How Business Insider investigated the state of the Eighth Amendment

In the 1980s and 1990s, a series of Supreme Court decisions and a new law sought to curb "frivolous" prisoner lawsuits and give more deference to prison officials. Together, they changed the legal landscape of the Eighth Amendment — the foundational constitutional protection against "cruel and unusual punishments."

Heavy criticism soon followed — from the American Bar Association, members of Congress, and Supreme Court justices themselves. Ted Kennedy, then the senior senator from Massachusetts, described one critical measure, the 1996 Prison Litigation Reform Act, as "a far-reaching effort to strip federal courts of the authority to remedy unconstitutional prison conditions."

In successive decades, judges and legal scholars have continued to argue that these developments erode protections for the 1.2 million people incarcerated nationwide. Top law-enforcement officials, on the other hand, have argued that prisoners were clogging up the courts with costly, meaningless actions. And federal judges have hesitated to substitute their own judgment for that of correctional personnel on the ground.

Little has been known about how the Supreme Court standards and the 1996 law shaped outcomes for cases alleging Eighth Amendment violations — and how many of the underlying claims dealt with frivolous matters.

Business Insider set out to investigate. Guided by advice from legal scholars including Margo Schlanger, a law professor at the University of Michigan, and attorneys who have argued Eighth Amendment cases, BI built and analyzed an original database of 1,488 federal Eighth Amendment lawsuits.

BI found that constitutional protections had been dramatically weakened. Cases settled a third as often as civil-rights suits filed by nonprisoners. And plaintiffs won their cases less than 1% of the time.

Here's how we arrived at our findings.

We started with cases in appellate courts

BI sampled lawsuits using Westlaw, a legal-research platform, searching for opinions that cite one of the two Supreme Court standards that now guide how federal courts evaluate Eighth Amendment claims. While Westlaw is one of the most complete databases of legal filings, it is, like the others, not completely comprehensive.

Claims that allege excessive force are governed by a standard called "malicious and sadistic," so to sample those claims BI searched Westlaw for opinions that mentioned the standard and cited one of the Supreme Court's precedent-setting decisions, Whitley v. Albers or Hudson v. McMillian.

Cases alleging constitutionally inadequate healthcare, a failure to protect, or inadequate conditions of confinement are governed by the "deliberate indifference" standard; to sample those claims, BI searched Westlaw for opinions that mentioned the standard and cited Farmer v. Brennan, Wilson v. Seiter, or the earlier Estelle v. Gamble, the relevant Supreme Court precedent-setting cases.

Because so few Eighth Amendment cases make it to the appeals stage, we were able to pull all opinions that fit these parameters over the course of five years — from 2018 to 2022 — spanning two presidential administrations and the COVID-19 pandemic, which affected prisons and courts nationwide.

Then we sampled cases at the district courts

A great many cases filed in district court never reach the appeals courts. In consultation with constitutional scholars, we pulled a sample of district-level cases to understand how Eighth Amendment cases founder earlier in the process. For these, we pulled all cases with judicial decisions filed in the third quarter of 2022 that fit the same search parameters.

Our district sample captured many more cases that failed under mechanisms imposed by the PLRA.

Together, our samples spanned cases from every circuit in the United States filed by prisoners held in hundreds of US prisons in 47 states, Washington, DC, and Puerto Rico. They are likely not comprehensive. If a case file did not mention the Supreme Court standards or the precedent-setting cases — as when a prisoner without counsel lacked knowledge of the law and the case was dismissed on technical grounds at prescreening — our search would not have captured that case. The impact of the PLRA may be greater than our dataset indicates.

About one in 10 cases contained additional claims unrelated to the Eighth Amendment, including claims filed under the First Amendment or the Americans with Disabilities Act. We tracked only the outcomes of their Eighth Amendment claims.

The sample is dominated by suits filed by people while they were incarcerated but also includes cases filed by formerly incarcerated people or their surviving families. BI excluded cases unrelated to incarceration and those pertaining to pretrial detainees. We excluded any case from our district sample that also appeared in the circuit sample. We also excluded cases that had not reached a final disposition, including subsequent appeals, by September 1, 2024.

Our data analysis

A team of researchers read federal court filings associated with each case and logged details in 59 variables including plaintiff names and incarceration status; the claim category and standard applied; defendant names, roles, and employers; alleged harm suffered, such as broken bone, amputation, or death; whether defendants asserted an affirmative defense, such as qualified immunity; whether the plaintiff had counsel; and judge and court names. If a plaintiff had a lawyer at any point during their case, we logged them as having legal representation.

Researchers assigned each case a final disposition category based on the standard or procedural mechanism under which the case was ultimately decided. BI counted all cases that failed under any mechanism of the PRLA — whether they were dismissed at screening, or were struck because prisoners failed to exhaust the prison's grievance process before filing suit, or fell short of any other of the law's requirements — as having failed under the PLRA. This tally included cases in which a judge granted a plaintiff leave to amend. If the plaintiff filed an amended complaint, we logged the case according to its final resolution.

When defendants voluntarily withdrew their cases for reasons clearly unrelated to reaching a settlement; died without a family member or estate to continue their case; or otherwise failed to respond to a court order resulting in dismissal, BI counted these cases as abandoned or withdrawn.

BI's sample included 86 cases with claims governed under both the "malicious and sadistic" and "deliberate indifference" standards. For those, we analyzed the case on both standards, recorded each claim's disposition, and assigned a final disposition category to the case based on the longest surviving claim. For example, if a judge dismissed a plaintiff's excessive-force claim but allowed the plaintiff's claim of inadequate medical care to go to a jury trial, we counted the case outcome under the "deliberate indifference" standard.

In 12 cases, judges granted summary judgment to claims governed by both standards in a single order. Ten of these cases alleged excessive force and alleged inadequate medical care for the injuries they said resulted; two claimed both excessive force and inadequate conditions of confinement. We assigned all these cases as having lost on the "malicious and sadistic" standard since the claims centered on an excessive-force complaint.

The sample includes 28 cases in which a jury found in favor of the defendants. BI counted these as cases in which the plaintiffs lost on the relevant standard.

BI counted cases as having failed on qualified immunity, the Supreme Court doctrine that can shield public officials from civil liability, only when that standard was decisive in the dismissal of a plaintiff's claim. Another Supreme Court precedent, known as Bivens, governs cases filed against federal officials; we tallied a case as a Bivens failure only when a judge dismissed a claim solely because it had not met Bivens standards.

All cases were recorded as having reached a settlement when the settlement was clearly stipulated in the docket or when the defendant and the plaintiff submitted a joint motion for dismissal pursuant to an out-of-court settlement agreement. To identify damages paid, BI reviewed the docket for disclosed settlement amounts and supplemented this research with news reports and public announcements.

BI counted two categories of cases as plaintiff wins: cases in which the plaintiff won at jury trial and cases in which judges made decisions in favor of the plaintiff — whether for damages or for injunctive relief.

To supplement and contextualize our analysis of the 1,488 lawsuits, containing tens of thousands of pages of court records, we reviewed hundreds of pages of training materials obtained through public-records requests from 37 states, as well as medical records, incident reports, and surveillance footage obtained through records requests and court filings. We reviewed scholarly literature on the Eighth Amendment; read Supreme Court decisions, dissents, and archives; and reviewed archived congressional debates. And we interviewed more than 170 people, including prisoners and their families, attorneys and legal scholars, correctional staff and prison healthcare providers, policymakers and law-enforcement officials, and current and former federal judges.