Appeals Council Practice: When Review Beats Re-Filing
After an ALJ denial, you choose between Appeals Council review and a fresh application. Here's how to read the AC standards, the base rates, and the onset-date math that should drive that call.
The unfavorable ALJ decision lands and the clock starts. You have two roads: request Appeals Council review on the existing record, or advise the client to file a new application and start over. They are not interchangeable, and the wrong reflex — "just re-file, it's faster" — can quietly forfeit years of back pay. The choice is a structured one, and the AC's own standards plus its workload data tell you most of what you need to make it.
This is a practitioner's read on when review beats re-filing, how to write the brief so it survives into federal court, and how to protect onset and entitlement across the whole decision tree.
The Appeals Council standards — what review is actually for
The Appeals Council is an error-correction body, not a second hearing. Under 20 CFR 404.970, it will grant review when there is an apparent error of law, when the ALJ's action, findings, or conclusions are not supported by substantial evidence, when there is a broad policy or procedural issue, or when the record contains new and material evidence that relates to the period on or before the hearing decision — and, post-2017, only where there is a reasonable probability the evidence would change the outcome and the claimant shows good cause for not submitting it earlier.
Frame your request to those hooks. The AC is not persuaded by a re-argument of the medical weight of the evidence; it responds to identifiable legal and evidentiary defects in the decision below.
AC review vs. starting a new application — the trade-offs
The single variable that should dominate this decision is time — specifically protected time. A request for review preserves the original application date, the alleged onset, and the full retroactive period in play. A new application abandons all of it: you reset to a new protective filing date, you re-expose insured-status questions, and any disability that pre-dates a now-expired DLI may become unprovable in the new posture.
Re-filing earns its place in a narrow band of cases: the strongest medical evidence is new and post-dates the hearing, onset is comfortably re-establishable inside insured status, and the legal defects in the ALJ decision are thin. Outside that band, conceding the prior period to chase a faster initial decision is usually a bad trade — you're spending guaranteed back pay to buy speed.
Where the error set is legal — RFC built without a logical bridge, a Step-5 finding unsupported by the vocational record, symptom evaluation that ignores the regulatory factors — review (and the federal-court appeal it preserves) is the road that keeps the whole claim alive.
The remand / deny / decide base rates
The AC does one of three things: it denies review (the ALJ decision stands), it remands for a new hearing, or — less often — it issues its own decision. The mix matters because it sets your realistic expected value before federal court is even on the table. SSA's own oversight, including the SSA OIG review of Appeals Council workloads, documents the volume and disposition pressures on the AC and the processing-time exposure that comes with requesting review.
Read the base rates honestly with the client. Denial of review is the statistical default; remand is the realistic win at this level and is far more common than an outright AC decision in the claimant's favor. That shapes the conversation: you're usually playing for a remand and a cleaner second hearing, not for the AC to award benefits outright. Knowing the disposition mix for cases like yours — and the added wait it carries — is exactly the input the review-vs-refile call needs.
Writing the AC brief to set up federal court
Treat the AC brief as the first draft of your district-court complaint, because exhaustion runs through it. Every error you want a federal judge to reach should be specifically and clearly raised here; issues you don't fairly present to the AC are the ones that get waived downstream.
Anchor each argument to a recognized, recurring defect rather than a bespoke theory. SSA publishes its Top 10 Court Remand Reasons, and the live error types — RFC and evidence-evaluation failures, Step-5 and DOT-conflict problems, opinion-evidence and symptom-evaluation errors — are the same ones the AC and the courts actually act on. Map your assignments of error to that list, cite the page of the decision and the record where each defect lives, and keep the medical re-argument to a minimum.
Protecting onset date and back pay across the decision
The decision you're appealing usually contains findings you want to keep — most often a favorable or unchallenged onset framing — alongside the error you're attacking. Be deliberate about scope. A remand reopens the period for re-adjudication, which is the point, but it can also put a previously workable onset back in play. Flag in the brief what should survive and why.
Run the back-pay exposure both ways before you advise. Preserving the original onset through review protects the full retroactive period; a new application caps you to the look-back the new filing date allows and may strand benefits behind an expired DLI. In long-backlog offices, the added months of AC and court processing change the math again — sometimes in favor of patience, sometimes not.
The discipline is the same one that governs every stage of this work: the path with the highest expected value isn't the fastest one, it's the one that protects the most provable benefit. Quantify it, brief the errors that travel, and keep the whole claim alive on the way to federal court.
Sources
This article is for general information and education only. It is not legal advice, and it does not create an attorney–client relationship. SSDI rules change and individual cases differ — for advice about your situation, consult a licensed attorney or accredited representative. AISSDI figures are built on public Social Security Administration data.