration. In its design guidance for states, the Center on Budget and Policy Priorities warns specifically against setting up duplicative processes for the work requirement alone, on the grounds that a separate step generates significant additional work for individuals and eligibility workers and can lead to applicant or enrollee drop-off.2 The added layer has to be embedded within the existing workflow or it makes the problem worse. The distinction that matters is between a community engagement layer that extends an existing environment and a second system a participant has to find on their own.
In practice, this is a well-understood integration pattern. The participant-facing experience is branded to the state, so the program remains the state’s rather than a vendor’s. Authentication runs through the state’s single sign-on (SSO). Opportunity data, attendance records, credentials and verified hours move over the platform’s application programming interface (API), with native connectors where the state or its partners already run a commercial system of record.
Verified hours land in the existing record of a verified individual beneficiary. There is no second database of truth, no reconciliation job, and no separate portal that a beneficiary has to find and learn on their own. For county-administered states, the same layer supports role-based access and sub-accounts, so each jurisdiction runs its own program under a central roll-up.
Because the integration points are deliberately limited, the timeline is realistic.
H.R. 1 appropriated $200 million in one-time fiscal 2026 grants, named the Rural Health Transformation fund, to the states and the District of Columbia, half divided equally and half distributed among expansion states in proportion to the residents subject to the requirement. A separate $200 million went to the Centers for Medicare & Medicaid Services (CMS) for federal implementation. Divided across states, that funding is unlikely to cover a state’s community engagement build on its own.3 4
The more consequential mechanism is the enhanced federal match, and the rule answers this rather than leaving it to interpretation. Its preamble tells states that where they must design, develop or install new or enhanced eligibility and enrollment systems to implement the requirement, those activities may be eligible for 90 percent federal financial participation (FFP) under 42 CFR 433.112, with 75 percent for ongoing operations.1 6 CMS built its own cost estimate on that assumption. To draw the enhanced match, a state files an advance planning document (APD) describing the project and its cost. Reviewing those filings from 21 states, CMS put the average reported eligibility system cost at $12.2 million, projected a one-time $15 million per state, and assumed 90/10 on implementation and 75/25 on maintenance throughout.1
Three details determine whether a layer that is purchased rather than custom built qualifies. Licensed software is not excluded: Section 433.112(c)(2) counts initial licensing of commercial off-the-shelf (COTS) software, plus the necessary costs to analyze, install, configure and integrate it, as design and development.6 A subscription is split rather than disqualified: CMS guidance directs states to allocate license cost across the license period, match the share attributable to implementation at 90 percent and the operating share at 75 percent, and states that those rates apply equally to COTS and all commercially available software including software as a service.7 Ownership works the way a vendor relationship requires, with the vendor retaining the underlying product and the state owning the configuration and customization built with enhanced funds.7
The binding constraint is procedural. Section 433.112(a) conditions the 90 percent rate on CMS approving the state’s APD before the state spends the money, and the rule repeats that in its own words.1 6 A state that contracts first and files afterward has a claiming problem; it does not matter how well the work fits the definition because the approval has to come before the spending, not after it.
For most states, this is not a filing they can assume is already handled. CMS found that roughly half of states have not reported any community engagement system costs in their APD at all.1 The first step is to pull the state’s current APD and see whether this work is described in it. If it is not, the APD needs an update filed and approved before the contract is signed. That sequencing, rather than the merits of the project, is what most often costs a state the 90 percent rate.
However, one qualification materially changes the math: The enhanced rates apply to systems costs. General Medicaid administrative expenditures, meaning staff, caseworkers, call centers, notice printing and mailing, and outreach, draw 50 percent of cost.4 That is where most states’ unbudgeted cost is landing, and a plan built on the assumption that 90/10 covers implementation generally will be wrong by a wide margin.
This is the part most vendor material skips, yet it often determines whether a program can move forward. The difference between a 90 percent federal share and a state general fund line is often what decides whether a project happens this year.
The same analysis notes that consolidating everything with a single vendor may reduce a state’s immediate administrative effort but can also raise costs and diminish control over implementation outcomes and vendor accountability.2 A narrowly scoped layer for one pathway, integrated into an incumbent environment, avoids that risk.
Making opportunities available is the one component that cannot be procured on a fixed timeline. A state cannot contract with every food bank, shelter, faith organization, animal rescue and service commission within its borders before January 1, and should not try.
The workable path is a platform that community organizations can adopt themselves, and that many already use to run their own volunteer programs. Those organizations get the tracking capability the rule now requires of them, and their qualifying opportunities become visible to participants in the state’s program.
That reverses the sequencing problem. It allows available opportunities to grow through the relationships a state and its service commission already have, rather than through a procurement cycle that does not fit inside the remaining calendar.
Standing it up takes four workstreams. They run in parallel rather than in sequence, so none of them waits on another to finish.
Golden is commercially available today rather than in development, and can be engaged directly or through many existing partners, contract vehicles, or reseller relationships. Golden’s technology is turnkey and can run in parallel with existing systems. When integrated with a state’s eligibility system, the time to go live depends on the state’s integration queue and on how quickly a technical owner is assigned, which is generally the binding constraint rather than anything on the vendor side.
The rule allows for a state experiencing genuine systems barriers to seek temporary relief, supported by a documented implementation plan showing steps taken, the barriers encountered, and a milestone timeline.1 That provision exists for good reason, and some states will need it.
Exemptions run six months at a time, require quarterly progress reporting, can be terminated for a missed report, and all of them expire no later than December 31, 2028 with no renewal.5 CMS’s own impact analysis assumes roughly two states are approved. This is not a broadly available off-ramp.
Requesting a Good Faith Exemption also means negotiating from a difficult position, and it draws federal attention to the parts of a program that are not yet working. A state that can demonstrate a functioning closed loop for community service, in which it has been integrated into its environment, does not need to have that conversation.
Community service, the only pathway with no national data source behind it, is the one most likely to lead to coverage losses in 2027. It is also the one that can be closed without touching the rest of the build.
If it would help, we will put together a one-page integration outline for your environment. Rather than a generic diagram, it will name the eligibility system you run and show where verified community service hours would enter the record. No demo is required.
Anish Das,
Chief of Staff
anish@goldenvolunteer.com
1. Medicaid Program; Community Engagement Requirement for Certain Individuals. Centers for Medicare & Medicaid Services, Federal Register, June 3, 2026. Accessed September 1, 2026. https://www.federalregister.gov/documents/2026/06/03/2026-11094/medicaid-program-community-engagement-requirement-for-certain-individuals.
2. Assessing the Medicaid Work Requirement Vendor Landscape. Center on Budget and Policy Priorities, 2026. Accessed September 1, 2026. https://www.cbpp.org/research/health/assessing-the-medicaid-work-requirement-vendor-landscape.
3. Preparing for Medicaid Community Engagement Requirements: Key Steps and Opportunities for States and Plans. Health Management Associates, 2026. Accessed September 1, 2026. https://www.healthmanagement.com/blog/preparing-for-medicaid-community-engagement-requirements-key-steps-and-opportunities-for-states-and-plans/.
4. H.R. 1 Medicaid Implementation: An Administrative Cost Monitoring Guide for States. State Health and Value Strategies. Accessed September 1, 2026. https://shvs.org/h-r-1-medicaid-implementation-an-administrative-cost-monitoring-guide-for-states/.
5. Medicaid Community Engagement Requirement for Certain Individuals Interim Final Rule with Comment Period (CMS-2454-IFC). Centers for Medicare & Medicaid Services. Accessed September 1, 2026. https://www.cms.gov/newsroom/fact-sheets/medicaid-community-engagement-requirement-certain-individuals-interim-final-rule-comment-period-cms.
6. 42 CFR part 433, subpart C: Mechanized Claims Processing and Information Retrieval Systems. Electronic Code of Federal Regulations. Accessed September 11, 2026. https://www.ecfr.gov/current/title-42/chapter-IV/subchapter-C/part-433/subpart-C.
7. Mechanized Claims Processing and Information Retrieval Systems: Enhanced Funding (SMD 16-004). Centers for Medicare & Medicaid Services, March 31, 2016. Accessed September 11, 2026. https://www.medicaid.gov/federal-policy-guidance/downloads/smd16004.pdf.
Of the four ways Medicaid expansion adults can meet H.R. 1’s 80-hour monthly requirement, three can be verified against data states already hold. Community service can’t: no national system records who volunteered where, or for how long. Here’s why that gap carries the most risk, and what closing it takes.
‹ Previous Post
Volunteer hours have no data source behind them the way wages and school enrollment do, so states are verifying them three ways: self-report, document upload with caseworker review, or capture at the point of service. Here’s how the three compare, and which can clear a renewal without contacting the beneficiary.
Next Post ›1:1 consultation aimed at transforming your volunteer program
Learn how to transform supporters into donors
Help us spread our message