Methodology

Data: CMS public enrollment files, in a monthly snapshot processed by TapQuality. Senior Deal Club is not affiliated with Medicare or any government agency.

The Centers for Medicare & Medicaid Services (CMS) posts enrollment files each month. TapQuality turns them into a monthly snapshot. Senior Deal Club reads that snapshot and builds every chart and sentence on these pages from it. We do not type numbers in by hand.

The snapshot is as of September 2026 (CMS report date 2026-09-01).

CMS counts Medicare Advantage members in more than one file. Each file answers a slightly different question. Every figure on these pages carries a label for its file, and we never put two files in one ratio or one chart.

County file (SCP)

CMS's state, county and plan-type file. All Medicare Advantage plans, with employer and union group plans included. We use it for state counts, national trends and 12-month change. 35,849,071 members in September 2026.

Contract file (SCC)

CMS's state, county and contract file. All Medicare Advantage contracts, with employer and union group plans included. We use it for carrier shares and for the headline total. 36,183,289 members in September 2026.

Individual plans (IND)

Individual Medicare Advantage plans from county data. Employer and union group plans, which a retiree cannot pick on their own, are left out. We use it for county tables and for the individual-plan change by state. 30,149,164 members in September 2026.

Plan by county (CPSC)

CMS's plan-by-county file. It is the most detailed file. This tracker does not use it yet.

Geographic Variation (GV)

A CMS file of Medicare use by county. We take the share of Medicare beneficiaries in Medicare Advantage from the 2024 edition. It is older than the member counts, so we print its year and never divide current members by it.

In September 2026 the contract file counts 36,183,289 members, the county file 35,849,071, and the individual-plan data 30,149,164. They differ for three reasons.

• The county file has no rows for Alaska and Connecticut in recent months. The contract file does.
• Individual-plan data leaves out employer and union group plans.
• CMS hides counts of 10 or fewer in each file.

None of the three is wrong. Quote the one that matches your question, with its label.

Alaska and Connecticut have no rows in the state file this tracker reads after July 2026. The contract file still lists them (September 2026: Alaska 2,273; Connecticut 446,529). This is a hole in one CMS file, not a fall in enrollment. The published national total falls 375,623 in August 2026; that drop is a gap in the source file, not a loss of members. The national line on these pages sums the 53 jurisdictions that report in every month, and those states show "no data after July 2026". Year over year, the panel reads +2.0% and the all-state contract file +2.0%.

For that reason the national line on these pages adds up only the 53jurisdictions that report in every month. We call it "same jurisdictions each month." The chart on the overview page also shows the total as CMS publishes it, so you can see the gap.

Missing months: July 2023 and December 2023 are absent from the state and county file; July 2023 is absent from the contract file. Lines break at a gap. Nothing is filled or averaged across it. Year-over-year changes compare the same calendar month, so July 2023 and December 2023 cannot be compared.

County file: July 2023 and December 2023. Contract file: July 2023. We do not fill, average or bridge a missing month. Month-over-month change is left blank across a gap.

Counts and shares step at each January, when plan-year contracts and annual elections take effect. A 12-month change is mostly one January step plus drift. CMS posts January counts late, so compare November with February or March, not January alone.

CMS also renumbers many contract IDs each January. Between 15% and 23% of members change contract from December to January, against under 1% in other months. A contract ID therefore cannot be followed across years. This tracker follows carriers, not contracts. If we ever cite a single contract, we name the file the figure comes from, because two CMS sources can differ by about 2% for the same contract.

Insurer not assigned in CMS data: 0.6% of members in September 2026 (225,918), against 10.7% in January 2023. These are mostly retired contracts with no carrier label. Shares divide by members with a label so old and new months compare. When the unassigned bucket is above 3% in either month, treat share changes as rough.

Carrier names are the labels in the data. Some carriers appear under several labels, and we do not merge them. The 22 Blue Cross Blue Shield labels stay separate, and one carrier name changed: "California Physicians' Service" is shown as Blue Shield of California. The level of market concentration depends on how carriers are grouped, so we show only whether it rose or fell. A contract that serves several states carries one carrier label, so a carrier can show members in a state where it is not a local plan. Small states are the most affected.

• Share.A carrier's members divided by members who have a carrier label, in the same file and month.
• 12-month change. A month against the same calendar month a year earlier. Seasonal swings cancel out.
• Net adds, November to February.February members minus the prior November's, so one late January posting does not decide the answer.
• Same jurisdictions each month. The national line adds only the jurisdictions that report in every month.
• Indexed lines.A state's members divided by its members in the start month, times 100, beside the same calculation for the nation.
• Individual-plan change by state. Rebuilt from county changes. State changes on the individual-plan basis are rebuilt from county changes. Where a state's individual plan ended in many counties (Vermont), the rebuilt change understates the drop and the state shows no percentage.
• Rankings. We leave American Samoa, Guam, the Northern Mariana Islands and the U.S. Virgin Islands out of rankings, because a handful of members moves them by whole percents.

We rebuild the snapshot once a month, after CMS posts the new file. Before a new snapshot goes live it must pass checks: every state still present, history older than three months moves less than half a percent, and carrier shares add up to the whole. If a check fails, the old snapshot stays.

• CMS hides counts of 10 or fewer, so plan and county totals are lower bounds.
• Small states and territories move on one plan. Read their percents with care.
• Member counts say who enrolled, not why. A change next to an event is not proof the event caused it.
• Counts are people enrolled, not people who use their plan, and not what a plan costs or covers.
• Share of Medicare in Medicare Advantage comes from the 2024 Geographic Variation file and is older than the member counts.
• Counts for the latest month can change when CMS revises a file. We update the pages on the next monthly build.

If a figure looks wrong, email support@seniordealclub.com with the page and the number. We check it against the CMS file. When we find an error we fix it, update the changelog below, and say what changed. Our editorial policy covers how we handle corrections across the site.

Senior Deal Club is a free educational site. It may refer readers to licensed insurance agents. These tracker pages carry no call buttons and no plan sales offers. Read the editorial policy for how we separate education from advertising.

Senior Deal Club is not affiliated with Medicare or any government agency. Pages are by the Senior Deal Club Standards Desk.

• CMS Medicare Advantage/Part D Contract and Enrollment Data. From this page: Monthly Enrollment by Contract/Plan/State/County (CPSC), MA Enrollment by State/County/Contract (SCC) and MA Enrollment by State/County/Plan Type (SCP).
• CMS Medicare Geographic Variation. Share of beneficiaries in Medicare Advantage by county.

• September 2026: snapshot updated with data through September 2026. First release of the tracker.