Medicaid work requirement: how to stress-test a behavioral health revenue forecast
How behavioral health finance teams can build a revenue range by payer, program and site before the Medicaid work requirement starts on 1 January 2027.
From 1 January 2027, certain Medicaid applicants and beneficiaries aged 19 to 64 must meet a work requirement unless they are excluded. On 13 November 2026, the Centers for Medicare and Medicaid Services (CMS) announced an optional hold-harmless period through 30 June 2027 for states that implement by 1 January. Your board wants to know what it does to next year’s numbers. The honest answer is that no one can give a single reliable figure yet, and a range with named assumptions will hold up better in the boardroom than a point estimate.
This post walks through how to build a defensible revenue range from payer, program, and site data, which assumptions to flex, which monthly signals to watch, and how to frame the range for the board.
Key takeaways
- The work requirement is one variable in a layered forecast. Build a range from payer, program, and site rather than a single policy-driven number.
- A defensible stress test flexes four assumptions independently: enrollment, payer mix, visit volume, and collections timing. Directional ranges matter more than fabricated percentages.
- The signals that tend to move before the revenue line are eligibility churn in your Medicaid book, average days to first paid claim, denial reasons tied to eligibility, and payer-mix drift at the program and site level.
- Bring the board mild, base and severe scenarios with the assumptions named. Under a policy change this large, a single-point forecast hides the range the board needs to see.
- Before the next budget cycle, ask whether your finance stack can tag revenue by payer, program, and site in one place, reforecast in days rather than weeks, and connect financial data to encounter and labor feeds without rebuilding the workbook each month.
What we’ll cover
- Why a single-number forecast fails here
- How to decompose the forecast by payer, program, and site
- The four assumptions to flex independently
- Monthly signals to watch before the revenue line moves
- How to present the range to the board
- Evaluating your finance stack for this kind of scenario work
- Frequently asked questions
Why a single-number forecast fails here
A policy change that touches Medicaid enrollment creates uncertainty at every layer of your revenue model simultaneously. Enrollment may fall, but by how much depends on how aggressively your state implements the requirement, how many of your current enrollees are already working or exempt, and how quickly your outreach team can help affected clients document their status. Payer mix shifts as some Medicaid enrollees move to self-pay or commercial coverage, or disenroll entirely. Visit volume may drop, hold steady, or even rise if crisis and outpatient demand increases among people navigating the transition. Collections timing may slip if managed care organizations (MCOs) see more eligibility-related denials. The CMS announcement fixes the dates. In states that use the hold-harmless period, applications won’t be denied and otherwise eligible people won’t be disenrolled solely for not meeting the requirement before 30 June 2027, so the timing of any impact depends partly on your state’s choice. The rest of the forecast rests on your own assumptions.
Each of those variables moves on a different timeline and in a different direction depending on your program mix and geography. A single number collapses all of that into one assumption, and when the assumption is wrong, the board loses confidence in the finance function. A range with named assumptions is harder to present but far more defensible.
How to decompose the forecast by payer, program, and site
Start by pulling your trailing revenue into three dimensions: payer, program, and site. Payer means Medicaid fee-for-service, Medicaid MCO by plan, commercial, self-pay, and grants. Program means outpatient, residential, medication-assisted treatment (MAT), crisis services, and any other distinct service line. Site means each physical or virtual location where you deliver care.
Once you have that matrix, identify which cells are most exposed. A residential program at a single site where Medicaid MCO revenue is the dominant payer carries a different risk profile from a multi-site outpatient program with a mixed payer base. The decomposition also tells you where a shift in one cell is likely to cascade: if your MAT program is heavily Medicaid and your crisis program draws from the same population, a drop in MAT enrollment may show up in crisis volume before it shows up in MAT revenue.
For background on how the funding streams interact across a behavioral health revenue mix, see the behavioral health funding streams post.
The four assumptions to flex independently
Once you have the decomposed matrix, build three scenarios by flexing these four assumptions one at a time before combining them:
Enrollment. How many of your current Medicaid enrollees are in the age range affected by the work requirement, and what share might disenroll or lose coverage? Use your own eligibility data, not a national average. Conservative, base, and severe values should reflect the range of outcomes your outreach and enrollment staff consider plausible, not a single midpoint.
Payer mix. Of those who disenroll from Medicaid, what share moves to commercial coverage, self-pay, or no coverage? A shift to self-pay has a different revenue impact than a shift to commercial, and both are different from a straight volume loss. Model each path separately.
Visit volume. Does a reduction in enrolled clients reduce visit volume proportionally, or does demand from other payer segments partially offset it? Crisis and walk-in programs often see demand that is less sensitive to enrollment status than scheduled outpatient programs.
Collections timing. Eligibility-related denials slow the revenue cycle. Even clients who remain enrolled may hit eligibility verification failures once enforcement begins in your state, lengthening average days to first paid claim and compressing cash flow before the revenue line reflects the full impact.
Present the scenarios as directional ranges. The board needs to understand the shape of the risk across each assumption, and a fabricated percentage obscures that.
Monthly signals to watch before the revenue line moves
Four indicators tend to move before your headline revenue line, and tracking them monthly gives you time to adjust the forecast before the board meeting rather than after.
Eligibility churn in your Medicaid book. Track the number of active Medicaid clients whose eligibility status changes each month, broken out by MCO and program. A rising churn rate is the earliest signal that enrollment is shifting.
Average days to first paid claim. When eligibility verification failures increase, this metric often lengthens early. A trend of even a few additional days across your Medicaid book is a leading indicator of a collections timing problem.
Denial reasons tied to eligibility. Separate eligibility-related denials from clinical and coding denials in your monthly denial analysis. A rising share of eligibility denials, even at stable total denial rates, is the cue to check whether the work requirement is beginning to affect your revenue cycle.
Payer-mix drift at the program and site level. Watch the share of visits by payer at each program and site, not just in aggregate. A shift that is invisible at the organization level may be concentrated in one program or one site, and catching it early lets you adjust staffing and capacity before the financial impact compounds.
Behavioral Health Resources (BHR) faced a comparable shift when Washington state prepared to move from county-run distribution of behavioral health funding to an integrated model in which managed care organizations process Medicaid claims. “We knew the new MCO model would create all sorts of new hoops we had to jump through for Medicaid claims, increasing our billing workload substantially,” said Eric Jensen, director of finance. BHR was running a version of Microsoft Dynamics GP that was no longer supported, and it moved to Sage Intacct ahead of the change.
How to present the range to the board
Structure the board presentation around three scenarios, mild, base and severe, and for each one name the enrollment, payer mix, visit volume, and collections timing assumptions explicitly. The board needs to understand the range of outcomes, the assumptions behind each scenario, and what would have to be true for the severe scenario to materialize.
Then name the triggers: what monthly signal, at what threshold, would cause you to move from the base to the severe scenario in your operating plan? That framing turns the forecast from a prediction into a decision framework, which is what the board needs from finance under uncertainty.
Present the work requirement as one variable among several. The board should understand that payer-mix drift, MCO contract terms, and grant funding cycles are all moving at the same time, and that the stress test is designed to hold those variables separately so the organization can respond to each one as it develops.
Evaluating your finance stack for this kind of scenario work
Before the next budget cycle, ask your team three concrete questions about your current finance infrastructure.
First, can you tag revenue by payer, program, and site in one place, without exporting to a spreadsheet and rebuilding the matrix each month? If the answer is no, the decomposition work described above becomes a manual project rather than a standing report, and the monthly signal tracking becomes impractical at the cadence it requires.
Second, can you reforecast in days rather than weeks? A policy environment that is still evolving through the hold-harmless period means your assumptions will need to change more than once before the end of the first half of next year. A finance team that needs three weeks to reforecast will struggle to keep pace with that.
Third, can you connect financial data to encounter data from your electronic health record (EHR) and labor data from payroll without rebuilding the workbook each month? Payer-mix drift at the program level is far easier to see when financial and clinical data sit in the same view. At Health Solutions, a community behavioral health center, the old software couldn’t produce a monthly balance sheet, and allocations for unit cost reporting were such a headache that the chief financial officer (CFO), Paige Oldham, did them only once a year, when the state required it. Under constraints like those, monthly signal tracking becomes very hard to sustain.
These are evaluation questions for your current infrastructure. If the answers reveal gaps, address them before the work requirement takes effect. Explore how healthcare organizations run finance on Sage Intacct to see what multi-dimensional visibility looks like in practice.
If your current finance stack cannot decompose revenue by payer, program, and site or reforecast quickly enough to keep pace with a changing policy environment, now is the time to evaluate what a more connected system would make possible.
Frequently asked questions
What is the Medicaid work requirement that takes effect in 2027?
From 1 January 2027, certain Medicaid applicants and beneficiaries aged 19 to 64 must meet a work requirement unless they are excluded. On 13 November 2026, CMS announced an optional hold-harmless period through 30 June 2027 for states that implement by 1 January. Eligibility and compliance questions should be directed to your legal and policy advisors; the finance team’s job is to model the revenue impact across scenarios.
How should a behavioral health CFO model the revenue impact of the work requirement?
Decompose your revenue by payer, program, and site, then flex four assumptions independently: enrollment, payer mix, visit volume, and collections timing. Build mild, base and severe scenarios with the assumptions named, and present a range to the board rather than a single number.
Which leading indicators signal Medicaid enrollment risk before it hits the revenue line?
Watch eligibility churn in your Medicaid book, average days to first paid claim, the share of denials tied to eligibility reasons, and payer-mix drift at the program and site level. Each of these tends to move before the headline revenue figure, giving the finance team time to adjust the forecast.
What should a behavioral health finance team present to the board under this kind of policy uncertainty?
Present three scenarios, mild, base and severe, with the enrollment, payer mix, visit volume, and collections timing assumptions named for each. Define the monthly triggers that would move you from one scenario to another, so the board has a decision framework rather than a single prediction.
What finance system capabilities matter most for scenario planning under a Medicaid policy change?
The ability to tag revenue by payer, program, and site in one place; to reforecast in days rather than weeks; and to connect financial data to encounter and labor feeds without manual workbook rebuilds. These capabilities determine whether monthly signal tracking is a standing report or a manual project.