Appendix C: Key Data by Circuit Court

This appendix breaks out data about qualified immunity appeals for each federal court of appeals except the Federal Circuit, which generally does not hear qualified immunity cases, and the District of Columbia Circuit, which had few relevant appeals. Disaggregating such a small sample would not yield reliable estimates. (The 41 qualified immunity appeals from the District of Columbia Circuit are included in the “all circuits” row of this appendix’s tables, however.) Because dividing our main dataset into 11 subsets magnifies the potential impact of incorrect algorithm predictions, we have largely included only fields with excellent algorithmic performance. 1  

Table C1: Basic Facts

Circuit Number of Appeals Number of Appeals per Million Pop. Opinion Publication Rate Percent With Plaintiffs Representing Themselves Percent Where Plaintiffs Appealed a District Court Loss Percent Where Defendants Appealed a District Court Loss
1st 105 7.5 86% 6% 64% 34%
2nd 382 16.0 35% 18% 61% 35%
3rd 335 14.8 25% 22% 68% 30%
4th 244 7.8 41% 13% 58% 36%
5th 708 20.2 29% 33% 70% 28%
6th 819 25.1 32% 9% 49% 46%
7th 273 10.8 77% 22% 71% 28%
8th 407 19.2 81% 11% 53% 44%
9th 936 14.4 22% 24% 63% 33%
10th 540 30.1 33% 25% 64% 34%
11th 736 20.9 17% 18% 60% 38%
All Circuits* 5,526 17.0 35% 20% 61% 36%

*The “all circuits” row reflects data for all qualified immunity appeals in our dataset, including those from the District of Columbia Circuit.

Note: The percentage of appeals that involved cross-appellants is not shown. Accordingly, the percentages of appeals where plaintiffs and defendants appealed a district court loss do not sum to 100%.

Table C2: Defendants and Allegations

Circuit Percent With Law Enforcement Defendants Percent With Prison Defendants Percent With Non-Police, Non-Prison Defendants Percent With Excessive Force Allegations Percent With First Amendment Allegations
1st 61% 10% 31% 18% 19%
2nd 57% 23% 32% 14% 21%
3rd 57% 20% 33% 18% 26%
4th 62% 24% 24% 29% 18%
5th 55% 30% 25% 29% 16%
6th 63% 19% 26% 34% 18%
7th 54% 29% 24% 16% 20%
8th 65% 25% 20% 33% 16%
9th 52% 25% 29% 28% 18%
10th 56% 31% 27% 23% 19%
11th 70% 21% 18% 36% 13%
All Circuits* 59% 24% 26% 27% 18%

*The “all circuits” row reflects data for all qualified immunity appeals in our dataset, including those from the District of Columbia Circuit.

Note: For simplicity, we have combined state and federal law enforcement defendants into a single law enforcement defendants column, and state and federal prison defendants into a single prison defendants column. These results should be considered reasonable approximations as opposed to precise values.

Table C3: Qualified Immunity Grants and Denials

Circuit Appeal With QI Grant Appeal With QI Denial
1st 56% 26%
2nd 56% 27%
3rd 56% 20%
4th 46% 41%
5th 67% 16%
6th 48% 41%
7th 52% 32%
8th 63% 30%
9th 61% 31%
10th 66% 26%
11th 60% 34%
All Circuits* 59% 30%

*The “all circuits” row reflects data for all qualified immunity appeals in our dataset, including those from the District of Columbia Circuit.

Table C4: Overall Prevailing Party

  Plaintiff Success Rates Defendant Success Rates
Circuit All Appeals When Appealing a District Court Loss When Defending a District Court Win All Appeals When Appealing a District Court Loss When Defending a District Court Win
1st 21% 5% 53% 61% 33% 76%
2nd 26% 11% 52% 57% 36% 69%
3rd 18% 8% 43% 64% 42% 74%
4th 34% 13% 67% 42% 18% 58%
5th 16% 4% 46% 72% 40% 86%
6th 34% 12% 59% 47% 23% 68%
7th 25% 14% 51% 59% 42% 66%
8th 22% 7% 40% 63% 43% 80%
9th 22% 10% 44% 57% 36% 67%
10th 19% 4% 49% 64% 34% 79%
11th 25% 6% 54% 61% 29% 82%
All Circuits* 24% 8% 51% 59% 33% 74%

*The “all circuits” row reflects data for all qualified immunity appeals in our dataset, including those from the District of Columbia Circuit.

Note: This table uses the “prevailing party” field, which records the party who won an appeal overall. Corresponding columns (i.e., the two “all appeals” columns; the plaintiff loss and defendant win columns; and the plaintiff win and defendant loss columns) will not sum to 100%, as we do not display mixed outcomes (where both plaintiffs and defendants won on different aspects of an appeal). Mixed outcomes can be inferred based on the data shown.