Across all case types in US federal courts, roughly 86% of civil appellants do not achieve reversal of the lower court’s judgment — the figure is most favorable for private civil disputes (approximately 14% reversal rate), worse across the overall caseload (8.2%). Spending the time and money to appeal results in the original ruling standing in the vast majority of cases. On the inaction side, a large share of litigants with potentially valid grounds for appeal never file one: access-to-justice research suggests cost barriers, lack of counsel, and low awareness of appellate rights prevent roughly a third of adverse-judgment holders from even attempting an appeal — leaving them with outcomes they believed were wrong but never contested.
The decision to appeal sits in a distinctive position among legal choices: it is almost always made after an already costly first proceeding, the odds are quantifiably long before the filing is made, and the costs of the action are immediate while potential benefits are speculative. Appeals courts are explicitly deferential to trial courts — the “any reasonable trier of fact” standard in civil cases means that many judgments that appear wrong to the losing party are legally unreviewable. The anti-plaintiff pattern in appellate outcomes is robust in the empirical literature: plaintiff appellants achieve reversal roughly half as often as defendant appellants across multiple studies, meaning individuals bringing claims face steeper uphill odds than institutional defendants with experienced appellate counsel.
The regret calculus here is complicated by the multiple functions an appeal can serve. Some appellants file to force a settlement renegotiation, not because they expect to win at the appellate level — in those cases, even an unsuccessful appeal may achieve the desired outcome and generate no regret. The inaction side is equally ambiguous: many litigants who accept an adverse judgment have correctly assessed that their grounds for appeal are thin, and acceptance is a rational response rather than resigned defeat. The Gilovich temporal model predicts that regret about inaction grows over time — people who never fought a ruling they thought unjust may carry that grievance for years — while the disappointment of a failed appeal fades as appellants adapt to the outcome. The point estimates here are the weakest in this dataset and should be read as directional rather than quantitative.