Federal Court Appeals: When to Take an SSDI Case Past the Appeals Council

Once the Appeals Council denies review, the administrative process is over. The only thing left is a civil action in federal district court — a different forum, a different standard, and a different economic calculus than anything that came before. The question is rarely whether you can file. It's whether this particular loss is one worth filing.

That decision is a screening problem, not a litigation problem. The cases that win at the district-court level are the ones with identifiable legal error in the ALJ's decision, surfaced before you commit the hours. Here's the framework for making that call.

The procedural gateway

Three conditions have to line up before a district court will hear you. First, you need a final decision — the Appeals Council either denied review of the ALJ's unfavorable decision or issued its own. Second, you have 60 days from receipt of the Appeals Council's notice to file the complaint, with the regulatory presumption that notice arrives five days after the date on the letter. Third, you file a civil action in the district where the claimant resides.

The 60-day clock is jurisdictional in practice and runs hard. If the deadline is tight, an extension request to the Appeals Council is the only safety valve, and it has to be filed for good cause before the period lapses. Treat the docketing of that deadline the way you'd treat a statute of limitations on any other matter.

Standard of review — substantial evidence and legal error

District courts do not retry the claim. The court reviews the administrative record on two grounds: whether the Commissioner's findings are supported by substantial evidence, and whether the decision applied the correct legal standards. Substantial evidence is the deferential prong — "more than a scintilla," enough that a reasonable mind might accept it. You rarely win by arguing the ALJ weighed the evidence wrong.

You win on the legal-error prong. An ALJ who failed to evaluate a medical opinion under the persuasiveness factors, ignored a line of evidence, posed a hypothetical that omitted credited limitations, or never resolved a vocational conflict has committed legal error a court can correct regardless of how much evidence sits on the other side. Case selection means reading the decision for those failures, not relitigating the diagnosis.

Sentence-four vs. sentence-six remands

The two remand types come from the same statute and mean very different things for your case and your fee.

A sentence-four remand accompanies a judgment affirming, modifying, or reversing the Commissioner's decision — the court finds error and sends the case back for further proceedings. This is the typical disability remand and, importantly, it makes your client a prevailing party, which opens the door to fees under the Equal Access to Justice Act.

A sentence-six remand is narrower: the court orders the agency to consider new and material evidence that there was good cause for not submitting earlier, or the Commissioner requests remand before answering. There's no merits judgment, so prevailing-party status — and the EAJA fee that rides on it — doesn't attach the same way. When you're scoping a case, know which remand the facts actually support; it changes the economics.

The remand / affirm / dismiss base rates

This is where case selection should be data-driven rather than anecdotal. Federal-court SSDI outcomes resolve into a few buckets — remand, affirmance, voluntary dismissal — and the remand share is meaningful but far from automatic. GAO's review of court dispositions documented both the volume of cases reaching federal court and the persistent share returned to the agency, with remand rates that vary by court and over time rather than holding to a single national number.

The driver behind a remandable case is almost always the type of error in the ALJ decision, and those errors cluster. RFC-formulation and evidence-evaluation failures, vocational/Step-5 problems, and symptom-evaluation shortcomings recur far more often than novel legal questions. If you can match the error in front of you to the categories that courts actually remand on, you're selecting from the high-probability pool instead of guessing.

Case selection — which losses are worth federal court

Run every Appeals Council denial through the same filter:

• Is there an isolable legal error? A clean, citable failure — uncredited opinion evidence, an unresolved DOT conflict, an RFC untethered to the record — beats a "the ALJ got it wrong overall" theory every time.
• Does the record support the remand type you'd be arguing? Sentence-four error on the existing record is the standard play; reserve sentence-six for genuinely new, material evidence with good cause.
• What does the error category's base rate look like? Weight your time toward errors that historically remand, not toward sympathetic facts that don't translate to legal error.
• Is the forum hospitable? Remand tendencies differ across districts and over time; know where you're filing.

A disciplined version of this filter is exactly what keeps a federal-court docket profitable: you're not appealing losses, you're appealing errors, and you can see the base rates before you commit.

Fee economics — EAJA and 406(b) make marginal cases viable

The reason federal-court work pencils out even on hard cases is the two-layer fee structure. A sentence-four remand that makes your client the prevailing party supports an EAJA fee, paid by the government and tied to your reasonable hours — recovery that does not come out of the claimant's benefits. Then, if the remand ultimately produces an award of past-due benefits, a Section 406(b) fee of up to 25% of past-due benefits, paid from the award and subject to the court's reasonableness review, compensates the merits work. Where both apply, you refund the smaller of the two to the client.

That structure is what lets a firm take cases that wouldn't survive a pure contingency analysis at the administrative level. The administrative fee-agreement cap of $9,200 governs the agency side; the court side runs on its own EAJA-plus-406(b) track. The marginal case becomes viable precisely because a remand can pay twice — once for the hours, once for the result.

None of that changes the threshold question. The fee economics only reward you if the case remands, and the case only remands if the legal error is real. Build the case-selection discipline first; the economics follow from it.

When you're triaging a batch of hearing-level losses, run them against the appeal-decision tool to weigh remand likelihood by error type, and use the judge data to understand the forum the decision came out of. For upstream intake — scoring claims before they ever reach this stage — the Lead Scorer applies the same evidence-first logic.