Do Appellate Courts Actually Correct Errors?

The conventional wisdom about appellate courts is straightforward: they exist to correct errors made by trial courts. This textbook description portrays appellate courts as institutional safeguards, carefully reviewing lower court decisions to ensure justice is properly administered. But does this conventional understanding match reality? A closer examination of how appellate courts actually function—particularly in criminal cases—reveals a troubling disconnect between theory and practice.

The Traditional Role of Appellate Courts

Appellate courts occupy a unique position in the judicial system. Unlike trial courts, which have original jurisdiction and conduct initial proceedings with witnesses, evidence, and jury deliberations, appellate courts have appellate jurisdiction—they review decisions made by lower courts. They do not retry cases or hear new evidence, but instead examine the legal record to determine whether errors occurred during the original proceedings.

The traditional scholarly consensus holds that appellate courts serve primarily as error-correcting institutions. Legal scholars have long characterized appellate review as essential for maintaining the integrity of the judicial system. See, e.g., Steven Shavell, The Appeals Process as a Means of Error Correction, 24 J. Legal Studies 379 (1995). This view sees appellate courts as institutional mechanisms for catching and remedying mistakes that inevitably occur in complex legal proceedings.

The Challenge of Defining “Error”

However, the conventional description immediately encounters a fundamental problem: what constitutes an “error” is often far from objective. Many appellate cases involve mixed questions of law and fact where reasonable people can disagree. Was a police search “reasonable” under Fourth Amendment standards? Did a trial judge “abuse discretion” in ruling on evidence? These questions resist simple classification as either “correct” or “erroneous.”

But there is one category of cases where we can objectively identify errors: criminal convictions where the defendant is factually innocent or guilty. Here, we can apply the framework of statistical hypothesis testing to understand two distinct types of errors that can occur in criminal trials.

Type I and Type II Errors in Criminal Justice

In a scientific framework, a Type I error occurs when we reject a true null hypothesis—in criminal law, this means convicting an innocent person. The null hypothesis in criminal proceedings is innocence, so a Type I error represents the wrongful conviction of someone who did not commit the crime.

A Type II error occurs when we fail to reject a false null hypothesis—in criminal law, this means acquitting a guilty person. This represents the failure to convict someone who actually committed the crime.

Both types of errors undermine the criminal justice system, but they present very different challenges for appellate review.

The Structural Limitation: Type II Errors

Appellate courts cannot correct Type II errors due to a fundamental structural constraint: double jeopardy protections. The Fifth Amendment prevents the government from appealing an acquittal, even when compelling evidence later emerges that the defendant was actually guilty. No matter how clear the evidence of guilt, no matter how obvious the trial court’s error in interpretation of law or fact, appellate courts have no power to overturn an acquittal.

This means that appellate courts can only potentially correct Type I errors—cases where innocent people have been wrongfully convicted.

Do Appellate Courts Correct Type I Errors?

The statistics on appellate review paint a stark picture. The overwhelming majority of criminal convictions are affirmed on appeal, with reversal rates typically hovering around 7-12% across different jurisdictions. This might suggest that trial courts rarely make significant errors, but the reality is more troubling.

We can test the error-correction hypothesis by examining cases where we know with certainty that Type I errors occurred: wrongful convictions later proven through DNA evidence or other definitive proof of innocence. These cases provide a natural experiment—if appellate courts effectively correct errors, they should have identified and overturned these wrongful convictions.

The Sobering Evidence

Brandon Garrett’s comprehensive study “Judging Innocence,” published in the Columbia Law Review, examined the first 200 DNA exonerations in the United States and reached a damning conclusion about appellate courts’ error-correction capabilities. Among these cases—where we now know with scientific certainty that innocent people were wrongfully convicted—appellate courts played an astonishingly limited role in correcting the errors.

Garrett found that appellate courts freed only a handful of these 200 innocent people. The vast majority of these wrongful convictions were affirmed on appeal, despite the fact that we now know the defendants were factually innocent. In many cases, appellate courts reviewed the same evidence that would later prove innocence and still upheld the convictions. “Strikingly, courts denied at least twelve exonerees relief despite at least preliminary DNA results excluding them. … Only two received DNA testing and a vacatur through federal habeas corpus” (Garrett 2009, p. 102). Comparing the treatment of 200 innocent people to a comparison sample, Garrett reports that appellate courts were no more likely to reverse the conviction of an innocent person than they were to reverse generally.

Alternative Paths to Justice

Perhaps most tellingly, Garrett’s research revealed that more wrongful convictions were corrected through executive action—gubernatorial pardons and commutations—than through the judicial branch’s appellate process. This suggests that political actors operating outside the formal legal system have been more effective at identifying and correcting judicial errors than the courts themselves.

The inadequacy of traditional appellate review has prompted the development of new institutional mechanisms. Conviction Review Units (CRUs) represent a growing recognition that the standard appellate process fails to catch errors. These specialized prosecutorial offices systematically review potentially wrongful convictions, often using investigative techniques unavailable to appellate courts. CRUs examine physical evidence, conduct new interviews, and apply modern forensic techniques to old cases.

The proliferation of CRUs across the country—from a handful in the early 2000s to dozens today—serves as an institutional admission that traditional appellate review is insufficient. These units exist precisely because the conventional error-correction mechanism has proven inadequate.

Rethinking Appellate Courts

The evidence suggests that our conventional understanding of appellate courts as error-correcting institutions may be more aspirational than descriptive. While appellate courts certainly serve important functions—ensuring legal consistency, clarifying doctrine, and providing procedural safeguards—their actual performance in correcting the most serious errors (wrongful convictions) has been disappointing.

This raises profound questions about how we structure criminal appeals and what realistic expectations we should have for appellate review. Rather than assuming that appellate courts actually correct errors, those interested in justice may need to confront unpleasant facts. Appellate review of criminal trials is slow, unreliable, and costly. While CRUs offer a back-up system to address wrongful convictions, the appellate system itself needs to reckon with its failures. When a plane crashes, there is an investigation to determine the cause and whether similar crashes can be prevented. Similar inquiries are needed in criminal justice.

The gap between the textbook description of appellate courts and their actual performance in correcting errors represents more than an academic curiosity. It reveals a fundamental challenge in designing institutions that can reliably identify and correct judicial mistakes, particularly in criminal cases where the stakes are highest and the consequences of error most severe.