For attorneys, journalists and researchersHurt and looking for an attorney? Start here
perfectplaintiff
  1. Home
  2. Newsroom
  3. Case path model

Case path model

Estimates of how a kind of case tends to end and how long it takes, each with a range, and a record of how those estimates held up.

What does the model do?

Our own model of how a kind of case tends to end and how long it takes. For one kind of case in one place, it takes every such case resolved in the federal courts there over the last five complete years and works out the share that ended each way (recorded as settled, dismissed voluntarily, dismissed by the court, decided on a motion before trial, decided at trial) and the share resolved within one, two and three years of filing. The value shown is the share over all five years together. The range starts from the lowest and highest single years and is then widened at each end by half the gap between them, because testing showed the unwidened range was too narrow: we rebuilt the model for each year from 2020 on using only the five years before it, and the share that actually followed fell inside the unwidened range about 64% of the time and inside the widened range about 88% of the time. The test results are published with the model. An estimate is produced only when each of the five years has at least 50 resolved cases. Limits: it describes cases as a group, as the courts recorded their endings. A voluntary dismissal often follows a settlement, and the record does not say which. It covers federal courts only; state courts may differ. It does not predict any single case, and it says nothing about amounts.

Tested against 2020 to 2025, the share that actually followed fell inside the model's range 88% of the time across 6,104 tests. It fell outside the range the rest of the time, so treat every estimate as a guide to what is typical, not a forecast for one case.

Personal injury cases nationwideEstimated

This is our own case path model. It estimates, for personal injury cases in federal courts nationwide, how such cases ended and how long they took, from the 98,741 such cases the courts resolved from 2021 to 2025. The mark is the estimate and the band is its range. It describes cases as a group; it does not predict any one case and says nothing about amounts.

How they ended

  • Recorded as settledAbout 48%range 45% to 51%
  • Dismissed voluntarilyAbout 19%range 16% to 21%
  • Dismissed by the courtAbout 22%range 17% to 25%
  • Decided on a motion before trialAbout 5.0%range 3.8% to 6.3%
  • Decided at trialAbout 1.2%range 0.6% to 1.7%

How long they took

  • Within 12 monthsAbout 53%range 43% to 61%
  • Within 24 monthsAbout 84%range 77% to 90%
  • Within 36 monthsAbout 94%range 92% to 97%

Endings are as the courts recorded them. Courts record "settled" unevenly, and a voluntary dismissal often follows a settlement; the record does not say which. The endings shown do not add up to 100% because a few less common endings are left out. Federal courts only. Method: Case path model. How the model did when tested against past years.

Class actions nationwideEstimated

This is our own case path model. It estimates, for federal cases with class action allegations nationwide, how such cases ended and how long they took, from the 44,541 such cases the courts resolved from 2021 to 2025. The mark is the estimate and the band is its range. It describes cases as a group; it does not predict any one case and says nothing about amounts.

How they ended

  • Recorded as settledAbout 29%range 21% to 37%
  • Dismissed voluntarilyAbout 34%range 29% to 42%
  • Dismissed by the courtAbout 16%range 15% to 19%
  • Decided on a motion before trialAbout 8.1%range 6.1% to 10%
  • Decided at trialAbout 0.3%range 0.1% to 0.4%

How long they took

  • Within 12 monthsAbout 67%range 64% to 69%
  • Within 24 monthsAbout 82%range 79% to 85%
  • Within 36 monthsAbout 90%range 88% to 92%

Endings are as the courts recorded them. Courts record "settled" unevenly, and a voluntary dismissal often follows a settlement; the record does not say which. The endings shown do not add up to 100% because a few less common endings are left out. Federal courts only. Method: Case path model. How the model did when tested against past years.

Every national and state estimate

Case path estimates: personal injury, nationwideEstimated

Kind of caseRecorded as settledDismissed voluntarilyDismissed by the courtDecided on a motion before trialDecided at trialResolved within 12 monthsResolved within 24 monthsResolved cases the estimate rests on
Personal injury (all types)47.9% (45.1% to 51.4%)18.8% (16.1% to 21.1%)21.6% (16.7% to 25.2%)5.0% (3.8% to 6.3%)1.2% (0.6% to 1.7%)52.8% (42.5% to 61.1%)83.8% (76.7% to 89.7%)98,741
Motor vehicle57.8% (55.5% to 60.7%)18.9% (17.5% to 20.9%)16.9% (12.9% to 20.1%)2.0% (1.4% to 2.5%)1.3% (0.8% to 1.7%)53.9% (48.6% to 59.4%)88.4% (84.1% to 92.5%)26,473
Motor vehicle product liability47.7% (40.4% to 54.1%)15.6% (9.9% to 21.9%)22.2% (13.6% to 29.3%)7.0% (5.0% to 9.2%)1.7% (0.3% to 3.4%)50.4% (32.1% to 66.5%)78.8% (66.9% to 87.9%)1,048
Medical malpractice35.9% (28.2% to 41.9%)17.8% (12.1% to 24.4%)28.0% (18.1% to 43.4%)9.1% (5.3% to 13.2%)3.0% (0.7% to 5.8%)42.4% (35.0% to 50.6%)72.0% (67.9% to 76.5%)4,301
Product liability38.4% (17.5% to 56.2%)29.9% (10.5% to 48.1%)22.0% (8.3% to 37.4%)4.1% (1.8% to 6.9%)0.8% (0.5% to 1.1%)47.5% (37.3% to 61.8%)76.9% (68.1% to 85.6%)9,643
Health care and pharmaceutical product liability30.1% (0.0% to 62.1%)19.9% (7.8% to 34.1%)39.0% (3.1% to 80.2%)4.5% (0.1% to 10.2%)0.3% (0.0% to 0.9%)39.5% (14.0% to 71.6%)64.8% (40.8% to 96.1%)3,345
Asbestos43.4% (35.5% to 50.5%)18.0% (10.7% to 28.5%)19.5% (2.8% to 33.6%)4.4% (0.0% to 12.7%)1.7% (0.0% to 7.4%)31.6% (14.4% to 48.0%)66.9% (49.0% to 80.9%)544
Aviation54.9% (48.9% to 64.2%)17.1% (14.1% to 21.0%)17.5% (5.1% to 29.0%)4.4% (0.8% to 7.8%)1.3% (0.0% to 2.6%)43.8% (17.7% to 70.5%)70.8% (34.2% to 100.0%)1,140
Marine60.4% (55.2% to 66.0%)12.3% (9.2% to 14.8%)13.3% (6.3% to 21.9%)3.6% (0.5% to 7.2%)1.4% (0.0% to 3.4%)58.6% (29.6% to 80.1%)84.8% (73.2% to 95.4%)5,336
Railroad employee injury60.7% (50.4% to 70.6%)17.2% (6.0% to 32.4%)14.0% (9.2% to 20.2%)3.9% (1.9% to 6.1%)2.0% (0.7% to 3.3%)37.0% (30.0% to 45.0%)74.0% (65.7% to 82.1%)1,790
Assault, libel and slander14.6% (9.5% to 19.6%)16.8% (11.2% to 26.1%)45.7% (38.3% to 53.1%)13.4% (10.9% to 16.0%)1.6% (0.4% to 3.0%)67.6% (60.2% to 73.1%)87.4% (82.7% to 91.5%)2,776
Other personal injury46.6% (41.9% to 50.1%)17.4% (13.7% to 20.9%)22.4% (18.9% to 25.4%)6.4% (4.4% to 8.7%)1.0% (0.3% to 1.6%)54.8% (44.2% to 64.5%)86.0% (80.8% to 90.4%)42,345

Each cell is the model's value with its range in brackets. Method: case path model, version 2. Inputs: yearly counts of resolved federal cases, 2021 to 2025.

Case path estimates: personal injury in ten large statesEstimated

StateRecorded as settledDismissed voluntarilyDismissed by the courtDecided on a motion before trialDecided at trialResolved within 12 monthsResolved within 24 monthsResolved cases the estimate rests on
California35.6% (24.9% to 48.4%)27.6% (16.3% to 40.5%)23.9% (20.3% to 27.8%)5.2% (2.3% to 7.4%)0.9% (0.2% to 1.6%)56.6% (41.7% to 71.8%)81.7% (70.7% to 94.1%)7,120
Texas40.1% (34.8% to 45.1%)16.9% (12.2% to 21.5%)29.6% (20.6% to 36.8%)4.6% (2.8% to 7.4%)1.0% (0.3% to 1.9%)52.6% (47.3% to 57.7%)90.3% (87.7% to 92.7%)10,005
Florida54.7% (40.2% to 68.6%)9.7% (6.3% to 13.9%)23.4% (14.8% to 32.1%)3.2% (0.4% to 5.6%)1.0% (0.0% to 1.9%)68.4% (46.5% to 88.0%)93.1% (85.5% to 100.0%)9,139
New York42.7% (36.4% to 48.4%)28.3% (22.3% to 33.8%)16.1% (13.7% to 19.1%)6.2% (3.6% to 8.6%)0.8% (0.2% to 1.3%)46.6% (38.0% to 54.4%)75.8% (70.2% to 80.6%)6,732
Pennsylvania41.5% (15.8% to 67.7%)9.5% (5.6% to 16.1%)40.2% (0.0% to 73.3%)0.8% (0.2% to 1.4%)0.8% (0.0% to 1.6%)55.2% (32.1% to 74.4%)80.2% (61.6% to 96.2%)6,968
Illinois34.7% (8.2% to 62.5%)36.5% (13.3% to 59.2%)21.4% (0.0% to 43.0%)3.1% (1.2% to 5.8%)0.7% (0.0% to 1.4%)45.5% (33.3% to 55.7%)75.8% (61.0% to 87.8%)5,150
Georgia54.0% (47.4% to 59.5%)15.8% (13.5% to 18.7%)12.9% (9.1% to 17.2%)8.1% (3.2% to 13.7%)1.3% (0.8% to 1.8%)56.8% (48.6% to 67.7%)88.1% (85.3% to 92.1%)4,146
New Jersey72.4% (70.4% to 76.0%)8.1% (6.7% to 9.6%)14.3% (10.4% to 17.3%)3.7% (2.9% to 4.6%)0.3% (0.0% to 1.0%)55.4% (44.3% to 64.6%)81.3% (70.7% to 90.1%)4,802
Ohio46.6% (35.6% to 62.3%)19.9% (12.7% to 26.6%)21.2% (11.5% to 31.2%)8.1% (4.5% to 12.7%)0.9% (0.0% to 2.1%)51.2% (41.0% to 62.0%)82.2% (75.1% to 88.6%)1,339
Michigan59.9% (48.2% to 69.0%)10.9% (2.4% to 20.1%)16.7% (7.5% to 27.9%)7.3% (0.0% to 17.2%)0.8% (0.4% to 1.6%)47.0% (41.2% to 57.5%)79.8% (72.2% to 85.4%)1,565

Each cell is the model's value with its range in brackets. Method: case path model, version 2. Inputs: yearly counts of resolved federal cases, 2021 to 2025.

Case path estimates: class actions, nationwideEstimated

Kind of class actionRecorded as settledDismissed voluntarilyDismissed by the courtDecided on a motion before trialDecided at trialResolved within 12 monthsResolved within 24 monthsResolved cases the estimate rests on
Class actions (all types)29.3% (20.8% to 36.9%)34.4% (29.0% to 41.7%)16.4% (14.7% to 18.5%)8.1% (6.1% to 10.2%)0.3% (0.1% to 0.4%)66.5% (64.3% to 69.2%)82.4% (79.2% to 85.1%)44,541
Wage, hour and labor class actions43.1% (36.0% to 51.3%)19.4% (13.8% to 23.3%)13.3% (12.1% to 15.2%)6.0% (3.9% to 8.7%)0.6% (0.0% to 1.5%)45.7% (38.6% to 51.3%)72.6% (70.8% to 74.9%)7,355
Employee benefit (ERISA) class actions25.9% (12.7% to 46.0%)12.5% (3.6% to 20.5%)15.0% (6.6% to 23.3%)22.1% (16.7% to 26.3%)2.3% (0.0% to 5.9%)30.7% (14.9% to 41.5%)55.1% (42.2% to 65.6%)479
Disability access class actions38.9% (30.4% to 45.5%)50.2% (44.3% to 56.1%)7.8% (5.9% to 9.1%)0.7% (0.2% to 1.0%)0.0% (0.0% to 0.1%)93.8% (92.4% to 95.4%)97.9% (94.5% to 100.0%)10,815
Other civil rights class actions26.6% (22.7% to 29.8%)17.5% (12.2% to 22.9%)25.8% (19.4% to 31.9%)15.9% (12.6% to 20.1%)0.9% (0.1% to 1.6%)47.1% (41.0% to 53.7%)69.4% (66.4% to 72.5%)1,875
Consumer credit class actions44.2% (30.0% to 52.6%)25.9% (18.8% to 30.8%)15.6% (5.5% to 26.7%)7.6% (4.7% to 11.4%)0.2% (0.0% to 0.7%)71.2% (53.1% to 81.9%)88.7% (76.8% to 96.6%)2,776
Telephone consumer protection class actions28.4% (20.8% to 35.9%)53.2% (34.4% to 67.6%)12.7% (7.8% to 19.8%)1.7% (0.0% to 3.9%)0.0% (0.0% to 0.0%)85.5% (67.8% to 100.0%)97.1% (92.4% to 100.0%)2,804
Securities class actions10.3% (4.0% to 16.3%)20.0% (3.2% to 39.1%)21.9% (7.1% to 41.6%)22.8% (15.7% to 30.8%)0.1% (0.0% to 0.6%)33.0% (20.2% to 50.6%)55.6% (38.2% to 77.1%)1,528
Antitrust class actions21.6% (2.7% to 41.8%)13.6% (0.8% to 29.9%)37.4% (0.0% to 77.1%)10.5% (0.0% to 22.0%)0.2% (0.0% to 1.7%)26.0% (2.3% to 60.7%)35.4% (5.9% to 67.8%)810
Contract class actions17.5% (12.6% to 22.9%)28.6% (14.8% to 43.1%)22.6% (13.0% to 35.2%)12.7% (0.0% to 29.7%)0.3% (0.1% to 0.4%)61.6% (34.8% to 82.8%)80.4% (68.9% to 89.7%)4,699
Fraud and property damage class actions14.6% (11.4% to 18.5%)41.3% (33.8% to 47.2%)17.9% (10.2% to 27.4%)12.5% (7.1% to 18.2%)0.2% (0.1% to 0.3%)60.6% (47.6% to 69.5%)81.6% (69.4% to 91.2%)4,287
Personal injury class actions12.9% (10.3% to 15.8%)38.0% (2.7% to 64.2%)24.5% (3.5% to 59.9%)6.6% (2.1% to 12.1%)0.2% (0.0% to 1.3%)66.1% (23.0% to 96.9%)79.9% (41.6% to 100.0%)2,974
Other class actions20.3% (13.3% to 28.6%)28.8% (10.1% to 45.4%)23.3% (18.7% to 29.1%)16.7% (6.3% to 30.6%)0.3% (0.0% to 0.7%)61.8% (50.0% to 71.0%)78.1% (67.4% to 86.9%)4,139

Each cell is the model's value with its range in brackets. Method: case path model, version 2. Inputs: yearly counts of resolved federal cases, 2021 to 2025.

How have the estimates held up?

Track record: what the model estimated and what happenedComputed

What was estimatedYearModel's estimate (range)What happenedInside the range
Recorded as settled202053.2% (46.0% to 59.1%)47.8%Yes
Recorded as settled202151.5% (44.1% to 58.6%)49.8%Yes
Recorded as settled202251.0% (44.1% to 58.6%)47.7%Yes
Recorded as settled202349.8% (44.1% to 58.7%)48.9%Yes
Recorded as settled202448.7% (46.7% to 50.9%)46.7%Yes
Recorded as settled202548.2% (45.1% to 51.4%)46.7%Yes
Resolved within 12 months202053.5% (33.7% to 70.2%)60.0%Yes
Resolved within 12 months202157.3% (47.1% to 65.7%)51.2%Yes
Resolved within 12 months202257.1% (46.3% to 66.0%)47.2%Yes
Resolved within 12 months202355.3% (40.2% to 68.0%)52.7%Yes
Resolved within 12 months202454.3% (40.2% to 68.0%)55.7%Yes
Resolved within 12 months202553.3% (40.8% to 66.4%)56.5%Yes

Personal injury cases in all federal courts, outside multidistrict litigation.

How the model did when tested against past yearsComputed

What was estimatedTestsActual share fell inside the rangeTypical gap
Dismissed by the court76387%3.5 points
Dismissed voluntarily76385%3.5 points
Decided on a motion before trial76391%1.5 points
Recorded as settled76389%5.1 points
Decided at trial76392%0.6 points
Resolved within 12 months76387%5.6 points
Resolved within 24 months76387%3.7 points
Resolved within 36 months76386%1.7 points
All measures6,10488%2.5 points

A test is one place, one kind of case and one year. The typical gap is the median difference between the model's value and the actual share, in percentage points.

Use it

The model makes no estimate about any named firm, attorney, judge, person or litigation, and none about amounts. When you quote it, say "estimated" and give the range.

Sources

Last updated 18 November 2026. That is the date we last retrieved the source files below.

  • Federal Judicial Center Integrated Database (civil, 1988 forward) File retrieved 18 November 2026 from fjc.gov. File fingerprint 74405231a9a3.

How the estimates on this page were made

Case path model. Our own model of how a kind of case tends to end and how long it takes. For one kind of case in one place, it takes every such case resolved in the federal courts there over the last five complete years and works out the share that ended each way (recorded as settled, dismissed voluntarily, dismissed by the court, decided on a motion before trial, decided at trial) and the share resolved within one, two and three years of filing. The value shown is the share over all five years together. The range starts from the lowest and highest single years and is then widened at each end by half the gap between them, because testing showed the unwidened range was too narrow: we rebuilt the model for each year from 2020 on using only the five years before it, and the share that actually followed fell inside the unwidened range about 64% of the time and inside the widened range about 88% of the time. The test results are published with the model. An estimate is produced only when each of the five years has at least 50 resolved cases. Limits: it describes cases as a group, as the courts recorded their endings. A voluntary dismissal often follows a settlement, and the record does not say which. It covers federal courts only; state courts may differ. It does not predict any single case, and it says nothing about amounts. More on this method.

The full address of every source file is on the sources page. Found something wrong? See corrections and removal requests. How figures are labelled and calculated is in the methodology.