The Disaster SEP Most Insurance Agents Never Check For

The client you don't know to call

On August 25, 2026, FEMA declared a major disaster for Indiana, DR-4933-IN, covering severe storms, straight-line winds, tornadoes, and flooding across 21 counties, with individual assistance designated for residents⁴. If you have even one client in those counties who missed a Medicare enrollment window this year, or who never finished an ACA application because the power was out and the internet was down, that client almost certainly qualifies for a Special Enrollment Period right now. Nobody sent you a list. Nothing in your CRM flagged it. The only way you'd know is if you happened to check FEMA's declarations page against your own book, county by county, this week.

Most agents don't do that, not because they don't care, but because nobody told them the rule exists in the first place, or because checking means opening a federal disaster feed and cross-referencing it against a spreadsheet by hand. So the SEP sits there, real and usable, until the four-month Medicare window or the 60-day ACA window closes, and the client either goes without coverage they were entitled to fix, or a competitor's agent, working the same disaster feed you weren't, calls first.

This isn't a rare edge case you'll hit once a career. FEMA declared 16 of these disasters with individual assistance in the first eight months of 2026 alone, an average of exactly one every 16 days⁴. If your book has clients in more than one state, the odds that at least one of them lives in a county that gets declared this year are not small. The gap isn't whether the SEP exists. It's whether anyone on your side is watching for it.

Part of why this stays invisible is structural. FEMA's job is disaster response, not insurance enrollment, so its declarations page is written for emergency managers and doesn't mention Medicare or the ACA anywhere on it. CMS's job is administering the SEP, not tracking which of your specific clients live in a newly declared county, so its guidance describes the rule in general terms and stops there. Neither agency owes you a personalized alert, and no third party sits in the middle connecting the two feeds for a working agent by default. The connection has to be made on your side, or it doesn't get made at all.

Why this SEP exists in the first place

CMS built this rule because enrollment deadlines don't pause for a hurricane. A Medicare beneficiary whose Aging and Disability Resource Center closed for two weeks, or an ACA applicant whose only internet access was the library that flooded, has a legitimate reason for missing a window that has nothing to do with indifference. The exceptional-circumstances SEP under 42 CFR 422.62(b)(4) and 423.38(c)(8)(ii) exists specifically for people who "were unable to and did not make an election during another valid election period as a result of the emergency or major disaster"¹.

The trigger is narrow and specific on purpose. To qualify, a person has to reside, or have resided at the start of the incident period, in an area FEMA has declared an emergency or major disaster for, with affected counties designated as eligible for individual or public level assistance. They also need to have had another valid election period open at the time, and to not have used it¹. The rule even extends to people who don't live in the affected area themselves but rely on a friend or family member there to help make healthcare decisions, which is a detail most agents have never heard, let alone used.

On the ACA side, CMS's current Special Enrollment Periods job aid for Navigators and Certified Application Counselors, version 3.0, dated March 2026, lists this under "Other Qualifying Changes": consumers who experience an exceptional circumstance, including a FEMA-declared national emergency or major disaster, that kept them from enrolling during OE or another SEP². CMS clarified this policy back in 2018, and its 2026 guidance names wildfires, hurricanes, and flooding specifically as the kind of events that trigger it, alongside the COVID-19 public health emergency as a past example².

Neither agency built this as a favor to insurance agents. CMS built it because the underlying enrollment deadlines are federal law and don't bend on their own, and a disaster SEP is the release valve that keeps a legitimate emergency from turning into a permanent coverage gap. The fact that it also happens to be an opportunity for an agent to reach a client who genuinely needs the call is incidental to why the rule exists, but it's exactly why it's worth knowing.

How often this actually happens

FEMA's OpenFEMA Disaster Declarations Summaries data shows 16 major disaster declarations with individual assistance in 2026 through August 28, spread across February, April, May, June, and August⁴. Those 16 declarations designated 191 counties and parishes across 12 states and territories, including Indiana, West Virginia, the Northern Mariana Islands, Mississippi, Louisiana, Michigan, Wisconsin, Montana, Arizona, Hawaii, Washington, and Tennessee⁴.

Some months carried no new declarations of this type at all, and some carried four. That unevenness is exactly why a manual, once-in-a-while check misses events: the pattern isn't a steady drip you can check quarterly, it's a cluster during storm season and a gap in between, and the gap is precisely when an agent stops looking.

Every one of the 16 disasters counted here carries a full major disaster (DR) designation. It's worth checking the other declaration type too: FEMA can also issue a lighter emergency (EM) declaration, and CMS's own rule covers "an emergency or a major disaster" equally, not major disasters alone¹. Querying the same OpenFEMA dataset for emergency declarations with individual assistance turns up zero for 2026 through August 28, so this year the full count and the major-disaster count happen to be the same number. That won't necessarily hold in a year with a different mix of events, which is one more reason to check the actual declaration type rather than assume.

Medicare vs. ACA: two different clocks

Whichever book you write, don't assume the Medicare version and the ACA version of this SEP run the same way. They share a trigger, a FEMA declaration with individual or public assistance designated, but the clock, the access point, and what counts as proof differ enough that using the wrong one costs a client their window.

What a missed window actually costs

Put a number on the Medicare side first, because it's the most concrete. A client who missed the Annual Enrollment Period because their county flooded in October and who never finds out about the disaster SEP stays on whatever plan they had, or has no Part D coverage at all, until the next AEP rolls around, up to a full year later. If that plan doesn't cover a medication they actually need, or carries a higher premium than a plan they would have picked with a clear head, that gap isn't abstract. It's a real coverage cost that a two-minute phone call, made in time, would have fixed.

On the ACA side the cost is often sharper because it's binary: a household that missed Open Enrollment because of a disaster and doesn't use this SEP simply goes uninsured for the plan year, absent another qualifying event. That's a heavier miss in 2026 than in a normal year: KFF found that letting the enhanced premium tax credits expire at the end of 2025 raises average annual premium payments for subsidized marketplace enrollees by 114 percent, from $888 in 2025 to $1,904 in 2026⁶, which makes a missed enrollment window a much more expensive gap to leave open than it would have been a year earlier.

There's a competitive cost too, and it's the one that should actually motivate you to build a habit around this. You are not the only agent who can see a FEMA declaration. If a competitor in your market runs even a rough manual check against a disaster feed and you don't, they get the call, not you, on a client who might have been in your book for years. The SEP doesn't belong to whichever agent originally wrote the policy. It belongs to whoever reaches the client first with the right information.

Put a rough range around the exposure at the agency level, not just the household level, and the reason to build a habit gets clearer. A single-office agency with a few hundred active households, spread across two or three counties, might have zero disaster-SEP-eligible clients in a quiet year and a double-digit number in a year like 2026, when 12 states and territories saw a qualifying declaration before September. An FMO or IMO with downline agents working a dozen states carries that same exposure multiplied by every county its book touches, which is exactly the kind of arithmetic that turns "check it sometimes" into "someone has to own this every week."

None of that is a precise forecast, and it shouldn't be treated as one. It's the honest shape of the problem: the wider your book's geography, the less realistic it becomes to trust a memory of "did anything happen near my clients" instead of an actual check against FEMA's list.

When this SEP doesn't apply

It's worth being precise about the edges of this rule, because misusing it wastes a client's time and yours. The disaster SEP doesn't apply if the client already used a valid election period on time. It's only for someone who had a window open and, because of the emergency or major disaster, didn't use it. A client who calmly re-shopped during a normal AEP and picked a plan has nothing to reopen.

It also doesn't apply outside the declared area. CMS's rule is tied specifically to residence in a county FEMA designated as eligible for individual or public level assistance during the incident period¹. A client one county over from a declared area, in a county that never got the individual assistance designation, doesn't qualify no matter how bad the weather looked on the news. This is the exact reason the county-level match matters more than a general sense that "there was a storm nearby."

And it isn't a plan-category loophole. On the ACA side, most disaster SEP enrollees who are already covered stay inside their existing plan's metal category the same way they would under most other SEP types, per CMS's plan category limitation rules, unless a separate qualifying change applies². The SEP reopens the window to enroll or change, it doesn't hand someone an unrelated upgrade.

How to check your own book

You don't need software to start. You need a habit and a source. Here's the method, run by hand, that works for a single-office agency with no automation at all.

List your book by county, not just by zip code. FEMA declarations are issued and designated at the county level, so a client record that only has a zip code makes matching slower and more error-prone than one that has the county name attached.

Pull FEMA's disaster declarations feed weekly. The public page at fema.gov/disaster lists every current declaration, and FEMA also publishes the same data through its OpenFEMA API, which you can query for free without an account⁴. Filter for declaration type "DR" (major disaster) and check whether individual assistance was designated. Weekly is frequent enough to catch a new declaration inside the same month it happens, which matters more on the ACA side where the window is shorter.

Match the declared counties against your client list. This is the manual step that actually takes time without a tool, and it's also the step most agents skip once the initial excitement of "I should check this" wears off after a month or two.

For every match, check whether the client actually missed a valid election period. The SEP only applies if there was a window open that the disaster kept them from using. A client who already enrolled or made their change on time doesn't need it, and telling them they do wastes both your time and theirs.

Confirm which deadline applies before you call. Four full calendar months from the incident start for Medicare, up to 60 days after the disaster ends for ACA. Know which one you're working with before you're on the phone, not during the call.

File it, and keep your documentation. The next section covers exactly what to have ready.

What CMS will actually ask for

Don't assume a disaster SEP is a rubber stamp. CMS's 2026 rules tightened verification generally: for plan year 2026 only, Exchanges on the federal platform must conduct pre-enrollment SEP verification for at least 75 percent of new enrollments across all SEP categories for consumers not already enrolled, a policy that CMS's own SEP Job Aid documents alongside an active piece of litigation, City of Columbus v. Kennedy, that has stayed some of the broader verification requirements while leaving loss-of-coverage verification in place².

Practically, that means you should have three things ready before you help a client file a disaster SEP: proof they lived in the declared county at the start of the incident period, evidence of the enrollment window they missed and why (a plan renewal notice, a documented AEP deadline, a Marketplace application started but not finished), and the specific FEMA declaration number and incident period dates. CMS gives consumers 30 days after plan selection to resolve any SEP verification issue that gets generated, so having the documentation ready up front avoids a pended enrollment sitting unresolved while the client thinks they're covered².

On the Medicare side, the process runs through 1-800-MEDICARE, the client's agent or broker, or the plan sponsor directly, and the beneficiary generally isn't required to prove they were personally affected beyond meeting the residency and missed-election criteria¹. Document it anyway. A missed premium payment or a lapsed enrollment that gets questioned later is much easier to sort out with a dated note in the file than without one.

If a client's ACA application does generate an SEP verification issue, CMS's SVI resolution process runs on a fixed sequence: the client submits the application with an SEP attestation, picks a plan, and gets a 30-day timer; the Marketplace sends a pended plan selection notice, then a warning notice at the 20-day mark if documents are still outstanding². Coach the client to upload documents rather than mail them, since CMS notes uploading is faster, and don't let the client assume they're covered just because they picked a plan. Until the SVI resolves, the enrollment sits pended and the client can't use the coverage yet.

Multi-county and multi-office agencies

Everything above assumes one person checking one book. That breaks down fast once an agency runs more than one office, or an FMO oversees a downline of agents who each carry their own book across different counties and, sometimes, different states. Nobody at the top of that structure has a single view of which counties the whole downline actually covers, so a FEMA declaration in a state where the FMO itself has no office can still be sitting on top of a dozen downline agents' clients without anyone at the FMO level ever seeing it.

The fix at that scale isn't "everyone checks their own book harder." It's a single county list, built once from wherever the downline's client data actually lives, checked against the FEMA feed in one place, with matches routed back down to whichever agent owns that client. That's a data and workflow problem before it's a compliance problem, and it's the same shape of problem as reactivating a dormant database or automating appointment reminders: real records, matched against a real external trigger, routed to the person who should act on it.

It also compounds the AEP timing pressure most agencies already feel. CMS's federal call recording rule change cut Medicare call retention from 10 years to 6, effective October 1, 2026, right before AEP opens, which is a separate compliance shift covered in our guide to that rule. A downline managing that change and a disaster-SEP check by hand, in the same six weeks, is exactly the kind of season where a manual process quietly stops happening.

How we automate the check instead

The manual method above works. It just doesn't scale, and scaling it by hand is exactly the kind of repetitive, data-matching work that shouldn't eat a producer's selling hours. That's the gap our custom builds close.

What we build for an agency that wants this automated is specific to that agency's actual book and CRM, not a generic add-on⁵. You describe the workflow, we build it on your accounts and your domain, and you own it when it's done⁵. For this particular problem, that typically means a workflow that pulls FEMA's disaster declarations on a schedule, matches the declared counties against your client list by county, and flags the matches for you to review and call, so the FEMA-checking step stops depending on someone remembering to do it every week.

It's the same underlying idea behind the database reactivation and appointment automation work we build for agencies generally: real data, matched against your real book, surfaced as a specific action instead of a general reminder to "check your list." Ambrose, our AI platform, keeps context per contact across calls, texts, and notes, which is what lets an automation like this attach the right client history to a flagged match instead of just handing you a bare name and county.

Concretely, that means the workflow pulls the FEMA declarations feed on a schedule you set, cross-checks the newly designated counties against your book's county field, and produces a short, specific list: this client, this county, this declaration, this deadline. It doesn't file anything or contact the client on its own. It hands you exactly what you need to make the call yourself, which keeps the actual enrollment decision, and the relationship, where it belongs.

What actually changes

Not a promised enrollment count or a guaranteed number of saved clients, because nobody can honestly promise you that. What changes is concrete and specific to this one problem.

You stop finding out about a disaster SEP by accident, or not at all. Clients in a declared county who missed a window get a call from you before the window closes, not a letter from Medicare.gov they never read. And you're the agent in your market who actually knew the rule existed, which is a smaller advantage than a lower premium but a real one, in a business where most competitive edges come down to who called first.

Questions agents ask